J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest.

Protected by copyright.
EVIDENCE-BASED CLINICAL MEDICINE

Vitamin D: An Evidence-Based Review

Teresa Kulie, MD, Amy Groff, DO, Jackie Redmer, MD, MPH, Jennie Hounshell, MD,
and Sarina Schrager, MD, MS

Vitamin D is a fat-soluble vitamin that plays an important role in bone metabolism and seems to have
some anti-inflammatory and immune-modulating properties. In addition, recent epidemiologic studies
have observed relationships between low vitamin D levels and multiple disease states. Low vitamin D
levels are associated with increased overall and cardiovascular mortality, cancer incidence and mortal-
ity, and autoimmune diseases such as multiple sclerosis. Although it is well known that the combination
of vitamin D and calcium is necessary to maintain bone density as people age, vitamin D may also be an
independent risk factor for falls among the elderly. New recommendations from the American Academy
of Pediatrics address the need for supplementation in breastfed newborns and many questions are
raised regarding the role of maternal supplementation during lactation. Unfortunately, little evidence
guides clinicians on when to screen for vitamin D deficiency or effective treatment options. (J Am Board
Fam Med 2009;22:698 –706.)

Background and Physiology duce adequate vitamin D from the skin depends on
Vitamin D is a hormone precursor that is present in the strength of the UVB rays (ie, place of resi-
2 forms. Ergocalciferol, or vitamin D2, is present in dence), the length of time spent in the sun, and the
plants and some fish. Cholecalciferol, or vitamin amount of pigment in the skin. Tanning beds pro-
D3, is synthesized in the skin by sunlight. Humans vide variable levels of UVA and UVB rays and are
can fulfill their vitamin D requirements by either therefore not a reliable source of vitamin D.
ingesting vitamin D or being exposed to the sun for Vitamin D3 is synthesized from 7-dehydrocho-
enough time to produce adequate amounts. Vita- lesterol in the skin. The vitamin D binding protein
min D controls calcium absorption in the small transports the vitamin D3 to the liver where it
intestine and works with parathyroid hormone to undergoes hydroxylation to 25(OH)D (the inactive
mediate skeletal mineralization and maintain cal- form of vitamin D) and then to the kidneys where
cium homeostasis in the blood stream. In addition, it is hydroxylated by the enzyme 1 ␣hydroxylase to
recent epidemiologic studies have observed rela- 1,25(OH)D, its active form.1 This enzyme is also
tionships between low vitamin D levels and multi- present in a variety of extrarenal sites, including
ple disease states, probably caused by its anti-in- osteoclasts, skin, colon, brain, and macrophages,
flammatory and immune-modulating properties which may be the cause of it’s broad-ranging ef-
and possible affects on cytokine levels. fects.1 The half-life of vitamin D in the liver is
Vitamin D3 can be manufactured in the skin by approximately 3 weeks, which underscores the need
way of ultraviolet (UV) B rays. UVB rays are for frequent replenishment of the body’s supply.
present only during midday at higher latitudes and
do not penetrate clouds. The time needed to pro- Vitamin D and Mortality
Vitamin D may be a determinant of mortality be-
cause of its anti-inflammatory and immune-modu-
This article was externally peer reviewed. lating effects. It has been used to treat secondary
Submitted 27 February 2009; revised 10 July 2009; ac-
cepted 13 July 2009. hyperparathyroidism in people on dialysis. Retro-
From the Department of Family Medicine, University of spective trials show that vitamin D supplementa-
Wisconsin, Madison.
Funding: none. tion is associated with decreased mortality in peo-
Conflict of interest: none declared. ple on dialysis.2 Low serum vitamin D levels are
Corresponding author: Sarina Schrager, MD, MS, Depart-
ment of Family Medicine, University of Wisconsin, 777 S. also related to increased mortality in most patients
Mills St., Madison, WI 53715 (E-mail: [email protected]). with chronic kidney disease before dialysis.3 How-

