Sickness absence and control of insulin-treated diabetes as assessed by glycosylated haemoglobin.

The sickness absence records for 1986 were obtained for 63 employed insulin-treated diabetic patients who had glycosylated haemoglobin (HbA1) measurements during that year. One subject whose absence was associated with attempts to improve control because of pregnancy was excluded from the analysis. Fifteen had good control (HbA1 8.5 per cent or less) and 47 had poor control (HbA1 greater than 8.5%). The groups were similar for age, sex, duration of diabetes and occupation. The distribution of sickness absence showed greater frequency of absence (p less than 0.05), greater number of working days lost (p less than 0.02), and greater average duration of absence (p less than 0.04) among those diabetic workers with poor control compared to those with good control. Because some individuals with poor control had no absence, HbA1 measurement cannot be recommended on its own to identify those workers who will be absent from work. Its use may lie in indicating a level below which sickness absence is minimized.