Sleep · Evidence identifiers checked 2026-10-11 · Version 0.1.0 · Not yet reviewed by a clinician
Light in the Bedroom at Night: What the Studies Show
What one night of light during sleep did in a controlled study, what eye masks and earplugs changed, and what a standby light is and is not shown to do.
Most people sleep with some light on: a streetlight through the curtain, a standby LED, a hallway light left on for someone else. A controlled study of one night found that moderate light during sleep raised night-time heart rate, lowered heart rate variability and increased next-morning insulin resistance compared with a dim room. A separate study found ordinary room light before bedtime delayed melatonin onset in 99% of participants and shortened melatonin duration by roughly 90 minutes. In a simulated intensive-care setting, earplugs and eye masks gave more REM sleep and fewer arousals. These are small, short studies. They support making the bedroom darker as a cheap, low-risk change; they do not show that a standby light is causing disease.
Not medical advice. This is a plain-language summary of published research for generally healthy adults. It is AI-drafted, has not been reviewed by a clinician, and nothing here diagnoses or treats anything.
What the evidence supports
One night of moderate light during sleep raised night-time heart rate, lowered heart rate variability and increased next-morning insulin resistance compared with a dim room in a controlled study. c-01
Ordinary room light before bedtime delayed melatonin onset in 99% of participants and shortened melatonin duration by roughly 90 minutes. c-02
With earplugs and eye masks in a simulated noisy, lit environment, people had more REM sleep, reached REM sooner and had fewer arousals. c-03
These are short, small studies. They support darkening the bedroom as a low-cost change, not a claim that a standby light causes disease. c-05
One next step: Make one change tonight that removes light from your sleeping hours, then keep it for a week.
The short version
What one night of light did
The most direct evidence is a controlled study in which healthy adults slept one night with moderate light on and one night in a dim room. With light, night-time heart rate was higher, heart rate variability was lower, and next-morning insulin resistance was higher. Those are physiological markers measured over a single night, not a disease diagnosis, and the study was small. It is enough to say that light during sleep is not inert.
A controlled study of ordinary room light before bedtime found melatonin onset was delayed in 99% of participants and melatonin duration shortened by roughly ninety minutes. Melatonin timing is a marker rather than a health outcome, but it is the mechanism the light reviews point to: evening and night light delays the circadian phase, morning light advances it, and even low light levels at night can produce a response.
In practice, the room you sit in for the last hour of the evening is often brighter than the bedroom you sleep in.
Not every bedroom can be darkened. Curtains you do not own, a partner's reading lamp, a streetlight outside, a hallway light left on for a child: these are the usual constraints. In a simulated environment with noise and light, earplugs and eye masks produced more REM sleep, faster time to REM and fewer arousals. The setting was deliberately disruptive, so an ordinary bedroom should show a smaller effect, but an eye mask is cheap and reversible.
A review of artificial light at night describes circadian disruption in observational studies but does not quantify a safe threshold. Nothing in this batch links a standby LED to a disease outcome over years. The reasonable reading is that darkening the sleeping hours is a low-cost, low-risk change with a plausible mechanism and some short-term physiological support, and that anyone telling you a night light is ruining your health is going past the evidence.
Three common beliefs, checked against the studies behind this page.
Pick what you believe first; the evidence answer is revealed after, so you can see where your assumption and the studies part company. Nothing is scored against you.
Sleeping with a light on is harmless because your eyes are closed.
In a controlled study, one night of moderate light during sleep raised night-time heart rate, lowered heart rate variability and increased next-morning insulin resistance versus a dim room. Eyes closed does not make light inert, though one night of markers is not a disease finding.
Only bright light affects your body clock; a dim lamp does nothing.
A systematic review found even low light levels at night can produce a circadian response, and ordinary room light before bedtime delayed melatonin onset in 99% of participants in a controlled study.
Mixed to false. In a simulated noisy, lit environment, earplugs and eye masks produced more REM sleep, faster REM onset and fewer arousals. The setting was deliberately disruptive, so the effect in a normal bedroom is likely smaller.
One night of light during sleep changed heart rate and next-morning insulin resistance in a controlled study. Certainty: low. c-01
Wear an eye mask(1 min a night; $5-15)
Low
Eye masks with earplugs increased REM sleep and reduced arousals in a simulated lit, noisy environment. Certainty: low. c-03
Dim the last hour before bed(10 min a night; $0)
Low
Ordinary room light before bed delayed melatonin onset in 99% of participants and shortened its duration by about 90 minutes. Certainty: moderate. c-02c-04
Leave it and watch(0 min; $0)
Low
No study in this batch links bedroom light to long-term disease, so waiting is defensible. Certainty: moderate. c-05
Darken the room itself
Blackout curtains, blinds, or tape over standby LEDs. This removes the exposure rather than shielding you from it, which is the version the light-during-sleep study speaks to.
Cost, and you may not control the curtains in a rented room.
