Yes, social changes
could produce dramatic changes. But social changes themselves are usually very slow, and I know of few exceptions to this in the context of health. Normally people blame obesity on social aspects (like exercise) which 1) have changed very gradually and 2) don't matter anyways (explaining that would make this a much longer discussion). If someone could come up with an example of just 1 social change that dramatically shifted in 1980 I would be very open to the idea. The only thing I know of is weight-training- the obesity numbers used are normally just based on BMI- they don't take into account that muscle isn't fat.
Here are some actual numbers:
http://en.wikipedia.org/wiki/File:USObesityRate1960-2004.gif
What we see is a linear increase starting around 1980, where before then rates appeared to be constant.
Before 1980, 1/3 of Americans were overweight. Every overweight person has the same metabolic problem as the obese, just to a lesser extent. The fact that 1/3 of Americans are now obese 20 years later is what I would expect due to a gradual decline of American health that hit a tipping point for obesity around 1980.