Birth weight centiles are available for developed countries [1,2] but are not easily available in developing countries although the importance of ethnic-specific standards has been acknowledged [3]. Sex- and parity-specific birthweight charts for gestational age have been published based on births in tertiary hospitals but community-based birth weight centile charts representing all births in a population are not available [4,5]. An earlier comparison of birth weights in rural south India against an Indian standard showed that the Indian standard was descriptive rather than normative as the proportion small-for-gestation (SGA, below the 10th percentile) was much lower than the proportion who were low birth weight (below 2500g), while comparison with a Canadian standard showed a high rate of SGA [6].

The primary objective of this study was to construct birthweight centiles for children born to healthy mothers in a rural block in southern India between 2003-2012 and to compare with the birth weights of developed countries as well as the WHO child growth standards. The secondary objective was to compare mortality rates within two months of birth, for children with birth weights below the 3rd centile and above the 97th centile to those with birthweights between the 3rd and 97th centiles.

The data on births occurring in the block was obtained from the computerised database maintained by the health information system of the secondary hospital [6]. Only births with recorded birth weights and gestational age at the time of delivery (determined by either the date of the last menstrual period or ultrasound scans) were included in the final database. Tukey’s method was followed to exclude extreme values/outliers of birthweights, which revealed seven outliers [7]. The upper limit (Tukey’s outer fence) was taken as the third quartile value plus 3 times the interquartile range while the lower limit of normal (Tukey’s inner fence) was taken as the first quartile minus three times the interquartile range.

As the birthweight centiles were meant to be descriptive of healthy pregnancies, mothers diagnosed with medical conditions known to have an effect on birthweight, including gestational diabetes, pregnancy induced hypertension, maternal heart disease, short stature (taken as maternal height less than 140 cm), and hemoglobin less than 10 g/dL were excluded [8].

Although most infant deaths occur in the first month of life, we used under-two month mortality rates as a means of validation of the birth weight centile groupings, in order to account for all deaths may be associated with low birth weight. The mortality rates were computed using available mortality data for children born to permanent residents of the study area, as follow up was done mainly for these children and not for mothers who were temporary residents (mothers who had only come for antenatal care/delivery to their maternal homes). The number of deaths in various centile groups among the children included in the birth weight centile data were used for obtaining these rates, thus giving under two month mortality rates for singleton live born children born to mothers without the antenatal high risk factors mentioned earlier.

The total number of births in Kaniyambadi block between 2003-2012 was 21 726, which included 356 stillbirths and 337 twins. The number of singleton live births between 2003 to 2012 was 21 054. During this period there was a 2.7 % increase in mean (SD) birth weight from 2.799 (0.464) kg in 2003 to 2.904 (0.471) kg in 2012 with an increase in hospital deliveries from 84% in 2003 to 99.4% in 2012, with an overall rate of 95% hospital deliveries.

Of the 21054 singleton live births, gestational age was unknown for 712 and birth weight was unknown for 692. The number of singleton live births of 28 weeks or more, with known gestational ages and birth weight was 19692. However, as the number of births according to gender was small between 28 and 31 weeks, birth weight centiles were only computed for the singleton live births between 32 and 42 completed weeks. The number of singleton live births between 32 and 42 completed weeks with known gestational ages and birth weights was 19545 and after excluding outliers there were 19538 singleton live births. Birthweight centiles were finally computed using gestational ages and birth weights of 15994 singleton live births, after omitting women with complications such as pregnancy induced hypertension (291, 1.5%), gestational diabetes (23, 0.1%), haemo-globin below 10 g% (2039, 10.9%), maternal heart disease (68, 0.3%), short stature (58, 1.1%) and missing values for hemoglobin (859, 4.4%) or height (631, 3.2%). The number of males and females born to nulliparous mothers was 4122 and 3894, respectively; while males and females born to multiparous women were 4189 and 3789, respectively. The average height of the mothers of the selected 15994 children was 155 cm (SD 6.7 cm).

Birth weight centile charts for both sexes were created by the LMS software and the corresponding values of L, M, S and centile values obtained. Birth weight centiles for males and females are depicted in Table I-II, and Fig. 1,2. The normality of Z scores calculated using the L, M and S values for each gestational age was confirmed by Normal Q-Q plots for both genders and gestational age categories within each gender.

TABLE I  Centile Values for Males (n=8311) 

Table II Centile Values for Females (n=7683)

Fig. 1 Birth weight centiles for males - 3rd, 10th, 25th, 50th, 75th, 90th, 97th

Fig. 2 Birth weight centiles for females - 3rd, 10th, 25th, 50th, 75th, 90th, 97th

While the low birth weight proportion was 14%, the proportion of small for gestational age was 8.1%. The median birth weights of the study population were compared to the median birth weights of US whites as well as UK white and Asian newborns (Web Fig. I) [12,13].

Mortality: Mortality data analyzed for the years 2003 to 2012 for children for whom follow up data was available showed that there were 207 deaths within the first two months among 14557 live born singleton births, giving an under-two month mortality rate of 14/1000 live births. Of these 207 deaths, 146 (70%) occurred among the children born to women without risk factors as included in the current study. A comparison of mortality rates between children with lower birth weight centiles and those in higher birth weight centile groups ( Web Table I) showed higher mortality at extremes of birth weight.

Previous Indian growth charts have been from tertiary centres [4,5], which are often referral centres for those with antenatal complications. However, the WHO growth standards provides median birth weights for both sexes, for children born at term to healthy non-smoking mothers [14], which can be taken as possibly representing ideal birthweights.

