The Sleep Protocol: 14 Days to Falling Asleep Faster

✓Moderate evidence

This protocol requires buying nothing. It requires fourteen days and logging two numbers, which is considerably harder than it sounds.

What do you need before starting?

You need an alarm clock, a room thermometer or a weather app if you have a thermostat, and somewhere to write. A sheet of paper on the fridge works better than an app, because you see it in the morning when the numbers need writing down.

The protocol assumes you have no diagnosed sleep apnoea or chronic insomnia. If you wake unrefreshed after eight hours in bed, snore with breathing pauses, or get morning headaches, start with a doctor. Sleep hygiene does not treat apnoea, and delaying diagnosis has a real cost.

The steps, and why in this order

The steps are in the block above, introduced one at a time on the days given. The order is deliberate: it starts with the free, best-documented changes and ends with the most inconvenient one.

Why that order? Because the first two steps, fixed wake time and morning light, act on the same mechanism: resetting the circadian clock through the suprachiasmatic nucleus. Introduced together they reinforce each other; introduced separately they work more slowly. Temperature has an immediate and independent effect, so it can arrive later without loss. Caffeine comes last because withdrawal costs a few days of feeling worse, and you do not want that in week one, when motivation decides whether you finish at all.

Minimal version: the one thing to do if you do only one

If you do one thing: a fixed wake time, seven days a week. It is the only step that resets your circadian rhythm on its own, and the only free one.

Full version

Add to the above: no screens in the bedroom, overhead lights dimmed from 9 pm, and the last large meal three hours before bed. These three have weaker evidence than the four core steps, which is why they are optional rather than required.

How do you adapt the protocol for an evening chronotype?

If you naturally fall asleep at 2 am and wake at 10, the protocol works but more slowly and with more friction. Advancing a circadian rhythm runs at roughly 15 to 30 minutes a day, so shifting by two hours means 5 to 8 days of shifting alone before the fourteen days of protocol even start counting.

Practical modification: in week one, move your wake time 15 minutes earlier each day rather than jumping straight to the target. In this variant morning light is not optional, it is the condition for success.

It is also worth asking why you are shifting. An evening chronotype is not a defect to be repaired. The problem is the mismatch between your chronotype and your obligations, and sometimes that is solvable on the obligations side.

How do you adapt the protocol for shift work?

In its basic form this protocol assumes a stable daily rhythm, so shift work requires rewriting it. A fixed wake time stops being feasible, and forcing it usually makes things worse.

What stays feasible: bedroom temperature, blackout, and cutting caffeine 8 hours before planned sleep, whenever that sleep falls. Add one thing specific to shift workers: controlling light on the way home after a night shift. Sunglasses when leaving work at 6 am are not an affectation, they stop morning light from shifting your rhythm in a direction you do not want.

I do not have a good protocol for shift work and I am not going to pretend otherwise. The literature is much weaker here and the trade-offs are larger.

What do you do when travelling across time zones?

For flights across up to three time zones, hold the destination schedule from day one and get outside for morning light. For a bigger difference, start shifting your wake time by an hour a day two or three days before departure.

During the trip itself, suspend the protocol and resume it afterwards. The numbers from a week with a flight will not be comparable to the rest anyway, so it is better to mark them as a gap than to pretend they measure something.

How do you do the light step in winter?

From November to February the morning light step gets hard above roughly 50 degrees latitude, because it is still dark at 7 am. Pre-dawn light intensity is not enough to shift the circadian rhythm, so stepping outside at that hour achieves little.

Two workable options. First: move the exposure to a break around 9 to 10 am, once it is properly light. Later light works less well than light right after waking, but it works. Second: a 10,000 lux lamp used for 20 to 30 minutes over breakfast, at the distance the manufacturer specifies.

Lamps of this type cost roughly £60 to £150. This is the only purchase in the protocol that is justified, and only in the winter variant for people who leave for work before sunrise.

What should you not change during the protocol?

Do not change your training, your diet or your dinner time if you can avoid it for fourteen days. Every extra variable lowers the chance that you draw any conclusion from those two weeks.

Do not start the protocol in a week with a trip, a wedding or a work deadline that ends in four hours of sleep. Better to push the start by a week than to burn fourteen days on data you will discard anyway.

How do you log the data so it is usable?

A sheet on the fridge with two columns is enough: date, minutes to fall asleep, number of awakenings. Write it in the morning, not the evening, because in the evening you will be guessing about the night before.

After fourteen days, take the median of the first three days and the median of the last three. Median, not mean: one night after a wedding should not decide the conclusion of the whole protocol.

If you also track with a device, treat its data as a second set rather than the deciding one. Trackers are decent at sleep-wake detection and clearly weaker at stage classification. I go into that in more detail in sleep and longevity.

When should you expect results?

The effect size quoted is the median from trials of the individual components, not the result of a trial on this specific combination. Nobody has tested these four together in a controlled experiment. Treat the numbers as an order of magnitude.

Why does the protocol most often fail?

First: the weekend. Two days of shifted waking is enough to cancel five days of regularity. This is by far the most common failure.

Second: no measurement. Without the two numbers you are relying on memory, and memory for sleep is notoriously bad. People systematically overestimate sleep onset on bad nights and underestimate it on good ones.

Third: introducing everything at once on day one. There is usually no protocol left by day five.

Fourth, the rarest and the most important: the protocol is aimed at the wrong problem. Sleep hygiene will not fix apnoea, iron deficiency or hyperthyroidism.

What next, if it worked

Hold the four steps and only then consider adding anything. The mechanisms and mortality data sit in sleep and longevity, and the breakdown of which hygiene rules have evidence is in sleep hygiene.

What next, if it did not

If the two logged numbers have not moved after fourteen days, the protocol did not work for you, and that is also a result. The next step is a consultation and diagnosis, not another supplement. Take your log with you: fourteen days of latency and awakening data is more than most patients bring to an appointment.