Well, I've always wanted to know this too. Jonathan Gruber and Michael Frakes do some research:

The strong negative correlation over time between smoking rates and obesity have led some to suggest that reduced smoking is increasing weight gain in the U.S.. This conclusion is supported by the findings of Chou et al. (2004), who conclude that higher cigarette prices lead to increased body weight. We investigate this issue and find no evidence that reduced smoking leads to weight gain. Using the cigarette tax rather than the cigarette price and controlling for non-linear time effects, we find a negative effect of cigarette taxes on body weight, implying that reduced smoking leads to lower body weights. Yet our results, as well as Chou et al., imply implausibly large effects of smoking on body weight. Thus, we cannot confirm that falling smoking leads in a major way to rising obesity rates in the U.S.

Can't seem to find the full paper, so I don't know exactly how they went about the study, but this seems like a very roundabout way of studying something that shouldn't be too hard to determine, no? Experiment group and control group? Oh I don't know, I'm just the blogger.