Image description: The picture shows a Xray view of a woman's body highlighting the adrenal glands in the body. MEA logo bottom right

Is Low Cortisol a Marker of Long Covid?

ME Association Comment

Lowered levels of the hormone cortisol have been reported in several research studies involving people with ME/CFS and treatment with low doses a steroid hormone called hydrocortisone have been assessed in clinical trials as a possible form of treatment.

Low levels of cortisol are also being reported in Long Covid. However, this review of the research to date is questioning the validity of these findings. The authors point out at the end that there are now far more accurate ways of assessing blood cortisol levels over a 24 hour period and it would be useful to do this in both Long Covid and ME/CFS

Dr Charles Shepherd, Hon Medical Advisor, MEA

The American Journal of Medicine

by Armin Alaedini, Stafford Lightman, Gary P Wormser

Is Low Cortisol a Marker of Long COVID?

Cortisol levels

In two recent studies (Klein et al. in 2023 and Fleischer et al. in 2024), researchers reported on the analysis of hormonal and other markers in blood samples from individuals with post-acute sequelae of SARS CoV-2 infection (PASC) — commonly known as long COVID (LC).

Long COVID is associated with a range of persistent symptoms, including fatigue, pain, and brain fog that can be debilitating. Notably, the most compelling and statistically significant finding of the Klein et al. study revolved around the levels of cortisol, which were reported to be much lower in long COVID patients than in controls without long COVID. The authors of the study concluded that “[s]erum cortisol was the most significant predictor of LC status in the model, and cortisol alone achieved an AUC [area under curve] of 0.96”.

The findings have been widely publicized in the media, engendering speculation regarding their therapeutic and biomarker potential in long COVID. On the other hand, a more recent study by Fleischer et al. found no significant difference or trends in the cortisol levels in a comparison of post-COVID-19 patients, with and without persistent symptoms, apparently contradicting the results reported in the earlier study.

While the cortisol assessments in both of these two studies may represent remarkable and consequential findings in the context of long COVID, unfortunately, they are impacted by significant limitations and concerns regarding how the cortisol levels were assessed and interpreted.

The levels of cortisol in the body are very dynamic, following a distinct circadian pattern throughout the day. Levels typically increase in the first half hour after awakening by up to 60%, in what has been called the cortisol awakening response, followed by a sharp drop over the next few hours, reaching their lowest point in the evening

Underlying this circadian rhythm, cortisol also exhibits ultradian fluctuations in a pulsatile pattern with several peaks and troughs throughout the day that change in amplitude. Because of these fluctuations and the wide and rapid changes in circulating levels, reliable and reproducible evaluation of cortisol secretion is not trivial. Such assessment cannot be based on only a single measurement, but rather requires the collection of multiple blood, saliva, or urine samples throughout the day

In assessing the difference in cortisol output between groups, multiple measurements in blood or saliva samples are usually used to analyze the change from baseline, the size of the cortisol awakening response, or the elevation and slope of the diurnal curve


Surprisingly, none of these approaches was taken in either of the two cited studies of cortisol levels in long COVID.

Both studies conducted a single measurement in blood plasma or serum, taking note of the time of day, but not the time since awakening, for the sample collection. More importantly, no attempt was made to take serial samples to overcome the problem associated with the very rapid changes occurring due to ultradian rhythmicity

Given the presence of persistent fatigue as a hallmark symptom of long COVID, the evaluation of HPA axis function through the assessment of cortisol output is a logical approach. However, considering the available data and the major methodological shortcomings noted in the two published studies it is clear that neither the utility of cortisol as a biomarker nor the therapeutic role of corticosteroids in treating long COVID can be justified at this juncture

The recent availability of new techniques to measure cortisol levels sequentially across the full 24 hours, including during sleep should provide the methodology needed for obtaining reliable and accurate data on HPA axis function in long COVID and other conditions.

Further information

Medical Matters:

Shopping Basket