Nutritional supplements for people being treated for active tuberculosis

Cochrane
Library
Cochrane Database of Systematic Reviews
Nutritional supplements for people being treated for active

tuberculosis (Review)
Grobler L, Nagpal S, Sudarsanam TD, Sinclair D
Grobler L, Nagpal S, Sudarsanam TD, Sinclair D.

Nutritional supplements for people being treated for active tuberculosis.
Cochrane Database of Systematic Reviews 2016, Issue 6. Art. No.: CD006086.
DOI: 10.1002/14651858.CD006086.pub4.
www.cochranelibrary.com
Nutritional supplements for people being treated for active tuberculosis (Review)
Copyright © 2016 The Authors. Cochrane Database of Systematic Reviews published by John Wiley & Sons, Ltd. on
behalf of The Cochrane Collaboration.
Cochrane Trusted evidence.
Informed decisions.
Library Better health. Cochrane Database of Systematic Reviews
TABLE OF CONTENTS
HEADER......................................................................................................................................................................................................... 1
ABSTRACT..................................................................................................................................................................................................... 1
PLAIN LANGUAGE SUMMARY....................................................................................................................................................................... 2
SUMMARY OF FINDINGS.............................................................................................................................................................................. 4
BACKGROUND.............................................................................................................................................................................................. 8
OBJECTIVES.................................................................................................................................................................................................. 9
METHODS..................................................................................................................................................................................................... 9
RESULTS........................................................................................................................................................................................................ 11
Figure 1.................................................................................................................................................................................................. 11
Figure 2.................................................................................................................................................................................................. 13
DISCUSSION.................................................................................................................................................................................................. 19
AUTHORS' CONCLUSIONS........................................................................................................................................................................... 20
ACKNOWLEDGEMENTS................................................................................................................................................................................ 21
REFERENCES................................................................................................................................................................................................ 22
CHARACTERISTICS OF STUDIES.................................................................................................................................................................. 28
DATA AND ANALYSES.................................................................................................................................................................................... 85
Analysis 1.1. Comparison 1 Macronutrient supplementation, Outcome 1 Death (1 year of follow-up)........................................... 86
Analysis 1.2. Comparison 1 Macronutrient supplementation, Outcome 2 Cured (at 6 months)...................................................... 87
Analysis 1.3. Comparison 1 Macronutrient supplementation, Outcome 3 Treatment completion.................................................. 88
Analysis 1.4. Comparison 1 Macronutrient supplementation, Outcome 4 Sputum negative at 8 weeks........................................ 88
Analysis 1.5. Comparison 1 Macronutrient supplementation, Outcome 5 Mean weight gain.......................................................... 88
Analysis 1.6. Comparison 1 Macronutrient supplementation, Outcome 6 Change in maximum grip strength (kg)........................ 89
Analysis 1.7. Comparison 1 Macronutrient supplementation, Outcome 7 Change in quality of life score...................................... 90
Analysis 2.1. Comparison 2 High cholesterol (850 mg/day) versus low cholesterol (250 mg/day) diet, Outcome 1 Sputum-culture 91
positive...................................................................................................................................................................................................
Analysis 3.1. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 1 Death during follow-up in adults 93
and children..........................................................................................................................................................................................
Analysis 3.2. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 2 Tuberculosis treatment 93
completion.............................................................................................................................................................................................
Analysis 3.3. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 3 Sputum-smear or sputum- 94
culture positive at 1 month..................................................................................................................................................................
Analysis 3.4. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 4 Sputum-smear or sputum- 94
culture positive at 2 months................................................................................................................................................................
Analysis 3.5. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 5 Clearance of chest X-ray at 6 95
months...................................................................................................................................................................................................
Analysis 3.6. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 6 Weight....................................... 95
Analysis 3.7. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 7 Anthropometrical changes at 96
follow-up................................................................................................................................................................................................
Analysis 3.8. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 8 Mean change in handgrip 96
strength (kg)..........................................................................................................................................................................................
Analysis 4.1. Comparison 4 Vitamin A versus placebo, Outcome 1 Children: mean serum retinol (normal range > 20 µg/L)......... 98
Analysis 4.2. Comparison 4 Vitamin A versus placebo, Outcome 2 Adults: mean serum retinol (normal range > 70 µmol/L)......... 99
Analysis 4.3. Comparison 4 Vitamin A versus placebo, Outcome 3 Death......................................................................................... 99
Analysis 4.4. Comparison 4 Vitamin A versus placebo, Outcome 4 Treatment completion............................................................. 100
Analysis 4.5. Comparison 4 Vitamin A versus placebo, Outcome 5 Symptomatic at 6 weeks.......................................................... 100
Analysis 4.6. Comparison 4 Vitamin A versus placebo, Outcome 6 Sputum-smear and sputum-culture positive during follow- 101
up...........................................................................................................................................................................................................
Analysis 4.7. Comparison 4 Vitamin A versus placebo, Outcome 7 BMI (kg/m2)............................................................................... 101
Analysis 4.8. Comparison 4 Vitamin A versus placebo, Outcome 8 Body fat (%).............................................................................. 102
Analysis 5.1. Comparison 5 Zinc versus placebo, Outcome 1 Serum zinc level (normal range > 10.7 µmol/L)................................ 105
Analysis 5.2. Comparison 5 Zinc versus placebo, Outcome 2 Death by 6 to 8 months..................................................................... 105
Analysis 5.3. Comparison 5 Zinc versus placebo, Outcome 3 Death by 6 to 8 months (subgrouped by HIV status)........................ 106
Nutritional supplements for people being treated for active tuberculosis (Review) i
Copyright © 2016 The Authors. Cochrane Database of Systematic Reviews published by John Wiley & Sons, Ltd. on behalf of The Cochrane
Collaboration.
Informed decisions.
Analysis 5.4. Comparison 5 Zinc versus placebo, Outcome 4 Treatment completion at 6 months.................................................. 107
Analysis 5.5. Comparison 5 Zinc versus placebo, Outcome 5 Sputum-smear or sputum-culture positive during follow-up.......... 107
Analysis 5.6. Comparison 5 Zinc versus placebo, Outcome 6 Clearance of chest X-ray at 6 months............................................... 108
Analysis 5.7. Comparison 5 Zinc versus placebo, Outcome 7 Weight at follow-up........................................................................... 108
Analysis 5.8. Comparison 5 Zinc versus placebo, Outcome 8 BMI (kg/m2)....................................................................................... 109
Analysis 5.9. Comparison 5 Zinc versus placebo, Outcome 9 Body fat (%)....................................................................................... 109
Analysis 5.10. Comparison 5 Zinc versus placebo, Outcome 10 Weight-for-age z score................................................................... 110
Analysis 5.11. Comparison 5 Zinc versus placebo, Outcome 11 BMI-for-age z score........................................................................ 110
Analysis 5.12. Comparison 5 Zinc versus placebo, Outcome 12 Height-for-age z score at follow-up.............................................. 110
Analysis 6.1. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 1 Death by 6 months.................................................... 113
Analysis 6.2. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 2 Treatment completion at 6 months.......................... 114
Analysis 6.3. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 3 Sputum-smear and sputum-culture positive during 114
follow-up................................................................................................................................................................................................
Analysis 6.4. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 4 Body weight (kg)....................................................... 115
Analysis 6.5. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 5 BMI (kg/m2)............................................................... 116
Analysis 6.6. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 6 Mid upper arm circumference (cm).......................... 116
Analysis 6.7. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 7 Biceps skinfold thickness (mm)................................ 117
Analysis 6.8. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 8 Triceps skinfold thickness (mm)............................... 117
Analysis 6.9. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 9 Subscapular skinfold thickness (mm)...................... 117
Analysis 6.10. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 10 Suprailiac skinfold thickness (mm)...................... 117
Analysis 6.11. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 11 Body fat (%)........................................................... 118
Analysis 6.12. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 12 Fat mass (kg)......................................................... 118
Analysis 6.13. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 13 Karnofsky score..................................................... 119
Analysis 7.1. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 1 Serum vitamin D levels (nmol/L)............... 120
Analysis 7.2. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 2 Death during follow-up (2 to 12 months)...... 121
Analysis 7.3. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 3 Death during follow-up (2 to 12 months)...... 122
Analysis 7.4. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 4 Cure at 6 months........................................ 123
Analysis 7.5. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 5 Tuberculosis score..................................... 123
Analysis 7.6. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 6 Sputum-smear or sputum-culture 124
positive...................................................................................................................................................................................................
Analysis 7.7. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 7 Body mass index........................................ 125
Analysis 7.8. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 8 Body weight (kg)........................................ 126
Analysis 7.9. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 9 Karnofsky score at 8 weeks........................ 126
Analysis 8.1. Comparison 8 Arginine versus placebo, Outcome 1 Death during treatment............................................................. 127
Analysis 8.2. Comparison 8 Arginine versus placebo, Outcome 2 Cured at 6/8 months................................................................... 128
Analysis 8.3. Comparison 8 Arginine versus placebo, Outcome 3 Sputum-smear or sputum-culture positive............................... 128
Analysis 8.4. Comparison 8 Arginine versus placebo, Outcome 4 Cough.......................................................................................... 129
Analysis 8.5. Comparison 8 Arginine versus placebo, Outcome 5 Weight gain > 10%...................................................................... 130
Analysis 9.1. Comparison 9 Vitamin E plus selenium versus placebo, Outcome 1 Sputum-smear positive at follow-up................ 131
ADDITIONAL TABLES.................................................................................................................................................................................... 131
APPENDICES................................................................................................................................................................................................. 139
WHAT'S NEW................................................................................................................................................................................................. 151
HISTORY........................................................................................................................................................................................................ 151
CONTRIBUTIONS OF AUTHORS................................................................................................................................................................... 152
DECLARATIONS OF INTEREST..................................................................................................................................................................... 152
SOURCES OF SUPPORT............................................................................................................................................................................... 152
DIFFERENCES BETWEEN PROTOCOL AND REVIEW.................................................................................................................................... 152
INDEX TERMS............................................................................................................................................................................................... 153
Nutritional supplements for people being treated for active tuberculosis (Review) ii
Collaboration.
Informed decisions.
[Intervention Review]
Nutritional supplements for people being treated for active tuberculosis
Liesl Grobler1, Sukrti Nagpal2, Thambu D Sudarsanam3, David Sinclair2
1Centre for Evidence-based Health Care, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa.
2Department of Clinical Sciences, Liverpool School of Tropical Medicine, Liverpool, UK. 3Medicine Unit 2 and Clinical Epidemiology Unit,
Christian Medical College, Vellore, India
Contact address: David Sinclair, Department of Clinical Sciences, Liverpool School of Tropical Medicine, Liverpool, UK.
[email protected].
Editorial group: Cochrane Infectious Diseases Group.

Publication status and date: Unchanged, published in Issue 6, 2016.
Citation: Grobler L, Nagpal S, Sudarsanam TD, Sinclair D. Nutritional supplements for people being treated for active tuberculosis.
Cochrane Database of Systematic Reviews 2016, Issue 6. Art. No.: CD006086. DOI: 10.1002/14651858.CD006086.pub4.
Copyright © 2016 The Authors. Cochrane Database of Systematic Reviews published by John Wiley & Sons, Ltd. on behalf of The
Cochrane Collaboration. This is an open access article under the terms of the Creative Commons Attribution-Non-Commercial Licence,
which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for
commercial purposes.
ABSTRACT
Background
Tuberculosis and malnutrition are linked in a complex relationship. Tuberculosis may cause undernutrition through increased metabolic
demands and decreased intake, and nutritional deficiencies may worsen the disease, or delay recovery by depressing important immune
functions. At present, there is no evidence-based nutritional guidance for adults and children being treated for tuberculosis.
Objectives
To assess the effects of oral nutritional supplements in people being treated with antituberculous drug therapy for active tuberculosis.
Search methods
We searched the Cochrane Infectious Disease Group Specialized Register, Cochrane Central Register of Controlled Trials (CENTRAL; Issue
1, 2016), MEDLINE (from 1946 to 4 February 2016), EMBASE (from 1980 to 4 February 2016), LILACS (from 1982 to 4 February 2016), the
metaRegister of Controlled Trials (mRCT), the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP), and
the Indian Journal of Tuberculosis up to 4 February 2016, and checked the reference lists of all included studies.
Selection criteria
Randomized controlled trials that compared any oral nutritional supplement given for at least four weeks with no nutritional intervention,
placebo, or dietary advice only for people being treated for active tuberculosis. The primary outcomes of interest were all-cause death,
and cure at six and 12 months.
Data collection and analysis

Two review authors independently selected trials for inclusion, and extracted data and assessed the risk of bias in the included trials.
We presented the results as risk ratios (RR) for dichotomous variables, and mean differences (MD) for continuous variables, with 95%
confidence intervals (CIs). Where appropriate, we pooled data from trials with similar interventions and outcomes. We assessed the quality
of the evidence using the Grading of Recommendation Assessment, Development and Evaluation (GRADE) approach.
Main results
Thirty-five trials, including 8283 participants, met the inclusion criteria of this review.
Nutritional supplements for people being treated for active tuberculosis (Review) 1
Collaboration.
Informed decisions.
Macronutrient supplementation
Six trials assessed the provision of free food, or high-energy supplements. Only two trials measured total dietary intake, and in both trials
the intervention increased calorie consumption compared to controls.
The available trials were too small to reliably prove or exclude clinically important benefits on mortality (RR 0.34, 95% CI 0.10 to 1.20; four
trials, 567 participants, very low quality evidence), cure (RR 0.91, 95% CI 0.59 to 1.41; one trial, 102 participants, very low quality evidence),
or treatment completion (data not pooled; two trials, 365 participants, very low quality evidence).
Supplementation probably produces a modest increase in weight gain during treatment for active tuberculosis, although this was not seen
consistently across all trials (data not pooled; five trials, 883 participants, moderate quality evidence). Two small studies provide some
evidence that quality of life may also be improved but the trials were too small to have much confidence in the result (data not pooled;
two trials, 134 participants, low quality evidence).
Micronutrient supplementation
Six trials assessed multi-micronutrient supplementation in doses up to 10 times the dietary reference intake, and 18 trials assessed single
or dual micronutrient supplementation.
Routine multi-micronutrient supplementation may have little or no effect on mortality in HIV-negative people with tuberculosis (RR
0.86, 95% CI 0.46 to 1.6; four trials, 1219 participants, low quality evidence), or HIV-positive people who are not taking antiretroviral
therapy (RR 0.92, 95% CI 0.69 to 1.23; three trials, 1429 participants, moderate quality evidence). There is insufficient evidence to know if
supplementation improves cure (no trials), treatment completion (RR 0.99, 95% CI 0.95 to 1.04; one trial, 302 participants, very low quality
evidence), or the proportion of people who remain sputum positive during the first eight weeks (RR 0.92, 95% CI 0.63 to 1.35; two trials,
1020 participants, very low quality evidence). However, supplementation may have little or no effect on weight gain during treatment (data
not pooled; five trials, 2940 participants, low quality evidence), and no studies have assessed the effect on quality of life.
Plasma levels of vitamin A appear to increase following initiation of tuberculosis treatment regardless of supplementation. In contrast,
supplementation probably does improve plasma levels of zinc, vitamin D, vitamin E, and selenium, but this has not been shown to have
clinically important benefits. Of note, despite multiple studies of vitamin D supplementation in different doses, statistically significant
benefits on sputum conversion have not been demonstrated.
Authors' conclusions
There is currently insufficient research to know whether routinely providing free food, or energy supplements improves tuberculosis
treatment outcomes, but it probably improves weight gain in some settings.
Although blood levels of some vitamins may be low in people starting treatment for active tuberculosis, there is currently no reliable
evidence that routinely supplementing above recommended daily amounts has clinical benefits.
17 April 2019
Update pending
Studies awaiting assessment
The CIDG is currently examining a new search conducted in April 2019 for potentially relevant studies. These studies have not yet been
incorporated into this Cochrane Review.
PLAIN LANGUAGE SUMMARY
Nutritional supplements for people being treated for active tuberculosis
Cochrane researchers conducted a review of the effects of nutritional supplements for people being treated for tuberculosis. After searching
for relevant studies up to 4 February 2016, they included 35 relevant studies with 8283 participants. Their findings are summarized below.
What is active tuberculosis and how might nutritional supplements work?
Tuberculosis is a bacterial infection which most commonly affects the lungs. Most people who get infected never develop symptoms as
their immune system manages to control the bacteria. Active tuberculosis occurs when the infection is no longer contained by the immune
system, and typical symptoms are cough, chest pain, fever, night sweats, weight loss, and sometimes coughing up blood. Treatment is with
a combination of antibiotic drugs, which must be taken for at least six months.
People with tuberculosis are often malnourished, and malnourished people are at higher risk of developing tuberculosis as their immune
system is weakened. Nutritional supplements could help people recover from the illness by strengthening their immune system, and
Collaboration.
Informed decisions.
by improving weight gain, and muscle strength, allowing them to return to an active life. Good nutrition requires a daily intake of
macronutrients (carbohydrate, protein, and fat), and micronutrients (essential vitamins and minerals).
What the research says
Effect of providing nutritional supplements to people being treated for tuberculosis
We currently don't know if providing free food to tuberculosis patients, as hot meals or ration parcels, reduces death or improves cure
(very low quality evidence). However, it probably does improve weight gain in some settings (moderate quality evidence), and may improve
quality of life (low quality evidence).
Routinely providing multi-micronutrient supplements may have little or no effect on deaths in HIV-negative people with tuberculosis (low
quality evidence), or HIV-positive people who are not taking anti-retroviral therapy (moderate quality evidence). We currently don't know if
micronutrient supplements have any effect on tuberculosis treatment outcomes (very low quality evidence), but they may have no effect
on weight gain (low quality evidence). No studies have assessed the effect on quality of life.
Plasma levels of vitamin A appear to increase after starting tuberculosis treatment regardless of supplementation. In contrast,
supplementation probably does improve plasma levels of zinc, vitamin D, vitamin E, and selenium, but this has not been shown to have
clinically important benefits. Despite multiple studies of vitamin D supplementation in different doses, statistically significant benefits on
sputum conversion have not been demonstrated.
Authors' conclusions
Food or energy supplements may improve weight gain during recovery from tuberculosis in some settings, but there is currently no
evidence that they improve tuberculosis treatment outcomes. There is also currently no reliable evidence that routinely supplementing
above recommended daily amounts has clinical benefits.
Collaboration.
Collaboration.
SUMMARY OF FINDINGS
Summary of findings for the main comparison. 'Summary of findings' table 1
Library
Cochrane
Food provision compared with nutritional advice or no intervention for patients with active tuberculosis
Patient or population: adults and children with active tuberculosis
Settings: low- and middle-income countries
Better health.
Informed decisions.
Trusted evidence.
Intervention: calorie supplementation as food or energy dense supplements
Comparison: nutritional advice, micronutrient supplement, or no intervention
Outcomes Illustrative comparative risks* Relative ef- Number of Quality of the Comments
(95% CI) fect participants evidence
(95% CI) (trials) (GRADE)
Assumed risk Corresponding
risk
Standard Increased calo-

care rie intake
Death 3 per 100 1 per 100 RR 0.34 567 ⊕⊝⊝⊝ We don't know if food supplementation reduces mortal-
(0 to 4) (0.10 to 1.20) (4 trials) very low 1,2,3 ity from tuberculosis in food-insecure settings
(at 6 months)
Cured 48 per 100 44 per 100 RR 0.91 102 ⊕⊝⊝⊝ We don't know if food supplementation increases cure
(28 to 68) (0.59 to 1.41) (1 trial) very low 2,3,4 in tuberculosis patients
(at 6 months)
Treatment com- 79 per 100 85 per 100 Not pooled 365 ⊕⊝⊝⊝ We don't know if food supplementation increases treat-
pletion (70 to 100) (2 trials) very low 3,5,6 ment completion in tuberculosis patients
(at 6 months)

Sputum negative 76 per 100 82 per 100 RR 1.08 222 ⊕⊝⊝⊝ We don't know if food supplementation reduces the du-
(65 to 100) (0.86 to 1.37) (3 trials) very low 3,5,6 ration of sputum positivity in tuberculosis patients
(at 8 weeks)
Mean weight gain — — Not pooled 883 ⊕⊕⊕⊝ Supplementation probably increases weight gain during
(5 trials) moderate 7,8 treatment
(At 8 weeks)
Quality of life — — Not pooled 134 ⊕⊕⊝⊝ Supplementation may increase quality of life scores dur-
(At 8 weeks) (2 trials) low 9,10 ing the first 2 months of treatment
4
Collaboration.
*The assumed risk is taken from the mean risk in the control groups in the included studies. The corresponding risk (and its 95% CI) is based on the assumed risk in the
comparison group and the relative effect of the intervention (and its 95% CI).
Abbreviations: CI = confidence interval; RR = risk ratio; GRADE = Grading of Recommendations Assessment, Development and Evaluation.
Library
Cochrane
GRADE Working Group grades of evidence
High quality: further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: we are very uncertain about the estimate.
Better health.
Informed decisions.
Trusted evidence.
1Three trials reported some deaths during the 6 months of treatment (Jahnavi 2010 IND; Jeremiah 2014 TZA; Sudarsanam 2010 IND), and 1 reported that no deaths occurred
(Martins 2009 TLS). The trials were conducted in Tanzania, Timor-Leste, and India in participants with signs of undernutrition. Martins 2009 TLS gave a daily hot meal, Sudarsanam
2010 IND gave monthly ration packs, Jahnavi 2010 IND gave daily locally appropriate supplements, and Jeremiah 2014 TZA gave high energy multivitamin enriched biscuits.
2Downgraded by 1 for indirectness: trials are only available from limited settings. Food supplementation would plausibly have its biggest effect in highly food-insecure or
emergency settings which are not reflected in these trials.
3Downgraded by 2 for imprecision: the trials and meta-analysis are significantly underpowered to either detect or exclude an effect if it exists.
4Data on successful cure at 6 months is only available from Sudarsanam 2010 IND which randomized tuberculosis patients in India to monthly ration packs or advice only.
5Two trials report on tuberculosis treatment completion at 6 months (Jahnavi 2010 IND; Martins 2009 TLS). One trial was conducted in India and 1 in Timor-Leste in participants
with signs of undernutrition. Both trials gave daily locally appropriate supplements.
6Downgraded by 1 for inconsistency. Jahnavi 2010 IND found a statistically significant benefit while the larger trial, Martins 2009 TLS, did not.
7Five studies reported measures of weight gain but at different time-points, which prevented meta-analysis.
8Downgraded by 1 for inconsistency. Praygod 2011b TZA included only HIV-positive patients and although the trend was towards a benefit this did not reach statistical significance.
Jeremiah 2014 TZA noted a greater increase in mean weight gain in the supplemented group compared to the non-supplemented group after 8 weeks; however the difference
was not appreciable (1.09 kg, P < 0.6, authors' own figures). The 3 other trials all demonstrated clinically important benefits.
9Downgraded by 1 for indirectness. Only 2 small trials, 1 from Singapore (Paton 2004 SGP) and 1 from India (Jahnavi 2010 IND) report quality of life scores. The results can not
be generalized to other populations or settings with any certainty.
10Downgraded by 1 for imprecision. The presented data appear highly skewed and could not be pooled.
Summary of findings 2. 'Summary of findings' table 2
Multi-micronutrient supplementation compared with placebo for patients with active tuberculosis

Patient or population: adults and children with active tuberculosis
Settings: low- and middle-income countries
Intervention: multi-micronutrient supplements
Comparison: placebo or no intervention
Outcomes Illustrative comparative risks* Relative ef- Number of Quality of the Comments
(95% CI) fect participants evidence
(95% CI) (studies) (GRADE)
5
Collaboration.
Assumed risk Corresponding
risk
Library
Cochrane
Placebo Multi-micronutri-
ents
Death HIV-negative participants RR 0.86 1219 ⊕⊕⊝⊝ Multi-micronutrient supplements may have little or no effect
(0.46 to 1.6) (4 trials) low 1,2,3 on mortality in HIV-negative tuberculosis patients
40 per 1000 34 per 1000
Better health.
Informed decisions.
Trusted evidence.
(18 to 64)
HIV-positive participants RR 0.92 1429 ⊕⊕⊕⊝ Multi-micronutrients probably have little or no effect on mor-
(0.69 to 1.23) (3 trials) moderate 4,5 tality in HIV-positive tuberculosis patients not on ARV therapy
357 per 1000 328 per 1000
(246 to 439)
Cure rate — — — (0 trials) — We don't know if multi-micronutrients improve cure in tuber-

culosis patients
Treatment 970 per 1000 960 per 1000 RR 0.99 (0.95 302 ⊕⊝⊝⊝ We don't know if multi-micronutrients improve treatment
completion to 1.04) very low 6,7 completion in tuberculosis patients
(920 to 101) (1 trial)
Remaining 309 per 1000 312 per 1000 RR 0.92 1020 ⊕⊝⊝⊝ We don't know if multi-micronutrients reduce the proportion
sputum posi- (263 to 371) (0.63 to 1.35) (2 studies) very low of patients still sputum positive at 4 weeks
tive 8,9,10
(at 4 weeks)
Weight gain — — Not pooled 2940 ⊕⊕⊝⊝ Multi-micronutrient supplements may not improve weight
(5 trials) low 11 gain in tuberculosis patients
Quality of life — — — (0 trials) — We don't know if multi-micronutrients improve quality of life

in tuberculosis patients

*The assumed risk is taken from the risk in the control groups of the included studies. The corresponding risk (and its 95% CI) is based on the assumed risk in the compari-
son group and the relative effect of the intervention (and its 95% CI).
Abbreviations: CI = confidence interval; RR = risk ratio; GRADE = Grading of Recommendations Assessment, Development and Evaluation.
GRADE Working Group grades of evidence

High quality: further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: we are very uncertain about the estimate.
6
Collaboration.
1Five RCTs comparing multi-micronutrient supplementation with placebo in adults (Range 2005 TZA; Semba 2007 MWI; Villamor 2008 TZA) and children (Lodha 2014 IND; Mehta
2011 TZA), reported deaths during treatment. The exact composition of nutrients varied from 1 to 10 times the DRI. Three studies are from Tanzania, 1 from Malawi, and 1 from
India. There is evidence of participants being significantly undernourished at baseline.
Library
Cochrane
2No serious inconsistency: statistical heterogeneity was low. Although the estimate of effect is trending towards a benefit this is due to 1 trial (Villamor 2008 TZA), the remaining
trials found non significant trends in the opposite direction. Not downgraded.
3Downgraded by 2 for imprecision: the 95% CI of the pooled effect crosses 1 and includes a 2% absolute reduction in death which might be considered beneficial given the low
cost of the intervention. The optimal information size to reliably detect a clinically beneficial effect if there is 1 is greater than 2000.
4Downgraded by 1 for indirectness: none of the participants in these trials were receiving antiretroviral therapy (Range 2005 TZA; Semba 2007 MWI; Villamor 2008 TZA). The exact
composition of nutrients varied from 1 to 10 times the DRI. Two studies are from Tanzania and 1 in Malawi. There is evidence of participants being significantly undernourished
at baseline.
Better health.
Informed decisions.
Trusted evidence.
5No serious imprecision: the 95% CI of the pooled effect crosses 1 and does includes a 5% absolute reduction in death which might be considered beneficial given the low cost
of the intervention. However, the effect estimate is of no difference between the treatments and the optimal information size to reliably detect a clinically beneficial effect is met.
6Downgraded by 2 for indirectness: current evidence is limited to 1 small trial (Lodha 2014 IND) conducted in moderately undernourished, HIV-negative children in India, where
treatment completion was very high in both groups. The result is not easily generalized to other settings.
7Downgraded by 1 for imprecision: although this trial approaches adequate power to detect a result, the fact that treatment completion was so high in both arm renders the
result irrelevant to settings where treatment completion is lower.
8Downgraded by 1 for inconsistency. Statistical heterogeneity is high; 1 study found a non-significant trend in favour of supplementation, and 1 study in favour of placebo.
9Downgraded by 1 for indirectness. Both studies were conducted in Tanzania. Different populations may differ in their micronutrient deficiency or requirement.
10Downgraded by 1 for imprecision. The 95% CI is wide and includes what may be clinically important effects.
11Downgraded by 2 for inconsistency: Statistical heterogeneity is very high. Four studies from Tanzania and India found no evidence of improved weight gain with supplements,
while 1 study from Tanzania found large increases in weight with micronutrients at 8 months.

7
Informed decisions.
BACKGROUND these recommended amounts. The two important questions are

therefore whether tuberculosis patients require more, and whether
Description of the condition these increased requirements should be provided as part of routine
health care.
Tuberculosis is an infection caused by the bacterium
Mycobacterium tuberculosis, which is spread from person to person The effects of supplementation in people with HIV (but without
by inhalation of respiratory droplets (Harries 2006). In 2013, the tuberculosis) is covered in two other Cochrane Reviews (Irlam 2010;
World Health Organization (WHO) estimated that there were nine Grobler 2013).
million new cases of active tuberculosis worldwide, and 1.5 million
deaths (WHO 2014). Description of the intervention
Most people who are infected with M. tuberculosis develop what Nutritional requirements can be broadly divided into
is known as latent tuberculosis. People with latent tuberculosis macronutrients (carbohydrate, protein, and fat) and
are infected with the bacterium Mycobacterium tuberculosis, but micronutrients (essential vitamins and trace elements).
the infection is contained by their immune system. These people
remain well and do not exhibit any clinical signs or symptoms of Macronutrients
illness (Barry 2009). The immune response to infection is complex, Each day the average 70 kg male requires approximately 2500
it initially involves the uptake of the bacterium into macrophage kilocalories (kcal) of energy to maintain body weight and
cells as part of the non-specific 'innate' immune response, and later composition; ideally consumed as 55% carbohydrate, 15% protein,
recruitment of both B- and T-lymphocytes of the cellular immune and 30% fat (Meyers 2006).
response (Schluger 1998). These cells isolate the bacterium as a
granuloma, typically in the lung (Saunders 2007). If it was shown that patients with active tuberculosis required
additional macronutrients; these could be purchased and
Active tuberculosis occurs when the infection is no longer consumed by the patient simply following nutritional advice.
contained by the immune system, and can occur at any time However, in many situations, especially in low- and middle-income
following infection. The lifetime risk of conversion from latent to countries, the patient may not be able to acquire this additional
active tuberculosis is around 5% to 10% in an otherwise healthy food due to economic hardship through illness and loss of work,
population (Harries 2006), but this can rise to around 50% in people or due to local food insecurity (Kamolratanakul 1999; Wyss 2001).
with severe impairment of their immune system, such as occurs In these situations healthcare services might provide increased
with human immunodeficiency virus (HIV) infection (Zumla 2000; nutrients through free provision of meals, take home rations, or
Aaron 2004). specific high energy supplements. In many crisis or low-income
settings this already happens and the World Food Programme
Tuberculosis most commonly affects the lungs (pulmonary (WFP) in particular is involved in many food support programmes
tuberculosis), but can also spread to affect the central nervous for tuberculosis patients (WFP 2007).
system, lymphatic system, circulatory system, genitourinary
system, and bones and joints. The symptoms of active pulmonary Micronutrients
tuberculosis include cough, chest pain, fever, night sweats, weight
loss, and sometimes coughing up blood (Harries 2006). The daily micronutrient requirements for an adult male are given
in Appendix 1. These are usually expressed as the 'dietary reference
Tuberculosis is treated with a combination of antibiotic drugs intake' (DRI), and this is different for each individual micronutrient
(antituberculous therapy), which must be taken for a period of at (Meyers 2006).
least six months to ensure success (WHO 2010). If left untreated,
around half of those with active tuberculosis will die of the disease These requirements can be gained from the consumption
(Corbett 2003). With adequate treatment the mortality is around 5% of a healthy, and varied diet, or through pharmaceutical
globally (WHO 2009), although this may be higher in HIV-positive supplementation as tablets, capsules, or powders. Any additional
people (Aaron 2004). The WHO target for successful cure in national requirements could be gained through increased consumption
tuberculosis control programmes is 85% (WHO 2009). following dietary advice, or through pharmaceutical provision via
the health service or tuberculosis programme.
Throughout the world, poor nutritional status is more common in
people with active tuberculosis than in people without tuberculosis In trials of macronutrient and micronutrient interventions two
(van Lettow 2003), and weight loss, including loss of lean body important factors should be noted.
mass, is a well-recognized symptom of the disease. Cohort and
• The intervention is a supplement and does not represent the
cross-sectional studies have suggested that active tuberculosis
total daily intake of that nutrient.
is commonly associated with low serum levels of important
micronutrients such as zinc (Taneja 1990), and vitamins A, C, • Any benefit derived from the intervention is likely to be
D, and E (Davies 1985; Plit 1998; Nnoaham 2008). However, the dependant on the initial nutritional status of the patient.
measurement of serum vitamin levels during an acute infection,
In order to accurately interpret data it is therefore essential to
such as tuberculosis, is known to be unreliable, as transient
consider both the baseline nutritional status, and the overall
abnormalities can occur (Louw 1992).
nutritional intake of the patients.
The daily nutritional requirements for healthy individuals of all
age groups are well described (Meyers 2006), and it is unlikely
that people with active tuberculosis would require less than
Collaboration.
Informed decisions.
How the intervention might work to gain weight and recover from tuberculosis, and highlight where
more research might be needed.
Tuberculosis and undernutrition interact in a two-way process.
Tuberculosis can lead to weight loss and micronutrient deficiencies OBJECTIVES
by increasing nutritional requirements, by changing metabolic
processes, or by decreasing appetite and causing a reduction in To assess the effects of oral nutritional supplements in people being
food intake (Macallan 1999). Alternatively, low body mass index treated with antituberculous drug therapy for active tuberculosis.
(BMI; a measure of weight for height that is indicative of nutritional
status) and some micronutrient deficiencies can depress cell- METHODS
mediated immunity, the key host defence against tuberculosis,
increasing the susceptibility to active tuberculosis and delaying Criteria for considering studies for this review
recovery (Chandra 1996; Zachariah 2002; Cegielski 2004). Types of studies
The micronutrients which have received the most attention are the Randomized controlled trials (RCTs).
following.
Types of participants
• Vitamin A, which is involved in both T- and B-lymphocyte
function, macrophage activity and the generation of antibody Children or adults being treated for active tuberculosis with or
responses (Semba 1998; Stephensen 2001). without concurrent HIV infection, and with or without a diagnosis
of being underweight, malnourished, or nutrient deficient.
• Vitamin D, which is involved in the function of macrophages,
a key component of the immune response to tuberculosis Types of interventions
(Wintergerst 2007).
Intervention
• Vitamin E, which has anti-oxidant properties and may protect
against T-lympocyte failure due to oxidative stress (Wintergerst Any oral macro or micronutrient supplement given for at least four
2007). weeks.
• Zinc, which is necessary for adequate functioning of many
aspects of human immunity (Shankar 1998). We excluded trials that assessed tube feeding or parenteral
nutrition, and trials that assessed dietary advice alone without the
• Selenium; which is essential for both cell-mediated and humoral
actual provision of supplements.
immunity (Arthur 2003).
Control
Nutritional interventions, in people with active tuberculosis,
therefore have the potential to do the following. No nutritional intervention, placebo, or dietary advice alone.
• Improve tuberculosis treatment outcomes; through restoration Types of outcome measures

of cell-mediated immunity, increasing the individuals' ability to
Primary outcomes
fight the infection and hastening recovery from the illness.
• Promote nutritional recovery; with improved weight gain, • All-cause death.
restoration of muscle strength, function, and quality of life. • Cure (completed treatment and sputum-smear or sputum-
Nutritional recovery is of great importance in tuberculosis culture negative) at six and 12 months.
treatment, allowing the patient to return to work, and recover
economically as well as physically. Secondary outcomes
Food may also be given to people with tuberculosis for quite • Completion of treatment.
different reasons, such as; to promote adherence to treatment, • Sputum positive at follow-up.
or to mitigate the financial consequences of prolonged illness. • Self-reported recovery from illness or resolution of symptoms.
Another Cochrane Review is addressing the use of food to promote • Change in weight, skinfold thickness, or other measure of lean
adherence (Lutge 2009). or total mass.
• Any measure of growth in children.
In addition, it is important to note that pathogens such as
tuberculosis also require certain micronutrients for their own • Any measures of physical functioning, quality of life, or ability to
metabolism, and greater availability of these nutrients through return to work.
supplementation could encourage their growth. There is some • Total calorie intake.
evidence for this in the case of iron (Lounis 2001), and so nutritional • Micronutrient levels before and after supplementation.
interventions cannot be considered entirely benign.
We intended to include cure assessed at six and 12 months, as
Why it is important to do this review is customary. For other outcome measures, we accepted data
presented at any time point.
There is currently no evidence-based guidance on food provision
or supplementation for adults or children being treated for Search methods for identification of studies
tuberculosis, with or without concurrent HIV infection. This
Cochrane Review seeks to assess the evidence for the effectiveness We attempted to identify all relevant trials regardless of language
of different food and nutritional supplements in helping people or publication status (published or unpublished, in press, or in
progress).
Collaboration.
Informed decisions.
Electronic searches adequate steps were taken to reduce the risk of bias across six
domains: sequence generation, allocation concealment, blinding
We searched the following databases using the search terms and
(of participants, personnel, and outcome assessors), incomplete
strategy described in Appendix 2.
outcome data, selective outcome reporting, and other sources
• Cochrane Infectious Disease Group Specialized Register. of bias. We categorized our judgements as either 'low', 'high',
or 'unclear' risk of bias. Where our judgment was 'unclear', we
• Cochrane Central Register of Controlled Trials (CENTRAL; Issue
attempted to contact the trial authors for clarification.
1, 2016).
• MEDLINE (from 1946 to 4 February 2016). Measures of treatment effect
• EMBASE (from 1980 to 4 February 2016).
We compared interventions using risk ratios (RR) for dichotomous
• LILACS (from 1982 to 4 February 2016). data, and mean difference (MD) values for continuous data. We
presented all results with 95% confidence intervals (CIs).
We checked the metaRegister of Controlled Trials (mRCT) and
the World Health Organization (WHO) International Clinical Trials Unit of analysis issues
Registry Platform (ICTRP) (www.who.int/ictrp/search/en) up to
4 February 2016, using 'tuberculosis' and 'supplementation' as We split trials that included more than two comparison groups
search terms. and analysed them as individual pair-wise comparisons. When
we conducted meta-analysis we ensured that we did not count
In addition we searched the Indian Journal of Tuberculosis using the participants and cases in the placebo group more than once, by
keywords given in the search strategy (Appendix 2) on 4 February dividing the placebo cases and participants evenly between the
2016. intervention groups.
Searching other resources Dealing with missing data

We also checked the reference lists of all studies identified by the If data from the trial reports were insufficient, unclear, or
above methods. missing, we attempted to contact the trial authors for additional
information.
Data collection and analysis
Assessment of heterogeneity
Selection of studies
We assessed heterogeneity amongst trials by inspection of the
Two review authors (Liesl Grobler (LG) and Sukrti Nagpal (SN)/ forest plots (to detect overlapping CIs), the I2 statistic with a level
Thambu Sudarsanam (TS)) independently screened all citations of 50% to denote moderate levels of heterogeneity, and application
and abstracts identified in the search for potentially eligible studies. of the Chi2 test with a P value of 0.10 to indicate appreciable
Full reports of potentially eligible studies were obtained and heterogeneity.
assessed for inclusion in the review by the same authors using a
pre-designed eligibility form based on the inclusion criteria. Where Assessment of reporting biases
it was unclear whether a study was eligible for the review, we
attempted to contact the authors for clarification. We resolved We planned to assess the likelihood of publication bias by
differences in opinion by discussion and, where necessary, by examining the funnel plots for asymmetry. However, there were too
discussion with a fourth author (David Sinclair). We screened all few trials to make this assessment meaningful.
papers for multiple publications. We excluded studies that did not
Data synthesis
meet the criteria and documented the reasons for their exclusion.
We analysed the data using Review Manager (RevMan) (RevMan
Data extraction and management 2014). Using pair-wise meta-analyses we compared the treatments.
Two review authors (LG and SN) independently extracted data We stratified meta-analyses by time-point or HIV status where
using a tailored data extraction form. We extracted data on study appropriate.
design, participant characteristics, interventions, and outcomes.
When there was no statistically significant heterogeneity we used
For dichotomous data, we extracted the number of participants the fixed-effect meta-analysis model. When we observed moderate
with the outcome and the total number analysed. For continuous statistically significant heterogeneity within groups that we could
data, we extracted the arithmetic mean and standard deviation not explain by subgroup or sensitivity analyses we used a random-
(SD) for each group. If medians were used, we also extracted the effects meta-analysis model to synthesize the data. When a pooled
ranges where possible. If there was skewed continuous data, we meta-analysis result was considered to be meaningless because
planned to extract geometric means where presented by the trial of clinical or substantial statistical heterogeneity the results are
author(s). We resolved any discrepancies regarding extracted data presented in a forest plot without a pooled estimate of effect.
by discussion between the review authors.
Data presented as medians and ranges are presented in tables and
Assessment of risk of bias in included studies described in the narrative. Highly skewed continuous data (where
the SDs were larger than the means) are only presented in tables.
Two review authors (LG and SN) independently assessed
components of risk of bias of the included trials using the Cochrane Subgroup analysis and investigation of heterogeneity
'Risk of bias' assessment tool (RevMan 2014), and discussed any
Due to the small number of included trials or each comparison,
differences of opinion. We followed the guidance to assess whether
the investigation of heterogeneity was not necessary or possible. In
Collaboration.
Informed decisions.
future updates of this Cochrane review we may perform subgroup RESULTS

analyses by: HIV status, nutritional status at baseline, presence of
co-morbidities, and specific micronutrient level at baseline. Description of studies
Sensitivity analysis Results of the search
We have planned to perform a sensitivity analysis to investigate the The original search (June 2008) identified 47 articles, and the
robustness of the results to the 'Risk of bias' components. However, updated search (July 2011) identified a further 17 articles. The
there were too few included trials for each comparison for this to searches conducted in December 2014 and updated in February
be possible. 2016 identified 48 published articles and 16 ongoing studies. In
total the current review includes data from 39 reports, covering 35
individually randomized controlled trials (RCTs) (see Figure 1).
Figure 1. Study flow diagram.
Included studies extrapulmonary tuberculosis (Morcos 1998 EGY; Mehta 2011 TZA;
Lodha 2014 IND).
Participants
The 35 trials included 8285 participants. Twenty-eight trials HIV status
included only adults being treated for pulmonary tuberculosis, and Eleven trials specifically included people with HIV, and presented
three trials also included adults with extrapulmonary tuberculosis some results separately for HIV positive and HIV negative
(Wejse 2008 GNB; Jahnavi 2010 IND; Sudarsanam 2010 IND). participants (Schön 2003 ETH; Range 2005 TZA; Semba 2007 MWI;
Likewise of four studies in children, one included only pulmonary Villamor 2008 TZA; Wejse 2008 GNB; Lawson 2010 NGA; Mehta
tuberculosis (Hanekom 1997 ZAF), and three also included 2011 TZA; Sudarsanam 2010 IND; Schön 2011 ETH; Ralph 2013 IDN;
Collaboration.
Informed decisions.
Jeremiah 2014 TZA). Four of the trials that presented results for HIV- Interventions
positive and HIV-negative participants separately used stratified
Seven trials assessed macronutrient supplementation (Paton 2004
randomization (Villamor 2008 TZA; Wejse 2008 GNB; Sudarsanam
SGP; Pérez-Guzmán 2005 MEX; Martins 2009 TLS; Jahnavi 2010 IND;
2010 IND; Jeremiah 2014 TZA), but six tested participants for HIV
Sudarsanam 2010 IND; Praygod 2011b TZA; Jeremiah 2014 TZA), six
following randomization (Schön 2003 ETH; Range 2005 TZA; Semba
trials assessed multi-micronutrient supplementation (Range 2005
2007 MWI; Lawson 2010 NGA; Mehta 2011 TZA; Schön 2011 ETH),
TZA; Semba 2007 MWI; Villamor 2008 TZA; Mehta 2011 TZA; Praygod
and therefore the subgroup analyses by HIV status cannot be said
2011a TZA; Lodha 2014 IND), and 21 trials assessed single or dual
to be truly randomized. Four trials included both HIV-positive and
micronutrient supplementation. The remaining trial assessed the
HIV-negative individuals and reported on numbers but did not
effect of supplementation with Channa striata capsules (Paliliewu
present results separately (Martineau 2011 GBR; Praygod 2011a
2013 IDN). C. striata is a fresh-water fish found in most tropical and
TZA; Visser 2011 ZAF; Tukvadze 2015 GEO). One trial included
subtropical countries.
only HIV-positive participants (Praygod 2011b TZA). None of the
HIV-positive participants were receiving antiretroviral treatment. Sample size
One trial was stopped early when antiretrovirals became locally
available (Semba 2007 MWI). Eleven trials excluded people with Eleven of the 35 trials included less than 100 participants in their
HIV infection (Hanekom 1997 ZAF; Paton 2004 SGP; Pérez-Guzmán final analysis. To aid interpretation and inform future research we
2005 MEX; Nursyam 2006 IDN; Armijos 2010 MEX; Jahnavi 2010 IND; have calculated the optimal sample size to reliably demonstrate
Paliliewu 2013 IDN; Lodha 2014 IND; Daley 2015 IND; Farazi 2015 some suggested clinically important results (Appendix 3; Appendix
IRN; Mily 2015 BGD), and eight trials did not mention HIV infection 4).
(Morcos 1998 EGY; Karyadi 2002 IDN; Seyedrezazadeh 2006 IRN;
Martins 2009 TLS; Pakasi 2010 IDN; Kota 2011 IND; Ginawi 2013 IND; As micronutrients are a cheap and easily administered
Singh 2013 IND). intervention, even a small effect on tuberculosis treatment
outcomes might be considered clinically important. For example,
Study site to demonstrate a reduction in death from the worldwide average
of 5% to just 4% (a relative risk reduction of 20%); a sample size
Trials were undertaken in the following locations. of over 13,000 participants would be necessary. This is far above
the data included in this Cochrane Review. Similarly an increase in
• Africa: Egypt (Morcos 1998 EGY), Guinnea-Bissau (Wejse 2008
successful cure rate from 80% to 84% would require almost 3000
GNB), Ethiopia (Schön 2003 ETH; Schön 2011 ETH), Tanzania
participants.
(Range 2005 TZA; Villamor 2008 TZA; Mehta 2011 TZA; Praygod
2011b TZA; Praygod 2011a TZA; Jeremiah 2014 TZA), Malawi For full details of the included trials see the 'Characteristics of
(Semba 2007 MWI), Nigeria (Lawson 2010 NGA), and South Africa included studies' table.
(Hanekom 1997 ZAF; Visser 2011 ZAF).
• Asia: Singapore (Paton 2004 SGP), Indonesia (Karyadi 2002 Excluded studies
IDN; Nursyam 2006 IDN; Pakasi 2010 IDN; Paliliewu 2013 IDN;
We excluded 17 studies that we had thought were eligible
Ralph 2013 IDN), Iran (Seyedrezazadeh 2006 IRN; Farazi 2015
after initial screening for the reasons we have given in the
IRN), Bangladesh (Mily 2015 BGD), and India (Jahnavi 2010 IND;
'Characteristics of excluded studies' table. Five trials are currently
Sudarsanam 2010 IND; Kota 2011 IND; Ginawi 2013 IND; Singh
awaiting classification pending further information from the author
2013 IND; Lodha 2014 IND; Daley 2015 IND).
(Chandra 2004; Guzman-Rivero 2013; Nagrale 2013; Nawas 2013; Al
• Oceania: Timor Leste (Martins 2009 TLS). Mamun 2014). In addition we are aware of 17 potentially relevant
• North America: Mexico (Pérez-Guzmán 2005 MEX; Armijos 2010 ongoing or unpublished trials; see the 'Characteristics of ongoing
MEX). studies' table.
• Europe: UK (Martineau 2011 GBR) and Georgia (Tukvadze 2015
GEO). Risk of bias in included studies
We have displayed the 'Risk of bias' assessments in a table and
summarised these results in Figure 2.
Collaboration.
Informed decisions.
Figure 2. 'Risk of bias' summary: review authors' judgements about each 'Risk of bias' item for each included
study.
Collaboration.
Informed decisions.
Figure 2. (Continued)
Allocation Twenty-one of the 27 trials assessing micronutrients used placebos

