Spatiotemporal trends and multilevel determinants of coexisting maternal and child undernutrition in Nepal: evidence from Nepal Demographic and Health Survey 2006–2022

Abstract

Abstract Background Coexisting maternal and child undernutrition (CMCU) is a critical aspect of intergenerational malnutrition. The evidence on its epidemiological trends and spatial distribution remains limited in Nepal. This study investigated the temporal trends, determinants, and spatial distribution of CMCU in Nepal from 2006 to 2022. Method We used data from the 2006, 2011, 2016, and 2022 Nepal Demographic and Health Survey (NDHS) databases. Maternal and child pairs with complete anthropometric records were included in the analysis. The co-occurrence of maternal undernutrition (height < 145 cm) and child stunting/wasting/underweight were defined as CMCU. Survey-weighted descriptive analysis was performed to assess prevalence trends. Individual- and community-level determinants of CMCU were assessed using multilevel logistic regression with a two-level hierarchy. Individual children and their households (level one) were grouped within community enumeration areas or clusters (level two). Global Moran’s I and Getis-Ord Gi* Hotspot analysis were performed to assess spatial autocorrelation and local clustering, respectively. Results The prevalence of CMCU decreased from 9.78% in 2006 to 4.68% in 2022, a 52.1% decrease over 16 years). A PCV of 0.802 in Model IV of the community level variation has been explained after adding both the individual and community factors. Even low ICC = 0.029 in Model IV shows that only 2.9% of the remaining, unexplained variances in Child and Maternal Co-undernutrition (CMCU) is caused by unmeasured community clustering factors as compared to ICC = 0.134 in model I which shows that 13.4% of the variance is explained by unobserved community level factors. This means Model IV shows that 97.1% variance is explained by individual and household level factors. CMCU odds were inversely associated with maternal education (AOR = 0.68(95% CI: 0.50–0.91), 0.39(95% CI: 0.26–0.58), 0.20(95% CI: 0.07–0.58) for primary, secondary, higher respectively relative to no education). Children aged 12–59 months had higher CMCU odds than < 12 months. Odds were lower in contraceptive users (AOR = 0.75(95% CI: 0.60–0.93)) and higher for perceived health facility distance (AOR = 1.24(95% CI: 1.00–1.52)). The spatial average was significantly clustered in 2022 (Moran’s I = 0.058, Z = 2.548) with hotspots in the southern Terai. Conclusion Despite an overall decline in CMCU during 2006–2022, socioeconomic and regional inequalities were observed in Nepal. CMCU was associated with maternal education, child age, reproductive health services utilization and access of health services. The new trend of spatial clustering in the southern provinces makes it important to develop geographically-targeted and equity-sensitive approaches to aid the implementation of multisectoral nutrition strategies in Nepal.