Consistent determinants of delayed initiation of breastfeeding among nursing mothers aged 15-49 years in Uganda

Abstract

Delayed initiation of breastfeeding (DIBF), defined as starting breastfeeding one hour or more after birth, remains a major contributor to neonatal morbidity and mortality in Uganda. Although early initiation increased from 32% in 2000–2001 to 67% in 2016, delays persist among certain maternal groups. Using pooled data from 17,357 mothers across four Uganda Demographic and Health Surveys (2000–2016), this study examined determinants of DIBF among mothers aged 15–49 years. Guided by the Social Cognitive Theory (SCT) and Health Belief Model (HBM), a Generalized Structural Equation Model (GSEM) estimated direct and indirect effects of socio-demographic, economic, and health facility related factors, enabling mediation analysis. Results showed no consistent factors directly linked to DIBF across all surveys. Rural residence (β = 0.38, p < 0.05) increased DIBF risk, while delivery from health facilities (β = –0.48, p < 0.001), maternal age 20–34 (β = –0.39, p < 0.05) or 35–49 years (β = –0.88, p < 0.01), and pregnancy desirability (β = –0.30, p < 0.05) reduced it. Mediation analysis revealed rural mothers were less likely to deliver from health facilities (β = –0.69, p < 0.001), indirectly increasing DIBF. Mothers aged 20-34 and 35-49 years and married mothers were more likely to deliver from health facilities, reducing delays. These findings highlight that socio-demographic factors influence breastfeeding both directly and indirectly through healthcare utilization. The study concludes that DIBF in Uganda is shaped by intertwined behavioral and structural determinants. Integrating SCT and HBM within a GSEM framework uncovered pathways linking social context, healthcare access, and breastfeeding behavior. Strengthening rural health infrastructure and enhancing antenatal education for mothers aged 15-19 years and rural mothers are recommended to accelerate progress toward SDG 3 on maternal and newborn health. Future research should use GSEM to capture complex direct and indirect effects in longitudinal or community-based studies.