Pre-Sleep Wind-Down Routine

Sleep · §189

The last hour before you sleep decides most of how the next day feels. Spend it on email, the phone, and the news, and you arrive in bed wired — sleep takes 30 to 60 minutes to come, the night runs shallow, and the morning starts in debt. Spend it on a fixed, repeatable sequence of low-stimulation things — dim light, a warm shower, a paper book, lights out at the same time — and you fall asleep faster, sleep more continuously, and wake into mornings you don't have to caffeine your way back from. The intervention is free, it takes a week or two to bed in, and it does its work through three independent mechanisms at once. The version of you that mainlines coffee to mask short, broken nights doesn't have to be the version that shows up at the desk.

Do · Daily Evidence Moderate Chapter Sleep

The cheapest sleep upgrade you can give yourself: a fixed sequence of low-stimulation things in the last hour before bed, repeated nightly. It works through three independent mechanisms — conditioning the bed to mean sleep, lowering the racing-thoughts and physical tension you'd otherwise drag into the night, and helping core body temperature drop the way sleep onset requires. The hard part is not the activities, which are pleasant. The hard part is claiming the hour back from work, the phone, and the news, and holding the line.

Three separate things have to happen for you to fall asleep easily. Most evenings that go badly are evenings where one of the three got blocked.

The bed has to mean sleep. Brains learn what places are for. If most of your evenings end with the phone in bed, work email at 11, or a half-finished argument with your partner, then the bed itself becomes a cue for being on alert — for thinking, for problem-solving, for vigilance. Sleep researchers have called this conditioned arousal for fifty years (Bootzin 1972), and it is why the same person can be exhausted on the sofa and instantly awake the moment their head hits the pillow. A fixed wind-down works because it teaches the bed the opposite lesson — same place, same low-key sequence, same clock time, every night, until the cues start pulling sleep toward you instead of pushing it away.

The racing-thoughts have to stop. Insomnia is well-described as a state of being too awake — cognitively (your mind is solving problems, rehearsing tomorrow, replaying the day) and physically (muscles tense, heart rate up, alertness high) (Harvey 2002). The two channels run independently, and both predict how long it takes you to fall asleep (Nicassio et al. 1985). Activities that drain those channels — slow breathing, a paper book, a quick brain-dump of tomorrow's three things — do the work the bed cannot do on its own.

The body's core has to cool. The signal that flips you from awake to asleep is a small drop in your core body temperature, driven by blood moving out to your hands and feet. A warm shower or bath an hour or two before bed pushes blood to the skin and accelerates that cooling, which is why the meta-analysis of warm-water bathing finds people fall asleep faster after one.

A wind-down routine is the cheap, simple thing that does all three at once. Nothing in it is novel by itself — humans have been dimming lamps and reading before bed forever. What is new is the framing that the routine itself is the intervention, not the activities inside it. The repetition is what conditions the bed; the activities are what defuse the arousal; the warm water (when included) does the temperature work. None of it requires you to be good at meditating, and none of it costs anything.

What the trials actually show

The honest version of the evidence: the wind-down has rarely been tested as one packaged thing in adults. What has been tested, repeatedly and well, is each piece of it.

Removing the screen hour before bed is one of the cleanest single-ingredient tests. A within-subject study in PNAS ran each participant through five nights of reading an e-book before bed and five nights of reading a paper book in the same dimmed conditions. The numbers were dramatic.

The bedtime-worry version of the evidence is just as crisp. People who write a five-minute to-do list for tomorrow before lights-out fall asleep faster — measurably, on polysomnography — than people who write about what they finished today; the more thoroughly the list off-loaded tomorrow's commitments, the bigger the gain (Scullin et al. 2018). The point is not the writing itself; it is that cognitive arousal is a real, measurable thing, and you can drain it.

Mindfulness components have been tested in head-to-head trials against sleep-education controls. In one randomised trial in older adults with moderate sleep disturbance, six weeks of mindfulness training improved sleep quality (and daytime fatigue) more than the active education control, by enough to move people across the threshold from poor to acceptable sleep (Black et al. 2015). A meta-analysis of 18 randomised mindfulness-for-sleep trials reached the same conclusion at the population level (Rusch et al. 2019).

What about the routine as a single package? The best-controlled trial of an explicit nightly routine is in young children — a two-week protocol of bath, then quiet activity, then bed reduced how long it took children to fall asleep, reduced their night wakings, and improved their mothers' mood as a side effect (Mindell et al. 2009). For healthy adults, the closest test is the meta-analysis of 47 behavioural sleep-improvement trials in non-clinical populations: a small-to-moderate improvement in sleep quality and ~5 to 10 minutes off sleep onset latency, with the bigger gains in trials that combined sleep-hygiene and relaxation components (Murawski et al. 2018).

