I believe your observation is not precise enough because of the limited sample and because of some specific sorroundings.
Not that is wrong, just it is incomplete, leading to some unrepresentative result.
I could proceed with further inferences, saying that your vapers were originally all part of the same group of smokers, and one day they decided to switch to vaping, making a subset out of the original set.
Therefore, I infer, all chubby people are more careful about their health.
You see from the above, that an observation remains mostly a perception affair. Correct to yourself, but unprovable to others.
Only a logically tought set of rules coupled with a mathematically certain sampling method - can change an observation into a statistically valid measurement.
Which takes a lot of time, money and knowledge to put together.