Headphone volume

A systematic review and meta-analysis in BMJ Global Health searched for studies of listening habits in 12- to 34-year-olds published between 2000 and 2021. Thirty-three studies met the bar, covering 35 records and 19,046 individuals in total. Of those, the 17 records that concerned personal listening devices covered 8,987 people - the rest were studied for loud venues - and across those 8,987 the pooled prevalence of exposure to unsafe listening from devices came out at 23.81% (95% CI 18.99% to 29.42%). It is worth being exact about what that means, because the number travels badly. It is a count of a BEHAVIOUR, not of an injury. Not one audiogram contributed to it. Nobody inside that 23.81% has been shown to have lost any hearing; they have been shown to be listening at levels the included studies classified as unsafe. What "unsafe" meant also varied - some studies measured the sound coming out of the earphones, others asked people about their own volume and hours, and self-reported loudness is a poor proxy for decibels. So the honest reading is that roughly a quarter of young people are running an exposure that standards would not permit, over two decades of studies from a limited set of countries.

Dillard LK et al. 2022

The prevalence figure says nothing about outcomes, so the second question is whether the hearing of long-term device users measurably differs. A systematic review and meta-analysis in the International Journal of Environmental Research and Public Health split 16 studies into two groups: 7 that measured hearing before and after a listening session, and 9 that compared habitual device users with non-users. The short-term group found no statistically significant change - the largest effect was at 4 kHz and its confidence interval crossed zero, and one of the included studies reported that such shifts had recovered within a few hours. That recovery window comes from that single study, not from anything measured across the pool; the review's own summary says only that after temporary shifts "their hearing is restored". The long-term group did find a difference. Pooled across those 9 studies, users had worse pure-tone thresholds overall (Hedges' g -0.419, 95% CI -0.632 to -0.207), with the gap reaching significance at 6 kHz (-0.525, 95% CI -0.897 to -0.154) and 8 kHz (-0.486, 95% CI -0.819 to -0.152). Hedges' g is a standardised difference between groups, not decibels, and the review does not convert it - so nobody can say from this how many decibels of hearing anyone has lost. Two more caveats belong here. The studies disagreed substantially at 6 kHz, which is where the largest estimate sits. And the groups were sorted by questionnaire, so this compares people who say they use headphones a lot with people who say they do not.

You S et al. 2020

This is the advice almost everyone has heard, and it is worth saying plainly what we found when we went looking for its source. Searching the peer-reviewed literature for a "60/60 rule" returns papers about COCHLEAR IMPLANT REFERRAL - an unrelated clinical criterion that happens to share the nickname, based on a 60 dB HL hearing level and a 60% word score. The headphone version appears in clinic blogs, news segments and public-awareness material; we could not trace it to a measurement, a trial or a standard. There is also a physical problem with it. "60% of maximum" is not a sound level. Maximum output differs between phones, between earphones and between the two combined, so the same slider position on two setups is not the same sound arriving at the ear. A rule stated as a fraction of an unknown quantity cannot be checked by anyone. That does not make the underlying advice wrong - turning it down and taking breaks are sensible - but the two numbers are not evidence, and this page does not carry them. Which is the practical reason not to use it: maximum output differs from device to device, so the same percentage is a different sound on every set of headphones.

Dillard LK et al. 2022

There is a real, measured standard, and it is written in decibels and hours rather than in percentages. Recommendation ITU-T H.870, developed jointly by the World Health Organization and the International Telecommunication Union, defines a weekly "sound allowance" for personal audio devices: 80 dB for 40 hours a week for adults, and 75 dB for 40 hours a week for children and other sensitive users. It also specifies a dosimetry function, so a device can track how much of that allowance you have used. The part worth internalising is the exchange rate. Permissible time halves for every 3 dB of extra level. At 83 dB the allowance is 20 hours a week; at 86 dB, 10 hours; at 92 dB, two and a half hours; at 98 dB, 38 minutes. That arithmetic is why "turn it down" and "listen for less time" are not interchangeable instructions. Dropping the level by a few decibels buys back hours; cutting your listening in half buys back 3 dB. If your phone shows headphone sound levels or warns you when you cross a threshold, that readout is this standard doing its job, and it is the only way to know your real exposure rather than a slider percentage. Two practical notes on that self-check, because it is a rule of thumb and this page is not going to dress it up as a measured threshold. It can fail in both directions: with in-ear or noise-isolating earphones you can pass it comfortably at a genuinely unsafe level, so passing is not evidence that you are safe. If your phone reports headphone sound levels in decibels, or warns you when a weekly allowance is used up, that feature is this standard implemented on hardware you already own, and it is the only way to see your real exposure. At home, or anywhere quiet, the volume that already works is fine - and none of this is a reason to buy anything.