and adequately blinded participants and study staff to be
Twenty-four trials described an adequate method of generating a
considered at low risk of bias.
truly random allocation sequence (Karyadi 2002 IDN; Schön 2003
ETH; Paton 2004 SGP; Range 2005 TZA; Semba 2007 MWI; Villamor Incomplete outcome data
2008 TZA; Wejse 2008 GNB; Martins 2009 TLS; Lawson 2010 NGA;
Pakasi 2010 IDN; Sudarsanam 2010 IND; Martineau 2011 GBR; Mehta We considered 12 studies at high risk of bias for some outcomes
2011 TZA; Praygod 2011a TZA; Praygod 2011b TZA; Schön 2011 ETH; due to high losses to follow-up (that is, loss to follow-up > 10%
Visser 2011 ZAF; Ralph 2013 IDN; Jeremiah 2014 TZA; Lodha 2014 overall, between groups or within a group (Hanekom 1997 ZAF;
IND; Daley 2015 IND; Farazi 2015 IRN; Mily 2015 BGD; Tukvadze 2015 Karyadi 2002 IDN; Paton 2004 SGP; Range 2005 TZA; Semba 2007
GEO). The other trials did not report how the random sequences MWI; Martins 2009 TLS; Pakasi 2010 IDN; Sudarsanam 2010 IND;
were generated although all were described as "randomized". Visser 2011 ZAF; Ginawi 2013 IND; Jeremiah 2014 TZA; Tukvadze
2015 GEO).
Eighteen trials described an adequate method of ensuring
allocation concealment (Schön 2003 ETH; Semba 2007 MWI; Wejse Selective reporting
2008 GNB; Martins 2009 TLS; Lawson 2010 NGA; Sudarsanam 2010 Although treatment outcomes in tuberculosis such as cure,
IND; Martineau 2011 GBR; Mehta 2011 TZA; Praygod 2011a TZA; and treatment completion are well established in tuberculosis
Praygod 2011b TZA; Schön 2011 ETH; Visser 2011 ZAF; Ralph 2013 programmes only one trial reported cure (Sudarsanam 2010 IND),
IDN; Lodha 2014 IND; Daley 2015 IND; Farazi 2015 IRN; Mily 2015 and five reported treatment completion (Martins 2009 TLS; Jahnavi
BGD; Tukvadze 2015 GEO). The other trials did not provide sufficient 2010 IND; Pakasi 2010 IDN; Sudarsanam 2010 IND; Farazi 2015 IRN).
information to determine if the allocation sequence was truly We were unable to retrieve trial protocols.
concealed from the person allocating participants to the treatment
groups. Other potential sources of bias
Blinding One trial did not adequately describe baseline nutritional status
(Morcos 1998 EGY), one trial had a large imbalance in HIV status at
It is generally not possible to blind patients to macronutrient baseline (Mehta 2011 TZA), and one trial had appreciable difference
supplementation. However, it is possible to blind the outcome in both HIV status and severity of chest X-ray at baseline (Ralph 2013
assessors but only one of the six studies reports an attempt to do IDN).
this (Martins 2009 TLS).
Collaboration.
Informed decisions.
Effects of interventions Jahnavi 2010 IND also found that more participants given
supplements were smear negative at eight weeks, while Martins
See: Summary of findings for the main comparison 'Summary 2009 TLS and Jeremiah 2014 TZA found no statistically significant
of findings' table 1; Summary of findings 2 'Summary of findings' difference (RR 1.08, 95% CI 0.86 to 1.37; three trials, 222
table 2 participants; Analysis 1.4).
1. Macronutrients Nutritional recovery and quality of life
1.1 Increased energy supplementation (average daily Effects on weight gain were mixed (five trials, 689 participants,
requirement for a male adult: 2500 kcal) Analysis 1.5). The daily hot meal was associated with an extra 1.7
Six trials examined the effects of providing macronutrient kg weight gain at eight weeks (95% CI 0.1 to 3.2; P = 0.04; authors'
supplements through the health service. own figures), and 2.6 kg at eight months (95% CI 0.1 to 5.2; P =
0.04; authors' own figures; Martins 2009 TLS), whereas the monthly
• In Timor Leste, adults with pulmonary tuberculosis and a mean ration pack was not associated with important differences (change
weight of 43.3 kg (31% of participants had a BMI of less than in lean body mass, and percentage body fat were not significantly
16 kg/m2) were randomized to nutritional advice plus a daily different between groups; P = 0.479 and P = 0.573 respectively;
cooked meal or nutritional advice alone (Martins 2009 TLS). The authors' own figures; Sudarsanam 2010 IND).
daily midday meal (administered for two months during the
intensive phase) consisted of a bowl of meat, kidney beans, Of the trials evaluating high energy supplements, the two smaller
and vegetable stew with rice. During the continuation phase trials in HIV-negative participants found that supplementation
participants in the supplement group also received a weekly resulted in significantly more weight gain than advice alone at six
food parcel containing unprepared red kidney beans, rice, and weeks (MD 1.73 kg, 95% CI 0.81 to 2.65; one trial, 34 participants;
oil adequate for one meal per day. Analysis 1.5) and 12 weeks, respectively (MD 2.6 kg, 95% CI 1.74
to 3.46; one trial, 100 participants; Analysis 1.5). Paton 2004 SGP
• In India, adults with pulmonary tuberculosis and a BMI of less
further quantified this as an increase in lean body mass (MD
than 19 kg/m2 were randomized to receive a macronutrient and
1.13 kg, 95% CI 0.37 to 1.89; one trial, 34 participants), with
micronutrient supplement plus standard care versus standard
no significant difference in total fat mass. However, in Tanzania
care alone (Sudarsanam 2010 IND). The supplement consisted of
Praygod 2011b TZA found no significant difference in weight gain
three daily servings of a cereal and lentil mixture (providing 930
with supplementation in HIV-positive participants at eight or 20
kcal and 31.5 g protein) and a once-a-day multivitamin tablet.
weeks (one trial, 332 participants, Analysis 1.5). Similarly, Jeremiah
Participants were given a month's supply of supplement at a
2014 TZA found no difference in final mean weight between
time.
the supplement and no supplement group after two months of
• The remaining four trials compared daily high energy supplementation (one trial, 92 participants, Analysis 1.5).
supplements with: nutritional advice alone (Paton 2004 SGP;
Jahnavi 2010 IND); a single multi-micronutrient biscuit (Praygod Three trials of high energy supplements (Paton 2004 SGP; Jahnavi
2011b TZA); or no supplement (Jeremiah 2014 TZA). The 2010 IND; Praygod 2011b TZA) report changes in maximum grip
supplements were described as 'sweetballs' (made from wheat strength, and again a statistically significant benefit was seen in the
flour, caramel, ground nuts, and vegetable ghee) (Jahnavi small trials of HIV-negative participants but not the larger trial of
2010 IND), high energy 'packets' (Paton 2004 SGP), and high HIV-positive participants although the data appear skewed (three
energy 'biscuits' (Praygod 2011b TZA; Jeremiah 2014 TZA) (see trials, 466 participants, Analysis 1.6).
Appendix 5). None of the studies estimated the total daily energy
intake, but the supplement provided between 600 and 3690 kcal Jahnavi 2010 IND and Paton 2004 SGP also reported that the
per day on top of the regular diet. All four studies recruited benefits on weight gain and grip strength were accompanied by
people with mean BMI below 20 kg/m2. improvements in some quality of life scores. It was not possible
to assess whether these difference were statistically significant
Only two trials measured total dietary intake. Both trials confirmed because the data appeared highly skewed (the SDs were larger than
that supplementation had increased nutritional intake compared the means for most outcomes); see Analysis 1.7 and Appendix 6.
to dietary advice alone, and not simply substituted food that
patients might have obtained elsewhere (Paton 2004 SGP; 1.2 Altered dietary composition
Sudarsanam 2010 IND).
One very small trial of 21 participants compared a high cholesterol
Tuberculosis treatment outcomes diet (2500 kcal per day with 800 mg cholesterol per day) with a diet
with a similar nutritional profile but lower in cholesterol (2500 kcal
The number of deaths reported from these trials was very low, per day with 250 mg cholesterol per day) for eight weeks in adults
and trials were too small to reliably detect or exclude important being treated for sputum-culture positive pulmonary tuberculosis
differences in mortality (risk ratio (RR) 0.34, 95% confidence interval (Pérez-Guzmán 2005 MEX).
(CI) 0.10 to 1.20; four trials, 567 participants; Analysis 1.1), or cure
(RR 0.91, 95% CI 0.59 to 1.41; one trial, 102 participants; Analysis Tuberculosis treatment outcomes
1.2). Jahnavi 2010 IND found a statistically significant difference in
This trial did not report on death, cure, or treatment completion.
treatment completion in favour of supplementation, but this was
not seen in the larger trial conducted in Timor Leste (two trials, 365 Fewer participants in the high cholesterol group were still sputum-
participants, Analysis 1.3). culture positive at two weeks compared with those in the normal
cholesterol group (RR of remaining sputum positive at two weeks
0.22, 95% CI 0.06 to 0.77; one trial, 20 participants, Analysis 2.1),
Collaboration.
Informed decisions.
the difference was not significant at four weeks, and at eight weeks Range 2005 TZA found that participants receiving multiple
all participants in both groups were sputum-culture negative. Self micronutrients had gained significantly more weight at seven
reported cough and dyspnoea are reported to have decreased at months than those in the placebo group. This was a 2 x 2
the same rate in both groups (figures not given). factorial study and the difference in weight was appreciable in
both treatment groups compared to placebo. In the treatment arm
Nutritional recovery and quality of life that received both high dose multivitamins and high dose zinc,
This trial did not report on any aspect of the participant's the weight gain appeared clinically important (MD 2.37 kg, 95% CI
nutritional recovery or quality of life. 2.21 to 2.53; one trial, 192 participants, Analysis 3.6). However, in
the treatment arm that received high-dose multivitamins alone,
2. Micronutrients the weight gain was minimal (MD 0.30 kg, 95% CI 0.17 to 0.43;
one trial, 198 participants, Analysis 3.6). In Lodha 2014 IND,
2.1 Multivitamins and trace elements
multi-micronutrient supplementation with or without zinc did not
Four trials in adults being treated for sputum test positive or consistently alter children's weight, BMI-for-age z score or height-
negative pulmonary tuberculosis (Range 2005 TZA; Semba 2007 for-age z score (one trial, 198 participants; Analysis 3.7).
MWI; Villamor 2008 TZA; Praygod 2011a TZA) and two trials in
children (Mehta 2011 TZA; Lodha 2014 IND) compared daily multiple One study, Praygod 2011a TZA, found an appreciable improvement
micronutrient supplements (containing a range of vitamins and in mean handgrip strength at two months but not five months
trace elements), with placebo. The exact doses of the individual (mean difference (change in handgrip strength) 1.22 kg, 95%
constituents ranged from one to 10 times the dietary reference CI 0.49 to 1.95; one trial, 771 participants; Analysis 3.8). The
intake (DRI), and are given in Appendix 1. In summary: vitamin A (1 clinical importance of this difference is unclear. Consistent with the
to 3 x DRI), B vitamins (1 to 10 x DRI), vitamin C (1 to 5 x DRI), vitamin change in weight, this increase was only present in HIV-negative
D (approximately 1 x DRI), vitamin E (1 to 10 x DRI), zinc (1 to 5 x DRI), participants.
and selenium (1 to 4 x DRI).
2.2 Individual micronutrients
In two trials participants received daily supplements for two Vitamin A (DRI: 900 µg/3000 IU per day)
months (Mehta 2011 TZA; Praygod 2011a TZA), one trial for six
Supplement dosing regimes
months (Lodha 2014 IND), one trial for eight months (Range 2005
TZA), and in two trials they received daily supplements for 24 Three trials directly compared vitamin A given alone versus placebo
months (Semba 2007 MWI; Villamor 2008 TZA). (Ginawi 2013 IND: vitamin A 5000 IU; Hanekom 1997 ZAF: vitamin
A 200,000 IU on Day 0 and Day 1; Pakasi 2010 IDN: vitamin A 5000
Tuberculosis treatment outcomes
IU daily). In addition, seven trials combined vitamin A with zinc
In trials with HIV negative people the number of deaths from (see Comparison 6: Zinc plus vitamin A versus placebo), and five
tuberculosis was low and the trials substantially underpowered studies gave vitamin A as part of a multi-micronutrient supplement
to demonstrate an effect, Consequently, the 95% CI is wide (see Comparison 3: Multivitamin and trace element tablets versus
and includes clinically important benefit and harm (RR 0.86, placebo).
95% CI 0.46 to 1.60, four trials, 1219 participants, Analysis 3.1).
Serum vitamin A concentrations at baseline and follow-up
Deaths were more common in trials with HIV positive people not
taking antiretroviral therapy, but again no differences between Seven studies report on measures of vitamin A status, but Pakasi
micronutrients and placebo were demonstrated (RR 0.92, 95% CI 2010 IDN and Visser 2011 ZAF reported data as median plasma
0.69 to 1.23, three trials, 1429 participants, Analysis 3.1). levels so could not be included in the meta-analysis, and Semba
2007 MWI only presented data graphically (see Table 2). Only Semba
Lodha 2014 IND reported no difference in tuberculosis treatment 2007 MWI reports a difference that was statistically significant in
completion at six months between micronutrients (with and favour of supplements, but this difference (at eight months) is
without zinc) and placebo (one trial, 302 participants, Analysis 3.2). unlikely to be of clinical significance.
There was no statistically significant difference between the In four trials in adults and one in children, mean serum
supplement and control groups in the numbers of participants who vitamin A level substantially increased in both intervention and
remained sputum-culture or sputum-smear positive at one month control arms regardless of supplementation, and there was no
(two trials, 1020 participants, Analysis 3.3) or two months (two statistically significant difference between the groups (one trial, 85
trials, 731 participants, Analysis 3.4). participants; Analysis 4.1; three trials, 242 participants; Analysis 4.2;
Table 2).
Lodha 2014 IND and Mehta 2011 TZA reported no appreciable
difference in chest x-ray clearance (after two or six months of Tuberculosis treatment outcomes
supplementation) between the supplemented children and the
placebo children at follow-up (two trials, 497 participants, Analysis Supplementation with vitamin A alone or in combination with other
3.5). micronutrients appears to have little or no effect on mortality (eight
trials, 3308 participants; Analysis 4.3), but these trials, even the
Nutritional recovery and quality of life larger ones, were significantly underpowered to rule out a clinically
important effect.
Five trials reported changes in weight or body mass using a variety
of parameters, and only Range 2005 TZA reported statistically Only Pakasi 2010 IDN reported on treatment completion and
significant effects (see Table 1). found no statistically significant effect (one trial, 158 participants,
Analysis 4.4). Hanekom 1997 ZAF reported that more children in
Collaboration.
Informed decisions.
the supplement group remained symptomatic after six weeks of larger ones, are significantly underpowered to rule out a clinically
tuberculosis treatment than in the control group, but this was not important effect.
statistically significant (one trial, 76 participants, Analysis 4.5). The
trial authors also reported no statistically significant differences in There was no appreciable differences in treatment completion
respiratory symptoms at three months, or in chest x-ray resolution; between the zinc alone and placebo groups (two trials, 353
but specific data on these outcomes were not provided. participants, Analysis 5.4), and there were no differences between
the groups in the numbers who remained sputum-culture positive
Supplementation with vitamin A alone had no effect on the number at four weeks (three trials, 783 participants, Analysis 5.5) or eight
of participants who were sputum smear negative after two weeks, weeks (five trials, 1076 participants, Analysis 5.5). Furthermore,
one month, or two months (two trials, 224 participants, Analysis Lodha 2014 IND reports no difference in chest X-ray clearance after
4.6). Pakasi 2010 IDN reported that all participants in the vitamin A six months between the zinc and placebo groups (one trial, 201
and placebo arm were smear negative at two months. participants, Analysis 5.6).
Nutritional recovery and quality of life Nutritional recovery and quality of life
Hanekom 1997 ZAF reported that the mean weight z score at One trial in children in India found no difference in mean weight
baseline was −1.41 (SD 1.41) in the supplement group, and −1.44 (kg), weight-for-age z score, BMI-for-age z score or height-for-age
(SD 1.34) in the placebo group (it is unclear whether this is weight z score after two months or six months supplementation with
for age or weight for height). The trial authors also reported that zinc alone (one trial, 201 participants, Analysis 5.7; Analysis 5.10;
no statistically significant differences in change-in-weight z scores Analysis 5.11; Analysis 5.12). Similarly, Pakasi 2010 IDN found no
were recorded at any time-point, but the specific data for this difference in BMI or body fat (%) in adults at two or six months
outcome was not provided. (one trial, 162 participants, Analysis 5.8 and Analysis 5.9; Pakasi
2010 IDN), and Lawson 2010 NGA presented data on changes in BMI
Pakasi 2010 IDN reports that the mean BMI at baseline was 16.5 graphically, and BMI appeared to improve at the same rate in all
kg/m2 (SD 2.2) in the supplement group and 16.4 kg/m2 (SD 2.5) in groups (Lawson 2010 NGA).
the placebo group. There was no statistically significant difference
in mean BMI between groups at two or six months (one trial, Range 2005 TZA reported a very small but statistically significant
158 participants, Analysis 4.7). Pakasi 2010 IDN also reported that decrease in weight gain with supplementation compared to
there were no statistically significant differences in mid upper arm placebo (MD −0.21 kg, 95% CI −0.36 to −0.06; one trial 183
circumference (MUAC) or percentage body fat (Analysis 4.8). participants; Analysis 5.7).
Zinc (DRI: 11 mg/day) Vitamin A plus zinc (DRI: vitamin A 900 µg/3000 IU, zinc 11 mg per day)
Supplement dosing regimes Supplement dosing regimes
Five trials directly compared daily zinc given alone versus placebo Seven studies in adults with sputum positive pulmonary
(Ginawi 2013 IND: 15 mg zinc sulphate; Lodha 2014 IND: 20 mg tuberculosis compared the combination of vitamin A and zinc
elemental zinc; Range 2005 TZA: zinc 45 mg daily; Lawson 2010 versus placebo (Ginawi 2013 IND: vitamin A 5000 IU and 15 mg
NGA: 90 mg elemental zinc weekly; Pakasi 2010 IDN: 15 mg zinc zinc sulphate; Singh 2013 IND: 25000 IU vitamin A and 50 mg zinc
sulphate daily). In addition, seven trials combined vitamin A with sulphate, once daily for 10 days and then thrice weekly until six
zinc (see Comparison 6: zinc plus vitamin A versus placebo), and months; Karyadi 2002 IDN: vitamin A 5000 IU and zinc 15 mg daily
four trials gave zinc as part of a multi-micronutrient supplement for six months; Armijos 2010 MEX: vitamin A 5000 IU plus zinc 50 mg
(see comparison 3: Multivitamin and trace element tablets versus daily for four months; Lawson 2010 NGA: vitamin A 5000 IU/day plus
placebo). 90 mg elemental zinc/week for six months; Pakasi 2010 IDN: vitamin
A 5000 IU plus 15 mg zinc sulphate daily for six months; Visser 2011
Serum zinc concentrations at baseline and follow-up ZAF: vitamin A 100,000 IU at baseline plus zinc 15 mg for five days
Five studies report mean plasma zinc levels at baseline and during per week for two months).
follow-up (see Table 3). Pakasi 2010 IDN and Armijos 2010 MEX
Tuberculosis treatment outcomes
report mean zinc levels within the normal range at baseline.
Ginawi 2013 IND and Karyadi 2002 IDN reported that 30% of the Four trials reported on deaths that occurred during the trial. In
participants had low zinc levels (less than 10.7 µmol/L) and in HIV negative participants there were no statistically significant
Lodha 2014 IND approximately 50% of the participants had zinc differences in risk of death between those who received zinc and
levels greater than 65 µg/dL. vitamin A or placebo (four trials, 430 participants, Analysis 6.1),
but in HIV-positive participants the effect did reach statistical
Overall, daily supplementation with zinc sulphate increased serum significance (RR 5.94, 95% CI 1.07 to 32.96; two trials, 136
zinc concentrations after two months and six months compared to participants; Analysis 6.1). However, both of these analyses are
placebo (four trials, 472 participants, Analysis 5.1; Table 3), with substantially underpowered to have confidence in these effects.
more consistent effects at six months.
Only Pakasi 2010 IDN reported on treatment completion and found
Tuberculosis treatment outcomes no statistically significant difference between the groups (one trial,
No effect on mortality has been seen with zinc alone or 152 participants; Analysis 6.2). Overall, there is no statistically
in combination with other micronutrients (seven trials, 2862 significant difference between supplementation and placebo in
participants, Analysis 5.2, Analysis 5.3). These trials, even the the number of participants who remain sputum-smear positive
at one month or two months (seven trials, 726 participants,
Collaboration.
Informed decisions.
Analysis 6.3). One study, Armijos 2010 MEX, did find a statistically serum vitamin D levels were in the normal or insufficient range at
significant difference in sputum positivity at three months in favour baseline, there were no statistically significant differences at eight
of supplementation (RR 0.12, 95% CI 0.02 to 0.84; one trial, 33 weeks (Wejse 2008 GNB: MD 2.00, 95% CI −7.76 to 11.76), or six
participants; Analysis 6.3), but the difference was not significant months (Daley 2015 IND: MD 8.60, 95%CI −6.29 to 23.49).
at two or four months. Visser 2011 ZAF found no statistically
significant difference in time to smear or culture conversion (one Tuberculosis treatment outcomes
trial, 154 participants; P = 0.15 and P = 0.38 respectively, authors' There were no statistically significant differences in the number
own figures). of deaths between those receiving vitamin D (any formulation) or
placebo regardless of HIV status (seven trials, 2649 participants,
Nutritional recovery and quality of life
Analysis 7.2; Analysis 7.3).
Karyadi 2002 IDN reported a statistically significant increase in
mean body weight at six months (MD 3.10 kg, 95% CI 0.74 to Only Ralph 2013 IDN reported on cure, for which there was no
5.46; one trial, 80 participants; Analysis 6.4), but there were no statistically significant difference between the vitamin D and the
differences in any other nutritional parameters (see Table 4 and placebo groups (one trial, 76 participants, Analysis 7.4). Wejse 2008
Analysis 6.5; Analysis 6.6; Analysis 6.7; Analysis 6.8; Analysis 6.9; GNB also found no statistically significant difference in recovery,
Analysis 6.10; Analysis 6.11; Analysis 6.12). There was no statistically as defined by a newly developed tuberculosis scoring system (one
significant differences between intervention and control arms in trial, 348 participants, Analysis 7.5). This system rates the patient's
any of the other three trials that reported changes in BMI or weight. condition on a scale of zero to 13, based on signs and symptoms
and anthropometric measurements (Wesje 2008).
Two trials reported on changes in Karnofsy score, a rating scale
of a person's ability to perform activities of daily living ranging Overall, there were no statistically significant differences in the
from 0 (dead) to 100 (normal). Karyadi 2002 IDN reported that proportion of people that remained sputum positive at any
supplementation resulted in a small but statistically significant time point from four weeks to eight months (seven trials, 1197
difference in Karnofsky score at six months (MD 2.5%, 95% CI participants, Analysis 7.6). One trial of daily supplementation,
0.91 to 4.09; one trial 80 participants; Analysis 6.13), while Lawson Nursyam 2006 IDN, showed a statistically significant difference in
2010 NGA found no difference at two or six months but only the proportion of participants who remained sputum positive at
presented data graphically (one trial, 233 participants). A difference six weeks (RR 0.06, 95% CI 0.00 to 0.95; one trial, 67 participants,
in Karnofsky score of 2.5% is unlikely to be of clinical significance. Analysis 7.6); but the difference was not significant two weeks
later. One additional trial, Kota 2011 IND, also reported finding a
Vitamin D (DRI for adults: 5 to 15 µg/200 to 600 IU per day) statistically significant difference but only presented the P value (P
Supplement dosing regimes = 0.067).
Eight trials evaluated vitamin D supplementation versus placebo: Nutritional recovery and quality of life
Morcos 1998 EGY: 1000 IU daily for eight weeks; Nursyam 2006 IDN: There were no statistically significant differences in mean BMI
250 µg daily for six weeks; Ralph 2013 IDN: 50,000 IU daily for eight after six to eight weeks of supplementation (four trials, 430
weeks; Mily 2015 BGD: 5000 IU daily for eight weeks; Tukvadze 2015 participants, Analysis 7.7), or in mean body weight (two trials,
GEO: 50000 IU three times a week for eight weeks, then every two 150 participants, Analysis 7.8). Wejse 2008 GNB also reported no
weeks for eight weeks; Wejse 2008 GNB: 100,000 IU at 0, 5, and 8 statistically significant difference in weight gain at eight months
months; Daley 2015 IND and Martineau 2011 GBR: 2.5 mg on days but only reported the P value (one trial, 359 participants, P = 0.9,
0, 14, 28, and 42). authors' own figures), and Ralph 2013 IDN reported no statistically
In addition, one trial evaluated vitamin D combined with arginine significant difference in the proportions gaining less than 5%, 5%
(Ralph 2013 IDN: 50,000 IU plus arginine 6 g daily for eight weeks), to 9.9%, or greater than 10% weight.
two trials evaluated vitamin D combined with calcium (Singh 2013 Daley 2015 IND reported that mean Karnofsky score increased in
IND: 250 IU plus calcium 500 mg daily for 10 days then three times a both groups, with no statistically significant difference at eight
week; Kota 2011 IND: 60,000 IU per week plus calcium 1000 mg per weeks (one trial, 212 participants, Analysis 7.9).
day for 3 months), and four trials of multi-micronutrients included
vitamin D in standard daily doses (see comparison 3). Adverse events
Vitamin D levels at baseline and follow-up (deficient ≤ 50 nmol/L, and Five trials reported adverse events, which we have summarized in
insufficient ≤ 75 nmol/L) Table 6. There were no important differences in reported adverse
events between the supplemented and the placebo groups.
Seven trials reported vitamin D status at baseline and during follow-
up, although only four trials provided data that we could include Vitamin E and selenium capsules (DRI: vitamin E 15 mg, selenium 55 µg
in a meta-analysis of the effect of Vitamin D supplementation on per day)
Vitamin D levels at follow-up (see Analysis 7.1; Table 5).
One trial compared a daily vitamin E (140 mg) and selenium (200 µg)
In four studies where the mean serum vitamin D levels were in the supplement with placebo in adults being treated for sputum-smear
deficient range at baseline, there were large improvements at eight positive pulmonary tuberculosis (Seyedrezazadeh 2006 IRN).
weeks with vitamin D compared to placebo (Martineau 2011 GBR:
The trial authors reported the median plasma vitamin E and
MD 78.60, 95% CI 54.17 to 103.03; Kota 2011 IND: MD 28.00, 95% CI
selenium levels at baseline and at eight weeks. They reported that
20.29 to 35.71; Mily 2015 BGD and Tukvadze 2015 GEO presented
the median level of both micronutrients rose in the supplement
graphically, see Table 5). In two additional studies where the mean
group and decreased in the placebo group. We were unable to
Collaboration.
Informed decisions.
assess whether these differences between groups were statistically Nutritional recovery and quality of life
significant (Appendix 8). In addition, one study that gave multi- In two studies there was no significant difference in weight gain
micronutrients, including vitamin E (133 mg) and selenium (65 µg), (Schön 2003 ETH only presented the data graphically) or the
measured the vitamin E and selenium levels at baseline and during proportion of participants with weight gain greater than 10%
follow-up (Semba 2007 MWI). The trial authors reported that both (Schön 2011 ETH; one trial, 170 participants, Analysis 8.5) between
vitamin E and selenium levels were "significantly higher" in the the arginine and placebo or low arginine groups. In a recent
supplement group after eight months, but only presented the data study, arginine supplementation significantly reduced the number
graphically. of participants with a BMI of less than 18.5 after one (P = 0.032) and
Tuberculosis treatment outcomes two (P = 0.04) months of antituberculous treatment compared to
placebo (Farazi 2015 IRN).
No deaths were reported and this trial did not report cure or
treatment completion. 3. Other trials
There was no statistically significant difference between the One small trial compared supplementation with C. striata capsules
supplement and placebo groups in the numbers of participants (6 g per day), for four months, with organoleptically-matched
who were sputum-smear positive at 15, 30, 45, and 60 days after placebo on sputum and cytokine conversion in 36 HIV-negative
the start of antituberculous treatment (one trial, 35 participants, participants with sputum smear positive pulmonary tuberculosis
Analysis 9.1). (Paliliewu 2013 IDN).
Nutritional recovery and quality of life Tuberculosis treatment outcomes
The trial authors reported a 'constant increment' in BMI for the This trial did not report on death, cure, or treatment completion.
two months of treatment with no statistically significant differences The rate of sputum smear conversion was greater in the
between the groups, but did not present these data. supplemented group compared to placebo group. However, these
differences were not statistically significant at any of the time points
Arginine (currently considered a conditionally essential amino acid, measured.
depending on the developmental stage and health status of the
individual) Nutritional recovery and quality of life
Four trials compared daily supplementation with arginine (1 mg This outcome was not reported.
daily: Schön 2003 ETH and Schön 2011 ETH; 1000 mg twice daily for
30 days: Farazi 2015 IRN; and 6 mg daily for eight weeks: Ralph 2013 DISCUSSION
IDN) with either placebo (Schön 2003 ETH; Ralph 2013 IDN; Farazi
2015 IRN) or a biscuit that contained trace amounts of arginine Summary of main results
(0.1 mg arginine, Schön 2011 ETH). The trials were all conducted
Macronutrient supplementation
in adults being treated for smear-positive pulmonary tuberculosis.
Farazi 2015 IRN only included HIV-negative participants. The The included trials were too small to reliably prove or exclude
percentage of HIV-positive participants in the other three trials clinically important benefits on mortality, cure, or treatment
ranged from 13% (Ralph 2013 IDN) to 52% (Schön 2003 ETH). completion (very low quality evidence). One small trial from India
did find a statistically significant benefit on treatment completion,
Tuberculosis treatment outcomes and clearance of the bacteria from the sputum, but these findings
The included trials reported a total of 12 deaths. There was no have not been confirmed in larger trials elsewhere.
significant difference in the risk of death between the arginine
supplemented group and the placebo group (three trials, 394 The provision of free food or high-energy nutritional products
participants, Analysis 8.1). There was also no significant difference probably does produce a modest increase in weight gain during
in the proportion of participants who were cured (two trials, 279 treatment for active tuberculosis, although this was not consistent
participants, Analysis 8.2), or sputum smear or culture positive across all included trials (moderate quality evidence). Two small
at four or eight weeks of follow-up (four trials, 464 participants, studies provide some evidence that quality of life may also be
Analysis 8.3). Schön 2003 ETH reported a statistically significant improved but the trials were too small to have much confidence in
increase in sputum-smear conversion in HIV-negative participants the result (low quality evidence).
receiving arginine; however, our analysis of the data showed Micronutrient supplementation
a non-significant difference between groups in the numbers of
participants still sputum-smear positive at eight weeks (one trial, Multi-micronutrients may have little or no effect on mortality
56 participants, Analysis 8.3). in HIV-negative people with tuberculosis (low quality evidence),
and probably have little or no effect on mortality in HIV-positive
At two weeks Schön 2003 ETH reported that compared to the patients who are not taking anti-retroviral therapy (moderate
placebo group, fewer HIV-negative participants in the arginine quality evidence). There is insufficient evidence to know if
group reported cough (RR 0.71, 95% CI 0.53 to 0.96; one trial, 56 multi-micronutrients improve cure, treatment completion, or the
participants, Analysis 8.4). Similarly, at eight weeks significantly proportion of tuberculosis patients remaining sputum positive
fewer participants in the arginine group reported cough compared during the first eight weeks (very low quality evidence).
to those in the placebo group (RR 0.78, 95% CI 0.61 to 0.99; three
trials, 348 participants, Analysis 8.4); this difference should be Multi-micronutrients may have little or no effect on weight gain
viewed with caution as only Farazi 2015 IRN provided baseline data during treatment (low quality evidence), although one of the three
that showed no difference in cough symptoms between groups. studies did find a substantial benefit in one study arm. No studies
Collaboration.
Informed decisions.
assessed the effect of multi-micronutrient supplementation on confidence in the result, but further research evidence would still
quality of life. be helpful. 'Low' and 'very low' quality reflect decreasing levels of
confidence in the result.
Individual micronutrients
The quality of evidence was mainly downgraded under 'directness'
Although low vitamin A levels are common in tuberculosis,
and 'precision'. Directness is an assessment of how well the
plasma levels probably increase following initiation of tuberculosis
evidence matches the PICO question being asked (population,
treatment regardless of supplementation. There is no evidence that
intervention, control, outcome). Because nutritional deficiencies
supplementation in doses up to three times the DRI has a beneficial
are likely to differ widely among populations, it is difficult to
effect on tuberculosis treatment outcomes, or nutritional recovery.
generalize the results of one or two trials to other settings or
B vitamins have been given in doses up to 10 times the DRI as part subgroups. Precision involves an assessment of the statistical and
of multi-micronutrient supplementation. There is no evidence of an clinical significance of the result, and also whether the sample size
effect on mortality from tuberculosis. Due to the paucity of trials of the trials is adequate to reliably detect an effect. Most of these
we cannot make any further conclusions about the effect of this trials were small and well below the optimal information size for the
vitamin on other tuberculosis outcomes. outcomes that were being measured.
Vitamin C has only been evaluated as part of multi-micronutrient Potential biases in the review process
supplements. Doses of vitamin C up to five times the DRI had We minimized biases in the review process by performing a
no effect on tuberculosis mortality. Due to the paucity of trials comprehensive search of the literature for relevant published,
we cannot make any further conclusions about the effect of this unpublished, and ongoing trials and by independently selecting
vitamin on other tuberculosis outcomes. and appraising the studies, and extracting the data in duplicate.
Where data was missing, we sought additional information and
Vitamin D supplementation appears to improve plasma levels
data directly from the study authors where this was possible to do
compared to placebo when participants are deficient at the start of
so. We did not conduct an extensive hand-search for grey literature.
supplementation, but may not have any clinically important effects
Therefore, the review is at risk of publication bias from less
on early sputum conversion.
prominent trials. We attempted to reduce this risk by identifying
Vitamin E has only been assessed in combination with other relevant conference abstracts and registered ongoing trials. The
vitamins or selenium. The dose used was up to 10 times the DRI. search of the trials registry to identify trial protocols and ongoing
None of these trials show any significant benefit or harm with trials yielded 16 potentially relevant trial protocols. These will
supplementation, but supplementation probably does improve require further assessment and exploration to either: 1) link them to
blood levels of both vitamin E and selenium. trials already included in the review; or 2) if not included, to attempt
to obtain the completed trial reports. This task is time-consuming
There is some evidence that plasma zinc levels increase during the and has to be balanced against feasibility and time constraints.
first six months of supplementation, but no convincing evidence of
other benefits. Agreements and disagreements with other studies or
reviews
Overall completeness and applicability of evidence
Nutritional supplementation in HIV-positive patients without
The included studies were generally too small and too limited to tuberculosis has been assessed by two further Cochrane reviews
make broad conclusions on the presence or absence of clinically (Irlam 2010; Grobler 2013).
important benefits of nutritional supplementation on tuberculosis
treatment outcomes. Grobler 2013 found 14 small trials that compared macronutrient
supplementation with no supplementation and concluded that
Although the included studies are from low- and middle-income although there was evidence that supplementation increased both
countries, they may not reflect the food-insecure settings where energy and protein intake, there was no evidence of a clinical
most supplementation programmes take place, and where the benefit on either nutritional- or HIV-related outcomes.
benefit may plausibly be greatest.
Irlam 2010 assessed 30 trials that compared micronutrients with
Where a supplement has so far not shown any benefit, this may placebo from both developed and developing countries. The
also be an issue relating to the dose used as people recovering authors concluded that multi-micronutrient supplements have
from tuberculosis may have micronutrient requirements which are significant benefits in HIV-positive pregnant women and children.
higher than healthy people, however there is currently no evidence In children there was also evidence of a reduction in mortality with
to support this. vitamin A, and a reduction in diarrhoeal morbidity with zinc.
In addition, it should be noted that most of the HIV-positive AUTHORS' CONCLUSIONS

