Analysis of the duration and extent of the legacy effect in patients with type 2 diabetes: A real-world longitudinal study

Abstract Aims To analyze the duration and extent of the legacy effect on diabetic complications in real-world patients with type 2 diabetes. Methods This was a retrospective cohort study. We included the following three cohorts of patients: diabetic retinopathy (DR) (n = 1107), diabetic kidney disease (DKD) (n = 1486), and cardiovascular disease (CVD) (n = 1485). Patients were enrolled from 1995 to 1999 and followed up to 2017. Endpoints were DR incidence, ≥40% decrease in estimated glomerular filtration rate, and CVD incidence. The relationships between HbA1c as a time-dependent variable and the risk of reaching each endpoint were analyzed using multivariate Cox regression models. Results A total of 313 patients developed DR, 316 developed DKD, and 177 developed CVD. Hazard ratios as a function of time-dependent HbA1c (moving mean) accumulated over time. This accumulation was largest for DR, followed by DKD and CVD. The hazard ratios for each endpoint reached a plateau during the preceding 14–19 years. Conclusions The effect of past glycemic control may continue during 14–19 years, with a greater effect during ≤10 years. Therefore, the end of the legacy effect could be 15–20 years. This effect may be the greatest for DR, followed by DKD, and the smallest for CVD.