Context: The duration of hypothalamic-pituitary-adrenal (HPA) axis suppression after glucocorticoid treatment is uncertain.
Objective: We aimed to determine the duration of HPA axis suppression in prednisolone-treated infants and the age at which circadian variation in salivary cortisol is established in healthy infants.
Design, setting, and participants: Before the adoption of propranolol treatment by the Vascular Birthmarks Clinic, 12 infants with infantile hemangioma received high-dose prednisolone for 12 to 25 weeks' duration, weaned over 4 to 6 weeks, and ceased at age 21 to 31 weeks. Parents collected serial salivary samples at two time points per day (before first and last feed) until circadian variation in salivary cortisol (measured by radioimmunoassay) was observed, when a confirmatory 1 μg Synacthen test was performed. Ten healthy control infants had serial salivary cortisol measurements to determine the age at which circadian variation is established.
Main outcome measure: We defined circadian variation as evening salivary cortisol <50% of the early morning level on two consecutive sampling weeks.
Results: Circadian variation appeared within 6 weeks (median 2.7, range 1.4-5.4) of prednisolone cessation. All confirmatory Synacthen tests were normal (peak serum cortisol >600 nmol/L) and were performed within 12 weeks of prednisolone cessation. Healthy controls developed circadian variation at median 16 weeks of age (range 8-24).
Conclusion: HPA recovery occurred within 6 to 12 weeks, shorter than empirical recommendations, to give stress cover for 6 to 12 months. Reduced duration of stress-cover precautions may reduce parental anxiety and side effects from unnecessary glucocorticoid use. Healthy control infants established circadian variation in salivary cortisol between 2 and 6 months of age.