Adenosine pressure, the clock and the axis together drive sleep onset and maintenance, and when the three are synced you fall asleep easily and wake naturally.Sleep is not "powering off" — it's an active neurochemical event. Three axes simultaneously drive sleep onset and maintenance:
① Adenosine pressure (homeostatic, the hourglass)
· The longer you're awake, the more brain adenosine rises. · Adenosine binds A1 + A2A receptors → inhibits arousal systems (locus coeruleus NE / forebrain ACh). · After 16-18 h of wakefulness, the adenosine pool is enough to trigger deep sleep. · Sleep → adenosine clearance → hourglass reset.
② clock (circadian, the metronome)
· The suprachiasmatic nucleus (SCN) receives blue-light signals (480 nm) via retinal ipRGC ganglion cells. · Morning light → adjusts CRY/PER gene expression → calibrates SCN phase. · Evening darkness → pineal synthesizes melatonin → release → "time to sleep" signal + core body temperature drops.
③ axis (stress, the off switch)
· Hypothalamic PVN → pituitary ACTH → adrenal cortisol. · Normal cortisol rhythm: morning peak (CAR) → afternoon low → lowest just before sleep. · Low pre-sleep cortisol + parasympathetic activation = the conditions for sleep onset.
Three axes synced: hourglass full + metronome on time + off switch closed → smooth sleep onset + 7-9 h of deep sleep + natural morning awakening.
2 · Single-axis failure
Any single broken axis can wreck sleep: afternoon coffee blocks adenosine, evening blue light tells the it is still day, and chronic stress keeps the axis on.Any single broken axis can wreck your sleep:
Scenario A · Adenosine blocked (afternoon / evening coffee)
· A large coffee at 14:00 (200 mg caffeine). · Caffeine half-life ~5 h (CYP1A2 slow metabolizers ~10 h). · At 22:00, still 50-100 mg in blood → A1/A2A receptors occupied. · Adenosine hourglass is full but the brain doesn't receive the signal → trouble falling asleep + reduced deep sleep. · CYP1A2 rs762551 slow-metabolizer genotype: the same coffee lingers markedly longer; subjectively you may notice nothing, while deep sleep objectively takes a hit — how large a hit varies by person, and nothing on this site supports a specific figure, so none is given.
Scenario B · pushed off by blue light (evening phone use)
· At 22:00 watching phone / Netflix / tablet. · Screen blue light (peak 480 nm) → ipRGCs fire → SCN concludes "still daytime." · Melatonin release delayed by 1-2 h. · Subjectively: "in bed but not sleepy" + struggling to wake up. · A common lifestyle trigger for DSPS (delayed sleep phase syndrome).
Scenario C · won't turn off (chronic stress / rumination)
· High-pressure work + ruminative thoughts. · Pre-sleep cortisol remains high (should be low). · The "sleep-wake" transition is blocked. · Early cortisol rise around 2-4 a.m. → early awakening. · A typical maintaining mechanism of chronic insomnia.
The three scenarios:
· Each alone can cause insomnia. · "Take melatonin" only helps B (and only if timed correctly). · "Take a sleeping pill" can short-term force sedation but doesn't repair any axis. · addresses all three axes at once (next step).
3 · Chronic — all 3 drift
Chronic insomnia is almost never a single-axis problem but all three axes drifting and reinforcing each other; sleeping pills fix none of them, while targets each.Chronic insomnia (≥ 3 months) is almost never a single-axis problem — it's all three axes drifting and reinforcing each other:
Typical course:
· Trigger (acute insomnia): one major stressor / jet lag / illness. · "Afraid of not sleeping" + anxiety → hard to shut off. · Insomnia → daytime fatigue → more coffee → adenosine more blocked. · Falling asleep on the couch → later bedtime → drift. · "I need to catch up tonight" → weekend sleeping till noon → harder to sleep Monday. · After 3 months: all three axes are drifting → chronic insomnia is established.
Conditioned (psychophysiological) insomnia:
· The conditioned reflex "bed = anxiety location" is built up. · Getting into bed itself triggers wakefulness (even though a second earlier you were nodding off on the couch). · This is the phenomenon 's "stimulus control" specifically targets.
