Cognitive Impairment

Summary

Sleep loss is not a willpower problem you can push through — it is a measurable, dose-dependent degradation of the brain's attention, working memory, reaction time and judgement, and its single most dangerous feature is that the impairment grows while your sense of being impaired flattens out, so the people who most need to stop are the least able to tell.

Why Strong

Tier 1 (Strong) because: the core claims — that sleep loss dose-dependently impairs attention, working memory, executive function, and reaction time, with the largest hit to sustained attention; that chronic restriction accumulates without adaptation; that subjective sleepiness uncouples from objective deficit; that recovery is real but slower than the debt — rest on multiple independent, government/institution-funded, randomised, controlled laboratory experiments with objective endpoints (PVT, working-memory tasks, EEG), converging meta-analytic effect sizes (Lim & Dinges g = -0.78 for the headline deficit), and a fully characterised mechanism (adenosine, local cortical sleep, prefrontal vulnerability). The funders (NIH, DoD, NASA, transport-safety bodies) have an operational stake in accuracy, not sales.

NOT Tier 0.5 because: Tier 0.5 is reserved for the foundational prescriptions (get adequate sleep — that's sleep_foundations_for_baseline / why_sleep_matters). This entry is the consequence/mechanism explainer that justifies those — strong evidence, but a specific causal claim rather than a universal first-principle.

NOT Tier 2 because: this is not "plausible mechanism + thin trials." It is one of the most experimentally replicated findings in cognitive neuroscience, with dose-response controlled trials and meta-analytic synthesis. Holding it at Tier 2 would understate genuinely strong evidence.

The one in-entry exception: the precise BAC-equivalence numbers (0.05% at 17h, 0.10% at 24h) are held a notch softer — the direction and rough magnitude replicate, but the exact mapping is task-dependent and rests on smaller older studies. Flagged in-line and in What Would Change Our Mind.

Practical takeaway

The recovery framing first. In Realised's register, the goal is not to "perform on less sleep" — that target is a category error, because the cost is structural, not motivational. The goal is to protect the sleep that protects the brain, and to recognise impairment you cannot feel. Everything below serves that.

Know the rough dose-response so you can self-locate:
• One full night, well-timed (7-9h for most adults): baseline. This is the target, not the luxury.
• Chronic 6h/night: you are accruing a real, accumulating deficit and you will not feel it after the first few days. "I'm fine on six" is the predicted feeling, not the fact.
• One all-nighter (24h awake): functioning roughly at the level of legal alcohol intoxication. Do not drive; do not make irreversible decisions; do not trust your self-assessment.
• 17+ hours awake (e.g. up since 6 am, now past 11 pm): measurably impaired, ~0.05% BAC ballpark — relevant for the late-night drive home.

What actually restores it (in order of power):
1. Adequate, regular night sleep. The only thing that clears the deficit rather than masking it. Consistency matters as much as duration (sleep_consistency).
2. Recovery sleep after a deficit — but budget more than you "owe." One long lie-in helps but does not fully reset a week of restriction; expect cognition to lag behind your feeling of recovery by a day or more. Pay debt down over several nights, not one (sleep_debt_payback).
3. A strategic nap. A 10-20 minute nap measurably restores alertness and reaction time for a few hours (short enough to avoid grogginess); a ~90-minute nap allows a full cycle. A nap is the best emergency countermeasure when night sleep isn't available — e.g. before a long drive.
4. Caffeine — a mask, not a fix, and time it. Caffeine genuinely rescues simple reaction time and suppresses the feeling of sleepiness for a few hours by blocking adenosine. It does not restore complex executive function, does not reliably prevent microsleeps, and the adenosine debt is still mounting underneath. Use it tactically, respect its long half-life so it doesn't wreck the next night's sleep (caffeine_timing_sleep), and never treat "I've had coffee" as "I'm safe to drive after no sleep."
5. Banking sleep before a known short night (extra sleep in the days before an anticipated deficit) buys real resilience and faster recovery — the only proactive lever (sleep_debt_payback).

What does NOT restore it: willpower, "getting used to it," energy drinks beyond their caffeine, a cold shower (transient arousal only), or the conviction that you personally are the exception.

