Indmedica Home | About Indmedica | Medical Jobs | Advertise On Indmedica
Search Indmedica Web
Indmedica - India's premier medical portal

Indian Journal of Community Medicine

A Study of Surrogate Parameters of Birth Weight

Author(s): Y. R. Kadam, P. Somaiya, S.V. Kakade

Vol. 30, No. 3 (2005-07 - 2005-09)


Research question: Which anthropometric parameter is correlating highly with birth weight? Can we use this parameter as a screening test for predicting birth weight? What is their cut-off value?

Hypothesis: Various anthropometric parameters of newborn correlate each other positively.

Objective: To find out the most effective anthropometric parameter in the newborn to assess birth weight so that newborn with LBW can be identified.

Study design: Hospital based cross-sectional study.

Participants: Newborn babies born in KIMS, Karad.

Results: Relatively highest correlation was observed between birth weight and thigh circumfirence (T.C.) (r = 0.8637) and next with chest ciramfirence (C.C.). (r = 0.8247) Cut-off values of T.C. and C.C. had better sensitivity, specificity and predictive value for identifying LBW babies.

Conclusion: T.C. is the best effective parameter to predict birth weight. Next to it is C.C.

Key words: Birth Weight, Low Birth Weight, Anthropometric Parameters


One in three babies in India are born with moderate to severe malnutrition with birth weight of 2500 gm or below. Babies having low birth weight are more susceptible to infection and they do not grow to their full potential of physical and mental abilities and start their life at disadvantage leading to high infant morbidity and mortality. Hence it is imperative to identify the newborns with low birth weight and to offer them adequate and needed care instantly for their survival. However in developing countries like India 70-90% deliveries occur at home where weighing facility does not exist or due to certain logistic problems1 and as a result babies with low birth weight could not be identified and these babies are deprived of much needed adquate and immediate care.

The logistic problems such as non-availability of weighing machine is also big hurdle in recording the birth weight, even though the deliveries are conducted by properly trained health personal. As a result in many cases LBW babies remain undetected, which also reflects adversely upon the 'high risk' aproach to ensure better child survival.

Material and Methods

This study was carried out in the Krishna Hospital and Medical Research Center (KH&MRC) Karad, for a period of six month. This hospital is attached to Krishna Medical College.

All deliveries that took place during this period were included in the study. Total 293 live births took place. Out of which 14 new borns were not included, since the worker could not record their birth weight and other anthropometric measurements during first 48 hours after their birth as per WHO recommendation due to certain unavoidable reasons2.Thus 279 newborns were included in the study. Equipments used during the study were of flexible, non-stretchable measuring tapes a spring dial weighing machine, a vertical measuring rod and a pediatric weighing machine capacity of measuring upto 0.1 cm, 50 gm 0.1 cm and 50 gms respectively which include general demographic information obstetric history and anthropometric of newborn.

Anthropometric measurements in respect to newborns were:

  1. Mid arm circumference (MAC)
  2. Maximum thigh circumference (TC)
  3. Calf circumference (CFC)
  4. Chest circumference (CC)
  5. Head circumference (HC) and
  6. Birth weigh (BW)

The methodology employed in respect of these anthropometric measurements was as per a standardized recommended procedures3-6.

Data was fed and analysed by computer with the help of statistical package 'STATPAL'.


It was observed in the study that the difference in the mean birth weight between male and female babies is not significant (t-1.94, p>0.05). Further, no significant difference was observed between the mean birth weight of male and female babies with weight less than 2500 gms (t=1.25, p>0.1) and less than 2000gms (t=0.66, p>0.1).

Table I shows than there is no significant difference in anthropometric parameters of all male and female newborns, male and female newborns with weight <2,000 gms and male and female newborns weight <2,500 gms.

It was observed that birth weight is significantly correlated with all studied anthropometric parameters (P<0.001). However the thigh circumference has a relatively higher correlation value than the other parameters. After TC, the next anthropometric parameter with a relatively higher correlation with birth weight was found to be the CC. (Table II)

Table III shows simple regression equation for estimating birth weight (y) for babies with birth weight less than 2500gms, in relation to various studied anthropometric variables with R2,

Table I : Anthropometric Measurement in Relation to Birth Table II : Correlation between birth weight and anthropometric Weight and Sex.

