Utility of anthropometric measurements to predict low birth weight newborns

Introduction: Birth weight is an important indicator of survival, future growth and overall development of the child. Timely care of a low birth weight newborn (LBW) is important but it's difficult in developing countries since many are home deliveries with inadequate facilities to weigh the newborn. Hence this study was conducted to find out the relationship between birth weight and different anthropometric measurements from which LBW babies could be identified reliably and managed accordingly. Methods: The study samples of 500 live newborns were selected by random sampling technique born at Mysore Medical College and Research Institute (MMCRI), Mysore between 1st December 2006 to 30th November 2007. The following anthropometric measurements were recorded within 24 hours of birth: Weight, Head circumference (HC), Chest circumference (CC), Mid arm circumference (MAC), Calf circumference (CaC), Crown heel length (CHL), Foot length (FL) and Thigh circumference (TC). Comparison between these measurements was done to find out the most suitable birth weight substitute in identification of LBW babies. Results: For determining LBW <2.5 kg, the critical limits for HC, CC, MAC,TC, CaC, FL, CHL were 33.82cm, 31.5cm, 9.99cm, 15.47cm, 10.25cm, 7.67cm and 48.45cm respectively. For determining a birth weight ≤2kg the critical limits for HC, CC, MAC,TC, CaC, FL, CHL were 31.93cm, 29.75cm, 9.03cm, 13.62cm, 9.5cm, 6.93cm and 44.8cm respectively. MAC of ≤9.99cm and ≤9.03cm for detection of birth weight <2.5kg and <2kg respectively have higher measures of validity. Conclusions: MAC is a simple, quick and reliable indicator for predicting LBW. Hence a simple tricolored tape for early detection of at risk newborn can be introduced in community for their timely management. ………………………………………………………………………………………………………………………………..


Introduction
Birth weight is the single most important indicator of survival, future growth and overall development of the child. LBW is associated with high neonatal morbidity and mortality due to susceptibility to adverse environmental influences, predilection to infections and under nutrition [1]. LBW is also associated with post neonatal mortality, infant and childhood morbidity. It also accounts for about 70% of perinatal and 50% of infant deaths in India [2,3]. In developing countries like India, majority of births are still conducted at home by the traditional birth attendants (TBA) where the estimation of birth weight is not done because of lack of weighing machines. Practically it is not possible to provide expensive weighing scales to the community members. Therefore it is essential to find out an alternative method for the estimation of birth weight.
Such anthropometric measurements must identify LBW babies reliably and be easily measurable using a simple and robust measuring instrument. Many of the birth weight substitutes that have been proposed are anthropometric measurements such as arm circumference, foot length, chest circumference, thigh circumference and calf circumference. The data was collected in a predesigned and pretested proforma after obtaining informed and written consent from the parents of the babies. Relevant antenatal histories with following anthropometric measurements were recorded in warm environment using standard technique and instruments. A. Birth weight: The naked birth weight of the babies was measured in the spring scale to the nearest 25g. B. Head circumference(HC) : It was measured using flexible non -stretchable measuring tape placed over the supra-orbital ridges in front and the maximum occipito-frontal circumference was recorded to the nearest 0.1cm.
[20] C. Chest circumference (CC) : It was measured to the nearest 0.1cm at the level of nipple/4th costosternal joint [6,21]. D. Mid arm circumference(MAC) : It was measured at the midpoint between the tip of acromion and the olecranon process in the left upper arm with the measuring tape to the nearest 0.1cm [1,6,22,23,24]. E. Thigh circumference (TC): In supine position the thigh circumference was measured to the nearest 0.1cm at the level of the lowest gluteal furrow of left thigh, the tape being placed perpendicular to the long axis of the left lower limb [7]. F. Calf circumference (CaC): It was measured at the most prominent point in semi-flexed position of the left leg with the measuring tape to the nearest 0.1cm [8]. G. Foot length (FL) : A 15cm long sliding gauge with divisions up to .05cm was prepared. Heel was stabilized against the fixed vertical end of the gauge and sliding end was adjusted against the tip of big toe after straightening the foot and foot length measurement was recorded to the nearest 0.1cm. [25,26]

Results
Among the total 500 newborns, 492 were singleton births and 8 twins. The birth weight and sex distribution is depicted in Table 1 below.    S -Significant; p < 0.001 From the table 3, it can be observed that correlation coefficients of all the parameters are highly positive and are statistically significant (p < 0.001). Maximum positive correlation was observed in case of thigh circumference (r = 0.857). The arrangement of anthropometric measurements in the descending order of "r" value is given below.  Inference-Mid-arm circumference with cut-off value of  9.99 cm has high measures of validity in the detection of birth weight of < 2.5 kg in comparison to other anthropometric measurements.