698 JABFM November–December 2009 Vol. 22 No. 6 http://www.jabfm.org

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
ever, there have been no randomized prospective vitamin D intake by infants may reduce the risk of the
trials examining this relationship.4 development of type 1 DM.13
In patients not on dialysis, low vitamin D levels Vitamin D has recently been associated with
are associated with increased levels of inflammation several of the contributing factors known to be
and oxidative load. A prospective study of more linked to the development of type 2 DM, including
than 3000 male and female patients scheduled for defects in pancreatic ␤cell function, insulin sensi-
coronary angiography found a positive association tivity, and systemic inflammation. Several physio-
between low vitamin D levels and cardiovascular as logic mechanisms have been proposed, including
well as all-cause mortality.5 Data analysis from the the effect of vitamin D on insulin secretion, the
National Health and Nutrition Examination Sur- direct effect of calcium and vitamin D on insulin
vey III (more than 13,000 adults) showed that peo- action, and the role of this hormone in cytokine
ple with vitamin D levels in the lowest quartile had regulation.9,12,13 Although most studies indicating
a mortality rate ratio of 1.26 (95% CI, 1.08 –1.46).6 this relationship are observational, one meta-anal-
A recent meta-analysis demonstrated that intake of ysis showed a relatively consistent association be-
a vitamin D supplement at normal doses also was tween low vitamin D status, calcium or dairy intake,
associated with decreased all-cause mortality rates.7 and prevalence of type 2 DM or metabolic syn-
These data suggest that vitamin D may play a part drome. The study concluded that the highest type
in multiple causes of death, although causality has 2 DM prevalence, 0.36 (95% CI, 0.16 – 0.80),
not been determined. among participants who were not black was associ-
ated with the lowest blood levels of 25-hydroxyvi-
tamin D. In addition, metabolic syndrome preva-
Vitamin D and Cardiovascular Disease
lence of 0.71 (95% CI, 0.57– 0.89) was highest
Vitamin D receptors are present in vascular smooth
among those with the lowest dairy intake. There
muscle, endothelium, and cardiomyocytes and may
was also an inverse relationship between type 2 DM
have an impact on cardiovascular disease. Observa-
and metabolic syndrome incidences and vitamin D
tional studies have shown a relationship between
and calcium intake.14
low vitamin D levels and blood pressure, coronary
artery calcification, and existing cardiovascular dis-
Vitamin D and Osteoporosis
ease. A large cohort study that included more than
Osteoporosis is the most common metabolic bone
1700 participants from the Framingham offspring
disease in the world. A low vitamin D level is an
study looked at vitamin D levels and incident car-
established risk factor for osteoporosis. Inadequate
diovascular events.8 During a period of 5 years,
serum vitamin D levels will decrease the active
participants who had 25-OH D levels of ⬍15 were
transcellular absorption of calcium.
more likely to experience cardiovascular events
Although combination calcium and vitamin D
(hazard ratio, 1.62; 95% CI, 1.11–2.36). The rela-
supplementation is associated with higher bone
tionship remained significant among people with
mineral density and decreased incidence of hip
hypertension but not among those without hyper-
fractures,15 the evidence for vitamin D supplemen-
tension.8
tation alone is less clear. A recent evidence sum-
mary found that vitamin D supplementation at
Vitamin D and Diabetes doses of more than 700 IU daily (plus calcium)
Recent studies in animal models and humans have prevented bone loss compared with placebo.16
suggested that vitamin D may also play a role in the However, vitamin D supplementation (300 to 400
homeostasis of glucose metabolism and the develop- IU daily) without calcium did not affect fractures.16
ment of type 1 and type 2 diabetes mellitus (DM). A Cochrane review found unclear evidence that
Epidemiologic data has long suggested a link between vitamin D alone affected hip, vertebral, or other
exposure to vitamin D early in life and the develop- fracture rates but supported the use of vitamin D
ment of type 1 DM.9,10 Vitamin D3 receptors have with calcium in frail, elderly nursing home resi-
strong immune-modulating effects. In some popula- dents.17 A subsequent meta-analysis of trials look-
tions the development of type 1 DM is associated ing at vitamin D and fracture rates concurred that
with polymorphisms in the vitamin D receptor calcium was also necessary to affect a significant
gene.11,12 There is also some evidence that increased difference.18

