Background
Low serum cortisol has been associated with long COVID (LC) in some but not all studies. However, these studies examined untimed cortisol levels, which may not adequately assess adrenal function. Here we assessed untimed, timed, and stimulated cortisol levels and tested their association with LC.
Methods
Untimed total cortisol levels were quantified in 711 serum samples obtained 1-12 months post-SARS-CoV-2 infection in 381 participants enrolled in two cohort studies between 2020-2022. LC at the time of blood draw was defined as reporting at least one symptom and not having returned to pre-COVID health status.
Additionally, we prospectively enrolled 44 adults with moderate-to-severe LC and 48 adults matched on age, sex, BMI, years of education, and months since acute COVID-19 who recovered fully from COVID-19 within 6 weeks (controls). In this third cohort, we excluded those with recent use of any type of corticosteroids including oral, inhaled, topical, or injectable. The cortisol awakening response was assessed in saliva (Salimetrics) at 5, 30, and 60 minutes after morning awakening in all, and diurnal cortisol was assessed after awakening, before lunch, before dinner, and before bedtime in half. The high-dose adrenocorticotropic hormone (ACTH) stimulation test with blood sampling before, 30-, and 60-minutes post-injection was used to assess adrenal function in 16 LC and 16 controls. Mann Whitney U test was used to test differences between groups.
Results
No significant differences in serum cortisol were observed between LC and never-LC participants among 381 participants with untimed blood draws, even if segregated by 2-hour time window of blood draw (example in Figure).
The morning cortisol awakening response and diurnal pattern of cortisol secretion were preserved and not significantly different between LC and controls at all time points. Peak morning salivary cortisol trended lower in LC than controls (p=0.11). Neither LC nor controls had abnormal ACTH stimulation test results, and no differences were observed in post-stimulation cortisol levels. There was a trend towards higher pre-stimulation endogenous ACTH in LC (p=0.06).
Conclusions
In people with LC not taking exogenous corticosteroids, we observed no abnormalities in serum total cortisol or salivary cortisol levels, with or without exogenous adrenal stimulation, suggesting that adrenal function is preserved in people with long COVID.