08 February 2024
The existence of smartwatch-type movement measuring devices and the recently established British repository of medical, biological and lifestyle data are making it possible to start assessing the impact of sleep patterns on people's risk of premature mortality.
For some time now, the IARC (International Agency for Research on Cancer), the world’s body that defines carcinogens, has considered night shift work as a potential carcinogen.
But what about the general population of adults that are not night-shift workers and that sleep too little or go to bed too late – or have “poor quality" sleep, waking up repeatedly during the night? Are they also at an increased risk of cardiovascular disease or cancer and are therefore more likely to die prematurely than people whose sleep patterns are considered healthier?
In an innovative approach, a study published some weeks ago in the scientific journal Sleep examined this question based on quantitative data collected using wrist-worn devices from a large sample of adults.
The study's first author is Pedro Saint-Maurice. This Portuguese researcher has spent the last seven years working at the National Cancer Institute of the National Institutes of Health (NIH, the biomedical research agency of the US Federal Government).
"The mission of NIH scientists is to carry out innovative biomedical research to inform the US population while promoting scientific excellence," explains Pedro Saint-Maurice. "I do research on the epidemiology of physical activity, cancer risk, and the risk for other chronic diseases as a function of the duration, quality and timing of physical activity. I'm also trying to understand the impact of exercise on people diagnosed with cancer. Recently, I've also started doing research on sleep patterns, another very important behaviour for a healthy lifestyle."
In October 2023, Pedro Saint-Maurice joined the Champalimaud Foundation's research team, with the aim of researching and defining recommendations for physical activity and sleep for cancer patients.
The aim of the study now published in SLEEP was to understand how sleep can influence health. Studies on the sleep/health association have been carried out for many years, but they have relied on questionnaires completed by participants about their own sleep patterns. Now, self-reporting your sleep isn't easy: science has shown that volunteers are likely to report the time they spend in bed, which doesn't necessarily correspond to their actual sleep time.
But now, thanks to smartwatch-type wearable devices for measuring body movements, it has become possible to start assessing the impact of sleep patterns on health in a much more objective fashion.
"In the world of epidemiology, one of the questions that arises is how sleep patterns can influence health and the onset of diseases in the long term," explains Pedro Saint-Maurice. "We use mortality data because mortality is an indicator of illness in general and is therefore very useful for studying these kinds of questions."
Another substantial limitation is that previous studies, despite being very valuable, have focused on small sample sizes (a few thousand participants). "These studies didn't allow us to analyse, for example, whether the associations between sleep and mortality observed were equally valid for men, women, older, and younger adults" notes Pedro Saint-Maurice.
Now that too has changed with the recently established UK Biobank, a giant repository of medical, biological and lifestyle data that was created for research purposes.
The UK Biobank currently contains data on half a million adults, 100,000 of whom volunteered to wear a smartwatch-like device on their wrist for seven days to allow measurements of their sleep characteristics. These volunteers also agreed to be followed for several years, through a linkage between the UK Biobank and national health registries in the UK. A wealth of information has thus become available to researchers from all over the world.
"The existence of the UK Biobank radically changed the situation," emphasises Pedro Saint-Maurice. "For the first time, it became possible to study epidemiological questions in a much more precise way. The UK Biobank is in fact the largest medical database containing wearable data and it allows us to answer lifestyle questions, including questions related to sleep. We've been able to apply algorithms developed by the scientific community to identify when a person is asleep or awake, or if they move around a lot at night in their sleep – and even at what time they fall asleep."
The studies of sleep/mortality associations had another important limitation: these studies looked at the relationship between sleep and mortality only in terms of the duration of sleep, but not in terms of the timing of sleep – in other words, the time at which someone goes to bed and falls asleep. What's more, the quality of sleep itself (which is measured by its continuity) hadn't been much investigated either.
"What we did [in the Sleep study] was to try to understand in more depth the impact of these three sleep characteristics," says Pedro Saint-Maurice. "Our hypothesis was that shorter sleep duration, intermittent sleep (which is related to sleep quality) and later timing were associated with an increased risk of death. It was this hypothesis that we decided to test.
