Although the development of retinopathy has been related to poor glycemic control in young patients with type 1 diabetes, the association is less clear for type 2 diabetes. Saum and colleagues used a collaboration between family physicians and community optometrists to study the association between glycemic control and retinal changes.

The five general practices and two optometrists serving an English town collaborated to develop a register of diabetic patients in 1996. Practice nurse specialists in diabetes compiled records of all patients with type 2 diabetes. All patients were examined by one of the optometrists, and a scoring system was developed for retinal changes. If changes were detected, the patient was independently examined by a second optometrist who did not know the results of the first assessment. The study compared significant factors in glycemic control between 38 patients with retinal changes and 222 without changes.

In both groups, the median age was 70 years (range, 28 to 98 years) and 57 percent of the cohort was male. All but one of the patients with retinal changes used oral hypoglycemic medications, compared with 70 percent of the comparison group. The median glycosylated hemoglobin A_1c (HbA_1c) was significantly higher in those with retinal changes (8.0 percent compared with 7.1 percent). Similarly, the proportion of patients above the critical HbA_1c value of 6.9 percent (recommended by the British diabetic expert groups) was 79 percent in those with retinal changes compared with 56 percent in those without changes.

Patients with retinopathy had significantly longer duration of diabetes (72 months compared with 48 months) and nonsignificantly higher median microalbumin levels (9.5 mg per L compared with 7.8 mg per L). The groups did not differ in cholesterol, blood pressure, or body mass index measurements. In statistical analyses, patients with HbA_1c levels greater than 6.9 percent or disease duration longer than four years were three times more likely than other patients to have retinal changes.

The authors conclude that patients at greatest risk for retinopathy resulting from type 2 diabetes can be identified in the community through collaboration and simple screening for poor glycemic control and long duration of disease. Conversely, finding retinopathy in a patient with type 2 diabetes should stimulate efforts to improve glycemic control.