The largest effect sizes in the literature come from cognitive behavioural therapy for insomnia (CBT-I), in which the wind-down's components — stimulus control, sleep hygiene, relaxation — are several of the active ingredients. A meta-analysis of 20 CBT-I trials in chronic insomnia found mean reductions of 19 minutes in time to fall asleep and 26 minutes in time spent awake during the night, with sleep efficiency rising 10 percentage points — effects equal to or larger than prescription sleeping pills and durable past the end of treatment (Trauer et al. 2015). CBT-I is first-line for chronic insomnia in every major guideline — the American College of Physicians (Qaseem et al. 2016), the American Academy of Sleep Medicine (Edinger et al. 2021), and the European Sleep Research Society (Riemann et al. 2017).

The honest summary: the components are well-supported and the mechanisms are clear. For a healthy adult sleeping badly, the realistic envelope is faster sleep onset by 5 to 15 minutes, lower pre-sleep arousal, and a felt improvement in next-day alertness within two weeks. For someone with chronic insomnia, the wind-down alone is not the treatment — the full CBT-I package is.

What the no-routine evening costs you

For most readers, the alternative to the routine is not nothing. It is the modal evening: the phone in bed, one more work message answered at 11.07, twenty minutes of scrolling that turn into forty, and a head that hits the pillow already rehearsing tomorrow's three difficult conversations.

Tonight. Sleep onset latency for a healthy adult is normally 5 to 20 minutes; the wired-into-bed version pushes it to 30 to 60 on the worse nights. Pre-sleep arousal is high — cognitive (problem-solving in the dark) and somatic (jaw tight, breath shallow) — and both independently delay sleep (Nicassio et al. 1985). When you do fall asleep, the night runs lighter, with more arousals; when the alarm goes off, total sleep time is short by 30 to 60 minutes because wake time stays fixed even when bedtime drifted late.

Tomorrow. The cost is the morning that starts with coffee as rescue rather than ritual, the 3pm crash that's actually a 1pm crash, the meeting where you watch yourself lose the thread halfway through. The Chang et al. e-reader study made this concrete — the screen-hour subjects performed measurably worse on next-morning alertness even with the same total sleep time (Chang et al. 2015). It is not just hours of sleep that matter; it is what kind.

This year. The same evening repeated three hundred nights stops being "a bad night" and becomes a base rate. The version of you that is irritable at the dinner table, that defers the harder project at work because you don't trust your own focus, that drinks more in the evenings because the day was draining — that version is partly downstream of the hour you keep handing to the phone. People around you read it before you do. Partners notice the tossing; colleagues notice the puffy Monday mornings; the friend who used to text at 11pm gives up because you replied incoherently or not at all.

The longer arc. Sleep is one of the catalogue's highest-leverage substrates — for mood reactivity, immune function, weight regulation, cognitive aging — and short, broken sleep across a decade trades against all of them. The wind-down does not single-handedly fix that arc; it is one of several inputs. But it is the input that costs nothing and works through three independent mechanisms, and skipping it means the arc bends the wrong way for free.

The actual sequence

The routine has three required pieces and a few high-leverage optional ones. The activities matter less than the consistency — same place, same sequence, same clock time, every night, until the bed starts pulling sleep toward you.

The minimum protocol.

About 60 minutes before lights-out: cut the cognitively activating input. Work email, news, social feeds, group chats, conflict-laden conversations — anything that engages problem-solving or threat-checking. Phones face-down or in another room is the cheapest way to enforce it. Dimming or grayscale is a half-measure; the content is the bigger problem, not the colour temperature.
The last 30 to 60 minutes: a low-arousal sequence, repeated. Anything genuinely low-stimulation. Paper book, light stretching, slow breathing, a quick brain-dump of tomorrow's three things, a shower, brushing teeth, a few pages of a journal. The activities can be boring — boring is the point. Pick a sequence and stop redesigning it.
Lights out at a fixed clock time, within about 30 minutes night-to-night. Wake time is the harder anchor (it sets the next night's sleep pressure), but a stable bedtime is what gives the wind-down a known endpoint.

High-leverage optional additions.

A warm shower or bath 1 to 2 hours before bed. Water at 40 to 42.5°C for 10 to 30 minutes, ending an hour or two before lights-out, is the sweet spot from the warm-bathing meta-analysis (Haghayegh et al. 2019). Hot enough that you exit pink-skinned; not so close to bed that you are still warm at lights-out.
Ten minutes of breathing or mindfulness, for racing-thoughts sleepers. Any guided practice will do; the meta-analytic evidence is for the modality, not the brand (Rusch et al. 2019).
A five-minute to-do list at lights-out, if your sleep onset is wrecked by tomorrow-rehearsing. Write tomorrow's three things on paper and put it on the bedside table. The polysomnography study found the more specific the list, the bigger the gain (Scullin et al. 2018).

