Abstract P436: Trait Anxiety and Anger Do Not Predict Incident Type 2 Diabetes Independent of Depression: the Multi-Ethnic Study of Atherosclerosis

Introduction: Prior studies have shown a bidirectional association between depression and type 2 diabetes mellitus (T2DM); however, the prospective association of anxiety and anger with T2DM is not established. We hypothesized that trait anxiety and anger would predict incident T2DM, independent of depressive symptoms, in a multi-ethnic cohort. Methods: In the Multi-ethnic Study of Atherosclerosis (MESA), we prospectively examined the association of trait anxiety and trait anger (assessed via the Spielberger Trait Anxiety and Trait Anger Scales, respectively) with incident T2DM over 7 years in 5,941 White, Black, Hispanic, and Chinese participants (53.6% women, mean age 61.8 years) at baseline without prevalent T2DM or cardiovascular disease. We used multivariable logistical regression models to calculate the odds ratios (OR) of incident T2DM by anxiety quartile and anger category (low, medium, high). Results: In the minimally adjusted model, compared to quartile 1, the OR for incident T2DM were 1.25 and 1.38 for trait anxiety quartiles 3 and 4, respectively and the OR for incident T2DM was 1.50 for those with high trait anger compared to low trait anger (Table). In both models, these associations were attenuated and became non-significant following additional adjustment for depressive symptoms (modeled as continuous Center for Epidemiological Studies-Depression score) and antidepressants. Additional adjustment for socioeconomic, behavioral, metabolic, and inflammatory factors did not change either association (Table). Conclusion: Higher trait anxiety and anger were associated with incident T2DM; however, these associations appeared to be explained by the presence of depressive symptoms and socioeconomic status. These data suggest that the emotional traits of anxiety and anger may impact T2DM risk due to their symptoms overlap with depression.