Why was the definition of chronic migraine changed?

This post is perhaps provocative for those migraineurs who experience a good effect from taking Aimovig and/or Ajovy. Still, I dare to stick my neck out, because there is – I think – pretty good evidence that a large part of the effect is placebo.

What I write here is my personal thoughts. So it’s me alone who should be blamed if you think I’m wrong or unreasonable.

Placebo is excellent medicine with very few side effects. However, the active CGRP medication has some side effects, whether it really reduces migraine or just gives a placebo effect.

I would like to draw attention to this – and at the same time maybe put some pressure on the manufacturers of Aimovig and the next CGRP medicines, to document what, if anything, characterizes migraineurs who get the real effect from the medicine. If, in the future, we know which special groups get a real (not placebo) effect from the drug then the drug can be targeted at this group.

I hope this post will help kick-start a debate about what we (the migraineurs), the doctors, the media and the pharmaceutical industry can do to target Aimovig and the upcoming new migraine medicines to those migrainers, that actually get an effect.

The number of chronic migraineurs more than doubled in 2018

The definition of chronic migraine was changed drastically on October 1, 2018. In fact, the number of chronic migraineurs more than doubled on that date.

Prior to this date, chronic migraines were defined as at least 15 MIGRAINE DAYS per month over a 3-month period. After October 1, 2018, the definition became that you must have at least 15 HEADACHE days a month, including at least 8 migraine days. Headache in this context is the common term for all kinds of headaches – ie. most often tension headaches and migraines.

On 1. October 2018 the changed definition of chronic migraine almost trebled the number of chronic migraines.


The change was a consensus decision in a working group of the International Headache Society. But there was no real explanation as to why the new definition is better than the old one, although there must, of course, have been some thoughts behind it.

It does not, however, take much thought to imagine, that the coincidence that the definition of chronic migraine was changed immediately before the publication of the results of a series of clinical trials of the new CGRP medicines is somewhat improbable.

Clinical trials are expensive

Doing a clinical trial is expensive and ressource demanding. Excellent researchers from several countries must be enrolled, and they must find test persons who have just the right characteristics. And the trials must be just big enough to ensure that the results are significant – ie. that we can trust that the effect is better than placebo.

The majority of the results of the first tests of the new CGRP agents were published with a graph that looked somewhat like the one in the figure below.

The results for the first 6 months showed a placebo effect of approx. 1.5 days a month. That means that the subjects who received the cheating medication received a reduction in migraine days of approx. 1.5 days a month.

Those who were given the true medication received a reduction of 3 – 4 days a month. This was a big news in the media – now we have new medicine that actually reduces the number of migraine days, said the press releases.

The graph is from Aimovig’s US leaflet https://www.rxlist.com/aimovig-drug.htm#description. So we can safely assume this is what the company wants us to see.

That was perfectly correct – the average number of migraine days was reduced and the reduction was significantly (that means we could trust it was true) greater than the placebo effect. However, it just wasn’t quite enough to convince doctors and patients with a critical sense to be enthusiastic.

A review of the published results from many clinical trials showed that the migraineurs ‘used’ in the trials were selected to have 8 to 11 migraine days per month. That is, according to the new definition, they could be used both as episodic migraineurs (if you just closed your eyes to the fact that they also had some days with other types of headaches) and as chronic migraineurs (with 8 or more migraine days per month).

There has been general skepticism once curves similar to the one above appeared in a number of trials. A reduction of approx. 2 days a month (when the 1.5-day placebo effect was deducted) does not make a big difference if you have 15 or 20 migraine days a month. Patients were unsure if the new medicine was worth the money.

And more thoughtful patients and doctors began to think that probably only a small group of test persons had a convincing real effect from the medicine. We just don’t know who the lucky ones are.

But the big pharmaceutical companies that had invested fortunes in the new medicines had saved a little on the clinical trials. The reaction meant, however, that they felt the need to find other good slogans for the sales material. A small average reduction in a large number of migraine days was not convincing.

How to save the situation for a new drug with poor results?

During 2019, a new slogan was coined to describe the excellence of the CGRP medicines. Now the slogan was: The medicine halves the number of migraine days for half of the subjects.

That sounded much better than telling that the medicine removed only 2 migraine days.

The graph is from D. W. Dodick, S. D. Silberstein et al. 2018. Effect of Fremanezumab Compared With Placebo for Prevention of Episodic Migraine A Randomized Clinical Trial. JAMA 319, 1999-2008. doi:10.1001/jama.2018.4853

The pharmaceutical companies had probably been thinking about the possibility of using this slogan long before the change of the definition of chronic migraine. It is hardly a coincidence that they chose the number 8 as the minimum limit for chronic migraine days per month. With 8 migraine days as the base limit for the participants in the clinical trials, the placebo effect (as shown above) was approx. 1.5 days a month. A little mathematics shows that halving the migraine days of half of the participants may, with a little bit of luck, reach this goal, if half of the participants ‘lose’ 5 migraine days – or put another way, where the average loss of migraine days in the group is 2.5 days per person. Now it is merely acknowledged that the ‘lost’ migraine days are not evenly dispersed within the test group, but is found mainly in half of the test persons.

But here too, it was still difficult to get the “right” documentation in place. The placebo treatment halved the number of migraine days for quite a large group of migraineurs. In the figure above (the one with grey and yellow columns) 25% of the placebo group halved their migraine days after a placebo injection in the first month (the light gray bar) and 37% had this effect in the third month. These are hard odds for the active medicine. After all, the medicine helped only about 15 – 20% of the participants, while the placebo helped about every 3rd participant.

This is an unexpectedly large placebo effect, compared to what we know from clinical trials of other medicines.

One might rightly say that the usual calculation rules for placebo effect cannot be used when counting people with the desired effect instead of migraine days per person. But the graph with the bars has been produced by Novartis to show us that yes, it is possible that half of the users may experience a halving of their migraine days. But then we have to ‘forget’ to correct that every third in the placebo group had the same effect. If we subtract a little over 30% from the good results for Aimovig, the active medicine managed to halve the migraine days for only about 20% of the migraineurs.

Migraineurs have previously been involved in trials that showed significantly greater placebo effect when the placebo tablets were labelled Rizatriptan with an exorbitant price tag. The new medicines will have a high price, in the region of 500 EURO per injection, so we must expect a large placebo effect solely because of the price. The placebo effect of the high price has been well supported by media hype, doctors who praise the medicine, and a recurring (and certainly necessary) conversation with a doctor or nurse every three months.

We also know that the effect of Botox injections is only marginally better than the placebo effect, so there also generally seems to be a large placebo effect when medicine is injected.

Is it all just a placebo?

Placebo effects tend to decrease over time. We’ll see if the CGRP antibodies retain the effect. Patients from the United States report that some users of Aimovig (which was the first CGRP agent on the market) seem to experience a diminishing effect and want to switch to Ajovy after a year. Then they get a renewed good effect.

So we cannot know whether a large percentage of Aimovig users get ‘only’ the placebo effect, or whether the new medicines are actually fantastic for some migraineurs. However, they are hyped, the high price increases the placebo effect and injections also have a fine placebo effect.

From a migraineur’s point of view, the best thing would be if ‘someone’ could find out why some, but not all migraineurs get fewer migraine days when treated with the CGRP medicines. Is it genetics, placebo or something completely different?

What do you think?

Back to menu (Blog posts in English)

Skriv et svar

Din e-mailadresse vil ikke blive publiceret. Krævede felter er markeret med *