Give the routine 10 to 14 days before judging it. The bed-as-sleep-cue mechanism is a conditioning effect, and conditioning needs repetition. The children's-routine trial got clear results at two weeks (Mindell et al. 2009); the adult literature is consistent with a similar timescale. Sporadic execution is essentially zero — five good nights followed by a phone-in-bed Sunday undoes the previous week.

What most guides get wrong

Sleep hygiene is not a treatment for insomnia. A widespread framing — in health-class posters, GP handouts, wellness blogs — presents the wind-down as the answer to insomnia. For functionally normal sleepers it improves things modestly. For diagnosed chronic insomnia disorder (sleep difficulty 3+ nights a week for 3+ months with daytime impairment), it underperforms full cognitive behavioural therapy and in head-to-head trials does about as well as the attention-control conditions (Edinger et al. 2021). If your sleep is genuinely broken, the wind-down helps, but the treatment is CBT-I.

The blue-light filter is not the fix. The screen-before-bed problem has two channels: light wavelength (the melatonin-suppression mechanism) and content (the cognitive arousal mechanism). The blue-light filter addresses only the first. A grayscale Twitter feed at midnight is still a problem because what is doing the work is the engagement, not the photons (Hale & Guan 2015). Behaviour change is the larger lever; light filters are a small correction on top.

Trying harder makes it worse. Sleep cannot be willed. If you treat the wind-down as a performance — checking the clock, grading your sleep tracker, lying in bed monitoring whether sleep is "working yet" — you have recreated the cognitive arousal the routine was meant to drain (Harvey 2002). The routine works because the cues do the work for you. Your job is consistent execution, not consistent striving.

The activities do not have to be impressive. The wellness genre often prescribes elaborate ten-step routines with gratitude journals and tongue scrapers and meditation timers. The trial evidence is that the activities mostly do not matter — what matters is that they are low-arousal and that you do them in the same order at the same time every night. A boring, repeatable routine you actually do beats a beautiful one you keep tweaking.

Where this goes wrong in practice

Routine in name only. The most common failure is high variance. The reader has a "routine" — but bedtime swings an hour and a half, the activities rotate, the phone reappears on Thursdays. The conditioning mechanism needs repetition; without it, the bed never learns what it is for, and none of the three mechanisms in the routine engage properly. If you cannot repeat the sequence at least five nights out of seven, you do not yet have a routine — you have an intention.

The pre-routine evening is hot. An intense workout that ends 30 minutes before bed, a heated argument at 10pm, a final work email at 10.45 — any of these can put your body and mind into a state that thirty minutes of low-stimulation activity cannot undo. Evening exercise broadly improves sleep, but vigorous training that ends within an hour of bedtime can extend sleep onset in the average case (Stutz et al. 2019). The wind-down works best when the hour leading into it is not a fire.

The household is incompatible. Routines that do not fit a shared household get unwound. A partner who watches news in bed, a roommate who comes home loud at 11.30, a child who wakes at irregular times — these are friction the protocol cannot just override. The fix is usually to negotiate the household rather than to try harder on the routine; with children, the adult routine typically anchors downstream of the child's routine (Mindell et al. 2009).

Treating the wind-down as a performance. A reader who buys a sleep tracker, scores their sleep onset latency every night, and grades their wind-down execution has reintroduced cognitive arousal under a new label. The studies that work on this (Harvey 2002) consistently find that effortful self-monitoring around sleep makes sleep worse. Set up the routine, do the activities, and stop measuring.

Quitting after a week. The conditioning effect takes 10 to 14 nights to start producing reliable returns. A reader who tries it for four nights, decides it "doesn't work," and goes back to the phone is judging the intervention before it can deliver. Give it two clean weeks.

If your evenings aren't your own

The protocol above assumes a reader with control over the hour before bed. A real share of readers does not have that hour — newborns, two jobs, caregiving, shift work, on-call schedules. The honest framing for those readers is: the optimum is out of reach this season; the harm-reduction version still works.

New parents are the worked case. Build the adult wind-down downstream of the child's bedtime routine — the child's routine is already a fixed sequence at a fixed time, which is half the protocol for free. Once the child is down, claim 15 to 20 minutes for the simplest version: no phone, one dim lamp, paper book or quiet stretch, lights out. Sleep is going to be broken regardless of how good the wind-down is; the routine's job in this season is to make the limited time you get count, not to optimise hours you don't have.

Shift workers and on-call schedules need the routine more, not less, because the rest of their sleep architecture is fighting them. The conditioning works regardless of clock time — same activities, same sequence, before whatever counts as "bed" — and matters most because the circadian system is not helping.

For everyone else, the practical block is rarely the activities. It is the decision to claim the hour from late-evening work, late-evening entertainment, and late-evening social demand, and to hold the line through the predictable pressure to give it back. The 9.30pm work email that "won't take five minutes," the show that "is just one more episode" — these are the small daily negotiations the routine has to win, and the only way they get won is by deciding once instead of every night.

