Study on the association between PHQ-9 score for depression and diabetic retinopathy based on NHANES

Objective  To investigate the association between Patient Health Questionnaire-9 (PHQ-9) score and diabetic retinopathy (DR).

Methods  This study was based on data from the National Health and Nutrition Examination Survey (NHANES) conducted between 2005 and 2018, with a total of 3 742 participants enrolled. PHQ-9 score was defined as the exposure variable, and DR diagnosis as the outcome variable. Covariates included age, gender, ethnicity, educational level, poverty-income ratio, body mass index, smoking status, alcohol consumption status, hypertension, diabetes duration, and glycated hemoglobin level. Weighted multivariate logistic regression, receiver operating characteristic (ROC) curve analysis, weighted restricted cubic spline (RCS) analysis, Spearman rank correlation analysis, mediation effect analysis, subgroup analysis, and interaction analysis were performed to evaluate the relationship between PHQ-9 scores and DR risk.

Results  Among the enrolled participants, 771 cases were diagnosed with DR, accounting for 18% of the total sample. After adjusting for confounding covariates, PHQ-9 scores were positively correlated with DR (odds ratio [ OR]=1.04, 95% confidence interval [ CI]: 1.01-1.06, P<0.001). Specifically, severe depression (PHQ-9 score: 20-27) was significantly associated with DR ( OR=2.46, 95% CI: 1.31-4.63, P=0.005). Moreover, individual items of PHQ-9, including depressed mood, sleep disturbance, fatigue, and difficulty concentrating, were positively correlated with DR. The ROC curve of the fully adjusted covariate model exhibited better predictive performance than the unadjusted and partially adjusted models. RCS analysis indicated a linear positive correlation between PHQ-9 scores and DR risk ( P for non-linearity=0.18). Spearman rank correlation analysis revealed varying degrees of positive correlations among the individual items of PHQ-9 in patients with depression. Mediation effect analysis demonstrated that the poverty-income ratio exerted a mediating effect of 10.70% on the positive association between PHQ-9 scores and DR, while no significant mediating effects were observed for other covariates. Additionally, subgroup analyses across different strata of age, sex, ethnicity, educational level, poverty-income ratio, body mass index, smoking status, alcohol consumption status, hypertension, diabetes duration, and glycated hemoglobin level consistently showed positive correlations between PHQ-9 scores and DR risk (all OR>1, all P<0.05), with no significant interaction detected among these subgroups (all P for interaction>0.05).

Conclusions  PHQ-9 scores for depression are linearly and positively correlated with the risk of DR. Depressive symptoms may serve as a risk factor for the onset and progression of DR, suggesting that mental health management may play an important role in DR care.