Types of studies like this don't actually prove causal relationships between the variables, only correlations, so the woman thing isn't that strange. It may be that women don't ride as hard, but it may be that the biking isn't the significant thing that leads to those men having cardiovascular benefits. We already know that exercise is helpful, but the question of whether this exercise is significantly beyond what the normal population does or is instead simply a substitute for some other exercise or generally not significant in its own right when done at the average commuting levels is not clear.

BTW, there were more women than men considered "active commuters" from the sample, so it isn't a sample size issue for the women.

There's a huge confounding variable in this study and those like it in that people who actively commute are obviously already more likely to be more fit before they start such activity (since to the unfit person, it is a more difficult/strange task to actively commute). If women are more likely to actively commute as a form of corrective exercise than men, that could explain these results in that women who actively commute are more likely than men to be in worse shape than the normal population in the first place.