The impact of treatment on HPA axis activity in unipolar major depression

Dysregulation of hypothalamic–pituitary–adrenal axis activity in major depressive disorder has been found to normalize with successful treatment, though inconsistencies exist. To determine the magnitude of change in cortisol levels from pre to post-treatment in individuals with unipolar depression quantitative methods of meta-analysis were applied. Thirty-four studies met inclusion criteria and consisted of a total of 1049 depressed patients across study samples. The overall mean effect size of pre–post-treatment cortisol measures indicated that approximately 56% of depressed participants had similar cortisol levels before and after treatment regardless of symptom improvement. The mean effect size of pre–post cortisol measures for those who responded to treatment was larger than the mean effect size of non-responders; however, this difference did not reach statistical significance. As well, no significant differences in mean effect size of pre–post cortisol measures based on type of treatment (e.g. antidepressant vs. ECT) were found. Subtype of depressive illness and length of treatment may contribute to the magnitude of change in cortisol measure before and after treatment. Inconsistent findings within the reviewed literature may confound the overall results. The type of treatment and response to treatment do not appear to impact the magnitude of change in cortisol level pre to post-treatment. Our findings suggest that the utility of cortisol as an outcome measure may be limited to specific subsets of the depressed population, and that given the variability in HPA results between studies, it is premature to state that cortisol is not a good outcome measure.