doi: 10.3122/jabfm.2009.06.090037 Vitamin D: An Evidence-Based Review 699

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
The most recent meta-analysis of 12 random- treat from the pooled results was 15 to prevent 1
ized, controlled trials that included more than person from falling22 Assessing vitamin D levels in
42,000 people found that vitamin D supplementa- a population at high risk for falling and supple-
tion of more than 400 IU daily slightly reduced menting with 800 to 1000 IU daily of vitamin D
incidence of nonvertebral fractures (rate ratio, 0.86; should be a part of any fall prevention program.
95% CI, 0.77– 0.96).19 The effect was dose depen-
dent and was not significant if doses were ⱕ400 IU Vitamin D and Cancer
daily. Both observational studies in humans and animal
models support that vitamin D has a beneficial role
Vitamin D and Falls among the Elderly in cancer prevention and survival. The mechanism
Vitamin D status is increasingly recognized as an of action is probably related to its role in the reg-
important factor in fall status among elderly pa- ulation of cell growth and differentiation.23 In the
tients. Several trials have demonstrated that vita- Health Professionals Follow-Up study (a cohort
min D supplementation decreases the risk of fall- study of 1095 men), each increment in 25(OH)D
ing. One proposed mechanism is that higher level of 25mmol/L was associated with a 17% re-
vitamin D levels are associated with improved mus- duction of total cancer cases.24 However, the Na-
cle function. tional Health and Nutrition Examination Survey of
A randomized, controlled trial from Australia 16,818 men and women did not find a relationship
evaluated women with at least one fall in the pre- between total cancer mortality and vitamin D level.
ceding 12 months and with a plasma 25-hyroxyvi- There was an inverse relationship between vitamin
tamin D level ⬍24.0 ng/mL.20 All women were D level and colorectal cancer, however. In this
given calcium 1000 mg per day and were random- study, serum 25(OH)D levels of ⱖ80 nmol/L con-
ized to receive either ergocalciferol 1000 IU per ferred a 72% reduction in risk of colorectal cancer
day or placebo. Women in the study group had compared with a level lower than 50 nmol/L.25
fewer falls after 12 months, but this was not a A recent meta-analysis of 63 observational stud-
significant difference (53% versus 62.9%; odds ra- ies looked at the relationship between vitamin D
tio, 0.66; 95% CI, 0.41–1.06). After correction for levels and cancer incidence and mortality.26
height difference in the 2 groups, the ergocalciferol Twenty of the 30 studies looking at vitamin D and
group had a significantly lower risk of falling (odds colon cancer showed that people with higher vita-
ratio, 0.61; 95% CI, 0.37– 0.99). min D levels had either a lower incidence of colon
A dose of 800 IU daily significantly reduced the cancer or decreased mortality. Similarly, 9 of the 13
risk of falling compared with a placebo in a dose- studies about breast cancer and 13 of the 26 studies
stratified analysis of the effect of 5 months of vita- about prostate cancer showed beneficial effects of
min D supplementation on fall risk (72% lower vitamin D levels on cancer incidence or mortality
incidence rate ratio; rate ratio, 0.28; 95% CI, 0.10 – (some of the studies included more than one type of
0.75). Lower doses of vitamin D, however, did not cancer).26
significantly change the rate of fall incidence com- A population-based randomized, control trial
pared with placebo.21 found that postmenopausal women who were sup-
A review of 12 randomized, controlled trials plemented with calcium and vitamin D had a re-
studying the effect of vitamin D supplementation duced risk of cancer after the first year of treatment
on fall risk among both nursing home residents and (rate ratio, 0.232; 95% CI, 0.09 – 0.60).27 There was
community dwellers found a small benefit of sup- not a group that was supplemented with vitamin D
plementation on fall risk (odds ratio, 0.89; 95% CI, alone.
0.80 – 0.99),9 an effect that was also shown in a
review of randomized, controlled trials with strict Vitamin D and Multiple Sclerosis
inclusion criteria, which included 1237 men and Multiple sclerosis (MS) is a neurodegenerative, T
women with a mean age of 70 years and supple- lymphocyte-mediated, autoimmune disease of un-
mentation for 2 months to 3 years. The pooled certain etiology. Although genetic susceptibility
results showed a significant 22% decrease in fall may be involved, epidemiologic studies suggest en-
risk among those treated with vitamin D versus vironmental influence because the development of
placebo or calcium only. The number needed to MS correlates most strongly with rising latitude in

700 JABFM November–December 2009 Vol. 22 No. 6 http://www.jabfm.org

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
both the northern and southern hemispheres.28 in the United States that followed 2 large cohorts
Migration studies show that risk can be modified at of women—the Nurses’ Health Study (92,253
an early age from both low to high and high to low women followed from 1980 to 2000) and the
prevalence rates.28 Exposure to sun in early child- Nurses’ Health Study II (95,310 women followed
hood is associated with reduced risk of developing from 1991 to 2001)—found that vitamin D supple-
MS29 and population-based studies about MS in mentation in the form of a multivitamin seemed to
Canada have also shown that birth timing is a risk lower their MS risk by 40%.41 However, several
factor for MS because there are statistically signif- methodological weaknesses in study design made
icantly fewer patients with MS born in November the results inconclusive.42
and more born in May compared with controls.30 A Despite the overwhelming amount of data de-
birth-timing association suggests that seasonality scribing the association between vitamin D and
and sunlight exposure may also have an effect on MS, there is a paucity of research describing the
the developing fetus in utero.30,31 benefit of vitamin D supplementation to these pa-
Several studies have shown that vitamin D af- tients. One small safety study of 12 patients taking
fects the growth and differentiation of immune- 1000 ␮g per day (40,000 IU) of vitamin D for 28
modulator cells such as macrophages, dendritic weeks showed a decline in the number of gadolin-
cells, T cells, and B cells.32–34 This immune-mod- ium-enhancing lesions on magnetic resonance im-
ulatory effect has implications for a variety of au- aging per patient; this led to a 25(OH)D serum
toimmune diseases including rheumatoid arthritis, concentration of 386 nmol/L (158 ng/mL) without
systemic lupus erythematosous, type I DM, inflam- causing hypercalcemia, hypercalciuria, or other
matory bowel disease, and MS.33 complication.43
Despite the wealth of epidemiologic studies sup-
porting a relationship between vitamin D and MS Vitamin D and Cognition
in humans, data showing a link between serum Observational studies have shown that people with
vitamin D levels and MS are only beginning to Alzheimer dementia have lower vitamin D levels
emerge. One prospective, nested, case-control than do matched controls without dementia.44 The
study examined the serum samples of 7 million biological plausibility of this relationship includes
military veterans and compared serum samples of vitamin D’s antioxidative effects and the presence
257 MS patients before diagnosis with those of of vitamin D receptors in the hippocampus, which
matched controls.35 An inverse relationship be- has been seen in rats and humans.44 A cross-sec-
tween vitamin D levels and MS risk was found, tional study of 225 outpatients diagnosed with Alz-
particularly for vitamin D levels measured in pa- heimer disease found a correlation between vitamin
tients younger than 20. Another case-control study D levels (but not other vitamin levels) and their
compared the serum vitamin D levels of 103 MS score on a Mini Mental Status Examination.45
patients with 110 controls and found that for every
10-nmol/L increase of serum 25(OH)D level the Vitamin D and Chronic Pain
odds of MS was reduced by 19% in women, sug- Because of the important role vitamin D plays in
gesting a “protective” effect of higher vitamin D bone homeostasis, some have questioned whether
levels.36 In addition, a negative correlation was vitamin D deficiency may also correlate with
found between Expanded Disability Status Scale chronic pain syndromes, including chronic low
scores among female MS patients and 25(OH)D back pain. Several case series and observational
levels. Several other studies have supported the studies have suggested that vitamin D inadequacy
finding that lower levels of vitamin D in MS pa- may represent a source of nociception and impaired
tients are associated with more severe disability.37 neuromuscular functioning among patients with
Lower levels during relapses have also been re- chronic pain.
ported in patients with relapse-remitting MS.38 – 40 The data to support this conclusion are mixed. A
The potential effects of oral vitamin D intake recent review of 22 relevant studies found no con-
have been observed in several different ways. A vincing link between prevalence and latitude and
Norwegian case-control study found that fish and no association between serum levels of 25-OH vi-
cod liver oil have a protective effect against the tamin D in chronic pain patients and controls.
development of MS.29 A large observational study Interestingly, though, there was a contrast in treat-