What "impaired" looks like (track these, because you can't feel the deficit directly):
• Re-reading the same line; losing the thread of a conversation or sentence.
• Reaction-time errors: missing a turn, fumbling, near-misses, "where did that car come from."
• Doing tasks but making more careless errors and noticing them late or not at all.
• Irritability, flat mood, impulsive choices (food, spending, snapping at people) — the prefrontal signature.
• The classic head-nod / "lost a few seconds" while driving or in a meeting — that was a microsleep; treat it as a hard stop signal, not a "push on" signal.

The one rule that survives everything: if you are deciding whether you're too tired to drive, and the deficit-recognition machinery is itself impaired, do not use your feeling as the gauge. Use the clock and the dose: hours of sleep last night, hours awake now, time of day. If those say impaired, you are impaired regardless of how alert you feel — pull over, nap, or hand over the keys.

Evidence detail

Why This Entry Exists

A Realised user says some version of: "I only got five hours but I'm fine," or "I've adjusted to six," or "I'm tired but I can still function." They are running on a deficit and rating themselves as coping. This entry exists to answer the question they are actually asking — "does my sleep loss matter, really?" — with the honest, slightly uncomfortable answer: yes, more than you can feel, and the not-feeling-it is itself part of the deficit.

It also exists because sleep loss is upstream of almost every other complaint that arrives at the platform. A user reports brain fog, irritability, poor decisions, "no motivation," can't-stick-to-the-plan, a workout that fell apart, an eating spiral. Before any of those gets a bespoke explanation, the cognitive cost of short sleep is the first thing to rule in or out. This entry is the mechanism explainer — the home of record for what sleep loss does to thinking, by how much, how fast, and what gives it back — that the more behavioural sleep entries point to when a user needs to understand why the foundation matters.

What bad advice does this protect against?
• "You can train yourself to need less sleep." You cannot train away the cognitive cost; you can only train away the feeling of it (which is worse — see below).
• "Coffee fixes it." Caffeine masks subjective sleepiness and partly rescues simple reaction time; it does not restore the more complex deficits, and it does nothing for the microsleeps that cause crashes.
• "I pulled an all-nighter and got loads done." Self-rated productivity under sleep loss is unreliable; the work is slower and more error-prone than it feels.
• "I caught up on the weekend, so I'm reset." One long lie-in does not fully reverse a week of restriction on objective measures — recovery is slower than the debt was acquired.
• The macho framing that treats running-on-no-sleep as toughness rather than as operating at the cognitive level of someone over the drink-drive limit.

Evidence

The evidence here is unusually strong for a wellness topic, for a simple reason: sleep deprivation is cheap and ethical to study experimentally. You can randomise healthy volunteers to a sleep dose, keep them in a lab, and measure their cognition directly — you do not have to infer from epidemiology. That gives this field something most nutrition or supplement claims never get: large, controlled, dose-response data with hard objective endpoints. Most of it is government- or institution-funded (US Department of Defense, NIH, NASA, aviation and transport safety bodies — all of whom have a hard operational interest in the truth, not in selling anything).

1. The meta-analytic anchor — effect sizes by cognitive domain.
Lim, J., & Dinges, D.F. (2010). A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychological Bulletin, 136(3), 375-389. 70 articles, 147 cognitive tests. Funding: NIH / US Department of Defense (government — independent of any product).
• The single largest deficit was lapses in simple attention (the Psychomotor Vigilance Task — see Mechanism): Hedges g = -0.776 (95% CI [-0.96, -0.60], p < .001) — a large effect.
• The biggest decrements overall were in sustained attention and working memory. More complex, engaging tasks were affected less — counter-intuitively, the brain can briefly rally for something demanding and interesting, but cannot hold a boring vigil. This is why monotony (motorway driving, monitoring tasks, long meetings) is where sleep loss bites hardest.
• The weakest / non-significant effect was reasoning accuracy — given enough time and effort, a tired brain can still reason; it just does so slower and with more lapses along the way.

