Geographical clustering of delayed motherhood in India: an application of multilevel and geospatial analyses
TL;DR
The findings underscore the need for region-specific and demographically targeted policy interventions, along with further research into behavioral and health system factors influencing delayed childbearing, as well as the importance of regional context.
Abstract
Background Rapid sociodemographic changes in India, including rising female education, workforce participation, urbanization, and changing gender norms, have contributed to delayed childbearing. Despite declining fertility, limited evidence exists on the district-level geographic variations and determinants of delayed motherhood (childbirth at age ≥35) in India. This study applies multilevel and geospatial analyses to examine the distribution, determinants, and district-level spatial clustering of delayed motherhood using National Family Health Survey data. Methods This study employed multilevel multivariable logistic regression to examine individual- and district-level determinants of delayed motherhood. Spatial clustering was assessed using Global Moran's I and Local Indicators of Spatial Association (LISA). To account for spatial dependence across districts, a comparative spatial modeling framework was applied using ordinary least squares regression as the baseline model, followed by the spatial lag model and the spatial error model. Results The prevalence of delayed motherhood varied widely, ranging from negligible levels in some districts to as high as 60.18% in South West Khasi Hills, with a national average of 8.7% [95% confidence interval (CI): 8.1–9.3]. Higher education [adjusted odds ratio (AOR): 2.00; 95% CI: 1.80–2.23] and age at marriage ≥21 years (AOR: 5.13; 95% CI: 4.65–5.67) were significant predictors. Additional factors such as higher parity, religious affiliation, and socioeconomic status also showed significant associations. A spatial analysis revealed strong clustering (Moran's I = 0.459), with high–high clusters in southern and northeastern regions and low–low clusters in central and northern India. Spatial models confirmed significant spillover effects across neighboring districts. Conclusion Delayed motherhood in India exhibits substantial sociodemographic and spatial heterogeneity. Key determinants include education, marriage timing, and parity patterns. The presence of distinct geographic clusters highlights the importance of regional context. These findings underscore the need for region-specific and demographically targeted policy interventions, along with further research into behavioral and health system factors influencing delayed childbearing.