An interview with weather-pain researcher Dr Geir Smedslund

Dr Geir Smedslund is a psychologist with a PhD in health psychology. From childhood, he has had an interest in weather and climate, and for a while wanted to become a meteorologist.

Since 2007, Dr Smedslund has held a part-time position as a senior researcher in the rheumatological department of Diakonhjemmet Hospital, Oslo, Norway. Here he has conducted research comparing weather data to the patient records of people with rheumatoid arthritis and fibromyalgia. He also produced a systematic review of nine studies that have examined the association between weather and pain.

In this week's Cloudy blog, Dr Smedslund tells us more about his research, the meteorological features that he believes are the most likely to have an impact on pain, and explains why studies up to now have struggled to discover whether a relationship between weather and pain exists.

Can you tell us about how you became interested in the relationship between weather and pain?

The patients at the hospital where I work are convinced that weather changes cause an increase in their pain, and so does my wife, who has rheumatoid arthritis (RA). At the hospital, we had an old dataset with a lot of pain reports. The original study (Smedslund et al 2009) was not concerned with weather, but with the 36 patients with rheumatoid arthritis who had reported their pain on a daily basis for about 100 consecutive days. We had the idea that we could pair the original pain assessments with weather data from the local weather station.

Can you share some of the findings from your work and others into weather and pain? What is the evidence for a connection so far?

We collected data on aspects about the weather such as temperature, humidity, air pressure, precipitation, sun and cloudiness, but also more global parameters such as number of sun spots and the NAO-index (North Atlantic Oscillation). So, for each patient and each day we had pain data and a lot of weather data, and we used a type of advanced statistics called times series analyses to explore whether pain co-varied with weather on the same day or the preceding or subsequent days. We found that the associations between weather and pain were highly individual. Some patients, for example, seemed to get more pain when it got colder, while other patients got more pain when it got warmer. Patients reacted to different weather variables, and some patients seemed to be more weather sensitive than others. Because of the individual differences, it proved difficult to establish strong associations on the group level. Later, we tried to sum up what other researchers had found, and we published a systematic review with nine included studies (Smedslund and Hagen 2011). In these nine studies, temperature, humidity, and atmospheric pressure were the three most studied variables, but the studies did not agree about the effects of these variables. My third study on weather and pain was partly a replication of my study on rheumatoid arthritis patients, but this time with patients diagnosed with fibromyalgia (Smedslund et al 2014). We also had data on mental health in order to explore whether this could moderate the weather-pain associations. We found a small inverse association between pain and atmospheric pressure, meaning that a drop in pressure was associated with a subsequent increase in pain, but this association was very small.

If there is a relationship between weather and pain, what meteorological factors do you think are likely to be involved e.g. sunlight, rain, pressure etc.?

I would say atmospheric pressure and temperature. In our study with fibromyalgia, we found pressure to be associated with pain. Other researchers found that rats showed increased pain-related behaviour when the pressure was lowered in the laboratory, and this association was much stronger than the one observed in humans. In fact, the rats responded to a reduction in pressure corresponding to what is typically observed when a front system is passing. Think of what happens if an astronaut's space suit is depressurized. We need atmospheric pressure to survive. So it should not be surprising that even slight changes in this pressure occur in people who have pain in their muscles, and/ or joints. Temperature could be involved because objects expand when the temperature increases and compress when it decreases. Maybe different body tissues expand and compress at different rates when the temperature changes. This would stress the muscles and joints and cause pain.

What have been the limitations of research that has been carried out to date?

Pain is subjective and it can't be measured, only assessed. People can also have pain in many places in the body, and so when asked to report their pain, they may attempt to mentally average the pain they have in many places. People also often have difficulty with rating scales. They might manage to report on whether their pain is high or low, or even high, medium or low but if they are asked to report on a scale from zero to ten, it can be extremely difficult to distinguish between a pain level of 7 and 8. Also, is a pain rating of 7 for one patient equal to 7 for another patient? In summary, I think the unreliability of pain is a limitation. Another limitation is that people stay indoors most of the time, while the weather obviously stays outdoors. Atmospheric pressure is an exception.

What value do you think that the Cloudy findings could add to the existing knowledge on the subject?

The Cloudy project's strength seems to be that it records a large amount of data in real time.