Dillard LK et al. 2022

A third systematic review, in the Journal of Speech, Language, and Hearing Research, screened 460 records and included 20 studies of recreational noise exposure and hearing function in young people. Its conclusion is a useful brake on everything above: the relationship between exposure and outcome is unclear, and although some studies found threshold shifts or reduced otoacoustic emissions, most of those changes were short term and confined to the extended high-frequency range. Its authors attribute much of the disagreement to inconsistent methods - researchers measure both the exposure and the effect in different ways - and call for long-term studies using real sound-level measurement. Take that seriously in both directions. It does not rescue loud listening; the two pooled analyses on this page point where they point, the physical mechanism is not in dispute, and an exposure standard exists precisely because the dose-response for noise is well established in occupational settings. What it does mean is that nobody can tell you, from this evidence, how much hearing a given number of hours of headphone use will cost you. Everything here is observational, none of it randomised anybody to a volume, and the people who listen loudly differ from those who do not in ways no questionnaire fully captures.

Elmazoska I et al. 2024 · You S et al. 2020

Nothing on this page is a way to work out whether your hearing is already damaged, and it is not written for anyone who suspects it is. If you have ringing, buzzing or hissing in your ears that does not settle, if sounds seem muffled after a loud night, if you find yourself asking people to repeat themselves or turning subtitles on, if you have ear pain, discharge, dizziness or sudden hearing change in one ear - that is a matter for a doctor or an audiologist, promptly, and sudden hearing loss in particular is treated as urgent. A hearing test is the only thing that answers the question, and no article, app or volume-slider habit substitutes for one. The differences described above sit at frequencies above the range that carries speech, which is exactly why they are found by testing rather than noticed by the person. This module is about a common exposure in people whose hearing is currently fine.

You S et al. 2020

Nothing on this page can answer that for you, and it is worth being blunt about why. The evidence here is observational: a meta-analysis whose 17 personal-device records covered 8,987 people found 23.81% of 12- to 34-year-olds exposed to unsafe listening from those devices (95% CI 18.99% to 29.42%), and a separate meta-analysis of 9 studies found long-term users had worse pure-tone thresholds than non-users at 6 kHz and 8 kHz. Neither randomised anyone to a volume, both sorted people by questionnaire, and people who listen loudly differ from those who do not in many other ways. A third systematic review of 20 studies concluded the relationship between recreational noise and hearing function is still unclear. So this describes a group-level association with a common exposure, not a verdict on your ears. A hearing test is what answers that.

We could not trace it to a measured source. Searching the peer-reviewed literature for a 60/60 rule returns work on cochlear implant referral criteria, which shares the name and nothing else; the headphone version lives in clinic blogs and awareness material. It is also unfalsifiable as written, because 60% of maximum volume is a fraction of a maximum that differs between phones and earphones, so it is not a sound level at all. The standard that does exist is WHO-ITU Recommendation H.870, which defines a weekly sound allowance of 80 dB across 40 hours for adults and 75 dB across 40 hours for children and sensitive users, and halves the permitted time for every 3 dB of extra level. Turning it down and taking breaks are still sensible; the two sixties are just not evidence.

See a doctor or an audiologist rather than changing a setting. Ringing that does not settle, muffled hearing after a loud night, needing people to repeat themselves, ear pain, discharge or dizziness are all reasons to get your hearing tested, and a sudden loss of hearing in one ear is treated as urgent - do not wait it out. Nothing on this page is a way to diagnose yourself, and none of the evidence here describes people who already have hearing difficulty. The differences the studies found sit at 6 and 8 kilohertz, above the frequencies that carry speech, which is precisely why they are picked up by a test rather than noticed by the person. If you are worried, the test is the answer, not the volume dial.