The authors of this paper suggest that the obesity trend started ramping up in the 70s - 80s and has continued unabated.
Sure enough, this is directly negatively correlated with tobacco consumption per capita.
https://cancercontrol.cancer.gov/sites/default/files/2020-06...
I'm surprised the authors didn't touch on this -- it's a really thorough research paper. I have to imagine that nicotine consumption (or lack thereof) is directly correlated with obesity considering it's both a neurochemical (something the authors were suggesting as a contaminant), we know it affects appetite (it's a stimulant), and those most prone to smoking are also most prone to obesity (remove cigarettes = more food intake).