doi: 10.3122/jabfm.2009.06.090037 Vitamin D: An Evidence-Based Review 701

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
ment effects between randomized, double-blind Table 1. Vitamin D Preparations for Newborns
trials that minimized bias and those with studies Preparation Dose Comments
known to be subject to bias. In those that blinded
the vitamin D therapy, only 10% of patients were Just D by Sunlight 400 IU in 1 mL
vitamins
in trials showing a benefit of vitamin D treatment,
Carson labs 400 IU gelcap
whereas among those who did not blind the treat-
Polyvisol 400 IU Also contains other
ment 93% were in trials showing a benefit of vita- vitamins
min D supplementation.46
A second review examined the role of vitamin
deficiency in patients from outpatient and inpatient ficient vitamin D levels in infants and their moth-
rehab units. Fifty-one articles were reviewed and a ers.52 Both bone mineral accrual in early child-
direct correlation was noted between vitamin D hood53 and the risk of recurrent wheezing episodes
deficiency and musculoskeletal pain. Treatment of in children at age 354 were linked to insufficient
vitamin D deficiency produced an increase in mus- vitamin D intake by women during pregnancy. If a
cle strength and a marked decrease in back and fetus or breastfeeding infant receives an inadequate
lower-limb pain within 6 months.47 Although these amount of vitamin D from its mother it can have a
data were suggestive of a link between vitamin D direct impact on the baby’s health as an adult.
and pain, the available evidence does not imply Because of these findings, in 2007 the Canadian
causality. The verdict on this topic will remain Pediatric Society recommended 2000 IU of vitamin
undecided until this is evaluated by double-blind, D3 for pregnant and lactating mothers with peri-
randomized, controlled trials stratified by baseline odic blood tests to check levels of 25 (OH)D and
vitamin D level with defined treatments and com- calcium.52 The American Academy of Pediatrics
parison placebo groups. recommendations focus on supplementing the in-
fant and make no specific recommendations about
universally supplementing breastfeeding mothers.48
Vitamin D Supplementation for Infants and
Breastfeeding Mothers Supplementing the Newborn: 2008
Breast milk is an ideal form of nourishment for a Recommendations from the American Academy of
newborn. Because of most nursing mother’s own Pediatrics
vitamin D deficiency, however, and despite the The American Academy of Pediatrics recommends
mother taking a prenatal vitamin, breast milk alone supplementing all children who are exclusively
is not sufficient to maintain newborn vitamin D breastfed with 400 IU of vitamin D from the first
levels within a normal range.48 Many nursing few days of life. Children who are fed by breast and
mothers or their infants require vitamin D supple- formula or who are exclusively formula fed should
mentation for optimal health.49 also be supplemented until they are consistently
In 2003, the American Academy of Pediatrics ingesting 1 L of formula a day (approximately 1
recommended that 200 IU of vitamin D be used as quart). The supplementation should continue until
supplementation for all infants beginning during 1 year of age, when children begin ingesting vita-
the first 2 months after birth.50 More recently, in min D-fortified milk.48 All formulas sold in the
2008 the recommendation has been increased to a United States contain at least 400 IU/L of vitamin
minimum of 400 IU daily during the first days of D3; therefore, 1 L per day would meet the vitamin
life to prevent vitamin D deficiency that may lead D recommendations set by the American Academy
to rickets.48 of Pediatrics.55
A 2004 systematic review looked at 166 cases of
nutritional rickets diagnosed between 1986 and Preparations for Supplementation
2004 in 17 states from the mid-Atlantic region to There are many available preparations for new-
Texas and Georgia. A disproportionate number of borns (Table 1). Some companies make a single-
rickets cases were found in African-American, drop preparation that contains 400 IU, but caution
breastfed infants.51 In addition to rickets and the should be used when prescribing this product be-
risk of developing type I DM, other pediatric and cause of the ease of dispensing too much vitamin D
adult health conditions may be impacted by insuf- to a newborn with just a few drops.48