Why sleeping pills don't work:
· Z-drugs / benzodiazepines = forced opening → forced sleep. · They repair none of the three axes: don't increase adenosine sensitivity, don't recalibrate SCN, don't lower HPA. · Tolerance begins in 2-4 weeks → dose creep upward. · Discontinuation rebound: the three axes are still drifting; stopping the drug exposes the original problem. · The FDA boxed warning covers exactly one topic — complex sleep behaviours (sleepwalking, sleep-driving, dangerous activity while not fully awake) — plus a contraindication in anyone who has had such an episode. Falls, dementia and dependence are not in the boxed warning; don't pack them into the same parenthesis. · Falls and dementia are a different class of evidence (observational association), and the dementia signal belongs mainly to benzodiazepines: a case-control study found risk rising from three months of cumulative use (Billioti de Gage 2014 BMJ). Z-drug-specific evidence is weaker and more mixed.
CBT-I components mapped to axes:
· Stimulus control (bed only for sleep) → rebuilds the adenosine-bed conditioned reflex. · Sleep restriction (limit time in bed) → strengthens adenosine pressure. · Fixed wake-up time → resets SCN. · Morning outdoor light (15-30 min) → advances the SCN into phase. · Cognitive restructuring ("a bad night tonight is not catastrophic") → lowers HPA. · Relaxation training (PMR / diaphragmatic breathing) → lowers HPA.
The firmest link in the chain: in a of randomized trials in adults with chronic insomnia, 6-8 weeks of CBT-I → sleep latency ↓ 19 min + WASO ↓ 26 min + efficiency ↑ 10% + still maintained 6 months later; the AASM guideline makes it a strong recommendation on that basis (Trauer 2015 + Edinger 2021).
4 · Tool map
Evidence-based sleep tools sort by axis: magnesium and glycine on the adenosine and axis, low-dose melatonin and morning light on the , and on the axis.The genuinely evidence-based "sleep tools," sorted by axis:
· Magnesium 300-400 mg (glycinate) — NMDA Mg²⁺ plug + SERCA — a few small randomized trials; the tolerable upper intake level () for supplemental magnesium is 350 mg a day, so go above it only under medical guidance. · Glycine 3 g — thermoregulation + GABA (Yamadera 2007) — one small randomized trial. · L-theanine 200-400 mg — α-brainwaves + reduced anxiety (Owen 2008) — a few small randomized trials. · Ashwagandha (KSM-66 / Sensoril) — chronic stress / indirect + warning — a few small randomized trials.
axis (clock signal)
· Melatonin 0.3-0.5 mg 30-60 min before bed — only for phase problems + jet lag + DSPS + shift work — several randomized trials and a . · 5-10 mg melatonin = overdose → roughly 50-100 × the dose needed to return plasma to the physiological nocturnal range (~ 0.1 mg), with no further gain in sleep onset (the Dollins 1994 dose-response). · Morning outdoor light 15-30 min — strongest SCN calibration signal — established physiology (light intensity outweighs any supplement).
HPA axis (stress / rumination)
· cognitive restructuring + relaxation training — first choice; pooled randomized trials agree. · Mindfulness meditation — reduces rumination — supported by small randomized trials. · Diaphragmatic 4-7-8 breathing — rapid parasympathetic activation — immediate.
Drugs (GABA bypass — use long-term with caution)
· Z-drugs (Ambien / Lunesta) — short-term < 4 weeks + watch for sleepwalking. · Benzodiazepines — avoid as first-line for insomnia + high dependence. · Antihistamines (Benadryl / doxylamine) — higher cumulative anticholinergic exposure tracks with higher incident dementia (Gray 2015: prospective cohort of 3,434 adults 65+, median 7.3 years) + not recommended chronically. · Age 65+: all three classes above (benzodiazepines / Z-drugs / first-generation antihistamines) appear on the AGS 2023 Beers Criteria as potentially inappropriate medications — cognitive impairment, delirium, falls and fractures. · Newer orexin antagonists (suvorexant / daridorexant) — better choice + expensive.
· Week 1: caffeine cutoff at noon + blue-light control + fixed wake time. · Week 2: digital CBT-I app (Sleepio / Somryst / CBT-I Coach) + Mg + Gly before bed. · Week 3: bring the sleep diary to a CBT-I therapist, or use the sleep restriction module inside a digital program — do not compress time in bed on your own: daytime sleepiness rises sharply in the first week or two and the driving / machinery risk is real; anyone with epilepsy or a seizure history, uncontrolled bipolar disorder, untreated sleep apnea or a parasomnia especially should not self-administer it. Add progressive muscle relaxation alongside. · Week 4: evaluate → improving / refer to sleep medicine.
Accepting imperfection is one of CBT-I's core wisdoms: "Tonight I may not sleep well, but life will go on, and tomorrow will too."