participants in these trials were not taking antiretroviral therapy.
Implications for practice
Quality of the evidence
There is insufficient research to know whether routinely providing
We assessed the quality of evidence using the GRADE methodology, free food or energy supplements results in better tuberculosis
and displayed the results in two 'Summary of findings' tables treatment outcomes, but limited evidence suggests it probably
(Summary of findings for the main comparison; Summary of improves weight gain.
findings 2). 'Moderate' quality evidence implies we can have some
Collaboration.
Informed decisions.
Although blood levels of some vitamins may be low in is also essential that measures of how this translates into improved
patients starting treatment for active tuberculosis, there is quality of life and physical functioning are made, as weight gain on
currently no reliable evidence that routinely supplementing above its own is of little interest.
recommended daily amounts has clinical benefits.
ACKNOWLEDGEMENTS
Implications for research
We are grateful to Professor Jimmy Volmink and Katharine Abba,
High quality studies of food provision to tuberculosis patients the authors of previous versions of this Cochrane review, who have
in food-insecure settings are urgently needed to support now stepped down from the author team for this review update.
the continued expenditure on food support to tuberculosis
programmes. In designing these studies, researchers, and This document is an output from a project funded by the
programme managers need to be clear about the aim of food UK Department for International Development (DFID) (Grant:
provision. Whilst an effect on mortality would provide strong 5242) for the benefit of low- and middle-income countries. The
advocacy for continued financial support, if the primary aim is views expressed are not necessarily those of DFID. Stellenbosch
to promote adherence, or to mitigate the catastrophic financial University, Stellenbosch, South Africa, and Christian Medical
consequences of the illness, then these are the outcomes that College, Vellore, India provided additional support for the review
should be measured, and appropriate comparison interventions authors.
should be selected. If the primary aim is to reduce mortality, then
future trials must be large enough to reliably detect or exclude an David Sinclair is supported by the Effective Health Care Research
effect. Consortium. This Consortium and the editorial base of the
Cochrane Infectious Diseases Group is funded by UK aid from the
The failure to demonstrate a beneficial effect with micronutrient UK Government for the benefit of developing countries (Grant:
supplementation does not imply that one does not exist. Further 5242). The views expressed in this publication do not necessarily
studies, perhaps using higher doses, would still be beneficial but reflect UK government policy.
should have adequate sample sizes to reliably detect or exclude
clinically important benefits. It would also be useful if some The Contact Editor for this review update was Dr Mical Paul.
standardization of outcome measurements could be made.
We thank the following authors of the included studies for
For nutritional recovery it seems important to assess changes providing us with vital additional (unpublished) information
in weight and lean body mass during the first two months of regarding their respective studies: Saurabh Mehta, Anna Ralph,
treatment rather than as a single measure at the end of treatment. It Sushil Kabra, Kidola Jeremiah, and Peter Daley.
Collaboration.
Informed decisions.
REFERENCES
References to studies included in this review tuberculosis. Tropical Medicine and International Health
2010;15(12):1481-90.
Armijos 2010 MEX {published data only}
Armijos RX, Weigel MM, Chacon R, Flores L, Campos A. Lodha 2014 IND {published data only}
Adjunctive micronutrient supplementation for pulmonary Lodha R, Mukherjee A, Singh V, Singh S, Friis H, Faurholt-
tuberculosis. Salud Pública de México 2010;52(3):185-9. Jepsen D, et al. Effect of micronutrient supplementation on
treatment outcomes in children with intrathoracic tuberculosis:
Daley 2015 IND {published data only}
a randomized controlled trial. American Journal of Clinical
Daley P, Jagannathan V, John KR, Sarojini J, Latha A, Vieth R, et Nutrition 2014;100(5):1287-97.
al. Adjunctive vitamin D for treatment of active tuberculosis in
India: a randomised, double-blind, placebo-controlled trial. The Martineau 2011 GBR {published data only}
Lancet. Infectious Diseases 2015;15(5):528-34. Martineau AR, Timms PM, Bothamley GH, Hanifa Y, Islam K,
Claxton AP, et al. High-dose vitamin D3 during intensive-
Farazi 2015 IRN {published data only}
phase antimicrobial treatment for pulmonary tuberculosis:
Farazi A, Shafaat O, Sofian M, Kahbazi M. Arginine adjunctive a double-blind randomised controlled trial. Lancet
therapy in active tuberculosis. Tuberculosis Research and 2011;377(9761):242-50.
Treatment 2015;2015:205016. [DOI: 10.1155/2015/205016]
Martins 2009 TLS {published data only}
Ginawi 2013 IND {published data only}
Martins N, Morris P, Kelly PM. Food incentives to improve
Ginawi IA, Ahmed MQ, Ahmad I, Al-Hazimi AM. Effect of zinc completion of tuberculosis treatment: randomised controlled
and vitamin a supplementation along with inter-tubercular trial in Dili, Timor-Leste. BMJ 2009;229:b4248.
treatment in pulmonary tuberculosis in north Indian patients.
International Journal of Pharmaceutical Sciences and Research Mehta 2011 TZA {published and unpublished data}
2013;4(9):3426-31. * Mehta S, Mugusi FM, Bosch RJ, Aboud S, Chatterjee A,
Finkelstein JL, et al. A randomized trial of multivitamin
Hanekom 1997 ZAF {published data only}
supplementation in children with tuberculosis in Tanzania.
Hanekom WA, Potgeiter S, Hughes EJ, Malan H, Nutrition Journal 2011;10:120.
Kessow G, Hussey GD. Vitamin A status and therapy in
childhood pulmonary tuberculosis. Journal of Pediatrics NCT00145184. Effect of Multivitamin Supplements on Clinical
1997;131(6):925-7. and Immunological Response in Childhood Tuberculosis.
https://clinicaltrials.gov/ct2/show/NCT00145184 (accessed 4
Jahnavi 2010 IND {published data only} February 2016).
Jahnavi G, Sudha CH. Randomised controlled trial of food
supplements in patients with newly diagnosed tuberculosis and Mily 2015 BGD {published data only}
wasting. Singapore Medical Journal 2010;51(12):957-62. * Mily A, Rekha RS, Kamal SM, Arifuzzaman AS, Rahim Z, Khan L,
et al. Significant effects of oral phenylbutyrate and vitamin D3
Jeremiah 2014 TZA {published data only} adjunctive therapy in pulmonary tuberculosis: a randomized
Jeremiah K, Denti P, Chigutsa E, Faurholt-Jepsen D, PrayGod G, controlled trial. PLoS One 2015;10(9):e0138340. [DOI: 10.1371/
Range N, et al. Nutritional supplementation increases rifampin journal.pone.0138340]
exposure among tuberculosis patients co-infected with HIV.
Antimicrobial Agents and Chemotherapy 2014;58(6):3468-74. Morcos 1998 EGY {published data only}
Morcos MM, Gabr AA, Samual S, Kamel M, el Baz M, el Beshry M,
Karyadi 2002 IDN {published data only} et al. Vitamin D administration to tuberculous children and its
Karyadi E, West CE, Schultink W, Nelwan RH, Gross R, Amin Z, value. Bollettino Chimico Farmaceutico 1998;137(5):157-64.
et al. A double-blind, placebo-controlled study of vitamin A and
zinc supplementation in persons with tuberculosis in Indonesia: Nursyam 2006 IDN {published data only}
effects on clinical response and nutritional status. American Nursyam EW, Amin Z, Rumende CA. The effects of vitamin
Journal of Clinical Nutrition 2002;75(4):720-7. D as supplementary treatment in patients with moderately
advanced pulmonary tuberculosis lesion. Acta Medica
Kota 2011 IND {published data only} Indonesiana 2006;38(1):3-5.
Kota SK, Jammula S, Kota SK, Tripathy PR, Panda S, Modi KD.
Effect of vitamin D supplementation in type 2 diabetes patients Pakasi 2010 IDN {published data only}
with pulmonary tuberculosis. Diabetes and Metabolic Syndrome: Pakasi TA, Karyadi E, Suratih NMD, Salean M, Darmawidjaja N,
Clinical Research & Reviews 2011;5(2):85-9. Bor H, et al. Zinc and vitamin A supplementation fails to reduce
sputum conversion time in severely malnourished pulmonary
Lawson 2010 NGA {published data only} tuberculosis patients in Indonesia. Nutrition Journal 2010;9:41.
Lawson L, Thacher TD, Yassin MA, Onuoha NA, Usman A,
Emenyonu NE, et al. Randomized controlled trial of zinc and
vitamin A as co-adjuvants for the treatment of pulmonary
Collaboration.
Informed decisions.
Paliliewu 2013 IDN {published data only} Schön 2003 ETH {published data only}
Paliliewu N, Datau EA, Matheos JC, Surachmanto EE. Channa Schön T, Elias D, Moges F, Melese E, Tessema T, Stendahl O, et
striatus capsules induces cytokine conversion in pulmonary al. Arginine as an adjuvant to chemotherapy improves clinical
tuberculosis patients. Journal of Experimental and Integrative outcomes in active tuberculosis. European Respiratory Journal
Medicine 2013;3(3):237-42. 2003;21(3):483-8.
Paton 2004 SGP {published data only} Schön 2011 ETH {published data only}
Paton NI, Chua YK, Earnest A, Chee CB. Randomized controlled Schön T, Idh J, Westman A, Elias D, Abate E, Diro E, et al.
trial of nutritional supplementation in patients with newly Effects of a food supplement rich in arginine in patients with
diagnosed tuberculosis and wasting. American Journal of smear positive pulmonary tuberculosis - a randomised trial.
Clinical Nutrition 2004;80(2):460-5. Tuberculosis 2011;91(5):370-7.
Pérez-Guzmán 2005 MEX {published data only} Semba 2007 MWI {published data only}
Pérez-Guzmán C, Vargas MH, Quiñonez F, Bazavilvazo N, Semba RD, Kumwenda J, Zijlstra E, Ricks MO, van Lettow M,
Aguilar A. A cholesterol-rich diet accelerates bacteriologic Whalen C, et al. Micronutrient supplements and mortality of
sterilization in pulmonary tuberculosis. Chest HIV-infected adults with pulmonary TB: a controlled clinical
2005;127(2):643-51. trial. International Journal of Tuberculosis and Lung Disease
2007;11(8):854-9.
Praygod 2011a TZA {published data only}
Praygod G, Range N, Faurholt-Jepsen D, Jeremiah K, Faurholt- Seyedrezazadeh 2006 IRN {published data only}
Jepsen M, Aabye MG, et al. Daily multi-micronutrient Seyedrezazadeh E, Ostadrahimi A, Mahboob S, Asadi Y,
supplementation during tuberculosis treatment increases Ghaemmagami J, Pourmogaddam M. Effect of vitamin E
weight and grip strength among HIV-uninfected but not HIV- and selenium supplementation on oxidative stress status in
infected patients in Mwanza, Tanzania. The Journal of Nutrition pulmonary tuberculosis patients. Respirology 2008;13(2):294-8.
2011;141(4):685-91.
* Seyedrezazadeh E, Ostradrahimi AR, Mahboob SA, Assadi Y,
Praygod 2011b TZA {published data only} Ansarin K, Sakoori P, et al. Vitamin E-selenium supplement and
* Praygod G, Range N, Faurholt-Jepsen D, Jeremiah K, clinical responses of active pulmonary tuberculosis. Tanaffos
Faurholt-Jepsen M, Aabye MG, et al. Daily multi-micronutrient 2006;5(2):49-55.
supplementation during tuberculosis treatment increases
Singh 2013 IND {published data only}
weight and grip strength among HIV-uninfected but not HIV-
infected patients in Mwanza, Tanzania. The Journal of Nutrition Singh AKr, Gogoi JB, Pant NC, Mittal P, Juyal V, Mukherjee S. A
2011;141:685-91. study on the role of vitamins and minerals supplementation in
the treatment of tuberculosis. Indian Journal of Public Health
Praygod G, Range N, Faurholt-Jepsen D, Jeremiah K, Faurholt- Research and Development 2013;4(2):26-30.
Jepsen M, Aabye MG, et al. The effect of energy-protein
supplementation on weight, body composition and handgrip Sudarsanam 2010 IND {published data only}
strength among pulmonary tuberculosis HIV-co-infected Sudarsanam TD, John J, Kang G, Mahendri V, Gerrior J,
patients: randomised controlled trial in Mwanza, Tanzania. The Franciosa M, et al. Pilot randomized trial of nutritional
British Journal of Nutrition 2012;107(2):263-71. supplementation in patients with tuberculosis and HIV-
tuberculosis coinfection receiving directly observed short-
Ralph 2013 IDN {published data only} course chemotherapy for tuberculosis. Tropical Medicine and
Ralph AP, Waramori G, Pontororing GJ, Kenangalem E, International Health 2011;16(6):699-706.
Wiguna A, Tjitra E, et al. L-arginine and vitamin D adjunctive
therapies in pulmonary tuberculosis: a randomised, double- Tukvadze 2015 GEO {published data only}
blind, placebo-controlled trial. PLoS One 2013;8(8):e70032. Frediani J, Tukvadze N, Sanikidze E, Kipiani M, Hebbar G,
Tangpricha V, et al. Drug susceptibility influences macronutrient
Range 2005 TZA {published data only} intake and body composition in tuberculosis patients. The
* Range N, Andersen AB, Magnussen P, Mugomela A, Friis H. The FASEB Journal 2014;28(Supplement 1):1014.4.
effect of micronutrient supplementation on treatment outcome
in patients with pulmonary tuberculosis: a randomised * Tukvadze N, Sanikidze E, Kipiani M, Hebbar G, Easley KA,
controlled trial in Mwanza, Tanzania. Tropical Medicine and Shenvi N, et al. High-dose vitamin D3 in adults with pulmonary
International Health 2005;10(9):826-32. tuberculosis: a double-blind randomised controlled trial.
American Journal of Clinical Nutrition 2015;102(5):1059-69.
Range N, Changalucha J, Krarup H, Magnussen P, Andersen AB,
Friis H. The effect of multi-vitamin/mineral supplementation Villamor 2008 TZA {published and unpublished data}
on mortality during treatment of pulmonary tuberculosis: a Villamor E, Mugusi F, Urassa W, Bosch R J, Saathoff E,
randomised two-by-two factorial trial in Mwanza, Tanzania. Matsumoto K, et al. A trial of the effect of micronutrient
British Journal of Nutrition 2006;95(4):762-70. supplementation on treatment outcome, T cell counts,
morbidity, and mortality in adults with pulmonary tuberculosis.
The Journal of Infectious Diseases 2008;197(11):1499-505.
Collaboration.
Informed decisions.
Visser 2011 ZAF {published data only} Africa: a pragmatic cluster-randomized controlled trial. Trials
Visser ME, Grewal HMS, Swart EC, Dhansay MA, Walzl G, 2013;14:154.
Swanevelder S, et al. The effect of vitamin A and zinc
Martineau 2009 {published data only}
supplementation on treatment outcomes in pulmonary
tuberculosis: a randomized controlled trial. American Journal of Martineau AR, Nanzer AM, Satkunam KR, Packe GE, Rainbow SJ,
Clinical Nutrition 2011;93(1):93-100. Maunsell ZJ, et al. Influence of a single oral dose of vitamin D(2)
on serum 25-hydroxyvitamin D concentrations in tuberculosis
Wejse 2008 GNB {published and unpublished data} patients. The International Journal of Tuberculosis and Lung
Wejse C, Gomes VF, Rabna P, Gustafson P, Aaby P, Lisse IM, et Disease 2009;13(1):119-25.
al. Vitamin D as supplementary treatment for tuberculosis
Mbala 1998 {published data only}
- a double-blind randomised placebo-controlled trial.
American Journal of Respiratory and Critical Care Medicine Mbala L, Matendo R, Nkailo R. Is vitamin B6 supplementation
2009;179(9):843-50. of isoniazid therapy useful in childhood tuberculosis. Tropical
Doctor 1998;28(2):103-4.
References to studies excluded from this review Narang 1984 {published data only}
Narang NK, Gupta RC, Jain MK. Role of vitamin D in pulmonary
Denti 2015 {published data only}
tuberculosis. Journal of the Association of Physicians of India
Denti P, Jeremiah K, Chigutsa E, Faurholt-Jepsen D, PrayGod G, 1984;32(2):185-8.
Range N, et al. Pharmacokinetics of isoniazid, pyrazinamide,
and ethambutol in newly diagnosed pulmonary TB patients in Oluboyede 1978 {published data only}
Tanzania. PLoS One 2015;10(10):e0141002. Oluboyede OA, Onadeko BO. Observation on haematological
patterns in pulmonary tuberculosis in Nigerians. Journal of
Dibari 2013 {published data only}
Tropical Medicine and Hygiene 1978;81(5):91-5.
Dibari F, Bahwere P, Huerga H, Irena A H, Owino V, Collins S, et
al. Development of a cross-over randomized trial method to Permatasari 2014 {published data only}
determine the acceptability and safety of novel ready-to-use Permatasari A, Hasan H. The effect of ethanol extract propolis
therapeutic foods. Nutrition 2013;29(1):107-12. (EEP) on the level of IFN-F and superoxide dismutase (SOD)
activities in patients with MDR tuberculosis. Respirology 2014;
Gwinup 1981 {published data only}
Vol. 19, issue Suppl. 3:1-62 (Abstract 0-Q-012).
Gwinup G, Randazzo G, Elias A. The influence of vitamin
D on serum calcium in tuberculosis. Acta Endocrinologica Ramakrishnan 1961 {published data only}
1981;97(1):114-7. Ramakrishnan CV, Rajendran K, Jacob PG, Fox W,
Radhakrishnan S. The role of diet in the treatment of pulmonary
Hasan 2015 {published data only}
tuberculosis. An evaluation in a controlled chemotherapy study
Hasan H, Rini YP. The effect of propolis on the level of in home and sanatorium patients in South India. Bulletin of the
interleukin-10 in multidrug resistance tuberculosis patient. World Health Organization 1961;25:339-59.
Respirology 2015; Vol. 20, issue Suppl. 3:1-160 (Abstract ID 869).
Samsidi 2013 {published data only}
Kawai 2014 {published data only}
Samsidi S, Ridwan H, Kusharto C M, Sulaeman A, Alisjahbana B.
Kawai K, Meydani SN, Urassa W, Wu D, Mugusi FM, Saathoff E, Efficacious of synbiotic and nutrients supplement on stimulated
et al. Micronutrient supplementation and T cell-mediated of secretory immunoglobulin a (SIGA) in treated pulmonary
immune responses in patients with tuberculosis in Tanzania. tuberculosis patients. Annals of Nutrition and Metabolism
Epidemiology and Infection 2014;142(7):1505-9. 2013;20th International Congress of Nutrition; Sept 15-20
2013; Granada, Spain:407.
Khandelwal 2014 {published data only}
Khandelwal D, Gupta N, Mukherjee A, Lodha R, Singh V, Shi 2001 {published data only}
Grewal HM, et al. Vitamin D levels in Indian children with Shi XY, Xiao JQ, Yi JY, Zhu YQ, Deng CS. Influence of partial
intrathoracic tuberculosis. Indian Journal of Medical Research parenteral nutrition with fat emulsion on nutritional status
2014;140(4):531-7. in patients with abdominal tuberculosis. Journal of Clinical
Internal Medicine 2001;18(2):120-1.
Lin 2014 {published data only}
Lin HC, Yu MC, Liu HJ, Bai KJ. Impact of food intake on the Srivastava 2011 {published data only}
pharmacokinetics of first-line antituberculosis drugs in Srivastava GN, Chaudhry S, Kumar H, Singh SK. Vitamin-
Taiwanese tuberculosis patients. Journal of the Formosan D supplementation in patients with new smear positive
Medical Association 2014;113(5):291-7. pulmonary tuberculosis (PTB) with reference to sputum
conversion [Abstract]. European Respiratory Journal
Lutge 2013 {published data only}
2011;38(Supplement 55):2588.
Lutge E, Lewin S, Volmink J, Friedman I, Lombard C. Economic
support to improve tuberculosis treatment outcomes in South
Collaboration.
Informed decisions.
References to studies awaiting assessment in Qom. http://www.irct.ir/searchresult.php?

keyword=&id=8429&number=1&prt=2840&total=10&m=1
Al Mamun 2014 {published data only}
(accessed 28 August 2015).
Al Mamun SMA. Effects and rationale of micronutrients
supplementation in the treatment of smear positive pulmonary IRCT201211179855N2 {unpublished data only}
tuberculosis patients. Chest Journal 2014; Vol. 145, issue IRCT201211179855N2. Randomized, double-blind, placebo-
3_Meeting Abstracts. controlled trial of L-Arginine supplementation for the treatment
of pulmonary tuberculosis. http://www.irct.ir/searchresult.php?
Chandra 2004 {published data only}
id=9855&number=2 (accessed 28 August 2015).
Chandra RK. Nutrient supplementation as adjunct therapy in
pulmonary tuberculosis. International Journal for Vitamin and ISRCTN16469166 {unpublished data only}
Nutritional Research 2004;74(2):144-6. ISRCTN16469166. Nutrition and wasting in tuberculosis (TB):
Can nutritional supplementation in TB patients improve body
Guzman-Rivero 2013 {published data only}
weight gain, body composition and treatment outcome?.
Guzman-Rivero M, Sejas E, Medina M, Verduguez-Orellana A, www.controlled-trials.com/ISRCTN16469166/ISRCTN16469166
Akesson B. Changes in clinical status, nutritional and (accessed 10 October 2007).
biochemical biomarkers in pulmonary tuberculosis patients
receiving zinc supplementation in addition to conventional NCT00507000 {unpublished data only}
treatment (Abstract). Annals of Nutrition and Metabolism NCT00507000. Role of oral vitamin D as an adjunct therapy
2013;63(Supplement 1):1104. in category I pulmonary tuberculosis along with assessment
of immunological parameters. clinicaltrials.gov/ct2/show/
Nagrale 2013 {published data only}
NCT00507000 (accessed 10 October 2007).
Nagrale D, Mahakalkar S, Gaur S. Supplementation of N-
acetylcysteine as an adjuvant in treatment of newly diagnosed NCT00698386 {unpublished data only}
pulmonary tuberculosis patients: A prospective, randomized NCT00698386. Efficacy of oral zinc administration as an adjunct
double blind, placebo controlled study (Abstract). European therapy in new pulmonary tuberculosis (Category I) patients.
Respiratory Journal 2013;42(Suppl 57):P2833. https://clinicaltrials.gov/ct2/show/NCT00698386 (accessed 21
July 2011).
Nawas 2013 {published data only}
Nawas MA, Burhan E, Soepandi P, Isbaniah F, Agustin H. Double NCT00788320 {unpublished data only}
blind randomized placebo-controlled trial in tuberculosis NCT00788320. Antimicrobial peptide LL-37 (cathelicidin)
patients with tuberculosis treatment supplemented with production in active tuberculosis disease: role of vitamin
morrinda-zinger extract: The speed of sputum conversion D supplementation. https://clinicaltrials.gov/ct2/show/
(Abstract). Respirology 2013;18(Supplement 4):9. NCT00788320 (accessed 21 July 2011).
NCT01635153 {published data only}

References to ongoing studies NCT01635153. Effects of a protein calorie supplement in
ChiCTR-IPR-15006395 {unpublished data only} HIV-infected women with tuberculosis (DarDar). https://
ChiCTR-IPR-15006395. The influence and mechanism of clinicaltrials.gov/ct2/show/NCT01635153 (accessed 31 August
vitamin D3 supplementation on the treatment outcomes of 2015).
tuberculosis patients of different glucose tolerance. http://
www.chictr.org.cn/showproj.aspx?proj=10964 (accessed 4
February 2016). NCT01657656. Vitamin D supplementations as adjunct to anti-
tuberculosis drugs in Mongolia. https://clinicaltrials.gov/ct2/
ChiCTR-TRC-12002546 {published and unpublished data} show/record/NCT01657656 (accessed 31 August 2015).
Wang Q, Ma A, Bygbjerg IC, Han X, Liu Y, Zhao S, et al. Rationale
and design of a randomized controlled trial of the effect of
retinol and vitamin D supplementation on treatment in active NCT01722396. Pharmacogenetics of vitamin D supplementation
pulmonary tuberculosis patients with diabetes. BMC Infectious in tuberculosis. https://clinicaltrials.gov/ct2/show/
Diseases 2013;13:104. [DOI: 10.1186/1471-2334-13-104] NCT01722396 (accessed 31 August 2015).
ChiCTR-TRC-14005241 {unpublished data only} NCT01992263 {unpublished data only}

ChiCTR-TRC-14005241. A prospective study of oral nutritional NCT01992263. A trial of vitamin D supplementation among
supplement in perioperative application with pulmonary tuberculosis patients in South India. https://clinicaltrials.gov/
tuberculosis patients. http://apps.who.int/trialsearch/ ct2/show/NCT01992263 (accessed 31 August 2015).
Trial2.aspx?trialid=ChiCTR-TRC-14005241 (accessed 28 August
NCT02169570 {unpublished data only}
2015).
NCT02169570. Effect of supplementary vitamin D in patients
IRCT201112178429N1 {unpublished data only} with diabetes mellitus and pulmonary tuberculosis (EVIDENT).
IRCT201112178429N1. Effect of zinc supplementation https://clinicaltrials.gov/ct2/show/NCT02169570 (accessed 31
in improving pulmonary tuberculosis patients August 2015).
Collaboration.
Informed decisions.
NCT02464683 {unpublished data only} Harries 2006

NCT02464683. Effect of vitamin D as adjunctive therapy in Harries AD, Dye C. Tuberculosis. Annals of Tropical Medicine and
patients with pulmonary evolution tuberculosis (Vitamin D). Parasitology 2006;100(5-6):415-31.
https://clinicaltrials.gov/ct2/show/NCT02464683 (accessed 4
February 2016). Irlam 2010
Irlam JH, Visser MME, Rollins NN, Siegfried N. Micronutrient
NCT02554318 {unpublished data only} supplementation in children and adults with HIV infection.
NCT02554318. The effect of fermented soybean Cochrane Database of Systematic Reviews 2010, Issue 12. [DOI:
supplementation on the body weight and physical function 10.1002/14651858.CD003650.pub3]
of tuberculosis patients with standard therapy in Indonesia.
https://clinicaltrials.gov/show/NCT02554318 (accessed 4 Kamolratanakul 1999
February 2016). Kamolratanakul P, Sawert H, Kongsin S, Lertmaharit S,
Sriwongsa J, Na-Songkhla S, et al. Economic impact of
tuberculosis at the household level. International Journal of
Additional references Tuberculosis and Lung Disease 1999;3(7):596-602.
Aaron 2004
Lefebvre 2011
Aaron L, Saadoun D, Calatroni I, Launay O, Mémain N, Vincent V,
Carol Lefebvre, Eric Manheimer, Julie Glanville, on behalf
et al. Tuberculosis in HIV-infected patients: a comprehensive
of the Cochrane Information Retrieval Methods Group.
review. Clinical Microbiology and Infection 2004;10(5):388-98.
Chapter 6: Searching for studies. In: Higgins J, Green S
Arthur 2003 (editors). Cochrane Handbook for Systematic Reviews of
Interventions Version 5.1.0 [updated March 2011]. The
Arthur JR, McKenziey RC, Beckett GJ. Selenium in the Immune
Cochrane Collaboration, 2011. http://handbook.cochrane.org/
System. The Journal of Nutrition 2003;133(5 Suppl 1):1457S-9S.
chapter_6/6_4_11_1_the_cochrane_highly_sensitive_search_strategies_for
Barry 2009
Lounis 2001
Barry CE 3rd, Boshoff HI, Dartois V, Dick T, Ehrt S, Flynn J, et
Lounis N, Truffot-Pernot C, Grosset J, Gordeuk VR, Boelaert JR.
al. The spectrum of latent tuberculosis: rethinking the biology
Iron and Mycobacterium tuberculosis infection. Journal of
and intervention strategies. Nature Reviews. Microbiology
Clinical Virology 2001;20(3):123-6.
2009;7(12):845-55.
Louw 1992
Cegielski 2004
Louw JA, Werbeck A, Louw MEJ, Kotze TJvW, Cooper R,
Cegielski JP, McMurray DN. The relationship between
Labadarios D. Blood vitamin concentrations during the acute
malnutrition and tuberculosis, evidence from studies in humans
phase response. Critical Care Medicine 1992;20(7):934-41.
and experimental animals. International Journal of Tuberculosis
and Lung Disease 2004;8(3):286-98. Lutge 2009
Chandra 1996 Lutge EE, Knight SE, Volmink J. Incentives for improving
patient adherence to anti-tuberculosis treatment. Cochrane
Chandra RK. Nutrition, immunity and infection: from basic
Database of Systematic Reviews 2009, Issue 3. [DOI:
knowledge of dietary manipulation of immune responses to
10.1002/14651858.CD007952]
practical application of ameliorating suffering and improving
survival. Proceedings of the National Academy of Sciences of the Macallan 1999
United States of America 1996;93(25):14304-7.
Macallan DC. Malnutrition in tuberculosis. Diagnostic
Corbett 2003 Microbiology and Infectious Disease 1999;34(2):153-7.
Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, Meyers 2006
Raviglione MC, et al. The growing burden of tuberculosis: global
National Academy of Sciences. Dietary Reference Intakes:
trends and interactions with the HIV epidemic. Archives of
The Essential Guide to Nutrient Requirements. 6th Edition.
Internal Medicine 2003;163(9):1009-21.
Washington DC: National Academies Press, 2006.
Davies 1985
nMaster 1.0 [Computer program]
Davies PD, Brown RC, Woodhead JS. Serum concentrations
Department of Biostatistics, Christian Medical College. nMaster
of vitamin D metabolites in untreated tuberculosis. Thorax
1.0. Version 1.0. Vellore: Department of Biostatistics, Christian
1985;40(3):187-90.
Medical College, (accessed 9 May 2016).
Grobler 2013
Nnoaham 2008
Grobler L, Siegfried N, Visser ME, Mahlungulu SS, Volmink J.
Nnoaham KE, Clarke A. Low serum vitamin D levels and
Nutritional interventions for reducing morbidity and mortality
tuberculosis: a systematic review and meta-analysis.
in people with HIV. Cochrane Database of Systematic Reviews
International Journal of Epidemiology 2008;37(1):113-9.
2013, Issue 2. [DOI: 10.1002/14651858.CD004536.pub3]
Collaboration.
Informed decisions.
Plit 1998 WHO 2009

Plit Ml, Theron AJ, Fickl H, van Rensburg CE, Pendel S, World Health Organization. Global Tuberculosis Control. A short
Anderson R. Influence of antimicrobial chemotherapy and update to the 2009 report. Geneva: World Health Organization,
smoking status on plasma concentrations of vitamin C, 2009.
vitamin E, beta-carotene, acute phase reactants, iron and
lipid peroxides in patients with pulmonary tuberculosis. WHO 2010
International Journal of Tuberculosis and Lung Disease World Health Organization. Treatment of Tuberculosis
1998;2(7):590-9. Guidelines. 4th Edition. Geneva: World Health Organization,
2010.
RevMan 2014 [Computer program]
The Nordic Cochrane Centre, The Cochrane Collaboration. WHO 2014
Review Manager (RevMan). Version 5.3. Copenhagen: The World Health Organization. Global Tuberculosis Report 2014.
Nordic Cochrane Centre, The Cochrane Collaboration, 2014. Geneva: World Health Organization, 2014.
Saunders 2007 Wintergerst 2007

Saunders BM, Britton WJ. Life and death in the granuloma: Wintergerst ES, Maggini S, Hornig DH. Contribution of Selected
immunopathology of tuberculosis. Immunology and Cell Biology Vitamins andTrace Elements to Immune Function. Annals of
2007;85(2):103-11. Nutrition & Metabolism 2007;51(4):301-23.
Schluger 1998 Wyss 2001

Schluger NW, Rom WN. The host immune response to Wyss K, Kilima P, Lorenz N. Costs of tuberculosis for households
tuberculosis. American Journal of Respiratory Critical Care and health care providers in Dar es Salaam, Tanzania. Tropical
Medicine 1998;157(3 Pt 1):679-91. Medicine & International Health 2001;6(1):60-8.
Semba 1998 Zachariah 2002

Semba RD. The role of vitamin A and related retinoids in Zachariah R, Spielmann MP, Harries AD, Salaniponi FM.
immune function. Nutrition Reviews 1998;56(1 Pt 2):S38-48. Moderate to severe malnutrition in patients with tuberculosis
is a risk factor associated with early death. Transactions of the
Shankar 1998 Royal Society of Tropical Medicine and Hygiene 2002;96(3):291-4.
Shankar AH, Prasad AS. Zinc and immune function: the
biological basis of altered resistance to infection. American Zumla 2000
Journal of Clinical Nutrition 1998;68(2 Suppl):447S-63S. Zumla A, Malon P, Henderson J, Grange JM. Impact of HIV
infection on tuberculosis. Postgraduate Medical Journal
Stephensen 2001 2000;76(895):259-68.
Stephensen CB. Vitamin A, infection, and immune function.
Annual Review of Nutrition 2001;21:167-92.
References to other published versions of this review
Taneja 1990
Abba 2006
Taneja DP. Observations of serum zinc in patients with
Abba K, Sudarsanam TD, Grobler L, Volmink JA. Nutritional
pulmonary tuberculosis. Journal of the Indian Medical
supplements for people being treated for active tuberculosis.
Association 1990;88(10):280-1.
Cochrane Database of Systematic Reviews 2006, Issue 3. [DOI:
van Lettow 2003 10.1002/14651858.CD006086]
van Lettow M, Fawzi WW, Semba RD. Triple trouble: the role of Abba 2008
malnutrition in tuberculosis and human immunodeficiency
Abba K, Sudarsanam TD, Grobler L, Volmink J. Nutritional
virus coinfection. Nutrition Reviews 2003;61(3):81-90.
Wesje 2008 Cochrane Database of Systematic Reviews 2008, Issue 4. [DOI:
10.1002/14651858.CD006086.pub2]
Wesje C, Gustafson P, Nielson J, Gomes VF, Aaby P, Anderson PL,
et al. TB score: Signs and symptoms from tuberculosis patients Sinclair 2011
in a low-resource setting have predictive value and may be used
Sinclair D, Abba K, Grobler L, Sudarsanam TD. Nutritional
to assess clinical course. Scandinavian Journal of Infectious
Diseases 2008;40(2):111-20.
Cochrane Database of Systematic Reviews 2011, Issue 11. [DOI:
WFP 2007 10.1002/14651858.CD006086.pub3]
Ba M, Strauss A. Getting started: WFP food assistance in the
context of tuberculosis care and treatment. World Food * Indicates the major publication for the study
Programme, 2007.
Collaboration.
Informed decisions.
CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]
Armijos 2010 MEX

Methods Study design: randomized controlled trial (RCT).
Study dates and duration: recruited August 2005 to July 2006, follow-up 6 months.
Standard care: all participants received directly observed antituberculous therapy, per Instituto Mex-
icano del Seguro Social (IMSS) guidelines: 2 months of isoniazid/rifampin/pyrazinamide/ethambutol
followed by 4 months of isoniazid and rifampin.
Participants Number: 39 enrolled; 33 analysed.
Inclusion criteria: adults aged 18 to 65 years, smear positive pulmonary tuberculosis, informed consent.
Exclusion criteria: pregnancy, breast feeding, used corticosteroids, human immunodeficiency virus
(HIV), diabetes, or another serious co-morbidity.
Baseline characteristics
• Nutritional status: mean body mass index (BMI) (standard deviation (SD)): 20.4 kg/m2 (5.0) treatment
group versus 22.6 kg/m2 (4.2) control group.
• HIV status: all negative.
• Multidrug-resistant/extensively drug-resistant tuberculosis (MDR/XDR-TB): not described.
• Vitamin A status µg/dL (SD): 29.4 (13.3) supplement group versus 29.8 (13.8) placebo group; normal
range.
• Zinc status µg/L (SD): 738 (168) supplement group versus 764 (137) placebo group; normal range.
Interventions Group 1: vitamin A 5000 IU/day (as retinyl acetate) plus zinc 50 mg/day (as zinc chelate) for 4 months.
Group 2: placebo with identical appearance.
Outcomes • Deaths during study.

• Percentage remaining smear positive at 1, 2, 3, 4, and 5 months.
• Mean zinc, retinol, and albumin levels at 0, 2, and 6 months.
Not included in review: mean monthly intake of zinc and vitamin A, measures of immune response at 0,
2, and 6 months.
Notes Location: Ciudad Juárez, México.
Setting: IMSS outpatient services.
Funding: Center for Border Health Research, UTEP-UTSPH Hispanic Health Disparities Research Center.
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence genera- Unclear risk The trial authors described it as "randomised", but gave no further details.
tion (selection bias)
Allocation concealment Unclear risk "A co-investigator not involved in data collection or analysis maintained the
(selection bias) study codes and allocated the supplements".
Blinding (performance Low risk "Placebo group subjects received organoleptically identical, matched place-
bias and detection bias) bo...Subject codes remained sealed until after data analysis".
All outcomes
Collaboration.
Informed decisions.
Armijos 2010 MEX (Continued)
Incomplete outcome data Low risk The analysis included 33/39 (85%) randomized participants. The trial authors
(attrition bias) clearly stated the reasons for drop-out.
All outcomes
Selective reporting (re- Unclear risk We could not retrieve the trial protocol. Trial participants were followed-up to
porting bias) 6 months but treatment completion and cure were not reported.
Other bias Low risk We did not identify any other sources of bias.
Daley 2015 IND

Methods Study design: RCT.
Study dates and duration: January 2010 to August 2011 (enrolment), follow-up completed by February
2012.
Standard care: both groups received standard category 1 tuberculosis treatment according to Indian
national guidelines.
Participants Number: 247 randomized (121 to vitamin D group, 126 to placebo).
Vitamin D group: 101/121 included in primary analysis.
Placebo group: 110/126 included in primary analysis.
Inclusion: HIV-negative participants aged 18 to 75 years with pulmonary tuberculosis (at least 1 docu-
mented positive sputum smear) who had received 1 dose or less of tuberculosis treatment
Exclusion: participants with multidrug-resistant tuberculosis, pre-existing liver or kidney disease, con-
current steroid or cytotoxic drug treatment, metastatic malignant disease, pregnancy or lactation, ac-
tive diarrhoea, hyper-calcaemia (corrected serum calcium > 2.62 mmol/L), or those who were not ex-
pected to survive for 180 days.
• Nutritional status: mean BMI ± SD: 18.2 ± 2.9 kg/m2 vitamin D group versus 17.8 ± 3.0 kg/m2 placebo.
• MDR/XDR-TB: excluded.
• Mean vitamin D status ± SD (mmol/L): 63.1 ± 46.6 vitamin D group versus 62.2 ± 51.0 placebo group.
Interventions Both groups received antituberculous treatment and were not provided with any advice about chang-
ing their diet or sun exposure.
Vitamin D group: 4 doses of tasteless, odourless 2.5 mg vitamin D3 oil (100,000 IU per dose) orally once
every 2 weeks for 8 weeks.
Placebo group: identical Miglyol (the dosing regime is not clearly stated but we assume that it was
identical to that of the vitamin D).
Outcomes Primary outcome: time to sputum culture conversion (time to first negative culture).
Secondary outcome: time to sputum culture conversion (time to first of 2 consecutive negative cul-
tures), time to smear conversion (time to first negative smear and time to first of 2 consecutive negative
smears), proportion of participants who had positive cultures at 56 days, Karnofsky performance sta-
tus and body mass index (BMI) at 56 days, rate of rise in time to detection in culture and 25-hydroxyvit-
amin-D concentration at day 180.
Primary safety outcome: incidence of hypercalcaemia (corrected serum calcium > 2.62 mmol/L).
Collaboration.
Informed decisions.
Daley 2015 IND (Continued)

Secondary safety outcome: rate of serious adverse events (death, admission to hospital, life-threaten-
ing illness, persistent disability, congenital anomaly, or predefined disease-related complications of tu-
berculosis infection), and adverse events (any untoward medical occurrence after study drug).
Notes Baseline characteristics: mostly similar between groups, but the placebo group had slightly more men
than the vitamin D group, and there were slightly more participants with isoniazid monoresistance in
the vitamin D group (14/121; 12%) compared to the placebo group (8/126; 6%). It is unclear whether
these differences were appreciably different or not.
Setting: 13 clinics in Vellore and Krishnagiri districts of Tamil Nadu, India.
Funding: Dalhousie University and Infectious Diseases Training and Research Centre.
Trial registration: NCT00366470.
Risk of bias
Random sequence genera- Low risk Bottles containing vitamin D or placebo were randomized by computer into
tion (selection bias) permuted blocks of 4 without stratification, then labelled with serial study
numbers in their randomized order. The randomization code was maintained
in Canada and only after the database was locked was the code broken.
Allocation concealment Low risk The trial authors stated that neither the participant nor the clinical and labora-
(selection bias) tory investigators and personnel in India were aware of treatment assignment.
As each participant was recruited, the next serially numbered bottle was as-
signed to the participant by study personnel. Both treatments were stored in
identical dark glass dropper bottles at room temperature and protected from
light.
Blinding (performance Low risk The trial authors state that neither the patient nor the clinical and laborato-
bias and detection bias) ry investigators and personnel in India were aware of treatment assignment.
All outcomes The randomization code was maintained in Canada by 1 investigator. After the
database was locked, the code was broken and analysis was done with knowl-
edge of assignment. Both treatments were stored in identical dark glass drop-
per bottles at room temperature and protected from light. Both treatments
were oil-based and were tasteless and odourless.
Incomplete outcome data Low risk Participants receiving 1 dose of intervention and with at least 2 sputum culture
(attrition bias) results were included in the primary analysis. Participants with only 1 culture
All outcomes result were excluded: Vitamin D group = 8/121 (7%) and placebo group = 5/126
(4%).
Selective reporting (re- Low risk Primary and secondary outcomes reported in the publication match those set
porting bias) out in the protocol.
Other bias Low risk Baseline and demographic characteristics were mostly similar between
groups, but the placebo group had slightly more men than the vitamin D
group, and there were slightly more participants with isoniazid mono-resis-
tance in the vitamin D group.
Funding source: Dalhousie University and Infectious Diseases Research and

Training Center.
Conflict of interest: Reinhold Vieth co-owns Ddrops a company that promotes

and sells Vitamin D. All other trial authors declared no conflicts of interest.
Collaboration.
Informed decisions.
Farazi 2015 IRN

Study dates and duration: recruited December 2012 to May 2014, follow-up 4 weeks.
Standard care: all participants received directly observed antituberculous therapy (isoniazid, pyrazi-
namide, rifampicin, and ethambutol for 2 months followed by isoniazid and ethambutol for 6 months).
Participants Number: 63 randomized (32 in arginine group, 31 in placebo group).
Sex: arginine group 46.9% male; placebo group 51.6% male.
Inclusion: new cases smear-positive pulmonary TB 15 years or older.
Exclusion: hospitalization, pregnancy, clinical signs of comorbidity (diabetes, malignancy, hepatic/re-

nal failure, HIV-positive), L-arginine supplementation during the past month, and allergic reactions to L-
arginine.
• Nutritional status (BMI < 18.5 kg/m2): arginine 46.9%; placebo 45.2%.
• HIV status: all HIV-negative.
• MDR/XDR-TB: not reported on.
Interventions Both groups received anti-TB treatment.
Arginine group: 1000 mg pure L-arginine hydrochloride; twice daily for 30 days.
Placebo group: 1000 mg sugar (capsules identical to L-arginine capsule); twice daily for 30 days.
Outcomes • Treatment success.

• Sputum conversion.
• Weight gain.
• Clinical symptoms at 1 and 2 months.
Not included in review: ESR, CRP.
Notes Location: Markazi Province, Iran.
Setting: clinics.
Funding: Arak University of Medical Science.
Risk of bias
Random sequence genera- Low risk "random number tables".

Allocation concealment Low risk "Preparation of the randomization envelopes was performed by a member of
(selection bias) the staff who was not directly involved in the study and a sealed copy of the
treatment code was kept by the project leader until all data had been collected
and analyzed".
Blinding (performance Low risk "identical capsules of L-arginine...or placebo".

bias and detection bias)
All outcomes
Incomplete outcome data Low risk Supplement: 2/34 (6%).

(attrition bias)
Placebo: 3/34 (9%).
Collaboration.
Informed decisions.
Farazi 2015 IRN (Continued)

All outcomes
Selective reporting (re- Low risk Primary and secondary outcomes reported in the publication match those set
porting bias) out in the protocol.
Other bias Low risk Baseline and demographic characteristics were similar between groups.
Funding source: Arak University of Medical Sciences.
Conflicts of interest: all trial authors declared no conflict of interest.
Ginawi 2013 IND

Study dates and duration: not stated.
Standard care: not stated.
Participants Number: 178 randomized, 127 analysed.
Inclusion criteria: category 1 pulmonary tuberculosis patients attending the DOTS centre.
Exclusion criteria: those who did not take medication regularly (missing even 1 dose in the first 2
months), had severe adverse reactions, or drug-resistant tuberculosis.
• Nutritional status: not stated.

• HIV status: not stated.
• Vitamin A status µg/dL (SD): unclear if it is baseline data or % increase from baseline.
• Zinc status µg/L (SD): unclear if it is baseline data or % increase from baseline.
Interventions Placebo group: lactose only.
Vitamin A group: 1500 retinol equivalents (5000 IU) vitamin A (as retinyl acetate).
Zinc group: 15 mg zinc (as zinc sulfate).
Vitamin A and zinc group: 1500 retinol equivalents (5000 IU) vitamin A (as retinyl acetate) and 15 mg
zinc (as zinc sulfate).
Supplement and placebo were given to the participants with the antituberculous drugs on DOTS day.
Outcomes • Sputum positivity: sputum smear negative for tubercle bacilli at 2 months.
• Blood chemistry: haemoglobin, white blood cell count, erythrocyte sedimentation rate (ESR), serum
albumin, serum retinol, and serum zinc concentration at 2 and 6 months.
Notes Location: Lucknow, North India.
Setting: not stated.
Funding: not stated.
Risk of bias
Collaboration.
Informed decisions.
Ginawi 2013 IND (Continued)
Random sequence genera- Unclear risk Not described.

Allocation concealment Unclear risk Not described.

(selection bias)
Blinding (performance Low risk The trial authors described the trial as a double blind, placebo-controlled trial.
bias and detection bias) Supplement and placebo capsules were indistinguishable in appearance both
All outcomes externally and internally.
Incomplete outcome data High risk The trial authors stated that they excluded non-compliant participants, partic-
(attrition bias) ipants who experienced severe adverse reactions, and participants with drug-
All outcomes resistant tuberculosis from the analysis. The trial authors did not describe rea-
sons for loss to follow-up of 51 participants.
178 participants randomized.
127 participants completed the trial.
28.6% loss to follow-up (> 10% therefore high risk of bias).
Selective reporting (re- Unclear risk We were unable to obtain the trial protocol so it is unclear if there is selective
porting bias) outcome reporting in this instance.
Other bias Unclear risk Baseline comparability: no significant difference in biochemical status at base-
line. No information on age, gender, or nutritional status at baseline.
Conflicts of interest: not reported.
Funding: no information provided on funding resources.
Hanekom 1997 ZAF

Study dates and duration: dates not stated, follow-up 3 months.
Standard care: all participants received directly observed antituberculous therapy, without any routine
multivitamin preparations.
Inclusion criteria: children aged under 15 years with pulmonary tuberculosis.
Exclusion criteria: HIV-positive.
• Nutritional status: mean weight z score (SD): −1.41 (1.41) treatment group versus −1.44 (1.34) control
group.
• MDR/XDR-TB: not described.
• Vitamin A status µg/dL (SD): 17.6 (10.1) supplement group versus 18.6 (10.6) placebo group; normal
range 20 to 80.
Interventions Group 1: vitamin A 200,000 IU; 2 doses before beginning antituberculous therapy.
Group 2: placebo with identical appearance.
Collaboration.
Informed decisions.
Hanekom 1997 ZAF (Continued)
Outcomes • Resolution of respiratory symptoms at 6 weeks and 3 months.

• Change-in-weight z scores.
• Vitamin A levels at baseline, 6 weeks, and 3 months.
Not included in review: chest X-ray changes, retinol binding protein, and prealbumin levels: at baseline,
6 weeks, and 3 months.
Notes Location: Cape Town, South Africa.
Setting: children's hospital (outpatient or inpatient not stated in report).
Funding: the Glaxo 'Action TB' International Research Initiative, Medical Research Council of South
Africa.
Risk of bias
Random sequence genera- Unclear risk The trial authors described the trial as "randomised"; but did not provide any
tion (selection bias) further details.
Allocation concealment Unclear risk None described.

(selection bias)
Blinding (performance Low risk The trial authors described the trial as "double blind", with an "identical ap-
bias and detection bias) pearing placebo". No comment on outcome assessors.
All outcomes
Incomplete outcome data High risk The trial enrolled 110 participants, of which 15 were later excluded for not at-
(attrition bias) tending the last follow-up appointment and 10 for not meeting the inclusion
All outcomes criteria. The trial authors presented data for 76 of the remaining 85 (85%) par-
ticipants.
Selective reporting (re- Unclear risk We were unable to retrieve the trial protocol. There was no evidence of selec-
porting bias) tive reporting.
Jahnavi 2010 IND

Study dates and duration: August to December 2005, follow-up 1-year.
Standard care: all participants received directly observed antituberculous therapy according to the Re-
vised National Tuberculosis Control Programme, India, without any routine multivitamin preparations.
Participants Number: 100 participants randomized.
Inclusion criteria: adults aged 18 to 65 years; with active tuberculosis; pulmonary or extrapulmonary
smear/culture positive, BMI < 20 kg/m2.
Exclusion criteria: HIV-positive, diabetes, other severe underlying disease.
• Nutritional status: mean weight kg (SD):43.0 (8.2) treatment group versus 44.2 (8.2) control group;
mean BMI (SD): 17.1 kg/m2 (2.8) treatment group versus 17.9 kg/m2 (2.1) control group.
Collaboration.
Informed decisions.
Jahnavi 2010 IND (Continued)

• MDR/XDR-TB: unknown.
• Micronutrient status: not assessed at baseline.
Interventions Group 1: Control group: standard tuberculosis treatment as per the Revised National Tuberculosis Con-
trol Programme. A general instruction to "increase food intake".
Group 2: targeted dietary advice to reach 35 kcal/kg/day and food supplements that consisted of
"sweet balls" made of wheat flour, caramel, groundnuts, vegetable ghee, sprouted gram, and nuts —
each containing 6 g protein and 600 kcal — consumed daily under supervision for 3 months.
Outcomes • Body weight at baseline and 3 months.

• Maximum grip strength and timed stand test at baseline and 3 months.
• Quality of life at baseline and 3 months.
• Sputum conversion at 2, 4, and 6 months.
• Completion of the entire course of treatment.
• Deaths during study.
Notes Location: Krishna district, Andhrapradesh, India.
Setting: community-based catchment area of 1 medical college and 1 DOTS centre.
Funding: University of Padova, Italy.
Risk of bias
Random sequence genera- High risk "randomisation performed by randomly shuffling the envelopes that con-
tion (selection bias) tained the study codes".
Allocation concealment Unclear risk The trial did not describe this, but given that the envelopes with codes were
(selection bias) shuffled it would seem adequate.
Blinding (performance High risk The trial did not make any attempt to blind the participants; the trial report
bias and detection bias) does not mention whether or not the outcome assessors were blinded.
All outcomes
Incomplete outcome data Unclear risk There were no apparent losses to follow-up.
(attrition bias)
All outcomes
Selective reporting (re- High risk Despite 1 year of follow-up, the trial authors only reported outcomes at 3
porting bias) months, and incompletely reported sputum conversion.
Jeremiah 2014 TZA

Methods Study design: open label RCT.
Study dates and duration: September 2010 to August 2011.
Standard care: treatment was given as FDC tablets, each contained Isoniazid (75 mg), rifampin (150
mg), pyrazinamide (400 mg), and ethambutol (275 mg) during the intensive phase, and for continua-
Collaboration.
Informed decisions.
Jeremiah 2014 TZA (Continued)

tion phase the tablet contained isoniazid (75 mg) and rifampicin (150 mg) only. If body weight < 50 kg
then 3 tablets, and > 50 kg 4 tablets.
Participants Number: 469 screened, 100 randomized (by HIV status).
Inclusion criteria: sputum smear positive pulmonary tuberculosis patients 15 years or older, regardless
of HIV status, with written informed consent.
Exclusion criteria: HIV-positive participants on ART, pregnant women, terminally sick patients unlikely
to survive > 48 hrs, and non-residents of Mwanza City.
• Nutritional status: median BMI (kg/m2) pulmonary tuberculosis-positive/HIV-positive supplement

group: 19.4; 18.3, 20.5 pulmonary tuberculosis positive/HIV-positive no supplement group: 18.8; 17.4,
20.1, pulmonary tuberculosis positive/HIV-negative supplement group: 18; 16.9, 19.7. Pulmonary
tuberculosis positive/HIV-negative no supplement group: 18.6; 17.3,19.6. Median weight (kg): pul-
monary tuberculosis positive/HIV-positive supplement group: 52.3; 49.9, 56.6. pulmonary tuberculo-
sis positive/HIV-positive no supplement group: 51.9; 49.2, 56.1. Pulmonary tuberculosis positive/HIV-
negative supplement group: 50.7; 48.2, 57.7. Pulmonary tuberculosis positive/HIV-negative no sup-
plement group: 52; 44.6, 58.8.
• HIV status: median CD4 cell count; IQR cells/µL pulmonary tuberculosis-positive/HIV-positive sup-
plement: 243.5; 140.5, 337.5 pulmonary tuberculosis-positive/HIV-positive no supplement: 168; 64,
338 pulmonary tuberculosis positive/HIV-negative supplement: 611; 457; 726 pulmonary tuberculosis
positive/HIV-negative no supplement: 637; 433, 812.
Interventions • Group 1: pulmonary tuberculosis-positive/HIV-positive and Group 2: pulmonary tuberculosis posi-

tive/HIV-negative: no supplement.
• Group 3: pulmonary tuberculosis-positive/HIV-positive and Group 4: pulmonary tuberculosis posi-
tive/HIV-negative: daily 5 high energy and vitamin/mineral enriched biscuit bars containing about
1000 kcal and additional vitamins and minerals (including zinc and selenium). Produced by Compact
A/S. Supplement provided during first 2 months of tuberculosis treatment. Intake monitored by pa-
tient treatment supporter.
Basic biscuit (30 g): 4.5 g protein, 615 kJ energy, 120 mg phosphorous, 120 mg calcium, 36 mg magne-
sium, 70 mg sodium, 150 mg potassium, and traces < 1 mg of iron and zinc.
Biscuit with additional micronutrients: as above plus 1.5 mg vitamin A, 20 mg thiamin, 20 mg riboflavin,
25 mg vitamin B6, 50 µg vitamin B12, 0.8 mg folic acid, 40 mg niacin, 200 mg vitamin C, 60 mg vitamin E,
5 µg vitamin D, 0.2 mg selenium, 5 mg copper, 30 mg zinc.
Outcomes • Mortality at 2 months.