An eye mask works in rooms you cannot change. The study environment was deliberately disruptive, so expect a smaller effect at home, and give it a few nights before judging comfort.
Comfort and heat; some people cannot sleep with anything on the face.
Switch the main light off an hour before bed and use a low, warm lamp. This targets the pre-sleep exposure that the melatonin study measured, and it costs nothing.
Harder in a shared household; you may need one lamp that everyone accepts.
If you sleep well and the light is not disturbing you, there is no evidence here that demands action. Keep an eye on whether you wake unrefreshed and revisit if that changes.
No change now; you lose a cheap, low-risk experiment.
List the light sources in your bedroom and evening, rank them by when they are on, and print the two worth removing first.
Free: full audit of bedroom and evening light sources; the two actions to try first; text export and print; all citations and cautions Plus (optional, $5 one-time, price being tested): calendar (.ics) evening dimming reminders; unlimited share images; share card without the 'Made with CortexLit' footer Works offline. No account. Every source and caution stays free.
Build a three-stage dimming sequence counted back from your bedtime, with the clock time for each stage.
Free: full countdown from bedtime; stage-by-stage clock times; text export and print; all citations and cautions Plus (optional, $5 one-time, price being tested): calendar (.ics) reminders per stage; unlimited share images; share card without the 'Made with CortexLit' footer Works offline. No account. Every source and caution stays free.
Commercial disclosure: The companion apps have an optional paid Plus tier. All evidence, cautions, and the main takeaway are free. No affiliate links, no sponsors. Plus checkout isn't live yet.
Evidence cards
Each card links a claim from this page to its sources. "Certainty" is our informal editorial judgement of how sure we can be, based on study design, size, consistency, and how closely it fits everyday life. Sources were weighted by type: Cochrane and systematic reviews 1.0, guidelines 0.9, randomized trials 0.8, observational studies 0.5, lab, mechanism or animal work 0.2. That ranking isn't the same thing as certainty.
c-01 In a controlled study, one night of moderate light exposure during sleep raised night-time heart rate, lowered heart rate variability and increased next-morning insulin resistance compared with sleeping in a dim room.
Certainty: lowHuman trials (cause and effect)
Who was studied, what was found, and the limits
Who: Healthy adults sleeping one night in a laboratory Measured: Night-time heart rate, heart rate variability and next-morning insulin resistance Finding: Higher heart rate, lower variability, higher insulin resistance with light
Why this certainty: A controlled comparison with physiological measures, but one night in a small study, and mechanistic interpretation is limited. (Informal editorial assessment, not a formal GRADE rating.)
Limits: One night; laboratory; no long-term outcomes measured. Applies to: Why darkening the bedroom is worth trying, not a disease claim.
Identifier carried over from the 2026-10-11 abstract-level verification batch; not re-queried in this run. Source record
c-02 Exposure to ordinary room light before bedtime suppressed melatonin onset in 99% of participants and shortened melatonin duration by roughly 90 minutes compared with dim light.
Certainty: moderateHuman trials (cause and effect)
Who was studied, what was found, and the limits
Who: Healthy adults in a controlled light-exposure study Measured: Melatonin onset and duration Finding: Onset delayed in 99%; duration about 90 minutes shorter
Why this certainty: A controlled study with a strong, consistent hormonal effect, though melatonin timing is a marker rather than a health outcome. (Informal editorial assessment, not a formal GRADE rating.)
Limits: Melatonin measures are not health outcomes; the study used ordinary room light, not a specific lux threshold. Applies to: The hour before bed, which is where most people's light exposure actually is.
Identifier carried over from the 2026-10-11 abstract-level verification batch; not re-queried in this run. Source record
c-03 In a simulated intensive-care environment with noise and light, earplugs and eye masks led to more REM sleep, faster time to REM and fewer arousals; simulated noise and light alone worsened sleep.
Certainty: lowHuman trials (cause and effect)
Who was studied, what was found, and the limits
Who: Healthy volunteers in a simulated intensive-care setting Measured: Sleep architecture, melatonin and cortisol Finding: More REM, faster REM onset, fewer arousals with earplugs and eye masks
Why this certainty: Controlled, but in a deliberately disruptive simulated environment, so effects may be smaller in an ordinary bedroom. (Informal editorial assessment, not a formal GRADE rating.)
Limits: Simulated ICU conditions; short duration. Applies to: Eye masks and earplugs as a cheap option in rooms you cannot change.
Identifier carried over from the 2026-10-11 abstract-level verification batch; not re-queried in this run. Source record
c-04 A systematic review found light in the evening and at night delays the circadian phase while morning light advances it, and that even low light levels at night can produce a response.
Certainty: moderateHuman trials (cause and effect)
Who was studied, what was found, and the limits
Who: Studies pooled in a systematic review of light and circadian rhythm Measured: Circadian phase shifts Finding: Evening and night light delays phase; morning light advances it; low levels can still act
Why this certainty: A systematic review of a well-established physiological effect. (Informal editorial assessment, not a formal GRADE rating.)
