The pathophysiology of depression has been associated to dysregulation of the hypothalamic-pituitary-adrenal axis and the use of salivary cortisol measures is increasingly being incorporated into research. The aim of the present study was to investigate whether salivary cortisol differs for patients with depression and control persons. We did a systematic review with sequential meta-analysis and meta-regression according to the PRISMA Statement based on comprehensive database searches for studies of depressed patients compared to control persons in whom salivary cortisol was measured. Twenty case-control studies, including 1354 patients with depression and 1052 control persons were identified. In a random-effects meta-analysis salivary cortisol was increased for depressed patients as compared to control persons on average 2.58 nmol/l (95% C.I.: 0.95-4.21) p=0.002 in the morning and on average 0.27 nmol/l (95% C.I.: 0.03-0.51) p=0.03 in the evening. In a fixed-effects model the mean difference was 0.58 nmol/l (95% C.I.). Study sequential cumulative meta-analyses suggested random error for the finding of this rather small difference between groups. The reference intervals for morning salivary cortisol in depressed patients (0-29 nmol/l) and control persons (1-23 nmol/l) showed substantial overlap suggesting lack of discriminative capacity. These results should be interpreted with caution as the heterogeneity for the morning analysis was large and a funnel plot, suggested presence of bias. Further, in meta-regression analyses higher intra-assay coefficients of variation in cortisol kits (p=0.07) and mean age (p=0.08) were associated with a higher mean difference of morning salivary cortisol between depressed and controls, while gender and depression severity were not. Based on the available studies there is not firm evidence for a difference of salivary cortisol in depressed patients and control persons and salivary cortisol is unable to discriminate between persons with and without depression.
Copyright © 2010 Elsevier Ltd. All rights reserved.