40 Hz Gamma Waves Side Effects: Headache, Fatigue, and Who Should Be Careful

Type "40 Hz side effects" into Google and every result says the same thing: "generally safe," "mild headaches may occur," "consult your doctor if you have epilepsy." Not one of them gives you a number. That's strange, because the numbers exist — the biggest controlled trial of 40 Hz stimulation published its full side-effect table, with a placebo column, and it's public. This article puts that table in front of you, explains which effects came from light versus sound, and tells you what to actually do if you feel something. If you're new to the topic, our main guide to what 40 Hz gamma waves are and what the science supports is the place to start; this one picks up on the safety question it left open.

One thing to be clear about up front: everything below is about the frequency — 40 Hz stimulation as studied in research. No consumer audio product, including the one reviewed on this site, has its own safety study. The data comes from clinical trials, in patient populations, using medical devices. That's the honest baseline, and it cuts both ways: it means the numbers are real, and it means they don't describe you exactly.

What the trials actually reported (with the placebo column)

The most useful safety data on 40 Hz comes from the OVERTURE trial (Cognito Therapeutics, published in Frontiers in Neurology, 2024). Seventy-four adults with mild-to-moderate Alzheimer's used a device delivering 40 Hz light and sound for one hour a day, every day, for six months. Forty-six got the real stimulation; twenty-eight got a sham device that looked and felt the same but delivered nothing. Here is what people reported:

Source: OVERTURE trial, Frontiers in Neurology, 2024 (NCT03556280).

Three more facts from the same paper matter as much as the table. Seven people left the study because of a side effect — four in the active group, three in the sham group. The early-termination rate was 28% active versus 29% sham — essentially the same. And the authors reported no unexpected serious treatment-related events, and nothing serious enough to stop the study.

Read that again slowly: after six months of daily hour-long sessions, people receiving real 40 Hz stimulation quit at the same rate as people receiving nothing.

Why the placebo column changes everything

Look at the dizziness row. More people got dizzy with the device switched off (7.1%) than with it on (4.3%). And a full quarter of the sham group — people receiving no stimulation whatsoever — reported a treatment-related side effect.

That's not a flaw in the study. It's the most useful thing in it. It tells you that a large share of what people attribute to "the sound" is expectation: you sit down braced to feel something, and you do. Researchers call it the nocebo effect, and it shows up in nearly every trial of anything.

This doesn't erase the real signal — headache was genuinely more common in the active group, and tinnitus was the one effect with a clean separation (15.2% vs 0%). We'll come back to both. But it should reset your baseline. When a website tells you "binaural beats can cause dizziness, nausea and anxiety," ask the question no one on that page asked: compared to what?

On anxiety specifically, there is a direct test. Leistiko, Madanat, Yeung and Stone (2023) played 40 Hz binaural beats to 58 healthy adults, against a control tone, and measured attention and self-reported anxiety. Result: no difference on either — 40 Hz didn't improve attention, and it didn't raise anxiety. The "it causes anxiety" claim you see repeated online has been tested at this frequency and came up empty. We look at whether binaural beats do anything for focus in our separate guide.

Light or sound? The distinction almost every article skips

Here's something that gets lost in every "is 40 Hz dangerous" post: nearly all the human trials — OVERTURE, the MIT GENUS studies — used flickering light and sound together. The side-effect tables lump both into one column.

That matters, because some effects can only come from the light. Eye discomfort, dry eyes, light sensitivity, and the visual disorientation in the table above are visual-pathway effects. An audio file has no visual component. Those rows don't apply to it.

It matters even more for the seizure question. The warning you see everywhere — "if you have epilepsy, avoid 40 Hz" — is borrowed from photosensitive epilepsy, which is, by definition, triggered by flashing light. Sound doesn't flash. The sound-based condition, musicogenic epilepsy, is a different and extremely rare thing, and the literature describes its triggers as specific, emotionally loaded pieces of music rather than a steady rhythmic tone.

There is even a direct data point. In work connected to MIT's Picower Institute, two epilepsy patients at the University of Iowa were monitored with implanted electrodes during 40 Hz stimulation. The recordings showed the expected rise in 40 Hz activity in deep brain regions — "with no adverse events including seizures," in the researchers' words.

So is 40 Hz audio safe for someone with epilepsy? The honest answer has two halves. The seizure concern in the literature is about the light, and audio has no light. And yet: the trials deliberately excluded people with a seizure history as a precaution, which means there is no controlled data on them. That's why the advice to talk to a professional first isn't boilerplate — it's the accurate description of a gap.

The most common side effect isn't headache. It's feeling sleepy.

Most people expect a "focus frequency" to make them wired. The trial data says the opposite is the more common complaint.

In MIT's first human safety study — 43 volunteers across ages, including 16 with early Alzheimer's — the researchers reported "a few minor but no major adverse effects," and the most common one was feeling "sleepy or drowsy." In the follow-up phase 2A pilot (Chan et al., 15 participants, one hour daily for three to four months), drowsiness was again the top item in the active group at 37.5%, ahead of dry eyes and light sensitivity. There were zero seizures in any group.

Why sleepy? Think about the setup: an hour of sitting still, eyes closed or fixed on a light, a monotonous stimulus, older participants. Under those conditions most people would drift off with or without 40 Hz. It's worth knowing so you don't misread it. If a short session leaves you a little flat instead of sharp, that's a known, mild, passing effect — and it usually points to the session being too long or badly timed, not to anything wrong with you.

Headache: it's almost always the how, not the what

Headache was the one common complaint that was genuinely more frequent with real stimulation (21.7% vs 10.7%). But notice the conditions: sixty minutes, daily, through headphones, for six months. The plausible drivers are all things you control.