• Median weight gain at 2 months.
• Sputum culture at 2 months.
Not included in review: pharmacokinetics of rifampin and genotype of organic anion-transporting

polypeptide encoded by SLCO1B1 rs149032 responsible for hepatic drug disposition.
Notes Location: Mwanza, Tanzania.
Setting: urban and suburban clinics.
Funding: Danish Ministry of Foreign Affairs (DANIDA, DFC file no. 09-026RH) through Denmark’s Interna-
tional Development Cooperation.
Risk of bias
Collaboration.
Informed decisions.
Jeremiah 2014 TZA (Continued)
Random sequence genera- Low risk The trial authors stated that simple randomization, stratified by HIV status,
tion (selection bias) was computed using the website www.randomization.com.
Allocation concealment High risk This was an open-label trial, therefore it is likely that the people allocating the
(selection bias) treatment where aware of which treatment group they were allocating the
participants to.
Blinding (performance High risk This was an open-label trial, therefore the participants and personnel were not
bias and detection bias) blinded to the treatment. It is unclear if the outcome assessors were blinded to
All outcomes the treatment.
Incomplete outcome data High risk The trial authors stated that they analysed all available data.
(attrition bias)
All outcomes Intervention arm: 2/51 (4%) lost to follow-up (died or defaulted); no supple-
ment arm: 6/49 (12%) lost to follow-up (died or defaulted).
Selective reporting (re- Low risk We obtained the protocol obtained. The primary and secondary outcomes re-
porting bias) ported in the publication are in line with those set out in the protocol.
Karyadi 2002 IDN

Study dates and duration: December 1997 to December 1998, follow-up 6 months.
Standard care: all participants received directly observed antituberculous therapy in line with WHO
recommendations: 2RHZE/4HR.
Inclusion criteria: adults aged 15 to 55 years, 3 positive sputum smears, clinical and radiological signs
consistent with pulmonary tuberculosis.
Exclusion criteria: previous antituberculous therapy, drug resistance, extra-pulmonary tuberculosis,

pregnancy, lactation, use of corticosteroids, vitamin A, zinc, or iron in the previous month, moderate to
severe injury or surgery in the previous month, diabetes, renal failure, liver disease, neoplasm, or con-
gestive heart failure.
• Nutritional status: BMI < 18.5 kg/m2: 25/40 treatment group versus 26/40 placebo group.
• HIV status: not mentioned.
• Vitamin A status: mean plasma retinol µmol/L (SD): 0.82 (0.04) supplement group versus 0.90 (0.04)
placebo group; normal range > 0.7.
• Zinc status: mean plasma zinc µmol/L (SD): 11.52 (0.26) supplement group versus 11.15 (0.28) placebo
group; normal range > 10.7.
Interventions Group 1: daily capsules of vitamin A 5000 IU and zinc 15 mg (as zinc sulphate) for 6 months.
Group 2: placebo (lactose).
Outcomes • Sputum-culture positive at 0, 2, and 6 months.

• Body weight at 0, 2, and 6 months.
• BMI at 0, 2, and 6 months.
Collaboration.
Informed decisions.
Karyadi 2002 IDN (Continued)

• MUAC, body fat percentage.
• Karnofsky score at 0, 2, and 6 months.
• Blood retinol and zinc levels.
Not included in the review: biceps, triceps, subscapular, and supra-iliac skinfold thickness; blood
haemoglobin, zinc protoporphyrin, x-ray changes.
Notes Location: Jakarta, Indonesia.
Setting: pulmonary outpatient clinics of 1 public hospital and 3 health centres.
Funding: Gesellschaft fur Technische Zusammenarbeit (GTZ) GmbH, Eschborn, Germany; the Neys-van
Hoogstraten Foundation; the Directorate General of Communicable Disease Control and Environmen-
tal Health, Indonesia; the Integrated Excellent Research Project, Ministry of Research and Technology
Indonesia; PT Kimia Farma, Indonesia provided the supplements and placebo; PT Indo Farma provided
the tuberculosis drugs.
Risk of bias
Random sequence genera- Low risk "We used a table with randomly assorted digits to allocate patients into 2
tion (selection bias) groups".
Allocation concealment Unclear risk The trial authors did not describe the method of allocation concealment but
(selection bias) seems likely given the description of blinding.
Blinding (performance Low risk "Supplements and placebo were indistinguishable". "The authors, health staff,
bias and detection bias) and patients were unaware of the treatment code until the study was complet-
All outcomes ed".
Incomplete outcome data High risk The trial authors appear to have added 2 exclusion criteria post-hoc: partic-
(attrition bias) ipants who missed even 1 day of treatment during the first 2 months, and
All outcomes participants who had severe adverse drug effects. The trial authors analysed
80/110 (72%) participants.
Selective reporting (re- Unclear risk We were unable to retrieve the trial protocol. Although cure is a standard out-
porting bias) come for tuberculosis programmes, this was not reported.
Kota 2011 IND

Standard care: all participants received intensive phase antimicrobial treatment comprising of isoni-
azid, rifampicin, pyrazinamide, and ethambutol. All participants were adjusted for oral hypoglycaemic
agents and insulin for glycaemic control.
Participants Number: 45 enrolled, (15 excluded), 30 randomized and included in the final analysis.
Inclusion criteria: older than 15 years, newly diagnosed pulmonary tuberculosis patients, poorly con-
trolled Type 2 diabetes mellitus (HbA1C > 7%) and serum 25(OH) D < 20 ng/mL. Pulmonary tuberculo-
sis diagnosis based on 1 of the following criteria: at least 2 positive sputum smears (from 3), 1 positive
smear and typical picture of lung infiltration on chest X-ray.
Collaboration.
Informed decisions.
Kota 2011 IND (Continued)

Exclusion criteria: younger than 15 years, patients already on ATT, serum 25(OH) D > 20ng/mL, diseases
affecting vitamin D metabolism such as malabsorption, renal failure, or with prolonged immobilization.
• Nutritional status: mean weight (kg ± SD) supplement group: 49.1kg ± 4.5, control: 44.6 ± 5.6.
• MDR/XDR-TB: not stated.
• Vitamin D 25-(OH) D (ng/mL) status: supplement group 12.8 ± 4.5, control group: 11.1 ± 4.7.
• Vitamin A and zinc status: not reported.
Interventions Group 1: vitamin D supplement along with intensive phase ATT. Oral cholecalciferol sachets (60,000 IU/
week) and calcium carbonate (1000 mg/day).
Group 2: no vitamin D supplement just intensive phase ATT.
The trial authors did not state the duration of the intervention.
Outcomes • Vitamin D levels every 4 weeks for 12 weeks.

• Sputum smear conversion every 4 weeks for 12 weeks.
• Fasting blood sugar and post lunch blood sugar monthly.
• HbA1C monthly.
• ESR monthly.
Notes Location: Medwin Hospital, Hyderabad, India.
Setting: hospital setting (inpatient).
Funding: not reported.
Risk of bias
Random sequence genera- Unclear risk The trial authors did not describe the randomization process.
Allocation concealment Unclear risk The trial authors did not mention whether treatment allocation was concealed
(selection bias) or not.
Blinding (performance Unclear risk The trial authors did not describe any aspect of blinding in the paper.
All outcomes
Incomplete outcome data Low risk According to data provided, the trial authors included all of the participants
(attrition bias) randomized at baseline in the final analysis at 12 weeks.
All outcomes
Selective reporting (re- Unclear risk We were unable to obtain the study protocol so it is unclear if all of the pro-
porting bias) posed outcomes have been reported on.
Lawson 2010 NGA

Collaboration.
Informed decisions.
Lawson 2010 NGA (Continued)

Study dates and duration: September 2003 to June 2005, follow-up 6 months.
Standard care: all participants received standard National Tuberculosis Programme treatment for tu-
berculosis: 2RHZE/4HE.
Participants Number: 350 enrolled.
Inclusion criteria: adults aged > 15 years with sputum positive pulmonary tuberculosis.
Exclusion criteria: previous antituberculous therapy, pregnancy, lactation, use of corticosteroids or

zinc in the previous month, major surgery in the previous month, diabetes, severe cardiovascular/renal
or hepatic disease, currently taking oral contraceptives, unable to return.
• Nutritional status: mean BMI (SD):19.8 kg/m2 (3.3) placebo group, 21.3 kg/m2 (4.7) zinc plus placebo
group, 19.6 (3.5) zinc plus vitamin A group.
• HIV status: 45% HIV-positive placebo group, 49% zinc plus placebo group, 42% zinc plus vitamin A
group.
• MDR/XDR-TB: not mentioned.
• Vitamin A status: not reported.
• Zinc status: not reported.
Interventions Group 1: placebos (similar in appearance to zinc and vitamin A tablets).
Group 2: 90 mg of elemental zinc per week plus daily placebo.
Group 3: 90 mg of elemental zinc per week plus 1500 mcg retinol equivalents (5000IU vitamin A) weekly.
Supplements were observed for 2 months then given as monthly supplies for the next 4 months.
Outcomes • Time to sputum smear conversion.

• Chest X-ray scores at 2 and 6 months.
• Clinical symptoms (cough, fever and night sweats) at 2 and 6 months.
• BMI at 2 and 6 months.
• Karnofsky score at 2 and 6 months.
• Deaths during the trial.
Not included in the review: ESR and haemoglobin values.
Notes Location: Abuja, Nigeria.
Setting: 8 district hospitals.
Funding: no sources stated.
Risk of bias
Random sequence genera- Low risk "Prepared the allocation sequence using random numbers generated in
tion (selection bias) Minitab...using permuted block randomisation with four different black sizes".
Allocation concealment Low risk "One of the investigators who was not on the site prepared the allocation se-
(selection bias) quence...and the randomisation sequence was kept at the Liverpool School of
Tropical Medicine Treatment...allocation was designated by a lettered code
and concealed from the investigators and subjects until data analysis was
completed".
Collaboration.
Informed decisions.
Lawson 2010 NGA (Continued)
Blinding (performance Low risk "Placebo tablets resembled the supplement tablets".
All outcomes
Incomplete outcome data Low risk There were 14 of 116 in the placebo group, 15 of 117 in the zinc group, and 21
(attrition bias) of 117 in the zinc and vitamin A group lost to follow-up.
All outcomes
Other bias Low risk We were unable to identify any other sources of bias.
Lodha 2014 IND

Study dates and duration: January 2008 to June 2012.
Standard care: all children had 2 months intensive phase that used 3 or 4 drugs (isoniazid, rifampicin,
pyrazinamide, and ethambutol) followed by 4 months of isoniazid and rifampicin. Category 3 treat-
ment was isoniazid, rifampicin, and pyrazinamide for 2 months followed by 4 months of rifampicin and
isoniazid.
Daily doses were based on weight: 5 to 7 mg isoniazid/kg; 10 to 13 mg rifampicin/kg; 35 to 40 mg pyrazi-

namide/kg; 20 to 25 mg ethambutol/kg.
In case of non-response (clinical or radiological) antituberculous therapy was changed to 2 months

streptomycin, isoniazid, rifampicin, pyrazinamide, and ethambutol; followed by 1 month of isoniazid,
rifampicin, and ethambutol; and 5 months of isoniazid, rifampicin, and ethambutol. Any child who did
not respond on this regimen and for whom an alternative diagnosis was ruled out was started on 2nd
line antituberculous therapy, which included an injection of kanamycin, ofloxacin, and ethionamide
and an additional drug (cycloserine or co-amoxiclav).
Participants were followed up every 2 weeks in the first 2 months, and then every 4 weeks for the re-
maining 4 months of ATT treatment.
Participants Number: 403 randomized. At 2 months: data on 393 children; at 6 months: data on 381 children.
Inclusion criteria: children 6 months to 15 years who presented with any of the following symptoms
were considered tuberculosis suspects and screened: cough > 2 weeks with no improvement after 7 to
10 days amoxicillin, fever > 2 weeks with no improvement during 7 to 10 days amoxicillin, recent un-
explained weight loss or failure to thrive, unusual or unexplained fatigue or lethargy or subtle clinical
symptoms and history of close contact with adult tuberculosis patient. All had chest X-ray and tuber-
culin skin test. If the chest X-ray was consistent with intrathoracic tuberculosis, the child was consid-
ered to have tuberculosis. The trial enrolled children with newly diagnosed probable intrathoracic tu-
berculosis with or without extrapulmonary lesions.
Exclusion criteria: bilateral pedal oedema; known HIV infection; history of antituberculous therapy or
isoniazid prophylaxis > 4 weeks; signs of upper airway obstruction; O2 saturation < 92%; signs of renal,
hepatic, or cardiovascular disease; inability to attend follow-up session for reading of the skin test; doc-
umented intake of zinc continuously > 2 weeks in the 4 weeks preceding enrolment; CNS, osteoarticu-
lar, pericardial or renal tuberculosis; history of contact with a documented case of drug-resistant tuber-
culosis; or were non-residents of Delhi.
• Nutritional status (weight-for-age z score): MN+Z group −2.72 ± 1.85; MN group −2.8 ± 1.6; zinc group
−2.5 ± 1.5; placebo −2.8 ± 1.6.
Collaboration.
Informed decisions.
Lodha 2014 IND (Continued)

• HIV status: HIV-negative only.
• MDR/XDR-TB (n/N tested): MN+Z group 1/18 (5.6%); MN group 1/29 (3.4%); zinc group 1: 1/21 (4.8%),
placebo group 2/26 (7.9%).
• Serum zinc concentration < 65 µg/dL (n/N): MN + Z 60/102; MN 51/100; zinc 53/101; placebo 55/100.
Interventions All children received antituberculous therapy.
Supplemented daily for 6 months.
Zinc group: zinc (20 mg elemental zinc).
Micronutrient group: micronutrients (vitamin A, thiamine, riboflavin, vitamins B6 and B12, folic acid,
niacin, vitamin C, vitamin E, vitamin D, selenium, and copper) without zinc.
Micronutrient + zinc group: micronutrients (as above) in combination with zinc (20 mg elemental zinc).
Placebo group: placebo.
Outcomes • Change in weight-for-age z score at 6 months and resolution of pulmonary lesions using chest X-ray
at 6 months.
• Height-for-age z score.
• Mid upper arm circumference.
• Triceps skin fold thickness.
• BMI-for-age z score at 2 and 6 months.
• Resolution of symptoms at 2 months (as reported by parents).
• Proportion of children requiring extension of intensive phase of therapy at 2 months.
• Improvement in chest X-ray at 2 months.
Notes Location: Delhi, India. All India Institute of Medical Sciences & Kalawati Saran Children Hospital associ-
ated with Lady Hardinge Medical College.
Setting: urban, hospital.
Funding: Norwegian Programme for Development, Research and Education and Research council of
Norway and Global Health and Vaccination Research GLOBVAC.
Risk of bias
Random sequence genera- Low risk Adequate randomization described: a scientist who was not involved in the da-
tion (selection bias) ta collection and analysis, generated random-allocation sequences in permut-
ed blocks of variable sizes separately for the 2 sites. Separate sequences were
generated for the age groups 6 months to 3 years, 4 to 6 years, 7 to 9 years, and
10 to 15 years.
Allocation concealment Low risk Bottles that contained the micronutrient supplements were serially numbered
(selection bias) for each stratum for the 2 sites. The 4 study syrup preparations had a similar
packaging, appearance, and smell. The participant, physician, and laboratory
personnel were blinded to the intervention.
Blinding (performance Low risk The 4 study syrup preparations had a similar packaging, appearance, and
bias and detection bias) smell. The participant, physician, and laboratory personnel were blinded to
All outcomes the intervention. They maintained masking during the data analysis by coding
treatment allocation with 4 letters.
Incomplete outcome data Low risk Loss to follow-up overall

(attrition bias)
All outcomes 2 months: 10/403 (2.5%); 6 months: 22/403 (5.5%)
Collaboration.
Informed decisions.
Lodha 2014 IND (Continued)

6 month loss to follow-up per group:
Micronutrient + zinc group: 8/102 (8%)
Micronutrient group: 4/100 (4%)
Zinc group: 7/101 (7%)
Placebo group: 3/100 (3%)
Selective reporting (re- Low risk We obtained the study protocol and the trial authors reported all outcomes in
porting bias) the paper that they listed in the protocol. Five additional outcomes were listed
in the protocol and will likely be reported in future papers.
• Interferon gamma responses to M. tuberculosis antigens ESAT6 and CF10 by

quantiferon assay at baseline, 2 months, and 6 months of treatment.
• To study effect of zinc supplementation on ocular toxicity in children receiv-
ing ethambutol by VER.
• To document drug resistance (S, I, R, E) patterns among children with culture
confirmed tuberculosis.
• To document genotypic strain diversity among children with culture con-
firmed tuberculosis, also associations between strain type and disease sever-
ity and drug resistance.
• To document the spectrum of mycobacterial species by culture in children
clinically suspected of having pulmonary tuberculosis.
Other bias Low risk Baseline characteristics: most demographic and clinical profiles were compa-
rable at baseline. However, there were some differences regarding sex distri-
bution, parental educational level, diagnosis, and energy and zinc intakes be-
fore enrolment.
Fundings: non-biased source of funding.
Martineau 2011 GBR

Study dates and duration: January 2007 to July 2009, follow-up 8 weeks.
Standard care: All participants received directly observed antituberculous therapy: 2RHZE.
Participants Number: 146 enrolled; 128 in primary analysis.
Inclusion criteria: adults aged over 18 years, newly diagnosed sputum positive pulmonary tuberculosis.
Exclusion criteria: known intolerance to vitamin D or antituberculous therapy, sarcoidosis, hyper-

parathyroidism, nephrolithiasis, HIV infection, pulmonary silicosis, malignancy, renal or hepatic fail-
ure, oral corticosteroids, cytotoxic drugs or other immunosuppressant therapy in the last month, an-
tituberculous therapy for > 7 days in the preceding 6 months, currently taking antituberculous thera-
py other than RHZE, known rifampicin resistance, serum corrected calcium > 2.65 mmol/L, aspartate
transaminase or alanine transaminase > 120 IU/L, total serum bilirubin > 40 µmol/L, serum creatinine >
250 µmol/L, pregnant or breastfeeding.
• Nutritional status: mean BMI (SD): 20.1 kg/m2 (3.1) supplement group versus 20.2 kg/m2 (2.7) placebo
group.
• HIV status: 5% positive supplement group versus 3% positive placebo group.
• MDR/XDR-TB: rifampicin resistant isolate: 0% supplement group versus 4% placebo group.
Collaboration.
Informed decisions.
Martineau 2011 GBR (Continued)

• Vitamin D status: mean serum 25-hydroxyvitamin D nmol/L (SD): 21.1 (20.0) supplement group versus
21.3 (19.0) placebo group; normal range > 75 nmol/L.
Interventions Group 1: 4 oral doses of 2.5 mg vitamin D3 (Viganol oil, Merck Serono) on or before day 7 after the start
of tuberculosis treatment, day 14, day 28, and day 42.
Group 2: an organoleptically identical placebo (Miglyol Oil, Caesar and Loretz) given as above.
Outcomes • Median time to sputum conversion.

• Serious adverse events.
• Other adverse events.
• Death.
• Mean serum 25-hydroxyvitamin D at baseline and day 56.
Not included in the review: haemoglobin level, mean corpuscular volume, total white blood cell count,
lymphocyte count, monocyte count, neutrophil count, platelet count, erythrocyte sedimentation rate,
C-reactive protein levels, chest radiography zones affected.
Notes Location: London, UK.
Setting: 10 National Health Service Trusts.
Funding: British Lung Foundation.
Risk of bias
Random sequence genera- Low risk "generated a random sequence using a computer program that assigned the
tion (selection bias) term active or placebo to the numbers 1 to 200 by permuted block randomisa-
tion with blocks of 10".
Allocation concealment Low risk "The packs were then assigned a randomisation number according to this
(selection bias) computer generated randomisation sequence", "Study staff who assigned pa-
tients to the active drug or placebo had no knowledge of the next assignment
in the sequence, because they did not have access to the study code", "Treat-
ment allocation was concealed from patients and study staff".
Blinding (performance Low risk "organoleptically identical placebo".

All outcomes "adverse events judged to be potentially related to vitamin D by physicians un-
aware of allocation".
"Those analysing the data were not masked to group assignment".
Incomplete outcome data Unclear risk The trial authors reported outcomes on an intention-to-treat basis. The num-
(attrition bias) ber of participants lost to follow-up (3 in the placebo group and 6 in the place-
All outcomes bo group) and the reasons were clearly reported for each group.
porting bias) tive outcome reporting.
Martins 2009 TLS

Collaboration.
Informed decisions.
Martins 2009 TLS (Continued)

Study dates and duration: March 2005 to November 2005, follow-up 8 months.
Standard care: all participants received routine care including drugs (DOTS), default tracing, and clini-
cal monitoring according to National TB Control Program guidelines: 2RHZE/6HE.
Participants Number: 270 enrolled; numbers presented varies for each outcome.
Inclusion criteria: adults aged > 18 years, pulmonary tuberculosis diagnosed by positive sputum criteria
or NTP guidelines, willingness to receive treatment from the clinic for 8 months.
Exclusion criteria: pregnancy, previous treatment for tuberculosis for > 1 month, unable to agree to
complete treatment, unwilling to attend for a daily meal, chose community DOT.
Description: adults aged > 18 years.
• Nutritional status: BMI < 18.5 kg/m2: 108/137 treatment group versus 105/133 control group.
• Described as poor; 80% had no formal income, food was readily available but expensive.
Interventions Group 1: a daily meal (intensive phase; week 1 to 8) followed by a food parcel (continuation phase;
week 9 to 32). The meal consisted of a bowl of meat, kidney beans, and vegetable stew with rice. The
food parcel contained unprepared red kidney beans, rice, and oil adequate for 1 meal per day.
Group 2: verbal and written nutritional advice concerning locally available food that would constitute a
balanced diet.
Outcomes • Treatment completion at 8 months.

• Mean weight gain (%) at 8 weeks and 32 weeks.
• Cough clearance at 4, 8, and 32 weeks.
• Sputum clearance.
• Adverse events.
Not included in review: mean compliance, default rate.
Notes Location: Dili, Timor-Leste.
Setting: 3 primary care clinics.
Funding: the UNICEF/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical
Diseases, Australian National Health and Research Council.
Risk of bias
Random sequence genera- Low risk "An independent statistician computer generated a random allocation se-
tion (selection bias) quence with randomly varying block sizes in Stata (version 8)".
Allocation concealment Low risk "The sequence was concealed from all investigators with sequentially num-
(selection bias) bered opaque sealed envelopes prepared distant from the study site".
Blinding (performance Low risk "Both participants and treatment providers were aware of an individual’s allo-
bias and detection bias) cation status after randomisation. An independent observer, who was blinded
All outcomes to the intervention received by the patients, however, determined the primary
outcome".
Collaboration.
Informed decisions.
Martins 2009 TLS (Continued)
Incomplete outcome data High risk High levels of missing data for the outcomes of cough clearance (28% inter-
(attrition bias) vention versus 27% control at 8 weeks) and weight gain (12% intervention ver-
All outcomes sus 9% control at 8 weeks, 29% intervention versus 33% control at 32 weeks).
These could significantly affect the result.
Selective reporting (re- Low risk We were unable to retrieve the trial protocol. There was no evidence of selec-
Mehta 2011 TZA

Standard care: all participants received antituberculous therapy (6 months DOT: isoniazid 50 mg, ri-
fampicin 200 mg, ethambutol 10 to 15 mg/kg, and pyrazinamide 20 to 30 mg/kg daily for 2 months; iso-
niazid 50 mg and rifampicin 200 mg daily for 4 months) and were visited at home by the study nurse
every 2 weeks.
Study dates and duration: May 2005 to September 2007. Follow-up 8 weeks.
Participants Number: 255 enrolled and randomized; outcomes presented for 237.
Inclusion criteria: aged 6 weeks to 5 years; loss of > 10% maximum weight or failure to gain weight for 2
months; cough with wheeze for ≥ 4 weeks; history of household contact with a probable or confirmed
tuberculosis case in the past 6 months; pyrexia of unknown origin; painless swelling in a group of cervi-
cal lymph nodes; children who were diagnosed with tuberculosis in the past 5 years and have received
antituberculous therapy for a period < 4 weeks. Positive TST (≥ 10 mm induration in HIV-negative and
≥ 5 mm in HIV-positive; after 48 to 72 hours) or with chest X-ray indicative of tuberculosis (based on un-
equivocal lymphadenopathy or military tuberculosis) eligible for enrolment.
Exclusion criteria: children who had been treated with antituberculous therapy exceeding 4 weeks in
the past 1 year.
• Nutritional status: median (IQR) MUAC cm: 13.0 (11.9 to 14.25) multivitamin group versus 13.0 (11.2 to
14.0) placebo group; weight (kg): 8.48 multivitamin group versus 7.95 placebo group.
• HIV status: 39.1% HIV-positive in the multivitamin group, 29.1% HIV-positive in the placebo group.
• Micronutrient status: not documented.
Interventions Group 1: multivitamin supplements daily for the first 2 months. Composition: vitamin B1 0.5 mg, vita-
min B2 0.6 mg, niacin 4 mg, vitamin B6 0.6 mg, folate 130 μg, vitamin B12 1 μg, vitamin C 60 mg, and vit-
amin E 8 mg.
• Age < 6 months: 1 capsule daily.

• Age 7 to 36 months: 2 capsules daily.
• Age > 36 months: 3 capsules daily.
Group 2: placebo daily for the first 2 months.
Outcomes • Weight gain at 2 months.

• Deaths during treatment.
Not included in this review: height, MUAC, and triceps skin-fold thickness changes at 2 months, clear-
ance on chest x-ray at 2 months, immunological outcomes.
Collaboration.
Informed decisions.
Mehta 2011 TZA (Continued)
Notes Location: Dar Es Salaam, Tanzania.
Setting: hospital paediatric clinic.
Registration number: NCT00145184.
Source of funding: National Institutes of Health Grant, the Harvard School of Public Health.
Risk of bias
Random sequence genera- Low risk An off-site statistician generated the randomization sequence; a list from 1 to
tion (selection bias) 400 was prepared according to this randomization sequence in blocks of 20.
Allocation concealment Low risk At enrolment, the trial staff assigned each eligible child to the next numbered
(selection bias) bottle of regimen at the site. The statistician kept the randomization list confi-
dential, with the exception of the pharmaceutical company that prepared the
blinded treatment.
Blinding (performance Low risk To minimize the risk of loss of blinding, the regimen bottles, with no visual dif-
bias and detection bias) ference between active regimen and placebo, were received from the man-
All outcomes ufacturer without any identification; the study staff then labelled the bottles
with the participant’s initials and identification number. Both the clinicians
and the participants were blinded to the study regimen.
Incomplete outcome data Low risk Losses to follow-up were low: 2.3% supplement group versus 1.6% placebo
(attrition bias) group.
All outcomes
Selective reporting (re- Low risk Although we were unable to retrieve the trial protocol, there was no evidence
porting bias) of selective reporting.
Other bias High risk There was a large difference in HIV status between the intervention and con-
trol groups at baseline.
Mily 2015 BGD

Standard care: all participants received antituberculous therapy (isonizaid 75 mg, rifampicin 150
mg, pyrazinamide 400 mg, and ethambutol 275 mg for 2 months, followed by rifampicin 150 mg plus
isonizaid 75 mg for 4 months), given under direct observation.
Study dates and duration: December 2010 to December 2013. Follow-up 24 weeks.
Participants Number: 288 enrolled and randomized.
Inclusion criteria: age > 18 years, newly diagnosed sputum smear-positive tuberculosis, consent.
Exclusion criteria: pregnancy and lactation, relapse tuberculosis, HIV infection, hypercalcaemia, regu-
lar intake of vitamin D, diabetes, cardiovascular, hepatic and renal diseases and malignancy, suspicion
of prolonged drug abuse.
• Nutritional status: mean weight (kg): males: 47.9 vitamin D group versus 46.3 placebo group; females:
38.4 vitamin D group versus 39.8 placebo group.
Collaboration.
Informed decisions.
Mily 2015 BGD (Continued)

• HIV status: excluded.
• Micronutrient status: mean plasma 25(OH)D3 level: 28.0 vitamin D group versus 28.1 placebo group.
Interventions Group 1: vitamin D 5000 IU daily for 8 weeks.
Group 2: placebo.
Not included in the review:
Group 3: 4-phenylbutyrate 500 mg twice daily.
Group 4: vitamin D plus 4-phenylbutyrate.
Outcomes • Culture conversion at 4 weeks.

• Time to sputum conversion.
• Vitamin D levels.
• Weight gain.
• Adverse events.
Not included in this review: Cough remission, CXR changes, normalization of fever, immunological mea-
sures.
Notes Location: Dhaka, Bangladesh.
Setting: National Institute of the Diseases of the Chest Hospital (NIDCH).
Source of funding: International Centre for Diarrheal Disease Research, Bangladesh, Sida Agreement
support Grant 384, and Swedish Strategic Foundation, and the Swedish Heart-Lung Foundation.
Risk of bias
Random sequence genera- Low risk "computer-generated randomization sequence".

Allocation concealment Low risk "Independent assistants from the Hospital pharmacy prepared the study med-
(selection bias) ication packs (PBA and placebo tablets; with identical appearance, colour and
taste), and labelled these tablets with a randomization number".
Blinding (performance Low risk "placebo with identical appearance, colour and taste".
All outcomes
Incomplete outcome data Low risk Losses to follow-up in the first 8 weeks were low.
(attrition bias)
All outcomes
Collaboration.
Informed decisions.
Morcos 1998 EGY

Standard care: all participants received standard anti-tuberculous therapy: HRS for 2 months (beyond
2 months therapy is not described).
Participants Number: 24 enrolled; outcomes presented for 24.
Inclusion criteria: children with active tuberculosis; pulmonary or extrapulmonary.
Exclusion criteria: none stated.

• Vitamin D1: the mean vitamin D level at baseline is not given for the separate groups.
Interventions Group 1: usual treatment plus vitamin D (1000 IU/day) for 8 weeks.
Group 2: usual treatment.
Outcomes • Body weight before and after treatment.

• Vitamin D levels.
Notes Location: Cairo, Egypt.
Setting: inpatients and outpatients at a children's hospital.
Funding: not stated.
Risk of bias
Random sequence genera- Unclear risk "randomly divided into 2 groups". The trial authors did not provide any further
tion (selection bias) details.
Allocation concealment Unclear risk The trial authors did not describe any allocation concealment.
(selection bias)
Blinding (performance High risk The trial authors did not describe any blinding and did not use a placebo.
All outcomes
Incomplete outcome data Low risk No losses to follow-up: 24/24 (100%) analysed.
(attrition bias)
All outcomes
Selective reporting (re- Unclear risk We were unable to retrieve the trial protocol.
porting bias)
Other bias High risk The trial authors did not state important baseline characteristics.
Collaboration.
Informed decisions.
Nursyam 2006 IDN

Study dates and duration: January 2001 to August 2001, follow-up 8 weeks.
Standard care: all participants received antituberculous therapy in accordance with DOTS program,
first category: 2RHZE/4RH.
Inclusion criteria: adults aged 15 to 59 with sputum-culture positive pulmonary tuberculosis and mod-
erately advanced lesion.
Exclusion criteria: corticosteroids or immunosuppressive treatment, AIDS, renal failure, diabetes melli-
tus, liver cirrhosis, measles, malignancies, leprosy, or severe nutritional deficiency.
• Nutritional status: mean BMI (SD): 16.87 kg/m2 (2.06) treatment group versus 17.68 kg/m2 (2.54) place-
bo group.
• HIV status: negative.
• Described as: 67.2% had low income, 71.6% were low in nutritional status.
• Vitamin D status at baseline not described.
Interventions Group 1: vitamin D (0.25 mg/day for the first 6 weeks).
Group 2: placebo.
Outcomes • Sputum-culture positive at 6 weeks.

• BMI at 0 and 6 weeks.
Outcomes not included in this review: x-ray improvement at 6 weeks, default rate.
Notes Location: Jakarta, Indonesia.
Setting: outpatients at a pulmonary clinic.
Risk of bias
Random sequence genera- Unclear risk The trial authors described it as "randomised"; but gave no further details.
Allocation concealment Unclear risk The trial authors did not describe allocation concealment.
(selection bias)
Blinding (performance Low risk The trial authors described the trial as "double blind" and stated "the placebo
bias and detection bias) were manufactured in the same shape and size". They did no provide any fur-
All outcomes ther details.
Incomplete outcome data Low risk There were no losses to follow-up. The trial authors analysed 67/67 (100%) for
(attrition bias) sputum conversion at 6 weeks.
All outcomes
Collaboration.
Informed decisions.
Pakasi 2010 IDN

Study dates and duration: January 2004 to December 2005, follow-up 6 months.
Standard care: all participants received antituberculous therapy in accordance with DOTS program;
2RHZE (daily)/4HR (thrice weekly), given by a treatment partner who was paid if the participant suc-
cessfully completed treatment.
Participants Number: 300 enrolled; 76 zinc; 72 vitamin A; 66 vitamin A + zinc; 86 placebo - 255 completed 6 months
and were analysed.
Inclusion criteria: adults aged 15 to 55, newly diagnosed sputum AFB positive pulmonary tuberculosis.
Exclusion criteria: pregnancy, lactation, underlying chronic, or degenerative diseases.
• Nutritional status: mean BMI (SD): 16.5 kg/m2 (2.2) Zinc; 16.5 kg/m2 (2.2) Vit A; 16.6 kg/m2 (2.1) zinc +
vitamin A, and 16.4 kg/m2 (2.5) in the placebo group.
• Vitamin A status: median plasma retinol µmol/L (IQR): 0.75 (0.5 to 1.0) zinc; 0.7(0.5 to 1.5) vitamin A;
0.7(0.4 to 1.1) zinc + vitamin A and 0.7(0.5 to 1.0) placebo group.
• Zinc status: mean plasma zinc µmol/L (SD): 11.6(2.2) zinc; 11.9(3.0) vitamin A; 12.1(3.0), zinc + vitamin
A and 11.8(2.4) placebo group.
Interventions Zinc group: 15 mg zinc sulphate daily for 6 months as a capsule.
Vitamin A group: 5000 IU (1500 retinol equivalents) daily as a capsule.
Zinc + vitamin A group: both the above capsules for 6 months.
Placebo group: lactose capsule daily for 6 months.
Outcomes • Sputum smear conversion weekly for 8 weeks.

• Anthropometry: weight, BMI, skin folds, percentage body fat.
• Blood tests: plasma zinc, vItamin A.
Not included in this review: chest X-ray cavity size, C-reactive protein, ESR, Hemoglobin, leukocyte
count, and serum albumin.
Notes Location: Nusa Tenggara Timur Province, Indonesia.
Setting: community-based trial.
Funding: Canadian International Development Agency through World Vision International Indonesi-
a:Food integrated to Hinder TB Project (WVI-FIGHT Project).
Risk of bias
Random sequence genera- Low risk "randomisation was done using a computer program".
Allocation concealment Unclear risk "a treatment code was given to each subject". The trial authors did not provide
(selection bias) any further details.
Collaboration.
Informed decisions.
Pakasi 2010 IDN (Continued)
Blinding (performance Low risk "All capsules were identical in shape ,colour and size".
All outcomes
Incomplete outcome data High risk There was greater than 10% loss to follow-up in each group.
(attrition bias)
All outcomes
Paliliewu 2013 IDN

Study dates and duration: not described.
Standard care: short course directly observed antibiotic therapy. Intensive 60-day treatment with isoni-
azid (300 mg/day), rifampicin (600 mg/day), pyrazinamide (1600 mg/day), and ethambutol (1200 mg/
day) followed by a sustained 45-dose therapeutic phase with isoniazid (800 mg/dose) and rifampicin
(600 mg/dose).
Inclusion criteria: 14 to 50 years, ≥ 2 sputum smear positive, minimal-medium radiological lesion, 16

to 23 mg/kg2 (BMI), ≥ 2.5 to 4.5 g/dL albumin levels and no prior history of tuberculosis or tuberculosis
treatment.
Exclusion criteria: pregnant, breast feeding, used corticosteroids, had HIV, diabetes or another serious
co morbidity.

• MDR/XDR-TB: not stated.
• Micronutrient levels not stated.
• Sputum positivity: 18 (100%) in intervention group, 18 (100%) in placebo group.
Interventions Channa striata group: C. striata capsules, 4 months supplementation with 2 g/day, 3 times/day of C.
striata for 12 weeks.
Placebo group: organoleptically matched placebo with identical dosage regime.
Outcomes • Percentage sputum conversion.

• Disease signs & symptoms – cough, haemoptysis, dyspnoea, fever, night sweats, fatigue.
• Laboratory indicators- AST, ALT, creatinine, uric acid, albumin.
• TNF-α, IFN-γ, and IL-10 levels.
Notes Location: Indonesia.
Setting: clinic, patients attending Department of Internal Medicine, University of Sam Ratulangi, Mana-
do, North Sulawesi, Indonesia.
Funding: research grant from PT. Royal Medicalink Pharmalab, Makassar South Sulawessi, Indonesia.
Collaboration.
Informed decisions.
Paliliewu 2013 IDN (Continued)
Risk of bias
Random sequence genera- Unclear risk The trial authors did not describe how the random sequence was generated.
Allocation concealment Unclear risk The trial authors did not explicitly describe this process.
(selection bias)
Blinding (performance Unclear risk Participants were blinded to the treatment allocation as organoleptically iden-
bias and detection bias) tical capsules were provided to both groups at similar times. Blinding of study
All outcomes personnel and outcome assessors was not described.
Incomplete outcome data Unclear risk The trial authors did not provide any information on the number of partici-
(attrition bias) pants lost to follow-up.
All outcomes
Selective reporting (re- Unclear risk We were unable to obtain a copy of the trial protocol so it is unclear if the au-
porting bias) thors presented data on all of the stipulated outcomes.
Other bias Low risk Baseline comparability: figure in paper provides information on the sex, clin-
ical signs, and symptoms and lab indicators at baseline. There was no differ-
ence in any of these variables at baseline.
Funding source: research grant from PT. Royal Medicalink Pharmalab.
Conflict of interest: not stated.
Paton 2004 SGP

Study dates and duration: November 2000 to July 2002. Follow-up 24 weeks.
Standard care: all participants received combination antituberculous drug treatment, ancillary care,
and follow-up according to standard protocols of the Tuberculosis Control Unit.
Inclusion criteria: adults aged 18 to 69 years, with pulmonary tuberculosis and a body mass index < 20
kg/m2.
Exclusion criteria: diabetes or other severe underlying disease, concomitant corticosteroid or immuno-
suppressive therapy, HIV-positive or considered to be at risk of HIV and refused testing, history of non-
compliance to tuberculosis therapy, unable to tolerate conventional regimen, required inpatient care.
• Nutritional status: mean BMI (SD): 16.7 kg/m2 (1.5) supplement group versus 17.9 kg/m2 (1.9) control
group.
• Description: no comment on the social economic status of participants.
Interventions Group 1: a target energy intake was calculated for each participant and advice given to them on how to
reach this target based on a 24-hour food diary; participants were also supplied with high-energy oral
nutritional supplements (6.25 g protein, 20.2 g carbohydrate, 4.29 g fat, 150 kcal/100 mL) and advised
Collaboration.
Informed decisions.
Paton 2004 SGP (Continued)

to consume 2 packets/day between meals (600 kcal total), which increased to 3 packets/day if tolerat-
ed, until they reached a body mass index of 20 or usual body weight.
Group 2: participants were advised to increase their food intake and given advice to address any imbal-
ances in their diet based on a 24-hour food diary.
Outcomes • Body weight, total lean mass, and total fat mass.
• Change in maximum grips strength and time stands test.
• Change in quality of life score.
Not included in the review: total energy intake; total energy intake from normal diet.
Outcomes measured at 6, 12, and 24 weeks.
Notes Location: Singapore.
Setting: outpatients attending a tuberculosis control unit.
Funding: National Medical Research Council of Singapore, Ensure donated by Abbott Laboratories.
Risk of bias
Random sequence genera- Low risk "The randomisation was 1:1 for the 2 groups and was performed by random-
tion (selection bias) ly shuffling opaque envelopes containing study codes. Preparation of the ran-
domisation envelopes was performed by a member of the staff who was not
directly involved in the study".
Allocation concealment Unclear risk See above. The trial authors did not report any further details.
(selection bias)
Blinding (performance High risk None described. Given the nature of the intervention, only outcome assessors
bias and detection bias) could reasonably have been blinded but the trial authors did not describe this.
All outcomes
Incomplete outcome data High risk Losses to follow-up were low at 6 weeks (3%), but high at 24 weeks: 21% in the
(attrition bias) supplement group versus 35% in the advice only group.
All outcomes
Praygod 2011a TZA

Study dates and duration: April 2006 to March 2009, follow-up 5 months.
Standard care: all participants received standardized antituberculous drug treatment for 6 to 8 months
according to national guidelines: 2RHZE/6HE.
Inclusion criteria: adults aged > 15 years, with new or relapse pulmonary tuberculosis.
Collaboration.
Informed decisions.
Praygod 2011a TZA (Continued)

Exclusion criteria: extra-pulmonary tuberculosis, terminal illness, pregnancy, sputum-positive patients
with HIV co-infection.
Note: the trial included patients with sputum-negative pulmonary tuberculosis and HIV co-infection.
• Nutritional status: mean BMI (SD): 18.9 kg/m2 (2.8) intervention group versus 18.9 kg/m2 (3.1) control
group.
• HIV status: 27% HIV-positive intervention group versus 29% control group.
• Micronutrient status: at baseline not reported.
Interventions Group 1: the intervention group received a daily energy-protein similar to the control biscuit but with
additional: 1.5 mg vitamin A, 20 mg thiamin, 20 mg riboflavin, 25 mg vitamin B6, 50 µg vitamin B12, 0.8
mg folic acid, 40 mg niacin, 200 mg vitamin C, 60 mg vitamin E, 5 µg vitamin D, 0.2 mg selenium, 5 mg
copper, 30 mg zinc.
Group 2: the control group received a daily energy-protein biscuit for the first 60 days of treatment.
Composition: 4.5 g protein, 615 kJ energy, 120 mg phosphorous, 120 mg calcium, 36 mg magnesium, 70
mg sodium, 150 mg potassium, and traces < 1 mg of iron and zinc
Outcomes • Body weight, arm fat area, arm muscle area at 0, 2, and 5 months.
• Maximum hand grips strength at 0, 2, and 5 months.
Setting: 4 tuberculosis clinics serving urban and suburban patients.
Funding: the Danish Council for Independant Research, Danida and the University of Copenhagen.
Clinical trial registry number: NCT00311298.
Risk of bias
Random sequence genera- Low risk "Allocation to treatment arms followed a computer-generated randomisation
tion (selection bias) sequence using permuted blocks of ten".
Allocation concealment Low risk "The random sequence was used by a designated research staff member who
(selection bias) was not involved in any clinical work in TB clinics to arrange and label the sup-
plement packs with identity numbers ranging from 1 to 1500. During the study,
the randomisation sequence and code were kept in a safe cabinet and were ac-
cessible only to the research staff. Recruitment of study participants was done
by clinic staff. Then the same designated research staff assigned the recruit-
ed patient an identity number and sent the corresponding nutritional supple-
ment pack to the respective clinic".
Blinding (performance Low risk "Both control and experimental supplements were of the same size and colour
bias and detection bias) but had slightly different tastes and were wrapped in grey-coloured paper box
All outcomes with 6 bars each".
Incomplete outcome data Low risk Loss to follow-up at 2 months and 5 months were 10.2% and 18.0% respec-
(attrition bias) tively. However, the proportion lost to follow-up did not differ significantly be-
All outcomes tween the groups.
Collaboration.
Informed decisions.
Praygod 2011a TZA (Continued)
Selective reporting (re- Unclear risk Outcomes reported in the publication match those set out in the protocol.
porting bias) However, mortality, a secondary outcome was not reported on.
Other bias Low risk The baseline characteristics of the 2 arms were comparable.
Funding: Danish council for independent research - Medical sciences; Danida

through Consultative Research Committee for Development Research; Univer-
sity of Copenhagen.
Conflict of interest: all trial authors reported no conflicts of interest.
Praygod 2011b TZA

Study dates and duration: April 2006 to March 2009, follow-up 5 months.
Standard care: all participants received standardized antituberculous drug treatment for 6 to 8 months
according to national guidelines: 2RHZE/6HE.
Participants Number: 377 randomized.
Inclusion criteria: adults aged > 15 years, with new or relapse sputum-positive pulmonary tuberculosis
with HIV co-infection.
Exclusion criteria: extra-pulmonary tuberculosis, terminal illness, pregnancy.
• Nutritional status: mean BMI (SD): 18.7 kg/m2 (2.9) intervention group versus 18.5 kg/m2 (2.8) control
group.
• HIV status: all HIV positive.
Interventions Group 1: the intervention group received 6 daily energy-protein biscuits for the first 60 days of treat-
ment, of which 1 contained the additional micronutrients.
Group 2: the control group received 1 daily energy-protein biscuit with additional micronutrients.
Basic biscuit: composition: 4.5 g protein, 615 kJ energy, 120 mg phosphorous, 120 mg calcium, 36 mg
magnesium, 70 mg sodium, 150 mg potassium, and traces < 1 mg of iron and zinc.
Biscuit with additional micronutrients: as above plus 1.5 mg vitamin A, 20 mg thiamin, 20 mg riboflavin,
25 mg vitamin B6, 50 µg vitamin B12, 0.8 mg folic acid, 40 mg niacin, 200 mg vitamin C, 60 mg vitamin E,
5 µg vitamin D, 0.2 mg selenium, 5 mg copper, 30 mg zinc.
Outcomes • Body weight, arm fat area, arm muscle area at 0, 2, and 5 months.
• Maximum hand grips strength at 0, 2, and 5 months.
Setting: 4 tuberculosis clinics serving urban and suburban patients.
Funding: The Danish Council for Independant Research, Danida and the University of Copenhagen.
Clinical trial registry number: NCT00311298.
Risk of bias
Collaboration.
Informed decisions.
Praygod 2011b TZA (Continued)
Random sequence genera- Low risk "Computer-generated randomisation sequence, using permuted blocks of
tion (selection bias) ten".
Allocation concealment Low risk "The allocation sequence was used by designated research staff to sequential-
(selection bias) ly arrange and label supplement packs with identity numbers ranging from 1
to 500. During the study the randomisation sequence and code were kept in
a safe cabinet only accessible by designated research staff. Recruitment was
done by clinic staff. The same designated research staff not employed at the
study clinics assigned an identity number to the recruited patient and sent the
corresponding nutritional pack to the respective clinic".
Blinding (performance High risk Blinding was not possible due to the nature of the intervention.
All outcomes
Incomplete outcome data Low risk At 2 and 5 months, 12.2% and 19.1% of participants were lost to follow-up.
(attrition bias) However, the proportions lost to follow-up were similar across the interven-
All outcomes tion and control arms of the study.
Selective reporting (re- Low risk Outcomes reported in the publication match those set out in the protocol.
porting bias) However, mortality, a secondary outcome was not reported on.
Other bias Low risk The baseline characteristics of the 2 arms were comparable.
Funding: Danish council for Independent Research - Medical sciences; Danida

through Consultative Research Committee for Development Research; Univer-
sity of Copenhagen.
Conflict of interest: all trial authors reported no conflicts of interest.
Pérez-Guzmán 2005 MEX

Study dates and duration: March 2001 to January 2002, follow-up 8 weeks.
Standard care: all participants received a short-course regimen with 4 antituberculous drugs under the
DOTS strategy according to WHO guidelines.
Inclusion criteria: adults aged 17 to 60 years with sputum-culture positive pulmonary tuberculosis.
Exclusion criteria: diabetes mellitus; HIV-positive; signs of coronary heart disease.
• Nutritional status: mean BMI (SEM): 19.9 kg/m2 (1.2) intervention group versus 19.2 kg/m2 (0.8) normal
diet group.
• HIV status: excluded.
• MDR/XDR-TB: not reported
Interventions Group 1: normal diet containing 2500 calories a day, 16% protein, 54% carbohydrate, and 30% lipids for
8 weeks; included 250 mg cholesterol/day.
Collaboration.
Informed decisions.
Pérez-Guzmán 2005 MEX (Continued)

Group 2: high cholesterol diet containing 2500 kcal per day, 16% protein, 54% carbohydrate, and 30%
lipids for 8 weeks; included 850 mg cholesterol/day.
Outcomes • Sputum-culture positive at end of weeks 2, 4, and 8.
Not included in the review: mean number of colony-forming units and acid-fast bacilli in sputum smear;
self reported severity of cough, sputum production, and dyspnoea.
Outcomes measured at weeks 1, 2, 3, 4, 5, 6, 7, and 8.
Notes Location: Mexico City, Mexico.
Setting: hospital inpatients.
Funding: none declared.
Risk of bias
Random sequence genera- Unclear risk The trial authors described the trial as "randomised"; but did not provide any
tion (selection bias) further details.
Allocation concealment Unclear risk The trial authors did not describe allocation concealment.
(selection bias)
Blinding (performance Unclear risk "Patients were unaware of the group to which they had been assigned".
All outcomes
Incomplete outcome data Low risk No loss to follow-up. 21/21 (100%).

