US deaths involving diabetes and endocarditis rose from 1,208 to 2,400 a year in 1999–2024, CDC data

An analysis of US death certificates published in BMC Cardiovascular Disorders examined deaths in adults aged 45 years and older in which diabetes mellitus and infective endocarditis were recorded together. The study covers 1999–2024 and identified 41,455 such deaths. Annual deaths rose from 1,208 in 1999 to 2,400 in 2024, and the age-adjusted mortality rate (AAMR) rose from 1.27 to 1.56 per 100,000 biomedcentral.comBMC Cardiovascular Disordersbiomedcentral.comOpen source ↗.

The authors used a population-based ecological time-trend design built on the CDC WONDER multiple-cause-of-death data. Diabetes was defined by ICD-10 codes E10–E14, and infective endocarditis by codes B37.6, I33.0, I33.9 or I38. Rates were calculated per 100,000 persons in the general US population aged 45 years and older, together with age-specific crude rates, and analysed with Joinpoint regression. Race- and ethnicity-specific trends were analysed for 1999–2020, while estimates for 2021–2024 were presented descriptively.

Trend over time

The AAMR declined during 2003–2016, with an annual percent change (APC) of −1.94% (P < 0.0001). It then increased during 2016–2024, with an APC of 4.56% (P < 0.0001). Over the full period the average annual percent change (AAPC) was 0.98% (95% CI, 0.13–1.83; P = 0.0234).

The authors ran sensitivity analyses: they reduced the maximum number of joinpoints from four to three, excluded 2020 and 2021, and restricted the data to the 1999–2020 bridged-race database period. The recent increase persisted in these checks. In the 1999–2020 analysis the final segment remained significantly increasing, although the full-period AAPC was not significant.

Differences by group and region

The full-period AAPC was significant among men (1.76%; 95% CI, 0.19–3.36) but not among women (−0.04%; 95% CI, −0.67 to 0.59). Adults aged 85 years and older showed a significant full-period increase (AAPC, 2.46%; 95% CI, 1.03–3.92) biomedcentral.comBMC Cardiovascular Disordersbiomedcentral.comOpen source ↗.

The study also observed regional and urban–rural variation, and pooled state-level AAMRs varied within each database period.

The authors state that the findings support continued population-level surveillance. They note that the design, which relies on co-recording of diagnoses on death certificates, does not establish individual-level causality or prognosis.