How Long to Hold Steroids Before Cortisol Testing
Hold glucocorticoids for at least 24 hours before measuring morning serum cortisol (with the important exception of dexamethasone, which requires longer discontinuation). 1
Key Testing Parameters
The 2024 European Society of Endocrinology and Endocrine Society joint guideline provides clear direction on this:
- Timing: Measure early-morning serum cortisol between 8:00-9:00 AM
- Holding period: At least 24 hours after the last glucocorticoid dose
- Exception: Dexamethasone requires longer discontinuation due to its extended half-life 1
Important Prerequisites Before Testing
Do not test until the patient has been tapered to physiologic replacement doses (prednisone 4-6 mg daily or hydrocortisone 15-25 mg daily). Testing on supraphysiologic doses is not recommended and provides no useful clinical information. 1
Interpreting the Results
The cortisol value should be viewed as a continuum rather than a binary cutoff:
- >300 nmol/L (10 μg/dL): HPA axis recovery confirmed; glucocorticoids can be safely stopped
- 150-300 nmol/L (5-10 μg/dL): Continue physiologic replacement; retest in weeks to months
- <150 nmol/L (5 μg/dL): Persistent adrenal insufficiency likely; continue replacement and retest in a few months 1
Critical Caveats
Factors that falsely elevate cortisol:
- Oral estrogen/contraceptives (raise cortisol-binding globulin)
- Pregnancy
- Disrupted circadian rhythm (night shift workers, jet lag, severe insomnia) 1
Factors that falsely lower cortisol:
- Low cortisol-binding globulin states
- Improper timing of blood draw
Special Consideration for Oral Contraceptives
If the patient is on combined oral contraceptives, these should be discontinued for at least 4 weeks before cortisol testing, as they significantly elevate cortisol-binding globulin and total cortisol concentrations. While older guidelines recommended 6 weeks, recent prospective data shows normalization occurs by 4 weeks. 2, 3
Common Pitfall to Avoid
Do not assume low cortisol values (<150 nmol/L) on physiologic doses mean the patient cannot successfully discontinue steroids. Recent evidence shows that 42% of patients who successfully weaned off glucocorticoids had nadir morning cortisol values <150 nmol/L, including some with values <28 nmol/L. The HPA axis recovers during physiological dose tapering, not before it. 4 This challenges the traditional guideline thresholds and suggests that symptom-guided tapering with close monitoring may be more effective than strict adherence to cortisol cutoffs.
When Testing May Not Be Necessary
For patients on glucocorticoid therapy <3-4 weeks regardless of dose, testing is unnecessary—simply stop the glucocorticoid as HPA axis suppression is unlikely. 1