Aims: To determine the prevalence of retinopathy and its associations in patients diagnosed clinically with Type 2 diabetes and serum antibodies to glutamic acid decarboxylase (GADA) from a community-based sample.
Methods: In a case-control design, 24 GADA-positive Type 2 patients from the Fremantle Diabetes Study (FDS) cohort were recruited and matched as closely as possible for age, sex and diabetes duration with 72 GADA-negative Type 2 patients from the FDS. Each patient had a detailed clinical and biochemical assessment including slit lamp biomicroscopy and colour fundus photography with Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) grading.
Results: The GADA-positive patients had a significantly higher HbA1c (median (interquartile range); 8.4 (7.3, 9.6)%) than those who were GADA-negative (7.2 (6.5, 8.1)%: P = 0.002). The overall prevalence of retinopathy amongst the 96 subjects was 26.0%. The majority (92%) of the retinopathy detected was mild and non-proliferative. GADA-positive patients had double the retinopathy prevalence of the GADA-negative group (41.7% vs. 20.8%; P = 0.044). In a logistic regression model, diabetes duration, HbA1c, systolic blood pressure and current smoking were each significantly and independently predictive of retinopathy (P < 0.025), but GADA status was not.
Conclusions: These data show that GADA-positive patients have an increased prevalence of retinopathy compared with GADA-negative controls with Type 2 diabetes from an urban Australian community. This increased prevalence is due mainly to relatively poor glycaemic control in the GADA-positive group.