Background Diabetes mellitus (DM) and depression commonly co-occur, worsening self-management, complications, and survival. Yet long-term, population-based surveillance of mortality that involves both conditions remains limited. Objective To characterize national trends in mortality co-listing diabetes and depression and to examine disparities by demographics and geography. Methods We analyzed U. S. Multiple Cause-of-Death records from CDC WONDER for 1999–2023. Deaths that listed diabetes (ICD-10 E10–E14) and depressive disorders (ICD-10 F32–F33, F41.2) as underlying or contributing causes were included. We calculated crude and age-adjusted mortality rates (AAMRs) per 100,000, directly standardized to the 2000 U. S. population, and fitted segmented Joinpoint regression to estimate annual percent change (APC) and average APC. Stratifications included sex, age group, race/ethnicity, Census region, state, urban–rural category (2020 NCHS scheme applied uniformly across years), and place of death, with adherence to CDC WONDER suppression rules. Results National AAMRs were low and broadly stable through the mid-to-late 2000s, rose in the late 2010s, and plateaued in the early 2020s. Rates were consistently higher in men than women and increased with age, peaking in adults aged 85 years or older. By race/ethnicity, non-Hispanic American Indian/Alaska Native populations had the highest AAMRs, followed by non-Hispanic Black and Hispanic groups, with non-Hispanic White and non-Hispanic Asian/Pacific Islander populations lower. Regionally, the West had the highest rates and the Northeast the lowest; nonmetropolitan counties exceeded metropolitan counties. States showed wide heterogeneity, with roughly threefold differences between the top and bottom deciles. Conclusion Mortality involving co-occurring depression and diabetes shows an upswing in the late 2010s and substantial demographic and geographic inequalities. These findings support targeted, place-based strategies and integrated diabetes–mental health care to reduce preventable deaths.
Wang et al. (Tue,) studied this question.