Back

Food Insecurity as a Moderator of Rural Mental Health: A County-Level Analysis

Krishna, E. S. C.; Shanavas, N.; Gavini, P.; Roso, C.

2026-08-27 public and global health
10.64898/2026.08.25.26361353 medRxiv
Show abstract

Objective: To examine if food insecurity moderates the relationship between rurality and mental health outcomes (suicide mortality, poor mental health days, frequent mental distress) and to assess if these effects vary across U.S. Census divisions. Methods: This county-level (n=2,397) cross-sectional study used OLS and spatial error regression to analyze public data from sources including the County Health Rankings and USDA. We modeled suicide mortality, poor mental health days, and frequent mental distress as functions of the Index of Relative Rurality (IRR) and food insecurity, controlling for median income and provider rates. The suicide model was also tested across nine U.S. Census divisions. Results: Baseline models revealed a paradox: rurality was a direct risk factor for suicide (B=0.400) but protective for poor mental health days (B=-0.224). The national multivariable model revealed a significant, positive rurality-food insecurity interaction for suicide mortality (B=0.861), indicating a synergistic risk. This interaction was not significant for general mental distress, which was more strongly predicted by income and food insecurity. Regional analysis confirmed the suicide interaction was potent in five divisions, including the Pacific (B=3.048) and Mountain (B=1.712) , but absent in others (e.g., South Atlantic). Conclusions: The drivers of suicide are distinct from those of general mental distress and are geographically heterogeneous. The interaction of rurality and food insecurity creates a compounded risk for suicide. Suicide prevention must be regionally-tailored and address structural inequalities, such as food insecurity, alongside clinical care.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.