Nothing about the routine itself is unsafe. Two practical caveats worth naming:

If you have chronic insomnia — sleep difficulty most nights for three months or more, with daytime impairment — the wind-down on its own is not the treatment. The first-line treatment is cognitive behavioural therapy for insomnia (CBT-I), which adds sleep restriction, cognitive therapy, and stricter stimulus-control instructions to the routine's ingredients. CBT-I is what the American College of Physicians, the European Sleep Research Society, and the American Academy of Sleep Medicine all recommend ahead of sleeping pills (Qaseem et al. 2016) (Riemann et al. 2017) (Edinger et al. 2021). Ask your clinician about CBT-I, or look for digital CBT-I programs.

If you snore loudly, your partner has seen you stop breathing, or you wake unrefreshed despite long time in bed, the issue may be sleep apnea, not behaviour. No routine fixes apnea — it needs evaluation.

What changes, and when

The first week. The hour before sleep starts to feel different — less leftover-anxiety, more cozy-ritual. You notice you are not dreading bedtime; the warm shower and the paper book are pleasant in their own right. Sleep onset is variable still — the conditioning has not had time to settle — but the racing-thoughts at lights-out are quieter.

Two weeks in. The bed-as-sleep-cue mechanism starts producing reliable returns. You fall asleep faster — by 5 to 15 minutes for healthy adults (Murawski et al. 2018) (Haghayegh et al. 2019), more if you were starting from a wired-into-bed baseline. The night runs more continuously — fewer arousals, more consolidated sleep. The mornings change first: you wake on the same alarm, but the first ten minutes of the day are less of a fight. Coffee becomes the ritual instead of the rescue.

One to three months. The 3pm crash that used to demand sugar and a second coffee shows up less reliably. Afternoon meetings run without the glassy stare. The work that needs sustained attention gets done — not because you got smarter, but because the focus stopped being rationed by a short night (Chang et al. 2015). People around you start to comment. The partner who used to be woken by your tossing sleeps through. The colleague says you look well on Monday morning. Pre-sleep arousal has dropped enough that you are not bringing the workday's threat-checking into bed (Black et al. 2015).

The year. The base rate of how you feel has moved. The hour you used to hand to the phone is now an hour you look forward to. The version of you that mainlined coffee and white-knuckled the afternoons recedes into the past — not because you took anything, not because you bought anything, but because you gave yourself an hour of evening every night and the body did the rest. The compounding effect is the part that is hard to see from the start: it is not one good night, it is three hundred and sixty-five of them, and the trajectory at the end of the year is different.

The honest ceiling: for readers whose sleep is already good, the delta is smaller. For readers with chronic insomnia, the wind-down on its own will not finish the job — the full CBT-I package will. For most readers in the middle, the realistic envelope above is what the research supports, and the modesty of the per-night gain is part of why it works at all — it stays sustainable because it does not ask much.

Related territory worth knowing about. Morning light exposure is the other end of the same circadian system — a few minutes outdoors in the first hour after waking does for the morning what the wind-down does for the evening. Caffeine timing is the daytime input that frequently sabotages an otherwise-good evening protocol. If your sleep is genuinely broken on most nights despite a clean wind-down, the full cognitive behavioural therapy for insomnia (CBT-I) is the treatment. And if you snore loudly or wake unrefreshed after long nights, suspected sleep apnea is its own track entirely, and no behaviour change substitutes for testing.

Substance and claimed effects

A pre-sleep wind-down routine is a fixed, repeatable sequence of low-stimulation activities performed in the hour (roughly 30–90 min) before bed: dimming household light, shutting off screens or switching them to grayscale/night mode, a warm shower or bath, a brief grooming/hygiene set, a quiet activity (paper book, stretching, slow breathing, journaling), and lights-out at a fixed clock time. It is the operational expression of three overlapping clinical tools used in CBT-I: stimulus control (the bed-cues-sleep conditioning logic of Bootzin 1972), sleep hygiene (Irish et al. 2015), and pre-sleep arousal reduction (Harvey 2002; Nicassio et al. 1985). Claimed effects, all covered holistically by this entry: (1) reduced sleep onset latency — falling asleep faster; (2) reduced pre-sleep cognitive and somatic arousal — less racing-thoughts and physical tension in bed; (3) improved subjective and objective sleep quality — fewer awakenings, more consolidated sleep; (4) improved next-day alertness, mood, and cognitive performance, mediated mostly through (1)–(3); (5) downstream contributions to energy, focus, mood, and the long-run health and beauty effects that any sleep improvement carries. The recommendation applies to functionally normal adults; chronic insomnia disorder is managed with the full CBT-I protocol (Qaseem et al. 2016; Edinger et al. 2021), of which the wind-down is one ingredient.