(attrition bias)
All outcomes
Ralph 2013 IDN

Study dates and duration: June 2008 to February 2010.
Standard care: participants received directly observed antituberculous therapy: weight-dosed ri-
fampicin, isoniazid, pyrazinamide, ethambutol daily for 2 months; then rifampicin, isoniazid 3 times a
week for 4 months.
Participants Number: 548 assessed for eligibility; 200 randomized (45% of planned sample size); 155 analysed.
Inclusion criteria: consecutive patients referred to Timika tuberculosis clinic with newly diagnosed with
pulmonary tuberculosis (≥ 2 direct sputum specimens positive for acid fast bacilli (AFB)) were assessed
for eligibility: sputum smear positive, > 15 years, not pregnant, without hypercalcaemia (ionized calci-
um ≤ 1.32 mmol/L), not previously treated for tuberculosis, agreeing to stay in Timika for 6 months and
providing written informed consent.
Exclusion criteria: not specified.
Collaboration.
Informed decisions.
Ralph 2013 IDN (Continued)

• Nutritional status: arginine + vitamin D group 49 ± 8.9 kg; arginine group 47.9 ± 9 kg; vitamin D group
48.1 ± 6.6 kg; placebo group 48.7 ± 5.5 kg.
• HIV status: arginine + vitamin D group 28% HIV-positive; arginine group 6% HIV-positive; vitamin D
group 15% HIV-positive; placebo group 5% HIV-positive.
• MDR/XDR-TB: not reported.
• Vitamin D levels: not reported.
• Arginine levels: not reported.
Difference in sex, HIV status, and X-ray severity at baseline.
Interventions L-Arginine + vitamin D group: active L-arginine (L-arginine hydrochloride, Argimax®) 6 g (6 tablets) daily
for 8 weeks and active cholecalciferol (vitamin D3, Calciferol Strong®) 50,000 IU (1250 mcg, 1 tablet) at
baseline and on day 28.
L-Arginine + placebo vitamin D group: active L-arginine (L-arginine hydrochloride, Argimax®) 6 g (6

tablets) daily for 8 weeks and placebo cholecalciferol: 1 tablet at baseline, 1 tablet at 28 days.
Placebo L-arginine + vitamin D group: placebo L-arginine: 6 tablets daily for 8 weeks and active chole-
calciferol (vitamin D3, Calciferol Strong®) 50,000 IU (1250 mcg, 1 tablet) at baseline and on day 28.
Placebo group: placebo L-arginine: 6 tablets daily for 8 weeks and placebo cholecalciferol: 1 tablet at
baseline, 1 tablet at 28 days.
Outcomes • Proportion of participants with negative sputum culture on liquid medium at 4 weeks.
• Composite clinical severity score at week 8 (points allocated on weight change, forced expiratory value
(FEV1), cough and presence/absence of sputum and haemoptysis.
• Safety (death, hospitalization, hypercalcaemia).
• Sputum smear conversion time (≥ 2 consecutive negative smears without a subsequent positive).
• Change in 6 minute walk test.
• Modified St George’s respiratory questionnaire.
• Chest X-ray severity score (0, 8, and 24 weeks).
• FEV1.
• Primary endpoint stratified by HIV and ethnicity.
• Serious adverse events: death, hospitalization, and life-threatening conditions.
• Adverse events: new symptoms or hypercalcaemia.
Notes Location: Timika, Indonesia.
Setting: Timika tuberculosis clinic and community hospital.
Funding: Australian Respiratory Council, Royal Australasian College of Physicians Covance Award
(APR), National Health and Medical Research Council of Australia.
Trial registration: NCT00677339.
Risk of bias
Random sequence genera- Low risk Block random allocation sequence stratified by ethnicity (Papuan/non-
tion (selection bias) Papuan) was generated and remained concealed from all investigators
throughout study. Independent assistants prepared packs labelling them with
a code corresponding to the randomization sequence.
Collaboration.
Informed decisions.
Ralph 2013 IDN (Continued)
Allocation concealment Low risk Participants were assigned the next sequential code and dispensed an opaque
(selection bias) envelope containing study medication. Active and placebo medications ap-
peared identical.
Blinding (performance Low risk Randomization sequence was unknown to all investigators. Independent as-
bias and detection bias) sistants labelled medication packs with codes corresponding to random se-
All outcomes quence. Active and placebo medications appeared identical.
Incomplete outcome data Low risk The trial authors stated that they conducted a modified intention-to-treat
(attrition bias) analysis. They kept participants in the arm to which they were randomized,
All outcomes but excluded protocol violators and participants lost to follow-up from the fi-
nal analysis.
Loss to follow-up per arm at end of 4 weeks (primary outcomes time point)
• L-arginine + vitamin D: 4/50 (8%).

• L-arginine: 3/50 (6%).
• Vitamin D: 3/50 (6%).
• Placebo: 2/50 (4%).
(Loss to follow-up < 10%).
Other bias High risk Baseline comparability: there were differences in sex, HIV status, and X-ray
severity between groups at baseline.
Funding sources: non-commercial funding sources.
Conflict of interest: the trial authors stated there were no competing interests.
Other: grouping the data could lead to confounders.
Range 2005 TZA

Methods Study design: RCT with 2 x 2 factorial design.
Study dates and duration: August 2001 to July 2002, follow-up 7 months.
Standard care: all participants received antituberculous therapy according to WHO guidelines.
Participants Number: 530 enrolled; 31 later found ineligible and excluded; number with available outcomes data
varied by outcome.
Inclusion criteria: adults aged > 15 years, sputum-culture or sputum-smear positive pulmonary tuber-
culosis.
Exclusion criteria: returning to treatment after default or previous treatment failure; thought unlikely
to survive; pregnant or lactating.
• Nutritional status: mean BMI (SD): 18.3 kg/m2 (2.5) zinc + multivitamins group versus 18.0 kg/m2 (2.5)
multivitamin + placebo group versus 17.8 kg/m2 (2.5) zinc+placebo group versus 18.7 kg/m2 (2.7)
placebo group.
• HIV status: 39% zinc + multivitamins group versus 46% multivitamin + placebo group versus 47% zinc
+ placebo group versus 38% placebo group.
Collaboration.
Informed decisions.
Range 2005 TZA (Continued)

• Zinc or micronutrient status at baseline not given.
Interventions Factorial design, giving daily supplements or placebo for 8 months.
Group 1: zinc 45 mg plus placebo.
Group 2: multivitamin and mineral tablet (vitamin A 1.5 mg, vitamin B1 20 mg, vitamin B2 20 mg, vita-
min B6 25 mg, vitamin B12 50 μg, folic acid 0.8 mg, niacin 40 mg, vitamin C 200 mg, vitamin D3 5 μg, vit-
amin E 60 mg, selenium 0.2 mg, and copper 5 mg) plus placebo.
Group 3: zinc 4 mg plus multivitamin and mineral tablet (as above).
Group 4: placebo plus placebo.
Outcomes • Death before 8 months.

• Sputum positive at weeks 2, 4, and 8.
• Weight gain at 8 weeks and 7 months.
Not included in the review: HIV viral load and CD4 count at baseline and 8 weeks.
Notes Location: Mwanza region, Tanzania.
Setting: outpatients at 5 health facilities.
Risk of bias
Random sequence genera- Low risk "Before recruitment of patients, all the letters on the containers were replaced
tion (selection bias) with study serial numbers from 1 to 550 based on the computer-generated
random sequences, using permuted blocks of four".
Allocation concealment Unclear risk "The codes for the MVM and Zn tablets remained in a sealed envelope, and
(selection bias) were only broken after completion of the initial data analysis".
Blinding (performance Low risk "Placebo tablets were identical in colour, shape and size to the corresponding
bias and detection bias) white Zn and green MVM tablets".
All outcomes
Incomplete outcome data High risk Loss to follow-up: mortality: 9% zinc group versus 14% placebo group, weight
(attrition bias) gain: 22% zinc group versus 22% placebo group. These could significantly alter
All outcomes the result.
Schön 2003 ETH

Study dates and duration: recruited December 2000 to December 2001. Follow-up 8 weeks.
Standard care: all participants received DOTS in line with Ethopian National Guidelines.
Participants Number: 120 enrolled, outcomes presented for 115.
Collaboration.
Informed decisions.
Schön 2003 ETH (Continued)

Inclusion criteria: adults aged 15 to 60 years, smear positive pulmonary tuberculosis.
Exclusion criteria: admitted to hospital; pregnant; signs of any concomitant disease other than HIV; or
lived too far away to take part in the directly observed therapy short Direct Observed Treatment Short-
Course (DOTS) programme.
• Nutritional status: Mean weight in HIV-negative group: 47.8 kg intervention group versus 45.3 kg con-
trol group, mean weight in HIV-positive group: 45.0 kg intervention group versus 45.3 kg control group.
• HIV status: 52% HIV-positive.
Interventions Group 1: arginine capsules (1 g arginine) daily for 4 weeks.
Group 2: placebo.
Outcomes • Death during 8 weeks follow-up.

• Sputum positive at week 8.
• Cough at weeks 2 and 8.
• Weight gain at 1, 2, 4, and 8 weeks.
Not included in the review: serum arginine, citrulline, and nitric oxide metabolite levels at weeks 0, 2,
and 8.
Notes Location: Ethiopia.
Setting: Outpatient DOTS programme.
Risk of bias
Random sequence genera- Low risk "Randomized in blocks of six (performed by the state pharmacy of Sweden)".
Allocation concealment Low risk "The study was double blinded and a sealed copy of the treatment code was
(selection bias) kept by the project leader until all data had been collected and analysed".
Blinding (performance Low risk Described as "double blind" and placebos were used: the trial authors did not
bias and detection bias) give any further details.
All outcomes
Incomplete outcome data Low risk There were 115/120 included in the final analysis(96%). Three of these died, 2
(attrition bias) moved out of the study area.
All outcomes
Other bias Low risk We did not identify any sources of bias.
Schön 2011 ETH

Study dates and duration: recruited February 2004 to December 2006. Follow-up until August 2007.
Collaboration.
Informed decisions.

Standard care: all participants received DOTS in line with Ethopian National Guidelines.
Participants Number randomized: 80.
Inclusion criteria: newly diagnosed sputum smear positive older than 15 years.
Exclusion criteria: admitted to hospital; peanut allergy, pregnant; signs of any concomitant disease
other than HIV; previous treatment for tuberculosis.
• Nutritional status: high arginine group - mean BMI: 16.9 95% CI 16.4 to 17.4; low arginine group - mean
BMI 16.4 95%CI 16 to 16.8.
• HIV status: high arginine group - 37% HIV-positive; low arginine group - 32% HIV-positive.
Interventions All participants received antituberculous therapy.
High arginine group: 30 g peanuts (1 g arginine, 750 kJ) daily for 4 weeks.
Low arginine group: 30 g wheat cracker (0.1 g arginine, 623 kJ) daily for 4 weeks.
Outcomes Primary outcome: cure rate (smear positive at start of Rx, completed Rx, and smear negative at end of
Rx and 1 previous occasion).
Secondary outcomes: sputum smear conversion, weight gain, sedimentation rate, reduction of cough,
and chest X-ray improvement at 2 months.
Not eligible for review: levels of nitric oxide (NO) metabolites in urine and NO in exhaled air at 2 weeks
and 2 months.
Notes Location: Gondar, Ethiopia.
Setting: DOTS clinic at Gondar University Hospital.
Funding: Swedish Heart and Lung Foundation.
Trial number: NCT00857402.
Risk of bias
Random sequence genera- Low risk The Department of Epidemiology at the Karolinska Institute, Sweden per-
tion (selection bias) formed randomization in blocks of 6.
Allocation concealment Low risk The randomization code was concealed in 180 sealed individual envelopes.
(selection bias) The envelopes were only opened when the person was enrolled in the study.
Blinding (performance High risk Blinding was not possible due to the nature of the intervention (wheat cracker
bias and detection bias) versus peanuts).
All outcomes
Incomplete outcome data Low risk Regarding loss to follow-up, 7/180 (4%) did not complete the supplement.
(attrition bias)
All outcomes
Selective reporting (re- Low risk Outcomes reported on in the published study match those set out in the study
porting bias) protocol.
Other bias Low risk There were no differences in baseline characteristics between the 2 groups.
Collaboration.
Informed decisions.

Funding: Swedish Heart and Lung Foundation, The Swedish SAREC/SIDA Foun-
dation and the Swedish Research Council.
Conflict of interest: the trial authors had no conflicts of interest.
Semba 2007 MWI

Study dates and duration: July 1999 to February 2005, follow-up 24 months.
Standard care: all participants received standard antituberculous therapy as recommended by the
Malawi NTP: 2RHZE/6HE.
Participants Number: 1148 enrolled; number with available outcomes data varied by outcome.
Inclusion criteria: adults aged 18 to 60 years, sputum positive pulmonary tuberculosis.
Exclusion criteria: planning to move away from the study area within the next 2 years; already taking vi-
tamin supplements; already being treated for tuberculosis; previously treated for tuberculosis.
• Nutritional status: geometric mean BMI (SD): HIV-positive participants 18.3 kg/m2 (2.5) supplement
group versus 18.1 kg/m2 (2.5) placebo group; HIV-negative participants 18.5 kg/m2 (2.6) supplement
group versus 18.8 kg/m2 (2.9) placebo group.
• HIV status: 71% HIV-positive supplement group versus 74% placebo group.
Interventions Group 1: daily micronutrient supplementation for 24 months; supplement consisted of vitamins A (8000
IU), C (500 mg), D (400 IU), E (200 IU), B6 (2 mg), B12 (6 μg), riboflavin (1.7 mg), thiamin (1.5 mg), niacin
(20 mg), folate (0.4 mg), zinc (10 mg), iodine (175 μg), and selenium (65 μg).
Group 2: daily placebo.
Outcomes • Death after 8 months follow-up.

• Vitamin A, vitamin E, and selenium levels at 8 months.
Notes Location: Zomba and Blantyre, Malawi.
Setting: participants diagnosed at a hospital clinic and treated at a community clinic near where they
lived.
Funding: the National Institutes of Health; the Fogarty International Centre.
Risk of bias
Random sequence genera- Low risk "A computer random number generator was used to generate a random allo-
tion (selection bias) cation schedule in permuted blocks of 10".
Allocation concealment Low risk "Treatment assignment was concealed by pre-packing study supplements in
(selection bias) sequentially numbered series according to the allocation schedule".
Blinding (performance Low risk "Both the study staff and the participants were blinded to the treatment as-
bias and detection bias) signment".
All outcomes
Collaboration.
Informed decisions.
Semba 2007 MWI (Continued)

"The placebo and active supplements were of identical appearance in size,
shape and colour. The placebo and active supplements were packed in identi-
cal opaque, white plastic bottles with sealed caps".
Incomplete outcome data High risk Losses to follow-up were high: 13.2% in the supplement group and 12.9% in
(attrition bias) the placebo group.
All outcomes
Other bias Low risk We did not identify any other potential sources of bias identified.
Seyedrezazadeh 2006 IRN

Study dates and duration: recruited April 2003 to May 2004. Follow-up 60 days.
Standard care: all participants received the same antituberculous standard therapy in accordance with
the DOTS strategy.
Participants Number: 42 enrolled; outcomes data available for 37.
Inclusion criteria: adults (age range not stated), with sputum positive tuberculosis.
Exclusion criteria: previous antituberculous treatment; concurrent use of supplements containing sele-
nium and vitamin E; illicit drug addiction; signs of severe effects of antituberculous drug treatment dur-
ing treatment.
• Nutritional status: median BMI (IQR): men 19.6 kg/m2 (16.0 to 27.1) supplements group versus 21.0
kg/m2 (16.8 to 23.1) placebo group; women 19.5 kg/m2 (16.6 to 25.8) supplements group versus 22.0
kg/m2 (17.0 to 26.7) in placebo group.
Interventions Group 1: daily supplements containing vitamin E (140 mg) and selenium (200 μg) for 4 months.
Group 2: placebo.
Outcomes • Sputum positive at days 15, 30, 45, and 60.

• Change in body mass index at 2 months.
Not included in the review: number of lung cavities; cavity surface area; and mean lesion area at 0, 2,
and 6 months.
Notes Location: Tabriz, Iran.
Setting: outpatient clinic at research centre.
Funding: the Tuberculosis and Lung Disease Research Centre.
Risk of bias
Collaboration.
Informed decisions.
Seyedrezazadeh 2006 IRN (Continued)
Random sequence genera- Unclear risk The trial authors described the sequence generation as "randomised", but did
tion (selection bias) not provide any further details.
(selection bias)
Blinding (performance High risk The trial authors described the trial as double blind, and capsules as being
bias and detection bias) "similar in size and red in colour", which makes true blinding unlikely.
All outcomes
Incomplete outcome data Low risk Regarding losses to follow-up, 37/37 (100%) participants were eligible.
(attrition bias)
All outcomes
Singh 2013 IND

Study dates and duration: September to October 2010 (enrolment), 6 months study period (Oct 2010 to
March 2011).
Standard care: initial 2-month phase of antituberculous therapy. The trial authors did not describe the
precise content of the tuberculosis treatment.
Participants Number: 40 enrolled and randomized. 37 included in the final analysis (3 excluded because of poor
compliance).
Inclusion criteria: newly diagnosed sputum smear positive pulmonary tuberculosis patients taking not
more than 7 days of antituberculous therapy aged 18 to 60 years.
Exclusion criteria: known history of drug resistant tuberculosis and renal disease.
• Nutritional status: mean BMI kg/m2 placebo group: 17.7; vitamin D and calcium group: 16.37, vitamin
A and zinc group: 17.52 (no SDs provided).
• HIV status: not reported.
• Micronutrient status at baseline not described.
Interventions All groups received antituberculous therapy.
Control group: antituberculous therapy only.
Vitamin D and calcium group: tablet containing vitamin D and calcium (250 IU of vitamin D3, and 500
mg as calcium carbonate). One tablet per day for first 10 days, then 3 tablets per week for the remain-
der of the 2 months.
Vitaimn A and zinc group: zinc tablet (50 mg elemental zinc as zinc sulphate) and vitamin A tablet
(25,000 IU vitamin A as vitamin A palmitate). One vitamin A tablet and 1 zinc tablet per day for first 10
days, then 3 tablets of each per week for the remainder of the 2 months.
Outcomes Sputum smear conversion rates every 20 days from enrolment.
Collaboration.
Informed decisions.
Singh 2013 IND (Continued)

Time to sputum smear conversion.
Weight gain at enrolment and 60 days (no SDs provided).
BMI at enrolment and 60 days (no SDs provided).
Not relevant to review: neutrophil count at enrolment and 60 days; lymphocyte count at enrolment and
60 days; serum alanine transaminase and aspartate transaminase at enrolment and 60 days; serum
haemoglobin at enrolment and end of 2 months (60 days).
Notes Location: India, HNB hospital attached to the Veer Chandra Singh Garhwali Govt. Medical Science and
Research Institute, Srinagar – Pauri Garhwal (Uttarakhand).
Setting: hospital.
Funding: not reported.
Risk of bias
Random sequence genera- Unclear risk The trial authors did not describe this and simply stated randomly assigned.
Allocation concealment Unclear risk The trial authors did not describe this.
(selection bias)
Blinding (performance Unclear risk The trial authors did not describe blinding of participants, personnel, and out-
bias and detection bias) come assessors.
All outcomes
Incomplete outcome data Low risk There were 40 patients randomized to the 3 groups. The trial excluded 3 pa-
(attrition bias) tients due to non-compliance and excluded them from the analysis (8% loss to
All outcomes follow-up).
Selective reporting (re- Unclear risk We were unable to obtain the study protocol, and therefore there is unclear
porting bias) risk of selective outcome reporting. In the abstract the trial authors stated that
the aim of the study was to assess sputum conversion and blood profiles dur-
ing the initial phase of tuberculosis treatment. They reported on these out-
comes.
Other bias Unclear risk The trial authors did not report baseline characteristics. There was no descrip-
tion of funding sources or conflict of interest statement.
Sudarsanam 2010 IND

Study dates and duration: recruitment January to Nov 2005, follow-up 1 year.
Standard care: All participants received the same antituberculous standard therapy in accordance with
the DOTS strategy.
Participants Number: 103 enrolled, 99 analysed at 6 months, 91 at 1 year.
Inclusion criteria: age > 12 years, either sputum positive tuberculosis or clinical and radiological evi-
dence of pulmonary tuberculosis or biopsy proven extra pulmonary tuberculosis, informed consent
Collaboration.
Informed decisions.
Sudarsanam 2010 IND (Continued)

Exclusion criteria: relapse of previous antituberculous treatment, end stage renal or liver disease, CD4
count > 200 (if HIV-positive), BMI < 19 kg/m2, patients not from Vellore.
• Nutritional status: mean BMI supplement: 17.2 kg/m2; mean BMI no supplement: 18.2 kg/m2.
• HIV status: 20 out of 103 were HIV co-infected.
• All participants belonged to lower socioeconomic strata.
Interventions Group 1: macronutrient and micronutrient supplementation for 6 months. The macronutrient was a
ready-to-serve powder, given as monthly rations to supply 930 kcal and 31.5 g protein per day 3 divid-
ed servings. The micronutrient as a once-a-day multivitamin tablet containing: copper sulphate 0.1 mg,
D-pantheol 1 mg, dibasic calcium phosphate 35 mg, folic acid 500 µg, magnesium oxide 0.15 mg, man-
ganese sulphate 0.01 mg, nicotinamide 25 mg, potassium iodide 0.025 mg, vitamin A 5000 IU, vitamin
B1 2.5 mg, vitamin B12 2.5 µg, vitamin B2 2.5 mg, vitamin B6 2.5 mg, vitamin C 40 mg, vitamin D3 200
IU, vitamin E 7.5 mg, zinc sulphate 50 mg.
Group 2: dietary advice alone.
Outcomes • Death.
• Cure.
• Treatment completion.
• Weight gain.
Not included in the review: adherence.
Notes Location: Vellore, India.
Setting: tuberculosis clinics.
Funding: The Fogarty AIDS International Research and Training Program.
Risk of bias
Random sequence genera- Low risk "a computer generated randomisation code".
Allocation concealment Low risk "Allocation was concealed, the randomisation codes were in opaque en-
(selection bias) velopes opened by the dietician after dietary counselling".
Blinding (performance High risk "There were no attempts made to blind any of the study team or participants".
All outcomes
Incomplete outcome data High risk 3.9% were lost to follow-up at 6 months and 11.7% at 1 year.
(attrition bias)
All outcomes
Collaboration.
Informed decisions.
Tukvadze 2015 GEO

Standard care: all participants received antituberculous therapy (isonizaid, rifampicin, pyrazinamide,
and ethambutol - no further details given), given under direct observation.
Study dates and duration: July 2009 to April 2012. Follow-up 16 weeks.
Participants Number: 199 enrolled and randomized.
Inclusion criteria: age > 18 years, newly diagnosed pulmonary tuberculosis disease with a positive AFB
sputum smear, < 7 days of antituberculous drug therapy before entry, informed consent.
Exclusion criteria: previous tuberculosis, extrapulmonary tuberculosis, pregnancy or lactation, a histo-
ry of hypercalcaemia, nephrolithiasis, hyperparathyroidism, sarcoidosis, organ transplant, hepatic cir-
rhosis, seizures, or cancer in the past 5 years, baseline plasma calcium concentration > 2.6 mmol/L, cre-
atinine concentration > 250 mmol/L, or aspartate aminotransferase concentrations > 3 times the upper
limit of normal, renal replacement therapy, corticosteroid use in the past 30 days, current use of cyto-
toxic or immunosuppressive drugs, known MDR-TB before study enrolment, current imprisonment.
• Nutritional status: BMI < 18.5 kg/m2: 18% vitamin D group versus 28 % placebo group.
• HIV status: 191/199 were tested. 1% HIV positive in the vitamin D group versus 3% HIV positive in the
placebo group.
• MDR/XDR-TB: isoniazid and rifampicin resistance: 12% vitamin D group versus 11% placebo group.
• Micronutrient status: plasma 25(OH)D3 below 30 nmol/L: 92% vitamin D group versus 92% placebo
group.
Interventions Vitamin D group: 50,000 IU vitamin D3 3 times a week for 8 consecutive weeks followed by 50,000 IU
every 2 weeks for a further 8 weeks.
Placebo group: placebo tablet for similar dosing regime.
Outcomes • Time to sputum culture conversion.

• Culture conversion at 8 weeks.
• Vitamin D levels at baseline and every 2 weeks of follow-up.
• Adverse effects.
Not included in this review: calcium level every 4 weeks, multiple subgroup analyses looking at predic-
tors of culture conversion.
Notes Location: Tbilisi, Georgia.
Setting: National Centre for TB and Lung Disease and affiliated clinics.
Source of funding: National Institutes of Health Grant, the Emory University Global Health Institute.
Additional publications: Frediani JK, et al (2015) Macronutrient intake and body composition
changes during antituberculous therapy in adults, Clinical Nutrition, http://dx.doi.org/10.1016/
j.clnu.2015.02.007
Risk of bias
Random sequence genera- Low risk "Treatment assignments were generated by the Emory University–based study
tion (selection bias) biostatistician and implemented locally through a distributed study database
with
Collaboration.
Informed decisions.
Tukvadze 2015 GEO (Continued)

the use of a randomized permuted block algorithm stratified by clinical center
site".
Allocation concealment Low risk "All study medication bottles had a unique bottle number to allow for blinded
(selection bias) dispensing".
Blinding (performance Low risk "The study drugs vitamin D3 and the placebo were identical in shape and
bias and detection bias) colour". The trial authors did not report blinding of outcome assessors.
All outcomes
Incomplete outcome data High risk 20/100 (20%) were lost to follow-up or discontinued intervention in the vita-
(attrition bias) min D group versus 30/99 (30%) in the placebo group.
All outcomes
Villamor 2008 TZA

Study dates and duration: April 2000 to April 2005, follow-up 24 months.
Standard care: all participants received standard antituberculous treatment following the DOTS
scheme: 2RHZE/6HE.
Participants Number: 887 enrolled, number with available outcomes data varied by outcome.
Inclusion criteria: adults aged 18 to 65 years, sputum positive pulmonary tuberculosis.
Exclusion criteria: Karnofsky score < 40%; pregnant; received more than 4 weeks of antituberculous
therapy in the past year; plans to move away from the study area within the next 2 years
Baseline characteristics.
• Nutritional status: mean BMI (SD): 18.9 kg/m2 (2.5) in HIV-negative placebo group versus 18.9 kg/m2
(2.5) in HIV-negative supplement group, 19.6 kg/m2 (2.9) in HIV-positive placebo group versus 19.3 kg/
m2 (2.8) in HIV-positive supplement group.
• HIV status: 53% HIV-positive in supplement group versus 53% in placebo group.
• Description: > 85% had primary school education.
Interventions For 24 months
Group 1: daily oral dose in tablet form of mixed micronutrients containing retinol (5000 IU), vitamin B1
(20 mg), vitamin B2 (20 mg), vitamin B6 (25 mg), niacin (100 mg), vitamin B12 (50 μg), vitamin C (500
mg), vitamin E (200 mg), folic acid (0.8 mg), and selenium (100 μg).
Group 2: placebo.
Outcomes • Sputum-culture positive at 1 month.

• Recurrence of positive culture between 1 and 8 months.
• Mortality within 2 years.
• Nutritional parameters: body weight, mid-upper arm circumference, fat-mass, fat-free mass.
Collaboration.
Informed decisions.
Villamor 2008 TZA (Continued)

Not included in the review: CD4, CD8, CD3 count and HIV viral load, haemoglobin, albumin levels, pe-
ripheral neuropathy, and genital ulcers.
Notes Location: Dar es Salaam, Tanzania.
Setting: 5 outpatient clinics.
Funding: The National Institute of Allergy and Infectious Diseases, US Department of Agriculture.
Risk of bias
Random sequence genera- Low risk "Consenting subjects were randomly assigned in computer-generated permut-
tion (selection bias) ed blocks of 20".
(selection bias)
Blinding (performance Low risk "Active tablets and placebo were indistinguishable in size, taste, and colour.
bias and detection bias) All clinical and research staff were unaware of the subjects’ treatment assign-
All outcomes ment".
Incomplete outcome data Unclear risk There were 628/725 (85%) for sputum-culture positive at 1 month; for mortal-
(attrition bias) ity the trial authors included all participants in the analysis up to the point of
All outcomes loss to follow-up if before end of trial, and presented hazard ratios.
Visser 2011 ZAF

Study dates and duration: May 2005 to August 2008, follow-up 8 weeks.
Standard care: all participants received antituberculous therapy: 2RHZE plus pyridoxine 25 mg/day.
Participants Number: 154 enrolled; 124 completed study.
Inclusion criteria: 18 to 60 years old, 2 positive sputum smears for acid-fast bacilli or 1 positive plus sug-
gestive chest X-ray findings, written informed consent to a HIV test.
Exclusion criteria: previous treatment of tuberculosis, known or suspected multidrug resistant tubercu-
losis, clinical evidence of extrapulmonary tuberculosis or liver disease, renal failure, congestive heart
failure or neoplasm, pregnancy.
• Nutritional status: mean BMI (SD): males; supplement group 18.9 kg/m2 (2.7) versus 19.0 kg/m2 (2)
placebo group, females; supplement group 23.0 kg/m2 (4.3) versus 21.6 kg/m2 (4.8) placebo group.
• HIV status: 9% HIV positive supplement group versus 11% HIV-positive placebo group.
• Median retinol (range): 21.1 µg/dL (15.1 to 27.8) supplement group versus 21.2 (15.7 to 28.9) placebo
group (normal range: > 20 µg/dL).
Collaboration.
Informed decisions.
Visser 2011 ZAF (Continued)

• Median Zinc (range): 62 µg/dL (53 to 71.8) supplement group versus 59 (51.8 to 65.3) placebo group
(normal range: > 70 µg/dL).
Interventions Group 1: a single capsule that contained 200,000 IU vit A (retinyl palmitate) within 24 hours of starting
tuberculosis treatment plus 15 mg zinc (zinc gluconate) daily for 5 days per week for 8 weeks.
Group 2: placebo capsule (sunflower oil) within 24 hours of starting tuberculosis treatment plus place-
bo tablet (starch/gelatin base) daily for 5 days per week for 8 weeks.
Outcomes • Death.
• Sputum-smear positive at baseline and 8 weeks.
• Adverse events.
• Weight gain at 2 months.
• Arm muscle circumference at 2 months.
• Vitamin A and zinc status at 2 and 8 weeks.
Outcomes not included in review: adherence, lung cavities at 2 months, copper, CRP, albumin, haemo-
globin, WBC, and neutrophil levels at 2 and 8 weeks.
Notes Location: Cape Town, South Africa.
Setting: primary care tuberculosis clinics.
Funding: research grants from the National Research Foundation South Africa, the Norwegian Pro-
gramme for Development, Research and Higher Education, the Research Council of Norway, the Na-
tional Research Foundation, the South African Sugar Association, and the Fogarty International Cen-
tre. The South African Department of Health and Pharma Natura Pty donated the vitamin A and place-
bo capsules respectively.
Risk of bias
Random sequence genera- Low risk "computer-generated permuted blocks of 8, generated by an independent epi-
tion (selection bias) demiologist".
Allocation concealment Low risk "Treatment allocations was concealed by prepackaging supplements in se-
(selection bias) quentially numbered packets according to the allocation schedule".
Blinding (performance Low risk "Active and placebo capsules and tablets for both groups were identical in
bias and detection bias) size, shape and colour. All research team members as well as laboratory staff
All outcomes involved in the trial were blinded".
Incomplete outcome data High risk Loss to follow-up was high in both groups: 12/77 (16%) supplement group ver-
(attrition bias) sus 19/77 placebo group (25%).
All outcomes
Other bias Low risk "Because of a change in national policy in April 2008 15 participants received
TB treatment 7 days per week with unsupervised weekend doses".
Comment: this was unlikely to introduce significant bias.
Collaboration.
Informed decisions.
Wejse 2008 GNB

Study dates and duration: November 2003 to December 2005, follow-up 12 months.
Standard care: all participants received antituberculous therapy: 2RHZE/6HE.
Participants Number: 367 enrolled; number with available outcomes data varied by outcome.
Inclusion criteria: adults aged > 15 years, diagnosis of tuberculosis by smear positive (pulmonary tuber-
culosis) or WHO clinical criteria (extrapulmonary tuberculosis).
Exclusion criteria: none stated.
• Nutritional status: mean BMI (range): 18.8 kg/m2 (12 to 33) treatment group versus 18.5 kg/m2 (12 to
27) placebo group.
• HIV status: 39% HIV-positive supplement group versus 33% HIV-positive placebo group.
• Mean serum 25-hydroxyvitamin D3 nmol/L (SD): 77.5 (23.8) supplement group versus 79.1 (21.8) place-
bo group (normal range: > 75 nmol/L).
• Described as a poor urban population.
Interventions Vitamin D group: 3 doses cholecalciferol (100,000 IU) in drinkable ampoules; given at start of treatment,
5 months, and 8 months.
Placebo group: vegetable oil without cholecalciferol.
Outcomes • Death at 12 months in HIV-positive and HIV-negative participants.

• Sputum-smear positive at baseline, 2 weeks, 4 weeks, 6 weeks, 2 months, 5 months, and 8 months.
• Adverse events.
• Weight gain at 8 months.
• Changes in tuberculosis score at 2, 5, and 8 months.
• Vitamin D status at 2 and 8 months.
Outcomes not included in review: change in CD4 count at 8 months.
Notes Location: Guinnea-Bassau.
Setting: 3 health centres and a tuberculosis hospital in a demographic surveillance area.
Risk of bias
Random sequence genera- Low risk "The random allocation sequence was computer generated; a list of continu-
tion (selection bias) ous study numbers was generated with a random allocation to treatment 1 or
2".
Allocation concealment Low risk "Study numbers were consecutive and given to patients by the field assistant
(selection bias) at inclusion, and patients were recorded in a book with pre-written study num-
bers and allocation sequence number 1 or 2", "A physician gave the trial infor-
mation and obtained consent...a trial nurse administered study medicine ac-
cording to sequence number".
Blinding (performance Low risk "Study medicine was provided in identical containers labelled lot 204 (alloca-
bias and detection bias) tion sequence number 1) or lot 205 (allocation sequence number 2)".
All outcomes
"Patients, staff, and researchers assessing outcome were blinded".
Collaboration.
Informed decisions.
Wejse 2008 GNB (Continued)
Incomplete outcome data Unclear risk Imbalanced attrition: 20/187 (11%) supplement group versus 10/180 placebo
(attrition bias) group (6%).
All outcomes
Abbreviations: RCT = randomised controlled trial; AIDS = acquired immune deficiency syndrome; HIV = human immunodeficiency virus;
E = ethambutol; H = isoniazid; R = rifampicin; S = streptomycin; Z = pyrazinamide; CRP = C-reactive protein; CD4/CD8/CD3 = a measure
of immunological function in HIV-positive people; MDR = multidrug-resistant TB; XDR = extensively drug-resistant TB; DOTS = directly
observed therapy short course; TST = tuberculin skin testing; BMI = body mass index; CXR = chest X-ray; ESR = erythrocyte sedimentation
rate; WBC = white blood cell count; 2RHZE/4HE = 2 months of isoniazid, rifampin, prazinamide, and ethambutol followed by 4 months of
ethambutol and isoniazid; 2RHZE/4HR = 2 months of isoniazid, rifampin, prazinamide, and ethambutol followed by 4 months of rifampin
and isoniazid; INH = isoniazid; 2RHZE/6HE = 2 months of isoniazid, rifampin, prazinamide, and ethambutol followed by 6 months of
ethambutol and isoniazid.
Characteristics of excluded studies [ordered by study ID]
Study Reason for exclusion
Denti 2015 Ineligible outcomes: pharmacokinetics.
Dibari 2013 Ineligible intervention. Looks at acceptability of ready to eat foods.
Gwinup 1981 Outcomes were irrelevant to this Cochrane review's objectives (assessed effect of vitamin supple-
mentation on serum calcium levels in people with tuberculosis).
Hasan 2015 Ineligible intervention: propolis.
Kawai 2014 Ineligible outcomes: linked to Villamor 2008 TZA.
Khandelwal 2014 Ineligible study design: cohort study.
Lin 2014 Ineligible intervention; no relevant outcomes.
Lutge 2013 Ineligible intervention; food vouchers.
Martineau 2009 Excluded as evaluates a single dose of vitamin D.
Mbala 1998 Outcomes were not relevant to review's objectives (assessed neurological and neuropsychiatric
symptoms in children being treated for tuberculosis using isoniazid with or without vitamin B6
supplementation).
Narang 1984 Groups were not randomly allocated, participants not being treated for active tuberculosis, and
outcomes were not relevant to review's objectives.
Oluboyede 1978 It was unclear whether the groups were randomized; we were unable to contact the trial authors.
Permatasari 2014 Ineligible intervention: propolis.
Ramakrishnan 1961 Groups were not randomized to different dietary interventions.
Samsidi 2013 Ineligible control/comparison; the control group received milk-based protein supplement.
Collaboration.
Informed decisions.
Study Reason for exclusion
Shi 2001 The study authors assessed only partial parenteral nutrition, which we excluded from this
Cochrane review.
Srivastava 2011 The study only provided vitamin D for 5 days.
Characteristics of studies awaiting assessment [ordered by study ID]
Al Mamun 2014
Methods RCT
Participants Smear-positive tuberculosis participants (N = 111)
Interventions Vitamin A (5000 IU daily) + zinc (15 mg daily)
Placebo
Outcomes Smear conversion, radiology, haemoglobin, erythrocyte sedimentation rate, C-reactive protein
Notes The abstract does not provide any data for any of the measured outcomes. We contacted the study
authors but did not receive a response
Chandra 2004
Methods RCT
Duration: not described
Participants Number: 44 enrolled (28 men and 16 women)
Interventions Group 1: 3 tablets 3 times weekly of multiple micronutrients
Group 2: placebo
Outcomes Sputum-smear test positive at 2, 3, 5, and 6 months
Chest x-ray findings positive at 3 and 6 months
Notes Location: India
Setting: unclear
Information still pending: Ranjit Kumar Chandra is a researcher in the field of nutrition and im-
munology who has been accused of committing scientific fraud by the British Medical Journal. A ju-
ry trial in July 2015 concluded that the allegations of fraud were truthful. Due to these allegations,
a number of his scientific articles have been subject to retraction (see https://en.wikipedia.org/wi-
ki/Ranjit_Chandra).
Guzman-Rivero 2013
Methods RCT (pilot study)
Collaboration.
Informed decisions.
Guzman-Rivero 2013 (Continued)
Participants Pulmonary tuberculosis patients
Interventions Antituberculous treatment combined with zinc (45 mg/day) or placebo for 3 months
Outcomes Clinical outcomes, sputum clearance, radiological improvement and nutritional status
Notes Conference abstracts (2)
Information still pending:
Neither of the abstracts provides data for any of the measured outcomes. We contacted the au-
thors for information but have not received a response.
Nagrale 2013
Methods RCT
Participants Newly diagnosed sputum positive pulmonary tuberculosis
Interventions Antituberculous treatment with N-acetylcysteine (600 mg, 2 tablets daily ) or placebo for 2 months
Outcomes Sputum conversion, radiological improvement, DTH response, glutathione peroxidase levels
Notes Conference abstract
Information still pending:
The abstract does not provide any data for any of the measured outcomes. We contacted the study
authors in May 2015 for full text but did not receive a response
Nawas 2013
Methods RCT; November 2010 to October 2011
Participants 100 newly diagnosed sputum positive pulmonary tuberculosis
Interventions Antituberculous treatment (fixed-dose combination) with Morinda citrifolia (125 mg) and Zinger of-
ficinale (125 mg) extract or placebo
Outcomes Sputum conversion at 6 months
Notes Conference abstract
Information still pending: the abstract does not provide any data for any of the measured out-
comes. We contacted the study authors in May 2015 for the full-text article but did not receive a re-
sponse
Abbreviations: RCT = randomised controlled trial; N = number of participants.
Characteristics of ongoing studies [ordered by study ID]
Collaboration.
Informed decisions.
ChiCTR-IPR-15006395
Trial name or title The influence and mechanism of vitamin D3 supplementation on the treatment outcomes of tuber-
culosis patients of different glucose tolerance.
Methods RCT.
Participants Inclusion criteria: new diagnosed tuberculosis, sputum smear positive, aged ≥ 18, stable address.
Exclusion criteria: HIV-positive, tumour, pregnant or lactating women, injured lately, adjusted cal-
cium concentration > 2.65 mmol/L.
Gender: both.
Interventions NTB: vitamin D3; HTB: vitamin D3; NTB: control; HTB: control.
Outcomes Primary: treatment outcomes; CD4+/CD8+; quantity of life (SF-36v2).
Secondary: oxidative stress; serum 25(OH)D.
Starting date 25 May 2015.
Contact information [email protected]
Notes China.
ChiCTR-TRC-12002546
Trial name or title The effect and mechanism of retinol and vitamin A supplementation in patients with diabetes and
pulmonary TB.
Methods RCT.
Participants Inclusion criteria: participants with pulmonary tuberculosis and diabetes, aged 18 to 75 years, diag-
nosed by the golden criteria of pulmonary tuberculosis and diabetes; no vitamin or mineral supple-
ment 1 month before the screening.
Exclusion criteria: no severe complications of diabetes including diabetic eye diseases, renal dis-
ease and foot disease, etc; pregnancy or lactation women; cancer; coronary heart disease; recent
suffered trauma or recent surgery.
Interventions Vitamin D (400 IU/d).
Vitamin A (2000 IU/d).
Vitamin D (400 IU/d) + vitamin A (2000 IU/d).
Placebo control.
Outcomes Retinol; vitamin D; fasting glucose; fasting insulin; CD4+/D8+; protein kinase C (PKC); blood lym-
phocyte proliferation.
Starting date 20 October 2012.
Contact information [email protected]; [email protected]
Notes China.
Collaboration.
Informed decisions.
ChiCTR-TRC-14005241
Trial name or title A prospective study of oral nutritional supplement in perioperative application with pulmonary TB
patients.
Methods RCT.
Participants Inclusion criteria: diseases mainly included tuberculosis cavity, tuberculous bronchiectasis dis-
ease, pulmonary tuberculosis and pneumothorax, tuberculosis and pulmonary bulla, tuberculous
pyothorax, who were all the hospitalized patients with video assisted thoracoscopic surgery; by
NRS2002 screening score greater than or equal to 3 points to consider the risk of malnutrition in
this study; ranging in age from 18 to 80.
Exclusion criteria: pregnancy and lactation women; having the injection contraindication; having
nutrition agent or medication allergy or a history of asthma; diabetes medication; fat metabolism,
liver and kidney dysfunction; on the day of surgery transfusion volume above 800 mL; unstable en-
docrine disease; unstable vital signs.
Interventions Nutrison Fibre + normal diet or normal diet.
Outcomes Serum albumin and pre-albumin, weight, T cell subset.
Starting date From 01 October 2014 to 01 October 2016.
Notes Shenzhen, China.
Hospital setting.
IRCT201112178429N1
Trial name or title Effect of zinc supplementation in improving pulmonary TB patients in Qom.
Methods RCT.
Participants Inclusion criteria: active tuberculosis and filling the informed consent.
Exclusion criteria: diseases such as cancer, stroke, immunosuppressive diseases, and not agreeing
to take mineral supplement or placebo.
Interventions Intervention 1: zinc supplement (30 mg zinc) every second day for 6 months.
Intervention 2: placebo every second day for 6 months.
Outcomes Radiological signs: timepoint: at baseline, 2 months, and 6 months later. Method of measurement:
chest radiology.
Smear negative: timepoint: at baseline, 2 months, and 6 months later. Method of measurement:
smear.
Stop coughing: timepoint: at baseline, 2 months and 6 months later. Method of measurement:
physical exam.
Stop fever: timepoint: at baseline, 2 months and 6 months later. Method of measurement: physical
exam.
Starting date 21 January 2012.
Collaboration.
Informed decisions.
IRCT201112178429N1 (Continued)
Notes Recruitment complete.
Iran.
IRCT201211179855N2
Trial name or title Randomized, double-blind, placebo-controlled trial of L-arginine supplementation for the treat-
ment of pulmonary TB.
Methods RCT.
Participants Inclusion criteria: adults over 15 years; new cases of positive sputum smear.
Exclusion criteria: pregnant women; aged under 15 years; patients who during the past month re-
ceived L-arginine; sensitivity to L-arginine.
Interventions Intervention 1: standard antituberculous treatment for 6 months plus L-arginine, 2 g daily in first
month.
Intervention 2: standard 6-month antituberculous treatment plus placebo 2 g daily in first month.
Outcomes Improved para-clinical tests. Timepoint: day 14, 28, 42, and 56. Measurement method: check of
CRP.
Improved para-clinical tests. Timepoint: day 14, 28, 42, and 56. Measurement method: check of
ESR.
Improved para-clinical tests. Timepoint: day 14, 28, 42, and 56. Measurement method: check of
CBC.
Weight gain: timepoint: 0 and 56. Measurement method: weight measurement.
Improve of life quality: timepoint: 0 and 56. Measurement method: standard questionnaire of
GHQ-28.
Starting date 21 December 2012.
[email protected]
Notes Recruitment complete.
Iran.
ISRCTN16469166
Trial name or title Nutrition and wasting in TB: Can nutritional supplementation in TB patients improve body weight
gain, body composition and treatment outcome?
Methods RCT.
Participants Expected enrolment: adults > 18 years with sputum positive pulmonary tuberculosis.
Interventions • Standard treatment plus nutritional supplementation (supplements containing 450 kcal energy
plus a wide range of vitamins and minerals in doses of approximately half the required daily in-
takes (but no iron)).
Collaboration.
Informed decisions.
ISRCTN16469166 (Continued)
• Standard treatment only.
Outcomes Primary: body weight gain and body composition.
Secondary: micronutrient status and production of inflammatory cytokines.
Starting date 15 May 2007.
Anticipated end date: 1 January 2008.
Contact information Dr Frank Wieringa ([email protected]), Dept. of Internal Medicine, Hasan Sadikin Hospital, Ban-
dung, Indonesia.
Notes We attempted to contact the author for results on 25 January 2010 and 05 May 2010 with no reply.
Location: Indonesia.
Registration number: ISRCTN16469166.
Source of funding: Netherlands Organisation for Scientific Research (The Netherlands).
We contacted the trial authors for information on 20 July 2015.
There was no sign of publication of the data (PubMed search).
NCT00507000
Trial name or title Role of oral vitamin D as an adjunct therapy in Category I pulmonary TB along with assessment of
immunological parameters.
Methods RCT.
Participants Adults aged 18 to 60 years with newly diagnosed sputum-smear positive pulmonary tuberculosis.
Interventions • Cholecalciferol (vitamin D) and calcium carbonate.