Anthropometric parameter All newborn
Newborn with
Newborn with
n =42
Mean (SD) Mean (SD) Mean (SD)
1 MAC Combined 09.54 (1.22) 8.61 (0.96) 7.75 (0.79)
Male 09.63 (1.20) 8.7 (0.97) 7.82 (0.88)
Female 09.44 (1.23) 8.53 (0.96) 7.70 (0.74)
't' value (male vs. famale)* 1.35   0.92   0.48  
2 TC Combined 15.07 (2.21) 13.27 (1.77) 11.69 (1.48)
Male 15.01 (2.13) 13.16 (1.69) 11.45 (1.72)
Female 15.03 (2.29) 13.37 (1.85) 11.84 (1.21)
't' value (male vs female)* 1.13   0.66   0.84  
3 CC Combined 29.89 (2.65) 27.86 (2.41) 25.66 (2.15)
Male 30.05 (2.41) 27.96 (1.1) 25.78 (2.12)
Female 29.72 (2.1) 27.78 (2.75) 25.58 (2.2)
't' value (male vs. female)* 1.04   0.38   0.3  
4 CFC Combined 9.86 (1.15) 8.97 (0.91) 8.17 (0.74)
Male 9.95 (1.09) 9.03 (0.9) 8.21 (0.74)
Female 9.77 (1.2) 8.90 (0.33) 8.15 (0.75)
't' value (male vs. female)* 0.74   0.74   0.75  
5 HC Combined 32.82 (1.91) 31.42 (1.78) 29.84 (1.75)
Male 33.14 (1.76) 31.73 (1.65) 30.32 (1.85)
Female 32.48 (2.01) 31.14 (1.86) 29.52 (1.63)
't' value (male vs. female)* 0.71   0.38   0.3  
P > 0.05

Table II: Correlation between birth weight & anthropometric

Anthropometric Parameter MAC Thigh Chest Calf Head
Birth weight 0.8140 0.8637 0.8427 0.8188 0.8274
95% C.I. 0.746- 0.804- 0.779- 0.751- 0.761-
  0.882 0.823 0.906 0.886 0.894

Table III: Cut off values for identifying LBW babies.

R2 Cut of values
Wt< 2500gm Wt< 2000gm
MAC Y=366.95 (MAC)-1048.27 66.3 9.67 8.31
TC Y=214.97 (TC)-787.57 74.6 15.29 12.97
CC Y=174.46 (CC)-2762.19 71 30.16 27.30
CFC Y=392.52 (CFC)-1418.13 67 9.98 8.70
HC Y=237.94 (HC)-5357.43 68.5 33.02 30.92

Table IV: Validity of Cult off Values for Birth Weight Less than 2500 gms.

Anthe.Para Sensitivity Specificity Predictive+ve Predictive -ve False+ve False-ve
MAC 88.23 73.12 75.94 89.31 26.88 11.77
TC 94.95 85.62 83.08 95.8 14.03 5.04
CC 90.76 84.37 81.2 92.46 15.6 9.24
CALF 84.87 88 84.17 88.68 11.87 15.12
HC 87.39 66.87 66.24 87.7 33.13 12.61

Table V: Comparability of Studies

Correlation Co-efficient between Birth Weight and Self J.N. Sharma et. al4 Sidharth Ramaji et. al.5 G.M. Matto et. al.10 B.D. Bhatia and Tyagi7 Bhargava et. al.1 WHO Collabo-rative Study in India (A)3 WHO Collabo-rative Study in India (B)3 M.L.Kulkarni M. Rehman8 R.K. Sachar R.K. Soni et. al.6 R.K.Sachar R.K. Soni et. al.9
MAC 0.814 - 0.8282 0.21 N.S. 0796 0.811 0.83 0.77 0.9329 - 0.6
TC 0.864 0.9201 0.9188 0.5 - - - - - - -
CC 0.843 - - - 0.679 0.8696 0.87 0.94 0.7904 - -
CFC 0.819 - - - - - - - - 0.74 0.76
HC 0.827 0.7257 - - 0.766 0.7264 - - 0.8071 - -

the predictive value, R2 of birth weight is higher in relation to TC followed by CC.

Table IV shows validity parameters of the cut off values of all studied anthropometric parameters for weight <2,500gms.The highest sensitivity and predictive -ve value is observed for TC followed by CC. Simpler findings was observed for various parameter when wt < 2,000 gm.

Table V shows correlation coefficient between birth weight and other anthropometric parameters obtained by various researchers. Other studies have also mentioned TC as the best surrogate parameter in estimating birth weight4,5,10.


The prime concern of the present study was to identify the best suitable surrogate parameter, proxy, to birth weight, which when used by the health personnel in domiciliary outreach will detect the maximum number of at risk infants for providing them with timely and needed intervention strategy.

