Objectives To assess the effect of diabetes on risk of occupational injury, we undertook a case-control analysis nested within the UK General Practice Research Database (GPRD).

Methods The GPRD logs all primary care data for participating general practices (6% of the population). Medical consultations are classified by the Read system and drug prescriptions according to the British National Formulary (BNF). We identified 1,348 patients aged 16–64 years consulting over a 10-year period with workplace injury (cases) and 6,652 age, sex, and practice-matched controls with no such consultation. Groups were compared in terms of their diabetic status (defined by 320 Read codes and 355 BNF drug codes); and for those with diabetes, according to risks from diabetic eye disease, other complications, blood sugar-lowering treatment (insulin or oral hypoglycaemics), and indices of sub-optimal control. Associations were explored using conditional logistic regression.

Results In all, 199 (2.5%) subjects were classed as diabetic before the index date, including 77 with eye involvement, 86 on insulin and 52 with poor diabetic control. Odds of occupational injury consultation were seldom elevated relative to non-diabetics (e.g. OR 1.01 overall, 1.02 in diabetics on insulin) and for some measures were lower (e.g. OR for eye involvement 0.72). Only suboptimal chemical control (HbA1C >7%) in the 12 months before the index date was associated with a slightly elevated risk (OR 1.35); no differences were statistically significant.

Conclusions These findings are reassuring. Lower risks in some comparisons may reflect chance or a degree of health selection out of hazardous work. In any event, current employment practices are not placing diabetic workers at particular risk of workplace injury.