Correlation is not causation . . .

. . . but this association is both alarming and worth a lot more study;

Plus, overall, people with long commutes are fatter, and national increases in commuting time are posited as one contributor to the obesity epidemic. Researchers at the University of California–Los Angeles, and Cal State–Long Beach, for instance, looked at the relationship between obesity and a number of lifestyle factors, such as physical activity. Vehicle-miles traveled had a stronger correlation with obesity than any other factor.

Unlike things like genetic factors (which tend to be reasonably stable within decades at the population level), the shift towards longer commutes has moved in parallel with increases in obesity. If the association persists at the individual level then it is a possible candidate for obesity prevention.

The problem, of course, is we continue to build low density, car centered cities which do not improve matters. It makes it hard to change average commutes on any sort of easy time scale. But a discovery that this association was causal would suggest all sorts of public health interventions.