Salivary cortisol, serum lipids, and adiposity in patients with depressive and anxiety disorders

Publication Type:

Journal Article


Metabolism, Volume 58, Number 6, p.821-827 (2009)

DOI Name (links to online publication)



AGE; alcohol; Anxiety; Anxiety Disorders; AREA; Cholesterol; chronic stress; cortisol; Disease; DISORDER; HPA axis; HPA-axis; Lipids; LOW-DENSITY-LIPOPROTEIN; Netherlands; Saliva; Smoking; Stress


Depressive and anxiety disorders are associated with an increased risk of cardiovascular disease. Chronic stress induces hypothalamus-pituitary-adrenal (HPA)-axis perturbations, which might subsequently induce atherogenic lipoprotein profiles and adiposity. The aim of the present study was to investigate the relationship between basal saliva cortisol levels and serum lipids and adiposity in psychiatric patients. Eight salivary cortisol samples (awakening; 30, 45, and 60 minutes after awakening; 11:00 AM, 3:00 PM, 7:00 PM, and 11:00 PM) on 2 consecutive days in medication-free outpatients with depressive and/or anxiety disorders (n = 72) and in healthy controls (n = 42) were used to derive 2 measures of HPA-axis function: basal cortisol concentrations (ie, area under the curve [AUC(cortisol)]) and circadian cortisol variability (variability(cortisol)). Index z scores were calculated for dyslipidemia (from serum triglycerides, inverse high-density lipoprotein cholesterol, and low-density lipoprotein cholesterol) and adiposity (from body mass index and waist-to-hip ratio). Regression analyses were conducted to determine the contribution of AUC(cortisol) and variability(cortisol) in explaining the variance of, respectively, the lipid and adiposity index. Patients showed a higher mean AUC(cortisol) compared with healthy controls (t = 2.7, P = .01). Both cortisol parameters were independently associated with dyslipidemia in patients after adjustment for age, alcohol use, and smoking habits (beta = .31, P = .02 and beta = -.29, P = .02, respectively), but not in controls. Cortisol measures were not associated with adiposity in either group. We conclude that elevated basal cortisol concentrations and lower circadian cortisol variability were independently associated with a less favorable lipoprotein profile in patients with depressive and/or anxiety disorders. These data lend support to the hypothesis that the relationship between affective disorders and cardiovascular disease is partly mediated by HPA-axis perturbations