Abstract
AIM: To examine whether the association between child marriage and maternal healthcare utilization differs between conflict and non-conflict settings, and whether armed conflict amplifies the negative effects of child marriage on maternal healthcare utilization. BACKGROUND: Armed conflicts hinder progress in reproductive and maternal health, particularly in low- and lower-middle-income countries, by weakening health systems, disrupting access to care, and increasing gender-based vulnerabilities. Child marriage, which is common in such contexts, may further limit women's ability to seek adequate maternal healthcare. While both conflict exposure and child marriage are known to adversely affect maternal health outcomes, evidence on their intersection remains limited. Understanding their combined influence is essential for designing effective primary healthcare and humanitarian interventions. METHODS: We used data from 82 Demographic and Health Surveys (1994-2020) across 49 countries, linked spatially and temporally with armed conflict information from the Uppsala Conflict Data Program. The sample included 452,192 women aged 15-49. Maternal healthcare utilization was measured using continuum-of-care indicators: at least one antenatal care (ANC) visit, four or more ANC visits, four or more ANC visits with institutional delivery, and four or more ANC visits with institutional delivery and postnatal care (PNC). Associations were estimated using binomial logistic regression models, with robustness checks including interaction effects, macro-level analyses, and mediation analyses. FINDINGS: Women married before age 18 had significantly lower odds of utilizing maternal healthcare compared to those married at 18 or older. These disparities were strongest in conflict-affected areas, where child brides consistently showed the lowest utilization of ANC, institutional delivery, and PNC. Maternal education, household wealth, urban residence, and media exposure partially mitigated these associations. Additional analyses confirmed the robustness of findings across alternative model specifications, conflict measures, and subgroups.