Abstract
Background: Data on mortality from Diabetes Mellitus (DM) in the United States (US) are available from various sources, including the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research (CDC WONDER) death certificate data and the Global Burden of Disease Study (GBD). This study aimed to compare DM mortality reporting between the CDC WONDER and GBD datasets.
Methods: In this retrospective cohort study, we calculated the absolute number of deaths and crude mortality rates attributed to DM as the underlying cause of death in CDC WONDER and GBD in the US from 1999 to 2021 (CMR with 95% Confidence Intervals (CI) or Uncertainty Intervals (UI)) per 100000 population. Because of methodological differences between datasets, results were also evaluated based on the inclusion and exclusion of mortality from DM with renal complications.
Results: The reported CMR for DM increased from 34 (95% CI: 34 to 35) to 41 (95% CI: 41to 42) in CDC WONDER but decreased from 36 (95% UI: 32 to 37) to 29 (95% UI: 27 to 31) in GBD from 1999 to 2021. When deaths from DM with renal complications were excluded from CDC WONDER to mirror GBD reporting, CMRs in 1999 and 2021 from CDC WONDER were 33 (95% CI: 33 to 34) and 29 (95% CI: 28 to 29), respectively, and trends were generally similar.
Conclusion: Estimates of DM mortality rates and temporal trends in the US vary across commonly utilized sources of mortality data. These results have important implications for epidemiological efforts to understand and interpret DM mortality reporting.