Abstract 658: New Onset Diabetes Does Not Pose Additional Risk - Analysis Of 15,111 Treated Hypertensive Patients With 40 Year Follow-up.

Abstract

Objective: It is unclear whether new onset diabetes (NOD) is a separate entity associated with excess risk in hypertensive patients. Methods: We studied 15111 hypertensive patients with up to 40 year follow-up at the Glasgow BP Clinic database. Diabetes status was defined based on hospital admissions for any diabetes related diagnosis or prescription of anti-diabetic drugs or diabetes monitoring materials. The date at first hospital encounter either for prescription or admission was considered as the onset of diabetes. NOD was classified into early and late (diagnosis <10yrs or >10years from first clinic visit). Cause-specific outcome analysis was performed using multivariate-adjusted Cox proportional hazards (Cox-PH) models. In order to address any potential competing risk introduced due to the long follow-up period, an additional composite end point of all-cause mortality+NOD was analysed. Results: There were 2521(17%) patients with DM, of whom 2061(14%) had NOD. The incidence rate of NOD was 9.2 per 1000 person-years. Prevalence of early NOD was 898 (6%) and late NOD 1163 (8%). The total time at risk was 239,952 person-years with a median survival time of 28.04 years (IQR: 16.24-39.95). There were 5225 deaths (52% from cardiovascular causes) during the follow-up period. Independent predictors of new-onset diabetes in order of decreasing significance were baseline glucose, BMI, age, alanine transaminase, alkaline phosphatase, gamma glutamyl transpeptidase and bilirubin. Of these age, glucose, BMI and alkaline phosphatase remained top predictors for the composite outcome of NOD+all-cause death. The mortality risk was the highest in those with prevalent DM (HR=1.5[95%CI=1.2;1.9]) and lowest in those with late NOD (0.79[0.68;0.92]). Early NOD and non-diabetic subjects had similar risks. Conclusions: Indices of liver function tests predict the risk of NOD and mortality in addition to BMI and baseline glucose. The risk posed by NOD is related to duration of diabetes primarily indicating the importance of efforts to delay onset of NOD.