• Placebo (lactose granules).
Outcomes Primary: time to becoming sputum-smear negative.
Secondary: relapse rate; safety assessment; immune function.
Starting date May 2008.
Anticipated end date: September 2010.
Contact information Dr Ravinder Goswami ([email protected]), Department of Endocrinology and Metabolism,

All India Institue of Medical Sciences, New Delhi, India.
Notes Location: India.
Source of funding: Indian Council of Medical Research; Ministry of Science and Technology, India.
We contacted the trial authors for information.
There was no sign of publication of the data (PubMed search).
Collaboration.
Informed decisions.
NCT00698386
Trial name or title Efficacy of oral zinc administration as an adjunct therapy in new pulmonary TB (Category I) pa-
tients.
Methods RCT.
Participants Adults with newly diagnosed smear positive pulmonary tuberculosis.
Interventions 50 mg zinc as zinc sulphate daily for 6 months during DOTS versus placebo.
Outcomes Time to sputum conversion, cure rate, relapse, adverse events, patient and physician global assess-
ment of cure.
Starting date February 2008.
Contact information Surendra K Sharma, MD, PhD, All India Institute of Medical Sciences, New Delhi.
[email protected], [email protected]
Notes Location: India.
Source of funding: Ministry of Science and Technology, India.
We contacted the trial authors for information.
No sign of publication of the data (PubMed search).
NCT00788320
Trial name or title Antimicrobial peptide LL-37 (cathelicidin) production in active TB disease: role of vitamin D supple-
mentation.
Methods RCT.
Participants Adults aged > 18 years with newly diagnosed pulmonary tuberculosis.
Interventions Drug: vitamin D3.

Drug: placebo.
Outcomes LL-37 level, time to sputum conversion, vitamin D and calcium levels.
Starting date October 2008.
Contact information Not stated.
Notes Location: USA.
Source of funding: Emory University.
This study has been withdrawn prior to enrolment (inadequate enrolment). See https://clinicaltri-
als.gov/ct2/show/NCT00788320 for details.
Collaboration.
Informed decisions.
NCT01635153
Trial name or title Effects of a protein calorie supplement in HIV-infected women with TB DarDar.
Methods RCT.
Participants Inclusion criteria: female, HIV, age > 18, CD4 > 50, BMI > 16 new tuberculosis diagnosis, not on an-
ti-retroviral therapy, residence in Dar es Salaam.
Exclusion Criteria: current anti-retroviral therapy, serious co-morbidities.
Interventions Group 1: micronutrient supplement.
Group 2: protein calorie supplement.
Outcomes Primary outcomes: change in CD4 count (time frame: baseline to 8 months).
Secondary outcomes: body mass index at 6 months (time frame: baseline to 6 months); propor-
tion of subjects who achieve 100 cell count increase in CD4 cell count (Time frame: baseline to 8
months).
Starting date May 2012.
Contact information Charles F von Reyn, MD, Geisel School of Medicine at Dartmouth.
Notes Active, not recruiting.
NCT01657656
Trial name or title Vitamin D supplementations as adjunct to anti-TB drugs in Mongolia.
Methods RCT.
Participants Inclusion criteria: sputum positive tuberculosis patients.

Exclusion criteria: abnormal LFTs at baseline (2.5 times upper limit of normal), as they will be at
higher risk of developing drug-induced hepatitis.
Interventions Vitamin D (does not state what comparator group receiving).
Outcomes The primary endpoint will be time to sputum culture conversion from positive to negative (time
frame: 8 weeks).
Contact information [email protected] (Ganmaa Davaasambuu).
Notes Study status: completed.
Mongolia.
NCT01722396
Trial name or title Pharmacogenetics of Vitamin D Supplementation in TB.
Collaboration.
Informed decisions.
NCT01722396 (Continued)
Methods Open label study of vitamin D supplementation.
Participants Patients with active or latent tuberculosis.
Interventions Active tuberculosis patients take 100,000 units of vitamin D every 8 weeks during their tuberculosis
treatment.
Outcomes • Ex vivo responses of monocytes to vitamin D (time frame: 0 and 8 weeks).

• Ex vivo responses of T cells to vitamin D (time frame: 0 and 8 weeks).
Starting date March 2011.
Contact information Alice Turner, University of Birmingham.
Notes Status: completed.
NCT01992263
Trial name or title A trial of vitamin D supplementation among TB patients in South India.
Methods RCT.
Participants Inclusion criteria: 18 to 60 years old, active tuberculosis diagnosis by GeneXpert®; HIV infection sta-
tus (according to AMC HIV clinic medical records of enzyme-linked immunosorbent assay (ELISA)
results).
Exclusion criteria: children (< 18 years of age) and older than 60 years, pregnant at baseline, oth-
er severe complications or illnesses requiring hospitalisation, received tuberculosis treatment for
greater than 4 weeks in the past 5 years, refused to participate, residing in a geographic location > 1
hour from AMC (by public transit).
Interventions Vitamin D (2000 IU/4000 IU/600 IU) or placebo.
Outcomes Primary outcomes
• Cell-mediated immunological markers (T cells) (time frame: 1 year).

• Immune function (time frame: 1 year).
• Serum 25(OH)D concentrations (time frame: 1 year).
Secondary outcomes
• HIV disease progression (time frame: 1 year).

• tuberculosis treatment outcomes (time frame: 1 year).
Starting date January 2015.
Notes Not yet recruiting.
NCT02169570
Trial name or title Effect of supplementary vitamin D in patients with diabetes mellitus and pulmonary TB (EVIDENT).
Collaboration.
Informed decisions.
Methods RCT.
Participants Inclusion criteria: 30 to 60 years; patients having both tuberculosis and type 2 DM; consenting to
participate; no history of previous ATT; plan to have ATT and DM treatment.
Exclusion criteria: pregnant women; patients with extra-pulmonary tuberculosis or multi-drug re-
sistant (MDR)-tuberculosis or relapse cases, hepatic or renal diseases or HIV infection, hypo- or hy-
per-parathyroidism; patients on corticosteroids, or immunosuppressive, or thiazides diuretics, or
any other drugs known to interfere with vitamin D levels.
Interventions Calcium; placebo calcium; placebo vitamin D; vitamin D.
Outcomes Primary outcomes
• Change acid fast bacilli (AFB) smear (sputum) (time frame: 0, 4, 8, 12, 16, and 24 weeks, and 6
months).
• Change in chest X-ray (time frame: 0, 8, 16, and 24 weeks, and 6 months).
• Change in tuberculosis score (time frame: 0, 4, 8, 12, 16, 20, and 24 weeks, and 6 months).
• Change in weight (time frame: 0, 4, 8, 12, 16, 20, and 24 weeks, and 6 months).
Secondary outcomes
• Change haemoglobin A1c (HbA1c) (time frame: 0, 8, 16, and 24 weeks, and 6 months).
• Change in fasting blood test (FBS) (time frame: 0, 8, 16, and 24 weeks, and 6 months).
• Change in random blood sugar (RBS) (time frame: 0, 8, 16, and 24 weeks, and 6 months).
Starting date December 2014.
Notes Pakistan.
Status: not yet recruiting.
NCT02464683
Trial name or title Effect of vitamin D as adjunctive therapy in patients with pulmonary evolution TB (Vitamin D).
Methods RCT.
Participants Inclusion criteria: men and women over 18 and under 65 years old, with tuberculosis confirmed by
positive smear and positive culture, without documented evidence of previous treatment for tu-
berculosis, with haemoglobin values greater than 10 g/dL, and written consent.
Exclusion criteria: HIV-positive, no written consent, chronic lung disease, clinical evidence of infec-
tious or chronic inflammatory disease processes such as: rheumatoid arthritis, systemic lupus ery-
thematosus (SLE), Sjögren Sx, dermatomyositis, scleroderma, seronegative arthritis, gout, inflam-
matory bowel disease, chronic active hepatitis, glomerulonephritis, rheumatic fever and cardiac
disease, cancer, and history of alcohol or drug abuse.
Interventions Vitamin D (200 IU) or placebo daily for 60 days.
Outcomes Determination of cytokines.
Starting date April 2014.
Contact information Martha Torres Rojas, PhD [email protected]
Collaboration.
Informed decisions.
Notes Mexico.
NCT02554318
Trial name or title The effect of fermented soybean supplementation on the body weight and physical function of TB
patients with standard therapy in Indonesia.
Methods RCT.
Participants Inclusion criteria: newly diagnosed adult male and female tuberculosis active patients with clinical
evidences of active tuberculosis symptoms (positive or negative sputum smears, positive chest X-
ray that compatible with a diagnosis of tuberculosis); no history of previous antituberculous treat-
ment.
Exclusion criteria: heavy smoker (> 20 cigarettes per day); pregnancy and lactation; extrapul-
monary tuberculosis; known allergy to soybean; clinical evidences of any underlying disease.
Interventions Fermented soybean.
Outcomes Body weight, body mass index, handgrip strength, 6-minute walk test.
Contact information Michael B Krawinkel, Prof Dr University of Gleesen.
Notes Uncertain what the comparison group received.
DATA AND ANALYSES
Comparison 1. Macronutrient supplementation
Outcome or subgroup title No. of No. of Statistical method Effect size

studies partici-
pants
1 Death (1 year of follow-up) 4 567 Risk Ratio (M-H, Fixed, 95% CI) 0.34 [0.10, 1.20]
1.1 HIV-positive 2 72 Risk Ratio (M-H, Fixed, 95% CI) 0.60 [0.12, 3.01]
1.2 HIV-negative 3 230 Risk Ratio (M-H, Fixed, 95% CI) 0.18 [0.02, 1.48]
1.3 Unknown HIV status 1 265 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2 Cured (at 6 months) 1 102 Risk Ratio (M-H, Random, 95% CI) 0.91 [0.59, 1.41]
2.1 HIV-positive 1 22 Risk Ratio (M-H, Random, 95% CI) 1.38 [0.46, 4.14]
2.2 HIV-negative 1 80 Risk Ratio (M-H, Random, 95% CI) 0.85 [0.53, 1.35]
3 Treatment completion 2 Risk Ratio (M-H, Random, 95% CI) Subtotals only
Collaboration.
Informed decisions.

studies partici-
pants
3.1 HIV-negative 1 100 Risk Ratio (M-H, Random, 95% CI) 1.20 [1.04, 1.37]
3.2 Unknown HIV status 1 265 Risk Ratio (M-H, Random, 95% CI) 0.98 [0.86, 1.12]
4 Sputum negative at 8 weeks 3 222 Risk Ratio (M-H, Random, 95% CI) 1.08 [0.86, 1.37]
5 Mean weight gain 5 Mean Difference (IV, Random, 95% CI) Subtotals only
5.1 After 6 weeks 1 34 Mean Difference (IV, Random, 95% CI) 1.73 [0.81, 2.65]
5.2 After 8 weeks 3 689 Mean Difference (IV, Random, 95% CI) 0.78 [-0.05, 1.60]
5.4 After 20 weeks (12 weeks post 1 306 Mean Difference (IV, Random, 95% CI) -0.20 [-1.34, 0.94]
supplementation)
5.5 After 24 weeks 1 26 Mean Difference (IV, Random, 95% CI) 1.78 [-0.25, 3.81]
6 Change in maximum grip 3 Mean Difference (IV, Random, 95% CI) Subtotals only
strength (kg)
6.1 At 6 weeks 1 34 Mean Difference (IV, Random, 95% CI) 3.44 [0.78, 6.10]
6.2 At 8 weeks 1 332 Mean Difference (IV, Random, 95% CI) 0.50 [-0.63, 1.63]
6.3 At 12 weeks 1 100 Mean Difference (IV, Random, 95% CI) 1.50 [1.08, 1.92]
7 Change in quality of life score 1 Mean Difference (IV, Random, 95% CI) Totals not selected
7.1 At 6 weeks 1 Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
7.2 At 24 weeks 1 Mean Difference (IV, Random, 95% CI) 0.0 [0.0, 0.0]
Analysis 1.1. Comparison 1 Macronutrient supplementation, Outcome 1 Death (1 year of follow-up).

Study or subgroup Supplement No supplement Risk Ratio Weight Risk Ratio
n/N n/N M-H, Fixed, 95% CI M-H, Fixed, 95% CI
1.1.1 HIV-positive
Jeremiah 2014 TZA 1/26 3/24 33.03% 0.31[0.03,2.76]
Sudarsanam 2010 IND 1/13 0/9 6.17% 2.14[0.1,47.38]
Subtotal (95% CI) 39 33 39.2% 0.6[0.12,3.01]
Total events: 2 (Supplement), 3 (No supplement)
Favours supplement 0.001 0.1 1 10 1000 Favours no supplement
Collaboration.
Informed decisions.

Heterogeneity: Tau2=0; Chi2=1.01, df=1(P=0.32); I2=0.51%
Test for overall effect: Z=0.63(P=0.53)
1.1.2 HIV-negative
Jahnavi 2010 IND 0/50 2/50 26.46% 0.2[0.01,4.06]
Jeremiah 2014 TZA 0/25 0/25 Not estimable
Sudarsanam 2010 IND 0/37 3/43 34.34% 0.17[0.01,3.1]
Subtotal (95% CI) 112 118 60.8% 0.18[0.02,1.48]
Heterogeneity: Tau2=0; Chi2=0.01, df=1(P=0.93); I2=0%
1.1.3 Unknown HIV status

Martins 2009 TLS 0/136 0/129 Not estimable
Subtotal (95% CI) 136 129 Not estimable
Heterogeneity: Not applicable
Test for overall effect: Not applicable
Total (95% CI) 287 280 100% 0.34[0.1,1.2]

Test for subgroup differences: Chi2=0.78, df=1 (P=0.38), I2=0%
Favours supplement 0.001 0.1 1 10 1000 Favours no supplement
Analysis 1.2. Comparison 1 Macronutrient supplementation, Outcome 2 Cured (at 6 months).

Study or subgroup Supplement Advice only Risk Ratio Weight Risk Ratio
n/N n/N M-H, Random, 95% CI M-H, Random, 95% CI
1.2.1 HIV-positive
Sudarsanam 2010 IND 6/13 3/9 15.6% 1.38[0.46,4.14]
Subtotal (95% CI) 13 9 15.6% 1.38[0.46,4.14]
Total events: 6 (Supplement), 3 (Advice only)
1.2.2 HIV-negative
Sudarsanam 2010 IND 16/37 22/43 84.4% 0.85[0.53,1.35]
Subtotal (95% CI) 37 43 84.4% 0.85[0.53,1.35]
Heterogeneity: Tau2=0; Chi2=0, df=0(P<0.0001); I2=100%
Total (95% CI) 50 52 100% 0.91[0.59,1.41]

Favours Advice only 0.01 0.1 1 10 100 Favours Supplement
Collaboration.
Informed decisions.
Analysis 1.3. Comparison 1 Macronutrient supplementation, Outcome 3 Treatment completion.

1.3.1 HIV-negative
Jahnavi 2010 IND 49/50 41/50 100% 1.2[1.04,1.37]
Subtotal (95% CI) 50 50 100% 1.2[1.04,1.37]

Martins 2009 TLS 103/136 100/129 100% 0.98[0.86,1.12]
Subtotal (95% CI) 136 129 100% 0.98[0.86,1.12]
Test for subgroup differences: Chi2=4.32, df=1 (P=0.04), I2=76.85%
Favours no supplement 0.5 0.7 1 1.5 2 Favours supplement
Analysis 1.4. Comparison 1 Macronutrient supplementation, Outcome 4 Sputum negative at 8 weeks.

Jahnavi 2010 IND 35/36 29/36 37.47% 1.21[1.02,1.43]
Jeremiah 2014 TZA 37/38 35/35 44.48% 0.98[0.91,1.05]
Martins 2009 TLS 22/40 18/37 18.05% 1.13[0.73,1.74]
Total (95% CI) 114 108 100% 1.08[0.86,1.37]

Heterogeneity: Tau2=0.03; Chi2=10.04, df=2(P=0.01); I2=80.08%
Favours no supplement 0.5 0.7 1 1.5 2 Favours supplement
Analysis 1.5. Comparison 1 Macronutrient supplementation, Outcome 5 Mean weight gain.

Study or subgroup Supplement No supplement Mean Difference Weight Mean Difference
N Mean(SD) N Mean(SD) Random, 95% CI Random, 95% CI
1.5.1 After 6 weeks
Paton 2004 SGP 19 2.6 (1.8) 15 0.8 (0.9) 100% 1.73[0.81,2.65]
Subtotal *** 19 15 100% 1.73[0.81,2.65]
Test for overall effect: Z=3.69(P=0)
1.5.2 After 8 weeks

Jeremiah 2014 TZA 48 56.2 (12.7) 44 55.1 (9.4) 3.29% 1.1[-3.44,5.64]
Martins 2009 TLS 136 5.2 (6.2) 129 3.5 (6.3) 27.1% 1.7[0.19,3.21]
Favours no supplement -5 -2.5 0 2.5 5 Favours supplement
Collaboration.
Informed decisions.
Study or subgroup Supplement No supplement Mean Difference Weight Mean Difference

Praygod 2011b TZA 166 2.9 (3.9) 166 2.5 (3.9) 69.61% 0.4[-0.45,1.25]
Subtotal *** 350 339 100% 0.78[-0.05,1.6]
1.5.3 After 12 weeks

Jahnavi 2010 IND 50 3.7 (2.1) 50 1.1 (2.3) 100% 2.6[1.74,3.46]
Subtotal *** 50 50 100% 2.6[1.74,3.46]
Test for overall effect: Z=5.9(P<0.0001)
1.5.4 After 20 weeks (12 weeks post supplementation)

Praygod 2011b TZA 151 6 (4.4) 155 6.2 (5.7) 100% -0.2[-1.34,0.94]
Subtotal *** 151 155 100% -0.2[-1.34,0.94]

Paton 2004 SGP 15 4.4 (2.7) 11 2.7 (2.5) 100% 1.78[-0.25,3.81]
Subtotal *** 15 11 100% 1.78[-0.25,3.81]

Martins 2009 TLS 136 10.1 (9.6) 129 7.5 (7.6) 100% 2.6[0.52,4.68]
Subtotal *** 136 129 100% 2.6[0.52,4.68]
Test for subgroup differences: Chi2=19.03, df=1 (P=0), I2=73.73%
Favours no supplement -5 -2.5 0 2.5 5 Favours supplement
Analysis 1.6. Comparison 1 Macronutrient supplementation, Outcome 6 Change in maximum grip strength (kg).
Study or subgroup Supplements Advice only Mean Difference Weight Mean Difference
1.6.1 At 6 weeks
Paton 2004 SGP 19 2.8 (3.1) 15 -0.6 (4.5) 100% 3.44[0.78,6.1]
Subtotal *** 19 15 100% 3.44[0.78,6.1]
1.6.2 At 8 weeks
Praygod 2011b TZA 165 2.3 (5.2) 167 1.8 (5.3) 100% 0.5[-0.63,1.63]
Subtotal *** 165 167 100% 0.5[-0.63,1.63]
1.6.3 At 12 weeks
Jahnavi 2010 IND 50 3.9 (1.2) 50 2.4 (0.9) 100% 1.5[1.08,1.92]
Favours advice only -10 -5 0 5 10 Favours supplements
Collaboration.
Informed decisions.
Study or subgroup Supplements Advice only Mean Difference Weight Mean Difference
Subtotal *** 50 50 100% 1.5[1.08,1.92]
1.6.4 At 20 weeks
Praygod 2011b TZA 150 6 (6.3) 153 4.7 (6.3) 100% 1.3[-0.11,2.71]
Subtotal *** 150 153 100% 1.3[-0.11,2.71]
1.6.5 At 24 weeks
Paton 2004 SGP 15 3.2 (3.9) 11 2.8 (4.8) 100% 0.39[-3.05,3.83]
Subtotal *** 15 11 100% 0.39[-3.05,3.83]
Favours advice only -10 -5 0 5 10 Favours supplements
Analysis 1.7. Comparison 1 Macronutrient supplementation, Outcome 7 Change in quality of life score.
Study or subgroup Supplements Advice only Mean Difference Mean Difference
1.7.1 At 6 weeks
Paton 2004 SGP 19 14.5 (25.4) 15 13.3 (24.8) 1.14[-15.82,18.1]
1.7.2 At 24 weeks
Paton 2004 SGP 15 8.3 (22.5) 11 18.8 (27.4) -10.42[-30.2,9.36]
Favours advice -50 -25 0 25 50 Favours supplements
Comparison 2. High cholesterol (850 mg/day) versus low cholesterol (250 mg/day) diet

studies partici-
pants
1 Sputum-culture positive 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected
1.1 At baseline 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
1.2 At 2 weeks 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
Collaboration.
Informed decisions.
Analysis 2.1. Comparison 2 High cholesterol (850 mg/day) versus low

cholesterol (250 mg/day) diet, Outcome 1 Sputum-culture positive.
Study or subgroup High cholesterol diet Low cholesterol diet Risk Ratio Risk Ratio
2.1.1 At baseline
Pérez-Guzmán 2005 MEX 10/10 11/11 1[0.84,1.19]
2.1.2 At 2 weeks
Pérez-Guzmán 2005 MEX 2/10 10/11 0.22[0.06,0.77]
2.1.3 At 4 weeks
Pérez-Guzmán 2005 MEX 1/10 3/11 0.37[0.05,2.98]
2.1.4 At 8 weeks
Pérez-Guzmán 2005 MEX 0/10 0/11 Not estimable
Favours high cholesterol 0.01 0.1 1 10 100 Favours low cholesterol
Comparison 3. Multivitamin and trace element tablets versus placebo

studies partici-
pants
1 Death during follow-up in 5 2902 Risk Ratio (M-H, Random, 95% CI) 0.96 [0.79, 1.18]
adults and children
1.1 HIV-negative individuals 4 1218 Risk Ratio (M-H, Random, 95% CI) 0.86 [0.46, 1.60]
1.2 HIV-positive individuals 3 1429 Risk Ratio (M-H, Random, 95% CI) 0.92 [0.69, 1.23]
1.3 HIV-positive and HIV-neg- 1 255 Risk Ratio (M-H, Random, 95% CI) 1.59 [0.53, 4.72]
ative individuals
2 Tuberculosis treatment 1 302 Risk Ratio (M-H, Fixed, 95% CI) 0.99 [0.95, 1.04]
completion
3 Sputum-smear or spu- 2 1020 Risk Ratio (M-H, Random, 95% CI) 0.92 [0.63, 1.35]
tum-culture positive at 1
month
3.1 Mixed HIV-positive and 1 392 Risk Ratio (M-H, Random, 95% CI) 1.20 [0.99, 1.45]
HIV-negative
3.2 HIV-negative individuals 1 306 Risk Ratio (M-H, Random, 95% CI) 0.73 [0.47, 1.13]
only
3.3 HIV-positive individuals 1 322 Risk Ratio (M-H, Random, 95% CI) 0.77 [0.45, 1.31]
only
4 Sputum-smear or spu- 2 731 Risk Ratio (M-H, Fixed, 95% CI) 0.89 [0.59, 1.35]
tum-culture positive at 2
months
Collaboration.
Informed decisions.

studies partici-
pants
5 Clearance of chest X-ray at 2 497 Risk Ratio (M-H, Fixed, 95% CI) 1.03 [0.90, 1.18]
6 months
6 Weight 3 Mean Difference (IV, Fixed, 95% CI) Totals not selected
6.1 After 2 months of supple- 2 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
mentation
6.2 After 5 months follow-up 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
(3 months post supplementa-
tion)
6.3 After 6 months supple- 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
mentation (children)
6.4 After 7 months of supple- 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
mentation
7 Anthropometrical changes 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
at follow-up
7.1 Weight-for-age z score at 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
2 months
7.2 Weight-for-age z score at 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
6 months
7.3 Weight (kg) at 2 months 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
7.4 Weight (kg) at 6 months 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
7.5 Height-for-age z score at 2 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
months
7.6 Height-for-age z score at 6 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
months
7.7 BMI z score at 2 months 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
7.8 BMI z score at 6 months 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
8 Mean change in handgrip 1 1480 Mean Difference (IV, Fixed, 95% CI) 1.16 [0.50, 1.81]
strength (kg)
8.1 At 2 months 1 771 Mean Difference (IV, Fixed, 95% CI) 1.22 [0.49, 1.95]
8.2 At 5 months 1 709 Mean Difference (IV, Fixed, 95% CI) 0.90 [-0.56, 2.36]
Collaboration.
Informed decisions.
Analysis 3.1. Comparison 3 Multivitamin and trace element tablets

versus placebo, Outcome 1 Death during follow-up in adults and children.
Study or subgroup Multimi- Placebo Risk Ratio Weight Risk Ratio
cronutrient
3.1.1 HIV-negative individuals
Range 2005 TZA 2/55 1/32 0.71% 1.16[0.11,12.33]
Range 2005 TZA 4/61 1/33 0.85% 2.16[0.25,18.58]
Semba 2007 MWI 9/169 8/150 4.37% 1[0.4,2.52]
Villamor 2008 TZA 4/208 11/208 3.01% 0.36[0.12,1.12]
Lodha 2014 IND 2/102 0/50 0.44% 2.48[0.12,50.62]
Lodha 2014 IND 2/100 0/50 0.44% 2.52[0.12,51.61]
Subtotal (95% CI) 695 523 9.81% 0.86[0.46,1.6]
Total events: 23 (Multimicronutrient), 21 (Placebo)
3.1.2 HIV-positive individuals

Range 2005 TZA 11/45 7/21 5.83% 0.73[0.33,1.62]
Range 2005 TZA 4/42 7/21 3.09% 0.29[0.09,0.87]
Semba 2007 MWI 157/406 171/423 47.92% 0.96[0.81,1.13]
Villamor 2008 TZA 74/233 66/238 30.14% 1.15[0.87,1.51]
Subtotal (95% CI) 726 703 86.99% 0.92[0.69,1.23]
3.1.3 HIV-positive and HIV-negative individuals

Mehta 2011 TZA 8/128 5/127 3.21% 1.59[0.53,4.72]
Subtotal (95% CI) 128 127 3.21% 1.59[0.53,4.72]
Total (95% CI) 1549 1353 100% 0.96[0.79,1.18]

Favours MMN 0.005 0.1 1 10 200 Favours placebo

versus placebo, Outcome 2 Tuberculosis treatment completion.
cronutrient
Lodha 2014 IND 97/102 48/50 49.65% 0.99[0.92,1.06]
Lodha 2014 IND 97/100 49/50 50.35% 0.99[0.94,1.04]
Total (95% CI) 202 100 100% 0.99[0.95,1.04]
Favours placebo 1 Favours MMN
Collaboration.
Informed decisions.

cronutrient
Heterogeneity: Tau2=0; Chi2=0, df=1(P=0.98); I2=0%
Favours placebo 1 Favours MMN
Analysis 3.3. Comparison 3 Multivitamin and trace element tablets versus

placebo, Outcome 3 Sputum-smear or sputum-culture positive at 1 month.
cronutrient
3.3.1 Mixed HIV-positive and HIV-negative
Range 2005 TZA 115/201 91/191 44.3% 1.2[0.99,1.45]
Subtotal (95% CI) 201 191 44.3% 1.2[0.99,1.45]
3.3.2 HIV-negative individuals only

Villamor 2008 TZA 27/149 39/157 30.26% 0.73[0.47,1.13]
Subtotal (95% CI) 149 157 30.26% 0.73[0.47,1.13]
3.3.3 HIV-positive individuals only

Villamor 2008 TZA 21/165 26/157 25.43% 0.77[0.45,1.31]
Subtotal (95% CI) 165 157 25.43% 0.77[0.45,1.31]
Total (95% CI) 515 505 100% 0.92[0.63,1.35]

Favours MMN 0.1 0.2 0.5 1 2 5 10 Favours placebo
Analysis 3.4. Comparison 3 Multivitamin and trace element tablets versus

placebo, Outcome 4 Sputum-smear or sputum-culture positive at 2 months.
cronutrient
Lodha 2014 IND 12/102 5/50 16.11% 1.18[0.44,3.16]
Collaboration.
Informed decisions.

cronutrient
Lodha 2014 IND 10/100 4/50 12.8% 1.25[0.41,3.79]
Range 2005 TZA 23/219 29/210 71.08% 0.76[0.46,1.27]
Total (95% CI) 421 310 100% 0.89[0.59,1.35]


versus placebo, Outcome 5 Clearance of chest X-ray at 6 months.
cronutrient
Lodha 2014 IND 75/99 36/50 33.35% 1.05[0.86,1.29]
Lodha 2014 IND 67/99 36/50 33.35% 0.94[0.75,1.17]
Mehta 2011 TZA 51/97 49/102 33.3% 1.09[0.83,1.44]
Total (95% CI) 295 202 100% 1.03[0.9,1.18]

Favours placebo 0.1 0.2 0.5 1 2 5 10 Favours MMN
Analysis 3.6. Comparison 3 Multivitamin and trace element tablets versus placebo, Outcome 6 Weight.
Study or subgroup Multimicronutrient Placebo Mean Difference Mean Difference
N Mean(SD) N Mean(SD) Fixed, 95% CI Fixed, 95% CI
3.6.1 After 2 months of supplementation
Lodha 2014 IND 102 22.9 (8.9) 100 22.3 (8.8) 0.57[-1.87,3.01]
Lodha 2014 IND 100 23 (10.2) 100 22.3 (8.8) 0.66[-1.99,3.31]
Praygod 2011a TZA 383 2.9 (3.6) 392 2.9 (4.2) 0.06[-0.49,0.61]
3.6.2 After 5 months follow-up (3 months post supplementation)

Praygod 2011a TZA 349 4.9 (4.6) 360 4.8 (4.9) 0.07[-0.63,0.77]
3.6.3 After 6 months supplementation (children)

Lodha 2014 IND 102 24.8 (9.7) 100 24 (9.4) 0.82[-1.8,3.44]
Lodha 2014 IND 100 24.6 (10.8) 100 24 (9.4) 0.61[-2.2,3.42]
3.6.4 After 7 months of supplementation

Range 2005 TZA 101 6.6 (0.4) 97 6.3 (0.5) 0.3[0.17,0.43]
Range 2005 TZA 95 8.7 (0.6) 97 6.3 (0.5) 2.37[2.21,2.53]
Favours placebo -2 -1 0 1 2 Favours MMN
Collaboration.
Informed decisions.

versus placebo, Outcome 7 Anthropometrical changes at follow-up.
Study or subgroup Multimicronutrient Placebo Mean Difference Mean Difference
3.7.1 Weight-for-age z score at 2 months
Lodha 2014 IND 99 -2.2 (1.7) 99 -2.4 (1.6) 0.21[-0.24,0.66]
Lodha 2014 IND 99 -2.4 (1.4) 99 -2.4 (1.6) 0.01[-0.4,0.42]
3.7.2 Weight-for-age z score at 6 months

Lodha 2014 IND 99 -2.1 (1.3) 99 -2.4 (1.6) 0.32[-0.08,0.72]
Lodha 2014 IND 99 -1.9 (1.5) 99 -2.4 (1.6) 0.52[0.09,0.95]
3.7.3 Weight (kg) at 2 months

Lodha 2014 IND 99 23 (10.2) 99 22.3 (8.8) 0.66[-2,3.32]
Lodha 2014 IND 99 22.9 (8.9) 99 22.3 (8.8) 0.57[-1.9,3.04]
3.7.4 Weight (kg) at 6 months

Lodha 2014 IND 99 24.6 (10.8) 99 24 (9.4) 0.61[-2.21,3.43]
Lodha 2014 IND 99 24.8 (9.7) 99 24 (9.4) 0.82[-1.83,3.47]
3.7.5 Height-for-age z score at 2 months

Lodha 2014 IND 99 -1.7 (1.2) 99 -1.6 (1.4) -0.02[-0.38,0.34]
Lodha 2014 IND 99 -1.5 (1.5) 99 -1.6 (1.4) 0.1[-0.3,0.5]
3.7.6 Height-for-age z score at 6 months

Lodha 2014 IND 99 -1.5 (1.5) 99 -1.6 (1.4) 0.18[-0.21,0.57]
Lodha 2014 IND 99 -1.5 (1.2) 99 -1.6 (1.4) 0.1[-0.26,0.46]
3.7.7 BMI z score at 2 months

Lodha 2014 IND 100 -1.7 (1.2) 100 -1.7 (1.3) 0.02[-0.33,0.37]
Lodha 2014 IND 102 -1.6 (1.2) 100 -1.7 (1.3) 0.16[-0.18,0.5]
3.7.8 BMI z score at 6 months

Lodha 2014 IND 102 -1.2 (1.2) 100 -1.3 (1.2) 0.09[-0.24,0.42]
Lodha 2014 IND 100 -1.4 (1.1) 100 -1.3 (1.2) -0.16[-0.48,0.16]
Favours placebo -10 -5 0 5 10 Favours MMN

versus placebo, Outcome 8 Mean change in handgrip strength (kg).
Study or subgroup Multimicronutrient Placebo Mean Difference Weight Mean Difference
3.8.1 At 2 months
Praygod 2011a TZA 379 2.2 (4.9) 392 1 (5.5) 79.94% 1.22[0.49,1.95]
Subtotal *** 379 392 79.94% 1.22[0.49,1.95]
3.8.2 At 5 months
Praygod 2011a TZA 348 5 (12.7) 361 4.1 (5.7) 20.06% 0.9[-0.56,2.36]
Favours placebo -5 -2.5 0 2.5 5 Favours MMN
Collaboration.
Informed decisions.
Study or subgroup Multimicronutrient Placebo Mean Difference Weight Mean Difference

Subtotal *** 348 361 20.06% 0.9[-0.56,2.36]
Total *** 727 753 100% 1.16[0.5,1.81]

Favours placebo -5 -2.5 0 2.5 5 Favours MMN
Comparison 4. Vitamin A versus placebo

studies partici-
pants
1 Children: mean serum retinol 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
(normal range > 20 µg/L)
1.1 At baseline 1 85 Mean Difference (IV, Fixed, 95% CI) -1.0 [-5.38, 3.38]
1.2 At 6 weeks 1 85 Mean Difference (IV, Fixed, 95% CI) 1.30 [-3.91, 6.51]
2 Adults: mean serum retinol (nor- 3 Mean Difference (IV, Fixed, 95% CI) Subtotals only
mal range > 70 µmol/L)
3 Death 8 3308 Risk Ratio (M-H, Fixed, 95% CI) 0.97 [0.84, 1.12]
3.1 Vitamin A alone 1 115 Risk Ratio (M-H, Fixed, 95% CI) 1.79 [0.19, 16.69]
3.2 Vitamin A plus zinc 4 535 Risk Ratio (M-H, Fixed, 95% CI) 2.38 [0.87, 6.53]
3.3 Vitamin A as part of a multi-mi- 4 2658 Risk Ratio (M-H, Fixed, 95% CI) 0.94 [0.81, 1.09]
cronutrient supplement
4 Treatment completion 1 158 Risk Ratio (M-H, Fixed, 95% CI) 0.92 [0.80, 1.04]
5 Symptomatic at 6 weeks 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected
6 Sputum-smear and sputum-cul- 2 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
ture positive during follow-up
6.1 Baseline 1 158 Risk Ratio (M-H, Fixed, 95% CI) 1.0 [0.98, 1.03]
Collaboration.
Informed decisions.

studies partici-
pants
6.2 2 weeks 1 158 Risk Ratio (M-H, Fixed, 95% CI) 0.80 [0.50, 1.28]
6.3 1 month 1 148 Risk Ratio (M-H, Fixed, 95% CI) 0.70 [0.33, 1.48]
6.4 2 months 2 199 Risk Ratio (M-H, Fixed, 95% CI) 1.03 [0.23, 4.73]
7 BMI (kg/m2) 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
7.1 Baseline 1 158 Mean Difference (IV, Fixed, 95% CI) 0.10 [-0.63, 0.83]
7.2 2 months 1 148 Mean Difference (IV, Fixed, 95% CI) 0.30 [-0.44, 1.04]
7.3 6 months 1 136 Mean Difference (IV, Fixed, 95% CI) -0.30 [-1.15, 0.55]
8 Body fat (%) 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
8.1 Baseline 1 158 Mean Difference (IV, Fixed, 95% CI) -0.90 [-2.80, 1.00]
Analysis 4.1. Comparison 4 Vitamin A versus placebo, Outcome

1 Children: mean serum retinol (normal range > 20 µg/L).
Study or subgroup Vitamin A Placebo Mean Difference Weight Mean Difference
4.1.1 At baseline
Hanekom 1997 ZAF 41 17.6 (10.1) 44 18.6 (10.5) 100% -1[-5.38,3.38]
Subtotal *** 41 44 100% -1[-5.38,3.38]
4.1.2 At 6 weeks
Hanekom 1997 ZAF 41 34.8 (10.1) 44 33.5 (14.2) 100% 1.3[-3.91,6.51]
Subtotal *** 41 44 100% 1.3[-3.91,6.51]
4.1.3 At 3 months
Hanekom 1997 ZAF 41 34.4 (10) 44 33.7 (11.4) 100% 0.7[-3.85,5.25]
Subtotal *** 41 44 100% 0.7[-3.85,5.25]
Favours placebo -20 -10 0 10 20 Favours vitamin A
Collaboration.
Informed decisions.

2 Adults: mean serum retinol (normal range > 70 µmol/L).
4.2.1 At baseline
Armijos 2010 MEX 17 1 (0.5) 16 1 (0.5) 5.02% -0.01[-0.33,0.31]
Karyadi 2002 IDN 40 0.8 (0.3) 40 0.9 (0.3) 43.15% -0.08[-0.19,0.03]
Ginawi 2013 IND 41 0.8 (0.2) 20 0.8 (0.3) 24.54% -0.04[-0.19,0.11]
Ginawi 2013 IND 47 0.8 (0.2) 21 0.8 (0.3) 27.28% -0.05[-0.19,0.09]
Subtotal *** 145 97 100% -0.06[-0.13,0.01]
4.2.2 At 2 months
Armijos 2010 MEX 17 1.4 (0.5) 16 1.3 (0.4) 8.5% 0.11[-0.17,0.39]
Karyadi 2002 IDN 40 1.1 (0.3) 40 1.1 (0.3) 42.58% 0.06[-0.07,0.19]
Ginawi 2013 IND 41 1.1 (0.3) 20 1.1 (0.3) 24.03% 0.01[-0.16,0.18]
Ginawi 2013 IND 47 1.2 (0.3) 21 1.1 (0.3) 24.88% 0.07[-0.09,0.23]
Subtotal *** 145 97 100% 0.05[-0.03,0.14]
4.2.3 At 6 months
Armijos 2010 MEX 17 1.6 (0.5) 16 1.6 (0.4) 7.27% 0[-0.33,0.33]
Karyadi 2002 IDN 40 1.4 (0.4) 40 1.3 (0.3) 33.11% 0.12[-0.03,0.27]
Ginawi 2013 IND 41 1.6 (0.2) 20 1.6 (0.3) 30.33% 0.03[-0.13,0.19]
Ginawi 2013 IND 47 1.7 (0.3) 21 1.6 (0.3) 29.29% 0.08[-0.08,0.24]
Subtotal *** 145 97 100% 0.07[-0.02,0.16]
Favours Placebo -0.5 -0.25 0 0.25 0.5 Favours Vitamin A
Analysis 4.3. Comparison 4 Vitamin A versus placebo, Outcome 3 Death.

Study or subgroup Vitamin A Placebo Risk Ratio Weight Risk Ratio
4.3.1 Vitamin A alone
Pakasi 2010 IDN 3/72 1/43 0.44% 1.79[0.19,16.69]
Subtotal (95% CI) 72 43 0.44% 1.79[0.19,16.69]
Total events: 3 (Vitamin A), 1 (Placebo)
4.3.2 Vitamin A plus zinc

Armijos 2010 MEX 0/20 1/19 0.54% 0.32[0.01,7.35]
Lawson 2010 NGA 9/117 2/116 0.71% 4.46[0.99,20.21]
Pakasi 2010 IDN 2/66 1/43 0.43% 1.3[0.12,13.93]
Visser 2011 ZAF 1/77 0/77 0.18% 3[0.12,72.52]
Subtotal (95% CI) 280 255 1.85% 2.38[0.87,6.53]
Favours vitamin A 0.01 0.1 1 10 100 Favours placebo
Collaboration.
Informed decisions.

4.3.3 Vitamin A as part of a multi-micronutrient supplement

Lodha 2014 IND 4/202 0/100 0.23% 4.48[0.24,82.36]
Range 2005 TZA 21/213 16/108 7.47% 0.67[0.36,1.22]
Semba 2007 MWI 166/575 179/573 63.07% 0.92[0.77,1.1]
Villamor 2008 TZA 78/441 77/446 26.93% 1.02[0.77,1.36]
Subtotal (95% CI) 1431 1227 97.71% 0.94[0.81,1.09]
Total (95% CI) 1783 1525 100% 0.97[0.84,1.12]

Favours vitamin A 0.01 0.1 1 10 100 Favours placebo
Analysis 4.4. Comparison 4 Vitamin A versus placebo, Outcome 4 Treatment completion.

Pakasi 2010 IDN 59/72 77/86 100% 0.92[0.8,1.04]
Total (95% CI) 72 86 100% 0.92[0.8,1.04]

Favours placebo 0.5 0.7 1 1.5 2 Favours vitamin A
Analysis 4.5. Comparison 4 Vitamin A versus placebo, Outcome 5 Symptomatic at 6 weeks.

Study or subgroup Vitamin A Placebo Risk Ratio Risk Ratio
Hanekom 1997 ZAF 29/38 21/38 1.38[0.99,1.93]
Favours vitamin A 0.1 0.2 0.5 1 2 5 10 Favours placebo
Collaboration.
Informed decisions.

6 Sputum-smear and sputum-culture positive during follow-up.
4.6.1 Baseline
Pakasi 2010 IDN 72/72 86/86 100% 1[0.98,1.03]
Subtotal (95% CI) 72 86 100% 1[0.98,1.03]
4.6.2 2 weeks
Pakasi 2010 IDN 20/72 30/86 100% 0.8[0.5,1.28]
Subtotal (95% CI) 72 86 100% 0.8[0.5,1.28]
4.6.3 1 month
Pakasi 2010 IDN 9/66 16/82 100% 0.7[0.33,1.48]
Subtotal (95% CI) 66 82 100% 0.7[0.33,1.48]
4.6.4 2 months
Ginawi 2013 IND 3/31 3/32 100% 1.03[0.23,4.73]
Pakasi 2010 IDN 0/59 0/77 Not estimable
Subtotal (95% CI) 90 109 100% 1.03[0.23,4.73]
Favours Vitamin A 0.01 0.1 1 10 100 Favours Placebo
Analysis 4.7. Comparison 4 Vitamin A versus placebo, Outcome 7 BMI (kg/m2).

4.7.1 Baseline
Pakasi 2010 IDN 72 16.5 (2.2) 86 16.4 (2.5) 100% 0.1[-0.63,0.83]
Subtotal *** 72 86 100% 0.1[-0.63,0.83]
4.7.2 2 months
Pakasi 2010 IDN 66 17.8 (2.2) 82 17.5 (2.4) 100% 0.3[-0.44,1.04]
Subtotal *** 66 82 100% 0.3[-0.44,1.04]
Favours placebo -5 -2.5 0 2.5 5 Favours vitamin A
Collaboration.
Informed decisions.