Evidence by addressing question

Mechanism

Three mechanisms converge on the hour before bed.

Conditioned arousal and stimulus control. Bedrooms and bed-time activities become conditioned stimuli: the brain pairs whatever you regularly do at bedtime with the physiological state you arrive in. If most evenings end with email, doomscrolling, or a half-finished argument, the bed itself becomes a cue for vigilance rather than for sleep — the central insight of stimulus-control therapy (Bootzin 1972) and the proximal target of a fixed wind-down: same place, same low-arousal sequence, same clock time, repeated until the cues pull sleep forward.

Pre-sleep arousal. Insomnia is well-modelled as a state of cognitive and physiological hyperarousal (Harvey 2002; Kalmbach et al. 2018). The Pre-Sleep Arousal Scale separates cognitive arousal (racing thoughts, problem-solving, intrusive plans) from somatic arousal (muscle tension, heart rate, perceived alertness) and both independently predict sleep onset latency (Nicassio et al. 1985). Wind-down activities — mindfulness, slow breathing, journaling, low-arousal reading — act directly on these two channels. The Scullin et al. 2018 polysomnography RCT made the cognitive-arousal mechanism unusually crisp: subjects who wrote a five-minute to-do list at bedtime fell asleep faster than those who wrote about completed tasks, with greater offloading associated with shorter sleep onset latency.

Thermoregulatory sleep-gating. Sleep onset is preceded by a 0.3–0.5°C drop in core body temperature, driven by distal vasodilation (hands and feet warm, core cools). A warm shower or bath roughly 1–2 h before bed flushes blood to the skin, accelerates that core-temperature fall, and shortens sleep onset latency — the meta-analytic finding of Haghayegh et al. 2019 covering 17 studies and 13 trials: water at 40–42.5°C, 10–30 min, 1–2 h before bed reduced sleep onset latency by about 8.6 min and improved sleep efficiency. Evening exercise produces a related effect via post-exercise vasodilation (Stutz et al. 2019) provided it ends ≥1 h before bed.

Light and screens. Evening light exposure — particularly short-wavelength blue light from screens at typical reading distance — suppresses melatonin and phase-delays circadian timing. Chang et al. 2015 ran a within-subjects cross-over: 5 nights of light-emitting e-reader reading vs 5 nights of paper book in identical otherwise-dim conditions delayed circadian phase by ~1.5 h, suppressed evening melatonin by >50%, lengthened sleep onset latency by ~10 min, reduced REM, and produced measurably greater next-morning sleepiness despite identical total sleep time. Screen content adds a second arousal channel orthogonal to light: the cognitive activation of news, social feeds, and work email (Hale & Guan 2015).

Evidence

The wind-down routine is rarely tested as a single packaged intervention in adults; the evidence base is the union of its components plus the CBT-I literature where the routine is one of several active ingredients.

CBT-I as a whole. The Trauer et al. 2015 meta-analysis of 20 RCTs of CBT-I in chronic insomnia found mean reductions of 19 min in sleep onset latency and 26 min in wake after sleep onset, with sleep efficiency rising 10 percentage points — effect sizes equal to or larger than hypnotic medication and durable past treatment end. Stimulus control and sleep hygiene (the wind-down's core) are standard components in every CBT-I protocol; both the American College of Physicians (Qaseem et al. 2016), the European Sleep Research Society (Riemann et al. 2017), and the AASM (Edinger et al. 2021) recommend CBT-I as first-line treatment for chronic insomnia, ahead of medication.

Behavioural interventions in non-disordered adults. The Murawski et al. 2018 meta-analysis of 47 trials of cognitive and behavioural sleep interventions in adults without diagnosed sleep disorders found small-to-moderate improvements in sleep quality (SMD ~0.39) and reductions in sleep onset latency (~5–10 min) — smaller than in clinical insomnia but real, with sleep-hygiene-plus-relaxation interventions performing best.

Single components. Warm bathing/showering 1–2 h before bed reduces sleep onset latency by 8.6 min in healthy adults (Haghayegh et al. 2019). Mindfulness meditation improves sleep quality with moderate effect sizes (SMD ~0.33) across 18 RCTs (Rusch et al. 2019); the Black et al. 2015 JAMA Internal Medicine RCT in older adults with moderate sleep disturbance found mindfulness training reduced Pittsburgh Sleep Quality Index by 2.8 points vs 1.1 in the sleep-education control, with parallel reductions in daytime fatigue. Bedtime worry-offloading (5-min to-do list) shortened sleep onset latency from 25 to 16 min in a controlled polysomnography study (Scullin et al. 2018). The Finucane et al. 2021 pragmatic online RCT (n=991) of reading a book in bed vs not reading found a modest but significant improvement in self-reported sleep quality in the reading arm. Evening screen displacement — replacing 30–60 min of pre-bed device use with non-screen activity — produces effects of similar magnitude to the Chang et al. 2015 cross-over: ~10 min shorter sleep onset latency and better next-day alertness when the screen hour is removed.

