Sleep hygiene: which rules from the list have actually been tested

The list of “sleep hygiene rules” travels from article to article almost unchanged: a dark room, coolness, no coffee after lunch, phone an hour before bed. The list looks homogeneous but is compiled from things of different weight: some have been tested in randomized trials, some only in surveys, and some have not been tested at all.

Sleep and food are measurably linked

The most verifiable claim in this topic is not about sleep quality, but about its duration. In a University of Chicago trial, 80 overweight adults who usually slept less than 6.5 hours were divided into two groups. One group received a single consultation on sleep hygiene and was asked to adjust their bedtime to allow for 8.5 hours in bed. The second group continued to sleep as usual. Diet and physical activity were not changed for anyone.

After two weeks, the first group slept an average of 1.2 hours longer and consumed 270 kilocalories less per day than the control group. Energy expenditure did not differ between the groups, so the difference resulted in weight loss. Energy intake was measured not by diaries, but by the doubly labeled water method — that is, by actual expenditure and changes in reserves, rather than by what a person recorded. The trial was published in JAMA Internal Medicine.

The mechanism was described earlier. In a 2004 study, sleep restriction in healthy men lowered leptin, increased ghrelin, and intensified hunger, especially for high-calorie foods — Annals of Internal Medicine. More on this in the article how sleep deprivation affects weight and appetite.

What is known about the list of rules itself

The evidence here is weaker. A review of empirical data on sleep hygiene examined the components of the list one by one — caffeine, alcohol, physical activity, stress, noise, routine — and concluded that as a set of recommendations for the healthy population, the list is poorly studied: some points are based on laboratory research, some on associations in surveys, and the effectiveness of the list as a whole as an intervention has almost never been measured. Review in Sleep Medicine Reviews.

This does not mean that the rules are useless. It means that "sleep hygiene" is not a single intervention, but a list of heterogeneous items, each of which has its own evidence base.

Glasses against blue light

A separate item with its own market. A systematic review with meta-analysis found only three double-blind crossover trials with actigraphy — 49 participants in total. Sleep onset latency decreased by 4.9 minutes, total sleep time increased by 8.8 minutes, and sleep efficiency and wake after sleep onset did not change. None of the results were statistically significant. The authors' conclusion: glasses may provide a small improvement, but the available data do not confirm significant effects. Review in Frontiers in Neurology.

The complexity here is also methodological. It is impossible to blind a person who has been given colored glasses, and objective actigraphy and subjective sleep assessment regularly diverge.

What works for insomnia

When it comes not to "I want to sleep better," but to chronic insomnia, sleep hygiene ceases to be the main focus. A meta-analysis of 20 trials showed that cognitive behavioral therapy for insomnia improves sleep onset latency, wake after sleep onset, and sleep efficiency, and the effect persists after the course ends — Annals of Internal Medicine.

A later component network meta-analysis of 241 trials with 31,452 participants examined which specific parts of the therapy yield results. Cognitive restructuring, "third-wave" components, time-in-bed restriction, and stimulus control were found to be significant. Sleep hygiene education was not significant — an incremental odds ratio of 1.01 with a confidence interval from 0.77 to 1.32, which is indistinguishable from no effect. Study in JAMA Psychiatry.

The difference is fundamental. Sleep hygiene is the background, which in trials often serves as a control condition, that is, what treatment is compared against.

Where the numbers on deep sleep come from

Many "sleep protocols" are built around numbers from a ring or bracelet: how much deep sleep there was, how much REM sleep. A multicenter validation of eleven consumer trackers compared them with polysomnography in a laboratory. Agreement on sleep stages turned out to be very different: the best macro F1 score was 0.69, the worst was 0.26. Study in JMIR mHealth and uHealth.

Devices usually estimate total sleep time significantly better than the breakdown into stages. This is important for interpretation: "low deep sleep" on the screen is an algorithm's estimate, not a measurement. More on the same difference between measurement and estimation in the article about fitness trackers.

Three layers in one list

It is convenient to read the sleep hygiene list not as a single whole, but as three different layers. Sleep duration has been verified in a trial with objective energy measurement. Individual habits from the list have been studied in different ways and mainly in the short term. Devices provide plausible but inaccurate numbers on sleep structure.

This method of sorting — by what was measured, on how many people, and for how long — is common to the entire section. It is described in the article about biohacking in general.

Count it from a photo

Frequently asked questions

Is it true that lack of sleep leads to overeating?

In a randomized trial of 80 overweight adults, extending sleep by approximately 1.2 hours was associated with a reduction in energy intake of 270 kilocalories per day compared to the control group. Intake was measured using the doubly labeled water method, rather than food diaries.

Do blue light blocking glasses help?

A meta-analysis of three double-blind crossover trials with actigraphy and 49 participants reported a reduction in sleep onset latency of 4.9 minutes and an increase in total sleep time of 8.8 minutes, both of which were statistically insignificant. The authors' conclusion: available data do not confirm significant effects.

Is sleep hygiene sufficient for insomnia?

In trials, sleep hygiene education often serves as a control condition. A component network meta-analysis of 241 trials showed that the significant parts of therapy are cognitive restructuring, "third-wave" components, sleep restriction, and stimulus control, whereas sleep hygiene education yielded an incremental odds ratio of 1.01, which is indistinguishable from no effect.

How accurate is a sleep tracker?

A multicenter validation of eleven devices against polysomnography showed agreement on sleep stages ranging from 0.26 to 0.69 on the macro F1 score. Total sleep time is estimated more accurately than the breakdown into deep and REM sleep.

Which "sleep rules" are best supported by evidence?

Sleep duration. Its increase was measured in a randomized trial with objective assessment of energy balance. The other items on the list have been studied less thoroughly and mostly in isolation.

Read next

• How sleep loss affects appetite and weight — Sleep usually turns up in the last paragraph of a weight-loss article, phrased vaguely enough to be safe.
• Meal timing and the body clock: why breakfast and midnight are not the same meal — The same plate is not handled the same way at eight in the morning and at midnight.
• Biohacking: separating what has been tested from what is being sold — There is no definition of biohacking that two people would both sign.

This article is for general information. It is not medical advice, a diagnosis, or a prescription for treatment or a diet, and it does not replace a consultation with your doctor. If you have a health condition, are pregnant, take medication, or follow a diet prescribed to you, decisions about food belong with your doctor.

Figures from regulations, guidelines and studies are given as they stood when this article was prepared and may since have changed; check them against the primary sources. This article is not advertising, an offer, or individual advice, and neither the author nor the site owner is responsible for decisions taken on the basis of it.