4.7.3 6 months
Pakasi 2010 IDN 59 18.1 (2.4) 77 18.4 (2.6) 100% -0.3[-1.15,0.55]
Subtotal *** 59 77 100% -0.3[-1.15,0.55]
Favours placebo -5 -2.5 0 2.5 5 Favours vitamin A
Analysis 4.8. Comparison 4 Vitamin A versus placebo, Outcome 8 Body fat (%).
4.8.1 Baseline
Pakasi 2010 IDN 72 11 (6.2) 86 11.9 (5.9) 100% -0.9[-2.8,1]
Subtotal *** 72 86 100% -0.9[-2.8,1]
4.8.2 2 months
Pakasi 2010 IDN 66 13.3 (5.9) 82 14.2 (6.1) 100% -0.9[-2.84,1.04]
Subtotal *** 66 82 100% -0.9[-2.84,1.04]
4.8.3 6 months
Pakasi 2010 IDN 59 14.2 (6.4) 77 16 (6.3) 100% -1.8[-3.96,0.36]
Subtotal *** 59 77 100% -1.8[-3.96,0.36]
Favours placebo -10 -5 0 5 10 Favours vitamin A
Comparison 5. Zinc versus placebo

studies partici-
pants
1 Serum zinc level (normal range > 10.7 4 Mean Difference (IV, Fixed, 95% Subtotals only
µmol/L) CI)
1.1 At baseline 4 472 Mean Difference (IV, Fixed, 95% -0.03 [-0.36, 0.30]
CI)
1.2 At 2 months 4 452 Mean Difference (IV, Fixed, 95% 0.59 [0.24, 0.93]
CI)
1.3 At 6 months 4 440 Mean Difference (IV, Fixed, 95% 0.60 [0.31, 0.88]
CI)
Collaboration.
Informed decisions.

studies partici-
pants
2 Death by 6 to 8 months 7 2862 Risk Ratio (M-H, Fixed, 95% CI) 0.96 [0.82, 1.13]
2.1 Zinc alone 4 714 Risk Ratio (M-H, Fixed, 95% CI) 1.25 [0.70, 2.22]
2.2 Zinc plus vitamin A 4 477 Risk Ratio (M-H, Fixed, 95% CI) 2.08 [0.65, 6.64]
2.3 Zinc as part of a multi-micronutrient 3 1671 Risk Ratio (M-H, Fixed, 95% CI) 0.91 [0.77, 1.07]
supplement
3 Death by 6 to 8 months (subgrouped by 4 815 Risk Ratio (M-H, Fixed, 95% CI) 1.25 [0.75, 2.09]
HIV status)
3.1 HIV-negative individuals 3 442 Risk Ratio (M-H, Fixed, 95% CI) 1.92 [0.53, 6.98]
3.2 HIV-positive individuals 2 211 Risk Ratio (M-H, Fixed, 95% CI) 1.06 [0.59, 1.91]
3.3 HIV status unknown 1 162 Risk Ratio (M-H, Fixed, 95% CI) 1.70 [0.29, 9.89]
4 Treatment completion at 6 months 2 353 Risk Ratio (M-H, Fixed, 95% CI) 0.98 [0.92, 1.04]
5 Sputum-smear or sputum-culture posi- 5 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
tive during follow-up
5.1 At baseline 3 596 Risk Ratio (M-H, Fixed, 95% CI) 0.97 [0.92, 1.03]
5.2 At 2 weeks 3 806 Risk Ratio (M-H, Fixed, 95% CI) 1.09 [1.02, 1.17]
6 Clearance of chest X-ray at 6 months 1 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
7 Weight at follow-up 2 Mean Difference (IV, Fixed, 95% Totals not selected
CI)
7.1 Baseline: weight (kg) 2 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
7.2 2 months: weight (kg) 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
CI)
CI)
8 BMI (kg/m2) 1 Mean Difference (IV, Fixed, 95% Subtotals only

CI)
8.1 Baseline 1 162 Mean Difference (IV, Fixed, 95% 0.10 [-0.62, 0.82]
CI)
Collaboration.
Informed decisions.

studies partici-
pants
8.2 2 months 1 151 Mean Difference (IV, Fixed, 95% 0.20 [-0.55, 0.95]
CI)
8.3 6 months 1 140 Mean Difference (IV, Fixed, 95% 0.10 [-0.70, 0.90]
CI)
9 Body fat (%) 1 Mean Difference (IV, Fixed, 95% Subtotals only
CI)
9.1 Baseline 1 162 Mean Difference (IV, Fixed, 95% -0.90 [-2.51, 0.71]
CI)
9.2 2 months 1 151 Mean Difference (IV, Fixed, 95% -1.30 [-3.07, 0.47]
CI)
9.3 6 months 1 140 Mean Difference (IV, Fixed, 95% -1.5 [-3.51, 0.51]
CI)
10 Weight-for-age z score 1 Mean Difference (IV, Fixed, 95% Totals not selected
CI)
10.1 Baseline: weight-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
10.2 2 months: weight-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
10.3 6 months: weight-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
11 BMI-for-age z score 1 Mean Difference (IV, Fixed, 95% Totals not selected
CI)
11.1 Baseline: BMI-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
11.2 2 months: BMI-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
11.3 6 months: BMI-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
12 Height-for-age z score at follow-up 1 Mean Difference (IV, Fixed, 95% Totals not selected
CI)
12.1 Baseline: height-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
12.2 2 months: height-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
12.3 6 months: height-for-age z score 1 Mean Difference (IV, Fixed, 95% 0.0 [0.0, 0.0]
CI)
Collaboration.
Informed decisions.
Analysis 5.1. Comparison 5 Zinc versus placebo, Outcome 1 Serum zinc level (normal range > 10.7 µmol/L).
Study or subgroup Zinc ± Vitamin A Placebo Mean Difference Weight Mean Difference
5.1.1 At baseline
Ginawi 2013 IND 49 9.9 (0.9) 21 9.9 (1.4) 26.1% -0.02[-0.67,0.63]
Pakasi 2010 IDN 76 11.6 (2.2) 43 11.8 (2.4) 14.37% -0.2[-1.07,0.67]
Armijos 2010 MEX 17 11.3 (2.6) 16 11.7 (2.1) 4.28% -0.4[-2,1.2]
Karyadi 2002 IDN 40 11.5 (1.6) 40 11.2 (1.8) 19.51% 0.37[-0.38,1.12]
Ginawi 2013 IND 41 9.6 (0.8) 20 9.9 (1.4) 25.25% -0.32[-0.98,0.34]
Pakasi 2010 IDN 66 12.1 (3) 43 11.8 (2.4) 10.5% 0.3[-0.72,1.32]
Subtotal *** 289 183 100% -0.03[-0.36,0.3]
5.1.2 At 2 months
Ginawi 2013 IND 49 12.2 (1.1) 21 11.2 (1.3) 28.05% 1[0.35,1.65]
Pakasi 2010 IDN 69 11.7 (2.4) 41 11.7 (2.3) 14.29% 0[-0.92,0.92]
Armijos 2010 MEX 17 12.9 (3.4) 16 10.1 (1.6) 3.7% 2.72[0.92,4.52]
Karyadi 2002 IDN 40 11.2 (2.4) 40 10.2 (2.3) 11.37% 1[-0.03,2.03]
Ginawi 2013 IND 41 11.2 (0.9) 20 11.2 (1.3) 28.9% -0.06[-0.7,0.58]
Pakasi 2010 IDN 57 12.5 (2.3) 41 11.7 (2.3) 13.69% 0.8[-0.14,1.74]
Subtotal *** 273 179 100% 0.59[0.24,0.93]
5.1.3 At 6 months
Ginawi 2013 IND 49 14 (0.8) 21 12.9 (1) 34.33% 1.12[0.63,1.61]
Pakasi 2010 IDN 63 13.1 (1.9) 39 12.6 (1.7) 16.14% 0.5[-0.22,1.22]
Armijos 2010 MEX 17 10.4 (1.5) 16 11.3 (2.1) 5.25% -0.9[-2.16,0.36]
Karyadi 2002 IDN 40 12.5 (2.2) 40 12.1 (2) 10% 0.37[-0.55,1.29]
Ginawi 2013 IND 41 13.3 (1.2) 20 12.9 (1) 24.75% 0.49[-0.09,1.07]
Pakasi 2010 IDN 56 12.8 (2.9) 38 12.6 (1.7) 9.53% 0.2[-0.74,1.14]
Subtotal *** 266 174 100% 0.6[0.31,0.88]
Favours placebo -4 -2 0 2 4 Favours zinc ± vitamin A
Analysis 5.2. Comparison 5 Zinc versus placebo, Outcome 2 Death by 6 to 8 months.

Study or subgroup Zinc ± vitamin A Placebo Risk Ratio Weight Risk Ratio
5.2.1 Zinc alone
Lawson 2010 NGA 9/117 1/58 0.59% 4.46[0.58,34.38]
Lodha 2014 IND 1/101 0/100 0.22% 2.97[0.12,72.06]
Pakasi 2010 IDN 3/76 1/43 0.57% 1.7[0.18,15.82]
Range 2005 TZA 15/112 16/107 7.27% 0.9[0.47,1.72]
Subtotal (95% CI) 406 308 8.65% 1.25[0.7,2.22]
Total events: 28 (Zinc ± vitamin A), 18 (Placebo)
Favours zinc ± vitamin A 0.005 0.1 1 10 200 Favours Placebo
Collaboration.
Informed decisions.
Study or subgroup Zinc ± vitamin A Placebo Risk Ratio Weight Risk Ratio
5.2.2 Zinc plus vitamin A

Armijos 2010 MEX 0/20 1/19 0.68% 0.32[0.01,7.35]
Lawson 2010 NGA 9/117 1/58 0.59% 4.46[0.58,34.38]
Pakasi 2010 IDN 2/66 1/43 0.54% 1.3[0.12,13.93]
Visser 2011 ZAF 1/77 0/77 0.22% 3[0.12,72.52]
Subtotal (95% CI) 280 197 2.04% 2.08[0.65,6.64]
5.2.3 Zinc as part of a multi-micronutrient supplement

Lodha 2014 IND 2/102 0/100 0.22% 4.9[0.24,100.86]
Range 2005 TZA 21/213 16/108 9.43% 0.67[0.36,1.22]
Semba 2007 MWI 166/575 179/573 79.65% 0.92[0.77,1.1]
Subtotal (95% CI) 890 781 89.31% 0.91[0.77,1.07]
Total (95% CI) 1576 1286 100% 0.96[0.82,1.13]

Favours zinc ± vitamin A 0.005 0.1 1 10 200 Favours Placebo
Analysis 5.3. Comparison 5 Zinc versus placebo, Outcome 3 Death by 6 to 8 months (subgrouped by HIV status).
Study or subgroup Zinc Placebo Risk Ratio Weight Risk Ratio
Lawson 2010 NGA 3/58 1/58 4.59% 3[0.32,28]
Lodha 2014 IND 1/101 0/100 2.31% 2.97[0.12,72.06]
Range 2005 TZA 2/60 2/65 8.81% 1.08[0.16,7.45]
Subtotal (95% CI) 219 223 15.7% 1.92[0.53,6.98]
Total events: 6 (Zinc), 3 (Placebo)

Lawson 2010 NGA 6/59 1/58 4.63% 5.9[0.73,47.49]
Range 2005 TZA 13/52 14/42 71.06% 0.75[0.4,1.42]
Subtotal (95% CI) 111 100 75.69% 1.06[0.59,1.91]
Favours Zinc 0.01 0.1 1 10 100 Favours Placebo
Collaboration.
Informed decisions.

5.3.3 HIV status unknown

Pakasi 2010 IDN 3/76 2/86 8.61% 1.7[0.29,9.89]
Subtotal (95% CI) 76 86 8.61% 1.7[0.29,9.89]
Total (95% CI) 406 409 100% 1.25[0.75,2.09]

Favours Zinc 0.01 0.1 1 10 100 Favours Placebo
Analysis 5.4. Comparison 5 Zinc versus placebo, Outcome 4 Treatment completion at 6 months.
Lodha 2014 IND 98/101 97/100 59.31% 1[0.95,1.05]
Pakasi 2010 IDN 56/66 77/86 40.69% 0.95[0.84,1.07]
Total (95% CI) 167 186 100% 0.98[0.92,1.04]

Favours placebo 0.5 0.7 1 1.5 2 Favours zinc
Analysis 5.5. Comparison 5 Zinc versus placebo, Outcome 5

Sputum-smear or sputum-culture positive during follow-up.
5.5.1 At baseline
Lawson 2010 NGA 117/117 116/116 48.26% 1[0.98,1.02]
Lodha 2014 IND 38/101 44/100 18.24% 0.86[0.61,1.19]
Pakasi 2010 IDN 76/76 86/86 33.5% 1[0.98,1.02]
Subtotal (95% CI) 294 302 100% 0.97[0.92,1.03]
Heterogeneity: Tau2=0; Chi2=15.18, df=2(P=0); I2=86.83%
5.5.2 At 2 weeks
Lawson 2010 NGA 109/117 100/116 34.06% 1.08[0.99,1.18]
Pakasi 2010 IDN 35/76 30/86 9.55% 1.32[0.9,1.93]
Range 2005 TZA 172/201 170/210 56.39% 1.06[0.97,1.15]
Subtotal (95% CI) 394 412 100% 1.09[1.02,1.17]
Favours zinc 0.5 0.7 1 1.5 2 Favours placebo
Collaboration.
Informed decisions.

5.5.3 At 4 weeks
Lawson 2010 NGA 89/117 80/116 40.28% 1.1[0.94,1.29]
Pakasi 2010 IDN 19/76 16/82 7.72% 1.28[0.71,2.31]
Range 2005 TZA 99/190 107/202 52% 0.98[0.81,1.19]
Subtotal (95% CI) 383 400 100% 1.05[0.93,1.2]
5.5.4 At 8 weeks
Ginawi 2013 IND 3/35 3/32 4.19% 0.91[0.2,4.21]
Lawson 2010 NGA 35/117 35/116 46.95% 0.99[0.67,1.47]
Lodha 2014 IND 10/101 9/100 12.08% 1.1[0.47,2.59]
Range 2005 TZA 24/211 28/218 36.79% 0.89[0.53,1.48]
Subtotal (95% CI) 533 543 100% 0.96[0.72,1.28]
Favours zinc 0.5 0.7 1 1.5 2 Favours placebo
Analysis 5.6. Comparison 5 Zinc versus placebo, Outcome 6 Clearance of chest X-ray at 6 months.
Lodha 2014 IND 71/101 72/100 0% 0.98[0.82,1.16]
Favours placebo 0.1 0.2 0.5 1 2 5 10 Favours zinc
Analysis 5.7. Comparison 5 Zinc versus placebo, Outcome 7 Weight at follow-up.

Study or subgroup Zinc Placebo Mean Difference Mean Difference
5.7.1 Baseline: weight (kg)
Lodha 2014 IND 101 21.3 (8.7) 100 20.7 (8.1) 0.6[-1.72,2.92]
Range 2005 TZA 96 51.6 (1) 97 50.2 (8.5) 1.4[-0.3,3.1]
5.7.2 2 months: weight (kg)

Lodha 2014 IND 101 23.1 (9.2) 100 22.3 (8.8) 0.74[-1.75,3.23]

Lodha 2014 IND 101 25.2 (9.7) 100 24 (9.4) 1.26[-1.37,3.89]
Favours Placebo -10 -5 0 5 10 Favours Zinc
Collaboration.
Informed decisions.

Range 2005 TZA 96 6.1 (0.5) 97 6.3 (0.5) -0.21[-0.36,-0.06]
Favours Placebo -10 -5 0 5 10 Favours Zinc
Analysis 5.8. Comparison 5 Zinc versus placebo, Outcome 8 BMI (kg/m2).

Study or subgroup Zinc Placebo Mean Difference Weight Mean Difference
5.8.1 Baseline
Pakasi 2010 IDN 76 16.5 (2.2) 86 16.4 (2.5) 100% 0.1[-0.62,0.82]
Subtotal *** 76 86 100% 0.1[-0.62,0.82]
5.8.2 2 months
Pakasi 2010 IDN 69 17.7 (2.3) 82 17.5 (2.4) 100% 0.2[-0.55,0.95]
Subtotal *** 69 82 100% 0.2[-0.55,0.95]
5.8.3 6 months
Pakasi 2010 IDN 63 18.5 (2.2) 77 18.4 (2.6) 100% 0.1[-0.7,0.9]
Subtotal *** 63 77 100% 0.1[-0.7,0.9]
Favours placebo -5 -2.5 0 2.5 5 Favours zinc
Analysis 5.9. Comparison 5 Zinc versus placebo, Outcome 9 Body fat (%).
5.9.1 Baseline
Pakasi 2010 IDN 76 11 (4.5) 86 11.9 (5.9) 100% -0.9[-2.51,0.71]
Subtotal *** 76 86 100% -0.9[-2.51,0.71]
5.9.2 2 months
Pakasi 2010 IDN 69 12.9 (5) 82 14.2 (6.1) 100% -1.3[-3.07,0.47]
Subtotal *** 69 82 100% -1.3[-3.07,0.47]
5.9.3 6 months
Pakasi 2010 IDN 63 14.5 (5.8) 77 16 (6.3) 100% -1.5[-3.51,0.51]
Subtotal *** 63 77 100% -1.5[-3.51,0.51]
Favours placebo -10 -5 0 5 10 Favours zinc
Collaboration.
Informed decisions.

Analysis 5.10. Comparison 5 Zinc versus placebo, Outcome 10 Weight-for-age z score.

5.10.1 Baseline: weight-for-age z score
Lodha 2014 IND 101 -2.5 (1.5) 100 -2.8 (1.6) 0.3[-0.13,0.73]
5.10.2 2 months: weight-for-age z score

Lodha 2014 IND 101 -2.1 (1.4) 100 -2.4 (1.6) 0.3[-0.11,0.71]
5.10.3 6 months: weight-for-age z score

Lodha 2014 IND 101 -1.7 (1.2) 100 -2.1 (1.5) 0.37[-0.01,0.75]
Analysis 5.11. Comparison 5 Zinc versus placebo, Outcome 11 BMI-for-age z score.

5.11.1 Baseline: BMI-for-age z score
Lodha 2014 IND 101 -2.3 (1.6) 100 -2.5 (1.5) 0.18[-0.25,0.61]
5.11.2 2 months: BMI-for-age z score

Lodha 2014 IND 101 -2.3 (7.8) 100 -1.7 (1.3) -0.58[-2.12,0.96]
5.11.3 6 months: BMI-for-age z score

Lodha 2014 IND 101 -1.1 (1.1) 100 -1.3 (1.2) 0.2[-0.12,0.52]
Analysis 5.12. Comparison 5 Zinc versus placebo, Outcome 12 Height-for-age z score at follow-up.
5.12.1 Baseline: height-for-age z score
Lodha 2014 IND 101 -1.4 (1.3) 100 -1.6 (1.3) 0.2[-0.16,0.56]
5.12.2 2 months: height-for-age z score

Lodha 2014 IND 101 -1.4 (1.4) 100 -1.6 (1.4) 0.21[-0.17,0.59]
5.12.3 6 months: height-for-age z score

Lodha 2014 IND 101 -1.4 (1.3) 100 -1.6 (1.4) 0.28[-0.09,0.65]
Collaboration.
Informed decisions.
Comparison 6. Zinc plus vitamin A versus placebo

studies partici-
pants
1 Death by 6 months 4 578 Risk Ratio (M-H, Fixed, 95% CI) 2.33 [0.90, 6.07]
1.3 Unknown HIV status 1 12 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
2 Treatment completion at 6 1 152 Risk Ratio (M-H, Fixed, 95% CI) 0.95 [0.84, 1.07]
months
3 Sputum-smear and spu- 7 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
tum-culture positive during
follow-up
3.2 1 month 4 485 Risk Ratio (M-H, Fixed, 95% CI) 1.01 [0.86, 1.17]
4 Body weight (kg) 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
4.1 Baseline 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
4.2 2 months 1 Mean Difference (IV, Fixed, 95% CI) 0.0 [0.0, 0.0]
5 BMI (kg/m2) 2 664 Mean Difference (IV, Fixed, 95% CI) -0.09 [-0.45, 0.28]
5.1 Baseline 2 232 Mean Difference (IV, Fixed, 95% CI) -0.00 [-0.62, 0.61]
6 Mid upper arm circumfer- 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
ence (cm)
Collaboration.
Informed decisions.

studies partici-
pants
7 Biceps skinfold thickness 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
(mm)
8 Triceps skinfold thickness 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
(mm)
9 Subscapular skinfold thick- 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
ness (mm)
10 Suprailiac skinfold thick- 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
ness (mm)
11 Body fat (%) 2 Mean Difference (IV, Fixed, 95% CI) Totals not selected
12 Fat mass (kg) 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
Collaboration.
Informed decisions.

studies partici-
pants
13 Karnofsky score 1 Mean Difference (IV, Fixed, 95% CI) Totals not selected
Analysis 6.1. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 1 Death by 6 months.
Study or subgroup Zinc + Vitamin A Placebo Risk Ratio Weight Risk Ratio
Armijos 2010 MEX 0/20 1/19 26.79% 0.32[0.01,7.35]
Lawson 2010 NGA 2/59 1/58 17.58% 1.97[0.18,21.09]
Pakasi 2010 IDN 2/66 2/86 30.28% 1.3[0.19,9.01]
Visser 2011 ZAF 0/63 0/59 Not estimable
Subtotal (95% CI) 208 222 74.64% 1.11[0.31,3.99]
Total events: 4 (Zinc + Vitamin A), 4 (Placebo)

Lawson 2010 NGA 7/58 1/58 17.43% 7[0.89,55.11]
Visser 2011 ZAF 1/9 0/11 7.92% 3.6[0.16,79.01]
Subtotal (95% CI) 67 69 25.36% 5.94[1.07,32.96]

Visser 2011 ZAF 0/5 0/7 Not estimable
Total (95% CI) 280 298 100% 2.33[0.9,6.07]

Favours Zinc plus Vit A 0.01 0.1 1 10 100 Favours Placebo
Collaboration.
Informed decisions.
Analysis 6.2. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 2 Treatment completion at 6 months.
Pakasi 2010 IDN 56/66 77/86 100% 0.95[0.84,1.07]
Total (95% CI) 66 86 100% 0.95[0.84,1.07]

Favours placebo 0.5 0.7 1 1.5 2 Favours Zinc + Vitamin A
Analysis 6.3. Comparison 6 Zinc plus vitamin A versus placebo, Outcome

3 Sputum-smear and sputum-culture positive during follow-up.
6.3.1 At baseline
Armijos 2010 MEX 17/17 16/16 5.19% 1[0.89,1.12]
Karyadi 2002 IDN 40/40 40/40 12.38% 1[0.95,1.05]
Lawson 2010 NGA 117/117 116/116 35.75% 1[0.98,1.02]
Pakasi 2010 IDN 66/66 86/86 23% 1[0.97,1.03]
Visser 2011 ZAF 77/77 77/77 23.68% 1[0.98,1.03]
Subtotal (95% CI) 317 335 100% 1[0.99,1.01]
Heterogeneity: Tau2=0; Chi2=0, df=4(P=1); I2=0%
6.3.2 1 month
Armijos 2010 MEX 10/17 10/16 8.46% 0.94[0.54,1.63]
Karyadi 2002 IDN 14/40 18/40 14.78% 0.78[0.45,1.34]
Lawson 2010 NGA 91/117 80/116 65.98% 1.13[0.97,1.32]
Pakasi 2010 IDN 7/57 16/82 10.78% 0.63[0.28,1.43]
Subtotal (95% CI) 231 254 100% 1.01[0.86,1.17]
6.3.3 2 months
Armijos 2010 MEX 4/17 8/16 9.5% 0.47[0.18,1.26]
Ginawi 2013 IND 2/29 3/32 3.29% 0.74[0.13,4.1]
Karyadi 2002 IDN 0/40 0/40 Not estimable
Lawson 2010 NGA 41/117 35/116 40.52% 1.16[0.8,1.68]
Singh 2013 IND 0/13 2/13 2.88% 0.2[0.01,3.8]
Visser 2011 ZAF 31/77 38/77 43.81% 0.82[0.57,1.16]
Subtotal (95% CI) 350 376 100% 0.9[0.71,1.15]
6.3.4 3 months
Favours Zinc + Vitamin A 0.005 0.1 1 10 200 Favours placebo
Collaboration.
Informed decisions.
Armijos 2010 MEX 1/17 8/16 26.3% 0.12[0.02,0.84]
Lawson 2010 NGA 29/117 23/116 73.7% 1.25[0.77,2.03]
Subtotal (95% CI) 134 132 100% 0.95[0.61,1.49]
6.3.5 4 months
Armijos 2010 MEX 1/17 3/16 18.02% 0.31[0.04,2.71]
Lawson 2010 NGA 25/117 14/116 81.98% 1.77[0.97,3.23]
Subtotal (95% CI) 134 132 100% 1.51[0.86,2.65]
6.3.6 5 months
Armijos 2010 MEX 0/17 0/16 Not estimable
6.3.7 6 months
Armijos 2010 MEX 0/17 0/16 Not estimable
Karyadi 2002 IDN 0/40 2/40 100% 0.2[0.01,4.04]
Subtotal (95% CI) 57 56 100% 0.2[0.01,4.04]
Favours Zinc + Vitamin A 0.005 0.1 1 10 200 Favours placebo
Analysis 6.4. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 4 Body weight (kg).
Study or subgroup Zinc + Vitamin A Placebo Mean Difference Mean Difference
6.4.1 Baseline
Karyadi 2002 IDN 40 44.7 (5.1) 40 43.6 (5.7) 1.1[-1.26,3.46]
6.4.2 2 months
Karyadi 2002 IDN 40 46.9 (5.7) 40 45.8 (5.7) 1.1[-1.39,3.59]
6.4.3 6 months
Karyadi 2002 IDN 40 51.6 (5.1) 40 48.5 (5.7) 3.1[0.74,5.46]
Favours placebo -10 -5 0 5 10 Favours Zinc + Vitamin A
Collaboration.
Informed decisions.
Analysis 6.5. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 5 BMI (kg/m2).
Study or subgroup Zinc + Vitamin A Placebo Mean Difference Weight Mean Difference
6.5.1 Baseline
Karyadi 2002 IDN 40 17.6 (1.9) 40 18.1 (3.2) 10.22% -0.5[-1.64,0.64]
Pakasi 2010 IDN 66 16.6 (2.1) 86 16.4 (2.5) 24.9% 0.2[-0.53,0.93]
Subtotal *** 106 126 35.12% -0[-0.62,0.61]
6.5.2 2 months
Karyadi 2002 IDN 40 18.4 (1.9) 40 19 (3.2) 10.22% -0.6[-1.74,0.54]
Pakasi 2010 IDN 57 17.7 (2) 82 17.5 (2.4) 24.74% 0.2[-0.53,0.93]
Subtotal *** 97 122 34.96% -0.03[-0.65,0.58]
6.5.3 6 months
Karyadi 2002 IDN 40 19.4 (2.5) 40 20 (3.2) 8.48% -0.6[-1.85,0.65]
Pakasi 2010 IDN 56 18.3 (2) 77 18.4 (2.6) 21.44% -0.1[-0.89,0.69]
Subtotal *** 96 117 29.92% -0.24[-0.91,0.43]
Total *** 299 365 100% -0.09[-0.45,0.28]

Favours Placebo -2 -1 0 1 2 Favours Zinc + Vitamin A
Analysis 6.6. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 6 Mid upper arm circumference (cm).
6.6.1 Baseline
Karyadi 2002 IDN 40 22.8 (1.9) 40 21.8 (3.8) 1[-0.31,2.31]
6.6.2 2 months
Karyadi 2002 IDN 40 23.4 (2.6) 40 22.8 (3.8) 0.6[-0.82,2.02]
6.6.3 6 months
Karyadi 2002 IDN 40 25.2 (1.9) 40 24 (3.8) 1.2[-0.11,2.51]
Collaboration.
Informed decisions.
Analysis 6.7. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 7 Biceps skinfold thickness (mm).
6.7.1 Baseline
Karyadi 2002 IDN 40 4.2 (1.3) 40 4.2 (1.9) 0[-0.71,0.71]
6.7.2 2 months
Karyadi 2002 IDN 40 4.9 (1.9) 40 4.9 (1.9) 0[-0.83,0.83]
6.7.3 6 months
Karyadi 2002 IDN 40 5.7 (3.2) 40 5.5 (2.5) 0.2[-1.05,1.45]
Analysis 6.8. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 8 Triceps skinfold thickness (mm).
6.8.1 Baseline
Karyadi 2002 IDN 40 5.8 (3.8) 40 5.5 (3.2) 0.3[-1.23,1.83]
6.8.2 2 months
Karyadi 2002 IDN 40 6.7 (4.4) 40 6.6 (0.1) 0.1[-1.27,1.47]
6.8.3 6 months
Karyadi 2002 IDN 40 8 (5.1) 40 7.5 (3.8) 0.5[-1.46,2.46]
Analysis 6.9. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 9 Subscapular skinfold thickness (mm).
6.9.1 Baseline
Karyadi 2002 IDN 40 6.6 (1.9) 40 7 (2.5) -0.4[-1.38,0.58]
6.9.2 2 months
Karyadi 2002 IDN 40 7.5 (2.5) 40 7.7 (2.5) -0.2[-1.31,0.91]
6.9.3 6 months
Karyadi 2002 IDN 40 8.6 (3.8) 40 8.6 (3.2) 0[-1.53,1.53]
Analysis 6.10. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 10 Suprailiac skinfold thickness (mm).
6.10.1 Baseline
Collaboration.
Informed decisions.

Karyadi 2002 IDN 40 5.2 (2.5) 40 5.7 (3.2) -0.5[-1.75,0.75]
6.10.2 2 months
Karyadi 2002 IDN 40 6.2 (3.2) 40 6.7 (3.2) -0.5[-1.88,0.88]
6.10.3 6 months
Karyadi 2002 IDN 40 7.2 (3.8) 40 7.8 (4.4) -0.6[-2.41,1.21]
Analysis 6.11. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 11 Body fat (%).
6.11.1 Baseline
Karyadi 2002 IDN 40 11.5 (6.3) 40 13 (6.3) -1.5[-4.27,1.27]
Pakasi 2010 IDN 66 10.8 (4.7) 86 11.9 (5.9) -1.1[-2.79,0.59]
6.11.2 2 months
Karyadi 2002 IDN 40 13.1 (7) 40 14.6 (6.3) -1.5[-4.41,1.41]
Pakasi 2010 IDN 57 12.9 (5.1) 82 14.2 (6.1) -1.3[-3.17,0.57]
6.11.3 6 months
Karyadi 2002 IDN 40 14.2 (7.5) 40 16 (6.3) -1.8[-4.84,1.24]
Pakasi 2010 IDN 56 14.2 (6.4) 77 16 (6.3) -1.8[-4,0.4]
Analysis 6.12. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 12 Fat mass (kg).
6.12.1 Baseline
Karyadi 2002 IDN 40 5.1 (3.2) 40 5.7 (3.2) -0.6[-1.98,0.78]
6.12.2 2 months
Karyadi 2002 IDN 40 6.1 (3.2) 40 6.7 (3.2) -0.6[-1.98,0.78]
6.12.3 6 months
Karyadi 2002 IDN 40 6.9 (3.2) 40 7.7 (3.2) -0.8[-2.18,0.58]
Collaboration.
Informed decisions.
Analysis 6.13. Comparison 6 Zinc plus vitamin A versus placebo, Outcome 13 Karnofsky score.
6.13.1 Baseline
Karyadi 2002 IDN 40 80.8 (3.8) 40 82 (4.4) -1.2[-3.01,0.61]
6.13.2 2 months
Karyadi 2002 IDN 40 90 (3.8) 40 90.2 (1.9) -0.2[-1.51,1.11]
6.13.3 6 months
Karyadi 2002 IDN 40 97.8 (0.8) 40 95.3 (5.1) 2.5[0.91,4.09]
Comparison 7. Vitamin D versus placebo or no supplement

studies partici-
pants
1 Serum vitamin D levels (nmol/L) 5 Mean Difference (IV, Random, 95% Subtotals only
CI)
1.1 At baseline 5 894 Mean Difference (IV, Random, 95% -0.10 [-3.01, 2.82]
CI)
1.2 At 8 weeks 3 460 Mean Difference (IV, Random, 95% 33.68 [3.99, 63.37]
CI)
1.3 At 6 to 8 months 2 376 Mean Difference (IV, Random, 95% 7.90 [0.52, 15.28]
CI)
2 Death during follow-up (2 to 12 7 2649 Risk Ratio (M-H, Fixed, 95% CI) 0.96 [0.81, 1.12]
months)
2.1 Vitamin D alone 4 814 Risk Ratio (M-H, Fixed, 95% CI) 1.26 [0.79, 2.02]
2.2 Vitamin D plus arginine 1 75 Risk Ratio (M-H, Fixed, 95% CI) 1.53 [0.06, 36.25]
2.3 Vitamin D as part of a multi-mi- 3 1760 Risk Ratio (M-H, Fixed, 95% CI) 0.91 [0.77, 1.08]
cronutrient supplement
3 Death during follow-up (2 to 12 7 2649 Risk Ratio (M-H, Fixed, 95% CI) 0.97 [0.83, 1.13]
months)
3.3 HIV status mixed or unknown 2 157 Risk Ratio (M-H, Fixed, 95% CI) 1.5 [0.30, 7.38]
4 Cure at 6 months 1 151 Risk Ratio (M-H, Fixed, 95% CI) 0.99 [0.75, 1.31]
5 Tuberculosis score 1 1142 Mean Difference (IV, Fixed, 95% CI) -0.13 [-0.32, 0.06]
Collaboration.
Informed decisions.

studies partici-
pants
6 Sputum-smear or sputum-cul- 7 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
ture positive
7 Body mass index 4 Mean Difference (IV, Fixed, 95% CI) Subtotals only
7.2 At 6 to 8 weeks 4 430 Mean Difference (IV, Fixed, 95% CI) -0.19 [-0.70, 0.32]
8 Body weight (kg) 2 Mean Difference (IV, Fixed, 95% CI) Subtotals only
8.1 Before treatment 2 150 Mean Difference (IV, Fixed, 95% CI) 0.20 [-2.43, 2.83]
9 Karnofsky score at 8 weeks 1 Mean Difference (IV, Fixed, 95% CI) Subtotals only
Analysis 7.1. Comparison 7 Vitamin D versus placebo or no

supplement, Outcome 1 Serum vitamin D levels (nmol/L).
Study or subgroup Vitamin D Placebo Mean Difference Weight Mean Difference
7.1.1 At baseline
Favours placebo -100 -50 0 50 100 Favours vitamin D
Collaboration.
Informed decisions.

Daley 2015 IND 121 63.1 (46.6) 126 62.2 (51) 5.73% 0.9[-11.28,13.08]
Kota 2011 IND 15 31.9 (11.2) 15 27.7 (11.7) 12.65% 4.2[-4,12.4]
Martineau 2011 GBR 62 21.1 (20) 64 21.3 (19) 18.3% -0.2[-7.02,6.62]
Mily 2015 BGD 62 28 (17.5) 64 28.1 (16.2) 24.48% -0.1[-5.99,5.79]
Wejse 2008 GNB 187 77.5 (23.8) 178 79.1 (21.8) 38.83% -1.6[-6.28,3.08]
Subtotal *** 447 447 100% -0.1[-3.01,2.82]
7.1.2 At 8 weeks
Kota 2011 IND 15 55.7 (14.7) 15 27.7 (4) 35.62% 28[20.29,35.71]
Martineau 2011 GBR 62 101.4 (96.1) 64 22.8 (20.3) 29.27% 78.6[54.17,103.03]
Wejse 2008 GNB 157 105 (38.4) 147 103 (47.6) 35.11% 2[-7.76,11.76]
Subtotal *** 234 226 100% 33.68[3.99,63.37]
Heterogeneity: Tau2=628.75; Chi2=39.04, df=2(P<0.0001); I2=94.88%
7.1.3 At 6 to 8 months
Daley 2015 IND 65 72.2 (53.3) 77 60.4 (49.5) 18.79% 11.79[-5.25,28.83]
Wejse 2008 GNB 125 102 (28.5) 109 95 (34.6) 81.21% 7[-1.19,15.19]
Subtotal *** 190 186 100% 7.9[0.52,15.28]

supplement, Outcome 2 Death during follow-up (2 to 12 months).
Study or subgroup Vitamin D Placebo Risk Ratio Weight Risk Ratio
7.2.1 Vitamin D alone
Daley 2015 IND 3/121 1/126 0.43% 3.12[0.33,29.62]
Martineau 2011 GBR 1/62 1/64 0.43% 1.03[0.07,16.14]
Ralph 2013 IDN 1/51 0/25 0.29% 1.5[0.06,35.56]
Wejse 2008 GNB 30/187 24/178 10.75% 1.19[0.72,1.95]
Subtotal (95% CI) 421 393 11.9% 1.26[0.79,2.02]
Total events: 35 (Vitamin D), 26 (Placebo)
7.2.2 Vitamin D plus arginine

Ralph 2013 IDN 1/50 0/25 0.29% 1.53[0.06,36.25]
Subtotal (95% CI) 50 25 0.29% 1.53[0.06,36.25]
7.2.3 Vitamin D as part of a multi-micronutrient supplement
Favours Vitamin D 0.01 0.1 1 10 100 Favours Placebo
Collaboration.
Informed decisions.

Lodha 2014 IND 4/202 0/100 0.29% 4.48[0.24,82.36]
Range 2005 TZA 21/203 16/107 9.16% 0.69[0.38,1.27]
Semba 2007 MWI 166/575 179/573 78.36% 0.92[0.77,1.1]
Subtotal (95% CI) 980 780 87.81% 0.91[0.77,1.08]
Total (95% CI) 1451 1198 100% 0.96[0.81,1.12]


supplement, Outcome 3 Death during follow-up (2 to 12 months).
Range 2005 TZA 15/87 14/42 8.34% 0.52[0.28,0.97]
Semba 2007 MWI 157/406 171/423 74.01% 0.96[0.81,1.13]
Wejse 2008 GNB 21/72 15/59 7.29% 1.15[0.65,2.02]
Subtotal (95% CI) 565 524 89.63% 0.93[0.8,1.09]

Daley 2015 IND 3/121 1/126 0.43% 3.12[0.33,29.62]
Lodha 2014 IND 4/202 0/100 0.3% 4.48[0.24,82.36]
Martineau 2011 GBR 1/62 1/64 0.43% 1.03[0.07,16.14]
Range 2005 TZA 6/116 2/65 1.13% 1.68[0.35,8.09]
Semba 2007 MWI 9/169 8/150 3.75% 1[0.4,2.52]
Wejse 2008 GNB 7/111 8/117 3.44% 0.92[0.35,2.46]
Subtotal (95% CI) 781 622 9.48% 1.26[0.72,2.21]
7.3.3 HIV status mixed or unknown

Ralph 2013 IDN 2/101 0/50 0.29% 2.5[0.12,51.11]
Wejse 2008 GNB 2/4 1/2 0.59% 1[0.18,5.46]
Subtotal (95% CI) 105 52 0.88% 1.5[0.3,7.38]
Collaboration.
Informed decisions.

Total (95% CI) 1451 1198 100% 0.97[0.83,1.13]
Analysis 7.4. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 4 Cure at 6 months.
Ralph 2013 IDN 29/51 15/25 50.16% 0.95[0.64,1.41]
Ralph 2013 IDN 31/50 15/25 49.84% 1.03[0.7,1.52]
Total (95% CI) 101 50 100% 0.99[0.75,1.31]

Favours placebo 0.01 0.1 1 10 100 Favours vitamin D
Analysis 7.5. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 5 Tuberculosis score.
7.5.1 At baseline
Wejse 2008 GNB 72 6.9 (2.1) 58 6.8 (2.1) 6.75% 0.1[-0.63,0.83]
Wejse 2008 GNB 104 6.5 (1.9) 114 6.7 (1.9) 13.95% -0.2[-0.7,0.3]
Subtotal *** 176 172 20.7% -0.1[-0.52,0.31]
7.5.2 2 months
Wejse 2008 GNB 94 1.9 (1.7) 97 2.1 (2) 12.87% -0.2[-0.73,0.33]
Wejse 2008 GNB 57 2.8 (2.3) 49 2.4 (2) 5.31% 0.4[-0.42,1.22]
Subtotal *** 151 146 18.17% -0.02[-0.47,0.42]
7.5.3 5 months
Wejse 2008 GNB 88 1 (1.4) 91 1.2 (1.3) 22.67% -0.2[-0.6,0.2]
Wejse 2008 GNB 49 1.6 (1.6) 43 1.6 (2.2) 5.62% 0[-0.8,0.8]
Subtotal *** 137 134 28.29% -0.16[-0.51,0.19]
7.5.4 8 months
Wejse 2008 GNB 81 0.7 (1) 74 1 (1.4) 23.85% -0.3[-0.69,0.09]
Favours vitamin D -2 -1 0 1 2 Favours placebo
Collaboration.
Informed decisions.

Wejse 2008 GNB 39 1.3 (1.4) 32 1.2 (1.3) 8.99% 0.1[-0.53,0.73]
Subtotal *** 120 106 32.83% -0.19[-0.52,0.14]
Total *** 584 558 100% -0.13[-0.32,0.06]

Favours vitamin D -2 -1 0 1 2 Favours placebo

supplement, Outcome 6 Sputum-smear or sputum-culture positive.
7.6.1 At baseline
Daley 2015 IND 101/121 108/126 24.76% 0.97[0.88,1.08]
Mily 2015 BGD 63/72 65/72 15.21% 0.97[0.86,1.09]
Nursyam 2006 IDN 34/34 33/33 7.95% 1[0.94,1.06]
Tukvadze 2015 GEO 97/100 95/99 22.34% 1.01[0.96,1.07]
Wejse 2008 GNB 123/187 124/178 29.73% 0.94[0.82,1.09]
Subtotal (95% CI) 514 508 100% 0.97[0.92,1.03]
7.6.2 4 weeks
Daley 2015 IND 41/121 46/126 25.13% 0.93[0.66,1.3]
Mily 2015 BGD 24/62 37/64 20.3% 0.67[0.46,0.98]
Ralph 2013 IDN 16/50 8/25 5.95% 1[0.5,2.01]
Ralph 2013 IDN 15/51 8/25 5.99% 0.92[0.45,1.87]
Tukvadze 2015 GEO 46/97 45/95 25.35% 1[0.74,1.35]
Wejse 2008 GNB 25/110 30/103 17.28% 0.78[0.49,1.23]
Subtotal (95% CI) 491 438 100% 0.87[0.74,1.03]
7.6.3 6 weeks
Daley 2015 IND 24/121 26/126 35.19% 0.96[0.59,1.58]
Nursyam 2006 IDN 0/34 8/33 11.91% 0.06[0,0.95]
Tukvadze 2015 GEO 24/97 28/95 39.08% 0.84[0.53,1.34]
Wejse 2008 GNB 6/75 10/75 13.81% 0.6[0.23,1.57]
Subtotal (95% CI) 327 329 100% 0.76[0.55,1.03]
Collaboration.
Informed decisions.

7.6.4 8 weeks
Daley 2015 IND 15/121 16/126 32.61% 0.98[0.51,1.89]
Mily 2015 BGD 1/62 7/64 14.33% 0.15[0.02,1.16]
Nursyam 2006 IDN 0/34 1/33 3.17% 0.32[0.01,7.68]
Singh 2013 IND 1/11 2/13 3.81% 0.59[0.06,5.68]
Tukvadze 2015 GEO 13/97 15/95 31.53% 0.85[0.43,1.69]
Wejse 2008 GNB 8/100 7/100 14.56% 1.14[0.43,3.03]
Subtotal (95% CI) 425 431 100% 0.81[0.54,1.2]
7.6.5 5 months
Wejse 2008 GNB 3/74 1/74 100% 3[0.32,28.18]
Subtotal (95% CI) 74 74 100% 3[0.32,28.18]
7.6.6 6 months
Daley 2015 IND 12/121 17/126 100% 0.74[0.37,1.47]
Subtotal (95% CI) 121 126 100% 0.74[0.37,1.47]
7.6.7 8 months
Wejse 2008 GNB 0/74 0/73 Not estimable
Analysis 7.7. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 7 Body mass index.
7.7.1 At baseline
Daley 2015 IND 121 18 (2.9) 126 17.8 (2.9) 47.47% 0.18[-0.55,0.91]
Martineau 2011 GBR 62 20.1 (3.1) 64 20.2 (2.7) 24.5% -0.1[-1.12,0.92]
Nursyam 2006 IDN 34 16.9 (2.1) 33 17.7 (2.5) 20.57% -0.81[-1.92,0.3]
Singh 2013 IND 11 16.4 (2.1) 13 17.7 (2.5) 7.47% -1.33[-3.17,0.51]
Subtotal *** 228 236 100% -0.2[-0.71,0.3]
7.7.2 At 6 to 8 weeks
Favours Placebo -4 -2 0 2 4 Favours Vitamin D
Collaboration.
Informed decisions.

Daley 2015 IND 104 18.8 (3.2) 109 18.6 (3.1) 37.64% 0.18[-0.66,1.02]
Martineau 2011 GBR 62 21.3 (2.7) 64 21.2 (2.8) 28.83% 0.11[-0.85,1.07]
Nursyam 2006 IDN 34 17.9 (1.8) 33 18.8 (2.5) 24.53% -0.89[-1.93,0.15]
Singh 2013 IND 11 17.5 (1.8) 13 18.2 (2.5) 9% -0.77[-2.48,0.94]
Subtotal *** 211 219 100% -0.19[-0.7,0.32]
Test for subgroup differences: Chi2=0, df=1 (P=0.96), I2=0%
Favours Placebo -4 -2 0 2 4 Favours Vitamin D
Analysis 7.8. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 8 Body weight (kg).
7.8.1 Before treatment
Mily 2015 BGD 62 44.2 (9.4) 64 43.7 (7.4) 79.02% 0.5[-2.46,3.46]
Morcos 1998 EGY 12 14.6 (5.4) 12 15.5 (8.6) 20.98% -0.92[-6.66,4.82]
Subtotal *** 74 76 100% 0.2[-2.43,2.83]
7.8.2 At 8 weeks
Mily 2015 BGD 62 46.3 (9.5) 64 44.9 (7.6) 79.9% 1.4[-1.61,4.41]
Morcos 1998 EGY 12 17.2 (6.3) 12 17.4 (8.5) 20.1% -0.18[-6.18,5.82]
Subtotal *** 74 76 100% 1.08[-1.61,3.77]
Analysis 7.9. Comparison 7 Vitamin D versus placebo or no supplement, Outcome 9 Karnofsky score at 8 weeks.
7.9.1 At baseline
Daley 2015 IND 121 70.9 (8.2) 126 72.4 (7.4) 100% -1.47[-3.42,0.48]
Subtotal *** 121 126 100% -1.47[-3.42,0.48]
7.9.2 At 8 weeks
Daley 2015 IND 103 83.8 (7.7) 109 82.9 (8.5) 100% 0.85[-1.33,3.03]
Subtotal *** 103 109 100% 0.85[-1.33,3.03]
Favours placebo -100 -50 0 50 100 Favours Vitamin D
Collaboration.
Informed decisions.
Comparison 8. Arginine versus placebo

studies partici-
pants
1 Death during treatment 3 394 Risk Ratio (M-H, Fixed, 95% CI) 0.67 [0.21, 2.09]
1.1 HIV-negative 1 56 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
1.3 HIV-negative and positive 2 279 Risk Ratio (M-H, Fixed, 95% CI) 0.5 [0.13, 1.94]
2 Cured at 6/8 months 2 279 Risk Ratio (M-H, Fixed, 95% CI) 1.05 [0.90, 1.22]
3 Sputum-smear or sputum-culture 4 Risk Ratio (M-H, Random, 95% CI) Subtotals only
positive
3.1 At baseline 4 464 Risk Ratio (M-H, Random, 95% CI) 1.0 [0.98, 1.02]
3.2 At 4 weeks 2 162 Risk Ratio (M-H, Random, 95% CI) 0.76 [0.51, 1.12]
3.3 At 8 weeks 3 351 Risk Ratio (M-H, Random, 95% CI) 0.76 [0.41, 1.42]
3.4 At 8 weeks (HIV-negative only) 1 56 Risk Ratio (M-H, Random, 95% CI) 0.12 [0.01, 2.07]
4 Cough 3 404 Risk Ratio (M-H, Fixed, 95% CI) 0.76 [0.62, 0.93]
5 Weight gain > 10% 1 Risk Ratio (M-H, Fixed, 95% CI) Subtotals only
5.1 HIV-positive and HIV-negative 1 170 Risk Ratio (M-H, Fixed, 95% CI) 1.17 [0.74, 1.84]
Analysis 8.1. Comparison 8 Arginine versus placebo, Outcome 1 Death during treatment.
Study or subgroup Arginine Placebo Risk Ratio Weight Risk Ratio
8.1.1 HIV-negative
Schön 2003 ETH 0/24 0/32 Not estimable
Total events: 0 (Arginine), 0 (Placebo)
8.1.2 HIV-positive
Schön 2003 ETH 2/33 1/26 15.71% 1.58[0.15,16.44]
Favours arginine 0.01 0.1 1 10 100 Favours placebo
Collaboration.
Informed decisions.