702 JABFM November–December 2009 Vol. 22 No. 6 http://www.jabfm.org

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
Checking Serum Levels in Infants Table 2. Causes of Vitamin D Deficiency58
Clinicians should obtain a serum vitamin D level Causes Example
(25-OH-D not 1,25-OH2-D) among infants with
malabsorption disorders or who take anticonvul- Reduced skin synthesis Sunscreen, skin pigment,
season/latitude/time of
sants because they may need additional supplemen- day, aging, skin grafts
tation above 400 IU daily. Actual values of 25- Decreaseed absorption Cystic fibrosis, celiac
OH-D that determine vitamin D insufficiency in disease, whipple disease,
Crohn disease, gastric
children have not been defined. The ⱖ20 ng/mL of bypass, medications that
25-OH-D that determines a sufficient vitamin D reduce cholesterol
absorption
level for adults has been used for children.48
Increased sesquestration Obesity
Increased catabolism Anticonvulsant,
Supplementing Breastfeeding Mothers glucocorticoid, highly
active antiretroviral
Mothers who were supplemented with 400 IU of treatment, and some
vitamin D daily produced milk with vitamin D immunosuppressants
levels that ranged from ⬍25 to 78 IU per liter.48 Breastfeeding
Decreased synthesis of 25- Hepatic failure
Supplementing the mother alone with 400 IU— hydroxyvitamin D
equivalent to a prenatal vitamin—produced inade- Increased urinary loss of Nephrotic proteinuria
quate vitamin D levels in the breastfed infants.55 A 25-hydroxyvitamin D
randomized, controlled trial evaluated 19 breast- Decreased synthesis of 1,25- Chronic renal failure
dihydroxyvitamin D
feeding mothers who were supplemented with 6000
Heritable disorders Genetic mutations causing
IU of vitamin D3 and a prenatal vitamin with 400 rickets, or vitamin D
IU of vitamin D. The vitamin D levels found in resistance
their breast milk and in the exclusively breastfed Acquired disorders Tumor-induced
osteomalacia, primary
infants themselves were found to be equivalent to hyperparathyroidism,
the infants who received oral supplementation (300 hyperthyroidism,
granulomatous disorders
IU per day). This level of maternal supplementa- such as sarcoidosis,
tion showed no toxic effects and provided adequate tuburculosis, and some
lymphomas
vitamin D to nursing infants without needing to
supplement the infant.56 Safety and efficacy of this
dosing during pregnancy and lactation has not been
(2) older people admitted to hospital; (3) patients
confirmed. In the meantime, screening high-risk
with hip fracture; (4) dark-skinned women (partic-
women is appropriate and supplementing breast-
ularly if veiled); and (5) mothers of infants with
feeding women who are vitamin D3 deficient is
rickets (particularly if dark-skinned or veiled).58
warranted.57
If electing to test vitamin D status, serum 25-
hydroxyvitamin D is the accepted biomarker.60 Al-
Testing for Vitamin D Deficiency though 1,25-OH-D is the active circulating form of
There are many causes of vitamin D deficiency, as vitamin D, measuring this level is not helpful be-
listed in (Table 2),59 and despite growing attention cause it is quickly and tightly regulated by the
to this deficiency, there are no established guide- kidney. True deficiency would be evident only by
lines to help clinicians decide which patients war- measuring 25-OH-D. Of note, questions have been
rant screening laboratory testing. The US Preven- raised regarding the need for standardization of
tive Services Task Force does not comment for or assays.61 A large laboratory (Quest Diagnostics)
against routine screening for vitamin D deficiency. recently reported the possibility of thousands of
One approach is to consider serum testing in pa- incorrect vitamin D level results.62 Sunlight expo-
tients at high risk for vitamin D deficiency but sure questionnaires are imprecise and are not cur-
treating without testing those at lower risk. rently recommended.63
An Australian working group issued a position Controversy exists regarding the optimum level
statement itemizing groups of people at risk for of serum 25-hydroxyvitamin D in a healthy popu-
vitamin D deficiency. The risk groups include: (1) lation. Most experts agree that serum vitamin D
older people in low- and high-level residential care; levels ⬍20 ng/mL represent deficiency. However,

