This week much of the country is seeing the frostiest temperatures in months: All 50 states hit freezing temperatures yesterday. This can only mean one thing, other than a spike in hot chocolate sales (and spiked-hot-chocolate recipes): We are staying indoors.
Or as public-health researchers would say, "people from a variety of populations perceive inclement weather to be a barrier to physical activity." Studies have shown that people tend to take fewer steps in the rain and wind, and more in the sunshine.
Now Jawbone, purveyor of the activity tracker UP, offers some additional support for this idea, courtesy of its users' wrists. The company recently analyzed hundreds of thousands of UP users' steps each day for a year and correlated them with the weather conditions in their area.
The results show, essentially, that we don't like to move when it's either too hot or too cold.
Weekday Steps Taken by Time and Temperature
The sweet spot for walking around was a temperature between the low 60s and high 70s. Weather effects were far less pronounced on weekdays—probably because people still had to go to work. UP users took 5 percent more steps on a 70-degree weekday than a 40-degree one, but the difference was 15 percent on weekends.
Weekend Steps Taken by Time and Temperature
There are a few caveats here: UP users are likely wealthier than average, and this dataset is probably skewed toward the kinds of healthy, tech-savvy people who would buy an UP in the first place. Still, apparently even the activity-obsessed among us can't be convinced to go for a run when it feels like all our mucous membranes are about to freeze over.
The weather is a boring topic for conversation, but it's an interesting impediment to health. One of my favorite theories about the geography of obesity in America that the Southern states, which have the highest rates of obesity, are also intolerably hot and humid in the summer months. Colorado, the leanest state, is comparatively mild, and in the winter it's at least snowy enough for skiing.
It will be interesting to see if hypotheses like that one gain more traction as the proliferation of activity monitors makes it possible for us to cross-reference our physical activity with say, gyms per capita, neighborhood crime, or even the locations of parks.