Medication-Overuse Headache (Rebound): How to Recognise It and Come Off the Drugs

The ICHD-3 criteria are specific and differ by type of drug. The key is to count the days per month on which a drug was taken, not the number of tablets: a day with one tablet counts the same as a day with four. The diagnosis requires headache on at least 15 days a month in a person with a pre-existing headache and regular overuse for more than 3 months, and the headache must not be better explained by another diagnosis.

This table explains why distinguishing simple from combination drugs matters so much. Someone who takes ibuprofen on 12 days a month does not exceed the threshold (15), whereas a person who takes a tablet containing paracetamol, caffeine and codeine on 12 days already does (10). Triptans and opioids have a lower threshold because they raise the risk of overuse more. If the number of days approaches the threshold, that is a signal to act before the headache becomes entrenched.

The first layer is central sensitisation. Frequent use of painkillers and migraine drugs changes how the nervous system processes pain: the excitability threshold of trigeminal neurons falls, pain-inhibiting systems weaken, and the serotonergic system and excitatory receptors are remodelled. As a result pain appears more easily and returns faster after a drug has worn off. This has been described in neurophysiological and imaging studies, but the details of the mechanism remain under investigation, so it is mechanistic knowledge, not an established model.

The second layer is a behavioural pattern: in people with frequent headaches a tablet is often taken 'just in case', at the first signals, on days when difficulty is expected, which increases the number of days with a drug. Caffeine adds to this: it not only strengthens the painkilling effect (see our article on tension-type headache) but also itself causes withdrawal headache after a sudden break; we write about caffeine and sleep in separate texts on coffee. In people taking opioids, physical dependence joins the mechanism, which is why opioids are treated separately.

Different groups of drugs carry different risks. Triptans and opioids cause MOH faster, after a dozen or so days a month for a few months, while simple drugs usually need longer overuse. That is why the threshold for simple drugs is higher. We do not know why only some people who overuse drugs develop MOH; genetic factors, coexisting depression and anxiety, lack of sleep and chronic stress probably matter.

The simplest diagnostic tool is a headache diary: for 4–8 weeks write down every day with pain and every day with a drug (name, dose). Count the days with drugs by group. Such counting often comes as a surprise — many people do not realise they took a tablet on 14 days a month. If the threshold is exceeded or approached, it is worth talking to a doctor before the headache becomes entrenched.

For years the rule was 'first stop everything, then start prevention'. Newer studies show that the order matters and that withdrawal need not be the only method. Two randomised trials from the Danish Headache Center compared three strategies in adults with MOH and migraine or tension-type headache.

Important caveats: both papers are open-label studies (patients knew what treatment they were receiving), carried out at one tertiary centre in patients without severe comorbid disease and without dependence on other substances. The results do not translate to people who overuse opioids or to more severe cases. High caffeine as a 'protective factor' in the one-year follow-up is an observational association — it does not follow that one should increase coffee intake; guidelines instead generally advise limiting excess caffeine.

The practical conclusion from these sources: it is best to agree with a doctor a plan in which acute drugs are limited in parallel with starting preventive treatment of the underlying pain, rather than counting on simply 'gritting your teeth'. Withdrawal can be hard: for a few days up to two weeks the pain may worsen, and nausea, insomnia, restlessness and irritability may appear. During that time the doctor may suggest symptomatic treatment. The guidelines state clearly that opioids are withdrawn in a hospital setting.

MOH is not a disease in itself but a complication of treating another: migraine or tension-type headache. Effective treatment therefore requires prevention of that disease. In migraine the guidelines list topiramate, botulinum toxin type A and anti-CGRP antibodies as effective also in people overusing drugs, and in tension-type headache amitriptyline (we write about the evidence for it in the article on tension-type headache). Prevention also includes non-drug methods: regular sleep, physical activity, relaxation techniques and cognitive-behavioural therapy.

Caffeine is present in this story twice over. First, it is an ingredient of many painkiller tablets; combination drugs with caffeine have a lower overuse threshold (10 days a month) because they are more easily used often. Second, abrupt withdrawal of caffeine after regular drinking causes a headache of its own (described separately in the article on insomnia after caffeine withdrawal), which can overlap with the headache from drug withdrawal.

It is therefore sensible not to change caffeine abruptly while acute drugs are being withdrawn, but to settle on a stable, moderate dose and reduce it gradually once the pain is under control. The finding that high caffeine consumption was linked to a lower risk of chronic headache after a year is observational and does not mean that drinking coffee protects — it may have arisen because people drinking more coffee reached less often for other drugs or had a milder course. We write about caffeine doses in a separate article.

Medication-overuse headache is one of the rare problems in which the first and cheapest tool is counting: on how many days a month do we take a drug, and which. When the threshold is exceeded, effective treatment exists and rests on three pillars — education, prevention of the underlying pain and, if necessary, a controlled break from acute drugs. The best available trial suggests starting prevention at once, together with withdrawal, rather than waiting. Abrupt, do-it-yourself abandonment of drugs can be dangerous, especially with opioids, so it is worth drawing up the plan with a doctor.