Detailed time course of the cortisol awakening response in healthy participants.

The cortisol awakening response (CAR) can be assessed from saliva samples collected at home, which confers ecological validity but lacks researcher oversight. Participant non-adherence to requested saliva sampling regimes leads to inaccurate CAR estimates. Moderate sampling delays of just 8 (5-15) min between awakening and commencement of saliva sampling are reported to result in over-estimated CAR magnitude and earlier peaking. This has been attributed to an observed 'latent' period in which cortisol secretion does not increase for up to 10-min after awakening. Replication of this finding is essential as the findings have considerable implications for CAR research. Healthy participants (n=26) collected saliva samples at 5-min intervals for 60min on 2 consecutive typical weekdays. Full electronic monitoring of awakening and sampling enabled exclusion of non-adherent data (i.e., delays of greater than 5min between awakening and collection of the first sample). In the 0-15min post awakening segment of the CAR a quadratic effect was observed, with no difference between the awakening and 5 and 10min samples. Moderate sampling delays will shift assessment of the CAR just sufficiently along the time axis to not impact upon measurement of the first sample but to remove the immediate post-awakening latent period from CAR estimates-whilst retaining later estimates of elevated cortisol secretion. The implication from these results is that accurate CAR measures can only be determined from data with strict adherence to commencement of saliva sampling following awakening.