COVID-19, lockdowns, and home workouts

(This post is an attempt at light-heartedness in the midst of a global tragedy. This is my way of keeping myself from thinking morbid thoughts. If this piece distracts you, makes you smile, even for a few moments, it would have served its purpose.)


COVID-19, lockdowns, and home workouts

If you have reached this page looking for ways to exercise in your home during your confinement, please accept my sincere apologies. The intent of this piece was to merely reassure myself that people like us aren’t freaks, and are to be found all over the world. Which, I found to my dismay, is not true. But the internet is a seductive tool, and offers up much else that appears to be grist for the mill, particularly to an idle (and biased) mind! So, it ended up being a little more fun than that.

As over half the world is in some form of lockdown, I wanted to see if people were looking for ways to exercise within their homes. As everyone knows, the go to place for indulging oneself in such inconsequential pursuits is the Google Trends website (https://trends.google.com/trends). The next step was to find that one critical search term that would encompass all the collective craving for a satisfying exercise regimen which would work within the confines of the home. I settled for the term “home workouts” (See footnote*). I looked for “interest over time” for this term over the last 90 days. The result was a nice pleasing graph which rose steeply in the beginning of March 2020 when most of the social distancing and movement restrictions came into full effect. (Upper panel in figure 1) But the kicker is that Google presents “interest over time” on a relative scale. So we don’t get an idea about the absolute numbers. One way of getting a sense of this is by comparing it to the interest in a really popular term. The lower panel in figure 1 compares the interest in “home workouts” relative to “COVID”. Clearly, exercise doesn’t seem to be on most people’s minds. The insignificance of that blue line compared to the red was crushing.

Figure 1: Interest over time in the term “home workouts”

But as I said earlier, the internet provides avenues for distraction and self-delusion. Google Trends also provides country rankings, in this case for 26 countries. There was wide variation between countries in their appetite for exercise. The top 5 included the usual suspects such as Australia, New Zealand, Canada and South Africa. The US was a surprise member of the top 5. Towards the bottom were Mexico, Indonesia and Brazil. I would’ve expected the South Asian countries to have been at the very bottom. But perhaps I am being unkind. Maybe the intent is there. Or maybe they get all their exercise working on the computer! (Figure 2)

Figure 2: Variations between countries in their interest in “home workouts”

Physical inactivity and interest in home workouts: You are what you search!

With large variation comes the possibility of delectable correlations (real or otherwise). If a larger proportion of a country’s population is ordinarily physically inactive, we would not expect them to start losing sleep over lack of exercise when under lockdown. I collected data on the proportion of population which was insufficiently physically active for all the countries (Lancet Glob Health 2018; doi: 10.1016/S2214-109X(18)30357-7) which were in a lockdown in their entirety (for example, I excluded the USA, as different states had different levels of restriction at the time of writing). Here again, the data on India were surprising. It appeared that only about a third of Indians were insufficiently physically active. Perhaps the surveyors mistook intent to exercise for actual physical activity! The Germans and Italians were on the other hand surprisingly lethargic. The scatter plot showed initially showed no correlation between the countries’ Google search interest rank and the proportion of their people who were physically inactive. (Figure 3A) But this was mainly because of two outliers, Spain and France, who despite having a physically active population, appeared not to be interested in home workouts.



Figure 3: Relationship between country Google interest rank and the proportion of the population with insufficient physical activity, with (A) and without (B) Spain and France

A. 

B.

It is possible that the French and the Spanish search for ways to do home workouts in their native languages too, which Google doesn’t track. The Germans are probably more liberal about using English. In any case, removing France and Spain produced a neat and satisfying relationship! (Figure 3B)

Now, though this relationship is quite plausible, there are several reasons why what I’ve shown you may not be true. For one, Google Trends data change unpredictably. When I repeated my search today, Guam was the top country which was interested in working out at home! The problem with the Trends result is that it is a relative, not an absolute metric. If for example a mere 100,000 people in Guam had searched for ‘home workouts” they would be undisputable number 1 because 61% of the entire country’s population was involved. Which is why I excluded the small countries from this analysis. Second, searches in a language other than English are probably poorly represented. And then there is the problem of drawing conclusions from aggregate level data (something called the ecological fallacy, touched upon in an earlier post http://randomramblings2018.blogspot.com/2019/06/eggs-meat-and-happiness.html). But all things said, you will agree that it makes for good copy to conclude that, “You are, indeed, what you Google”!

__
*If you came here looking for detailed methodology on how I chose the search term, I am sorry to disappoint you. It was pure whimsy.