Nature Reviews Endocrinology 8, 22-32 (January 2012) | doi:10.1038/nrendo.2011.153
Corrected online: 26 June 2012

Subject Categories: Neuroendocrinology (including the hypothalamus) | Pituitary gland (including the pineal gland) | Adrenal gland (including endocrine hypertension)

Hypothalamic–pituitary–adrenal axis dysfunction in chronic fatigue syndrome

Andrew S. Papadopoulos & Anthony J. Cleare  About the authors


The weight of current evidence supports the presence of the following factors related to hypothalamic–pituitary–adrenal (HPA) axis dysfunction in patients with chronic fatigue syndrome (CFS): mild hypocortisolism; attenuated diurnal variation of cortisol; enhanced negative feedback to the HPA axis; and blunted HPA axis responsiveness. Furthermore, HPA axis changes seem clinically relevant, as they are associated with worse symptoms and/or disability and with poorer outcomes to standard treatments for CFS. Regarding etiology, women with CFS are more likely to have reduced cortisol levels. Studies published in the past 8 years provide further support for a multifactorial model in which several factors interact to moderate HPA axis changes. In particular, low activity levels, depression and early-life stress appear to reduce cortisol levels, whereas the use of psychotropic medication can increase cortisol. Addressing these factors—for example, with cognitive behavioral therapy—can increase cortisol levels and is probably the first-line approach for correcting HPA axis dysfunction at present, as steroid replacement is not recommended. Given what is now a fairly consistent pattern of findings for the type of HPA axis changes found in CFS, we recommend that future work focuses on improving our understanding of the cause and relevance of these observed changes.

Author affiliations

A. S. Papadopoulos & A. J. Cleare
Department of Psychological Medicine, Institute of Psychiatry, Box P074, 103 Denmark Hill, London SE5 8AZ, UK (A. S. Papadopoulos, A. J. Cleare).

Correspondence to: A. J. Cleare

Published online 27 September 2011

* In the version of this article initially published online in Table 3 the findings in relation to the study by Van den Eede et al. should have read "Patients without early-life stress had lower mean cortisol post DEX and lower AUC total post DEX plus CRH than both controls and patients with early-life stress". The error has been corrected for the HTML and PDF versions of the article.