Does meditation make pain, tinnitus or anxiety worse?

It is ten past three in the morning. No traffic, no fan, nothing running. The ringing in your ears has nowhere left to hide, and the last thing anyone told you to do about it was to pay attention to it on purpose.

Or the body scan has arrived at the exact region you spent all day steering around. Or it is a good day, no pain anywhere, and something in you has started looking for where it went.

One question sits underneath all of those. It gets asked in a dozen different rooms by people who never find out they are asking the same thing. If I turn my attention toward this, will I make it worse?

It deserves better than the answer it usually gets.

Rule this out before you rule out meditation

A symptom that began during meditation is not the same as a symptom caused by meditation. That distinction is the part most articles skip, and skipping it costs people years.

A few years ago someone posted to a meditation forum describing constant pressure in the middle of his face, three years after a silent retreat, worse when he was tired, bad enough to keep him awake. He assumed the meditation had done it. Hundreds of people replied, and the single most upvoted reply was not about meditation at all. Talk to your doctor.

He did. Scans were clear. It took several more specialists, and eventually a sinus operation that changed nothing, before an orthoptist found long-sightedness his eyes had been straining against the whole time. Glasses helped within weeks.

That is not a reason to distrust meditation. It is a reason to distrust the sequence your mind builds when something starts hurting shortly after you began paying attention to yourself. Meditation increases how much you notice. Noticing more is not the same as there being more, and it is also not proof that there is nothing.

This matters because the research on meditation adverse effects is real and worth taking seriously, and it is also the first place people file a symptom that has nothing to do with their practice. Both mistakes cost time. Only one of them is dangerous.

New symptoms, changing symptoms, one-sided symptoms, anything with a pattern nobody can explain. Those go to a clinician first. Nothing on this page, and nothing in any meditation app, replaces that.

The real question, and why nobody answers it whole

Search for whether attention makes tinnitus worse and you will find a good answer, written for tinnitus.

Search for whether focusing on pain makes it worse and you will find a good answer, written for chronic pain.

Search for whether body scanning is safe with health anxiety and you will find a good answer, written by anxiety clinicians, and it points the opposite way to the first two.

All three are correct. That is the strange part. The pain literature and the health-anxiety literature are not disagreeing by accident, and neither one is wrong. They are describing two genuinely different activities that look identical from outside, because in both cases a person is sitting still with their attention on a sensation.

Nobody publishes the page that stands between them. What separates the two is not the sensation, and it is not you.

Attention with a question mark on the end

Attention is not one activity. It has punctuation.

Some attention ends in a question mark. Is it louder tonight than last night? Is that the same twinge as this morning? Is this dangerous? That attention is not looking, it is asking, and it will not settle until it gets a reply. When the reply is reassuring it buys a few minutes. When it is not, it escalates. Either way the question gets asked again shortly afterward, because the reply was never the point.

Some attention ends in a full stop. There is ringing. There is an ache in the left hip. There is the weight of the chair. Nothing is pending. Nothing is being decided.

Same person, same room, same sensation. Completely different activity.

The sensation is not what makes the difference. The punctuation on the end of your attention is what makes the difference.

The same distinction has been drawn separately in three clinical settings, each in its own vocabulary, without any of them talking to the others.

Read down the second column. Fighting, monitoring, surveillance. Every one of those is attention that is waiting for an answer. Read down the third. Allowing, registering, noticing. Every one of those is attention with the question taken off the end.

Whose words these are, since it matters. Fighting and allowing are the tinnitus researchers’ own terms, taken from participants describing the change. Surveillance and checking belong to the cognitive-behavioural model of health anxiety, where checking the body is formally classified as a safety-seeking behaviour. The middle row is ours: no pain paper we read uses this vocabulary. What the pain research does supply is a measurement of the thing the row points at, and that is the next section but one.

The tinnitus researchers put this most directly. In a study of people who had completed mindfulness-based cognitive therapy for tinnitus, the authors raise our exact objection themselves and then answer it. One might reasonably worry that paying more attention to tinnitus would make the distress worse. On their reading it is the distressing thoughts and the stress arousal that drive the distress, and those are what narrow attention down onto the sound in the first place. Mindfulness, they argue, works by widening the attentional spotlight instead of aiming it.

That paper is nine people describing their own experience six months after a course. It is not proof that the practice works, and we are not offering it as proof. It is the clearest available statement of the mechanism, from the people best placed to describe it.

A test you can run mid-session

Ask yourself what would count as an answer. If you can name one, you are checking. Better than yesterday is an answer. Still there is an answer. There is ringing is not an answer, it is a description.

That takes four seconds and you can run it without opening your eyes. Almost every article on this subject says the quality of your attention matters. Almost none say how to tell which kind yours is while you are sitting there.