The sex-specific growth charts produced by the present study are descriptive of singleton live births of rural antenatal women with no major antenatal risk factors, good antenatal care, a high rate of institutional deliveries and a low birth weight rate of 15% similar to the rest of Tamil Nadu (17.2%), but lower than that of the entire country (22%). However, these growth charts are not suitable as standards as the population is not the ideal one required for the creation of standards.

Although data regarding nutritional status by daily caloric intake of this population was not recorded over the years, while a previous survey from Tamil Nadu [15] showed that the median intake of daily calories by pregnant women was below the recommended allowance. The average height of the study women of 155 cm [6] was higher than the national average of 152 cm, [16] but lower than that of developed countries [17]. The shorter stature of Indian women while partly explained by genetic variation, may also be due to under nutrition, as even in developed countries, 20% of the variation in height is thought to be due to environmental factors, with this proportion expected to be higher for less developed countries [18].

The difference between the birth weights in the WHO growth standards and our data illustrates that rural antenatal women even without major antenatal risk factors had children with lower birth weights. While major risk factors for low birth weight have been excluded, low caloric intake remains a possibility which needs to be further studied.

A limitation of the data was that gestational age was based mostly on last menstrual period, as performing ultrasound scans for dating has not been a routine practice in rural areas. However, we have attempted to reduce the error by removing birth weight outliers (abnormally high/low birth weights for a gestation) and restricting to an upper gestation of 42 weeks. Although we excluded mothers with known antenatal risk factors affecting birth weight, it is possible that undiagnosed risk factors for e.g. undiagnosed gestational diabetes could have contributed to mortality among those with high birth weights. Children with anomalies were also not excluded as this information was not available.

The study findings raise the possibility that increase in birthweights to match the WHO standards alone may not reduce infant mortality in rural India as birthweights seem to be following a different norm. While increase in birth weights would be of use in preventing deaths in those in the lowest birth weight categories it may not be necessary or even be harmful beyond a certain limit in its effect on neonatal mortality. Comprehensive measures are needed to address both outcomes for children with higher but so-called normal birth weights, along with attempts to decrease low birth weights.

Contributors: All the authors have written, contributed, designed and approved the study.

Funding: None; Competing interests: None stated.

1. Bonellie S, Chalmers J, Gray R, Greer I, Jarvis S, Williams C. Centile charts for birthweight for gestational age for Scottish singleton births. BMC Pregnancy Childbirth. 2008;8:5.

2. Seaton SE, Yadav KD, Field DJ, Khunti K, Manktelow BN. Birthweight centile charts for South Asian infants born in the UK. Neonatology. 2011;100:398-403.

3. Kierans WJ, Joseph KS, Luo ZC, Platt R, Wilkins R, Kramer MS. Does one size fit all? The case for ethnic-specific standards of fetal growth. BMC Pregnancy Childbirth. 2008; 8:1.　

4. Mathai M, Jacob S, Karthikeyan NG. Birthweight standards for south Indian babies. Indian Pediatr. 1993;33:203-9.

5. Kandraju H, Agrawal S, Geetha K, Sujatha L, Subramanian S, Murki S. Gestational age-specific centile charts for anthropometry at birth for South Indian infants. Indian Pediatr. 2012;49:199-202.

6. George K, Prasad J, Singh D, Minz S, Albert DS, Muliyil J, et al. Perinatal outcomes in a South Asian setting with high rates of low birth weight. BMC Pregnancy Childbirth. 2009 Feb 9;9:5.

7. Exploratory data analysis. J W Tukey. Addison-Wesley;1977.

8. Francis MR, Rakesh PS, Mohan VR, Balraj V, George K. Examining spatial patterns in the distribution of low birth weight babies in Southern India- the role of maternal, socio-economic and environmental factors. Int J Biol Med Res. 2012; 3:1255-59.

9. Cole TJ. Fitting smoothed centile curves to reference data. J Royal Stat Soc. 1988;151:385-418.

10. Cole TJ. The use and construction of anthropometric growth reference standards. Nutr Res Rev. 1993;6:19-50.

11. Cole TJ, Green PJ. Smoothing reference centile curves: the LMS method and penalized likelihood. Stat Med. 1992;11:1305-19.

12. Oken E, Kleinman KP, Rich-Edwards J, Gillman MW. A nearly continuous measure of birth weight for gestational age using a United States national reference. BMC Pediatr. 2003;3:6.

13. WHO child growth standards: length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: methods and development. Geneva: World Health Organization, 2006. Available from: http://www.who.int/childgrowth/standards/Technical_ report.pdf. Accessed 24 May, 2012.

14. WHO child growth standards, weight-for- age. Available from: http://www.who.int/childgrowth/standards/weight_for_age/en/index.html. Accessed 24 May, 2012.

15. Diet and Nutritional Status of Rural Population. National Nutrition Monitoring Bureau. NNMB Technical Report 21. Available from: http://www.nnmbindia.org/NNMBREPORT2001-web.pdf. Accessed 24 May, 2012.

16. Mamidi RS, Kulkarni B, Singh A. Secular trends in height in different states of India in relation to socioeconomic characteristics and dietary intakes. Food Nutr Bull. 2011;32:23-34.

17. Ogden CL, Fryar CD, Carroll MD, Flegal KM. Mean body weight, height, and body mass index, United States 1960-2002. Adv Data. 2004;(347):1-17.

18. Silventoinen K. Determinants of variation in adult body height. J Biosoc Sci. 2003;35:263-85.

19. Wilcox AJ, Russell IT. Birthweight and perinatal mortality: II. On weight-specific mortality. Int J Epidemiol 1983;12:319-25.