Could Humans Ever Need Less Sleep?

Why We Sleep (And Why It’s More Complicated Than You Think)

A typical person spends roughly 26 years of their life asleep. Another seven years trying to get there. That’s a third of a human life — unconscious, unproductive, and unavoidable. Or so the story goes.

The science has always been clearer on what happens when you don’t sleep than on precisely what sleep actually does. We know the consequences of deprivation in horrifying detail: after 24 hours without sleep, cognitive performance degrades to the level of someone who is legally drunk. After 72 hours, hallucinations arrive reliably. Rats forced into total sleep deprivation die within two to three weeks, from causes researchers still argue about.

What sleep does at a mechanistic level, though — the precise biological transactions that make it non-negotiable — has only begun to come into focus in the last decade. And as that picture sharpens, something counterintuitive emerges: many of the specific jobs sleep performs can, at least in principle, be performed by other means.

Sleep appears to handle several distinct functions simultaneously. Memory consolidation — the transfer of short-term information into durable long-term storage — happens primarily during slow-wave and REM sleep. Hormonal regulation, particularly growth hormone and cortisol cycling, is tightly coupled to sleep architecture. The immune system, rather quietly, does a great deal of its routine maintenance during the night. And then there is the brain’s plumbing system, which turns out to be the most consequential and the least understood.

But “doesn’t have an obvious equivalent yet” is doing a lot of work in that sentence. Science has a long history of declaring biological functions irreplaceable right up until someone figures out the underlying mechanism. The kidney was once a black box. So was the liver. So was the immune system.

The reasonable question — and the one researchers are now starting to ask seriously — is whether sleep’s constituent functions can be unbundled, and each addressed individually through targeted intervention. The answer, increasingly, looks like a cautious yes.

The Glymphatic System: Sleep’s Real Superpower

If there’s one discovery that frames the whole debate about sleep’s future, it’s the glymphatic system. Described in detail by neuroscientist Maiken Nedergaard and her team in 2013, it’s a network of perivascular channels through which cerebrospinal fluid flows during sleep, carrying metabolic waste — including amyloid-beta and tau proteins, the hallmarks of Alzheimer’s disease — out of the brain tissue and into the broader lymphatic circulation.

The implications were immediately recognized as enormous. Poor sleep accelerates amyloid accumulation. Even a single bad night measurably increases tau in the cerebrospinal fluid. People with Alzheimer’s and Parkinson’s both show markedly impaired glymphatic function. The pathway between chronic sleep deprivation and neurodegeneration suddenly had a plausible mechanism.

For years, however, the glymphatic system had been demonstrated only in animal models — primarily rodents. That changed in late 2024, when a team at Oregon Health and Science University used contrast MRI in five neurosurgical patients to become the first to definitively visualize the glymphatic network in living human beings, confirming perivascular fluid flow in real time.

The follow-up question arrived almost immediately: what exactly powers the system? In January 2025, Nedergaard’s group published in Cell their identification of the mechanism: tightly synchronized oscillations in norepinephrine, blood vessel volume, and cerebrospinal fluid constitute the pump that drives glymphatic clearance during non-REM sleep.¹ The brain’s blood vessels rhythmically expand and contract, creating a piston-like pumping action — and this happens because of slow, cyclical fluctuations in norepinephrine released by the locus coeruleus.

This is a significant finding for people interested in bypassing sleep. If the pump is driven by a specific neurochemical oscillation, and that oscillation can be artificially induced while a person is awake, you could theoretically run the brain’s waste-disposal system during waking hours. It hasn’t been done — but it’s no longer merely speculative. Researchers at several institutions are already designing experiments to attempt exactly this, targeting the norepinephrine rhythms pharmacologically and through transcranial stimulation.

The Short Sleeper Genes: Nature’s Proof of Concept

Perhaps the strongest argument that sleep’s minimum requirement is not biologically fixed is the existence of natural short sleepers. Not the person who sets their alarm for 6 a.m. and tells themselves they’re fine — but people with specific inherited mutations who genuinely thrive on dramatically less sleep with no measurable deficit, across their entire lifetime.

The first breakthrough came in 2009, when Ying-Hui Fu’s laboratory at UCSF identified a mutation in the DEC2 gene (also known as BHLHE41) in a mother and daughter who naturally slept only 6.25 hours per night. Family members without the mutation averaged over 8 hours. The short sleepers showed no daytime fatigue, no cognitive impairment, and no health consequences typically associated with chronic short sleep.²

Since then, researchers have identified four more genetic variants associated with natural short sleep. The list is growing:

What’s particularly striking about the DEC2 mutation is its effect on sleep efficiency rather than just duration. The mutation appears to increase the density of restorative processes per hour of sleep, so that six hours of DEC2-mutant sleep delivers roughly the same biological outcomes as eight hours of normal sleep. Three of the four originally identified mutations also appear to suppress amyloid and tau accumulation through independent pathways — meaning they’re not just getting by with less sleep, they’re actively protecting against the neurodegenerative consequences that short sleep typically carries.⁴

The prevalence of the DEC2 mutation is estimated at roughly 1 in 1,000 people. That’s rare, but it’s not vanishingly rare — and importantly, it proves that human biology can accommodate significantly compressed sleep without apparent harm. The question for researchers isn’t whether such a state is biologically possible (clearly it is), but whether the underlying mechanism can be pharmacologically or genetically induced in people who didn’t inherit the mutation.