2. The dose-response landmark — chronic restriction accumulates like a debt, and you stop feeling it.
Van Dongen, H.P.A., Maislin, G., Mullington, J.M., & Dinges, D.F. (2003). The cumulative cost of additional wakefulness: dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep, 26(2), 117-126. Healthy adults randomised to 4h, 6h, or 8h time-in-bed for 14 consecutive nights. Funding: NIH (government).
• Restriction to 6 hours a night for two weeks produced cognitive deficits (PVT lapses, working memory) equivalent to one to two full nights of total sleep deprivation. Restriction to 4 hours matched two to three nights of total deprivation.
• Crucially, the deficits kept accumulating across the two weeks with no sign of adaptation — the brain did not "get used to it."
• The single most important finding in this whole field: the 4h and 6h groups' subjective sleepiness ratings rose for a few days and then plateaued — they stopped feeling much worse — while their objective performance kept declining. The feeling of impairment uncoupled from the fact of it. This is why chronic short sleep is so widely tolerated: people genuinely believe they have adapted, because their internal sleepiness gauge has gone quiet while their brain has not recovered.

3. The wakefulness-vs-alcohol equivalence (the public-safety headline).
Dawson, D., & Reid, K. (1997). Fatigue, alcohol and performance impairment. Nature, 388(6639), 235. And the confirmatory replications: Williamson & Feyer (2000), Occupational & Environmental Medicine; Lamond & Dawson (1999), J Sleep Research; Falleti et al. (2003), J Sleep Research. Funding: government / academic road-safety bodies (independent).
• After ~17 hours awake, performance on hand-eye coordination and reaction-time tasks degrades to roughly the level seen at a blood alcohol concentration of 0.05% — at or above the drink-drive limit in much of the world.
• After ~24 hours awake, the equivalent is roughly 0.10% BAC — well over the limit everywhere.
• [VERIFY — magnitude] Treat the specific BAC numbers as a vivid, directionally-correct illustration rather than a precise law: they come from particular psychomotor tasks, the mapping is task-dependent, and individuals vary widely. The direction and rough magnitude are well replicated; the exact "0.05% at 17h" figure is a useful headline, not a constant. (This is the one sub-claim held a notch below the rest of the entry.)

4. Recovery is real but slow and asymmetric.
Banks, S., Van Dongen, H.P.A., Maislin, G., & Dinges, D.F. (2010). Neurobehavioral dynamics following chronic sleep restriction: dose-response effects of one night for recovery. Sleep, 33(8), 1013-1026. After five nights at 4h, subjects got a single recovery night of 0, 2, 4, 6, 8, or 10 hours. Funding: NIH (government).
• A single recovery night helped in a dose-dependent way — but even a 10-hour recovery night did not return PVT performance to baseline after just five nights of restriction. The debt comes on faster than it pays off.
• Other work (Banks & Van Dongen) shows that people who go in better-rested both deteriorate less under subsequent restriction and recover faster — i.e. a fuller "tank" buys resilience (the practical basis for the bank sleep before known short nights tactic; see sleep_debt_payback).

5. Microsleeps — the mechanism behind sudden failures.
Under sleep pressure, the brain inserts brief, involuntary intrusions of sleep — microsleeps, typically 3-15 seconds, often with the eyes open and no awareness they occurred. On the PVT these show up as long lapses (reaction time >500 ms, or a missed response entirely). At motorway speed, a single 4-second microsleep is roughly 100+ metres travelled with nobody driving. This is the proximate cause of fall-asleep-at-the-wheel crashes and is not something effort can reliably prevent once pressure is high enough (Dorrian, Rogers & Dinges, 2005, in Kushida ed., Sleep Deprivation; Poudel et al. on behavioural microsleeps — academic/independent).

Population caveats. Most of this is from healthy young/working-age adults in lab conditions. Real life adds caffeine, circadian timing, motivation, and stimulation that can transiently mask deficits — none of which removes them. Older adults show somewhat fewer microsleeps on a sleepless night than younger adults (a partial protection), but also worse sleep quality overall. Effects are larger at the circadian low (roughly 3-6 am) — the same hours when shift workers and long-haul drivers are most exposed.

Mechanism

Why thinking degrades — three layers, all well-characterised.

