The early shift sleep schedule, for 4am, 5am and 6am starts

The subtraction is the easy half and it is the same for everyone: shift start, minus the commute, minus however long you need to get out of the door, gives the alarm, and seven and a half hours back from the alarm gives lights out. Do that for a 5am start and it lands in the early evening. The hard half is that the answer is usually a bedtime you cannot reach, because for most people it falls inside the two to three hour window before their normal sleep onset when the body clock is pushing hardest to keep them awake. Going to bed earlier walks further into that window rather than out of it. Morning light timed against your own temperature minimum is what moves the window, and that is the number worth having, not the bedtime.

A 6am start typically needs an alarm between 4:30am and 5am once you count the commute, which puts the sleep window somewhere around 9pm to 9:30pm for seven and a half hours. That is roughly 45 minutes easier than a 5am start, and the research says so in numbers: between 04:30 and 09:00 starts, every hour later the shift begins is worth about 42 minutes more sleep. It is the single biggest lever anyone rostering early shifts has, and it is worth raising with whoever writes the roster.

Accept up front that the sleep will be short. Shifts starting before 04:30 average about five hours in the research, against about eight for a 10am start, so a 4am roster is running a three hour deficit every night and no amount of routine closes that. What is left to manage is where the shortfall hurts. Get bright light early in the shift to hold your clock where the roster needs it, take the caffeine early rather than in the last hours, protect the evening from light so the bedtime is reachable, and treat the mid-afternoon as the dangerous part of the day rather than the morning.

It is usually called the forbidden zone, and it is a window of roughly two to three hours immediately before your habitual sleep onset when falling asleep is at its hardest. Lavie identified it in the 1980s and Strogatz, Kronauer and Czeisler described the mechanism as the circadian pacemaker interfering with sleep onset at that specific time each day. It matters on an early start more than on any other roster, because the bedtime an early shift requires often lands right in it. Of 135 early-start rosters this tool models, the prescribed bedtime falls inside the zone on 44 percent of them.

Better than most, with one specific catch. It is a fixed start, so your clock can settle, which beats anything rotating. It is eight hours rather than twelve. And it frees the afternoon. The catch is that a 5am start costs you roughly an hour of sleep a night against a 6am one and about two against a 7am one, and finishing at 1pm puts you home before your afternoon dip rather than driving through it, which is the good version of that problem. Whether it lands before or during the dip depends on your own clock, and that is worth checking rather than assuming.

Seven hours in bed is at the bottom of the healthy adult range and it is a defensible target rather than a good one. Two things usually make the real figure smaller. Time in bed is not time asleep, so 10pm to 5am is closer to six and a half hours slept for most people. And if 10pm is your normal bedtime, then a 5am alarm is cutting the last cycle short at the point where most of your dreaming sleep sits. It is enough to work on. It is not enough to keep doing indefinitely without the afternoons getting harder.