It may be difficult to believe, but there are no definitive answers to either of these questions. Neither normal sleep nor how long one should sleep has ever been formally defined. Instead, in sleep medicine we work mostly with what is not normal in terms of the quality and quantity of sleep, after which through diagnostics and therapeutics we attempt to transform the “not normal” conditions into something less “not normal.”
If you suffer from sleep apnea, we try to reduce most of your breathing events, but we often do not succeed in eliminating all abnormal breathing. Thus, although the definition of less than 5 breathing events indicates no sleep breathing disorder, this data point does not reveal whether the breathing would be considered normal, compared to people who have no breathing events whatsoever. Would you think that individuals with perfectly normal sleep breathing are actually experiencing “super-normal” breathing, because so few people demonstrate no breathing events at all? Reminder, all breathing events must be scored, including the respiratory effort-related arousals (RERAs); and a further reminder, there are individuals whose RERAs are ridiculously subtle such that most sleep centers do not acknowledge let alone score them. Yet, we have treated many of these patients with PAP therapy to “normalize” their breathing as best we could, and most patients with these sub-threshold sleep breathing disorders reported dramatic improvements in sleepiness or insomnia or both.
So, what is normal breathing during sleep? I don’t know, but I do approach the problem by examining what normal breathing looks like on someone receiving an optimal response to PAP therapy. The shape of the airflow curve appears rounded on both inspiration and expiration, and this appearance to our eyes appears identical to the few “normal” sleepers we have been able to coax into the sleep lab to undergo sleep studies. In sum, a nearly perfect to totally perfect PAP response looks the same as what a normal sleeper’s breathing looks like without PAP therapy.
Such findings should lead sleep professionals to question whether or not the current system of measuring breathing is adequate, especially in the context of understanding and appropriately treating sub-threshold cases of OSA/UARS. In my opinion, the current system is inadequate to this task, because currently, many patients are informed they do not suffer from a sleep breathing disorder, because they do not meet certain criteria. And, because these criteria are insurance-driven and not evidence-based, we are led to miscommunicate with our patients on whether or not their sleep breathing is abnormal or diagnostic of a disorder or both. As you can imagine, this conundrum does not instill confidence in describing normal breathing and therefore further confounds efforts in describing normal sleep.
Which brings us to sleep architecture, or more precisely, the sleep fragmentation effects on sleep architecture that occur during sleep breathing and sleep movement disorders. When someone suffers from OSA/UARS or suffers from an independent, non-respiratory-related, periodic limb movement disorder, arousal activity triggered by these disrupters to sleep alter sleep architecture. Some changes include transitioning too rapidly between difference stages of sleep, no longer maintaining a healthy amount of time in deeper or more restorative stages of sleep, spending too much time in lighter stages of sleep, experiencing repeated episodes of frank awakenings that may require excess time to fall back asleep , and last, although somewhat speculatively, fracturing of REM sleep in ways that interfere with the typical contiguous pattern of consolidated REM periods. Instead, REM sleep appears as multiple discrete episodes lasting only a few minutes at a time, and then the sleep cycle transitions into wakefulness or light sleep before returning to REM.
To sum up the above points: we cannot really know what truly normal sleep is until we establish definitions for truly normal breathing (or the absence of other pathophysiological signs, e.g. leg jerks) and establish definitions for sleep architecture. The latter point is especially problematic among elderly patients where it is often claimed they experience reductions in deep or REM sleep, and yet based on the evidence it seems entirely plausible that “normal” elderly patients were tested who may in reality have suffered from undiagnosed UARS—a condition that like OSA seems to kick patients out of deep or REM sleep.
All of the above should make clear that we cannot definitively describe all the facets of normal sleep, and therefore how can we describe the “normal” duration or quantity of sleep? In short we cannot. The most obvious example of this problem is the insomniac or sleep apnea patient who complains of feeling awful no matter how many hours of sleep he or she gains each night. In both cases, the duration might seem abnormal (say 5 for the insomniac and 9 for the sleep apneic), but the real problem is that these numbers may be artificially inflated by not taking into account the quality of those periods of slumber. If sleep fragmentation were pervasive in either case, then the true hours of sleep for the insomniac might be as low as 2 or 4, and the true hours of sleep for the sleep apneic might be as low as 5 to 7. Thus, sleep fragmentation so disrupts normal sleep periods among those with sleep disorders, it is a given that their actual duration of sleep is far less than what they can perceive.
With so little objective information at the disposable of the patient, it is nearly an irrational process to attempt to determine the right number of hours of sleep for someone. There is no way to count what hours were achieved, because sleep quality disruption prevents that determination, therefore how could measure what hours are needed? Some people will suggest that when they sleep 8 to 10 hours it seems better than when they sleep 4 to 7 hours, and under certain circumstances they may be correct…in the short-term. However, if there goal were to determine the best number of hours of sleep on a typical night, then learning to achieve a consistently higher quality of slumber is the most clarifying step to sort out the number of hours.
A great example of this phenomenon over the long-term is the individual who achieves an excellent response to PAP therapy for the treatment of OSA. Initially, she may report sleeping more hours than prior to treatment as she seems to be catching up on sleep debt. Then, quite remarkably, she could report 6 to 12 months later, perhaps sooner in some folks, she no longer needs as much sleep. Some of these patients actually develop insomnia, because they go by their old numbers, thinking they should still shoot for 7.5 to 8.5 hours per night. Yet, their bodies are telling them they only need 6.5 to 7.0 hours per night now that the quality of sleep is so much better.
Unfortunately, because there is so little discussion about the nexus between quality and quantity of sleep, most patients do not have interactions with their sleep doctors on these finer points. If they are insomniacs, they may hear more about setting regular schedules, restricting sleep a bit to build sleep pressure, and trying to learn not to worry about losing some sleep on an occasional night. If they are sleep apnea patients, they may be told paradoxical statements like “try to get at least 5 hours a night on the device” just to satisfy some insurance standards. Notice how neither of these discussions for either type of patient addresses the heart of the problem, namely, learn to gain the higher quality of slumber first, and then let’s talk about the right sleep duration to fit that high quality slumber. We want all sleep apnea patients to use their devices all night long and gain whatever is the correct amount of sleep form them on an individual basis. The same holds for insomnia patients in teaching them how to sleep all through the night. Either way, starting out the discussion with quality builds the foundation for measuring and the aiming for the right quantity of sleep.