doi: 10.3122/jabfm.2009.06.090037 Vitamin D: An Evidence-Based Review 703

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
some experts recommend aiming for a higher min- mental intakes of 400 IU per day of vitamin D
imum target level of 30 ng/mL of 25-hydroxyvita- increase 25(OH)D concentrations by only 2.8 to
min D49 in a healthy population. Vitamin D intox- 4.8 ng/mL (7–12 nmol/L) and that daily intakes of
ication can occur when serum levels are greater approximately 1700 IU are needed to raise these
than 150 ng/mL. Symptoms of hypervitaminosis D concentrations from 20 to 32 ng/mL (50 – 80 nmol/
include fatigue, nausea, vomiting, and weakness L).65 Responses to vitamin D supplementation or
probably caused by the resultant hypercalcemia. Of sun exposure may vary by patient, so clinicians may
note, sun exposure alone cannot lead to vitamin D need to continue to monitor abnormal levels.
intoxication as excess vitamin D3 is destroyed by
sunlight. References
1. Brannon PM, Yetley EA, Bailey RL, Picciano MF.
Overview of the conference “Vitamin D and Health
Treatment in the 21st Century: an Update”. Am J Clin Nutr
2008;88(Suppl):483S–90S.
Given concern about skin cancer, many patients
2. Wolf M, Shah A, Gutierrez O, et al. Vitamin D
and clinicians are cautious regarding sun exposure
levels and early mortality among incident hemodial-
recommendations. However, exposure of arms and ysis patients. Kidney Int 2007;72:1004 –13.
legs for 5 to 30 minutes between the hours of 10 am 3. Inaguma D, Nagaya H, Hara K, et al. Relationship
and 3 pm twice a week can be adequate to prevent between serum 1,25-dihydroxyvitamin D and mor-
vitamin D deficiency.59 tality in patients with pre-dialysis chronic kidney
Natural dietary sources of vitamin D include disease. Clin Exp Nephrol 2008;12:126 –31.
salmon, sardines, mackerel, tuna, cod liver oil, shi- 4. Al-Aly Z. Vitamin D as a novel nontraditional risk
itake mushrooms, and egg yolk.58 Fortified foods factor for mortality in hemodialysis patients: the
need for randomized trials. Kidney Int 2007;72:909 –
include milk, orange juice, infant formulas, yogurts, 11.
butter, margarine, cheeses, and breakfast cereals.59 5. Dobnig H, Pilz S, Scharnagl H, et al. Independent
Over-the-counter multivitamin supplements fre- association of low serum 25-hydroxyvitamin D and
quently contain 400 IU of vitamins D1, D2, or D3. 1,25-dihydroxyvitamin D levels with all-cause and
Alternatively, over-the-counter vitamin D3 supple- cardiovascular mortality. Arch Intern Med 2008;168:
ments can be found in 400, 800, 1000, and 2000 IU 1340 –9.
strengths. Prescription-strength supplementation 6. Melamed ML, Michos ED, Post W, Astor B. 25-
hydroxyvitamin D levels and the risk of mortality in
choices include vitamin D2 (ergocalciferol), which
the general population. Arch Intern Med 2008;168:
provides 50,000 IU per capsule, and vitamin D2 liquid 1629 –37.
(drisdol) at 8000 IU/mL.59 7. Autier P, Gandini S. Vitamin D supplementation and
To prevent vitamin D deficiency in healthy pa- total mortality. Arch Intern Med 2007;167:1730 –7.
tients, the 1997 Institute of Medicine recommen- 8. Wang TJ, Pencina MJ, Booth SL, et al. Vitamin D
dations suggested a daily vitamin D intake of 200 deficiency and risk of cardiovascular disease. Circu-
IU for children and adults up to 50 years of age; lation 2008;117:503–11.
400 IU for adults 51 to 70 years of age; and 600 IU 9. Mathieu C, Gysemans C, Giulietti A, Bouillon R. Vi-
tamin D and diabetes. Diabetologia 2005;48:1247–57.
for adults 71 years or older.64 The upper limit
10. Sloka S, Grant M, Newhook L. The geospatial re-
recommended was 2000 IU daily. However, some
lation between UV solar radiation and type 1 diabe-
experts consider this to be too low and recommend tes in Newfoundland. Acta Diabetol 2009; epub
that children and adults without adequate sun ex- ahead of print.
posure consume 800 to 1000 IU daily to achieve 11. Mathieu C, van Etten E, Decallonne B, et al. Vita-
adequate serum vitamin D levels.59 min D and 1,25-dihydroxyvitamin D3 as modulators
Treatment recommendations vary depending on in the immunesystem. J Steroid Biochem Mol Bio
the cause of the deficiency. For example, patients 2004;89 –90:449 –52.
12. Palomer X, González-Clemente JM, Blanco-Vaca F,
with chronic kidney disease are recommended to
Mauricio D. Role of vitamin D in the pathogenesis
have 1000 IU of vitamin D3 daily.59 The expected of type 2 diabetes mellitus. Diabetes Obes Metab
blood level response to a given vitamin D dose 2008;10:185–97.
varies, probably because of differences in the cause 13. Danescu LG, Levy S, Levy J. Vitamin D and diabe-
of the deficit as well as the starting point for cor- tes mellitus. Endocrine 2009;35:11–7.
rection. A recent editorial reported that supple- 14. Pittas AG, Lau J, Hu FB, Dawson-Hughes B. The

