Spatial clustering and geographically varying correlates of household water disruption across India

Reliable drinking-water services require water to be available when needed, yet conventional measures of access do not adequately capture interruptions in household supply. This study examined the spatial distribution and district-level correlates of household water disruption across India using nationally representative National Family Health Survey-5 data. Household information was aggregated to 707 districts and analysed using Global Moran’s I, Local Indicators of Spatial Association, Ordinary Least Squares (OLS), Spatial Lag Model (SLM), Spatial Error Model (SEM), and Geographically Weighted Regression (GWR). The mean district-level prevalence of water disruption was 20.52%, ranging from 1.11% to 72.18%, with significant spatial clustering and higher-burden clusters concentrated across western, peninsular, Himalayan, and northeastern India. Greater piped-water coverage was positively associated with water disruption, with the coefficient remaining positive after accounting for spatial dependence (OLS β = 0.234; SLM β = 0.135; SEM β = 0.207; all p < 0.001). Reliance on off-premises water sources and a higher proportion of rural households were also positively associated with disruption in the global models. The GWR substantially improved model fit (R² = 0.801; adjusted R² = 0.733; AIC = 4651.41) and demonstrated significant spatial non-stationarity for most coefficients, indicating considerable geographic variation in the magnitude and direction of district-level associations. These findings demonstrate that household water disruption is spatially patterned and that its correlates cannot be adequately represented through geographically constant relationships. The results support complementing measures of household water access with systematic monitoring of service continuity and geographically differentiated water planning to advance equitable progress towards safely managed drinking-water services and Sustainable Development Goal 6.1.