How big is the effect, really?

Odds ratio 0.962. That doesn't sound like much — and that's also the point. Here's what the number means once you translate it into a day of your life.

0.962. That's the single most important figure on this page, and it's disappointing to look at. The number comes from the largest study of weather and pain there is: 2,658 Britons, 9,695 pain episodes weighed against 81,727 control periods, fifteen months from February 2016 to April 2017. The conclusion for air pressure is an odds ratio of 0.962 per 10 mbar the pressure rises.

In plain terms: if air pressure rises 10 mbar, the odds of a pain day fall by just under 4%. If pressure falls by the same amount, the odds rise by roughly the same. That's a link. It's also a weak one — weaker than humidity, which gave an odds ratio of 1.139 for an equivalent change, and which per standard deviation pulled around 12% against pressure's just under 4%.

The worst day imaginable

A single number from a regression model doesn't say much about how a real day feels. So the authors behind the study worked it through themselves: put low pressure, high humidity and high wind together — the worst plausible combination — and hold it up against an average day. The result is that the odds of a pain day rise by just over 20%.

20% is not nothing. It's also not much. It's the kind of number that disappears if you only look at one day, and becomes visible if you look at hundreds.

Wind counted, temperature didn't

The same model also looked at the other weather variables. Wind speed was significantly linked to pain, if weakly: an odds ratio of 1.02 per 1 m/s. Temperature wasn't linked at all — odds ratio 0.996, not statistically significant. And humidity correlated more strongly with osteoarthritis than with the other diagnoses in the study, which suggests that even a weak average effect can hide differences between conditions, not just between people. The authors themselves stress that the effect sizes are modest, and that the findings apply to British climate — they can't simply be carried over to different weather.

Another study lands on the same number

It's not just one study pointing this way. A 2024 review of the literature on weather and migraine concludes flatly: "the predictive value of these weather events is just 20%." Two independent figures, from two different methods and two different datasets, land on roughly the same ceiling. Weather alone can explain a fifth of the story — not more.

What the average hides

Here's the part that's easy to forget when you only look at one number: a small average effect is not the same as a small effect on everyone. When the authors behind the citizen-science study held temperature fixed and let the other variables vary, the odds of a pain episode swung from around 0.8 to 1.2 — a considerable spread, hidden inside a single average figure.

The 2024 review points to the same mechanism from a different angle: some patients react to low air pressure, others to high air pressure — put them together in one model, and they can partly cancel each other out. An average of "almost nothing" can easily cover a group that feels it clearly, and a group that feels nothing at all.

That's not an excuse to overinterpret 0.962. It's a reason not to let that number decide the matter for you personally.