Barometric pressure and migraine: what the studies show

What people report

• Among 1,207 patients of a headache clinic in Atlanta, 53% named weather as a trigger. Stress (80%), hormones in women (65%) and not eating (57%) were named more often (Kelman 2007).
• In a Boston diary cohort, 49% believed weather triggered their attacks (reported in Denney 2024).

What diary studies found

Study	Design	People	Result
Kimoto 2011, Japan	1-year diaries, nearby weather station	28 with migraine	More migraine when pressure fell by more than 5 hPa between the headache day and the next day. Weather change was linked to migraine in 18 of 28 people.
Okuma 2015, Japan	Diaries over 19 days as a typhoon approached	34 with migraine, 28 with tension-type headache	Most migraine when pressure was 6–10 hPa below the standard 1013 hPa. Headache linked to the fall in 73–75% of people with migraine and 21% of the comparison group.
Zebenholzer 2011, Austria	90-day diaries, 11 weather variables	238	Early associations were not significant in further analysis. Subjective weather perceptions did not correlate with the occurrence or persistence of migraine.
Hoffmann 2015, Germany	12-month diaries matched to hourly weather	100 with migraine	No link across the group. At least 13% were individually weather-sensitive, in no consistent direction. Forecasting a person’s attacks from weather was not possible.
Li 2019, USA	Prospective diaries	98	Higher humidity linked to attack onset in warm months. No link with pressure or temperature.
Katsuki 2023, Japan	Smartphone headache app	App users, 37% with diagnosed migraine	Lower pressure linked to more headaches.
Portt 2026, Canada	Smartphone diary, case-time-series	7,418 users, 69,808 attacks	Both drops and rises in pressure linked to higher odds of an attack, strongest within the first day. Cold and some air pollutants were also linked.

A 2024 review concluded that recent results are inconsistent (Denney 2024).

Why the results disagree

• Different definitions. One study counts a day-to-day fall of more than 5 hPa, another a level below 1013 hPa, a third the change over 24 hours. The numbers cannot be compared directly.
• Small groups. Several studies have under 100 participants and a few months of data.
• Individual sensitivity. In the Berlin study at least 13% were sensitive, and not to the same kind of change. Group averages hide them.
• Perception. In the Austrian study, people’s own view of their weather sensitivity did not correlate with when migraine occurred or how long it lasted. Expectation and memory can make a link look stronger than it is.

Units

Weather services report pressure in different units. The study thresholds above, converted:

hPa (millibar)	inches of mercury	mm of mercury
1013 (standard)	29.91	760
5	0.15	3.8
10	0.30	7.5

How to test it on yourself

1. Record attacks and weather prospectively, not from memory. NICE advises considering a headache diary for a minimum of 8 weeks, including possible precipitants. The printable diary has a column for them.
2. Count days, not impressions. Compare how often attacks start on days after a pressure fall with how often they start on other days.
3. Expect to need months. The Berlin study used 12 months per person and still could not forecast single attacks.
4. Plan around what you can change. The Migraine Trust notes that weather cannot be controlled and that focusing on triggers too much can be unhelpful. A known link helps with timing, for example keeping acute medicine at hand on a forecast pressure drop.