1. Time series measurements of adrenal hormones, measured in microdialysis fluid
2. Time series measurements of adrenal hormones and ACTH, measured in blood plasma
3. Supporting metadata
4. Non-validated questionnaire responses

Rhythms are intrinsic to endocrine systems, and disruption of these hormone oscillations occurs at very early stages of disease. Since adrenal hormones are secreted with both circadian and ultradian periods, conventional single time point measures provide limited information about rhythmicity, and crucially do not provide information during sleep when many hormones fluctuate from nadir to peak concentrations. If blood sampling is attempted overnight, this necessitates admission to a clinical research unit, can be stressful and disturbs sleep. To overcome this problem, and to measure free hormones within their target tissues, we used microdialysis, an ambulatory fraction collector and liquid-chromatography tandem mass-spectrometry (LC-MS/MS) to obtain high-resolution profiles of tissue adrenal steroids over 24 hours in 214 healthy volunteers. For validation, we compared tissue against plasma measurements in a further seven healthy volunteers. Sample collection from subcutaneous tissue was safe, well tolerated and allowed most normal activities to continue. In addition to cortisol, we identified daily and ultradian variation in free cortisone, corticosterone (CCS), 18-hydroxycortisol (18-OHF), aldosterone, tetrahydrocortisol (THF), allotetrahydrocortisol (aTHF), and the presence of dehydroepiandrosterone sulphate (DHEA-S). We used mathematical and computational methods to quantify the interindividual variability of hormones at different times of the day and develop “dynamic biomarkers” of normality in healthy individuals stratified by sex, age and body mass index (BMI). Our results provide, for the first time, an insight into the dynamics of adrenal steroids in tissue in real world settings and will serve as a normative reference for novel and improved biomarkers of endocrine disorders. (ULTRADIAN, NCT02934399).

Python, 3.9 or later