Geographic variation in male suicide rates in the United States

This study examines geographic variability of factors associated with male suicide in the United States using county-level data covering 2000 to 2006. Three variables are used as indicators of social isolation: separated/divorced marital status, migration status, and unemployment. A geographically weighted regression analysis shows variation from analogous global ordinary least squares and spatial error regression analyses. Separated/divorced marital status demonstrated a global positive influence. Migration and unemployment effects ranged from positive to negative across the United States, showing some geographic clustering. The findings suggest regional variation is masked by global models and the effect of social isolation indicators have on suicide may vary with geographic context. Any detection of at-risk population will require careful evaluation of privacy issues given the sensitive nature of the health topic.