Elmsword
← Sleep

Sleep · Research · 5 min read

Pink noise vs earplugs: earplugs protected deep sleep, pink noise cut REM

In a sleep-lab study of 25 healthy adults, pink noise did not offset intermittent environmental noise and cut REM sleep, while foam earplugs recovered more than 70% of the N3 deep sleep lost to noise.

Many people play a steady sound at night to cover traffic or other noise. A 2026 sleep-lab study tested that idea against plain foam earplugs. Intermittent environmental noise (EN), such as passing jets, cars and trains, mainly cut N3 deep sleep. Adding steady pink noise (PN) on top did not undo that loss and also cut REM sleep, so overall sleep structure got worse. Foam earplugs (EP) recovered most of the lost N3 sleep. This was a short lab study in young, healthy adults, so it can't say what happens over months at home. [1]

What the researchers did

This was a randomized, controlled, cross-over study in a sleep lab (trial registration NCT05774977). Twenty-seven healthy adults were enrolled. Two withdrew early and did not contribute data, so 25 were analysed: mean age 28.5 ± 5.9 years, seven male. None had used broadband sound during sleep before. Sleep was recorded with polysomnography, the overnight recording that sleep labs use to score sleep stages such as N3 and REM. [1]

The planned protocol comprised seven lab nights: a noise-free adaptation night followed by six conditions scheduled in a counterbalanced Latin square design. Not everyone completed all nights, and there were recording failures and condition substitutions: [1]

  • CTRL: a quiet control night.
  • EN: 93 intermittent noise events through the night, with maximum levels (LAS,max) of 45 to 65 dB. The sounds were jets, helicopters, drones, a low sonic boom, a car, a train, a baby crying and a fire alarm.
  • PN50: constant pink noise at 50 dB, with no EN.
  • EN + PN40 and EN + PN50: the same EN with constant pink noise at 40 or 50 dB.
  • EN + EP: the same EN while wearing standard foam earplugs, offered in three sizes.

All levels are A-weighted dB. Pink noise is one kind of broadband noise (BN), a sound spread across a wide range of frequencies; the paper lists white noise, fan or air-conditioner sounds and nature sounds as other examples. The average level across the night (LA,eq) was 23.7 dB on the quiet night, 43.2 dB with EN and 50.0 dB with PN50 alone. The earplugs lowered sound by 25.5 dB on average. [1]

Key findings

Estimated difference in minutes compared with the quiet control night (CTRL), Table 3. * Statistically significant after adjustment for multiple comparisons. EN + EP vs. EN: +16.9 min of N3. [1]
NightN3 deep sleep vs. CTRLREM sleep vs. CTRL
EN alone−23.4 min*−7.1 min
PN50 alone−4.5 min−18.6 min*
EN + PN40−21.2 min*−21.5 min*
EN + PN50−15.7 min*−29.0 min*
EN + EP (earplugs)−6.5 min−3.2 min
Pink noise vs earplugs against intermittent noisePlanned seven-night protocol: one noise-free adaptation night, then six conditions in Latin square order; not everyone completed all nights, and condition substitutions occurred. Compared with the quiet control night, EN alone cut N3 by 23.4 minutes; PN50 alone cut REM by 18.6 minutes; EN + PN40 cut N3 by 21.2 and REM by 21.5 minutes; EN + PN50 cut N3 by 15.7 and REM by 29.0 minutes; EN + EP differences (6.5 and 3.2 minutes) were not significant.Pink noise vs earplugs against intermittent noiseRandomized cross-over sleep-lab study, 25 healthy adults, polysomnographyA. Planned seven-night protocolNight 1noise-free adaptationNights 2–76 conditions, Latin square orderCTRL · EN · PN50EN + PN40 · EN + PN50 · EN + EPEN: 93 noise events, max 45–65 dBA · PN: constant pink noise, 40 or 50 dBA · EP: foam earplugsB. Minutes of N3 and REM sleep compared with the quiet control night (CTRL)N3 (deep sleep)REM sleepEN alone23.4 min less*7.1 min lessPN50 alone4.5 min less18.6 min less*EN + PN4021.2 min less*21.5 min less*EN + PN5015.7 min less*29.0 min less*EN + EP6.5 min less3.2 min less* Statistically significant vs. CTRL after adjustment. EN + EP vs. EN: +16.9 min N3.Short-term lab study in young, healthy adults new to broadband sound during sleep.Source: Basner et al., Sleep 2026, Table 3.
Source: Basner et al., Sleep 2026, Table 3. Bars show minutes of N3 or REM sleep lost compared with the quiet control night. [1]

Noise and pink noise hit different sleep stages. Compared with the quiet night, EN alone cut N3 deep sleep by 23.4 minutes on average, mostly swapped for lighter N2 sleep. Pink noise alone at 50 dB cut REM sleep by 18.6 minutes. [1]

“Rather than mitigating EN effects on sleep, PN further deteriorated sleep structure.”

