Spatial Variation In Legal Opioids and Its Influence On Suicide Rates

Keywords: suicide, opioids, spatial variation
Session Type: Virtual Paper Abstract
Day: Tuesday
Session Start / End Time: 3/1/2022 09:40 AM (Eastern Time (US & Canada)) - 3/1/2022 11:00 AM (Eastern Time (US & Canada))
Room: Virtual 16

Abstract

There are many covariates posited to be associated with suicide. However, the variation in mortality and the association with these covariates is modest when compared with models of other major causes of mortality (e.g., cancer). Large regional differences in suicide do exist, which have been partially explained in the literature by differences in a variety of ecological and demographic variables, including altitude, ethnicity, and a host of other factors. This variation ranges from 8 to 10 suicides per year per 100k population in the US Atlantic seaboard states of NY/NJ/DE to three times that in the mountain west states. I will present a set of negative binomial mixed effects models nesting the 2554 counties with suicides reported by the CDC in their states on the most commonly hypothesized covariates of suicide, adding per capita, legally prescribed opioids. This variable is significant and meaningfully large. This relationship between prescribed opioids and suicide counts holds across various rural/urban definitions. Further, I will show that when the suite of covariates are considered together (rather than independently, as has been commonly done), multiple broadly accepted covariates are either insignificant or appear to have little explanatory power. One such finding is ethnicity, where I will show that urban counties with higher populations of non-Hispanic whites have lower suicide counts than baseline.