Deconstructing the cortisol awakening response (CAR): Post-peak decline predicts health vulnerability in healthy...

The cortisol awakening response (CAR) has two regulatory pathways underlying its components typically not discriminated in CAR studies. The timing of the peak, post-peak secretion and overall levels were assessed in relation to health in this directional hypothesis-driven study.

Healthy females (N = 61, aged 18-38 years) collected saliva at awakening and three further 15-min intervals over seven consecutive days in their domestic setting. Awakening and saliva sampling times were electronically monitored (EM). Participants provided self-reported sampling times as well as demographics and health status. Analyses were conducted for 56 participants providing full cortisol and health data. Participants were categorised as 'healthy' (N = 34; 61%) or 'less healthy' (N = 22; 39%) based on their self-ratings.

Multilevel modelling using 'real time' sampling (EM-estimated) revealed significant cortisol growth curve shape differences between health groups. In comparison to the less healthy group, the healthy group showed lower secretion rates in the final sampling (3rd to 4th) interval with more negative rates (declines) and lower mean cortisol (over 7 study days and 4 samples). The less healthy had more final sample (later) peak cortisol concentrations. Receiver Operating Characteristic (ROC) analyses of CAR composite measures (decline, peak time, and mean levels) were conducted to optimise health status discrimination. Lower third to fourth sample secretion rate was the strongest discriminator, correctly detecting the health status of 71% of all participants. Lower cortisol concentration explained 68% and higher percentage of days with final (fourth) sample peaks explained 61%. Maximum detection was achieved by a combined decision criterion which correctly identified exactly three-quarters (75%) of cases. Findings were robust to known cortisol covariates.

This novel CAR analysis offers insight into potential health vulnerability in younger healthy populations, perhaps an early indicator of hypothalamic-pituitary-adrenal (HPA) axis dysfunction.