Longitudinal bidirectional associations between diabetic retinopathy and diagnosed depression: Results from a Danish nationwide registry-based cohort study.

Objective: Diabetic retinopathy (DR) is a feared complication and a leading course of visual impairment, but the connection between DR and depression including the direction has never been studied in a nationwide cohort. We aimed to assess, whether the associations between DR and diagnosed depression are bidirectional.

Methods: We performed a national register-based cohort study of individuals with type 2 diabetes, who attended diabetic eye screening between January 2013 and June 2022. Level of DR was extracted from the Danish Registry of Diabetic Retinopathy. The severity of DR was assessed according to the International Clinical Diabetic Retinopathy severity scale. Diagnosed depression was ascertained by physician diagnostic codes of unipolar depression (F32), recurrent depression (F33) or dysthymia (F34.1) from the Danish National Patient Register. We estimated presence of diagnosed depression according to DR level at index date and risk of diagnosed depression during follow-up using multivariable logistic and Cox regression, respectively. Secondly, we assessed whether diagnosed depression at index date could predict incident DR.

Results: We included 240,893 individuals with type 2 diabetes with baseline rates of diagnosed depression ranging from 5.2 to 6.0 % for DR level 1-4. At index date, individuals with type 2 diabetes and DR were less likely to have a history of diagnosed depression (multivariable adjusted OR, 0.77 [95 % CI 0.73-0.82]). In 226,523 individuals with type 2 diabetes followed for 1,159,755 person-years, 1.7 % developed at least one episode of diagnosed depression. In a model adjusted for age and sex, individuals with DR at index date had an increased risk of incident diagnosed depression compared to those without DR (HR 1.25 [95 % CI 1.16-1.36]). Adjusting for marital status, use of glucose-, lipid- and blood pressure lowering medication, HbA1c, diabetic neuropathy and Charlson comorbidity index waived the above risk (multivariable adjusted HR 1.02 [95 % CI 0.93-1.12]). Furthermore a previous history of diagnosed depression was not associated with increased risk of incident DR (multivariable adjusted HR 0.89 [95 % CI 0.77-1.03]).

Conclusion: In this nationwide cohort study, individuals with DR at first screening were 23 % less likely to have a history of depression, but our data did not support a bidirectional association between DR and depression. Selection bias may have occurred as diagnosed depression is a known barrier for attending DR-screening.