Socioeconomic inequalities in child growth and development in Sri Lanka compared to other South Asian countries

Background

Under-5 malnutrition is a major public health problem in South Asia. However, the prevalence of stunting in Sri Lanka (15.9% in 2022) is much lower than that observed in other South Asian countries (e.g., 31.7% in India in 2022). Better socioeconomic position (SEP) is associated with lower risk of childhood growth and development in South Asia. There is a lack of evidence on the association between SEP and early child development (ECD) in South Asia. No studies have tested whether the relationship between SEP and under-5 malnutrition (stunting) in Sri Lanka is weaker than other South Asian countries, or whether the associations of SEP with stunting, and wasting are mediated by child nutrition and household environment in Sri Lanka.

Aim and objectives

Aim: To identify the association between socioeconomic factors and early child development (ECD) and growth in the South Asian region, with a particular focus on Sri Lanka.

Objective 1 (Paper 1): Synthesise existing evidence on the association between household SEP and ECD and growth (under 3-years) in South Asian countries.

Objective 2 (Paper 2): Investigate whether the association of childhood stunting (under 3-years) with household SEP differs in Sri Lanka compared to other South Asian countries.

Objective 3 (Paper 3): Investigate how the association of socioeconomic position (SEP) with childhood stunting and wasting (under 3-years) is mediated by minimum acceptable diet (MAD) and the childcare environment in Sri Lanka.

Methods and results

Paper 1– Systematic review on the SEP-ECD association in South Asia

This paper was a systematic review of the existing evidence on socioeconomic inequalities in ECD (0- 35 months) in South Asia. Four databases (PubMed, Scopus, MEDLINE, Cochrane Library) were systematically searched using terms related to ECD and SEP. Study quality was assessed using the QualSyst tool. A narrative review for the association of motor, cognitive, language and socioemotional development outcomes with SEP measures (mother’s education, wealth, sanitary facilities, drinking water) was constructed. 12 articles were reviewed. The majority (n=10, 83%) reported language development outcomes. Fewer articles (n=3, 25%) assessed socioemotional development than other ECD outcomes. Adverse SEP was associated with poor ECD in most studies. The commonly reported associated SEP measures were mother’s education and family income. The risk of bias assessment rated 7 out of 12 articles as high risk.

Paper 2– Inequalities in stunting in Sri Lanka

This paper investigated whether the association of childhood stunting with household SEP differs in Sri Lanka compared to other South Asian countries. Anthropometric data (for children aged 0-35 months) and SEP measures (mother’s education, wealth, sanitary facilities, drinking water) from DHS 2016-2018 surveys in Bangladesh, India, Maldives, Nepal, Pakistan and Sri Lanka were used (n = 134,765). Binary logistic regression models incorporated SEP, Country, and SEP-by-Country interaction terms were employed. Outcome was stunting. Exposures were mother’s education, wealth, sanitary facilities, and drinking water. Models were adjusted for confounders (age and sex). The association of SEP with odds of stunting was similar in Sri Lanka, compared to other South Asian countries (e.g., India vs Sri Lanka difference AOR of stunting in the mother’s secondary education level was 1.03 (95%CI= 0.85, 1.25)).

Paper 3– Mediation of the SEP-Stunting/Wasting association by MAD/FCI in Sri Lanka

This paper investigated the extent to which the association of composite SEP (household wealth and mother’s education) with childhood stunting and wasting is mediated by family care indicator (FCI) and minimum acceptable diet (MAD) in Sri Lanka. DHS data for Sri Lanka 2016 (n=4236) was used. Counterfactual mediation models (adjusted for sex, age and residence place) were developed to estimate the natural direct and natural indirect effects of SEP on stunting and wasting via MAD or FCI. Lower SEP was associated with higher odds of stunting and wasting. Meeting MAD or FCI was not associated with stunting or wasting, other than a significant association between not meeting FCI and stunting, (OR of FCI=1.47, 95% CI=1.24, 1.74). All estimates of the proportion mediated, and natural indirect effect were not statistically significant at alpha 5%. For example, the proportion mediated by FCI on the association between the lowest composite SEP variable and stunting was 13% (OR=0.13, 95%CI= -0.64, 0.90).

Conclusions

Paper 1 reveals a strong association between SEP and ECD outcomes and the scarcity of evidence exploring the associations between household SEP and ECD in children under 36 months in South Asia. Paper 2 showed that the lower prevalence of stunting in Sri Lanka is unlikely to be due to less inequality. It is more likely that the lower prevalence of stunting in Sri Lanka is related to there being fewer mothers belonging to the lowest maternal education groups. Paper 3 showed that an arguably high proportion of the association between SEP and stunting and wasting in Sri Lanka was mediated by household nutrition or caregiving environment. Therefore, stronger evidence for the association between SEP and ECD, introducing gender balance in access to education in the region, and national nurturing care programs targeting socially and economically disadvantaged households with young children are needed to reduce poor child growth and ECD delays associated with a disadvantaged SEP in the South Asian region.