Abstract
Purpose :
Earlier studies have indicated that the associations between depression and diabetic microvascular complications may be bidirectional, but evidence from population-based studies is still scarce. We aim to test whether the associations between diabetic retinopathy (DR) and major depression are bidirectional in a 5-year prospective cohort study.
Methods :
In Denmark, DR screening has been nationally implemented and results have mandatorily been registered in the Danish Registry of Diabetic Retinopathy (DiaBase) database since 2013. We performed a register-based national cohort study including persons with type 2 diabetes who had attended DR screening with data collected between January 2, 2013 and December 28, 2018. Major depression was defined according International Classification of Disease 10 codes: F32 (unipolar depression), F33 (recurrent depression) or F34.1 (dysthymia) from the Danish National Patient Register. Multivariable adjusted Cox proportional hazard models were used to evaluate the risk of incident major depression among persons with DR along the reverse association.
Results :
In 164,452 patients with type 2 diabetes followed for 508,133 persons-years, 1.2% developed at least one episode of major depression. In a model adjusted for age and gender, hazard ratio (HR) of incident major depression was 1.39 (95% CI 1.24-1.56) for patients with DR at baseline compared to those without. However, this was no longer statistically significant after adjusting for insulin use (HR 1.07 95% CI 0.95-1.21). Conversely, major depression did not predict 5-year incident of DR (adjusted HR 0.97 95% CI 0.88-1.06).
Conclusions :
Although DR and major depression were not longitudinally linked, the use of insulin was a strong predictor of incident major depression in a nationwide cohort of persons with type 2 diabetes. Insulin use may be a proxy of dysregulated blood glucose or more advanced disease, which may cause the increased risk. Psychological problems due to insulin injections should also be suspected.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.