Wakefulness and the brain · stay-awake brain v1 · ZH / RO / EN
Stay-Awake Brain: 24-hour timeline, BAC comparison cards, PVT, caffeine calculator
In which waking hour does your brain “give up”? Sleep-deprivation research as an interactive 24-hour timeline: the wake slider, comparison cards, PVT, a caffeine calculator — the meaning of “17–19 hours ≈ the drink-driving line” and its limits (night testing, group averages, no individual prediction).
Public education, not medical advice · Group averages, no individual calibration
Three sentences first: every BAC card is a group average from night testing, never a prediction about you; the calculator's lines are authorized paraphrase, and the milligrams you enter are label values; the PVT here has no reading, only a reaction-time distribution.
The 24-hour timeline slider
Hour ticks from 0 to 24. Drag the slider to 17–19 and watch which card lights up; cards without an hour anchor never light — that is not a bug, it is the edge of the evidence.
Comparison cards: BAC, simulated driving, records
Every card names its source and scope (night / group average / single cohort). Without JS, all cards show statically.
17–19 hours awake ≈ BAC 0.05% (some tests)
Williamson 2000 (n=39: 30 transport workers + 9 army): after 17–19 hours awake, some cognitive/motor tests match or fall below BAC 0.05%. Night testing (clock times 22:30/01:00), group average, no individual prediction, no daytime extrapolation.
Longer wakefulness: some results ≈ BAC 0.10%
The same study found that after longer wakefulness, some results reached the level observed at its maximum alcohol dose (BAC 0.10%). Testing took place at night; these are group averages, not individual predictions.
Simulated driving: 18.5/21 hours awake ≈ BAC 0.05%/0.08%
Arnedt 2001: 18 men aged 19–35 completed a 30-minute simulated drive. The sleep-deprivation group was tested at night after staying awake all night. For mean lane position, lane-position variability, and speed variability, results after 18.5 and 21 hours awake were similar in magnitude to those at BAC 0.05% and 0.08%, respectively. Alcohol had larger effects on speed deviation and off-road events. The simulation had no other traffic.
Dawson & Reid (1997): publication record has no abstract
The Nature record describes a one-page letter but includes no abstract. The publication record alone does not establish its findings, so this card gives no wake-hour or BAC comparison.
Caffeine timing, the adenosine mechanism, one-night effects
The timing trial, the mechanism framing, and the meta-analysis effects card. The Drake card carries the conflict note; the Wüst card covers one restricted night only, no slopes.
Caffeine taken six hours before bed still disrupted sleep
Drake 2013: participants received a fixed 400 mg dose of caffeine or placebo at home and reported how they slept; researchers also used a portable sleep monitor. Caffeine taken zero, three, or six hours before bed disrupted sleep significantly more than placebo (p<0.05). The authors noted an important effect even when caffeine was taken six hours before bed. The study included one experimental group.
Conflict note: a Zeo Inc investigator-initiated grant + author pharma ties (Merck/Teva/Zeo/Pfizer/Sanofi/Jazz/Purdue; funding page verified).
Adenosine mechanism box (framing, not an evidence chain)
Reichert 2016 review (PMID 26861410, Biology 5(1):11): adenosine accumulation → the sleepiness signal, mechanism framing only. The A1-receptor paper's specific claims are unverified, not on this page.
One-restricted-night effects card (k=44 meta-analysis)
Wüst 2024 (one restricted night with 2–6 h sleep opportunity; k=44): subjective sleepiness SMD=0.986; sustained attention impaired (10-min tasks: reaction times SMD=0.512, lapses SMD=0.489); no significant effects on choice reaction time, cognitive throughput, working memory, or inhibitory control. Sleep duration was not significantly associated with sleepiness (β=-0.214 at a 0.01 threshold) or attention (p>0.090).
After 20–25 hours awake, some task results ≈ BAC 0.10%
Lamond 1999: 22 healthy participants aged 19–26 completed three counterbalanced conditions: 28 hours awake, alcohol, and placebo, using four standardized tasks. Depending on the task, about 20–25 hours awake produced performance decrements equivalent to those at BAC 0.10%. Results differed by task and do not predict an individual's response.
Tools: PVT and the caffeine calculator
Two independent tools, each with its own limits. Read the limits, then enter.
Method box
- The Drake trial: single cohort + actigraphy reading; the conflict note sits on the card (Zeo/pharma ties).
- Group-average limit: this tool never learns you, it only repeats group results; no individual predictions, no overnight learning.
- Gate status: Wüst closed at abstract level; the Dawson full-text gate stays open (the card stays a record); Arnedt hour anchors come from the abstract; the FDA page fetched and checked 2026-09-20 (400 mg line confirmed; no pregnancy number line, the current page states none); PVT reading copy under review (no reading in this version).
Source notes
S-Williamson-abs (PMID 10984335) · S-Arnedt-abs (PMID 11235795) · S-Lamond-abs (PMID 10646165) · S-Dawson-record (PMID 9230429) · S-Drake-abs (PMID 24235903) + S-Drake-COI (PMC3805807 disclosure) · S-Reichert-page (PMID 26861410) · S-Horne-record (PMID 7888930, qualitative basis of drowsy-driving warnings) · S-FDA-record (400 mg line paraphrase) · S-TFDA-quote (Taiwan labeling rules) · S-Wuest-fulltext (PMID 38759474). Full D1–D10 + D13 on the method page.