A 2025 paper in Current Biology confirmed that the dec2-P384R mutation extends lifespan and promotes healthy aging in Drosophila models, partly through enhanced mitochondrial fitness and upregulated stress response pathways. This is a model organism, not a human trial — but it’s a meaningful signal that compressed sleep via genetic means carries metabolic advantages, not just neutral tolerance.

DARPA’s Bet: Sleep as an Engineering Problem

Government research agencies don’t often tip their hand about what they think is genuinely achievable versus speculative. So it’s worth paying attention when the Defense Advanced Research Projects Agency — the organisation that funded the precursors to GPS, the internet, and stealth aircraft — launches not one but two formal programs aimed at deconstructing human sleep requirements.

In 2024, DARPA issued a solicitation for the Alert WARfighter Enablement (AWARE) program. The goal: develop a combination of photoswitchable drugs and wearable near-infrared light emitters that boost alertness in sleep-deprived military personnel without the addiction risk or emotional volatility of amphetamines. The target is soldiers who have been awake for 36 or more hours and need to make high-stakes decisions. A 2024 Government Accountability Office report found that 93% of active-duty US service members regularly sleep seven hours or less and report feeling sleep-deprived, with the problem linked to fatal accidents and hundreds of millions of dollars in equipment damage.⁵

Then, in February 2025, DARPA went further with the Reengineering Enabling Sleep Transitions in Operationally Restrictive Environments (RESTORE) program. This one is more ambitious. RESTORE’s stated goal is to demonstrate “precision control of sleep” sufficient to optimize full cognitive performance following just three hours of sleep per day. The approach focuses on non-invasive neuromodulation — essentially, using external brain stimulation to compress sleep architecture so densely that three hours delivers what eight hours currently provides.

The military context matters for a simple reason: armed forces have the resources, the institutional tolerance for experimental risk, and the operational urgency to push this research faster than civilian health systems would. The US Air Force already routinely provides modafinil to pilots on extended missions — a practice normalized over two decades. The pressure to move beyond modafinil is real, and DARPA is where that pressure usually goes to get solved.

Pharmacology: From Modafinil to Orexin Manipulation

Modafinil is, in some ways, the prototype of what sleep replacement might look like. It promotes wakefulness through a complex mechanism that isn’t fully understood — it’s not simply a stimulant in the amphetamine sense. Studies consistently show it preserves cognitive performance during sleep deprivation better than caffeine or dextroamphetamine, without the cardiovascular stress or addiction liability. The US Air Force has used it operationally since the early 1990s.

But modafinil doesn’t actually replace sleep’s restorative functions — it just delays the cognitive consequences of not sleeping. After two or three days, even modafinil-dosed subjects show significant impairment on tasks requiring complex judgment. It’s a bridge, not a solution.

The more promising pharmacological frontier is the orexin system. Orexin (also called hypocretin) is a neuropeptide produced in the hypothalamus that acts as the brain’s primary wakefulness promoter. People with narcolepsy have lost their orexin-producing neurons and collapse involuntarily into sleep — a natural experiment that confirmed orexin’s centrality to conscious wakefulness. The DEC2 short-sleep gene works partly by dysregulating orexin — the mutation removes a brake on orexin expression, which is part of why DEC2 carriers function well on less sleep.

A class of drugs called dual orexin receptor antagonists (DORAs) — suvorexant, lemborexant, and daridorexant — now approved for insomnia, work by blocking orexin’s wakefulness signal at its receptor sites, promoting sleep. The inverse approach — orexin agonists, drugs that amplify orexin signaling — could in theory extend wakefulness while preserving some of sleep’s biochemical maintenance functions. These don’t exist as approved drugs yet, but they’re in early-stage research.⁶

There’s also the separate question of the adenosine system. Adenosine is the chemical pressure that accumulates during waking hours and makes you sleepy — it’s why caffeine works (caffeine blocks adenosine receptors). A drug that could selectively clear adenosine, or prevent its buildup, while separately triggering glymphatic clearance and memory consolidation, would be a closer approximation of what sleep actually does. This is speculative for now, but it’s a coherent target.

Neurostimulation: Inducing Sleep’s Benefits Without Sleep

Pharmacology might not even be the primary route here. Neurostimulation — using electromagnetic fields, ultrasound, or light to alter brain activity non-invasively — is advancing rapidly and, in some respects, offers finer control than drugs.

Transcranial alternating current stimulation (tACS) can entrain the brain into specific oscillatory patterns. Delta frequency stimulation can induce and sustain slow-wave activity — the hallmark of deep sleep — without the person actually being asleep. A 2025 meta-analysis of tACS for chronic insomnia found significant improvements in sleep quality, latency, and total sleep time across randomized controlled trials. That’s using stimulation to improve sleep, not replace it — but the underlying principle is that the brain’s electrical architecture is malleable from the outside.

