Magliano · The lancet. Diabetes & endocrinology 2026 · Multinational population-based registry cohort study · n=13700000 deaths (1.7 billion person-years)

Trends in cause-specific mortality among people with and without diabetes in high-income settings: a multinational, population-based study.

Cited 2 times in the scientific literature.

Level 3 - non-randomized controlled study

Multinational population-based observational cohort study using administrative registries.

PubMed 41819111 · doi:10.1016/S2213-8587(25)00398-5 · record verified 2026-08-28

What was done

Researchers analyzed aggregated annual mortality data collected between 2000 and 2023 from administrative datasets across 11 high-income jurisdictions. Eligible datasets required continuous enrolment of diagnosed type 1 and type 2 diabetes and provided cause-specific death counts stratified by sex and 10-year age groups. Age- and sex-standardised Poisson regression models were used to estimate cause-specific mortality rates, proportional mortality, and mortality rate ratios comparing people with versus without diabetes.

What was found

Across 1.7 billion person-years of follow-up, 2.7 million deaths occurred in people with diabetes and 11.0 million in people without diabetes. Cardiovascular disease mortality declined in all jurisdictions in both groups; mean 5-year declines in people with diabetes ranged from 8.3% (95% CI 5.9 to 10.7) to 25.4% (95% CI 22.8 to 28.0). Dementia mortality increased in both groups in six of seven jurisdictions. Cancer mortality declined in people with diabetes in three of nine jurisdictions (33%) versus six of nine (67%) in people without diabetes; by the end of the period, cancer was the leading cause of death among people with diabetes in four of 11 jurisdictions (36%). Mortality rate ratios were largely stable, with exceptions including a faster drop in cardiovascular mortality for diabetes in Lithuania (5-year change -7.6%, 95% CI -10.1 to -5.1) and rising dementia mortality rate ratios in Denmark (8.0%, 95% CI 5.0 to 11.1) and Scotland (11.4%, 95% CI 8.5 to 14.3).

Why it matters

Cardiovascular disease is no longer universally the leading cause of death among people with diabetes in high-income settings. Clinical guidelines and preventive strategies need to adapt to the growing burden of cancer and dementia in this population.

Limits

The study is restricted to high-income jurisdictions and may not generalize globally. Cause of death data relied on administrative coding and death certificates, which are vulnerable to diagnostic practice changes over time (notably for dementia). Type 1 and type 2 diabetes were combined in the primary analyses rather than evaluated separately.

Cited by