Skip to main content

All-cause and cardiorenal mortality in 6 million adults with and without type 2 diabetes: a comparative, trend analysis in Canada, Spain, and the UK

Ling S, Zaccardi F, Vlacho B, Li P, Gatius JR, Mata-Casas M, Franch-Nadal J, Kosiborod MN, Gillies C, Fenici P, Mauricio D, Shah BR, Khunti K. Diabetes Obes Metab. 2022; Sep 3 [Epub ahead of print]. DOI:

Aims — To understand geographical and temporal patterns in the diabetes gap – the excess mortality risk associated with type 2 diabetes (T2D) – in three high-income countries.

Methods — Using databases from Canada (Ontario), Spain (Catalonia), and the UK (England), we harmonised the study design and the analytical strategy to extract information on subjects aged over 35 years with incident T2D between 1998 and 2018 matched to up to five subjects without diabetes. We used Poisson models to estimate age-specific mortality trends by diabetes status and rate ratios and rate differences associated with T2D.

Results — In more than 6 million people, 694,454 deaths occurred during a follow-up of 52 million person-years. Trends in all-cause mortality rates differed between Ontario and England; yet, the diabetes gaps were very similar in recent years: in 2018, we estimated 1.3 (95% confidence interval: 0.8, 1.8) and 0.8 (0.2, 1.5) more deaths per 1000 person-years in 50-year-old men with diabetes in Ontario and England, respectively, and 8.9 (6.1, 11.7) and 12.1 (9.1, 15.1) in 80-year-old; between-country differences were small also in women. In Catalonia, rate ratios comparing T2D to no diabetes in 2018 were 1.53 (1.11, 2.11) in 50-year-old men, 0.88 (0.72, 1.06) in 60-year-old, 0.74 (0.60, 0.90) in 70-year-old, and 0.81 (0.66, 1.00) in 80-year-old, indicating lower mortality rates in men with T2D from the age of 60 years; rates were similar in women with and without diabetes at all ages. The diabetes gaps in cardiorenal mortality mirrored those of all-cause mortality: we observed consistent reductions in the proportions of cardiorenal deaths in subjects aged 80 years but variations in subjects aged ≤70 years, regardless of the presence of diabetes.

Conclusions — By reducing the confounding impact of epidemiological and analytical differences, this study showed geographical similarities and differences in the diabetes gap: an excess risk of all-cause and cardiorenal mortality in subjects with T2D is still present in Ontario and England in recent years, particularly in elderly subjects. Conversely, there were very small gaps in young men with T2D or even lower mortality rates in older subjects with T2D in Catalonia.