Basner et al. · Statement of Significance · full text

Adding pink noise to EN made things worse overall. On EN + PN40 and EN + PN50 nights, both N3 and REM sleep were lower than on the quiet night, total sleep time was shorter, sleep efficiency was lower and time awake was longer. The louder pink noise did bring small benefits: a lower awakening frequency during environmental-noise playback in EN + PN50 than in EN + PN40 nights, not a demonstrated whole-night improvement. Those gains didn't offset the losses. [1]

“Thus, adding PN to EN clearly worsened sleep architecture instead of improving it.”

Basner et al. · Discussion · full text

Earplugs recovered most of the lost deep sleep. Sleep structure on EN + EP nights did not differ significantly from the quiet night. [1]

“Furthermore, more than 70% of the reduction in stage N3 in EN nights was recovered by wearing earplugs (+16.9 min, z = 0.59; p = .0003).”

Basner et al. · Results · full text

Earplugs had a ceiling. In exploratory event-related analyses, EN + EP did not differ significantly from CTRL at the tested event levels up to 55 dBA, but sleep responses remained at 65 dBA. These analyses were not adjusted for multiple testing and do not establish a universal earplug failure threshold. [1]

“Only at the highest LAS,max of 65 dB did effects of EN become visible in EN + EP nights.”

Basner et al. · Discussion · full text

Most people found them comfortable. Only two participants rated the earplugs as somewhat uncomfortable, and nobody rated them as very uncomfortable. [1]

“This study suggests that participants overwhelmingly found earplugs comfortable to wear, and the majority of subjects thought they slept better wearing earplugs.”

Basner et al. · Discussion · full text

What this doesn't tell us

The protocol planned one night per condition in a small group, and nobody had slept with broadband sound before. People who have used pink noise for months could respond differently. [1]

“This was a short-term exposure study in a relatively small sample of participants naïve to broadband sounds during sleep.”

Basner et al. · Limitations · full text

The participants were young and healthy, so the results may not apply to younger or older people, or to people who are not healthy. [1]

“We investigated a young and healthy population and thus findings may not generalize to younger, older, or non-healthy populations.”

Basner et al. · Limitations · full text

Only pink noise was tested, at two levels. White, brown or nature sounds were not. The lab loudspeakers did not reproduce the recorded noise perfectly, especially at very low frequencies, which the authors say limits how well this carries over to real bedrooms. [1]

REM sleep rose across the study nights. The authors can't tell whether that was REM rebound after nights of lost REM or people getting used to pink noise. [1]

“Unfortunately, the current study design does not allow us to disentangle REM sleep deprivation versus PN habituation effects.”

Basner et al. · Limitations · full text

No statistically significant between-condition differences were detected in morning cognition, cardiovascular measures or hearing. The study was not powered for the morning cardiovascular and cognitive outcomes, so this does not establish absence of effects or long-term safety. Mood and sleep ratings were worse after EN, EN + PN40 and EN + PN50 nights than after CTRL nights; this statement does not include EN + EP nights. So the study can't say whether these sleep-stage changes matter for long-term health. The research was funded by the U.S. Federal Aviation Administration (FAA) through ASCENT. [1]

What to take away

Opinion, not a study result: If I were in your spot and woken by street noise, I would try foam earplugs before buying a pink-noise machine. This study was short, in a lab and in young, healthy adults, so results may differ from person to person. If sleep problems persist, talk to a doctor.

Reference

  1. Sleep 2026[1] Basner M, Smith MG, Cordoza M, Kayser MS, Carlin M, Ecker AJ, Gilad Y, Park-Chavar S, Rennie K, Schneller V, Walsh S, Shou H, Cao Q, Younes M, Aeschbach D, Jones CW. Efficacy of pink noise and earplugs for mitigating the effects of intermittent environmental noise exposure on sleep. Sleep. 2026;49(5):zsag001.
    DOI: 10.1093/sleep/zsag001 · Full text

This article is for education only and is not medical advice. If sleep problems persist, talk to a doctor about your options.