Abstract
In 203 patients attending a diabetes clinic over five years, the death rate was significantly higher (p = 0.005) among those with typical diabetic retinopathy (DR) (12/46; 26%) than among those without diabetic retinopathy (NDR) at 14/157; 9%. This difference remains significant even when adjusted for age and other potential risk factors such as smoking, male sex, hypertension and greater duration of diabetes (see also first of these two articles in the March/April issue of this journal (pages 35–37)).
Congestive cardiac failure, myocardial infarction and stroke were the main causes of death in all people with diabetes in the study. Cancer was another main cause in those with retinopathy.
Differences in HbA~1~ and lipid measurements between deceased retinopathy and survivors were non‐significant, though the available sample was admittedly small. The most significant difference between surviving and deceased retinopathy was that there were more survivors on insulin (74% v 42%, p = 0.077) but this did not quite attain statistical significance. None of these factors significantly improved our selected logistic model for death rate, which depends only on retinopathy status and age. Eight out of 12 deaths in the DR group occurred in those with mild, non‐eyesight‐threatening, ‘background’ retinopathy plus four with eyesight‐threatening retinopathy, three with maculopathy and one with proliferative retinopathy.
The conclusions drawn from this study are that there is a suggestion that the presence of even mild background retinopathy is associated with a significant risk for future mortality, which persists when confounding established risk factors for mortality are allowed for. The fact that the majority of the deaths were vascular suggests that this subgroup would benefit from further increased efforts at preventing its occurrence and also a more aggressive approach to treating risk factors for vascular disease, when retinopathy of any degree is present.