There is a widespread myth that the obesity epidemic started in or around 1980. This is based on a misunderstanding of the relationship between body fat percentage and BMI, which is used to classify someone as "obese". 🧵

You can see this nonlinearity replicate in numerous contexts. For example here it is in the Heritage Family Study.

The distribution of BMIs shifts right as bodyfat percentages increase across the distribution, but the mean and variance increase faster than body fatness does due to that nonlinearity, which shows up because it's part of how BMIs are constructed.

Body fat percentages are much closer to being normally distributed, on the other hand. With these things stats in mind, it should be clear how, in time series, you can get an apparent acceleration of the increase in BMI even as body fatness linearly increases.

A simple way to grasp what's going on here is to look across different BMI deciles to see how those have trended over time. For White men, here's the picture:

For White women, we see something similar: smooth increases!

These authors explained that, contrary to the idea of an emergent epidemic in the 1980s, the persistent increase in BMI that we're familiar with today was already underway, what we see is just masking due to nonlinearities that their method makes apparent!

1980 was therefore not a special year for the obesity epidemic. But there's more and very interesting data that should also help us speak to the causes. For example, well-off men attending Yale and Amherst College started seeing sharply rising BMIs long before 1980:

Those men weren't exposed to today's various "toxins", "processed foods", and high-fructose corn syrup wasn't even invented until 1957, when this series actually ends!

You know another well-off group that had a rapidly rising BMI before all of the modern post-1980 contamination-based explanations for the obesity crisis? (Native-born) Major League Baseball players!

Somewhat less elite, but still up there: High school All Americans.

But you know what's really inconsistent with a post-1980 acceleration of BMI gain? There was nothing for kids in general. Their BMI changes, if anything, contradict acceleration in the suggested time period. Consider White girls first:

Now look at Black girls:

White boys:

Black boys:

You know what actually predicts rising obesity? The introduction of television. Increased food availability. Aging. The wealth of society! Increasing wealth has been a thing for a while now, so we should not be surprised to learn people have been getting fatter all the while.

Sources:

yes