Vol.3, No.

12, 752-756 (2011) Health

doi:10.4236/health.2011.312125

Anthropometric measurements at birth as predictor of

low birth weight
Negar Sajjadian1, Hamideh Shajari1, Farnoosh Rahimi1, Ramin Jahadi2, Michael G. Barakat3
1
Department of Neonatology, Shariati Hospital, Tehran University of Medical Sciences, Tehran, Iran;
[email protected], [email protected], [email protected]
2
Departmnt of Plastic and Reconstructive Surgery, 15 Khordad Hospital, Shahid Beheshti Medical University, Tehran, Iran;
[email protected]
3
Biomedical Studies, La Sierra University, Riverside, USA.

Received 22 May 2011; revised 16 July 2011; accepted 20 September 2011.

ABSTRACT as the optimal surrogate indicators of LBW

babies. The optimal cut-points for chest circum-
In developing countries, low birth weight (BW <
ference and arm circumference to identify LBW
2500 grams) accounts for 60% - 80% of neonatal
newborns were ≥31.2 cm and ≥10.2 cm respec-
deaths. Early identification and referral of LBW
tively. Conclusions: Chest and mid arm circum-
babies for extra essential newborn care is vital
ferences were the best anthropometric surrogates
in preventing neonatal deaths. Studies carried
of LBW among studied Iranian population. Further
out in different populations have suggested that
studies are needed in the field to cross-validate
the use of newborn anthropometric surrogates
our results. anthropometric values are simple,
of birth weight may be a simple and reliable
practicable, quick and reliable indicator for pre-
method to identify LBW babies. previous studies
dicting LBW newborns in the community and can
reported correlation between birth weight to
be easily measured by paramedical workers in
several anthropometric measurements and their
developing nation.
predictive value. We aimed to evaluate the
correlation between birth length, head, chest,
Keywords: Low Birth Weight; Anthropometric
and mid arm circumferences to birth weight.
Measurements; Neonatal Mortality
Methods: A cross sectional study has been
conducted in SHARIATI Hospital in Tehran, from
1. INTRODUCTION
September 2008 to February 2009. All Consecu-
tive full-term. Single ton, live born babies were Of the approximately four million global neonatal
included and anthropometric measurements deaths that occur annually, 98% occur in developing
carried out within 48 hours after birth by authors. countries, where most newborns die at home while they
Birth weight was measured by digital scale are being cared by mothers, relatives, and traditional
within the first 24 hours after delivery. Birth birth attendants [1]. About 38% of total under-five mor-
length by somatometer and head, chest, mid tality occurs during the neonatal period and nearly three
arm circumferences were measured 2 times by quarters of these deaths occur during the first week of
using plastic measuring tape. Result: Out of 500 life [2]. Globally, about one-sixth of all newborns are
newborn studied. 52.2% were male and 47.8% low birth weight (LBW, <2500 grams), which is single
were female. The mean birth weight was 3195.4 most important underlying risk factor for neonatal deaths
± 399.9 gram and 3.8% of newborns were low [1,3]. Only about half of the newborns are weighed at
birth weight. It was evident a positive correlation birth and for a smaller proportion of them gestational
of birth weight to all such anthropometric mea- age is known [4]. An estimated 18 million babies are
surements with the highest correlation coeffi- born with LBW [5]. They account for 60% - 80% of
cient for chest circumference (r: 0.74). By ROC- neonatal deaths [6]. Moreover, LBW babies who survive
AUC analyses, chest circumference (AUC = 0.91, the critical neonatal period may suffer impaired physical
95% CI 0.84 to 0.97) and arm circumference and mental growth. Therefore, an early identification and
(AUC = 0.87, 95% CI 0.79 to 0.95) were identified prompt referral of LBW newborns is vital in preventing

Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/

 N. Sajjadian et al. / Health 3 (2011) 752-756 753
753

neonatal deaths. Available evidence from resource-poor sented complications during pregnancy (preclampsia,
settings shows that extra essential newborn care for Diabetes, addiction and smoking) or newborns with ma-
LBW babies can reduce the number of neonatal deaths jor congenital anomaly, hydropic feature or intrauterine
by 20% - 40% [7]. Most neonatal deaths occur in the growth retardation (IUGR) were excluded.
community and some interventions, including vitamin A Birth weight with naked neonate in supine position was
supplementation, newborn skin cleansing with chlor- obtained soon after birth by digital scale with 10 gram
hexidine and topical emollient therapy may be targeted subdivision. All other anthropometric Variable including
preferentially to LBW infants to reduce mortality risk. chest, head, mid arm Circumferences were measured by
Thus, continued efforts are required to describe optimal non extendable measuring tape, with a width of 1.0 cm
methods for identifying these high-risk infants in the and subdivisions of 0.1 cm. and birth length was meas-
community. In resource-poor settings, a large proportion ured by somatometer
of deliveries take place at home and birth-weight is most Head circumference was obtained by placing tape
often not recorded. Therefore, there is a need to develop along the largest occipitofrontal diameter along over the
simple, inexpensive and practical methods to identify occiput and eyebrow.
LBW newborns soon after birth [8]. One such method The chest circumference was measured by placing
may be the use of anthropometric surrogates to identify measuring tape along the point of nipples. The mid arm
LBW babies. A number of studies have focused on circumference was obtained from the left arm with el-
measuring the circumference of the head, chest, mid bow at the mid point between acromion and olecranon,
upper-arm, thigh or calf and observed the correlation with the newborn was located in dorsal decubitus with
with continuous measurements on a gold standard arm lying laterally to the trunk.
weighing scale (Bhargava et al., 1985; Singh et al., 1988; The length was measured with the newborn in supine
WHO Collaborative study of birth weight surrogates, position with full extension of knee and distance between
1993; Dhar et al., 2002). In general, chest circumference top of head and heel when pressed against a vertical surface
has performed better than other measures and has been and role on a stabilizing board was measured.
recommended for continued investigation, although in- A total of two consecutive measurements were taken
vestigators have demonstrated correlations between birth for each variable and the mean values were recorded.
weight and mid upper-arm circumference (Sauerborn et Continuous variables are reported as mean and stan-
al., 1990, calf (Gupta et al., 1996) (Samal and Swain, dard deviation while between-gender comparisons of
2001 or thigh (Sharma et al., 1989) that are as strong as continuous variables were performed using independent
with chest circumference. sample t-test. Pearson’s product-moment correlation co-
Most suitable and reliable anthropometric surrogate to efficient was used to assess the association between an-
identify LBW Iranian newborns and its cut-off point to thropometric measurements. Receiver operating charac-
identify LBW newborns is not known. Therefore, we teristic (ROC) curves were used to evaluate the accuracy
carried out this study with following objectives: of different anthropometric measurements to predict
1) to identify a suitable anthropometric surrogate to LBW. Non parametric receiver operating characteristic
identify LBW babies and analysis was done to compare the overall utility of an-
2) to determine its cut-off value to identify LBW ba- thropometric measurements for Identifying LBW infant.
bies. Multivariate linear regression with backward stepwise
method was used for estimation of birth weight by an-
2. MATERIALS AND METHODS thropometric measurment.
Sensitivity, specificity were calculated at all cut-points
A cross sectional study was performed among all term
for any anthropometric measurement. We choose as “op-
single tone neonate born at SHARIATI hospital (Tehran,
timum” the cut-point with the highest [(sensitivity + speci-
Iran), a reference center for high risk pregnancies, from
ficity)/2] ratio. This criterion was chosen to allow compari-
September 2008 to February 2009.
son with previous studies available in the literature. The
The study group consisted of all consecutive full
data analysis was done by the spss version 11.5 and A
term―single tone newborn with gestational age of be-
P-value <0.05 was considered statistically significant.
tween 37 weeks and 41 weeks and 6 days as estimate by
maternal last menstrual period (LMP) date and first tri-
3. RESULTS
mester sonography when their differences are not more
than week. Preterms were excluded because they are A total of 500 newborns (52.2% male and 47.8% fe-
probably low birth weight. male) were studied.
These newborn were examined by the authors within The mean birth weight was 3195.48 ± 399.92 gram.
their first 48 hours of life. Newborns whose mother pre- 3.8% of patients were LBW. Summary measures of

Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/

754 N. Sajjadian et al. / Health 3 (2011) 752-756

weight, and anthropometric variable are presented in for predicting birth weight by anthropometric measure-
Table 1. In our study there were significant differences ments (Table 2).
in birth weight and anthropometric measurements be- The best discrimination of LBW, as detected by ROC-
tween male and female newborns (P-value <0.05) the AUC, was obtained by chest circumference (AUC =
males had higher birth weight and all anthropometric 0.91, 95% CI 0.84 to 0.97) followed by lenght (AUC =
variable except mid arm to head circumference ratio. 0.89, 95% CI 0.79 to 0.99), head circumference (AUC =
The anthropometric measurements were correlated 0.87, 95% CI 0.80 to 0.94), mid arm circumference
with birth weight with significant p value, the maximum (AUC = 0.87, 95% CI 0.79 to 0.95) and mid arm to head
correlation with birth weight was observed for chest cir- circumference ratio (AUC = 0.74, 95% CI 0.63 to 0.85)
cumference (r = 0.74) and the mid arm circumference (r The sensitivity, specificity, for classifying infants into
= 0.70), head circumference (r = 0.65). Length (r = 0.61) LBW status was shown in Table 3. An optimum cut off
and a weak correlation was seen with mid arm circum- point identifying LBW were 31.2 cm for chest circum-
ference to head circumference (r = 0.44) with the best ference 10.2 cm for mid arm circumference, 33.2 cm for
correlation coefficient observed for the weight-chest head circumference, 48.5 cm for length and 0.3 for mid
circumference association. head circumference, 48.5 cm for length and 0.3 for mid
Multivariate Linear regression analysis was conducted arm to head circumference ratio (Table 3).

Table 1. Descriptive statistics of birth weight and anthropometric measurements.

Variable Sex Mean Standard Deviation Minimum Maximum

Combined 3195.48 399.92 2100 4550

Weight Male 3269.43 408.20 2200 4550

Female 3114.73 375.18 2100 4150

Combined 10.79 0.89 9 13

Mid Arm Circumference Male 10.88 0.87 9 13

Female 10.70 0.91 9 13

Combined 32.78 1.75 28 38

Chest Circumference Male 33.9 1.73 28 38

Female 32.43 1.72 28 37

Combined 34.67 1.33 31 38.5

Head Circumference Male 35.01 1.34 31 38.5

Female 34.29 1.21 31 38

Combined 0.31 0.02 0.25 0.37

MIC/HC Ratio Male 0.31 0.02 0.26 0.37

Female 0.31 0.02 0.25 0.37

Combined 50.43 3.92 44 58

Length Male 50.81 2.01 46 58

Female 50.01 1.85 44 56

Table 2. Linear regression equation for estimation of birth weight.

Variable Regresion Equation F (P-value) Adjusted R2

MAC BW = –162.58 + 311.19 (MAC) 477.48 (0.00) 0.488
MAC/HC BW = 839.45 + 7569.4 (MAC/HC) 121.210 (0.00) 0.194
CC BW = –2329.13 + 168.55 (CC) 606.38 (0.00) 0.548
HC BW = –3596.83 + 195.90 (HC) 371.23 (0.00) 0.426
LE BW = 3041.77 + 123.67 (L) 298.8 (0.00) 0.374

Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/

 N. Sajjadian et al. / Health 3 (2011) 752-756 755
755

Table 3. Sensitivity and specifity of optimum cutoff points of have attempted to identify a suitable anthropometric
anthropometrics. surrogate to identify LBW babies which is reliable, sim-
Variable Cut off limit Sensitivity Specifity ple, and logistically feasible in field conditions. Some
Chest circumference 31.25 84% 81% studies have recommended that CHC, MUAC and HC
Head circumference 33.25 73% 85% may be used as anthropometric surrogates to identify
Mid arm circumference 10.25 94% 68% LBW babies [11-15,17-20]. Therefore we considered all
Length 48.5 84% 88% these anthropometric measurement. In our study CHC
Mid arm circumference was identified as a suitable surrogate to identify LBW
0.303 73% 62%
Head circumference babies.
In the present study, the maximum correlation with
4. DISCUSSION birth weight was observed for chest circumference (r =
0.74) and the mid arm circumference (r = 0.70) so chest
The prime concern of the present study was to identify and mid-arm circumference had the best correlation with
the best suitable surrogate parameter, proxy, to birth birth weight and are good predictors of LBW neonates.
weight, which when used by the health personnel in do- According to other studies (mullany et al., 2007 WHO
miciliary outreach will detect the maximum number of at Collaborative study of birth weight surrogates, 1993)
risk infants for providing them with timely and needed The correlations between birth weight, arm circumfer-
intervention strategy. ence and chest circumference are high ranging from
The mean birth weight and anthropometrics in our 0.60 to 0.95 and suggested that chest circumference is
population is higher than some previous mentioned study. the Optimal anthropometric measure for establishing
A WHO multicenter study reported that the average birth cutoffs for the identification of LBW infants [9,19].
weight was 2630, 2780 and 3840 for newborns in India, The preset study shows that chest circumference at a
Nepal and Sri Lanka respectively [14]. Higher mean cut off limit of 31.2 cm is affective in detecting low birth
birth weight may be because only the full term singleton weight infant with a sensitivity of 84% and specifity of
live births were included in our study. Previous studies 81%. Cupta et al. showed cut off limit 30.1 cm with
did not specify such criteria [11,16,19]. Birth weights of specifity 69 and sensitivity of 83% and Virdi et al. study
the newborns born before completion of 37 weeks of cut off point of 30 with sensitivity 60 and specifity 0.3
gestation (full term) may also have been included in the [12,13]. A WHO collaborative study has recommended
studies cited above that CHC of 29 centimeters and 30 centimeters may
Our data relieved that positive correlation between all identify “highly at risk” and “at risk” newborns respec-
studied parameters and birth weight is present. In our tively [19]. In our study maximum sensitivity and speci-
study there were significant differences in birth weight ficity for CHC was at CHC of 30.8 centimeters. The
and anthropometric measurements between male and higher mean birth weight of newborns may be the reason
female newborns. Males had higher mean birth weight for a slightly higher cut-off point obtained in our study.
and anthropometrics than females. This finding is similar We considered only full-term deliveries, which was
to Dhar study that strongest correlation was present be- unlike earlier studies [12,17,20].
tween CHC and birth weight was observed (r = 0.84). It Thus, it is evident from analysis of our data that chest
may be due to large size of our population that make circumference is the best suitable and simple surrogate
these difference significant however they lack clinical parameter that could be used in the domiciliary outreach
significance. when it is impossible to record weight of baby at birth.
The percent of LBW in our population were lower For health personnel who are working in the community
than the Nepal (8.5%) and Tanzania (18% - 8%) study can use color coded tapes indicting weight <2500 grams.
[8,9]. The reason of these finding may be related to dif- We recommend the use of chest rather than arm cir-
ferent characteristics of population studied (genetic, nu- cumference as a surrogate for birth weight for two rea-
tritional, environmental background) and because we sons. First, it is simpler to measure identification of the
exclude all preterm neonates. In WHO collaborative nipple line is easier, making measurement more opera-
study of birth weight surrogates Clear differences were tionally feasible than that of mid-arm circumference.
seen between the centers in terms of the means and tenth Second, our findings suggest that measurement of both
centiles of both birth weight and the anthropometric arm circumference and chest circumference is of little
measures. The values confirm the expected regional dif- additional value in predicting low-birth-weight babies.
ferences, since centers in South Asia, such as Delhi and We conclude that simple measurements such as chest
Chandigarh, have on average the lowest values, whereas circumference and also other mentioned anthropometric
those in Europe, such as St. Peters. Many researchers measure can simply and practicably identify infants with

Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/

756 N. Sajjadian et al. / Health 3 (2011) 752-756

LBW when recording birth weight is not feasible. these cumference for identification of low birth weight babies.
measurements are easy to learn and can conveniently be Indian Pediatrics, 38, 934-935.
[14] World Health Organization (1994) Multi-centre study on
introduced in to the existing system of health care to use
low birth weight and infant mortality in India, Nepal and
by paramedical workers to detect neonate who are LBW Sri Lanka. Southeast Asia Regional Office, World Health
and at risk. A color coded, measuring tape may be sug- Organization, New Delhi.
gested for use by health workers or family members to [15] Ahmed, F.U., Karim, E. and Bhuiyan, S.N. (2000)
identify LBW newborns in home setting. Mid-arm circumference at birth as predictor of low birth
weight and neonatal mortality. Journal of Biosocial Sci-
ence, 32, 487-493. doi:10.1017/S0021932000004879
REFERENCES [16] Das, J.C., Afroze, A., Khanam, S.T. and Paul, N. (2005)
Mid-arm circumference: an alternative measure for scree-
[1] The world health report. The newoborn health that went ning low birth weight babies. Bangladesh Medical Re-
unnoticed, perinatal mortality. A listing of available in- search Council Bulletin, 31, 1-6.
formation. World Health Organization, Geneva, 1996. [17] Ezeaka, V.C., Egri-Okwaji, M.T., Renner, J.K. and
[2] Lawn, J.E., Cousens, S. and Zupan, J. (2005) Lancet Grange, A.O. (2003) Anthropometric measurements in
neonatal survival steering team: 4 million neonatal deaths: the detection of low birth weight infants in Lagos. The
when? Where? Why? Lancet, 365, 891-900. Nigerian Postgraduate Medical Journal, 10, 168-172.
[3] Gogia, S. and Sachdev, H.S. (2010) Home visits by [18] Hossain, M.M., Habib, M. and DuPont, H.L. (1994) As-
community health workers to prevent neonatal deaths in sociation between birth weight and birth arm circumfer-
developing countries: A systematic review. Bulletin of the ence of neonates in rural Egypt. Indian Journal of Pedi-
World Health Organization, 88, 658 atrics, 61, 81-87. doi:10.1007/BF02753565
[4] Blanc, A.K. and Wardlaw, T. (2005) Monitoring low birth [19] (1993) Use of a simple anthropometric measurement to
weight: An evaluation of international estimates and an predict birth weight. WHO collaborative study of birth
updated estimation procedure. Bulletin of the World weight surrogates. Bulletin of the World Health Organi-
Health Organization, 83, 178-185. zation, 71, 157-163.
[5] United Nations Children’s Fund (UNICEF) (2005) The [20] Mullany, L.C., Darmstadt, G.L., Jhatry, S.K. et. al. (2007)
state of the world’s children. New York. Relationship between the surrogate anthropometric
[6] Bang, A., Reddy, M.H. and Deshmukh, M.D. (2002) measures, foot length and chest circumference and birth
Child mortality in Maharashtra. Economic Political weight among newborn of sarlahi, Nepal. European
Weekly, 37, 4947-4965. Journal of Clinical Nutrition, 61, 40-46.
[7] Darmstadt, G.L., Bhutta, Z.A., Cousens, S., Adam, T., doi:10.1038/sj.ejcn.1602504
Walker, N. and de Bernis, L. (2005) Lancet neonatal sur- [21] Hanimi, R.M. and Abhay T. (2005) How to identify neo-
vival steering team: Evidence-based, cost-effective in- nate at risk of death in rural India: clinical criteria for the
terventions: How many newborn babies can we save? risk approach. Journal of Perinatology, 25, s44-s50.
Lancet, 365, 977-998. doi:10.1038/sj.jp.7211272
doi:10.1016/S0140-6736(05)71088-6 [22] Kliegman, R.M., Jenson, H.B., Marcdanle, K.J. and Be-
[8] Mullany, L.C., Darmstadt, G.L., Coffey, P., Khatry, S.K., harman, R.E. (2006) Nelson essentials of pediatrics. 5th
LeClerq, S.C. and Tielsch, J.M. (2006) A low cost, colour Edition, Elsevier Saunders Press, Amsterdam.
coded, hand held spring scale accurately categorises birth [23] Eghba, L.F. (2007) Evaluation of etiology of low birth
weight in low resource settings. Archives of Disease in weight. Iranian Pediatric Journal, 17, 27-33.
Childhood, 91, 410-413. doi:10.1136/adc.2005.088781 [24] Bindu, N.R., Elizabeth, K.E., Varghese, G. and Varghese,
[9] Ramaiya, C., Msamanga, G., Massawe, S., Mpanju, W. S. (2006) Mid arm circumference (MAC) and body mass
and Nywalle, E. (1994) Newborn’s arm circumference as index (BMJ)―The tow important auxologic parameters
screening tool of low birth weight in temeke pistric, in neonates. Journal of tropical pediatrics, 52, 341-345.
peres salaam, Tunzania. Tropical and Geographical [25] Seeramarddy, C.T., Chuni, N., Patil, R., Singh, D. and
Medicine, 46, 318-321. Shakya, B. (2008) Anthropometric surrogates to identify
[10] Naik, D.B., Kulkarni, A.P. and Aswar, N.R. (2003) Birth low birth weight Nepalese newborn: A hospital-based
weight and anthropometry of newborns. Indian Journal study. BMC Pediatrics, 8, 16.
of Pediatrics, 2003, 70, 145-146. doi:10.1186/1471-2431-8-16
doi:10.1007/BF02723742 [26] (1993) WHO Collaborative study of birth weight surro-
[11] Samal, G.C. and Swain, A.K. (2001) Calf circumference gates―Use of a simple anthropometric measurement to
as an alternative to birth weight to predict low birth predict birth weight. Bulletin of the World Health Or-
weight babies. Indian Pediatrics, 38, 275-277. ganization, 71, 157-163.
[12] Gupta, V., Hatwal, S.K., Mathur, S., Tripathi, V.N., [27] Gupta, V., Hatwal, S.K., Mathur, S., et al. (1996) Calf
Sharma, S.N., Saxena, S.C. and Khadwal, A. (1996) Calf circumference as a predictor of low birth babies. Indian
circumference as a predictor of low birth weight babies. Pediatrics, 33, 119-121.
Indian Pediatrics, 33, 119-121.
[13] Virdi, U.S., Jain, B.K. and Singh, H. (2001) Calf cir-

Copyright © 2011 SciRes. Openly accessible at http://www.scirp.org/journal/HEALTH/