Children. The strongest single-package RCT of a routine per se is Mindell et al. 2009 in young children (n=405): instituting a three-step bath/massage/quiet-activity routine over two weeks produced significantly shorter sleep onset latency, fewer night wakings, and improved maternal mood. The adult evidence is more diffuse, but the routine principle (consistent cues, low-arousal sequence, fixed timing) transfers.

Quality of the evidence. Mechanism is well-characterised. Component RCTs are common and broadly consistent. The package as a single trial-tested intervention in healthy adults is rarer than the components — most adult trials test CBT-I (multi-component) or a single ingredient. Effect sizes are real but modest in non-clinical adults, larger in clinical insomnia.

Protocol

The minimum viable protocol is three things, repeated nightly at the same clock time:

A hard stop on cognitively activating input roughly 60 min before lights-out: work, news, social feeds, group chats, conflict-laden conversations. The Chang et al. 2015 finding gives the upper-bound case (full e-reader use suppresses melatonin >50%); dimming screens, switching to paper, or moving to non-screen activity captures most of the gain.
A low-arousal sequence in the last 30–60 min: any low-arousal-low-light activity will do (paper reading, light stretching, slow breathing, a brief gratitude/to-do list per Scullin et al. 2018). The activities matter less than their cognitive-arousal level and the consistency of the sequence; novelty defeats the conditioning.
A fixed lights-out time within ~30 min night-to-night. Wake time is the more important anchor in stimulus-control protocols (it sets the next night's sleep pressure), but a stable bedtime is what gives the wind-down a known endpoint.

Optional, high-leverage additions: a warm shower/bath 1–2 h before bed at 40–42.5°C for 10–30 min (Haghayegh et al. 2019); 10 min of mindfulness or paced breathing for cognitive-arousal-prone sleepers (Rusch et al. 2019); a 5-min to-do offload at lights-out for racing thoughts (Scullin et al. 2018). The package is permissive; the discipline is in the timing and the repetition.

Contraindications

No contraindications in the closed vocabulary apply. The wind-down is non-pharmacological and low-stakes in itself. Practical caveats: chronic insomnia disorder (sleep difficulty ≥3 nights/week for ≥3 months with daytime impairment) is not adequately treated by sleep hygiene alone; the consensus across Qaseem et al. 2016, Riemann et al. 2017, and Edinger et al. 2021 is that sleep hygiene as standalone monotherapy underperforms multi-component CBT-I, which adds sleep restriction, cognitive therapy, and structured stimulus-control instructions. Suspected sleep apnea (loud snoring, witnessed apneas, daytime sleepiness despite adequate time in bed) needs evaluation, not a better routine.

Misconceptions

Three persistent ones.

Sleep hygiene = sleep treatment. Public health framings often present sleep hygiene as the insomnia intervention. As monotherapy in chronic insomnia it underperforms full CBT-I and is roughly equivalent to attention-control conditions in head-to-head trials (Edinger et al. 2021). For healthy adults it improves sleep modestly (Murawski et al. 2018); for chronic insomnia it is a useful adjunct, not a treatment.

Blue-light filters fix the screen problem. The Chang et al. 2015 mechanism is light-mediated, but the cognitive-arousal mechanism of screen content (Hale & Guan 2015) is orthogonal to wavelength. A grayscale Twitter feed at midnight still suppresses sleep onset via engagement. The blue-light filter is a small correction; behaviour change is the larger one.

Trying harder makes it work. Effort and self-monitoring in bed are forms of cognitive arousal (Harvey 2002): readers who try to force sleep, measure their sleep latency, or grade their wind-down execution generally sleep worse. The routine works because the cues do the work; the reader's job is consistent execution, not consistent striving.

Failure-modes

Common ways the routine fails in practice:

Routine in name only. The reader nominally has a wind-down but the variance is high — bedtime swings 90 min, the activities change, the phone reappears. Conditioning needs repetition; without it, none of the mechanisms above engage.
Pre-routine activation. A high-arousal evening (intense workout 30 min before bed, work email at 11pm, conflict) overrides whatever low-stimulation thirty minutes follow. Stutz et al. 2019 finds late vigorous exercise tolerable for trained populations but raises sleep onset latency in the average case if it ends within an hour of bed.
Wind-down as a second shift. Caregivers, parents of young children, shift workers, and people with two jobs frequently cannot construct a fixed hour. The routine still helps in shortened form — ten minutes of consistent transition is better than zero — but the optimal protocol presumes some control over the late evening.
Sleep effort. Treating the wind-down as a performance, with sleep-tracker grades and pressure to fall asleep within X minutes, recreates the arousal it was meant to defuse (Harvey 2002).

