Abstract
Both time of awakening and stress are thought to influence the magnitude of the cortisol awakening response (CAR), but the relative importance of these factors is unclear. This study assessed these influences in a combined within and between-subject design. Data were collected from 32 men and women working as station staff in the London underground railway system in three conditions: earlyshift days, day-shift days, and control days. Saliva samples were obtained on waking, 30 and 60 min later, together with measures of concurrent subjective stress, sleep quality the night before, and accumulated stress at the end of the day. Participants woke up more than 3.5 h earlier on average on early-shift than day-shift or control days, and cortisol levels on waking were lower in the early-shift condition. The CAR (assessed both with increases from waking to 30 min and with area under the curve measures) was greater on early-shift days. However, respondents were more stressed over the hour after waking and reported more sleep disturbance on early-shift days; when these factors were taken into account, the difference in CAR related to experimental condition was no longer significant. Comparisons were also made between individuals who started their day-shifts in the morning and afternoon. The morning shift group woke an average of 2 h earlier than did the afternoon shift group, but did not differ on stress, sleep quality, or CAR. Stress assessed retrospectively at the end of the day was not associated with the CAR. We conclude that early waking, stress early in the day, and sleep disturbance often coincide, but need to be distinguished in order accurately to interpret differences in CAR magnitude.