1.gif (1892 bytes)

Brief Reports

Indian Pediatrics 2000;37: 872-876

Socioeconomic and Demographic Factors Associated with Birth Weight: A Community Based Study in Kerala

 

T. Radhakrishnan
K.R. Thankappan
R.S. Vasan
P.S. Sarma

From the Achutha Menon Center for Health Science Studies, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram.

Reprint requests: Dr. K.R. Thankappan, Associate Professor, Achutha Menon Center for Health Science Studies, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram 695 011, India.

E-mail: [email protected]
Manuscript received: October 5, 1999;
Initial review completed: November 26, 1999;
Revision accepted: February 15, 2000


The World Health Organization defines low birth weight (LBW) as less than 2500 grams at birth, irrespective of the region, the community or culture. LBW has been associated with a high infant mortality, morbidity in childhood and with an elevated risk of diabetes mellitus, hypertension and other cardiovascular disease in adulthood(1).

Kerala State is noted for its high health achievements at low cost; most health indices are comparable to those of developed countries(2). However, Kerala has been reported to have a high proportion of LBW compared to developed countries; the National Family Health Survey of India (1992) reported that 18.2% of newborns in Kerala have a LBW(3). The present investigation was designed to determine the proportion of LBW infants in a community based sample in Kerala and to evaluate potential socioeconomic and demographic correlates of the condition.

 Subjects and Methods

Our sampling frame consisted of non-coastal village Panchayaths of Thiruvanan-thapuram district in Kerala state. We chose this district because of its well functional Integrated Child Development Services (ICDS) Program, a feature that ensures adequacy of birth records of infants born in the district and facilitates tracking of the mothers(4). Kadakampally village Panchayath was selected for the present investigation.

We performed a cross sectional survey of all infants born in Kadakampally Panchayath during the reference period from 1 January, 1997 to 31 December, 1997. Based on its population size of 24540(5), and the state-wide birth rate of 17/1000, we estimated that a sample size of about 400 live births would yield reliable estimates of the frequency of LBW and would be adequate to assess the socioeconomic and demographic factors associated with LBW.

We contacted the ICDS centers in Kadakampally Panchayath and collected the birth records of all infants born in 1997 and the addresses of their mothers. There were 294 mothers who delivered 301 infants during the reference period. After exclusion of 21 infants (14 twins, 7 others without birth records), 280 infants and their mothers constituted our study sample.

Mothers of these 280 infants were contacted by one of the investigators (TR) and informa-tion collected on socioeconomic factors, demographic characteristics, and medical history using a pre tested questionnaire. We collected information regarding the education of the mother and the father of the infant, family income, housing conditions like roof of the house, wall of the house, floor of the house and floor area of the house. Socio-economic status (SES) of the households was assessed as a composite index incorporating several of these variables. Briefly, a weighted score (range 7 to 21, median 13), based on the variables listed above was used to categorize our sample into two groups Low SES below the median and high SES above the median score. Previous studies in Kerala have validated such an approach for determining socioeconomic status of households in Kerala(6). The questionnaire also sought the following additional demo-graphic and medical information: maternal age at delivery, occupation of mother, religion, place of delivery, antenatal care including history of sonographic examination during pregnancy, birth order, gestational age and sex of the baby.

Bivariate and multivariate analyses were done on 265 term babies only while other descriptive data analysis was done on all babies (n = 280). Chi square tests and chi square trend test was used to assess bivariate associations of select predictor variables with low birth weight. We constructed multiple logistic regression models to assess the association of SES with the occurrence of a LBW infant adjusting for other predictor variables. A p value <0.05 was considered as statistically significant.

 Results

All but two women had institutional deliveries. Most deliveries took place in government hospitals. Seventy five per cent of the mothers had more than ten antenatal visits to the doctors, 21% had between 5 to 10 visits, while the remainder had between 3 to 5 visits. Seventy four per cent of mothers underwent prenatal sonographic examination. About 4% of the women were less than 20 years of age, and another 4% were over the age of 35 years.

Fifty infants had a LBW (17.9%; 95% CI = 13.4-22.4%). Among term babies (gestational age >37 weeks, n = 265), 16.2% (95% CI = 11.7-20.7) had a LBW. Among 15 pre-term babies 54% had a LBW.

In the bivariate analysis SES of family was found to be significantly associated with LBW (Table I). In multivariate logistic regression model also maternal SES emerged as the principal determinant of LBW. A low SES was associated with a 3.5 fold (95% CI = 1.6-7.6) elevated risk of LBW in the baby compared to a high SES of the mother. There was no association found between LBW and other variables included in the multivariable analyses (Table II ).

Table I - Bivariate Associations of Birth-Weight and Selected Characteristics

Variable
Low birth weight
(n = 43)
Normal birth weight
(n = 222)
Chi square
p value
Sex of the infant
 
	Males
18 (41.8)
114 (51.4)
 
	Females
25 (58.1)
108 (48.6)
0.331y
Socio-economic status@
 
	Low
30 (69.8)
101 (45.5)
 
	High
13 (30.2)
121 (54.5)
0.006y
Number of antenatal check-ups
 
	<10
8 (18.6)
56 (25.2)
 
	³10
35 (81.4)
166 (74.8)
0.463y
Birth order of the infant
 
	 1
22 (51.2)
115 (51.8)
 
	 2
18 (41.8)
91 (41.0)
 
	³3
3 (7.0)
16 (7.2)
0.969*
Maternal age at delivery (yr)
 
	<20
3 (7.0)
6 (2.7)
 
	³20
40 (93.0)
216 (97.3)
0.165#
Maternal height@
 
	<152 cm
16 (37.2)
109 (49.1)
 
	³152 cm
27 (62.8)
113 (50.9)
0.207y
Religion
 
	Hindus
39 (90.7)
206 (92.8)
 
	Others
4 (9.3)
16 (7.2)
0.544#
Mother’s occupation
 
	Housewife
43 (100.0)
208 (93.7)
 
	Others
0 (0.0)
14 (6.3)
0.136#

'Y' Yates corrected; *Trend test;.: # Fishers exact test;@ Grouping at the median cut off point. Figures in parentheses indicate percentages.