Stakes

Without a wind-down, the modal evening for the modal reader ends in the activating zone: phone in bed, last work message answered after 10pm, racing-thoughts inventory of tomorrow's tasks. Pre-sleep arousal rises (Nicassio et al. 1985; Harvey 2002); sleep onset latency climbs from a normal 5–20 min to 30–60 min on bad nights; total sleep time loses 30–60 min on the back end as wake time stays fixed; sleep architecture skews lighter (less consolidated, more arousals); next-day alertness, mood, and cognitive performance pay the bill (Trauer et al. 2015; Chang et al. 2015). Compounded across years, this is the trajectory the catalogue's other sleep entries care about: cumulative sleep debt, mood reactivity, weight regulation, immune function.

Payoff

The realistic envelope, anchored in the meta-analyses: 5–15 min faster sleep onset in healthy adults (Murawski et al. 2018; Haghayegh et al. 2019), measurably lower pre-sleep arousal (Scullin et al. 2018; Rusch et al. 2019), better-rated sleep quality (SMD ~0.39 in Murawski et al. 2018), and a felt improvement in next-day alertness (Chang et al. 2015; Black et al. 2015). Onset: a stable routine takes 1–2 weeks to start producing reliable conditioning effects (Mindell et al. 2009 showed clear effects at 2 weeks in children). For clinical insomnia, the larger CBT-I package produces 19-min sleep-onset improvements and 10-point sleep-efficiency gains (Trauer et al. 2015).

Practicalities

Cost is effectively zero (a paper book, an alarm to start the routine). Effort is real but bounded: 30–60 min of evening discretion. The hardest practical block is not the activities but the decision to claim the hour from late-evening work, social, and entertainment demands. Partners and household-mates make a difference: routines that don't fit a shared household struggle. For parents of young children the pediatric-routine literature (Mindell et al. 2009) suggests building the adult routine downstream of the child's bedtime routine.

Out-of-scope

The wind-down routine intersects with several adjacent topics that this entry treats as forward-pointing pointers rather than as in-scope content: evening light hygiene and morning light exposure (the circadian entry point), suspected sleep apnea (medical evaluation, not behavioural), CBT-I as a full multi-component treatment (clinical insomnia), naps and napping protocols (separate timing question), caffeine timing (a pre-evening question that overlaps but is its own substance).

The credibility range

Optimist case. A pre-sleep wind-down routine is essentially free, low-effort, mechanism-coherent across three independent pathways (conditioned arousal, hyperarousal reduction, thermoregulation), backed by component RCTs and large meta-analyses, and is the active ingredient in the first-line non-drug treatment for the most common adult sleep complaint. Sleep is among the highest-leverage substrates for everything the catalogue cares about — cognition, mood, longevity, weight regulation, immune function — and the wind-down is the single behaviour-change with the most evidence per dollar/minute. Recommend universally.

Skeptic case. Sleep hygiene as a packaged intervention has consistently underperformed CBT-I in head-to-head trials and is no better than attention controls in clinical insomnia (Edinger et al. 2021). Adult trials of the wind-down per se are scarce; most "evidence" is component-level. Effect sizes in non-clinical populations are modest (~5–10 min sleep onset, small-to-moderate sleep quality SMDs in Murawski et al. 2018) — respectable but not transformative. The routine probably gets disproportionate cultural attention because it's actionable and free, not because its effect dominates. Real sleep problems need real treatment.

Author's call. Both cases hold. The wind-down is a default intervention for a healthy adult who wants better sleep without medication; it is not a cure for chronic insomnia disorder, which needs full CBT-I. Score the evidence as solid (multiple meta-analyses on components, clinical guidelines on the package as part of CBT-I) but flag the effect sizes honestly: modest in healthy adults, substantial as part of CBT-I in clinical populations. The article should advertise the routine to the healthy reader as a high-leverage, low-cost win and route clinical insomnia readers toward CBT-I.

Stakeholder + incentive map

Sleep clinicians and AASM/ESRS/ACP guideline bodies: support the wind-down as a component of CBT-I; consistently warn against sleep hygiene as standalone monotherapy for chronic insomnia.
Sleep-tech industry (wearables, smart lights, sleep-tracking apps): commercial incentive to frame the wind-down through their devices — the routine becomes a feature pitch (smart-bulb dimming schedules, wind-down notifications, breathing apps). Useful framing aids in some cases; the routine itself is free and analog.
Pharmaceutical industry (hypnotics, OTC sleep aids): counter-incentive — non-drug behavioural interventions reduce the drug market.
Wellness influencers and sleep coaches: enthusiastic boosters, often overstating effect size and prescribing rigid 10-step protocols where the evidence supports flexibility and consistency.
Public health bodies (CDC, NHS): include wind-down/sleep-hygiene messaging in general sleep guidance; broadly aligned with the moderate effect-size literature.

