Abstract:

Objective To investigate the associations of depressive symptoms and depressive episodes with renal hyperfiltration (RHF) in middle-aged and older adults. Methods This cross-sectional study was conducted among 9469 middle-aged and older participants from Lanzhou site of the REACTION study. After excluding individuals with estimated glomerular filtration rate (eGFR) <60 mL·min^-1·1.73 m^-2, urine albumin-to-creatinine ratio ≥30 mg/g, malignant tumors, incomplete depression questionnaire data, or antidepressant use, the participants were assessed for depressive symptoms using the Patient Health Questionnaire-9 (PHQ-9) and categorized in <5, 5-9, and ≥10 groups. Depressive episode status was also evaluated. RHF was defined as residuals above the 95th percentile from a regression model of log-transformed eGFR adjusted for age, sex, height, weight, hypertension, and diabetes. The associations of depressive status with RHF and the dose-response relationship between PHQ-9 total score and RHF risk were assessed using multivariable logistic regression and restricted cubic spline analysis, respectively. Random forest, SHAP, and Boruta were used for exploratory analyses. Results Restricted cubic spline analysis showed a significant nonlinear association between PHQ-9 total score and RHF risk. After multivariable adjustment, a significantly increased RHF risk was found in the PHQ-9 ≥10 group (OR=1.854, 95% CI: 1.074-3.201; P=0.027) but not in the PHQ-9 5-9 and PHQ-9 0-4 groups. When PHQ-9 was dichotomized as <10 and ≥10, a PHQ-9 score ≥10 was associated with a higher risk of RHF (OR=1.864, 95% CI: 1.080-3.210; P=0.025). Mild depressive episodes were not associated with RHF, whereas major depressive episodes were independently associated with increased RHF risks (OR=1.547, 95% CI: 1.082-2.212; P=0.017). Exploratory machine learning analyses showed that depressive episodes and clinically relevant depressive symptoms were stable features for RHF risk identification but with weaker contributions than age, fasting blood glucose, and systolic blood pressure. Conclusion In middle-aged and older adults, depressive symptoms are nonlinearly associated with RHF risks, which increased significantly when PHQ-9 score reached a clinically meaningful threshold. Major depressive episodes are also associated with increased RHF risks, and their temporal relationship and underlying mechanisms require further investigation in prospective studies.

Key words: depressive symptoms, depressive episode, renal hyperfiltration, middle-aged and older adults, Patient Health Questionnaire-9