Predictors of mortality from type 2 diabetes mellitus in Canterbury, New Zealand; a ten-year cohort study.

The aim was to establish mortality rates in a cohort of subjects with type 2 diabetes mellitus over 10 years in Canterbury, New Zealand (NZ) and to determine baseline prognostic factors. Subjects (447) with type 2 diabetes (208 male, 239 female; age range 30-82 years, median 62 years; of predominantly European origin) were characterised in a clinic survey in 1989. Individual status (dead or alive) at June 1 1999 (10 year follow-up) was ascertained. Mortality rates were compared with the general NZ population and the relative risk (RR) of baseline prognostic factors evaluated with Cox's proportional hazards model. At 10 years, 232 subjects were confirmed as alive and 187 as dead - only 28 were untraceable. Ten year survival was 55% (95% CI: 50-60) for the cohort, compared with 70% (95% CI: 65-75) at 6 years. Factors assessed at baseline (1989), that were independently prognostic of total mortality, included age (RR 2.0, 95% CI: 1.6-2.5), pre-existing coronary artery disease (CAD; RR 1.7, 95% CI: 1.2-2.4) and albuminuria (RR 1.58, 95% CI: 1.1-2.3). Glycated haemoglobin was not a significant predictor of total mortality, although was a predictor of CAD mortality in those subjects free of CAD in 1989 (RR 1.6, 95% CI: 1.1-2.3). In the latter subset, independent prognostic factors for CAD mortality also included age (RR 2.5, 95% CI: 1.7-3.8), hypertension (RR 1.9, 95% CI: 1.0-3.7), peripheral vascular disease (RR 2.4, 95% CI: 1.3-4.5) and smoking (RR 2.6, 95% CI: 1.2-5.8). Increased mortality in type 2 diabetic subjects is therefore attributable to multiple risk factors. Improved outcomes will depend on interventions targeted at glycaemic and all other remediable factors.