Determinants and temporal trajectories of exclusive breastfeeding cessation in 16 sub-Saharan African countries: a pooled multivariable survival analysis of the demographic and health survey (2019-2024).

backgroundExclusive breastfeeding (EBF) for the first six months of life is the globally recommended standard of infant nutrition, yet early cessation remains prevalent and highly variable across sub-Saharan Africa (SSA). Despite growing evidence on its determinants, few studies have applied pooled survival analysis with community-level contextual variables to examine both the timing and determinants of EBF cessation across multiple countries simultaneously. This study aimed to determine the time to EBF cessation and its associated factors among infants across 16 SSA countries, and to characterize the country-specific temporal trajectories of cessation hazard.

methodsWe analysed pooled Demographic and Health Survey (DHS) data from 16 SSA countries (2019-2024). The analytical sample comprised 20,830 last-born infants (weighted N = 20,085) aged 0-5 months. Cessation of EBF was the time-to-event outcome, defined via 24-hour dietary recall. Kaplan-Meier curves and log-rank tests were used for descriptive survival analysis. A multivariable Cox proportional hazards model stratified by country and adjusted for primary sampling unit clustering was fitted to identify independent determinants of EBF cessation. A Gompertz parametric model was additionally fitted, yielding country-specific shape parameters (γ) that characterise the temporal trajectory of cessation hazard within each country.

resultsOverall, 75.74% of infants were exclusively breastfed for up to five months, with a pooled cessation incidence rate of 0.093 per person-month (95% CI: 0.090-0.095) across 51,047 person-months of observation. Country-specific incidence rates ranged from 0.020 per person-month in Rwanda to 0.195 per person-month in Kenya, reflecting substantial cross-country heterogeneity. Kaplan-Meier analysis revealed that the probability of remaining exclusively breastfed declined from 0.971 at month 1 to 0.360 at month 5, with the steepest decline occurring between months three and five. In the adjusted Cox model, five independent predictors were identified: higher maternal education (aHR = 0.82; 95% CI: 0.69-0.98), Muslim religious affiliation (aHR = 0.73; 95% CI: 0.60-0.88), and early breastfeeding initiation within one hour of birth (aHR = 0.89; 95% CI: 0.84-0.94) were independently protective against cessation, while high birth weight (≥ 4,000 g) (aHR = 1.22; 95% CI: 1.07-1.39) and attendance at eight or more antenatal care visits (aHR = 1.16; 95% CI: 1.03-1.30) were associated with greater cessation hazard. No community-level variable reached statistical significance in the adjusted model. Gompertz analysis revealed significantly increasing cessation hazard over time in 11 of 16 countries, with Rwanda recording the steepest temporal acceleration (γ = 1.299; p = 0.015); four countries showed no statistically detectable change in hazard across the observation period.

conclusionsEBF cessation in SSA is driven primarily by individual-level factors, with the period between three and five months representing a critical window of vulnerability in most countries. Beyond identifying determinants of cessation, this study demonstrates substantial cross-country variation in the temporal trajectory of cessation risk. The country-specific Gompertz shape parameters provide novel evidence on when the hazard of EBF cessation accelerates, offering temporal insights that conventional prevalence-based analyses cannot capture. These findings suggest that breastfeeding interventions should not only target high-risk populations but also be timed according to country-specific cessation trajectories. Strengthening early breastfeeding initiation, integrating breastfeeding counselling into antenatal and postnatal care, and providing targeted support for mothers of high-birth-weight infants may improve EBF retention across SSA.