Spatial Clusters and Predictors of Substance-Related Mortality in Iran: A Geographically Weighted Regression Analysis

BackgroundSubstance abuse is a critical public health concern worldwide, causing significant social, economic, and health burdens. In Iran this issue is driven by diverse socioeconomic conditions and its position along regional drug trafficking routes. This study aims to explore spatial variations and identify factors contributing to substance-related mortality across provinces and counties. MethodsAn ecological study was conducted using a 5-year data from the Iran Legal Medicine Organization (ILMO) in 31 provinces and 424 counties. Descriptive statistics and univariate regression analysis determined the key variables. Geographically Weighted Poisson Regression (GWPR) was employed to assess local impacts and spatial clustering methods were developed to visualize influential variables. ResultsA total of 12,632 substance-related deaths were recorded during the study period, with over 90% of cases among men and 36% aged 30-39 years. Mortality rates ranged from 24.6 per million population in Kohgiluyeh and Boyer-Ahmad to 302.13 per million in Hamedan. Spatial clustering of mortality was significant (Morans I = 0.015, p = 0.014). The GWPR demonstrated superior model performance than ordinary least squares (AIC: 3351 vs. 6764), revealing substantial spatial heterogeneity in risk factors. Percentages of tenant families, older adults, and substance abuse arrests had the strongest positive association. Conversely, substance smuggling arrests, provincial gross income, and distance from key points demonstrated negative associations (p < 0.05). ConclusionSubstance-related mortality in Iran shows significant spatial clustering with geographically varying socioeconomic determinants, indicating need for region-specific prevention and treatment public health strategies rather than uniform national interventions.