Four more signs, all of them common and none of them a failure:

1. You are comparing. Any sentence with than in it. Louder than, worse than, more than this morning. Comparison needs two measurements, which means one of them was taken on purpose.
2. You feel relief when it is quieter. Relief is the tell. You only get relief from a question you were waiting on. Noticing that a sensation has faded feels like nothing much.
3. You are timing it. Counting how long the good stretch lasted. Estimating when the next one is due.
4. You would be annoyed to be interrupted. Not because the session was going well, but because you had not finished establishing the answer yet.

None of that means stop. It means you have found the thing the practice is actually about, several layers earlier than most people find it. The instruction is not notice harder. It is: leave the question where it is, and let the attention carry on without it.

The left-hand shape closes. That is the whole problem with it: it returns to where it started, so it can run all night. The right-hand shape does not close, which is why it does not need an answer to stop.

Tinnitus: the silence removed the masking on purpose

Everything about a quiet room is designed to take away the things that were covering the sound. That is what makes meditation hard here, and it is not a misunderstanding on your part.

People who have done long silent retreats with tinnitus describe the sound becoming the loudest it has ever been. That is a real and common experience, and the research does not pretend otherwise. In the mindfulness-based cognitive therapy for tinnitus study, the practices that involved turning toward sound were described by participants as genuinely difficult, and the authors are explicit that increased awareness of tinnitus provoked discomfort. Their argument is not that this does not happen. It is that the discomfort was part of how the relationship changed, and that what shifted was the move from fighting the sound to allowing it.

Now the part we would rather tell you than not.

The UK’s clinical guideline does not recommend doing this alone with an app. NICE guideline NG155 sets out a stepped approach for tinnitus-related distress. It applies only to people whose tinnitus is still affecting their wellbeing and daily life after they have received tinnitus support. The first step is digital tinnitus-related cognitive behavioural therapy provided by psychologists. Mindfulness-based cognitive therapy appears at the second step, group-based, and the guideline specifies it should be delivered by appropriately trained and supervised practitioners.

Read that as it is written. Mindfulness is in the pathway, by name, which is more than most practices manage. It is also second, group-based, supervised, and downstream of tinnitus support. If your tinnitus is distressing enough that you are searching this question at three in the morning, the honest recommendation is an audiology referral and a conversation about that pathway, not a meditation app.

Where a generated session can help is narrower and more ordinary. If sitting in silence is what makes the evenings unbearable, you do not have to sit in silence. You can ask for a session that never sends you to listen for sound at all. One that stays with the weight of the body, or the breath, or contact with the chair, and leaves the ears out of it.

Chronic pain: attention and monitoring are not the same move

Ask a room full of people with chronic pain what they do when it climbs, and you will get one answer with a hundred faces. Crochet. Sudoku. Washing up in water hot enough to occupy the nerves. A thriller audiobook loud enough to cover the nausea. Rearranging the freezer.

Distraction is the entire working toolkit for a lot of people, and it works, and it took years to build. Then meditation arrives and appears to propose the exact opposite. Of course the question comes up.

Underneath it sits something people describe more precisely than any article does. Days without pain can bring on panic, because the mind reads the absence as suspicious. Coming out of a flare, attention keeps going back to confirm it is really over, and the confirming brings it back. One person summarised the double bind in a sentence: being in pain makes them anxious and not being in pain makes them anxious.

That is the checking loop, and it is running whether or not anyone sits down to meditate. It runs during the crochet.

One experiment measured this directly, and what it measured matters as much as what it found.

Those two numbers are the argument. Pain intensity fell by 40%. How unpleasant the pain was fell by 57%. If attention amplified sensation and nothing else, both would have moved together, and they did not. Something separated the sensation from what was attached to it.

The chronic pain guide goes further into the practice side, and why generic pain meditation fails with fibromyalgia covers what goes wrong when a script assumes the wrong body.

None of which obliges you to point attention at the sore place. If a body scan is going to walk you straight into the region that is flaring, you can name that region and have the session route around it. You can also do the opposite on purpose, and spend the whole session there, on a day when that is what you want. Both of those are instructions rather than settings: you say them, and the session gets written around them.

Health anxiety: when the scan is the symptom

Here the answer turns around, and it turns for a reason built into how the disorder works.

In the cognitive-behavioural model of health anxiety set out by Salkovskis, Warwick and Deale, checking your body and its functions is not a neutral act of observation. It is classified as a safety-seeking behaviour, in the same family as seeking reassurance, repeat medical consultations and searching symptoms online. The model’s central claim about safety behaviours is that they reinforce the very threat appraisals they were performed to settle. The relief is real and it is short, and the belief underneath it, that danger requires checking, never gets tested.

Now put a daily formal body scan on top of that. A practice whose instruction is to move attention systematically through every region of the body and note what is there. From the outside it is meditation. From the inside it can be the same checking, with better lighting.

