Tanzania Study Reveals Hidden Regional Infant Mortality Risks

A new study published in the Archives of Public Health highlights significant regional disparities in infant mortality across Tanzania, suggesting that national averages may obscure critical subnational inequities. By examining data from the 2022 Tanzania Demographic and Health Survey (TDHS), the authors demonstrate that geographic location remains a significant factor in infant mortality risk, even after accounting for individual-level characteristics.

The study focused on live births occurring in the five years preceding the 2022 TDHS. The researchers employed two distinct modeling approaches to analyze the data: survey-weighted logistic regression and regional-level logistic frailty models. The survey-weighted model was designed to evaluate population-average associations, offering a broad view of how various factors influence infant mortality across the entire country. In contrast, the regional frailty model was specifically constructed to measure residual regional variability, allowing the team to identify latent geographic risk structures that might remain undetected in standard population-average analyses.

Calibration was assessed using calibration curves and Brier scores, which evaluate how well the predicted risks align with the observed outcomes. The results indicated that the regional frailty model demonstrated improved discrimination compared to the survey-weighted model, with an AUC difference of 0.017 and a p-value of 0.0166.

Among the individual-level factors examined, several variables were associated with higher odds of infant mortality. The analysis revealed that more recent birth cohorts, specifically those from 2020 to 2023, experienced higher odds of infant death compared to earlier cohorts within the study period. These individual-level associations provide important context for understanding the baseline risks that infants face, but they do not fully account for the geographic disparities identified in the study.

The regional frailty model identified specific regions in Tanzania that were linked to significantly higher frailty odds ratios, indicating that certain areas carry a persistent, unexplained risk of infant mortality. The study concludes that significant contextual and geographic heterogeneity persists in Tanzania, extending beyond the individual-level factors typically considered in public health analyses. By identifying these latent geographic risk structures, regional frailty modelling offers significant methodological and policy insights.

The authors argue that subnational risk modeling should be incorporated into equity-oriented public health planning to hasten the achievement of SDG 3.2. The study suggests that a deeper understanding of geographic risk structures can help mitigate the inequities that currently jeopardize progress toward reducing infant mortality in Tanzania.

Subject of Research: Regional disparities in infant mortality in Tanzania using survey-weighted and regional frailty models

Article Title: Regional disparities in infant mortality in Tanzania: uncovering contextual heterogeneity using survey-weighted and regional frailty models

Article References: Dlamini, W. J., Sibiya, S., Nombebe, T., Honwana, F. E., & Gaston, R. T. (2026). Regional disparities in infant mortality in Tanzania: uncovering contextual heterogeneity using survey-weighted and regional frailty models. Archives of Public Health. https://doi.org/10.1186/s13690-026-02072-1

Image Credits: AI Generated

DOI: 10.1186/s13690-026-02072-1

Keywords: infant mortality, Tanzania, regional disparities, frailty models, public health, SDG 3.2, demographic survey, geographic heterogeneity, statistical modeling, health equity, Regional, disparities