Noise-Induced Hearing Loss: The Silent Injury

Noise-induced hearing loss is the most common permanent occupational injury in the world, and almost nobody notices it happening. No bleeding, no pain, no bad day at the end of which you know something went wrong. There is a shift, then another shift, then twenty years, and then a conversation in a restaurant you cannot follow.

It is also entirely preventable. Here are the numbers, an honest comparison of protection, and what a hearing conservation program needs.

What is actually being damaged#

Sound travels down the ear canal, vibrates the eardrum, moves three small bones and arrives in the cochlea. A fluid-filled spiral. Along it sit rows of sensory cells topped with bundles called stereocilia; when the fluid moves, the bundles bend and that bending becomes a nerve signal. High frequencies are detected at the base of the spiral, low frequencies at the tip.

Loud noise damages these cells two ways. Intense impulses, a blast, a gunshot, a press strike, shear the structures mechanically. Sustained noise is slower: it drives the cells so hard their metabolism cannot keep up, and they die from oxidative stress over the hours and days after exposure ends.

In humans, these cells do not regenerate. Every one lost is lost permanently, and no drug or surgery changes that.

There is a further layer, appreciated only in the last decade. Before hair cells die, the synapses connecting them to the auditory nerve can be lost: cochlear synaptopathy, or "hidden hearing loss". The audiogram can be normal, so the workplace test passes, while the person still cannot follow speech in a noisy room. If you have been told your hearing is fine but you cannot hear in a crowded canteen, this is probably part of it.

Why you do not notice#

Three reasons.

There are no pain fibers in the cochlea. Nothing hurts because nothing there can.

The loss starts where you are not listening. Noise damage characteristically begins around 3,000 to 6,000 Hz, producing the classic "4 kHz notch" on an audiogram. Above the range carrying most of the energy in speech, so conversation sounds normal for years.

The consonants go first, and the brain covers for it. As the notch widens, high-frequency consonants, s, f, th, sh, t, k, become indistinct while low-frequency vowels remain. Speech does not go quiet; it goes mushy. The brain fills the gaps from context well enough that people blame the other person for mumbling, the restaurant for being loud, the television for bad sound mixing. The first person to notice noise-induced hearing loss is almost always someone else in the household.

The numbers: how loud, for how long#

Damage depends on intensity and duration. Regulators express this as a time-weighted average over eight hours, written dBA. The two main systems differ, and the difference matters:

• NIOSH (recommended, US): 85 dBA over 8 hours, with a 3 dB exchange rate. Every 3 dB increase halves the permitted time. This matches the physics of sound energy.
• OSHA (legally enforceable, US): permissible exposure limit 90 dBA over 8 hours, 5 dB exchange rate, with an 85 dBA action level triggering a hearing conservation program.
• UK and EU (Control of Noise at Work Regulations): lower action value 80 dB(A) or 135 dB(C) peak; upper action value 85 dB(A) or 137 dB(C) peak; absolute exposure limit 87 dB(A) or 140 dB(C) peak, measured allowing for hearing protection.

The 3 dB rule is the one to internalise, because it is unforgiving:

Note how fast the budget disappears: fifteen minutes of unprotected grinding uses a whole day's safe exposure. Dose accumulates too. Two hours at 91 dBA plus fifteen minutes at 100 dBA is a full day, and anything after is an overdose.

The one-meter voice test is the practical version. At arm's length from a colleague: if you must raise your voice to be understood, the background is probably 85 dBA or higher; if you must shout, likely 95 dBA. Crude, and right often enough to be useful.

Controlling it: the hierarchy, in the right order#

Hearing protection is the last line, not the first. It is where most workplaces start, which is why noise-induced hearing loss has barely improved in decades.

1. Elimination. Does the noisy process need to exist? Can the part be cast rather than ground, bolted rather than hammered?
2. Substitution and buy-quiet. Purchase specifications requiring a maximum sound power level are the highest-leverage move an EHS lead has: noise designed out for the life of the asset. A quieter compressor bought once protects everyone; earplugs must be reissued every week.
3. Engineering controls. Enclosures, vibration isolation mounts, damping on panels, silencers on exhausts and air lines, rubber in place of metal-on-metal impact. And, routinely missed: maintenance. Worn bearings, unbalanced fans and loose guards get louder as they degrade, so a rising noise reading is a maintenance signal before it is a hearing signal.
4. Administrative controls. Distance is free: in open space sound pressure falls about 6 dB per doubling of distance. Job rotation, scheduling loud work when fewest people are present, signed noise zones, control rooms off the plant floor.
5. Hearing protection. Necessary, but the weakest control, because it depends on human behavior every hour.

Earplugs versus earmuffs: the honest comparison#

Every protector carries a laboratory rating. Noise Reduction Rating (NRR) in the US, Single Number Rating (SNR) in Europe. Those numbers come from ideal conditions with expert fitting and substantially overstate real plant floor performance.

Regulators are open about this. OSHA's convention is to halve the NRR after a correction. NIOSH derates by type: about 25% for earmuffs, 50% for formable foam plugs, 70% for other plugs. Field studies find many workers achieve under 10 dB of real attenuation from a plug rated in the low thirties, almost entirely because of insertion technique.

How to insert a foam plug properly, because most people are never taught: roll it into a thin, crease-free cylinder, reach over your head with the opposite hand and pull the top of the ear up and back to straighten the canal, insert deeply, hold 20 to 30 seconds while it expands. When it is right your own voice sounds hollow and in a mirror you barely see the end of the plug. A plug sitting visibly in the ear opening does perhaps a third of its job.