Table II Multiple Logistic Regression of Low Birth Weight (Dependent Variable) on Select Predictor Variables

Variable
b
SE (b)
Odds ratio
(95% CI)
SES
1.2504
0.3904
3.5 (1.6-7.6)

Variables considered but not significant: age at delivery, antenatal check-up, birth order, height, sex, religion. SES = socio-economic status, b = regression coefficient, SE = standard error, CI = confidence interval.

 Discussion

Our estimate of 17.9% LBW babies in a randomly selected community-based sample in Thiruvananthapuram district in Kerala is consistent with prior reports from Kerala. Our investigation suggests that a low maternal socioeconomic status was the principal determi-nant of a LBW. We did not find an association of LBW with level of maternal education, perhaps because most women were literate. There was also no association between birth order and LBW. We did not detect any association of gender of the infant and LBW, contrary to some other reports(7). There was no association between frequency of antenatal visits and LBW probably because all mothers had a least 3 visits.

The association of a low SES with LBW has been reported previously(8,9). Such an association may be related to several potential mechanisms. A poor maternal nutritional intake during pregnancy (more likely in the low SES groups and related to certain cultural practices) is an important mechanism(10). Unfortunately, we did not obtain information regarding maternal nutrition during pregnancy in the present study to investigate this mechanism further.

A striking observation in our study was the high prevalence of LBW babies in the face of a very high frequency of antenatal examinations and prenatal ultrasound examinations. These findings suggest that perhaps maternal weight gain during pregnancy and fetal growth are inadequately monitored. It is also possible that practitioners ‘accept’ prevalent birth weight of infants and current maternal weight gain in pregnancy as the ‘norm’ for a ‘developing’ country and do not intervene. It is also possible that practitioners consider it futile to provide advice regarding additional nutritional supple-mentation to women with inadequate weight gain in pregnancy, given their awareness that women with a low SES are likely to lack the economic means to comply with such advice. Our results also call into question the adequacy of performance of the Anganwadi workers in ensuring maternal nutritional supplementation during pregnancy.

Since Kerala is a poor state with a per capita income of US $ 275(2), economic development would be the key factor for reducing the prevalence of LBW babies in the state. Improving the quality of ICDS program, antenatal care, and providing supplementary feeding for the pregnant women of low SES may be a short-term option.

We did not check the accuracy of the birth weight record in different institutions which is a limitation of this study. We believe that this factor may not significantly affect our study findings because we used only logistic models in which the dependent variable is dichotomous (presence of low versus normal birth weight).

Contributors: TR proposed the idea and conducted the actual cross-sectional survey. KRT and RSV assisted in designing the research protocol and preparing the question-naire; they supervised the data collection and the data analysis. PSS assisted in formulating the sampling strategy, preparing the compu-terized data entry form, and did the statistical data anlaysis. All investigators contributed to the writing of the paper.

Funding: None.
Competing interests:
None stated.

 

Key Messages

• A low maternal socioeconomic status was noted to be the principal determinant of a low birth weight baby.

• A high prevalence of low birth weight babies was observed despite a very high frequency of ante-natal examinations and prenatal ultrasound examinations.

• No association was observed between low birth weight and the following variables: age at delivery, birth order, frequency of ante-natal check ups, height of the mother, sex of the baby and religion of parents.

 

 References
  1. Barker DJP. Fetal origins of coronary heart disease. BMJ 1995; 311: 171-174.

  2. Thankappan KR, Valiathan MS. Health at low cost - the Kerala model. Lancet 1998; 351: 1274-1275.

  3. National Family Health Survey 1992-93, Kerala. Population Research Centre, University of Kerala and International Institute of Population Sciences, Bombay, 1995.

  4. Sadka NL. Integrated Child Development Services in India. New Delhi, United Nations Childrens Fund, 1984.

  5. Census of India. Series 12, Kerala. Final Popula-tion Totals. Directorate of Census Operations, Kerala 1991; p 184.

  6. Raman Kutty V, Thankappan KR, Kannan KP, Aravindan KP. How socio-economic status affects birth and death rates in rural Kerala, India: Results of a health study. Int J Health Serv 1993; 23: 373-386.

  7. Kramer MS Determinants of low birth weight: Methodological assessment and meta-analysis. Bull World Health Organ 1987; 65: 663-737.

  8. Deshmukh JS, Motghare DD, Zodpey SP, Wadhva SK. Low birth weight and associated maternal factors in an urban area. Indian Pediatr 1998; 35: 33-36.

  9. Hirve SS, Ganatra BR. Determinants of low birth weight: A community based prospective cohort study. Indian Pediatr 1994; 31: 1221-1225.

  10. Panikar PGK, Soman CR. Health Status of Kerala. Paradox of Economic Backwardness and Health Development. Center for Development Studies,Thiruvanthapuram, 1984; p 26.