JAMA Netw Open. 2026 Apr 1;9(4):e264864. doi: 10.1001/jamanetworkopen.2026.4864.
ABSTRACT
IMPORTANCE: Rural-urban differences in health outcomes have worsened over time, yet the variation in exposure to social determinants of health (SDOH), which are key drivers of these outcomes, has not been adequately quantified.
OBJECTIVE: To explore variation in exposure to a sample of beneficial and adverse community-level SDOH across the rural-urban continuum.
DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study included all residential addresses in Maryland using GPS-Health (Geographic Patterns of SDOH), a 2025 statewide address-level dataset.
MAIN OUTCOMES AND MEASURES: The straight-line (great-arc) distance to the nearest resource (hospital, federally qualified health center [FQHC], pharmacy, Supplemental Nutrition Assistance Program [SNAP] retailer, civic center, and major roadway) or hazard (gun violence incident, eviction site, and Environmental Protection Agency [EPA]-designated site) was estimated. Addresses were categorized using Rural-Urban Commuting Area codes as urban (1-3), large rural (4-6), small rural (7-9), and isolated rural (10). Linear mixed-effects models estimated adjusted differences by rurality.
RESULTS: The study included 2 070 970 addresses: 1 933 793 urban addresses (93.4%), 86 270 large-rural addresses (4.2%), 17 594 small-rural addresses (0.8%), and 33 313 isolated-rural addresses (1.6%). Adults aged 65 years and older were most prevalent in large-rural areas (30.8%), disability was more prevalent in small-rural areas (17.9%), and both were more common in isolated-rural than urban areas (older adults: 27.7% vs 17.1%; disability: 15.6% vs 11.6%). Compared with urban addresses in an adjusted model, isolated-rural addresses were farther from hospitals (estimated difference, 4.22 [95% CI, 3.32 to 5.13] miles), pharmacies (estimated difference, 2.16 [95% CI, 1.54 to 2.80] miles), SNAP retailers (estimated difference, 1.15 [95% CI, 0.83 to 1.47] miles), and civic centers (estimated difference, 1.09 [95% CI, 0.67 to 1.50] miles). FQHCs were closest to large-rural addresses (estimated difference, -2.13 [95% CI, -2.50 to -1.76] miles). Urban addresses were closest to hazards.
CONCLUSIONS AND RELEVANCE: In this cross-sectional study of all residential addresses in Maryland, distance from health-promoting resources increased with rurality, with isolated-rural areas farthest, despite a higher representation of populations with greater health care needs. Our findings support examining rurality as a continuum rather than a dichotomy to better characterize health outcomes and SDOH variation and to inform place-specific interventions.
PMID:41931292 | DOI:10.1001/jamanetworkopen.2026.4864