According to the authors, the data they used in their analysis included 88,282 adults, aged between 40 and 69, who wore a wrist device for seven days (as already mentioned) between 2013 and 2015. The wearable included a triaxial accelerometer and was able to capture and analyse movement in three dimensions.
The data from this so-called "actigraphy" (a non-invasive method for monitoring human rest/activity cycles) was processed to generate estimates of the duration and other characteristics of sleep – in particular, the frequency with which the person was awake at night for five minutes or more, and also the midpoint between sleep onset and waking up.
These data were crossed with mortality data during the participants' follow-up (from baseline through October 31st of 2021) and analysed with statistical models to quantify the associations between sleep and mortality. During those seven years, almost 3,000 deaths occurred (including 1,700 deaths from cancer and 586 deaths from cardiovascular disease).
"Our results show that less than six hours of sleep per night is significantly associated with a higher risk of mortality - from all causes, cancer and cardiovascular disease," says Pedro Saint-Maurice. More precisely, sleeping only five hours a night instead of seven is associated with a 29 per cent higher risk of all-cause mortality.
Timing also has an impact: people who went to bed very late or woke up very early had a 20 per cent higher risk of all-cause mortality when compared to those who went to sleep and woke up at less extreme hours.
"This is very interesting," adds the researcher, "because we know that late sleep times present an increased risk for a series of metabolic diseases and various types of cancer, prostate and breast being the two most frequently mentioned."
The main hypothesis regarding the influence of sleep timing has been, in particular, the idea that late bedtimes may contribute to the deregulation of melatonin cycles. Melatonin is the hormone that "prepares" us for sleep and whose levels increase at nightfall. "Melatonin is a tumour suppressor and hence the hypothesis that increased risk of cancer may be due to the deregulation of this hormone," explains Pedro Saint-Maurice.
As for waking up too early in the morning, this could interfere with the cycles of other hormones – such as cortisol (which also has well-defined 24-hour, or "circadian", cycles). "In general, altered sleep schedules can interfere with the patterns of these hormones and have an impact on various metabolic processes," notes Pedro Saint-Maurice. "Most importantly, sleep schedules can interfere with the body's alignment with daylight, because we know that daylight is the main entrainer of our circadian cycles." Extreme sleep timings disturb our biological rhythms by misaligning our physiology with our daylight exposure patterns.
"This interpretation of our results is still very exploratory," says the researcher. "The impact of sleep duration is the most well-established. But on the question of timing, and in particular of waking up earlier or later, I think we still need to understand how the relationship with daylight might increase, in particular, the risk of cancer," admits Pedro Saint-Maurice.
Finally, with regard to indicators of sleep quality, such as the frequency of awakenings during the night, in this study we did not find any associations with an increased risk of mortality – contrary to what a number of previous studies have concluded. "We found no association," says Pedro Saint-Maurice, "between this type of indicator – more movements during the night, or periods when the person was awake – and the risk of death," he emphasises.
What particular finding of the study could immediately be considered as a practical recommendation for a healthy sleep? "Make sure you sleep between seven and nine hours a night, as the National Sleep Foundation currently recommends," replies Pedro Saint-Maurice. "We have confirmed that this is extremely important for health."
As a second point, the researcher recommends paying attention to the timing of sleep – the time of the day when sleep happens. "It seems to be an important factor in maintaining a healthy life. It hasn’t been confirmed yet, we don't know what the ideal timing would be. It will take more work to determine this. We do know that there is an association and that timing is important, but we haven't yet been able to be more specific."
To conclude, Pedro Saint-Maurice points out that cancer patients can have their sleep patterns perturbed due to the treatments they undergo. "For example, chemotherapy can alter sleep quality," he says. "And we need to think about how we can help by designing behavioural interventions that can improve sleep patterns among cancer patients."
Scientific article here.