Transcranial focused ultrasound (TUS) is another candidate. It can reach deep brain structures non-invasively — structures like the locus coeruleus that control the norepinephrine oscillations driving glymphatic clearance. A 2025 review in the Journal of Sleep Research identified TUS and temporal interference stimulation (TIS) as the two technologies most likely to enable real control over deep sleep oscillations in clinical settings within the next decade.⁷

Transcranial photobiomodulation (t-PBM) — directing specific wavelengths of near-infrared light at brain tissue — is a third avenue. A July 2025 systematic review from NYU and Mount Sinai found evidence that t-PBM increases cerebral mitochondrial activity and blood flow, with downstream improvements in both sleep quality and waking alertness, and tentative evidence of effects on glymphatic function.⁸

The $411 Billion Incentive

Science follows funding, and funding follows incentive. The economic case for cracking sleep is staggering.

RAND Corporation’s analysis found that insufficient sleep costs the US economy up to $411 billion annually in lost productivity, with Japan losing up to $138 billion and the UK £50 billion. Across just five OECD countries, the Lancet put the 2025 projected figure at up to $718 billion when absenteeism and presenteeism are both counted.⁹ Sleep deprivation accounts for roughly 1.86–2.92% of Japan’s GDP annually — more, proportionally, than the US.

On the other side of the ledger, the global sleep economy — sleep aids, tracking devices, pharmaceuticals, smart mattresses — is estimated at nearly $600 billion and growing. Venture capital funding for sleep technology roughly doubled between 2017 and 2021. The global sleep tech market is projected to grow from $5.4 billion in 2024 to $9.1 billion by 2030.¹⁰

The combination of massive economic loss from bad sleep and a fast-growing commercial interest in solving it creates an unusually well-funded research environment. Unlike, say, curing aging — where the fundamental biology is still poorly understood — sleep manipulation has specific, identifiable molecular targets, several of which have already been validated by genetic evidence. That makes it commercially investable in a way that pure longevity science isn’t yet.

A Realistic 2026–2050 Roadmap

Let’s be specific about what “might not need sleep by 2050” actually means in practice — because the headline is doing some work here that deserves unpacking.

The most extreme reading — complete, permanent elimination of sleep — is almost certainly not where this goes. Sleep is deeply integrated into hormonal rhythms, immune function, memory consolidation, and tissue repair in ways that we can’t yet fully disentangle. Eliminating it entirely without consequence would require essentially reengineering human physiology from the ground up.

What does look achievable, based on the current research trajectory, is a future where sleep is optional in its current form — replaced by shorter, artificially optimized rest combined with waking-hour interventions that handle the specific functions sleep currently monopolizes.

This roadmap is optimistic in some respects — it assumes smooth regulatory progress and no catastrophic safety discoveries — but it’s grounded in research programs that actually exist and are funded right now. It’s not science fiction; it’s applied neuroscience with a long development timeline.

What Could Go Wrong: The Honest Counterarguments

Any serious treatment of this subject has to grapple with the objections, and there are genuine ones.

The glymphatic problem may not be solvable

We now know that the glymphatic system is driven by norepinephrine oscillations tied to non-REM sleep physiology. But we also know — from the January 2025 Cell paper — that common sleep aids can disrupt this oscillation, slowing clearance. The drugs we currently use to help people sleep are, in some cases, making the brain’s cleanup less efficient. Engineering around this without disrupting it further will require extraordinary precision.

Immune suppression is underexplored

Sleep deprivation consistently degrades immune function — reducing natural killer cell activity, disrupting cytokine production, and impairing vaccine response. A 2025 study in Sleep Advances from Oregon Health & Science University found that insufficient sleep was the second strongest predictor of reduced life expectancy after smoking — stronger than physical inactivity or social isolation.¹¹ We don’t know whether sleep compression technologies, even sophisticated ones, can fully preserve the immune maintenance functions that natural sleep provides.

Climate is actively making sleep worse

Even as we discuss engineering humans out of their sleep requirement, rising temperatures are eroding sleep quality globally. A 2022 study in One Earth projected that temperature-attributed short nights will increase from roughly 11 per person per year in 2010 to 12 by 2050 under moderate emissions scenarios. A 2025 Nature Communications analysis of 23 million sleep records found that for every 10°C increase in ambient temperature, odds of sleep insufficiency increase by 20.1%.¹² The population sleeping badly is growing at the same time the technology to address it is developing — which creates both urgency and risk that interventions will be applied before they’re adequately tested.

The equity problem is real

Every technology of this nature arrives first for the wealthy and well-connected. If sleep compression becomes a competitive advantage — more waking hours, sharper cognition — the pressure to adopt it will be coercive in practice, even if technically voluntary. Workers in high-performance fields will face the same implicit pressure people currently face around using caffeine or smartphones. The gap between those who can afford engineered sleep and those who can’t, or who lack access to the monitoring technology to do it safely, could become a genuine inequality axis.