Association between cardiorenal syndrome and depressive symptoms among the US population: a mediation analysis via lipid indices

Abstract

Background: Cardiovascular diseases (CVD), chronic kidney disease (CKD), and lipids are positively correlated with the presence of depressive symptoms. However, investigation into the complex link that exists between cardiorenal syndrome (CRS) and lipid indices and depression remains scarce.

Methods: This study analyzed data from 11, 729 adults in the National Health and Nutritional Examination Surveys from 2005 to 2018. Weighted regression analysis was employed to examine the relationships between CRS and depression, CRS and the Patient Health Questionnaire-9 score, and lipid indices with depression. The restricted cubic spline analysis was used to determine whether there is a linear association between lipid indices and depression. Smooth curve fitting was employed to illustrate the relationship between lipids, depression, and cardiorenal diseases. Subgroup and sensitivity analyses were also conducted to enhance the stability of the results. Finally, we applied mediation analysis to explore whether the Atherogenic Index of Plasma (AIP), triglyceride glucose (TyG) index, and remnant cholesterol (RC) mediate the association between CRS and depression.

Results: After applying propensity score matching (PSM), 1,509 adults remained in the study. After PSM, more remarkable results were rendered that CRS was associated with depression compared with non-CRS (OR: 1. 240, 95% CI: 1. 237 ~ 1. 243), only-CVD (OR: 0. 646, 95% CI: 0. 644 ~ 0. 649), and only-CKD (OR: 1.432, 95% CI: 1.428 ~ 1.437) in a fully corrected model. Smooth curve fitting shows that the intersection point of the lines of CRS and non-CRS occurs at a higher value on the horizontal axis than the intersection point of the lines representing CVD and non-CVD. In the fully corrected model, AIP, TyG, and RC did not independently mediate the association between CRS and depression.

Conclusion: There was a significant association between CRS and depression and a linear relationship between AIP, TyG, and RC and depression. However, the above lipid indicators did not mediate the association between CRS and depression.

Keywords: Cardiorenal syndrome; Depression; Lipids; Mediation analysis; NHANES.

Fig. 1

Flowchart for selecting the study population

Fig. 2

Restricted cubic spline plots on the relationship between AIP, TyG, and RC and depression Caption: The navy-blue line in the figure represents OR, and the gray line represents the 95% CI

Fig. 3

Smoothed curve fitting of the relationship between lipid indices and depression across various disease states Caption: A The link between AIP and depression is between CRS and non-CRS, CVD and non-CVD, and CKD and non-CKD.; B The link between TyG and depression is between CRS and non-CRS, CVD and non-CVD, and CKD and non-CKD; C The link between RC and depression is between CRS and non-CRS, CVD and non-CVD, and CKD and non-CKD

Fig. 4

Subgroup analysis of the association between CRS and depression before and after PSM

Fig. 5

Mediation effects of AIP, TyG, and RC in the association of CRS with depression Caption: ACME, average causal mediation effects (indirect effect); ADE, average direct effects; Dep: depression. * P < 0.05, ** P < 0.01, and *** P < 0.001. The model adjusted for age, gender, race, education level, marital status, PIR, drinking, smoking, diabetes, hypertension, BMI, PA, HDL-C, LDL-C, TG, FPG, HbA1c, UA, BUN, eGFR, UACR, antihyperglycemic drugs, antihypertensive drug, antithrombotic drugs, lipid-lowering drugs, and antidepressant drugs