Some people react to the opposite

The literature says falling pressure. But within the same studies sit patients who feel worse when pressure rises. They disappear into the average.

Most of what's been written about air pressure and migraine points the same way: falling pressure, more attacks. But inside those same studies sit patients for whom it's the opposite. They feel worse when pressure rises.

That's not a footnote. It's one of the most cited explanations for why research in the field looks so messy when you read it from the outside.

Two groups, one line in the conclusion

A review from 2024 in Current Pain and Headache Reports puts it directly:

"Individual variation may play into how barometric pressure affects migraine as … some patients were susceptible to low barometric pressure while others were susceptible to high barometric pressure."

In other words: some patients react to low pressure. Others to high pressure. Same diagnosis, same field of research, opposite direction.

The same review gathers examples in both directions. Low pressure was linked to more headache episodes in one app-based study, where 37.2% had a doctor-diagnosed migraine. Chinook winds in Canada produced significantly more attacks both the day before the wind and on the chinook day itself. In a study from North Carolina, by contrast, there was no statistically significant link at all between air mass changes, air pressure, and acute migraine visits. And in a Boston cohort, it was humidity, not pressure or temperature, that made a difference.

How an average can erase two groups

Picture a hundred patients in a study. Half get worse when pressure falls. The other half get worse when pressure rises. Every single one of the hundred has a real, stable link to the weather.

Put them together in one statistical model, and ask: is low pressure linked to more attacks in this group? The answer comes back no — or a number so small it drowns in the noise. The two halves pull in opposite directions and cancel each other out almost completely. The study concludes "no link", and the conclusion is true for the average and wrong for almost everyone in it.

It's one of the most likely explanations for why a systematic review across 14 studies and 2,696 participants finds mixed, mutually contradictory results on frequency — and itself points to "failure to account for individual variation in weather sensitivity" as one of its own weaknesses.

Even within one and the same study

You don't need to stack several studies on top of each other to see it. In a Japanese study of 28 migraine patients, 18 (64%) got migraine in connection with a change in the weather. 14 of them pointed to low pressure themselves as the cause. The study nonetheless measured the monthly average pressure against headache frequency over a year — and found no link. A monthly average is a coarser measure than 14 patients' own, specific sensations, and an even coarser one if some of the remaining 28 were reacting to something entirely different from low pressure.

Another review states the underlying problem itself: many studies in the field produce inconsistent results regarding the direction of the link. That sounds like a weakness of the research. It could just as easily be a description of reality: there isn't one direction to find, because there isn't one group.

That's Aneroid's whole premise

That's why the app doesn't assume a direction in advance. It logs your pressure and your state, whether pressure is rising or falling — and lets your own data show whether there's a pattern at all, and if so, which way it points. An average of 2,696 people can't answer that for you. At best, it can show that the question is worth asking.