People with health anxiety can meditate. What the model means for this one technique is that the body scan runs closer to the mechanism of the disorder than anything else you could sit down to do, so the four-second test matters more here than anywhere else on this page.

And here we have to stop and tell you what we do not know. We looked for a guideline, trial or review that settles whether formal body-scan practice is safe or harmful for health anxiety specifically. We did not find one. The safety-behaviour model is well established, and the body-scan question sitting on top of it appears not to have been directly adjudicated. Anyone telling you confidently in either direction is going beyond what has been shown.

The anxiety guide covers the broader picture, and why silence isn’t the point is useful if the quiet itself is what tips you into scanning.

What follows from an absent answer is caution rather than prohibition. If you want to practise, a practice that does not enter the body at all sidesteps the whole question. Attention on sound in the room. Contact with the floor. The breath at the nostrils rather than a survey of the torso. Asking for a session that stays out of the body takes one sentence.

OCD: the one where we would stop and ask someone

Someone posted to an OCD forum that mindfulness had made things worse, that sitting alone with their own thoughts was triggering, and that they could not see how it was supposed to help. Roughly twenty replies arrived. Almost every one of them, kindly, explained that they had misunderstood the practice.

By the end of the thread, the person who asked had concluded that they had probably been doing it wrong. They had not been doing it wrong. Not interacting with your thoughts is precisely the capacity OCD interferes with. Being told that the solution is to not interact with them is a bit like being told the solution to a limp is walking normally.

The thread also produced two flatly contradictory instructions, both upvoted: one commenter advised saying STOP internally when the thoughts start, another pointed out that this is the opposite of the practice. The second one was right, and worth knowing: telling yourself to stop is itself a mental compulsion, which means the advice offered as a fix was an instance of the problem.

The most sensible thing anyone said in that entire thread came from the person who asked the question. After reading all of it, they said they would try again with someone who specialised in OCD, because they did not trust themselves to do it alone. That is a good instinct and we are not going to talk anyone out of it.

For OCD, the psychological treatment recommended by NICE guideline CG31 is cognitive behavioural therapy including exposure and response prevention, delivered in structured formats and stepped by how much the condition is affecting daily life. We searched the guideline’s recommendations for mindfulness and meditation and found neither term. That does not make mindfulness harmful. It does mean that the intervention with the evidence behind it is a specific, structured, clinician-led therapy, and that a meditation app is not a version of it.

So this is the section with no product suggestion in it. If the practice has become somewhere your OCD lives, the next step is a therapist who works with OCD, not a better app and not a better technique.

You cannot answer this from one session

Everything above is about a single sitting. The question people actually want answered is bigger than that. When meditation makes symptoms worse in one session, does that mean the practice is sensitising you over months, or was that just a bad Tuesday?

That is not knowable in the moment, and the moment is a bad place to ask. A session that felt awful can be followed by a quiet week. A session that felt fine can be the third in a run where things have been slowly tightening. Judging a practice by how the last twenty minutes went is reading noise.

What answers it is a record. Not a streak, and not a count of sessions. Something with how you actually felt attached to it, kept long enough that a trend can show through the variation. That is why StillMind’s journal carries mood and emotion tags and not only a tally. Eight weeks of tagged entries can tell you something eight weeks of remembering cannot, because the remembering is done by the same mind that has been doing the checking. The same limitation applies when nothing is going wrong: meditation gives you almost no signal from inside a session, so a record beats an impression whichever direction things are moving.

A record also lets you stop on evidence rather than guilt. The forum poster whose facial pressure faded is worth returning to here. What he changed, alongside everything else, was practising less. He backed off, the pressure settled, and he eventually came back to a lighter practice a few times a week.

Stopping is a legitimate result. So is a smaller version of it. How long you should meditate covers why shorter is not the compromise it sounds like.

When it is the emotional kind of harder

One thing this page is deliberately not about. Sometimes what surfaces when you sit down is not a sensation but a feeling: agitation you cannot place, unexpected crying, old material arriving without an invitation. That is a different question with a different answer, and the checking test above will not help you with it.

It is also extremely common and well described in both contemplative and clinical traditions. When meditation makes things harder covers it properly, and meditation when you feel nothing takes the opposite complaint, the difference between distance and contact.

If the practice keeps asking you to picture something you cannot picture, meditating without visualisation is the one you want.

The question mark test is built for sensations. Feelings need a different instrument.

Frequently asked questions

Going deeper

The rest of this series takes the specific cases. Why meditation failed you is the long view, and why meditation scripts fail is for when the guidance itself is the mismatch.

For the technique taught properly, the body scan guide includes its own section on when to skip it. If the practice has become dependent on the thing that taught it to you, read what to do when you cannot meditate without an app.

The question this page opened with does not have one answer, and any page that gives you one is selling something. What it has is a discriminator. Attention that is waiting for a reply behaves one way. Attention that is not behaves another. Everything else here is that distinction, applied four times.