A lot of the literature about the impact of sleep duration on physical performance is plagued by a few common issues. First, controlled studies on the effects of sleep duration tend to use interventions lasting a single night – in other words, they generally just look at how one night of reduced sleep affects performance. Second, studies with longer durations tend to be observational in nature. In other words, they examine differences between individuals who typically sleep more versus individuals who sleep less, but we can’t be sure that sleep-related associations are caused by sleep differences – people who sleep less may simply systematically differ from people who sleep more in a variety of ways. Third, much of the sleep duration research focuses on the negative impacts of sleeping less, but very little research investigates the potential benefits of sleeping more than normal. A systematic review (which we covered in MASS) has suggested that sleep extension might be the most effective sleep-related intervention for athletes, but the body of sleep extension literature is still quite small.
With that in mind, a 2019 study by Roberts and colleagues offers valuable insight into the effects of sleep restriction and extension on performance. Research Spotlights generally cover studies that are hot off the press, but I’m covering a 2019 study for a good reason. I intended to cover a new study from the same research group, but I noticed that it presented secondary analyses of an already-published experiment. The original study (which I’m reviewing here) was more focused on performance outcomes, so I decided that it made more sense to center the discussion on the 2019 paper.
In this study, nine endurance-trained athletes (average VO2max = 63 ± 6 ml/kg/min) completed a crossover study consisting of three conditions – a normal sleep condition, a sleep restriction condition, and a sleep extension condition. In all conditions, subjects completed a cycling time trial every day for four days. The individualized workload completed during the time trials was the equivalent of cycling for one hour at each subject’s anaerobic threshold. Before each time trial, subjects completed a psychomotor vigilance task (which assessed reaction times to a visual stimulus) and filled out a Profile of Mood States questionnaire.
Before the start of the intervention, subjects’ habitual sleep habits were monitored for four nights to establish a baseline. During the normal sleep condition, subjects were instructed to spend a typical amount of time in bed. During the sleep restriction and sleep extension conditions, subjects were instructed to decrease or increase their time in bed by 30%. Subjects wore accelerometers to verify that they were spending the appropriate amount of time in bed during each condition. The researchers also used accelerometry data to calculate total sleep time (i.e. how long the subjects actually slept, independent of how long they spent in bed) and sleep efficiency (the proportion of time in bed spent sleeping). Subjects also reported their subjective sleep quality each day using a 5-point Likert scale (1 = very good sleep and 5 = very poor sleep). An overview of the study protocol can be seen in Figure 1.
As intended, subjects slept the most during the sleep extension condition (averaging 8.2-8.6 hours per night), followed by the normal sleep condition (averaging about 7 hours per night), followed by the sleep restriction condition (averaging 4.7-4.9 hours per night). Furthermore, while sleep efficiency was a bit lower at some time points in the sleep extension condition (indicating that subjects spent a slightly higher proportion of their night awake in bed), differences between conditions weren’t particularly large. Reported sleep quality also didn’t substantially differ between conditions.
Time trial performance didn’t change much across the four consecutive testing days during the normal sleep and sleep extension conditions. However, time trial performance got progressively worse in the sleep restriction condition. During the first day of testing in the sleep restriction condition, subjects completed the time trial in 57.6 minutes on average; by the fourth day of testing (i.e. following three nights of reduced sleep), it took them an average of 62.0 minutes to complete the same workload. Furthermore, by the fourth day of testing, subjects completed the time trial significantly faster in the sleep extension condition than either the normal sleep condition or the sleep restriction condition.
Data related to mood disturbances and psychomotor vigilance both suggest that subjects were becoming progressively fatigued over time in the normal sleep and sleep restriction conditions, even though time trial performance wasn’t negatively affected in the normal sleep condition. Total mood disturbance tended to increase (though the effect wasn’t statistically significant in the normal sleep condition), vigor tended to decrease (again, not statistically significant in the normal sleep condition), reported fatigue significantly increased in both conditions, and mean response time during the psychomotor vigilance task significantly increased in both conditions. Conversely, in the sleep extension condition, total mood disturbance and vigor didn’t meaningfully change, the increase in fatigue tended to be smaller (2 points, versus 5 points in the normal sleep condition and 10 points in the sleep restriction condition), and mean response time during the psychomotor vigilance task significantly decreased (which is a positive outcome).
Overall, these results paint a positive picture for sleep extension. Most of the very positive results in favor of sleep extension come from studies on high-level collegiate athletes. I’ve been somewhat concerned that those results wouldn’t generalize to other populations, for a couple of reasons. First, I was concerned that sleep extension might only have a positive effect in athletes who were actively engaged in super strenuous training. I’m certainly not saying that an hour-long time trial is easy, but I imagine it’s a much smaller workload than Division I collegiate swimmers are dealing with. Second, I’ve been somewhat concerned that the effects of sleep extension might only be present in athletes who are quite young. The subjects in the present study weren’t old by any means (30 ± 6 years old), but sleep duration and quality tend to decrease throughout the lifespan, so I wondered if attempts at sleep extension for people who weren’t quite young (i.e., collegiate athletes or younger) would simply result in steep declines in sleep efficiency, negligible changes in total sleep time, and no net ergogenic effect. So, this study reassures me that the benefits of sleep extension for athletes are probably fairly generalizable. With that being said, the performance-related effects in the present study were smaller than those observed in prior sleep extension studies. More research is needed to better understand who can benefit from sleep extension (at least for performance-related outcomes), and the degree to which sleep extension is likely to improve their performance.
It’s also worth noting that the largest differences between the normal sleep and sleep extension conditions were related to psychomotor vigilance and mood states. In other words, sleep extension may have only had a small positive effect on maintaining (or improving) performance, but it had larger positive effects on reaction times (and therefore general mental acuity, I suspect) and how the athletes generally felt. For my money, as someone who’s not a professional athlete, those benefits would probably be the ones I cared about the most in day-to-day life. If you spend a bit more time in bed, you’ll probably perform a bit better, but you’ll probably feel noticeably better, even if you’re already getting the recommended seven hours of sleep per night.
This Research Spotlight was originally published in MASS Research Review. Subscribe to MASS to get a monthly publication with breakdowns of recent exercise and nutrition studies.
Credit: Graphics by Kat Whitfield.