Population variability

Heterogeneity is real. Cognitively-aroused sleepers (worry-driven, work-anxious, ruminative) get the largest gains from cognitive-load-shedding components — mindfulness, journaling, paced breathing (Black et al. 2015; Scullin et al. 2018). Somatically-aroused sleepers (muscle tension, restless) respond more to thermoregulation and physical-relaxation components (Haghayegh et al. 2019). Older adults with mild sleep disturbance respond well to mindfulness-based routines (Black et al. 2015). Shift workers, new parents, and caregivers have constrained evening control; the abbreviated 10-min routine is the relevant intervention, not the optimal 90-min one. Clinical insomnia patients respond more to full CBT-I than to wind-down alone (Edinger et al. 2021). Adolescents have stronger circadian phase delay; screen-curfew is more impactful than the rest of the routine (Hale & Guan 2015). Healthy sleepers — people whose sleep is already good — gain less in absolute terms but gain stability and resilience to bad nights.

Knowledge gaps

What hasn't been adequately tested: the wind-down as a single packaged intervention in healthy adults with a true active-control arm (most studies test components or test CBT-I bundles). Component dismantling within the routine — how much variance is explained by thermoregulation vs cognitive arousal vs stimulus-control conditioning — is under-quantified. Optimal duration (10 vs 30 vs 60 min) lacks head-to-head data. Generalisation across cultural contexts where late-evening social and work norms differ has not been systematically studied. Whether a wind-down can prevent insomnia onset in at-risk populations (high sleep reactivity per Kalmbach et al. 2018) is plausible but unstudied. What would change the call: a large multi-arm RCT comparing healthy adults randomised to specific wind-down components vs a credible attention control with PSG endpoints would tighten the dose-response question.

Brief framing. The brief named four consequences (sleep onset latency, pre-sleep arousal, sleep quality, next-day alertness). All four are covered: mechanism and evidence cover onset latency and arousal directly; stakes and payoff cover quality and next-day alertness in felt-experience voice. No narrowing relative to the brief.

Sleep scored 3, not 4. The substantial-to-transformative sleep gains in the literature come from full CBT-I in clinical insomnia (Trauer et al. 2015: 19 min faster onset, 26 min less wake-after-onset). The wind-down as one component, in healthy adults, produces meta-analytic effects in the 5–15 min onset range (Murawski et al. 2018; Haghayegh et al. 2019) and small-to-moderate sleep-quality SMDs. That's a clear, named effect — 3 — not a substantial transformation. Calling it 4 would conflate the wind-down with the larger CBT-I package.
Evidence scored 4, not 5. Component meta-analyses are strong and guidelines name stimulus control and sleep hygiene as first-line CBT-I ingredients; the wind-down as a single packaged intervention in healthy adults has thinner direct RCT evidence. 4 is honest.
Beauty dimensions kept at 1. Better sleep does improve appearance (less puffy, better skin tone, slower-aging trajectory), but the routine's marginal contribution beyond the rest of any reader's sleep stack is small. Scoring 2 or 3 would inflate the routine's specific aesthetic claim.
Effort_burden scored 2, not 3. The activities are mostly displacement of what readers would be doing anyway (relaxing in the evening); the friction is the willpower of holding the line on the phone and late-evening work, not the time itself. "Mild lifestyle shift" maps closer than "sustained willpower" — but the call was borderline.
Routing clinical insomnia readers to CBT-I. The misconceptions section and the contraindications warning callout both name CBT-I explicitly. The temptation to oversell sleep hygiene as "the" insomnia treatment is the dominant cultural framing; the article deliberately pushes against it, which is the responsible call even at some cost to the entry's apparent power.
Audience scoping removed. An earlier draft wrapped the new-parents passage in data-audience-ages="18-39"; new parents span all adult age bands, so the scoping was misleading. Left the content in the audience+practicalities section as plain prose.

Future links. When the following entries exist, this one should cross-link to them: morning light exposure (the circadian-pair entry the out-of-scope section names), caffeine timing, CBT-I, sleep apnea screening / STOP-BANG, naps, fixed wake time. The out-of-scope section is written to make those wires easy to add.

Separate-entry candidates. Two pieces of territory the dossier touched but did not absorb: (1) the screen-curfew question deserves its own entry — Chang et al. 2015 is one of the cleanest single-RCT mechanisms in the sleep literature and the topic is broader than wind-down; (2) bedtime worry-offloading / journaling deserves a stand-alone entry on the Scullin et al. 2018 evidence base if a journaling entry is in scope.

Rating difficulty. The hardest call was sleep at 3 vs 4. Settled on 3 because the wind-down's specific contribution beyond the rest of the catalogue's sleep stack is meaningful but not dominant, and 4 ("substantial sleep transformation") would be borrowing credit from CBT-I.