Sleep research has produced an enormous amount of certainty: We know that sleep is vital, we know roughly how much is needed, and we know a great deal about the mechanisms of circadian rhythm and sleep architecture. What is discussed far less frequently is how much about sleep actually remains unexplained, even among the researchers who study it professionally.
Why we sleep at all is still not definitively settled
It sounds like a resolved question, but it isn't. There are several competing and partially overlapping theories: that sleep primarily serves physical restoration, that it primarily serves memory consolidation and brain cleansing, that it originated as an energy-saving strategy, and that it fulfills an immunological and metabolic regulatory function. The evidence supports parts of all four explanations, and no single theory currently covers the whole picture. Researchers largely agree that sleep performs all of these to some extent. But which function is primary, and why such an energy-intensive and vulnerable activity has arisen in almost every animal species studied, remains an active, unresolved field of research.
The function of dreams is truly unknown
REM sleep and dreaming are well-documented in terms of timing and active brain regions. But why the brain creates the subjective experience of dreaming, instead of simply consolidating memories and regulating emotions without a narrative inner world, is not understood. Some researchers consider dreams a byproduct of memory processing without their own function. Others argue that dreaming, especially with its emotionally intense and often threat-simulating content, serves its own evolutionary purpose for emotional regulation or danger rehearsal. Neither position is currently decisively proven, and this remains one of the truly open questions in neuroscience.
Why individual sleep needs vary so much is poorly explained
The range of seven to nine hours is well-documented at the population level. But why some people function well at the lower end of this range and others need the upper end or occasionally more is not fully mechanistically understood. Genetics has identified a small number of rare gene variants that are associated with naturally shorter sleep needs in a tiny fraction of the population. However, this only explains a fraction of the observed differences, and the bold claims of some public figures that one can train oneself to sleep less are not supported by current evidence for the general population.
The exact mechanism between sleep deprivation and disease risk is incomplete
The connections are well-documented: Poor sleep correlates with an increased risk of cardiovascular diseases, metabolic disorders, and cognitive decline. Less clear is the complete causal pathway that explains at the cellular and molecular level how disturbed sleep concretely produces each of these consequences. Researchers have identified several plausible mechanisms, inflammatory signals, disturbed glymphatic clearance, hormonal dysregulation, but the field does not yet have a unified model that links a specific pattern of sleep disturbance with a specific, predictable disease progression with the precision of more mature medical research fields.
Why sleep needs and sleep architecture change so much over a lifetime
Infants spend a dramatically higher proportion of their sleep in REM than adults, and sleep architecture shifts significantly through childhood, adolescence, and again in old age. The broad pattern is documented, but the biological logic behind why the brain needs exactly this developmental curve of sleep stages and why aging produces exactly the pattern of lighter, more fragmented sleep is only partially understood.
Whether one can "bank" sleep in advance is unclear
Individual studies suggest that several nights of extended sleep before an anticipated period of sleep deprivation can partially protect subsequent cognitive performance, a concept researchers call prophylactic sleep banking. However, the size, duration, and reliability of this protective effect are not well-established, and this is far from a solid recommendation. Most researchers are cautious about presenting this as a strategy rather than a preliminary, still developing field of research.
Why the field is still so open
Sleep research is a comparatively young field. Polysomnography, the technology that allows precise measurement of sleep stages, has only existed since the mid-20th century, and large human studies with the statistical power to clarify deeper mechanistic questions have only become common in the last few decades. The honest state of the field is substantial, well-documented knowledge about what sleep does and how to protect it, alongside truly open questions about why the underlying biology works the way it does. Both are true simultaneously, and any portrayal of sleep science as definitively settled omits the second half of this picture.