Dual protection does not add up. Plugs under muffs typically gain around 5 dB over the better single protector. Worth doing above roughly 100 dBA. It is not 30 plus 25.

Over-protection is a genuine problem. Too much attenuation makes speech and reversing alarms inaudible, so people remove protectors to communicate, and a protector worn for 90% of a shift loses much of its protection, because the unprotected minutes dominate the dose. In moderate noise the answer is often a lower-rated protector, or a level-dependent electronic muff that passes speech while clipping impulses.

Fit testing changes everything. Field attenuation estimation systems measure what a worker actually achieves with a given protector, in minutes. It turns an argument into a number, finds people unprotected despite being compliant, and matches protector to person. If a program can afford one upgrade, this is it.

Audiometry: what the program should do#

Hearing tests do not prevent hearing loss. They detect it early enough to change something, and they are the only feedback loop a noise program has.

A credible program includes:

• A baseline audiogram early in employment, OSHA requires it within six months of first exposure, after at least 14 hours away from workplace noise, so a temporary shift is not baked in.
• Annual audiometry for everyone at or above the action level, in a calibrated booth.
• Comparison against baseline for a standard threshold shift: an average worsening of 10 dB or more at 2,000, 3,000 and 4,000 Hz in either ear. Age correction is permitted under some rules, but a shift is worth acting on regardless.
• A defined response, not just a letter: retest within 30 days, refit or change protection, review the exposure and task, re-train, refer where indicated.
• Results explained to the worker in plain language, with the audiogram shown. People protect their hearing far better once they have seen their own notch on a chart.
• Program-level analysis. A rising rate of shifts in one department is a failing control, not a group of careless people.

Tinnitus#

Tinnitus, ringing, hissing, buzzing or whistling with no external source, very often accompanies noise damage. When the cochlea stops sending signals at certain frequencies, the central auditory system turns up its own gain, and that amplified internal activity is perceived as sound. It is generated largely in the brain, not the ear, which is why treatments aimed only at the ear disappoint.

What is true:

• There is no cure, and no supplement has convincing evidence. Ginkgo biloba has been studied repeatedly without reliable benefit.
• Most people habituate. Over months the brain gives it less attention, provided noise exposure is controlled.
• Cognitive behavioral therapy has the best evidence for reducing tinnitus distress. It does not make the sound quieter; it changes how much of your life it occupies.
• Sound enrichment helps. Silence is the worst environment for tinnitus; a fan or low background music at night genuinely helps.
• Hearing aids often reduce it when hearing loss is present, by restoring input and lowering central gain.
• Sleep and stress feed it. Poor sleep worsens tinnitus and tinnitus worsens sleep; breaking that loop is often the highest-value intervention.

Tinnitus in one ear only, pulsing with your heartbeat, or with vertigo or facial weakness needs medical assessment rather than reassurance.

At work#

A few things frequently missed.

Noise surveys go stale. Re-survey after any change of process, equipment or layout, and periodically regardless. Use dosimetry worn by real workers on real shifts, not a spot reading in a quiet moment. Peaks need C-weighting; impulse noise from presses and pneumatic tools is invisible on an A-weighted average.

Chemicals make noise worse. Solvents such as toluene, styrene and xylene, plus carbon monoxide and lead, are ototoxic, and exposure alongside noise causes more hearing loss than either alone. Where workers face both, the effective noise threshold should be lower and audiometry more frequent. Some medicines are ototoxic too, certain aminoglycoside antibiotics, some chemotherapy agents, high-dose loop diuretics, a matter for the prescribing doctor, never a reason to stop a prescribed drug.

Count the whole dose. Occupational exposure sits on top of shooting, motorcycles, home power tools, live music and personal audio. Someone at 84 dBA at work who rides a motorcycle an hour each way is not under threshold.

Keep protection where the noise is, not in a store room three departments away. Dispensers at every noise zone entrance remove the main excuse.

Watch visitors and contractors. Permanent staff are enrolled in the program. The contractor doing two weeks of grinding often is not, nor is the manager who walks the floor daily without plugs.

What I actually see#

What I find hardest is that this injury is discovered late, and by then there is nothing I can offer beyond referral and preventing further loss. There is no version of the conversation where I get to fix it.

What I have changed is showing people their own audiogram rather than telling them the result. The notch at 4 kHz is visible, and persuasive in a way no toolbox talk has ever been. Workers who have seen their own curve wear their protection differently the next week. I have also stopped assuming a compliant worker is a protected worker: the men most likely to have a genuine threshold shift are not those who refuse plugs but those who wear them badly, all day, every day, believing they are covered. Watching someone insert a foam plug takes ten seconds and tells you more than a compliance audit.

The question I am asked most is whether it gets worse if the person stays in the job. Noise-induced loss does not progress once exposure stops, but it never recovers, and continued unprotected exposure adds to it, which is why the remaining hearing is worth defending seriously even after a diagnosis.

The bottom line#

Noise damages the ear silently and permanently, starting at frequencies you do not consciously use, which is why it is caught years too late. The dose rule is simple: 85 dBA for eight hours, halving the safe time for every 3 dB louder, and if you must raise your voice at arm's length you are probably over it. Quieter equipment beats protectors, and protectors deliver far less in the field than on the packet unless fitted properly and worn every minute. Get a baseline audiogram, look at your own results, and treat sudden hearing loss in one ear as a same-day emergency.