Most Advice About Your Sleep Is Ridiculous

The reason that most sleep tips are so exhausting.

Pictured: An awake person. I hope. It would be REALLY weird if someone took stock photos while their subject was actually asleep. Photo by Isabella Fischer on Unsplash

As any parent of young children or person with insomnia will tell you, sleep is vital. Without sleep, people simply cannot function. It is remarkable how quickly you can go from a normal, happy human being to a jittery mess with just a few wakeful nights where you didn’t get your usual 8 hours.

And as everyone knows, sleep is also vital for your long-term health. There is a new headline every week pointing this out. In the most recent news, sleep has been linked to a staggering 172 diseases, with a range of sleep traits responsible. The evidence seems overwhelming that getting better sleep can lead you to good health.

The problem with all the sleep advice - whether it’s from this new study or just the general background noise - is that it ignores how sleep works. It’s easy to say that getting more sleep might be beneficial for your health, but the reality is that sleep is interwoven with almost every other aspect of your life, and getting more/less of it isn’t nearly as easy as online gurus pretend.

Pictured: Not always easy. Photo by Kate Stone Matheson on Unsplash

The most recent study is a perfect example of this. Researchers took the UK Biobank, which is a large cohort study with a huge amount of data on individuals. One aspect of this database is what’s called actigraphy data, which is basically information from a smartwatch or fitness tracking device that can be used to estimate when people fall asleep and wake up.

They then looked at 6 different sleep traits. These are ways of measuring either how much sleep people get or how good that sleep is - for example sleep latency or sleep onset timing, which is how long it takes people to go to sleep after they lie down in bed.

They compared these 6 traits with every disease category that was included in the Biobank dataset. This amounted to more than 500 separate disease codes. Using a fairly complex statistical model, they found that 172 of these codes were associated with at least one of the 6 measured sleep traits, with a total of 349 statistically significant associations.

These results come with the usual caveats that observational data has. The authors had some ability to control for confounding, but it’s hard to exclude the possibility that there are issues that the authors simply could not correct for in their analysis. Even with actigraphy, measuring sleep is a very difficult thing to do, and most smartwatches and similar are very inaccurate when it comes to sleep timing.

It’s also hard to know cause and effect in situations like this. The authors say that it’s likely that sleep is causing the health issues that they measured, such as Parkinson disease, but it’s equally likely that people’s Parkinson’s makes their sleep worse and that is what they have measured in the study itself. This is known as reverse causality, and it’s very hard to exclude in a study like this.

But by far the biggest issue I have with this study is the same as every similar study on sleep that I’ve seen - the results are totally disconnected from how people actually sleep.

For example, take Parkinson’s. The results suggest that people who have high relative amplitude are less likely to get Parkinson’s disease. Relative amplitude is a measure of how much you move about when you sleep compared to how much you move during the day, which combines exercise and restful sleep into one metric. It’s unclear if improving your relative amplitude will definitely reduce your risk of Parkinson’s, but even if it does it’s very hard to actually implement this. Exercising more should increase the metric, unless it makes your sleep more disturbed. If your sleep is disturbed for an external reason - say, you have a newborn or live under a flight path - then there’s not much you can do to change this metric anyway.

And this is true for almost every metric that we can measure about sleep. Taking a long time to fall asleep is associated with bad health outcomes but it’s also a really difficult thing to change. The main reason that I often take a while to go to sleep is stress. In 2020 and 2021, when I was an epidemiologist working during the COVID-19 pandemic, it often took me much longer to get to sleep than it does today. That had nothing to do with my own actions and everything to do with external stressors that I could not control. Even if it did increase my risk of chronic disease there wasn’t much I could do to change that.

The problem is that sleep is one of those aspects of health that is completely intertwined with everything else in our lives. Bad sleep may cause disease, but disease definitely causes bad sleep. Pretty much every negative thing that happens in day to day life can make your sleep disturbed. Losing hours on your shifts. Rent going up. Your teenager coming home at 2am smelling of booze and regret.

If you want to sleep a bit better, the basic recommendations have not changed in decades. Unless you have severe sleep apnea, the advice basically boils down to having a dark, quiet room and not using your bed for things other than amorous activities and sleep.

Beyond that, there really isn’t that much you can do to improve your sleep. Being rich helps a lot, as does being fit and having no chronic diseases. Being young. Having no children living in your house. Not smoking, drinking alcohol, or doing drugs. You’ll notice many of these are things that are outside of individual control, which is precisely the point.

You should already follow basic sleep hygiene recommendations, especially the one about not doomscrolling on your phone in bed. It’s just basic common sense. It’s also a good idea to try and get as much exercise as you can, because exercise helps with literally every disease that we know about. Similarly, smoking is bad for you, alcohol isn’t great, and most illegal pharmaceuticals are illegal for a reason.

But don’t worry too much about headlines linking sleep to various diseases. It’s not clear which way the relationship runs, and even if bad sleep is causing ill health it’s pretty tough to do anything about it anyway.