Salience versus magnitude in the measurement of the cortisol awakening response

Abstract

Pulsatile ultradian secretion of cortisol, rarely studied in salivary data, has functional importance in hypothalamic pituitary adrenal (HPA) axis regulation. The first daily ultradian episode, the cortisol awakening response (CAR), was examined in healthy adults, in 5-min secretion rates of salivary cortisol from electronically monitored awakening time to 1.25 h. Aggregated rates revealed a cubic trend, with wave-length of almost exactly 1 h, as predicted from known ultradian periodicity. Peak secretion rate occurred 20-min post-awakening. Peak (20-min) to trough (59-min) amplitude (PTA) expressed a salient signal shape. Rates rose steeply to and from peak, and major secretion was packaged into a few 5-min intervals, inconsistent with normal or uniform distribution of 5-min rates, but consistent with known pulsatile cortisol delivery. Null hypotheses asserting normal or uniform distributions were rejected. Maximal rates overwhelmingly occurred before and minimal rates after 30-mins, with degree of extremity at each polarity significantly positively correlated. To demonstrate utility and reliability of PTA estimation in a clinically relevant domain, re- analyses of a previously published study were conducted. Data from only three saliva samples were used, given importance of cost considerations for many CAR researchers. Difference between mean rates before and after 30-min yielded a simple salience index, highly correlated with PTA derived from full 5-min interval data. CAR salience performed significantly better than traditional AUCi magnitude in discriminating control cases (higher inferred amplitude) and cases with Seasonal Affective Disorder (lower inferred amplitude). Evidence suggested that low AUCi may be more sensitive in identifying within-subject changes (e.g. more depressed mood in winter among SAD cases) and low CAR salience better at revealing enduring between-subjects associations (e.g. underlying disorder vulnerability). Since both PTA salience and AUCi magnitude can be analysed and compared using exactly the same data from the same commonly used saliva sampling points, further research is warranted into the importance of individual differences in patterns of cortisol delivery, not just how much is delivered.

Keywords: CAR; HPA axis; Pulsatility; Secretion rate; Ultradian; saliva.