Volume. This is the big one, and most sites quote the wrong number. The World Health Organization's safe-listening guidance for headphones is 80 dB for up to 40 hours a week for adults (75 dB for children) — not the 85 dB figure copied everywhere, which is an occupational workplace limit from NIOSH. The WHO number is a budget, not a ceiling: at 90 dB you burn through it in about four hours a week. If a session gives you a headache, the volume is the first suspect, and a loud sustained tone doesn't need to sound "loud" to be over budget. Whether you need headphones at all is its own question, which we cover separately.

Session length. One hour a day was the clinical dose for a clinical population. Focus research on healthy adults used 5 to 20 minutes. How long you should actually listen is a question we've answered in detail — the short version is that shorter and earlier beats longer and constant.

The headphones themselves. Pressure from a tight headband or deep-seated earbuds over a long session produces a headache that has nothing to do with frequency.

The tone. A raw, sustained buzzing tone is unpleasant for a lot of people, and irritation with the sound itself is an underrated side effect. Tracks that carry the 40 Hz modulation on top of music tend to be far easier to tolerate than a bare tone — which is exactly how the one sound-only safety study in healthy people was designed (more on that below). Most commercial tracks take the music-based route rather than a bare tone; The Genius Song, which we've covered in our review, is one of the most searched of them.

Who should be careful (a short list, on purpose)

You'll find lists online with fifteen groups on them. Most of the entries are copied from one blog to the next. Here are the four that actually have a reason behind them.

1. Anyone with epilepsy or a seizure history. Not because audio has been shown to trigger seizures — it hasn't — but because the trials excluded you, so nobody can say it's been tested on you. Talk to your neurologist first.

2. Anyone with tinnitus or hyperacusis. Tinnitus is the one effect that separated cleanly from placebo (15.2% vs 0%), under an hour a day for six months. If you already have ringing or sound sensitivity, a constant modulated tone is the kind of stimulus most likely to bother you. Start very short and very quiet, and stop if it worsens.

3. Pregnancy. Here's a distinction no one makes: there is no evidence of harm and no safety study. Nobody is going to randomize pregnant women to listen to modulated tones, so the caution comes from absence of data, not from any signal of risk. That's a very different statement from "it's dangerous."

4. Children. Same logic — the trials were in adults, mostly older ones — plus a lower WHO volume budget (75 dB). Also, a child can't tell you their ears hurt as clearly as you'd like.

And two you can cross off. Pacemakers appear on nearly every warning list, and not one page explains why — because there isn't a why. That warning was inherited from electrical stimulation devices and electromagnetic fields. An audio file playing through headphones doesn't interact with a pacemaker. "Interacts with medication" is the same story: no mechanism, no source. What is sensible is that if you take medication for a neurological or psychiatric condition, mentioning it to whoever prescribes is prudent. That's ordinary care, not a drug interaction.

What to do if you feel something

Every site says "stop if you feel discomfort." Nobody says which discomfort calls for what. This is the table we'd want if we were the one pressing play.

Notice that five of the six rows end in an adjustment, not a goodbye. The one exception — tinnitus — is the one the data justifies. As for how long an effect lingers: the trials didn't measure this, but the effects reported were the kind that resolve when the stimulus stops. If something persists for hours after you take the headphones off, treat it as a signal, not a phase.

One more practical thing that only one page on the internet mentions: nobody has studied what 40 Hz entrainment does to reaction time or situational awareness. Don't listen while driving or operating anything that can hurt you. Not because it's been shown to be dangerous — because it's been shown to be unstudied.

What the science still doesn't know

An honest side-effects article owes you the gaps.

Sound-only data in healthy people is thin — but it exists. Sato, Maeda and colleagues (Healthcare, 2025) ran the one study that matches a real listener's situation: healthy adults 65 and older, sound only, through a phone and earphones, daily for 28 days, with the 40 Hz modulation built onto popular music. It was well tolerated, with no meaningful safety concerns. That's reassuring, and it's also the longest sound-only safety data on healthy adults there is: four weeks. The 30-month reports of daily use being "safe and well-tolerated" (Alzheimer's & Dementia, 2025) come from patients using light and sound through a clinical device. Between four weeks in healthy people and thirty months in patients, nobody has filled in the middle.

The negative evidence is real, and it's about effectiveness, not safety. Klichowski and colleagues (Scientific Reports, 2023) had 1,000 people take a fluid-intelligence test at home; those who listened to binaural beats during the second half performed measurably worse, while silence and other sounds made no difference. The authors' own line: self-applied binaural beats "may not only be unhelpful, but also harmful" — harmful to performance, in that context, not to health. That study covered binaural beats broadly, not 40 Hz specifically, so we won't stretch it further than that. It sits alongside the null result we discuss in our theta vs gamma comparison as a reminder that the honest position on 40 Hz is "promising and unproven," in both directions.

No one has studied the casual listener. Short track, a few times a week, before work — the most common real-world pattern — has never been the subject of a safety trial. What we can say is that it's a small fraction of the exposure that produced the numbers above.

That last point is the one to carry away. The side effects you read about came from older patients using a medical device for an hour a day, every day, for six months to two and a half years, with flashing light included — and even then they were mild and the dropout rate matched placebo. A short audio session is a different world of exposure, minus the component behind a good share of the complaints. For the full picture of what 40 Hz research does and doesn't show, our main guide is the place to go next.

This article is educational and is not medical advice. Safety figures cited come from clinical trials in patient populations using medical devices, not from testing of any consumer audio product. If you have epilepsy, a seizure history, tinnitus, or any neurological condition, talk to a qualified professional before using rhythmic sound stimulation.

Frequently asked questions