Health Disparities
Revisiting the theory of “deaths of despair”: a causes of cases and causes of incidence approach Luis E Segura* Luis Segura Seth J Prins Rodrigo Zepeda-Tello Silvia S Martins
Background: Since its publication, the theory of “deaths of despair” has been widely accepted. However, the central claim of this theory—whether increasing prevalences of despair (ΔDespair) led to increased all-cause mortality rates (ΔDeaths), between 1999 and 2014, only among NH white individuals in midlife, higher among those with high education—remains untested. Using Rose’s causes of cases and causes of incidence framework, we outline and test the individual and between-group claims made by the authors of this theory. Methods: We used Cox proportional hazard models to test the individual-level claim: despair increasing the risk of death using data from the NHIS. For the between-group claim, we fitted bootstrapped race- and education-specific autoregressive models with all-cause mortality rates per 100,000 individuals as our outcome adjusted for a 1-year lagged indicator of all-cause mortality rates, changing despair prevalence, and 2-year lagged indicators of confounder rates. We obtained yearly prevalences of despair from the NHIS, yearly rates of all-cause mortality from the CDC detailed mortality files, and confounder rates from the Current Population Survey. Results: At the individual-level, we found that despair increases the risk of death (HR= 1.24 [1.16; 1.31]) for individuals of all race and ethnic backgrounds. At the between-group level, was associated with a yearly increase of 0.17 (-0.75; 1.09) deaths per 100,000 among NH white individuals with high education, and a yearly decrease of -0.25 (-0.87; 0.38) deaths from all causes per 100,000 among NH white individuals with low education. Moreover, these associations were imprecise and not exclusive for NH white individuals in midlife. Conclusion: This study challenges the theory’s central claim, emphasizing the need to explore alternative explanations for the observed mortality increase among white Americans in midlife, such as racial resentment and anxieties.