Three-week actigraphy to assess sleep behaviour and circadian rest-activity patterns in suspected and confirmed Cushing’s syndrome: an exploratory prospective cohort study

PURPOSE: Sleep disturbances are common in endogenous Cushing's syndrome (CS), impair quality of life, and often persist despite remission. We evaluated the diagnostic value of actigraphy in suspected CS, characterised sleep and rest-activity patterns during remission, and examined associations between late-night salivary cortisol (LNSC) and objective sleep measures. METHODS: Exploratory prospective single-centre cohort study at LMU Hospital Munich with two cohorts: (1) individuals evaluated for suspected CS and (2) patients in remission, stratified by recovered versus persistent adrenal insufficiency. Participants wore an ActTrust2 actigraph for 21 days, provided daily LNSC samples, and completed the Munich Chronotype Questionnaire. RESULTS: Cohort 1 included 16 patients with confirmed CS and 13 with exclusion of CS. Actigraphy detected comparable sleep efficiency, fragmentation and circadian rest-activity rhythms in both groups, without distinct alterations attributable to active CS, except for a significantly earlier chronotype (p=0.022). Cohort 2 comprised 23 patients in remission (13 with persistent, 10 with recovered adrenal insufficiency) and showed similar sleep characteristics across subgroups. LNSC revealed high inter- and intraindividual variability without consistent associations with actigraphy-derived sleep parameters. Statistically significant effects were confined to patients with persistent adrenal insufficiency, in whom higher LNSC (likely reflecting non-physiological replacement therapy) was associated with earlier bedtime, longer time in bed, prolonged sleep-onset latency, and reduced sleep efficiency. CONCLUSION: Apart from chronotype, actigraphy showed limited discriminatory value between clinical groups. LNSC was highly variable and showed no consistent associations with sleep parameters, underscoring the need for improved biomarkers and monitoring strategies in CS.