Differential prognostic burden of cardiovascular disease and lower-limb amputation on the risk of all-cause death in people with long-standing type 1 diabetes

Abstract

Abstract Background Cardiovascular disease (CVD) and nontraumatic lower-limb amputation (LLA) each results in reduced life expectancy in patients with type 1 diabetes, but the differential burden between these conditions is unknown. We compared the effects of CVD and LLA on the risk of mortality in people with type 1 diabetes. Methods We used pooled data from the SURGENE, GENEDIAB, and GENESIS prospective cohorts. Data were divided into: 1/absence of CVD (myocardial infarction and/or stroke) nor LLA, 2/history of CVD alone without LLA, 3/LLA alone without CVD or 4/both conditions at baseline. Participants with baseline history of peripheral artery disease were excluded from groups 1 and 2. The study endpoint was any death occurring during follow-up, regardless of the causes. Results Among 1169 participants (male 55%, age 40 ± 13 years, diabetes duration 23 ± 11 years), CVD, LLA or both were present at baseline in 49 (4.2%), 62 (5.3%) and 20 (1.7%) subjects, respectively. All-cause death occurred in 304 (26%) participants during 17-year follow-up, corresponding to 18,426 person-years and an incidence rate of 16 (95%CI, 15–18) per 1000 person-years. The risk of death increased in individuals with baseline history of CVD (adjusted HR 2.00 [95% CI 1.34–3.01], p = 0.0008) or LLA (2.26 [1.56–3.28], p < 0.0001), versus no condition, with an additive effect in people with both conditions (5.32 [3.14–9.00], p < 0.0001). No incremental risk of death was observed in people with CVD versus LLA (0.87 [0.54–1.41]). Compared with no condition, CVD and LLA were similarly associated with reduced life expectancy during follow-up: 2.79 (95% CI 1.26–4.32) and 3.38 (1.87–4.88) years, respectively. Combined conditions expose to 7.04 (4.76–9.31) less years of life expectancy (all p < 0.0001). Conclusions CVD and LLA conferred a similar burden regarding mortality in type 1 diabetes population. Our findings encourage a careful consideration of people with type 1 diabetes and LLA as usually recommended for those with CVD, in terms of management of risk factors, treatments and prevention.