I think it's troublesome to compare general disappearance data from so long ago. There can be long-term trends of things like waste and changing methodologies that mess things up.

I think your point -- that overall carbs are not the issue -- is right on, though. Looking at wheat data from Statistics Canada going back about as far, people seem to have been eating a ton more wheat a hundred years ago than they do today, like 30% more. And yet in those days obesity had a <1% incidence, by some accounts.

I think that's not inconsistent with fingering particular carbohydrates (and even the way they're consumed) as the cause of modern weight problems. That is, I don't think that falsifies the Sunny Delight hypothesis I proposed earlier. There's a reason the Kool-Aid Man is that large -- as mentioned by others, it seems like fructose is the trans fat of carbohydrates.

And, I would say, food disappearance data doesn't help at all in answering the question of the best way for the obese to lose fat. After all, in the 1910s, there almost weren't any obese people! Now that over a third of Americans are statistically obese, this is its own highly relevant issue.

For the record, I realized something interesting about wheat consumption in the USA, which is that sugar follows wheat very often in the modern diet. It's very challenging to find packaged sandwich bread in a USA supermarket which doesn't contain a significant amount of sugar or HFCS. So today, wheat consumption tells you a lot about sugar consumption, but that may not have been the case a hundred years ago.

Again, you can't lump all carbs together. You can't just say carbs are bad for you. Fiber, for example, is a carb, but isn't going to make you fatter.