Subtotal (95% CI) 33 26 15.71% 1.58[0.15,16.44]
8.1.3 HIV-negative and positive

Ralph 2013 IDN 0/49 0/50 Not estimable
Schön 2011 ETH 3/90 6/90 84.29% 0.5[0.13,1.94]
Subtotal (95% CI) 139 140 84.29% 0.5[0.13,1.94]
Test for overall effect: Z=1(P=0.32)
Total (95% CI) 196 198 100% 0.67[0.21,2.09]

Analysis 8.2. Comparison 8 Arginine versus placebo, Outcome 2 Cured at 6/8 months.
Ralph 2013 IDN 28/49 30/50 31.36% 0.95[0.68,1.33]
Schön 2011 ETH 71/90 65/90 68.64% 1.09[0.92,1.29]
Total (95% CI) 139 140 100% 1.05[0.9,1.22]

Favours placebo 0.01 0.1 1 10 100 Favours arginine
Analysis 8.3. Comparison 8 Arginine versus placebo, Outcome 3 Sputum-smear or sputum-culture positive.
8.3.1 At baseline
Farazi 2015 IRN 32/32 31/31 6.67% 1[0.94,1.06]
Ralph 2013 IDN 50/50 51/51 16.84% 1[0.96,1.04]
Schön 2003 ETH 60/60 60/60 23.67% 1[0.97,1.03]
Schön 2011 ETH 90/90 90/90 52.82% 1[0.98,1.02]
Subtotal (95% CI) 232 232 100% 1[0.98,1.02]
Heterogeneity: Tau2=0; Chi2=0, df=3(P=1); I2=0%
Collaboration.
Informed decisions.

8.3.2 At 4 weeks
Farazi 2015 IRN 14/32 18/31 62.42% 0.75[0.46,1.23]
Ralph 2013 IDN 12/49 16/50 37.58% 0.77[0.41,1.45]
Subtotal (95% CI) 81 81 100% 0.76[0.51,1.12]
Heterogeneity: Tau2=0; Chi2=0, df=1(P=0.97); I2=0%
8.3.3 At 8 weeks
Farazi 2015 IRN 1/32 4/31 8.48% 0.24[0.03,2.05]
Schön 2003 ETH 5/57 8/58 34.66% 0.64[0.22,1.83]
Schön 2011 ETH 10/86 10/87 56.87% 1.01[0.44,2.31]
Subtotal (95% CI) 175 176 100% 0.76[0.41,1.42]
8.3.4 At 8 weeks (HIV-negative only)

Schön 2003 ETH 0/24 5/32 100% 0.12[0.01,2.07]
Subtotal (95% CI) 24 32 100% 0.12[0.01,2.07]
Analysis 8.4. Comparison 8 Arginine versus placebo, Outcome 4 Cough.

8.4.1 At 2 weeks
Schön 2003 ETH 16/24 30/32 23.98% 0.71[0.53,0.96]
Subtotal (95% CI) 24 32 23.98% 0.71[0.53,0.96]
8.4.2 At 8 weeks
Farazi 2015 IRN 3/32 5/31 4.74% 0.58[0.15,2.23]
Schön 2003 ETH 6/24 21/32 16.79% 0.38[0.18,0.8]
Schön 2003 ETH 13/33 12/26 12.52% 0.85[0.47,1.54]
Schön 2011 ETH 42/85 45/85 41.97% 0.93[0.7,1.25]
Subtotal (95% CI) 174 174 76.02% 0.78[0.61,0.99]
Total (95% CI) 198 206 100% 0.76[0.62,0.93]

Collaboration.
Informed decisions.

Analysis 8.5. Comparison 8 Arginine versus placebo, Outcome 5 Weight gain > 10%.
8.5.1 HIV-positive and HIV-negative
Schön 2011 ETH 28/85 24/85 100% 1.17[0.74,1.84]
Subtotal (95% CI) 85 85 100% 1.17[0.74,1.84]
8.5.2 HIV-positive
Schön 2011 ETH 16/36 6/29 100% 2.15[0.96,4.78]
Subtotal (95% CI) 36 29 100% 2.15[0.96,4.78]
Favours placebo 0.01 0.1 1 10 100 Favours arginine
Comparison 9. Vitamin E plus selenium versus placebo

studies partici-
pants
1 Sputum-smear positive at fol- 1 Risk Ratio (M-H, Fixed, 95% CI) Totals not selected
low-up
1.1 15 days 1 Risk Ratio (M-H, Fixed, 95% CI) 0.0 [0.0, 0.0]
Collaboration.
Informed decisions.
Analysis 9.1. Comparison 9 Vitamin E plus selenium versus

placebo, Outcome 1 Sputum-smear positive at follow-up.
Study or subgroup Vitamin E + selenium Placebo Risk Ratio Risk Ratio
9.1.1 15 days
Seyedrezazadeh 2006 IRN 14/17 15/18 0.99[0.73,1.34]
9.1.2 30 days
9.1.3 45 days
9.1.4 60 days
Favours Vitamin E + selenium 0.01 0.1 1 10 100 Favours placebo
ADDITIONAL TABLES
Table 1. Effect of multi-micronutrient supplementation on nutritional recovery

Trial ID Popula- Weight Baseline nutritional status Duration Comment
tion measure- of supple-
ment Supplement Placebo ment and
follow-up
Mehta HIV-pos- Weight MUAC (cm) MUAC (cm) Supple- The trial authors reported no effect of sup-
2011 TZA itive and gain me- ment and plement on median weight, height, or mid-
HIV-neg- dian (IQR) 13.0 (11.9 to 13.0 (11.2 follow-up 2 upper arm circumference at 2 months
ative chil- 14.25) to 14.0) months
dren
Weight: 8.48kg Weight kg:
(NR) 7.95 kg (NR)
Lodha HIV-neg- Change Weight-for- Weight-for- Supple- The trial authors reported no difference in
2014 IND ative chil- in weight- age z score age z score ment and weight (kg), weight-for-age z score, height-
dren for-age z (SD): (SD): −2.8 follow-up 6 for-age z score, or BMI z score after 2 or 6
score (1.6) months months supplementation with micronutri-
−2.72 (1.85, ents with or without zinc (1 trial, 302 partici-
MN+Z group); pants, Analysis 3.7)
−2.8 (1.6, MN
group)
Praygod HIV-pos- Mean Mean BMI mean BMI Supple- There was no statistically significant differ-
2011a TZA itive and weight (SD): 18.9 kg/ (SD): 18.9 ment for 2 ence in weight gain at 2 or 5 months. Sub-
1 negative gain (SD) m2 (2.8) kg/m2 (3.1) months group analysis by HIV status found a statisti-
adults cally significant difference in weight gain in
Follow-up favour of supplements in HIV-negative par-
at 2 and 5 ticipants, and in favour of placebo in HIV-
months positive participants (trial authors' own fig-
ures)
Villamor HIV-pos- Mean BMI HIV-negative HIV-nega- Supple- The trial authors reported no statistical-
2008 TZA itive and (SD) mean BMI tive mean ment for ly significant effect of supplement on BMI,
BMI (SD): 24 months mid-upper arm circumference, fat mass, or
Collaboration.
Informed decisions.
Table 1. Effect of multi-micronutrient supplementation on nutritional recovery (Continued)

negative (SD): 18.9 kg/ 18.9 kg/m2 and fol- fat free mass at 8 or 24 months, regardless of
adults m2 (2.5) (2.5) low-up at HIV status (no figures reported)
8 and 24
HIV-positive HIV-posi- months
mean BMI tive mean
(SD): 19.3 kg/ BMI (SD):
m2 (2.8) 19.6 kg/m2
(2.9)
Range HIV-pos- Mean Mean BMI Mean BMI Supple- There was a statistically significantly greater
2005 TZA 2 itive and weight (SD): 18.3 kg/ (SD): 18.7 ment for weight gain in supplemented group at 7
negative gain (SD) m2 (2.5) zinc + kg/m2 (2.7) 8 months months. While the difference in weight was
adults MMN group and fol- appreciable in both treatment (zinc + MMN
low-up 8 and MMN groups) arms compared to place-
weeks and bo, the weight gain in the zinc + MMN arm
7 months appeared to be clinically important
Abbreviations: SD = standard deviation; BMI = body mass index; MMN = multi-micronutrient; HIV = human immunodeficiency virus; IQR =
interquartile range; MUAC = mid-upper arm circumference; NR = not reported;
1The supplement administered by Praygod 2011a TZA included a similar dose of zinc to that used by Range 2005 TZA.
22 X 2 factorial design; Group 1: zinc plus placebo; Group 2: multivitamin and mineral tablet plus placebo; Group 3: zinc plus multivitamin
and mineral tablet; Group 4: placebo plus placebo.
Collaboration.
Collaboration.
Table 2. Effects of vitamin A supplementation on serum vitamin A levels
Trial ID Supplement dose Measure Baseline Follow-up Follow-up Comment
Library
Cochrane
Supple- Control Supple- Control
ment ment
Hanekom 200 000 IU vitamin A 2 doses at Mean µg/L 17.6 18.6 34.8 33.5 6 weeks No statistically significant differ-
1997 ZAF baseline ence in serum vitamin A levels at 6
(SD) (10.1) (10.5) (10.1) (14.2) weeks or 2 months
Better health.
Informed decisions.
Trusted evidence.
Ginawi 5000 IU vitamin A daily Mean 0.77 0.82 1.2 1.13 2 months No statistically significant differ-
2013 IND µmol/L ence in serum vitamin A levels at 2
(0.21) (0.29) (0.27) (0.34) or 6 months
(SD)
Pakasi 5000 IU vitamin A daily for 6 Median 0.7 0.7 1.5 1.2 2 months No statistically significant differ-
2010 IDN months µmol/L ence at 2 or 6 months
(0.5 to 1.5) (0.5 to 1.0) (1.0 to 2.0) (0.9 to 1.6)
(IQR)
Armijos 5000 IU vitamin A plus Mean 1.03 1.04 1.4 1.29 2 months No statistically significant differ-
2010 MEX µmol/L ence in serum vitamin A levels at 2
50 mg zinc daily for 4 months (0.46) (0.48) (0.47) (0.35) or 6 months
(SD)
Karyadi 5000 IU vitamin A plus 15 mg zinc Mean 0.82 0.9 1.14 1.08 2 months No statistically significant differ-
2002 IDN daily for 6 months µmol/L ence in serum vitamin A levels at 2
(0.25) (0.25) (0.32) (0.25) or 6 months
(SD)
Ginawi 5000 IU vitamin A plus 15 mg zinc Mean 0.78 0.82 1.14 1.13 2 months No statistically significant differ-
2013 IND sulphate daily µmol/L ence in serum vitamin A levels at 2
(0.23) (0.29) (0.25) (0.34) or 6 months
(SD)

Pakasi 5000 IU vitamin A plus 15 mg zinc Median 0.7 0.7 1.3 1.2 2 months No statistically significant differ-
2010 IDN sulphate daily for 6 months µmol/L ence in serum vitamin A levels at 2
(0.4 to 1.1) (0.5 to 1.0) (1.0 to 1.9) (0.9 to 1.6) or 6 months
(IQR)
Visser 200,000 IU vitamin A once plus 15 Median 21.1 (15.1 21.2 (15.7 40.3 (28.7 35.8 (27.7 2 months No statistically significant differ-
2011 ZAF mg zinc sulphate 5 days per week µd/dL to 27.8) to 28.9) to 48.5) to 43.2) ence in serum vitamin A levels at 2
for 8 weeks or 8 weeks
(IQR)
133
Collaboration.
Table 2. Effects of vitamin A supplementation on serum vitamin A levels (Continued)
Semba 8000 IU vitamin A daily as part of a Geomet- 0.60 (0.54 0.69 (0.63 NR NR 8 months Presented graphically with an in-
2007 MWI multi-micronutrient supplement ric mean to 0.66)1 to 0.75)1 crease in serum vitamin A levels in
Library
Cochrane
µmol/L both groups
(95% CI)
Abbreviations: IU = international units; IQR = interquartile range; CI = confidence interval; SD = standard deviation.
1Serum vitamin A levels for HIV-negative participants.
Better health.
Informed decisions.
Trusted evidence.
Table 3. Effects of zinc supplementation on serum zinc levels

ment ment
Ginawi 15 mg zinc sulphate daily Mean 9.86 9.88 12.23 11.23 2 months Statistically significant increase in
2013 IND µmol/L serum zinc levels in the zinc group at
(0.87) (1.44) (1.12) (1.34) 2 and 6 months
(SD)
Pakasi 15 mg zinc sulphate daily for 6 Mean 11.6 11.8 11.7 11.7 2 months No statistically significant difference
2010 IDN months µmol/L in serum zinc levels at 2 months
(2.2) (2.4) (2.43) (2.33)
(SD)
Armijos 50 mg zinc plus 5000 IU vitamin A Mean 11.29 11.69 12.85 10.13 2 months Statistically significant increase in
2010 MEX daily for 4 months µmol/L serum zinc levels in zinc plus vitamin
(2.57) (2.1) (3.4) (1.61) A group at 2 months. Increase not
(SD) sustained post supplementation.
Karyadi 15 mg zinc plus 5000 IU vitamin A Mean 11.52 11.15 11.22 10.22 2 months No statistically significant difference

2002 IDN daily for 6 months µmol/L in serum zinc levels at 2 months
(1.64) (1.77) (2.4) (2.28)
(SD)
Ginawi 15 mg zinc sulphate plus 5000 IU Mean 9.56 9.88 11.17 11.23 2 months No statistically significant difference
2013 IND vitamin A daily µmol/L in serum zinc levels at 2 months
(0.77) (1.4) (0.86) (1.34)
(SD)
Pakasi 15 mg zinc sulphate plus 5000 IU Mean 12.1 11.8 12.5 11.7 2 months No statistically significant difference
2010 IDN vitamin A daily for 6 months µmol/L in serum zinc levels at 2 months
(3.0) (2.4) (2.33) (2.33)
134
Informed decisions.
(Continued)
Table 3. Effects of zinc supplementation on serum zinc levels
Abbreviations: IU = international unit; SD = standard deviation.

(SD)
Collaboration.
Informed decisions.
Table 4. Effect of vitamin A plus zinc supplementation on nutritional recovery

Trial ID Popula- Baseline nutritional Endpoint nutritional Comments
tion status status
Mean BMI (SD) kg/m2
Vitamin A Placebo Vitamin A Placebo

plus zinc plus zinc
Karyadi Adults HIV 17.6 (1.9) 18.1 (3.2) Mean BMI Mean BMI Statistically significantly greater body weight (kg)
2002 IDN status un- (SD) 6 (SD) 6 in supplemented group at 6 months (3.10 kg, 95%
known months: months: CI 0.74 to 5.46). No statistically significant differ-
19.4 (2.5) 20.0 (3.2) ences in BMI, mid-upper arm circumference, biceps
skinfold thickness, triceps skinfold thickness, sub-
scapular skinfold thickness, supra-iliac skinfold
thickness, body fat (%), or fat mass (kg) between
groups at 2 or 6 months
Lawson Adults 19.6 (3.5) 19.8 (3.3) BMI data reported BMI appears to increase along a similar trajecto-
2010 NGA HIV-pos- graphically ry for both supplemented and placebo groups at 2
itive and and 6 months
HIV-nega-
tive
Pakasi Adults HIV 16.6 (2.1) 16.4 (2.5) Mean BMI Mean BMI No statistically significant differences in BMI be-
2010 IDN status un- (SD) 6 (SD) 6 tween the supplement and placebo group at 2 or 6
known months: months: months
18.3 (2.0) 18.4 (2.6)
Visser Adults Male: 18.9 Male: 19.0 Mean Mean No statistically significant difference in weight gain
2011 ZAF HIV-pos- (2.7) (2) weight weight during the first 2 months of treatment (P = 0.68, tri-
itive and gain 2.3 gain 2.2 al authors' own figures)
HIV-nega- Female: Female: kg kg
tive 23.0 (4.3) 21.6 (4.8)
Abbreviations: HIV = human immunodeficiency virus; BMI = body mass index; SD = standard deviation.
Collaboration.
Collaboration.
Table 5. Effects of vitamin D supplementation on serum vitamin D levels
Library
Cochrane
ment ment
Mily 2015 5000 IU daily for 8 Mean 28.0 28.1 102 30 8 weeks The follow-up mean values inserted here have
BGD weeks nmol/L been approximated from a graph in the pub-
(17.5) (16.2) (NR) (NR) lished paper. The trial authors reported that
Better health.
Informed decisions.
Trusted evidence.
(SD) the group receiving vitamin D3 supplementa-
tion exhibited significantly higher concentra-
tion of plasma 25-hydroxyvitamin D[25(OH) D3]
at week 8 compared to placebo after initiation
of therapy (P < 0.000)
Tukvadze 50000 IU 3 times a Mean 30 30 250 40 8 weeks The mean values inserted here have been ap-
2015 GEO week for 8 weeks, then nmol/L proximated from a graph in the published pa-
every 2 weeks for 8 (NR) (NR) (NR) (NR) per. The trial authors reported that high-dose
weeks (SD) vitamin D3 resulted in a significant increase in
plasma 25(OH)D concentrations to 250 nmol/L
at study week 8
Daley 2015 2.5 mg vitamin D3 on Mean 63.1 62.2 (51.0) 72.2 60.4 26 weeks Authors reported significant increase serum vi-
IND 1 days 0, 14, 28, and 42 nmol/L tamin D levels in supplement group (P = 0.001)
(46.6) (NR) (NR) but not in the placebo group (P = 0.15). No dif-
(SD) ference in serum vitamin D levels between
groups at week 26 (P = 0.24)
Kota 2011 60,000 IU vitamin D3 Mean ng/ 12.8 11.1 (4.7) 22.3 11.1 8 weeks Statistically significant increase in serum vi-
IND 2 per week plus 1000 mg ml tamin D level in the supplement group (P =
calcium carbonate per (4.5) (5.9) (1.6) 0.0001) at 4, 8, and 12 weeks
day (SD)

Martineau 2.5 mg on days 0, 14, Mean 21.1 21.3 (19.0) 101.4 22.8 8 weeks Statistically significant increase in serum vita-
2011 GBR 28, and 42 nmol/L min D level in the supplement group (P < 0.001)
3 (20.0) (NR) (NR)
(SD)
Wejse 100,000 IU vitamin D3 Mean 77.5 79.1 (21.8) 105 103 8 weeks There was no statistically significant differ-
2008 GNB at 0, 5, and 8 months nmol/L ence in serum vitamin D levels at 8 weeks or 8
4 (23.8) (95% CI 99 (95% CI 96 months between the groups
(SD) to 110) to 110)
137
Collaboration.
Table 5. Effects of vitamin D supplementation on serum vitamin D levels (Continued)
Morcos 1000 IU daily for 8 NR NR NR NR NR 16 weeks The trial authors reported that the serum vit-
1998 EGY 5 weeks amin D level rose in both groups and was not
Library
Cochrane
statistically different between groups (P > 0.05).
Abbreviations: NR = not reported; IU = international units; SD = standard deviation; 25(OH) D3 = 25-hydroxyvitamin D.

1 Daley 2015 IND only included HIV-negative participants.
2 Kota 2011 IND only enrolled newly diagnosed tuberculosis cases with uncontrolled diabetes and serum vitamin D < 20 ng/mL.
Better health.
Informed decisions.
Trusted evidence.
3 Martineau 2011 GBR reported that at baseline almost all participants had serum vitamin D levels below the definition of insufficiency used by Wejse 2008 GNB (95% supplement
group versus 98% control group).
4 Wejse 2008 GNB reports that at baseline approximately 10% of participants were defined as being vitamin D deficient (serum 25 (OH)D < 50 nmol/L), and 45% as vitamin D
3
insufficient (serum 25(OH)D3 < 75 nmol/L).
5 Morcos 1998 EGY reported a mean vitamin D level of 17.91 pg/mL at baseline which is below the normal reference range (20 to 42 pg/mL).

138
Informed decisions.
Table 6. Adverse effects of vitamin D

Trial ID Dose Severe ad- Effects on calcium Other
verse events
Mily 2015 5000 IU daily for 8 None in the vit- Hypercalcemia: none No differences between study
BGD weeks amin D group arms
Hypocalcemia: common but no
differences between groups
Tukvadze 50,000 IU 3 times a 5/100 vitamin Hypercalcemia: 3/100 vitamin D No differences between study
2015 GEO week for 8 weeks, then D versus 15/99 versus 7/99 placebo arms
every 2 weeks for 8 placebo
weeks
Daley 2.5 mg vitamin D3 on None in the vit- Hypercalcaemia: none 4/121 vitamin D versus 3/126
2015 IND days 0, 14, 28, and 42 amin D group placebo
(none required a change in med-

ical therapy)
Martineau 2.5 mg on days 0, 14, 28, 7/71 vitamin Hypercalcemia: 2/71 vitamin D ver- No differences between study
2011 GBR and 42 D versus 2/70 sus 0/70 placebo arms
placebo
Hypocalcaemia: 5/71 vitamin D
versus 2/70 placebo
Wejse 100,000 IU vitamin D3 at No comment Hypercalcemia: at 2 months: 1/157 At 2 months: only 24 reported any
2008 GNB 0, 5, and 8 months vitamin D versus 2/147 placebo symptom, most commonly exces-
sive thirst: 10/157 vitamin D versus
At 8 months: none 14/147 placebo (P = 0.31)
Abbreviations: IU = international units.
APPENDICES
Appendix 1. Composition of multi-micronutrient supplements: adults
Nutrient Adults
DRI for males Semba 2007 Range 2005 Villamor Praygod 2011a
MWI TZA 2008 TZA TZA
aged 19 to 70 years
Vitamin A 900 µg 2400 µg 1500 µg 1500 µg 1500 µg

(3000 IU) (8000 IU) (5000 IU) (5000 IU) (5000 IU)
Vitamin B1 (thiamine) 1.2 mg 1.5 mg 20 mg 20 mg 20 mg
Vitamin B2 (riboflavin) 1.3 mg 1.7 mg 20 mg 20 mg 20 mg
Vitamin B3 (niacin) 16 mg 20 mg 40 mg 100 mg 40 mg
Vitamin B6 (pyridoxine) 1.3 to 1.7 mg 2 mg 25 mg 25 mg 25 mg
Collaboration.
Informed decisions.
(Continued)
Vitamin B9 400 µg 400 µg 800 µg 800 µg 800 µg
(folic acid)
Vitamin B12 2.4 µg 6 µg 50 µg 50 µg 50 µg
Vitamin C 90 mg 500 mg 200 mg 500 mg 200 mg
Vitamin D 5 to 15 µg 10 µg 5 µg — 5 µg
(400 IU)
Vitamin E 15 mg 133 mg 60 mg 200 mg 60 mg

(200 IU)
Selenium 55 µg 65 µg 200 µg 100 µg 200 µg
Copper 0.9 mg — 5 mg — 5 mg
Zinc 11 mg 10 mg ± 45 mg (ele- — 30 mg
mentary zinc)
(as acetate)
Iodine 150 µg 175 µg — — —
Calcium 1000 mg — — — —
Manganese 2.3 mg — — — —
Magnesium 410 to 420 mg — — — —
D-panthenol — — — — —
Abbreviations: DRI = Dietary Reference Intake; IU = international unit.

Standards taken from the US Department of Agriculture Dietary Guidance available at http://fnic.nal.usda.gov/nal_display/index
Appendix 2. Search strategies for databases
Search CIDG SRa CENTRAL MEDLINEb EMBASEb LILACSb

set
1 tuberculosis tuberculosis tuberculosis tuberculosis tuberculosis
2 dietary supple- DIETARY SUPPLEMENTS DIETARY SUPPLE- dietary supplement$ dietary supple-
ments MENTS ments
3 macronutrients food supplement* food supplement* DIET-SUPPLEMENTATION macronutrients
4 micronutrients FOOD, FORTIFIED FOOD, FORTIFIED MACRONUTRIENT micronutrients
5 zinc macronutrients macronutrients micronutrient$ zinc
6 2 or 3 or 4 or 5 MICRONUTRIENTS MICRONUTRIENTS VITAMIN-SUPPLEMEN- 2 or 3 or 4 or 5

TATION
Collaboration.
Informed decisions.
(Continued)
7 1 and 6 TRACE ELEMENTS TRACE ELEMENTS IRON 1 and 6
8 — VITAMINS VITAMINS ZINC —
9 — vitamin* vitamin* TRACE-ELEMENT —
10 — zinc zinc 2-9 —
11 — iron iron 1 and 10 —
12 — 2-11/OR 2-11/OR Limit 11 to human —
13 — 1 and 12 1 and 12 — —
14 — — Limit 13 to human — —
aCochrane Infectious Diseases Group Specialized Register.

bSearch terms used in combination with the search strategy for retrieving trials developed by Cochrane (Lefebvre 2011); upper case: MeSH
or EMTREE heading, lower case: free text term.
Appendix 3. Optimal information size calculations (tuberculosis treatment outcomes)
Collaboration.
Collaboration.
Outcome Power Two-sided sig- Ratio of Risk in control RR deemed clinically Risk in interven- Sample size (to-
nificance level group 1: group significant tion group tal)
Library
Cochrane
group 2 (examples)
Death 80% 95% 1 5%1 0.8 4%2 13,890
Death 80% 95% 1 5%1 0.5 2.5%3 1968
Better health.
Informed decisions.
Trusted evidence.
Death 80% 95% 1 40%4 0.8 32% 1028
Death 80% 95% 1 40%4 0.5 20% 162
Cure 80% 95% 1 80%5 1.1 88%6 658
Cure 80% 95% 1 80%5 1.05 84%6 2894
Sputum positive at 1 80% 95% 1 30%7 0.75 22.5% 1078

month
Sputum positive at 2 80% 95% 1 10%7 0.75 7.5% 4010

months

142
Informed decisions.
Abbreviations: RR = risk ratio.
1Globally the risk of death in tuberculosis patients receiving treatment for tuberculosis is around 5%.
2Vitamins are relatively cheap and safe interventions, therefore even very modest reductions in the risk of death might be considered
important.
3A sample size of 2000 participants (higher than any of the included studies) would be necessary to reliably detect even a very large relative
reduction in death (50%).
4A very high risk of death was seen in HIV-positive participants in some trials due to antiretrovirals being unavailable at the study site at
the time. Death rates this high should not be seen in patients taking antiretrovirals.
5The target cure rate for directly observed treatment, short course (DOTs) programmes is 80%. The current global average is 86%.
6Vitamins are relatively cheap and safe interventions, therefore even very modest increases in successful cure might be considered
important.
7Sputum positivity rates in these trials were very variable. These examples are for illustrative purposes only.
We performed the calculations using nMaster 1.0 (nMaster 1.0), a sample size software that incorporates sample size calculation(STATA,
EpiInfo, nQuery, etc.), in terms of contents, each of use and the cost.
Appendix 4. Optimal information size calculations (nutritional and quality of life outcomes)
Collaboration.
Collaboration.
Outcome Power Two-sided Ratio of Mean in control SD Mean in supple- SD Mean differ- Sample
significance group 1: group ment group ence size (to-
Library
Cochrane
level group 2 tal)
Mean vitamin level1 80% 95% 1 35 µg/L 10 40 µg/L 10 5 µg/L 126
Mean increase in weight at 8 80% 95% 1 3.5 kg 6.3 5.5 kg 6.3 2 kg 312
weeks2
Better health.
Informed decisions.
Trusted evidence.
Mean increase in weight at 8 80% 95% 1 3.5 kg 6.3 8.5 kg 6.3 5 kg 50
weeks2
Mean BMI at 8 weeks3 80% 95% 1 18.5 kg/m2 2.6 19.5 kg/m2 2.6 1 kg/m2 214
Mean change in BMI at 8 weeks3 80% 95% 1 18.5 kg/m2 2.6 20.5 kg/m2 2.6 2 kg/m2 54
Mean Karnofsky score4 80% 95% 1 80 4.0 85 4.0 5 points 20

144
Informed decisions.
Abbreviations: BMI = body mass index; SD = standard deviation.
1We have taken this example from Hanekom 1997 ZAF

2We have taken this example from Martins 2009 TLS. This trial showed a 1.7 kg mean difference.
3This example uses the SD from Martins 2009 TLS but uses a 5 kg mean difference. This is for illustrative purposes.
4This example uses the SD taken from Karyadi 2002 IDN.
Appendix 5. Macronutrient supplements
Study ID Jahnavi 2010 Paton 2004 Jeremiah 2014 TZA Praygod 2011b TZA Martins 2009 TLS Sudarsanam 2010
IND SGP IND
Descrip- Advice on Advice on 4 high energy-pro- 5 high energy-pro- A daily meal (in- Macronutrient
tion of in- target en- target en- tein biscuits plus tein biscuits plus tensive phase; (ready-to-serve
terven- ergy intake ergy intake 1 vitamin/mineral 1 vitamin/mineral week 1 to 8) fol- powder, given as
tion and on how and on how fortified biscuit fortified biscuit lowed by a food monthly rations)
to achieve to achieve parcel (continua- and micronutri-
this with nor- this with nor- tion phase; week ent (daily multivi-
mal diet plus mal diet plus 9 to 32). The meal tamin tablet) sup-
food supple- 2 to 3 high- consisted of a plementation
ments (daily: energy oral bowl of meat, kid-
sweet balls nutritional ney beans, and Control: dietary
made from supplements vegetable stew advice alone
wheat flour, with rice. The food
caramel, parcel contained
ground- unprepared red
nuts, and kidney beans, rice,
vegetable and oil adequate
ghee as well for 1 meal per day.
as 100 g of
sprouted Control: nutrition-
grams and al advice alone.
nuts for vit-
amins and
minerals)
Duration 3 months Until they 2 months 60 days 8 months 6 months

of supple- reached a
menta- body mass
tion index of 20
or usual
body weight
Average Not stated 1560 kcal/ Not stated Not stated Not stated 2129 kcal
dietary day
energy 2072 kcal
intake 1502 kcal/
per day day
Total dai- Not stated 600 to 900 3075 kcal 3690 kcal 1800 kJ per daily 930 kcal
ly ener- kcal meal
gy intake (600 kcal per (615 kcal per bis- (615 kcal per bis-
through sweetball) (300 kcal per cuit) cuit)
supple- packet)
menta-
tion
Carbohy- Not stated 40.4 to 60.6 g Not stated Not stated 55.6 g per daily Not stated
drate meal
Collaboration.
Informed decisions.
(Continued)
(20.2 g per
packet)
Fat Not stated 8.58 to 12.87 Not stated Not stated Not stated Not stated
g
(4.29 g per
packet)
Protein Not stated 12.5 to 18.75 22.5 g 27g 18.4 g per daily 31.5 g protein
g meal
(6 g per (4.5 g per biscuit) (4.5 g per biscuit)
sweetball) (6.25 g per
packet)
Micronu- Not stated Not stated 600 mg phospho- 720 mg phospho- 24.2 mg vitamin Copper sulphate
trient rous, rous, C, 363 µg vitamin 0.1 mg, D-pan-
A, 3.1 mg iron, 3 g theol 1 mg, diba-
600 mg calcium, 720 mg calcium, zinc, 60 µg folate sic calcium phos-
phate 35 mg, folic
180 mg magne- 216 mg magne-
acid 500 µg, mag-
sium, sium,
nesium oxide 0.15
350 mg sodium, 420 mg sodium, mg, manganese
sulphate 0.01 mg,
750 mg potassium, 900 mg potassium, nicotinamide 25
mg, potassium io-
and traces (< 1 mg) and traces (< 1 mg) dide 0.025 mg, vit-
of iron and zinc, of iron and zinc, amin A 5000 IU, vi-
tamin B1 2.5 mg,
1.5 mg vitamin A, 1.5 mg vitamin A, vitamin B12 2.5
µg, vitamin B2 2.5
20 mg thiamin, 20 mg thiamin,
mg, vitamin B6 2.5
20 mg riboflavin, 20 mg riboflavin, mg, vitamin C 40
mg, vitamin D3
25 mg vitamin B6, 25 mg vitamin B6, 200 IU, vitamin E
7.5 mg, zinc sul-
50 µg vitamin B12, 50 µg vitamin B12, phate 50 mg
0.8 mg folic acid, 0.8 mg folic acid,
40 mg niacin, 40 mg niacin,
200 mg vitamin C, 200 mg vitamin C,
60 mg vitamin E, 60 mg vitamin E,
5 µg vitamin D, 5 µg vitamin D,
0.2 mg selenium, 0.2 mg selenium,
5 mg copper, 5 mg copper,
30 mg zinc 30 mg zinc
Appendix 6. High energy oral supplements versus dietary advice (additional data from Jahnavi 2010 IND and Paton
2004 SGP)
Collaboration.
Collaboration.
Outcome Timepoint Supplements Dietary advice P value
Library
Cochrane
Mean SD n Mean SD n
Change in physical function score (Jahnavi At 3 months 23.34 33.87 50 6.70 31.27 50 > 0.05
2010 IND)
Change in physical function score (Paton At 6 weeks 11.84 30.21 19 1.67 17.59 15 0.48
Better health.
Informed decisions.
Trusted evidence.
2004 SGP)
At 12 weeks 24.44 26.44 15 6.41 16.37 11 0.052
At 24 weeks 22.78 32.80 15 12.00 23.66 11 0.500
Change in emotional well-being score (Jah- At 3 months 22.32 22.69 50 2.56 21.45 50 > 0.05
navi 2010 IND)
Change in mental health score (Paton 2004 At 6 weeks 9.05 16.48 19 10.07 22.20 15 0.781
SGP)
At 12 weeks 10.13 26.74 15 8.92 20.08 11 0.581
At 24 weeks 11.20 24.94 15 11.00 23.01 11 0.810
Change in general health score (Jahnavi At 3 months 32.5 22.17 50 6.50 19.42 50 > 0.05
2010 IND)
Change in overall health score (Paton 2004 At 6 weeks 27.63 21.88 19 6.67 24.02 15 0.053
SGP)
At 12 weeks 30.00 19.36 15 25.00 27.00 11 0.544
At 24 weeks 36.67 18.58 15 29.17 33.43 11 0.465

147
Informed decisions.
Abbreviations: SD = standard deviation.
In total Jahnavi 2010 IND reported 8 and Paton 2004 SGP reported 12 quality of life/physical function scores. See the original papers for
full results.
Appendix 7. Composition of multi-micronutrient supplements: children
Collaboration.
Collaboration.
Children
Library
Cochrane
Nutrient Daily DRI for chil- Lodha 2014 IND Mehta 2011 TZA
dren
6 to 36 4 to 6 years 7 to 9 years 10 to 15 years <6 7 to 36 > 36
aged 1 to 3 years months months months months
Vitamin A 300 µg 0.8 mg1 0.9 mg1 1 mg1 1.2 mg1 — — —
Better health.
Informed decisions.
Trusted evidence.
Vitamin B1 (thiamine) 0.5 mg 1 mg 1.2 mg 1.8 mg 2.4 mg 0.5 mg 1 mg 1.5 mg
Vitamin B2 (riboflavin) 0.5 mg 1 mg 1.2 mg 1.8 mg 2.6 mg 0.6 mg 1.2 mg 1.8 mg
Vitamin B3 (niacin) 6 mg 12 mg2 16 mg2 24 mg2 32 mg2 4 mg 8 mg 12 mg
Vitamin B6 (pyridoxine) 0.5 mg 1 mg 1.2 mg 2 mg 2.6 mg 0.6 mg 1.2 mg 1.8 mg
Vitamin B9 (folic acid) 150 µg 300 µg3 400 µg3 600 µg3 800 µg3 130 µg 260 µg 390 µg
Vitamin B12 0.9 µg 3.8 µg 2.4 µg 3.6 µg 4.8 µg 1 µg 2 µg 3 µg
Vitamin C 15 mg 60 mg 60 mg 70 mg 80 mg 60 mg 120 mg 180 mg
Vitamin D 5 µg 10 µg 10 µg 10 µg 10 µg — — —
Vitamin E 6 mg 10 mg 10 mg 14 mg 20 mg 8 mg 16 mg 24 mg
Selenium 17 µg 10 µg 10 µg 10 µg 10 µg — — —
Copper 260 µg 2 mg 3 mg 4 mg 5 mg — — —

Zinc 2.5 mg 20 mg4 20 mg4 20 mg4 20 mg4 — — —
Iodine 65 µg — — — — — — —
Calcium 500 mg — — — — — — —
Manganese — — — — — — — —
Magnesium 65 mg — — — — — —
D-panthenol — — — — — — — —
149
Collaboration.
(Continued)
Abbreviations: DRI = Dietary Reference Intake. Standards taken from the US Department of Agriculture Dietary Guidance available at https://fnic.nal.usda.gov/dietary-guid-
ance/dietary-reference-intakes
Library
Cochrane
1mg retinol equivalent.
2mg niacin equivalent.
3µg dietary folate equivalent.
Better health.
Informed decisions.
Trusted evidence.
4One trial arm received zinc only; 1 arm received micronutrients plus zinc and 1 arm received micronutrients without zinc.

150
Informed decisions.
Appendix 8. Vitamin E and selenium levels (Seyedrezazadeh 2006)
Time Plasma vitamin E levels (µmol/L) Serum selenium levels (µmol/L)

point
Median value (range) Median value (range)
Supplement group (N = Placebo group (N = 18) Supplement group (N = 17) Placebo group (N = 18)
17)
At base- 24.7 (0 to 87) 20.2 (5.1 to 49) 1.0 (0.34 to 2.5) 0.93 (0.1 to 1.9)
line
At 8 weeks 28.2 (10.5 to 86.5) 19.3 (5.1 to 48.6) Not reported Not reported
Abbreviations: N = number of participants.
WHAT'S NEW
Date Event Description
28 June 2016 New citation required but conclusions We performed a new literature search and included 12 new trials
have not changed (from 15 articles). The author team has been updated.
28 June 2016 New search has been performed We amended the objective of this Cochrane Review to simpli-
fy the wording and to reflect the need to assess any differences
in response depending on HIV status. We amended some of the
outcomes. Also, we added total calorie intake and micronutrient
levels before and after supplementation as outcomes in the re-
view as there are important explanatory factors for any effects
seen. Also, we adapted the methods to reflect changes in the
methods of assessing and reporting risk of bias.
HISTORY
Protocol first published: Issue 3, 2006
Review first published: Issue 4, 2008
30 September 2011 New citation required but conclusions We added eight new trials and have considered more carefully
have not changed the nutritional status at baseline (both weight and biochemical
status of individual micronutrients). The author team has also
been updated.
20 September 2011 New search has been performed We performed a new search, and included eight new trials. We
constructed 'Summary of findings' tables, which summarize the
quality of the evidence. Also we performed a calculation of the
optimal information size to reliably detect clinically important
effects if they exist.
We updated the 'Risk of bias' assessments to the new format.
Collaboration.
Informed decisions.
10 November 2009 Amended An observant reader noted that there was an error in referencing
in the 'Risk of bias' tables. This error has now been corrected.
10 November 2008 Amended We corrected minor errors. There were no changes to the conclu-
sions.
CONTRIBUTIONS OF AUTHORS
Thambu David Sudarsanam conceived this Cochrane protocol, and designed it in collaboration with all review authors (Abba 2006). We
performed the selection of trials for inclusion, 'Risk of bias' assessments, and data extraction as indicated in the Methods. Katharine Abba
and David Sinclair mainly undertook the analyses of previous versions of the review, in consultation with the other review authors. In the
most recent update of this Cochrane review, Liesl Grobler, Sukrti Nagpal, and Thambu David Sudarsanam screened the search results and
extracted the data from the eligible studies. Liesl Grobler and David Sinclair, in consultation with the other review authors, analysed the
data and wrote the review.
DECLARATIONS OF INTEREST
Liesl Grobler has no known conflicts of interest.
David Sinclair was previously a member of the World Health Organization (WHO) Technical Advisory Group on Nutrition. This work may
contribute to future recommendations on nutritional care in tuberculosis.
Sukrti Nagpal has no known conflicts of interest.
Thambu D Sudarsanam has no known conflicts of interest.
SOURCES OF SUPPORT
Internal sources
• Liverpool School of Tropical Medicine, UK.
• Stellenbosch University, South Africa.
• Christian Medical College Vellore, India.
External sources
• Department for International Development (DFID), UK.
Grant: 5242
DIFFERENCES BETWEEN PROTOCOL AND REVIEW

We amended the objective of this Cochrane Review from "To assess the provision of oral nutritional supplements to promote the recovery
of people being treated with anti-TB drug therapy for active TB" to the current objective to simplify the wording and reflect the need to
assess any differences in response depending on HIV status.
Also we changed some of the outcomes, as follows: we expanded "change in weight or skinfold thickness" to "change in weight, skinfold
thickness, or other measure of lean or total mass" because we became aware that other measures, besides skinfold thickness, are equally
valid indicators of overall nutritional status; we added "any measure of growth in children" because it is a useful indicator of health and
nutritional status in children, which we had overlooked at the protocol stage; and included "sputum positive at follow-up" because it was
a primary outcome of many included trials, and it became apparent that it is a meaningful outcome as, depending on the period of follow-
up, it may be used as a proxy for cure or as an indicator of the time taken to become sputum-smear negative. Early sputum conversion is
a desirable outcome because on becoming smear-negative, patients become less ill and less infectious to those around them.
Between the publication of the original review (Abba 2008; Sinclair 2011), and this review update, the method of assessing and reporting
risk of bias has changed slightly. We have adapted the methods to reflect this.
For this review update, we included additional details on the nutritional and micronutrient status at baseline. Where reported, we have
also included plasma micronutrient levels during follow-up as an outcome. This outcome is not a patient important outcome, and is of
little interest on its own, but does contribute to the understanding of the results.
Collaboration.
Informed decisions.
INDEX TERMS
Medical Subject Headings (MeSH)

*Dietary Supplements; Antitubercular Agents [therapeutic use]; Energy Intake; HIV Infections [complications] [mortality];
Malnutrition [complications] [*diet therapy]; Micronutrients [administration & dosage]; Randomized Controlled Trials as Topic;
Tuberculosis [complications] [*diet therapy] [drug therapy] [mortality]
MeSH check words

Adult; Child; Humans
Collaboration.

Nutritional supplements for people being treated for active tuberculosis

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Nutritional supplements for people being treated for active tuberculosis

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nutritional supplements for people being treated for active tuberculosis

Uploaded by

Copyright:

Available Formats

Cochrane

Nutritional supplements for people being treated for active

Grobler L, Nagpal S, Sudarsanam TD, Sinclair D

Grobler L, Nagpal S, Sudarsanam TD, Sinclair D.

Nutritional supplements for people being treated for active tuberculosis

Liesl Grobler1, Sukrti Nagpal2, Thambu D Sudarsanam3, David Sinclair2

Editorial group: Cochrane Infectious Diseases Group.

Data collection and analysis

Studies awaiting assessment

PLAIN LANGUAGE SUMMARY

Nutritional supplements for people being treated for active tuberculosis

What is active tuberculosis and how might nutritional supplements work?

What the research says

Effect of providing nutritional supplements to people being treated for tuberculosis

Summary of findings for the main comparison. 'Summary of findings' table 1

Patient or population: adults and children with active tuberculosis

Settings: low- and middle-income countries

Comparison: nutritional advice, micronutrient supplement, or no intervention

Standard Increased calo-

Cochrane Database of Systematic Reviews

Summary of findings 2. 'Summary of findings' table 2

Cochrane Database of Systematic Reviews

Settings: low- and middle-income countries

Intervention: multi-micronutrient supplements

Comparison: placebo or no intervention

Cure rate — — — (0 trials) — We don't know if multi-micronutrients improve cure in tuber-

Quality of life — — — (0 trials) — We don't know if multi-micronutrients improve quality of life

Cochrane Database of Systematic Reviews

GRADE Working Group grades of evidence

Cochrane Database of Systematic Reviews

BACKGROUND these recommended amounts. The two important questions are

• Improve tuberculosis treatment outcomes; through restoration Types of outcome measures

Searching other resources Dealing with missing data

future updates of this Cochrane review we may perform subgroup RESULTS

Figure 1. Study flow diagram.

Allocation Twenty-one of the 27 trials assessing micronutrients used placebos

Nutritional recovery and quality of life Tuberculosis treatment outcomes

In addition, it should be noted that most of the HIV-positive AUTHORS' CONCLUSIONS

References to studies awaiting assessment in Qom. http://www.irct.ir/searchresult.php?

NCT01635153 {published data only}

ChiCTR-TRC-14005241 {unpublished data only} NCT01992263 {unpublished data only}

NCT02464683 {unpublished data only} Harries 2006

Plit 1998 WHO 2009

Saunders 2007 Wintergerst 2007

Schluger 1998 Wyss 2001

Semba 1998 Zachariah 2002

Characteristics of included studies [ordered by study ID]

Armijos 2010 MEX

Participants Number: 39 enrolled; 33 analysed.

Group 2: placebo with identical appearance.

Outcomes • Deaths during study.

Notes Location: Ciudad Juárez, México.

Setting: IMSS outpatient services.

Bias Authors' judgement Support for judgement

Armijos 2010 MEX (Continued)

Daley 2015 IND

Participants Number: 247 randomized (121 to vitamin D group, 126 to placebo).

Vitamin D group: 101/121 included in primary analysis.