Ill-health retirement and diabetes mellitus.

Five hundred and five subjects retired for medical reasons in 1987. Two hundred and ninety-nine were male and 206 female. Thirty (5.94 per cent) subjects had diabetes of whom 4 were insulin-treated. Twelve of these subjects retired because of diabetes and diabetic complications; 10 retired because of the former and other pathology and 8 retired because of other pathology but had diabetes incidentally at the time of assessment. The expected prevalence of diabetes in the 20-69-year-old group from recent population studies is 9.2-10.1/1000 population. The figure of 59.4/1000 is significantly greater than this. Excluding those who had diabetes incidentally at the time of assessment reduces this figure to 43.6/1000 which still remains significant. After age-standardization the results remain significant. Morbidity from diabetes as assessed by numbers of retirements with diabetes indicates that this is greater than expected.