1. Homeostatic sleep pressure: adenosine accumulates while you're awake. Every waking hour, the brain accumulates metabolic by-products of neural activity — chief among them adenosine, which builds up in the cortex and basal forebrain and dampens the arousal systems that keep you alert. The longer you're awake, the higher it climbs; sleep is what clears it. After even one night without sleep, adenosine A1-receptor binding in the human frontal cortex measurably rises. Caffeine works by blocking these adenosine receptors — which is exactly why caffeine masks sleepiness without clearing the underlying pressure: the adenosine is still there, you just can't feel it pushing.

2. "Local sleep" — bits of the cortex go offline while you're still awake. This is the striking finding that reframes the whole topic. In sleep-deprived animals (and, by inference, humans), individual patches of cortex briefly switch "OFF" — neurons go silent in slow-wave-like bursts — even though the animal is awake and behaving (Vyazovskiy et al., 2011, Nature; local-sleep literature). The incidence rises the longer you've been awake, and different cortical areas go offline at different moments. Subjectively you feel "awake"; functionally, the specific circuit you needed for that judgement or that response was momentarily asleep. A lapse of attention is, quite literally, a fragment of your brain having taken a micro-nap. This is why sleep-deprived performance is so variable — not uniformly slow, but normal-normal-normal-catastrophic-normal, as different regions flicker offline.

3. The prefrontal cortex takes the biggest hit. The frontal lobes — executive function, working memory, impulse control, judgement, emotional regulation — are disproportionately vulnerable to sleep loss (high metabolic demand, high adenosine sensitivity). This is the neural reason that a tired person is not just slower but worse at deciding, more impulsive, more emotionally reactive, worse at planning, and worse at recognising they're impaired (the self-monitoring that would tell you you're compromised runs on the very tissue that's most compromised). It is a closed loop: the part of the brain that should raise the alarm is the part that's been switched off.

Why the feeling uncouples from the fact (the load-bearing insight). Subjective sleepiness tracks the acute arousal state — how recently you slept, time of day, whether you just had coffee or stood up or got a hit of stimulation. Objective impairment tracks the accumulated homeostatic and circadian debt. Under chronic restriction these two diverge: the arousal-driven feeling habituates and plateaus, while the debt-driven deficit keeps climbing. You stop feeling worse; you keep getting worse. This is not a quirk — it is the central public-health danger of sleep loss, because it disables the one signal a person would use to decide to stop.

Risks And Contraindications

This entry is descriptive (what sleep loss does), so the "risks" are the risks of the state it describes, not of an intervention:
• Operational / driving risk is the headline. Drowsy driving is a leading, under-counted cause of fatal crashes precisely because microsleeps give no warning and the driver feels capable. Never drive at the circadian low (3-6 am) or after extended wakefulness if it can be avoided; nap first.
• Decision and emotional risk. Avoid making irreversible high-stakes decisions, having difficult emotional conversations, or operating dangerous equipment while significantly sleep-deprived. The judgement and the self-monitoring of judgement are both degraded.
• Do not weaponise this against someone with insomnia. A person lying awake terrified about the cognitive cost of not sleeping is now in a vicious loop — the anxiety worsens the insomnia. The honest counter-message: a single poor night is recoverable and the brain is resilient to occasional loss; the concern is chronic restriction, not one bad night. For genuine sleep-onset/insomnia distress, route to the behavioural sleep entries, not to alarm.
• Underlying medical causes. Persistent unrefreshing sleep, cognitive impairment despite adequate sleep opportunity, or heavy daytime sleepiness can signal sleep apnoea, thyroid dysfunction, depression, or other conditions — not simple sleep debt. Persistent fog that doesn't lift with more sleep warrants medical assessment, not just "sleep more" (fatigue_cross_pillar_diagnostic, osa_diagnostic_lifestyle, thyroid_dysfunction).
• Stimulant misuse risk. Using escalating caffeine or other stimulants to mask chronic deficit hides the accumulating debt and can disrupt the very sleep that would clear it — a self-reinforcing trap.

