Vision-Focused Nomogram for Assessing the Risk of Depression in Diabetic Retinopathy: Integrating Visual Function,...

Area of Science:

• Ophthalmology
• Psychiatry
• Medical Informatics

Background:

• Diabetic retinopathy (DR) is a leading cause of vision loss.
• Depression is a common comorbidity in patients with DR, impacting quality of life and disease management.
• Accurate risk stratification for depression in DR patients is crucial for timely intervention.

Purpose of the Study:

• To develop and validate a comprehensive nomogram for assessing depression risk in type 2 diabetic retinopathy (DR) patients.
• Integrate visual function, metabolic factors, and psychosocial measures using machine learning for feature selection.
• Create an interpretable tool for clinical use in early depression screening.

Main Methods:

• A cross-sectional study of 490 DR patients, split into training (n=343) and internal validation (n=147) cohorts.
• Multi-step feature selection using LASSO regression, Stepwise Forward Logistic Regression, and Random Forest importance ranking.
• Nomogram development via multivariate logistic regression, validated internally and temporally (n=104).

Main Results:

• The prevalence of depressive symptoms was 29.18%.
• Four key predictors identified: glycated hemoglobin (HbA1c), sleep quality (PSQI), social support (SSRS), and vision-related quality of life (VRQoL).
• The nomogram showed excellent discrimination (AUC training=0.918, validation=0.867) and maintained performance in temporal validation (AUC=0.875).

Conclusions:

• A vision-focused nomogram effectively stratifies depression risk in DR patients.
• The nomogram integrates clinical and patient-reported outcomes, offering an interpretable tool for early screening.
• This tool can assist clinicians in identifying high-risk individuals for prompt intervention.