On the survival of individuals diagnosed with type 2 diabetes mellitus in the United Kingdom: A retrospective matched cohort study

Aims: To estimate long-term hazards of all-cause mortality following a diagnosis of type 2 diabetes mellitus (T2DM) using electronic primary care data. Methods: Retrospective matched cohort study using electronic health records from THIN primary care database. Individuals born between 1930 and 1960, diagnosed with T2DM between 2000 and 2016 and aged 50–74 years were selected and followed up to 1 January 2017. Individuals with pre-existing selected severe medical conditions were excluded. The Gompertz-double-Cox model was used to estimate all-cause mortality hazards, adjusting for medical history, socio-demographic and lifestyle factors. Results: A total of 221 182 (57.6% Males, 30.8% T2DM) individuals were selected for the study of whom 29 618 (13.4%) died during follow-up. The adjusted mortality hazard of type 2 diabetes mellitus (T2DM) was estimated to be 1.21[1.12–1.3] and 1.52[1.44–1.6] among individuals diagnosed at 50–59 years and 60–74 years, respectively, compared to controls. Compared to the 1930–39 birth cohort, all-cause mortality hazards were reduced in the 1940–49 cohort, but increased at older ages in the 1950–60 birth cohort for both cases and controls. Conclusion: These hazards associated with T2DM which increase with age at diagnosis are constant across all birth cohorts demonstrating a lack of progress over time in reducing the relative risks of all-cause mortality associated with T2DM. A further study that includes people born after 1960 is needed to fully understand the emerging higher mortality hazards among the younger birth cohorts.