Survival analysis of environmental determinants related to child mortality in Liberia

Although Liberia’s under-five mortality decreased from 268 deaths per 1000 live births in 1990 to 192 in 2000 and 73 in 2023, progress remains suboptimal. Evidence on the impact of environmental factors on child survival remains limited in settings with a history of infectious zoonotic outbreaks. The study employed the life-course approach to examine how environmental factors affected under-five mortality over a 59-month life-course. Retrospective data from the 2019–2020 Liberia Demographic and Health Survey were used. We analyzed data for 5704 under-fives. Descriptive statistics, person-time prevalence rates, Kaplan-Meier survival estimates and Cox proportional hazards model were performed to determine the time to death. Of the 5704 under-fives analyzed, 459 (8.1%) died. Most deaths occurred on the day of birth (23%) or within the next 27 days (25%). In the base model, environmental factors with a higher relative hazard of mortality included rural residence [aHR = 1.38, 95% CI: 1.10–1.74, p = 0.006)], use of “other sources” of drinking water [aHR = 2.08 (95%CI: 1.01–4.32, p = 0.048], exposure to hunted wild animals by household members [aHR = 1.25, 95% CI:1.03–1.51, p = 0.023], and coming from North Western region. These associations were attenuated and no longer statistically significant after controlling for child, maternal, and household factors. A lower relative hazard of mortality was observed among female children, those whose mothers had longer inter-birth intervals, whose mothers’ age at birth were 20–24 years, had normal weight at birth, and were singleton births. Although environmental factors were not statistically significant in the final model, findings from the base model suggest they remain practically important. Interventions targeting environmental determinants of child mortality should also incorporate child, maternal, and household factors.