Cross-Pillar Connections

• Sleep (home pillar) — why_sleep_matters, sleep_foundations_for_baseline: this entry is the cognitive-cost mechanism those prescriptive entries point to. sleep_debt_payback: the recovery-dynamics and bank-before-deficit tactics live there; this entry supplies the why. sleep_consistency and sleep_duration: the dose side of the dose-response. caffeine_timing_sleep: caffeine as mask-not-fix and the half-life trap.
• Mental — cognitive_bandwidth_as_finite_resource: sleep loss is one of the largest single drains on the finite working-memory/attention budget that entry describes; they are mechanistically linked (prefrontal capacity). mental_adhd_comprehensive: sleep loss mimics and worsens attentional-control deficits — rule out short sleep before attributing focus problems to anything else.
• Cross-pillar diagnostic — fatigue_cross_pillar_diagnostic: persistent cognitive fog despite adequate sleep opportunity routes here for differential (apnoea, thyroid, mood, etc.) rather than "just sleep more."
• Physical / Diet (brief): sleep loss degrades the executive control that holds training plans and eating intentions together — many "discipline" failures are downstream of short sleep, not character.

What would change our mind

Sources (11)

• Lim, J., & Dinges, D.F. (2010). A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychological Bulletin, 136(3), 375-389. (70 studies, 147 tests; NIH/US DoD — government, independent. Largest deficit: simple-attention lapses, g = -0.776.)↗
• Van Dongen, H.P.A., Maislin, G., Mullington, J.M., & Dinges, D.F. (2003). The cumulative cost of additional wakefulness: dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep, 26(2), 117-126. (4h/6h/8h × 14 nights; NIH — government. The subjective-plateau-vs-objective-decline dissociation.)↗
• Banks, S., Van Dongen, H.P.A., Maislin, G., & Dinges, D.F. (2010). Neurobehavioral dynamics following chronic sleep restriction: dose-response effects of one night for recovery. Sleep, 33(8), 1013-1026. (Single recovery night, 0-10h; NIH — government. 10h recovery did not fully restore PVT.)↗
• Dawson, D., & Reid, K. (1997). Fatigue, alcohol and performance impairment. Nature, 388(6639), 235. (~17h awake ≈ 0.05% BAC, ~24h ≈ 0.10%; academic/road-safety — independent.)↗
• Williamson, A.M., & Feyer, A.-M. (2000). Moderate sleep deprivation produces impairments in cognitive and motor performance equivalent to legally prescribed levels of alcohol intoxication. Occupational & Environmental Medicine, 57(10), 649-655. (Replication; government/academic — independent.)↗
• Falleti, M.G., et al. (2003). Qualitative similarities in cognitive impairment associated with 24 h of sustained wakefulness and a blood alcohol concentration of 0.05%. Journal of Sleep Research, 12(4), 265-274. (Replication — independent.)↗
• Vyazovskiy, V.V., Olcese, U., Hanlon, E.C., Nir, Y., Cirelli, C., & Tononi, G. (2011). Local sleep in awake rats. Nature, 472(7344), 443-447. (Local-sleep / cortical-OFF mechanism; NIH — government.)↗
• Dorrian, J., Rogers, N.L., & Dinges, D.F. (2005). Psychomotor vigilance performance: neurocognitive assay sensitive to sleep loss. In Kushida (ed.), Sleep Deprivation. (PVT, lapses, microsleeps; academic — independent.)↗
• Goel, N., Rao, H., Durmer, J.S., & Dinges, D.F. (2009). Neurocognitive consequences of sleep deprivation. Seminars in Neurology, 29(4), 320-339. (Review incl. trait-like vulnerability, adenosine, prefrontal effects; NIH — government.)↗
• Krause, A.J., et al. (2017). The sleep-deprived human brain. Nature Reviews Neuroscience, 18(7), 404-418. (Mechanistic review; academic/NIH — independent.)↗
• Funding notation: this body of work is overwhelmingly government- and institution-funded (NIH, US Department of Defense, NASA, transport- and occupational-safety agencies). These funders have an operational stake in accuracy, not in selling a product — among the cleanest evidence environments in the whole knowledge base.*↗

Related entries