Simon and Masters Respond to “Small Correlations Among Deaths of Despair”

We thank Siddiqi and Sod-Erdene (1) for a thoughtful discussion of our paper (2). Given that the “deaths of despair” (DOD) perspective (3 –5) first emerged as a backend theory to explain empirical anomalies in US mortality trends (i.e., rising midlife mortality among White Americans), it is important to investigate the “empirical uniformities” advanced by the theory’s central assumptions (6, p. 39). Indeed, proponents of other middle-range theories have welcomed and responded to empirical tests of their core claims (e.g., Fundamental Cause Theory (7 –11)). In our analyses, we considered the central DOD claim that among White Americans, deaths from suicides, alcohol use, and drug use constitute a composite DOD category because the deaths move together “both temporally and spatially” and are thus “likely symptoms of the same underlying epidemic” (3, p. 15081). Illustrating small county-level correlations between the separate causes of death (COD) was not intended to be dispositive of all DOD scholarship. Rather, it is just one datum to consider among other nonsupportive data. For example, evidence suggests that widening educational gradients in US White life expectancy are primarily driven by deaths from drug use, cardiovascular disease, and other internal causes of death, not by suicide or alcohol-related deaths (12). Similarly, absolute increases in summed DOD death rates among US White men and women overwhelmingly reflect increases in drug-related death...
Source: American Journal of Epidemiology - Category: Epidemiology Source Type: research