LAUSR

In 2003, Roenneberg and colleagues defined the Social Jet Lag (SJL) as the difference between the mid sleep times on free days and on work days (Roenneberg et al., 2003). In the recent article, Jankowski described the SJL as groundbreaking and complemented its concept by correcting the sleep length on free days and on work days (Jankowski, 2017). This correction is based on the assumption that during free days the sleep debt accumulated during work days is slept off. We totally agree with Jankowski that sleep length on a free day will always be influenced by the previous work days and we further state that sleep timing will – depending, e.g., on the shift schedule – be influenced, too. Hence, we agree that sleep behaviour on free days cannot be seen as “isolated” from the work days. However, we disagree with the assumption that on free days “any impact of social obligations on sleep” could be excluded. As already described by Aschoff (1954), social zeitgebers have an enormous influence even on behaviour in animals. Accordingly, in humans the influence on sleep timing resulting from various social zeitgebers such as social events, surrounding persons’ chronotypes and cultural effects can definitely not be ignored (Gross et al., 2017). Hence, sleep behaviour on a free day is likely to be affected by precedent work days as well as by various social zeitgebers and does not necessarily represent the pristine internal time – defined as the physiological internal time without non-meteorological external zeitgebers (Gross et al., 2017). Hence, via the SJL it is not possible to compare effective sleep behaviour with the pristine internal time, but, from our point of view, the SJL encompasses the stability of sleep rhythms between work days and free days (even though not considering any information on sleep length) while ignoring the pristine internal time as a reference. In 2017, the Perfect Day approach was published as a tool to estimate pristine internal time, by asking for sleep behaviour on a day of complete preference (Gross et al., 2017). The resulting information on the mid sleep time of a Perfect Day could be compared to themid sleep times of the free days and work days. These comparisons result in what we define as the Work Day Jet Lag (the difference between the mid sleep time of a Perfect Day and aWork Day) and the Free Day Jet Lag (the difference between the mid sleep time of a Perfect Day and a free day).(Figure 1) If we want to use the SJL as one possible risk factor of sleep behaviour on health, the Work Day Jet Lag and the Free Day Jet Lag could complement the SJL and enable supporting individual preventive strategies, as referring to the SJL alone could result in misleading information and entail counterproductive prevention strategies. To give an example, imagine a man with a late chronotype (on his perfect day he would go to sleep at 01:00 and wake up at 10:00). On work days he goes to bed at 22:00 h and sleeps until 6 o’clock in the morning. On free days, he goes to bed at midnight and gets up at 8 o’clock to care for his children. This sleep behaviour results in an SJL of 2 hours. To reduce the SJL the man could for instance get up earlier on free day – if he got up at 6:00 h he would experience 1 hour of SJL only. However, this change would enlarge the chronobiological strain, as his sleep behaviour would differ even more from his pristine internal time. Besides the fact that the SJL has to be seen as limited because of missing information on sleep length, this example reveals that the SJL can provide misleading information when assessing risk of sleep behaviour and preventive strategies. Hence, changes in sleep