Area-Level Geospatial Analysis to Examine Smoking During Pregnancy with Social Risk Factors in North Carolina, United States

Background Smoking during pregnancy remains a public health concern in the United States, with disproportionately higher rates observed in rural populations. These differences are influenced by factors at system and community levels. This study examines the association between smoking during pregnancy, social risk factors, and geographic context in North Carolina, a state in the Southeast United States. Methods An ecological cross-sectional study was conducted at the county-level using publicly available secondary data from 2018–2022, incorporating spatial statistical methods to examine geographic variation and area-level drivers of smoking during pregnancy. Key independent variables included rural-urban status (2023 Rural-Urban Continuum Codes), managed care regions (North Carolina Department of Health and Human Services designations), and county-level indicators of socioeconomic instability (e.g., median household income), healthcare access (e.g., insurance coverage, OB-GYN physician rate), and neighborhood/built environment (e.g., broadband access, food availability). Descriptive analyses, Spearman correlations, and ordinary least squares (OLS) regression were conducted. Spatial dependence was evaluated using Global Moran’s I and Lagrange Multiplier tests, with a spatial lag regression model applied to adjust for spatial autocorrelation. Results The average county-level rate of smoking during pregnancy was 10.3%, ranging from 1.6% to 21.6%. Significant spatial clustering was observed (Global Moran’s I = 0.55, p < 0.01), with high-high clusters in western counties (Regions 1 and 2) and low-low clusters in central counties (Region 4). Rural counties had significantly higher smoking rates (11.98%) than urban counties (8.19%; p  < 0.0001) and significantly greater socio-economic disadvantage. In the OLS model (R² = 0.58), rurality, lower median household income, fewer OB-GYNs, and regional location were significantly associated with higher smoking rates. Spatial lag modeling improved model fit (R² = 0.71) and confirmed spatial dependence (ρ = 0.53, p < 0.001). Rurality, OB-GYN rates, and regional location remained significant predictors. Conclusions Smoking during pregnancy demonstrates distinct geographic clustering in North Carolina, in the Southeast United States. Rates are influenced by rurality, healthcare access, and regional context. Spatial models are crucial for informing place-based prevention efforts. Leveraging tobacco taxes, Master Settlement Agreement funds, and investments in rural economic development promise to reduce tobacco-related maternal and child health disparities.