Anthropometric measurement 'r' value
From the table 5 it can be observed that mid-arm circumference with a cut-off value of  9.03 cm has maximum sensitivity (100%) and specificity (71.22%).
Thigh circumference, calf circumference also has high sensitivity of 98.55% and 100% respectively. But lower specificity and positive predictive value as compared to that of mid-arm circumference.
Inference-Mid-arm circumference with a cut-off value of  9.03 cm has high measures of validity in the detection of birth weight < 2 kg.

Discussion
The early identification of low birth weight babies is an important pre-requisite of any initiative to reduce mortality. In many developing countries including India, widespread accurate measurement of birth weight is not practicable; easily measurable substitutes for birth weight are therefore needed.  [15], Sreeramreddy, [19] and Alves JGB et al [29].

Comparison of correlation coefficients of various anthropometric measurements in relation to birth weight of the present study with other studies.
Among all anthropometric measurements higher correlations was found with thigh circumference, calf circumference and mid-arm circumference with r = 0.    The cut-off value of MAC of 9.99 in the detection of birth weight<2.5 kg is comparable to study by Sauerborn et al [24] 7. Comparison of cut-off values of thigh circumference in the detection of birth weights < 2.5 kg and < 2 kg  The cut-off value of  15.47 cm of the present study is marginally higher than the values obtained in the study by Ramji S et al [7] and Huque F et al (1). However the sensitivity is higher than that of the other studies.

Comparison of cut-off values of mid-arm circumference in the detection of birth weights of < 2.5 kg and < 2 kg
The cut-off value of  13.62 cm in the detection of birth weight < 2 kg is comparable to study by Ramji S et al (7). Sensitivity is comparable to both studies done by Ramji S et al [7] and Huque F et al [1]. The cut-off value of  10.25 and its sensitivity in the detection of birth weight < 2.5 kg is comparable to study by Neela J et al [15] Data for comparison of cut-off value of calf circumference for detection of birth weight < 2 kg is not available.

Cut-off values of foot length in the detection of birth weights < 2.5 kg and < 2 kg
The cut-off value of foot length  7.67cm in detection of <2.5 kg has sensitivity and specificity of 75.94% and 93.7% respectively.
The cut-off value of foot length  6.93cm has sensitivity and specificity of 36.23% and 95.12 % respectively. Data for comparison is not available.

Cut-off values of head circumference in the detection of birth weights < 2.5 kg and < 2 kg
The cut-off value of head circumference of  33.82 cm is slightly lower than the value ( 33.9 cm) obtained in the study by Samal GC et al [14].
The measure of validity that is sensitivity is same in both studies but specificity is lower (48.6%) that of study done by Samal GC et al [14] (81.2%).
The cut-off value of head circumference of  31.93 in detection of birth weight < 2 kg has higher specificity of 76.79% with lower sensitivity 36.23%. Data for comparison is not available.

Conclusion
It was observed that all the anthropometric measurements studied had positive correlation with birth weight with statistical significance (p < 0.001).
Thigh circumference (r = 0.857) and calf circumference (r = 0.809) have higher correlation coefficient values and higher sensitivity than that of mid-arm circumference (r = 0.799), but specificity and positive predictive values are lower. Hence overall mid-arm circumference has higher measures of validity. Mid-arm circumference of  9.99 cm and  9.03 cm have higher measures of validity in the detection of birth weight < 2.5 kg and < 2 kg respectively than any other anthropometric measurement.
Measurement of mid-arm circumference can substitute for actual recording of birth weight. Mid arm circumference is a simple, quick and reliable indicator for predicting low birth weight. It can be easily measured by medical practitioners and traditional birth attendants in the community.
For this purpose a special measuring tape may be introduced for identifying infants of low birth weight. The device should be flat, flexible, non-stretchable and suitably coloured in red, yellow and green, so that these can be used and understood easily by the illiterate Traditional Birth Attendants (TBA).
1. If the MAC reading falls in the red zone the weight of the baby will be  2 kg.
2. If the MAC reading falls in the yellow zone the weight of the baby will be 2.1-2.49 kg.
3. If the MAC reading falls in the green zone the weight of the baby will be > 2.5 kg.
Funding: Nil, Conflict of interest: Nil Permission from IRB: Yes