One of the mysteries mentioned in the SMTM article was that the obesity epidemic began abruptly some time around 1980. The authors state,
This wasn’t a steady, gentle trend as food got better, or diets got worse. People had access to plenty of delicious, high-calorie foods back in 1965. Doritos were invented in 1966, Twinkies in 1930, Oreos in 1912, and Coca-Cola all the way back in 1886. So what changed in 1980?
This article argues (convincingly, in my opinion) that the obesity epidemic began much earlier, probably around the turn of the 20th century. If you track the evolution of BMI over time in different birth cohorts, you get a very continuous, smooth rise in obesity.
Figure 2. Trend of BMI decile curves of US-born white men by birth cohorts
Guyenet, who’s palatability theory you seem to prefer, shares the same interpretation as SMTM. His chart from The Hungry Brain:
And his commentary:
In 1960, one out of seven US adults had obesity. By 2010, that number had increased to one out of three (see figure 2). The prevalence of extreme obesity increased even more remarkably over that time period, from one out of 111 to one out of 17. Ominously, the prevalence of obesity in children also increased nearly fivefold. Most of these changes occurred after 1978 and happened with dizzying speed. [emaphasis mine]
Note: I’ve partially retracted this comment as indicated by the ETA comments.
I don’t think the quote from Guyenet’s supports SMTM’s argument. Guyenet is merely saying that the trend in extreme obesity has accelerated since 1978. This is consistent with my chart, showing a non-linear (but still quite smooth) line for the 90th percentile in BMI throughout the 20th century.
SMTM says, “Another thing that many people are not aware of is just how abrupt this change was. Between 1890 and 1976, people got a little heavier. The average BMI went from about 23 to about 26. This corresponds with rates of obesity going from about 3% to about 10%. The rate of obesity in most developed countries was steady at around 10% until 1980, when it suddenly began to rise.”
The median person did not merely get “a little heavier”. I think SMTM is misinterpreting the data, and introducing a mystery where one does not really exist. From the chart I showed, the trend in BMIs among average people (say, between say the 20th and 80th percentiles) remained relatively stable over time. There’s neither a hard jump at 1980 nor radical acceleration, though there is some slight acceleration if you squint (and it’s not particularly centered around 1980) [ETA: OK I realized I was being unfair here given that the series stops at around 1986. I’d welcome if someone could find a longer data series. See edit on the other comment too.]. This is also true for the other birth cohorts in the article I cited.
We could do a similar exercise for all sorts of phenomena. Would we say that the Flynn effect began abruptly at some point in the 20th century, simply because a chart showing “percentage of people who are gifted” has a non-linear curve, and picks up speed, say, around 1950? Did economic growth begin abruptly because the percentage of people who are “rich” (defined by some arbitrary cutoff) picked up around 1970?
The abrupt change in obesity is only a strong candidate for a mystery in the sense that it’s not immediately clear why it happened. But once provided the data for BMI over time, it quickly becomes clear that this provides no compelling reason to abandon the palatability theory. [ETA: At least, I have yet to see any compelling reason why the uptick after can’t be explained by slowly rising BMI over the 20th century]
Wait what? This isn’t even a chart of the same thing.
It’s a chart of data from 1880 to ~1980, whereas SMTM as looking specifically as the change in 1980
It’s a charge of BMI, whereas the SMTM chart is looking at growth in extreme outcomes. Say you have a normal distribution with mean 24 and SD 2. Only 0.13% of the population will have BMI over 30. But as the mean BMI slowly increases, you see rapid growth in extreme outcomes. At mean BMI of 26, you’re up to 2% over 30, at mean BMI 28 you’re up to 16%. So the fact that BMI growth is smooth doesn’t imply that obesity growth is smooth too.
It’s a chart of data from 1880 to ~1980, whereas SMTM as looking specifically as the change in 1980
The x-axis in Matthew’s chart is each cohort’s birth year, not the year their BMI was measured. Matthew’s source says the chart includes BMI data from 1959 to 2006.
It’s a charge of BMI, whereas the SMTM chart is looking at growth in extreme outcomes. Say you have a normal distribution with mean 24 and SD 2. Only 0.13% of the population will have BMI over 30. But as the mean BMI slowly increases, you see rapid growth in extreme outcomes. At mean BMI of 26, you’re up to 2% over 30, at mean BMI 28 you’re up to 16%. So the fact that BMI growth is smooth doesn’t imply that obesity growth is smooth too.
Note that it’s a chart of several BMI percentiles, not only mean or median BMI, and it shows a smooth (though accelerating) growth in the highest percentiles of BMI.
This chart of rate of change of BMI (from the same source) is instructive:
It seems that there was a mid-century slowdown in BMI growth starting with cohorts born during the Great Depression, which makes sense, and makes the 1980 acceleration seem more historically unusual than it actually is.
The BMI cutoff at 25 for overweight and 30 for obese is fairly arbitrary. People in different nations, and people in different time periods, have different standards. And we know that the effect of adiposity on lifespan and health is continuous with respect to body fat percentage, at levels higher than those characteristic of very lean bodies.