704 JABFM November–December 2009 Vol. 22 No. 6 http://www.jabfm.org

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
role of vitamin D andcalcium in type 2 diabetes. A 29. Kampman MT, Wilsgaard T, Mellgren SI. Outdoor
systematic review and meta-analysis. J Clin Endocri- activities and diet in childhood and adolescence re-
nol Metab 2007;92:2017–29. late to MS risk above the Arctic Circle. J Neurol
15. Rodriguez-Martinez MA, Garcia-Cohen EC. Role 2007;254:471–7.
of Ca⫹⫹ and vitamin D in the prevention and treat- 30. Willer CJ, Dyment DA, Sadovnick AD, et al. Tim-
ment of osteoporosis. Pharmacol Ther 2002;93:37– ing of birth and risk of multiple sclerosis: popula-
49. tion-based study. BMJ 2005;330:120.
16. Cranney A, Weiler HA, O’Donnell S, Puil L. Sum- 31. van der Mei IAF, Ponsonby A, Dwyer T, et al. Past
mary of evidence-based review on vitamin D efficacy exposure to sun, skin phenotype, and risk of multiple
and safety in relation to bone health. Am J Clin Nutr sclerosis: case-control study. BMJ 2003;327:316.
2008;88(Suppl):513S–9S. 32. Holick MF. Sunlight and vitamin D for bone health
17. Avenell A, Gillespie WJ, O’Connell DC. Vitamin D and prevention of autoimmune diseases, cancers, and
and vitamin D analogues for preventing fractures cardiovascular disease. Am J Clin Nutr 2004;80(6
associated with involutional and postmenopausal os- Suppl):1678S– 88S.
teoporosis. Cochrane Database Syst Rev 2005;(3): 33. Adorini L, Penna G. Control of autoimmune dis-
CD000227. eases by the vitamin D endocrine system. Nat Clin
18. Boonen S, Lips P, Bouillon R, Bischoff-Ferrari HA, Pract Rheumatol 2008;4:404 –12.
Vanderschueren D, Haetiens P. Need for additional 34. Szodoray P, Nakken B, Gaal J, et al. The complex
calcium to reduce the risk of hip fracture with vita- role of vitamin D in autoimmune diseases. Scand
min D supplementation: evidence from a compara- J Immunol 2008;68:261–9.
tive metaanalysis of randomized controlled trials.
35. Munger KL, Levin LI, Hollis BW, Howard NS,
J Clin Endocrinol Metab 2007;92:1415–23.
Ascherio A. Serum 25-hydroxyvitamin D levels and
19. Bischoff-Ferrari HA, Willett WC, Wong JB, et al.
risk of multiple sclerosis. JAMA 2006;296:2832– 8.
Prevention of nonvertebral fractures with oral vita-
min D dose dependency. A meta-analysis of random- 36. Kragt JJ, van Amerongen BM, Killestein J, et al.
ized controlled trials. Arch Intern Med 2009;169: Higher levels of 25-hydroxyvitamin D are associated
551– 61. with a lower incidence of multiple sclerosis only in
women. Mult Scler 2009;15:9 –15.
20. Prince RL, Adustin N, Devine A, et al. Effects of
ergocalciferol added to calcium on the risk of falls in 37. van der Mei IAF, Ponsonby A, Dwyer T, et al.
elderly high-risk women. Arch Intern Med 2008; Vitamin D levels in people with multiple sclerosis
168:103– 8. and community controls in Tasmania, Australia.
J Neurol 2007;254:581–90.
21. Broe KE, Chen TC, Weinberg J, et al. A higher dose
of vitamin D reduces the risk of falls in nursing home 38. Soilu-Hänninen M, Airas L, Mononen I, Heikkilä A,
residents: a randomized multiple-dose study. J Am Viljanen M, Hänninen A. 25-Hydroxyvitamin D lev-
Geriatr Soc 2007;55:234 –9. els in serum at the onset of multiple sclerosis. Mult
Scler 2005;11:266 –71.
22. Bischoff-Ferrari HA, Dawson-Hughes B, Willett
WC, et al. Effect of vitamin D on falls: a meta- 39. Smolders J, Menheere P, Kessels A, Damoiseaux J,
analysis. JAMA 2004;291:1999 –2006. Hupperts R. Association of vitamin D metabolite
levels with relapse rate and disability in multiple
23. Osborne JE, Hutchinson PE. Vitamin D and sys-
sclerosis. Mult Scler 2008;14:1220 – 4.
temic cancer: is this relevant to malignant mela-
noma? Br J Dermatol 2002;147:197–213. 40. Brown SJ. The role of vitamin D in multiple sclero-
24. Giovannucci E, Liu Y, Rimm EB, et al. Prospective sis. Ann Pharmacother 2006;40:1158 – 61.
study of predictors of vitamin D statius and cancer 41. Munger KL, Zhang S.M, O’Reilly E, et al. Vitamin
incidence and mortality in men. J Natl Cancer Inst D intake and incidence of multiple sclerosis. Neu-
2006;98:451–9. rology 2004;62:60 –5.
25. Freedman DM, Looker AC, Chang SC, Graubard 42. Smolders J, Damoiseaux J, Menheere P, Hupperts R.
BI. Prospective study of serum vitamin D and cancer Vitamin D as an immune modulator in multiple
mortality in the United States. J Natl Cancer Inst sclerosis, a review. J Neuroimmunol 2008;194:7–17.
2007;99:1594 – 602. 43. Kimball SM, Ursell MR, O’Connor P, Vieth R.
26. Garland CF, Garland FC, Gorham ED, et al. The Safety of vitamin D3 in adults with multiple sclero-
role of vitamin D in cancer prevention. Am J Public sis. Am J Clin Nutr 2007;86:645–51.
Health 2006;96:252– 61. 44. Buell JS, Dawson-Hughes B. Vitamin D and neuro-
27. Lappe JM, Travers-Gustafson D, Davies KM, cognitive dysfunction: preventing “D”ecline? Mol
Recker RR, Heaney RP. Vitamin D and calcium Aspects Med 2008;29:415–22.
supplementation reduces cancer risk: results of a 45. Oudshoorn C, Mattace-Raso FU, van der Velde N,
randomized trial. Am J Clin Nutr 2007;85:1586 –91. Colin EM, van der Cammen TJ. Higher serum vi-
28. Ebers GC. Environmental factors and multiple scle- tamin D3 levels are associated with better cognitive
rosis. Lancet Neurol 2008;7:268 –77. test performance in patients with Alzheimer’s dis-

