Some older people say, "My knees hurt, so it will rain tomorrow." Is this true? A big study in the United States looked at about 11 million visits to doctors. It found no link between rainy days and joint pain.

But in another study, 2,658 people wrote down their pain every day on their smartphones. The researchers found a small link between pain and the weather, such as humidity and wind.

"My knees always hurt before it rains." You may have heard someone say this. In 2017, Harvard researchers compared rainfall data with about 11.67 million doctor visits by older Americans. They found no meaningful difference in visits for joint or back pain between rainy days and dry days.

But the study only counted visits. It did not measure how much pain people felt. In 2019, a British study asked 2,658 people with long-term pain to record their pain every day on a smartphone app.

This time, the researchers found links between pain and humidity, air pressure and wind. The links were small, but they were real. So the two studies gave different answers, partly because they measured pain in different ways.

The folk belief that aching joints can forecast rain has proved surprisingly difficult to test. A 2017 Harvard analysis matched roughly 11.67 million outpatient visits by about 1.55 million Americans aged 65 or older with local rainfall records and found no meaningful link between rainy days and visits for joint or back pain, even among people with rheumatoid arthritis. The authors acknowledged, however, that visits are an imperfect proxy for pain, since the claims data had no detail on disease severity.

A 2019 British citizen-science project addressed this by having 2,658 people with chronic pain log their symptoms daily on a smartphone, which also recorded their location. It detected statistically significant but modest associations between pain and humidity, air pressure, and wind strength, even after accounting for mood and physical activity, although participants who already believed in the link may have been more likely to join. Psychology offers another explanation. In a 1996 experiment, many college students perceived relationships in completely random sequences of numbers. Finally, a 2011 review of rheumatoid arthritis studies concluded that weather has almost no effect at the population level but may affect a minority of individuals.