The phenomenon in need of explanation is why our wastelines got larger over time. Once we provide an explanation for that phenomenon, it’s no mystery why there is a discontinuity in the obesity chart at 1980. That’s merely an artifact of the arbitary cutoff. Overall I see no deep mystery that demands explanation above and beyond rising adiposity during the 20th century. Is it your position that we ought to be find the discontinuity at 1980 mysterious anyway?
ETA: I realize now I gave short shrift to your point about the data series ending around 1980. I actually think it’s ~1986 but it’s still hard to interpret the last six years. I’ll edit my other comment to reflect this oversight.
One of the mysteries mentioned in the SMTM article was that the obesity epidemic began abruptly some time around 1980. The authors state,
This article argues (convincingly, in my opinion) that the obesity epidemic began much earlier, probably around the turn of the 20th century. If you track the evolution of BMI over time in different birth cohorts, you get a very continuous, smooth rise in obesity.
Guyenet, who’s palatability theory you seem to prefer, shares the same interpretation as SMTM. His chart from The Hungry Brain:
And his commentary:
Note: I’ve partially retracted this comment as indicated by the ETA comments.
I don’t think the quote from Guyenet’s supports SMTM’s argument. Guyenet is merely saying that the trend in extreme obesity has accelerated since 1978. This is consistent with my chart, showing a non-linear (but still quite smooth) line for the 90th percentile in BMI throughout the 20th century.
SMTM says, “Another thing that many people are not aware of is just how abrupt this change was. Between 1890 and 1976, people got a little heavier. The average BMI went from about 23 to about 26. This corresponds with rates of obesity going from about 3% to about 10%. The rate of obesity in most developed countries was steady at around 10% until 1980, when it suddenly began to rise.”
The median person did not merely get “a little heavier”. I think SMTM is misinterpreting the data, and introducing a mystery where one does not really exist. From the chart I showed, the trend in BMIs among average people (say, between say the 20th and 80th percentiles) remained relatively stable over time. There’s neither a hard jump at 1980 nor radical acceleration, though there is some slight acceleration if you squint (and it’s not particularly centered around 1980) [ETA: OK I realized I was being unfair here given that the series stops at around 1986. I’d welcome if someone could find a longer data series. See edit on the other comment too.]. This is also true for the other birth cohorts in the article I cited.
We could do a similar exercise for all sorts of phenomena. Would we say that the Flynn effect began abruptly at some point in the 20th century, simply because a chart showing “percentage of people who are gifted” has a non-linear curve, and picks up speed, say, around 1950? Did economic growth begin abruptly because the percentage of people who are “rich” (defined by some arbitrary cutoff) picked up around 1970?
The abrupt change in obesity is only a strong candidate for a mystery in the sense that it’s not immediately clear why it happened. But once provided the data for BMI over time, it quickly becomes clear that this provides no compelling reason to abandon the palatability theory. [ETA: At least, I have yet to see any compelling reason why the uptick after can’t be explained by slowly rising BMI over the 20th century]
See this comment.
Wait what? This isn’t even a chart of the same thing.
It’s a chart of data from 1880 to ~1980, whereas SMTM as looking specifically as the change in 1980
It’s a charge of BMI, whereas the SMTM chart is looking at growth in extreme outcomes. Say you have a normal distribution with mean 24 and SD 2. Only 0.13% of the population will have BMI over 30. But as the mean BMI slowly increases, you see rapid growth in extreme outcomes. At mean BMI of 26, you’re up to 2% over 30, at mean BMI 28 you’re up to 16%. So the fact that BMI growth is smooth doesn’t imply that obesity growth is smooth too.
The x-axis in Matthew’s chart is each cohort’s birth year, not the year their BMI was measured. Matthew’s source says the chart includes BMI data from 1959 to 2006.
Note that it’s a chart of several BMI percentiles, not only mean or median BMI, and it shows a smooth (though accelerating) growth in the highest percentiles of BMI.
This chart of rate of change of BMI (from the same source) is instructive:
It seems that there was a mid-century slowdown in BMI growth starting with cohorts born during the Great Depression, which makes sense, and makes the 1980 acceleration seem more historically unusual than it actually is.
The BMI cutoff at 25 for overweight and 30 for obese is fairly arbitrary. People in different nations, and people in different time periods, have different standards. And we know that the effect of adiposity on lifespan and health is continuous with respect to body fat percentage, at levels higher than those characteristic of very lean bodies.
The phenomenon in need of explanation is why our wastelines got larger over time. Once we provide an explanation for that phenomenon, it’s no mystery why there is a discontinuity in the obesity chart at 1980. That’s merely an artifact of the arbitary cutoff. Overall I see no deep mystery that demands explanation above and beyond rising adiposity during the 20th century. Is it your position that we ought to be find the discontinuity at 1980 mysterious anyway?
ETA: I realize now I gave short shrift to your point about the data series ending around 1980. I actually think it’s ~1986 but it’s still hard to interpret the last six years. I’ll edit my other comment to reflect this oversight.