doi: 10.3122/jabfm.2009.06.090037 Vitamin D: An Evidence-Based Review 705

 J Am Board Fam Med: first published as 10.3122/jabfm.2009.06.090037 on 6 November 2009. Downloaded from http://www.jabfm.org/ on 17 May 2019 by guest. Protected by copyright.
ease. Dement Geriatr Cogn Disord 2008;25:539 – 56. Wagner CL, Hulsey TC, Fanning D, Ebeling M,
43. Hollis BW. High-dose vitamin D3 supplementation
46. Straube S, Andrew Moore R, McQuay HJ. Vitamin in a cohort of breastfeeding mothers and their in-
D and chronic pain. Pain 2009;141:10 –3. fants: a 6-month follow-up pilot study. Breastfeed
Med 2006;1:59 –70.
47. Heath KM, Elovic EP. Vitamin D deficiency: impli-
cations in the rehabilitation setting. Am J Phys Med 57. Hollis BW, Taylor SN. Vitamin D requirements in
Rehabil 203;85:916 –23. pregnancy and lactation. Poster abstract at the 135th
American Public Health Association Annual Meet-
48. Wagner CL, Greer FR, American Academy of Pe-
ing and Exposition; November 2007; Washington,
diatrics Section on Breastfeeding, American Acad-
DC.
emy of Pediatrics Committee on Nutrition. Preven-
tion of rickets and vitamin D deficiency in infants, 58. Working Group of the Australian and New Zealand
children, and adolescents. Pediatrics 2008;122:1142– Bone and Mineral Society, Endocrine Society of
52. Australia, Osteoporosis Australia. Vitamin D and
adult bone health in Australia and New Zealand: a
49. Hollis BW, Wagner CL. Assessment of dietary vi- position statement. Med J Aust 2005;182:281–5.
tamin D requirements during pregnancy and lacta-
59. Holick MF. Vitamin D deficiency. N Engl J Med
tion. Am J Clin Nutr 2004;79:717–26.
2007;357:266 – 81.
50. Gartner LM, Greer FR, American Academy of Pe-
60. Millen AE, Bodnar LM Vitamin D assessment in
diatrics Section on Breastfeeding, American Acad-
population-based studies: a review of the issues. Am J
emy of Pediatrics Committee on Nutrition. Preven-
Clin Nutr 2008;87(suppl):1102S–5S.
tion of rickets and vitamin D deficiency: new
61. Binkley N, Krueger D, Cowgill CS, et al. Assay
guidelines for vitamin D intake. Pediatrics 2003;111:
variation confounds the diagnosis of hypovitaminosis
908 –10.
D: a call for standardization. J Clin Endocrinol
51. Weisberg P, Scanlon K, Li R, Cogswell ME. Nutri- Metab 2004;89:3152–7.
tional rickets among children in the United States:
62. Pollack A. Quest acknowledges errors in vitamin D
review of cases reported between 1986 and 2003.
tests. Available at: http://www.nytimes.com/2009/
Am J Clin Nutr 2004;80(6 Suppl):1697S–1705S.
01/08/business/08labtest.html. Accessed 17 Septem-
52. Vitamin D supplementation: recommendations for ber 2009.
Canadian mothers and infants. Paediatr Child 63. McCarty CA. Sunlight exposure assessment: can we
Health 2007;12:583–9. accurately assess vitamin D exposure from sunlight
53. Javaid MK, Crozier SR, Harvey NC, et al. Maternal questionnaires? Am J Clin Nutr 2008;87(Suppl):
vitamin D status during pregnancy and childhood 1097S–101S.
bone mass at age 9 years: a longitudinal study. Lan- 64. Standing Committee on the Scientific Evaluation of
cet 2006;367:36 – 43. Dietary Reference Intakes Food and Nutrition
54. Camargo CA Jr, Rifas-Shiman SL, Litonjua AA, et Board, Institute of Medicine. Vitamin D. In: Dietary
al. Maternal intake of vitamin D during pregnancy reference intakes for calcium, phosphorus, magne-
and risk of recurrent wheeze at 3 y of age. Am J Clin sium, vitamin D, and fluoride. Washington, DC:
Nutr 2007;85:788 –95. National Academy Press; 1997:250 – 87.
55. Tsang R, Zlotkin S, Nichols B, Hansen J. Nutrition 65. Vieth R, Bischoff-Ferrari H, Boucher BJ, et al. The
during infancy: rinciples and ractice, 2nd ed. Cincin- urgent need to recommend an intake of vitamin D
nati, OH: Digital Education Publishing; 1997. that is effective. Am J Clin Nutr 2007;85:649 –50.

706 JABFM November–December 2009 Vol. 22 No. 6 http://www.jabfm.org