The effect of glycaemic control and the introduction of insulin therapy on retinopathy in non-insulin-dependent diabetes mellitus.

To study the progression of diabetic retinography in relation to diabetes treatment and glycaemic control in patients with non-insulin dependent (Type 2) diabetes mellitus (NIDDM), we performed a prospective study in a cohort of 1378 diabetic patients, aged > or = 40 years at diagnosis, of whom 333 were treated with insulin, and 1045 with oral antihyperglycaemic agents or diet alone. In the latter group 174 patients changed to insulin therapy during follow-up. We used the Wisconsin scale to grade retinopathy, recorded blindness (visual acuity < or = 0.1) and visual impairment (visual acuity 0.2-0.4), and measured the average HbA1c for each patient during a mean 3.1 year study period. In a multivariate analysis, patients who changed treatment from oral agents or diet alone to insulin therapy had a relative risk of 2.0 (95% confidence interval 1.7-2.3) for progression of retinopathy > or = 3 levels compared with all other patients in the study. The increase in risk remained even after controlling for mean HbA1c (relative risk 1.6; 95% confidence interval 1.3-1.9). Progression > or = 3 levels was significantly associated with a higher incidence of macular oedema and deterioration of visual acuity (p < 0.001). The relative risk for blindness/visual impairment due to retinopathy was 2.7 (95% confidence interval 1.8-4.0) in the group with changed treatment compared with all the other patients in the study. Poor glycaemic control (Hba1c%) before the start of insulin therapy and any retinopathy at baseline were significant risk factors for progression in the group with changed treatment (both p < 0.01). In the whole study group, poor glycaemic control was significantly associated with retinopathy progression > or = 3 levels; the relative risk for those having mean HbA1c above the median being 1.7 (95% confidence interval 1.4-2.1), compared to those with a HbA1c value below the median. Moderate non-proliferative diabetic retinopathy at baseline was also associated with progression (relative risk 2.5; 95% confidence interval 1.4-4.5). In contrast, insulin treatment at baseline was not associated with an increased risk of retinopathy progression. In conclusion, while hyperglycaemia was a risk factor for the progression of retinopathy in all patients, change of treatment from oral drugs to insulin was associated with a 100% increased risk of retinopathy progression and a 3-fold increased risk of blindness/visual impairment.