Sleep Part I: Long-Term Health Benefits of Sleeping 7 Hours per Night

After posting about the natural experiment that showed everybody is more productive if everybody sleeps more together, I read the book Why We Sleep by Matthew Walker, a professor at UC Berkeley. It’s an exciting read and I decided to track down the original studies behind the parts of the book about living a longer and healthier life. I’m planning three posts: this one on the long-term health benefits of sleeping enough, one on how to sleep better, and a third on what Walker describes as “embryonic” research into what I’ll call super sleeping.

1. All-cause mortality – I’ll start with the overall effect on lifespan.
   1. Capuccio 2010 conducts a meta-analysis of prospective trials and finds that short-sleepers have 12% higher mortality. This result was not significantly different for people above or below 60. It was also not significantly different for studies that defined short-sleeping as below 7, 6, or even 5 hours.
   2. The same meta-analysis found that long-sleepers have a more dramatic 30% higher mortality. Sub-group analysis makes a bigger difference here. Long-sleeping is a stronger indicator of mortality risk in people 60 and over than people under 60. Also, the effect size grows to a 54% when long-sleeping is defined as over 10 hours, compared to a 20% increase for studies that define it as more than 8 hours. Since the latter studies will of course consider both someone who sleeps 8 hours and 1 minute and someone who sleeps 11 hours as long sleepers, this may suggest that sleeping 8-10 hours is not a problem at all. Walker says there’s a strong consensus that long-sleep is just a warning sign of ill health. There is no mechanism by which sleeping too much is bad for you. When you have slept enough, you wake up. (Sleeping is of course not the same as lying in bed awake.)
   3. Kripke 2002 analyzes data from the Cancer Prevention Study II. The survey covered 1.1 million Americans. For both men and women, after adjusting for many covariates from the survey, people who average 7 hours of sleep (which covered 6.5 to 7.5) have the lowest mortality. Sleeping 6 hours was associated with an 8% (7%) increase in mortality in men (women).
2. CVD
   1. Capuccio 2011 conducts a meta-analysis of prospective trials and finds that short-sleepers are at 48% greater risk of coronary heart disease and a 15% greater risk of stroke, but no increase in the total risk of cardiovascular disease (I had to Google this myself to check that CHD is a subset of CVD).
   2. Long-sleepers have 38% higher risk of CHD, a gigantic 65% increased risk of stroke, and a 41% increase in total CVD.
3. Cancer – the relationship between sleep and cancer seems to be more ambiguous than that between sleep and CVD.
   1. Null results
      1. Data from the Nurses’ Health Study showed no connection between sleep duration and breast cancer.
      2. Survey data from Finland show no connection between short-sleeping and risk of breast cancer.
      3. The Singapore Chinese Health Study showed no connection between short-sleeping and risk of breast cancer.
   2. Positive results (where positive means bad for your health)
      1. The European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study found that sleeping less than 6 hours per night is associated with increased risk of cancer. Above 6 hours, there was no difference by hour.
      2. The Ohsaki Cohort Study of women found that sleeping 6 hours or less is associated with greater risk of breast cancer, but found no difference between those who sleep 7 or 8 hours.
   3. In between
      1. The Ohsaki Cohort Study of men found that sleeping 6 hours or less may be associated with greater risk of prostate cancer, but it’s not statistically significant. When they include their finding that men who sleep over 9 hours have a greatly decreased risk of prostate cancer they find strong evidence of a linear trend for great sleep decreasing prostate cancer risk. It’s not clear what this means for people who naturally wake up in under 9 hours.
4. Dementia – Walker says “a lack of sleep is fast becoming recognized as a key lifestyle factor determining whether or not you will develop Alzheimer’s disease.” In the opposite causal direction, Walker describes strong evidence that dementia disrupts sleep and that sleep disruption often precedes diagnosis of dementia by years. That means we should be very careful about interpreting observational correlations between poor sleep and later appearance of dementia as evidence that short sleep duration causes dementia. Let’s look at the six studies Walker cites in two gruops:
   1. Prospective trials: Walker claims that “getting too little sleep across the adult life span will significantly raise your risk of developing Alzheimer’s disease” and then lists references for four prospective trials. I don’t think his references back up his claim. All four of them are to studies in people who average over 80 years old and live in senior living facilities. It’s true that the ones who slept worse showed signs of dementia sooner, but this is exactly the group of people Walker warned us about in the immediately preceding paragraph: the poor sleep was likely an early symptom of dementia and not a cause. Plus, I’d hardly describe a study of people in their 80’s in senior living facilities as “across the adult lifespan.”
   2. Randomized trials: Walker claims that “by improving someone’s sleep, we should be able to reduce their risk of developing Alzheimer’s disease – or at least delay its onset. Tentative support has emerged from clinical studies in which middle- and older-age adults have had their sleep disorders successfully treated.” But, again, the references he gives for that claim don’t match his summary. He gives two, and both are studies of people who already have Alzheimer’s. These studies prove that treating sleep disorders in people with Alzheimer’s will improve their memory. But a large portion of the book, which I’ll cover in another post, demonstrates that short sleeping is bad for memory in healthy adults. These experiments demonstrate that sleep disruption can be a warning sign of Alzheimer’s and can exacerbate the symptoms of Alzheimer’s. They do not provide any suggestion that short sleep causes buildup of amyloid plaques in the brain. Such research may exist, but Walker never cites it.

The punchline is that sleeping at least 7 hours each night is good for your health. If you feel like you need more sleep than that, you should take it. Although it probably won’t protect you from Alzheimer’s, it’ll make you live longer, it’ll be good for your heart, and it might help your body fight off cancer.