Do retinopathy signs in non‐diabetic individuals predict the subsequent risk of diabetes?

We read with interest the article by Wong et al,¹ which studied a non‐diabetic population consisting of 7992 people aged 49–73 years. Non‐mydriatic retinal photographs of one eye were taken and graded for retinopathy lesions using standardised protocols. Surprisingly, the presence of typical retinopathy lesions (microaneurysms or retinal haemorrhages) in people without diabetes did not significantly predict subsequent development of diabetes over a period of 3.5 years. Incident diabetes developed in 4.7% and 3.6% of people with and without retinopathy lesions at baseline, respectively, with a multivariate adjusted odds ratio, OR, 1.1, 95% confidence interval, CI, 0.7 to 1.9. However, in people with a family history of diabetes, presence of retinopathy lesions at baseline predicted a doubling of the risk of incident diabetes (OR 2.0, CI 1.1 to 3.8).

We previously reported findings from the Blue Mountains Eye Study cohort (BMES, n = 3654) that the 5 year risk of developing diabetes in people without diabetes but with retinopathy lesions at baseline was 3.5% (7/202).² The BMES examined 3654 participants at baseline (1992–4) and re‐examined 2335 participants (75% of survivors) 5 years later (1997–9). Dilated six field retinal photographs of all participants were taken at the baseline and follow up examinations. Diabetes was diagnosed either from medical history or fasting blood glucose ⩾7.0 mmol/l at examination. Of the baseline participants without diabetes, 202 had retinopathy lesions (haemorrhages, microaneurysms, soft and hard exudates). Our 3.5% diabetes incidence over 5 years in this group is relatively similar to the 4.7% diabetes incidence over 3.5 years reported by Wong et al¹ in people with retinopathy lesions at baseline. These consistent findings suggest that retinopathy lesions occurring in people without diabetes are likely to have multiple aetiologies. In people with a family history of diabetes, retinopathy lesions may indicate a preclinical stage of diabetes. In the great majority of people without diabetes, however, retinopathy lesions are not necessarily linked to blood glucose. Reports from the BMES³ and Hoorn Study⁴ showed that baseline impaired fasting glucose or impaired glucose metabolism did not predict incident retinopathy lesions in people without diabetes. Older age and blood pressure, however, were strongly related.²,⁵ It is possible that the same phenotype can result from different pathogenic conditions (such as hypertension⁶) that damage the microvasculature. Given that retinopathy lesions predict systemic vascular outcomes,⁷ further research to clarify the causes of these lesions is warranted.