Scrutiny of collected data reveals that none of the studied parameters have shown the highest value of all validity parameters; thought positive correlation between all studied parameters and birth weight was observed. Hence the suggestion of best surrogate parameter has been done by choosing the combination of the validity parameters.

At weight <2,500 gm, <2,000 gm, TC is suggested as the best surrogate parameter to predict baby as LBW. It shows the highest sensitivity 94.95% and negative predictive value 95.8%.The next best surrogate parameter is CC.

In this situation increase in the false positive diagnosis is inevitable. But considering the problem of LBW a) the neonate with false negative result generally cause no concern either for parents or for health personal. But in reality LBW baby is at greater risk of high morbidity and high mortality. b) Whereas by false positive diagnosis, healthy neonates will be subjected to further screening at nearby referral center causing some inconvenience, discomfort and expenses to the family. Thus increase in the number of false positive will enhance the burden at the secondary screening center, which is mainly in terms of efforts of weighing the referred neonate. Hence it is sincerely felt that the penalty of false negative diagnosis is much more painful than the increased burden of weighing the neonates and declaring them as a baby having a normal weight. This opportunity may be exploited at the referral center for educating mother or father on immunization regarding feeding of baby, nutrition, family planning, importance of personal hygiene etc.

Thus, it is evident from anlysis of our data that TC is the best suitable and simple surrogate parameter that could be used in the domiciliary outreach when it is impossible to record weight of baby at birth. For health personnel who are working in the community can use color coded tapes indicting weight <2,000 gms, between 2,000 gms and 2,500 gms and above 2,500 gms. With this technique, measurement errors are likely to be less as measurements of thigh is taken at it.s maximum circumference, since at that section, larger amount of soft tissue is present.

In conclusion, TC is the best suitable parameter for predicting LBW (<2,500 gms and <2,000 gms) neonate while CC is the next best parameter.The use of such indicator would help to identify neonates requiring intervention to prevent deaths, diseases and disability. This will be a viable example of practical and cost effective health care strategy.


On evaluation of the validity of these cut off values if was observed that TC of ≤15.29 cm. and ≤12.97 cm. and CC of ≤30.16 cm and ≤27.3 cm. had the better sensitivity, specificity and predictive value for identifying infants weighing ≤2,500 gms. and ≤2,000gms. respectively.


TC and CC can be used as a surrogate parameter of birth weight.

Tapes with different colored risk zones could be devised and tested for reliability so that they can be used in community.


  1. Bhargava S.K. Ramji S, Kumar A, Manmohan, Marwah J Sachdev HPS. Mid Arm and Chest circumference at birth as predictors of Low Birth Weight and Neonatal Mortality in the community. British Medical Journal 1985; 291: 1617-19.
  2. Bhargava et. al, Survival, Growth and Development in babies weighing 2000 gms or less. Indian Paediatrics; 7: 132.
  3. WHO, Use of a simple Anthropometric Measurements to predict birth weight. Report prepared by Dr. I. Diamond et. al. as a part of the WHO collaborative study of Birth Weight Surrogates. Bulletin of WHO 1993; 71 (2): pg 157-163.
  4. Sharma J. N., Saxena S., Sharma U., Thigh Circumference at birth as the best predictor of low birth weight babies. Indian Paediatrics 1989; 26: 18-21.
  5. Ramji S. Marwah J., Satyanarayan L., Kapani V., Manmohan, Bhargava S.K. Alternative indicator of Low Birth Weight. Indian Journal Paediatrics 1986: 653-654.
  6. Sachar R.K. et. al. Calf Circumference as a Birth Weight surrogate. Indian Journal Maternal and Child Health. 1994; 5(2): 39-40.
  7. Bhatia B.D., Tyagi N.K. Birth Weight Relationship with other foetal anthropometric parameters. Indian Paediatrics 1984: Vol. 21, 833-838.
  8. Kulkarni M.L. et al. Neonatal Assessment Beyond Birth Weight. Indian Paediatrics 1991: 28; 929-931.
  9. Sachar R.K. et al. Consistent accuracy of Mid Arm Circumference and Calf Circumference as birth weight surrogate during first few days. Indian Journal Maternal and Child Health, 1994; 5(2): 41-44.
  10. Matto G.M. et al. Maximum Thigh Circumference as an Indicator of Birth Weight. Indian Journal Maternal and Child Health. 1991; 2(2): 40-42.
Access free medical resources from Wiley-Blackwell now!

About Indmedica - Conditions of Usage - Advertise On Indmedica - Contact Us

Copyright © 2005 Indmedica