Determinants of birthweight in rural Sri Lanka; a cohort study

BACKGROUND: Identifying determinants of birthweight among disadvantaged communities is critical to further reducing the inequitable burden of perinatal health issues in low-and-middle income settings. Therefore, we adopted a bio-psycho-social approach to identify the determinants of birthweight in a mother-infant cohort from a rural setting in Sri Lanka, a lower-middle-income country. METHODS: All third-trimester pregnant women with a singleton pregnancy registered for the national antenatal care programme at Ipalogama health division in 2017 were invited for a prospective cohort study. Data was collected using a self-completed questionnaire and data extraction from health records. The mother-infant cohort was followed up until one month after delivery. A principal component analysis was performed using economic, social, and psychological variables, and two composite variables were achieved. Care from husband and household members, perceived wellbeing, frequency of abuse, and affect during the third trimester strongly loaded to the variable 'psychosocial wellbeing'. Monthly income, husband's education level, and use of biomass fuel strongly loaded to the variable 'socioeconomic status'. Hierarchical logistic regression was used to predict factors associated with birthweight. Maternal age, parity, baby's sex, and gestational period at pregnancy registration were entered at the first step. BMI, psychosocial wellbeing, socioeconomic status, hypertensive disorders, and gestational/chronic diabetes were entered at step two. Preterm birth was entered at step three. RESULTS: 532 women were recruited, and 495 were retained at the postpartum follow-up. 421 (74.8%) had reported being abused at least once during the preceding month. Birthweight was approximately normally distributed (mean 2912 g, SD 456.6 g). Low birthweight was present in 72 (14.6%, 95% CI 11.7,17.9), and 46 (9.3%, 95% CI 7.0,12.1) had birthweights > 3500 g. The regression model explained 13.2% of the variance in birthweight. Preterm birth, maternal BMI, and mid-pregnancy psychosocial wellbeing could explain 6.9%(p < 0.001), 3.9(p < 0.001), and 1.2%(p = 0.02) of unique variance, respectively. CONCLUSIONS: In a setting where a large proportion of pregnant women suffer 'abuse' in their homes, psychosocial wellbeing during pregnancy was an important determinant of birthweight of babies. Expanding routine maternal care services, especially at the primary care level, to cater to the psychosocial issues of pregnant women would help reduce inequities in perinatal health.