How place shapes diarrheal risk: An eco-social analysis of environmental and social infrastructures in riverine Indonesia.

Diarrheal disease remains strongly patterned by environmental and social inequalities in place-based riverine settings, where daily interaction with contaminated water environments shapes exposure risk. Guided by eco-social theory, this study examines how environmental and social infrastructures jointly structure diarrheal vulnerability in riverine communities of Jambi Province, Indonesia. A cross-sectional survey of 463 adult household heads was integrated with river water quality assessment and analyzed using a parsimonious structural equation model to estimate pathways linking social capital, environmental sanitation, hygiene behavior, and diarrheal outcomes. RESULTS: indicated that social capital operated as an upstream social resource associated with improved household sanitation (β = 0.21) and hygiene practices (β = 0.26). Environmental sanitation and personal hygiene were independently associated with lower diarrheal risk (β = -0.19 and β = -0.12, respectively), while degraded river water quality constrained sanitation conditions. Age showed an inverse association with diarrheal outcomes, partly mediated through improved sanitation. The final model demonstrated good fit (RMSEA = 0.042; CFI = 0.949; SRMR = 0.037), supporting an integrated eco-social pathway linking ecological conditions, community organization, and household behavior. Rather than arising solely from individual practices or infrastructure deficits, diarrheal vulnerability reflects the interaction between environmental exposure and socially mediated capacity to maintain sanitation. These findings align with Indonesia's Community-Based Total Sanitation (STBM) strategy, highlighting how collective social processes support sustainable sanitation behaviors in ecologically vulnerable settlements. Strengthening community-based environmental governance alongside infrastructure investment may reduce enteric disease risks in river-dependent populations.