Limits: Phase shifting is not the same as a health outcome; thresholds vary between people. Applies to: Why timing of light matters more than a single brightness rule.
Identifier carried over from the 2026-10-11 abstract-level verification batch; not re-queried in this run. Source record
c-05 The light-at-night studies in this batch are short and small, and an observational review of artificial light at night describes circadian disruption without quantifying a safe threshold; no trial in this batch linked bedroom light to long-term disease, so the evidence supports a low-cost change rather than alarm.
Certainty: moderateEvidence gap
Who was studied, what was found, and the limits
Who: The light-at-night evidence base used here Measured: Whether bedroom light is linked with long-term disease Finding: No such trial found in the sources used for this entry
Why this certainty: The available studies measure one-night physiology and associations, not outcomes. (Informal editorial assessment, not a formal GRADE rating.)
Limits: Absence of evidence in this batch is not proof none exists. Applies to: Framing: a cheap change worth making, not a hazard to fear.
Identifier carried over from the 2026-10-11 abstract-level verification batch; not re-queried in this run. Source record
c-06 Making an if-then plan improved prospective memory with a medium effect (d = 0.445) in healthy young adults, which is the basis for tying a darkening step to an existing evening cue.
Certainty: lowHuman trials (cause and effect)
Who was studied, what was found, and the limits
Who: Healthy young adults across 36 comparisons Measured: Prospective memory Finding: d = 0.445
Why this certainty: Meta-analysis of lab tasks. (Informal editorial assessment, not a formal GRADE rating.)
Limits: Lab tasks; not tested on evening light routines. Applies to: Supports the plan builder only.
Identifier carried over from the 2026-10-11 abstract-level verification batch; not re-queried in this run. Source record
Where studies disagree
How much light is too much: The controlled studies used moderate light during sleep and ordinary room light before bed, and both found effects; a systematic review notes even low light levels at night can produce a circadian response, while an observational review describes disruption without quantifying a safe threshold. There is no agreed brightness cut-off in these sources. (src-01, src-02, src-04)
Sources for this entry
Every identifier below was verified against PubMed on 2026-10-11 in an earlier batch. This run carried the identifiers over rather than re-querying them, and no retraction database was checked. Verification was at abstract level; full text was not read.
Mason 2022 Mason IC, Grimaldi D, Reid KJ, et al. Light exposure during sleep impairs cardiometabolic function. Proc Natl Acad Sci U S A. 2022;119(12):e2113290119. doi:10.1073/pnas.2113290119. PMID: 35286195. PMID 35286195 · mechanistic human lab · confidence weight 0.2 · abstract level
One night of moderate light during sleep raised night-time heart rate, lowered heart rate variability and increased next-morning insulin resistance compared with a dim room.
Gooley 2011 Gooley JJ, Chamberlain K, Smith KA, et al. Exposure to room light before bedtime suppresses melatonin onset and shortens melatonin duration in humans. J Clin Endocrinol Metab. 2011;96(3):E463-E472. PMID 21193540 · randomized controlled trial · confidence weight 0.8 · abstract level
Ordinary room light before bed delayed melatonin onset in 99% of participants and shortened melatonin duration by roughly 90 minutes.
Hu 2010 Hu RF, Jiang XY, Zeng YM, et al. Effects of earplugs and eye masks on nocturnal sleep, melatonin and cortisol in a simulated intensive care unit environment. Crit Care. 2010;14(2):R66. doi:10.1186/cc8965. PMID: 20398302. PMID 20398302 · non randomized trial · confidence weight 0.5 · abstract level
Simulated noise and light worsened sleep. With earplugs and eye masks, people had more REM sleep, reached REM sooner and had fewer arousals.
Tähkämö 2019 Tähkämö L, Partonen T, Pesonen AK. Systematic review of light exposure impact on human circadian rhythm. Chronobiol Int. 2019;36(2):151-170. PMID 30311830 · systematic review meta analysis · confidence weight 1.0 · abstract level
Light in the evening and at night delays the circadian phase; morning light advances it. Even low light levels at night can produce a response.
Cho 2015 Cho Y, Ryu SH, Lee BR, et al. Effects of artificial light at night on human health: A literature review of observational and experimental studies applied to exposure assessment. Chronobiol Int. 2015;32(9):1294-1310. PMID 26375320 · observational or preprint · confidence weight 0.5 · abstract level
Artificial light at night is associated with circadian disruption; the review is descriptive and does not quantify a safe threshold.
Chen 2015 Chen XJ, Wang Y, Liu LL, et al. The effect of implementation intention on prospective memory: a systematic and meta-analytic review. Psychiatry Res. 2015;226(1):14-22. doi:10.1016/j.psychres.2015.01.011. PMID: 25639373. PMID 25639373 · systematic review meta analysis · confidence weight 1.0 · abstract level
In healthy young adults, forming an if-then plan improved prospective memory with a medium effect (d = 0.445); combined verbal and imagery plans had a larger effect (d = 0.590).