It’s an oft-repeated assertion in health magazines, websites and within the fitness community; not getting enough sleep can lead to weight gain. The implication is often that sleeping more will lead to effortless weight loss. Diet not working? You probably just need more time under your duvet. Better go out and buy one of those new desk-bed things for your office.

Before you do, though, let’s have a look at what the evidence actually says.

Artichokes and epidemiology

Most of what we know about this comes from observational research. Observational (or epidemiological) studies observe what people do in their everyday lives and then analyze the data collected. If we wanted to investigate whether there’s a connection between eating artichokes and performance on math exams (for example) we could just interview participants and ask them questions about how often they eat artichokes, and their test scores, and then do the statistics to see whether there’s a link.

That sort of study has the advantage of being much easier to conduct than interventional studies that require us to round people up, change their behavior, and then measure the results. To do a randomized trial of our artichoke hypothesis, we’d have to persuade half our participants to eat artichokes regularly, while the other half are put on an artichoke-free diet. We could then observe their exam results after a period of time and see whether the artichokes made a difference. (It’s a bit more complicated than this, of course, but that’s the general idea.)

Interventional studies are not just more difficult and time consuming, though; sometimes they’re impossible or unethical. Persuading people to change their sleep habits or diet for even a short period can be tricky, but when you want to look at the long-term effects over a period of months or years, it’s often completely unrealistic.

Much of our knowledge about how we’re affected by lifestyle factors such as diet, sleep. or physical activity comes from observational studies, for this very reason. And therein lies a problem.

Suppose our study shows that people who eat more artichokes do indeed score better on math exams. Does that prove that eating artichokes will improve your mathematical ability? Nope!

Correlation is not causation

Just because two things are correlated with each other, it doesn’t mean that there’s a causal link. (The website Spurious Correlations has examples of some entertaining, and unlikely, variables that correlate with each other).

It could equally be that for some unknown reason, people who are academically gifted tend also to like artichokes; or (more realistically) that both artichoke consumption and math results are connected to a “confounding” factor - such as higher socio-economic status. If that’s the case, then force-feeding your kids artichokes won’t, after all, improve their mathematical performance. It’s the hidden confounding factor that makes the difference, and has nothing at all to do with the artichokes.

So, with that in mind let’s move away from artichokes, and get back to the original question about sleep.

There have been quite a number of studies investigating the link between sleep and body weight. A systematic review in 2008 concluded that there was evidence for a modest link between short sleep duration and obesity in children. There was also some evidence in support of a link in adults, but this was more mixed. There was also an indication that people sleeping for longer than 8 hours may be more likely to be obese.

A more recent review in 2014 concluded that there was a statistically significant link between not getting enough sleep and a person’s future risk of becoming obese. Longitudinal studies that followed people over time, and compared their sleep habits at the start with their weight at a later time showed a modest, but reliable effect. Those who slept least gained the most weight.

Hang on a second, though—before you reach for your pillow, remember the artichokes. Even though two things are linked, it doesn’t mean there’s a causal relationship. It could be that there’s a common factor we don’t know about which affects both sleep and body composition. Perhaps there’s a subgroup of people who start off with better willpower; those people might report getting more sleep because they’re not staying up half the night on Facebook, and also be less likely to gain weight. Or maybe people with stressful jobs who work long hours are likely to get less sleep, but also more likely to get takeout in the evenings instead of shopping and cooking for themselves.

A 2012 paper looked at what happened when people changed how much they slept. It took adults who started out sleeping for less than six hours a night, and compared those who increased that to 7-8 hours with those who didn’t. Those who began sleeping more gained less weight than those who slept less, even though they’d started out in the same place. Although at a casual glance, this looks more like an interventional study (what happens to weight gain if we change sleep duration?), it’s important to notice that it’s not. Some of the study participants started sleeping for longer, but crucially, they made that decision for reasons of their own, not because they were told to by researchers. Although it’s a step in the direction of narrowing down a causal link, if we’re being picky (and we should), then we must still consider the possibility that there are confounding factors we haven’t identified yet.

How might it work?

There’s been some speculation as to the mechanisms by which not getting enough sleep might contribute to weight gain. Some have suggested that it’s a simple case of tiredness leading to reduced willpower, making the sleep-deprived more likely to snack on doughnuts or chocolate bars rather than sticking to their diet. This is supported by a 2011 study, which links short duration sleep to “disinhibited eating behavior,” and shows that it is the combination of the two which is most likely to lead to obesity

There’s also an increasing amount of evidence to suggest that sleep deprivation is related to impaired glucose metabolism and insulin resistance, and is a potential factor in the development of type 2 diabetes as well as obesity.

Other studies have found a link between lack of sleep, and ghrelin and leptin, the hormones that affect hunger and satiety. In other words, less sleep may lead directly to an increased appetite, especially for foods containing fat and sugar.

Conclusion

The verdict on this one is probably true. It’s often hard to prove beyond doubt that one thing causes another, but the accumulation of epidemiological evidence here, together with some plausible mechanisms of causation, are fairly convincing. Before you blame all your body composition issues on your insomnia, though, bear in mind that the effect size is probably quite modest. Kicking the Krispy Kreme habit should still be top of your agenda, I’m afraid.