If you're a serious health nut, you probably exercise regularly and take anti-oxidants to ward off those nasty free radicals.  Only it turns out you probably have to choose:

And as it turns out, antioxidant supplements appear to cancel out many of the beneficial effects of exercise. Soaking up those transient bursts of reactive oxygen species keeps them from signaling. Looked at the other way, oxidative stress could be a key to preventing type II diabetes. Glucose uptake and insulin sensitivity aren't affected by exercise if you're taking supplementary amounts of vitamins C and E, and this effect is seen all the way down to molecular markers such as the PPAR coactivator proteins PGC1 alpha and beta. In fact, this paper seems to constitute strong evidence that ROS are the key mediators for the effects of exercise, and that this process is mediated through PGC1 and PPAR-gamma. (Note that PPAR-gamma is the target of the glitazone class of drugs for type II diabetes, although signaling in this area is notoriously complex).

Interestingly, exercise also increases the body's endogenous antioxidant systems - superoxide dismutase and so on. These are some of the gene targets of PPAR-gamma, suggesting that these are downstream effects. Taking antioxidant supplements kept these from going up, too. All these effects were slightly more pronounced in the group that hadn't been exercising before, but were still very strong across the board.

This confirms the suspicions raised by a paper from a group in Valencia last year, which showed that vitamin C supplementation seemed to decrease the development of endurance capacity during an exercise program. I think that there's enough evidence to go ahead and say it: exercise and antioxidants work against each other. The whole take-antioxidants-for-better-health idea, which has been taking some hits in recent years, has just taken another big one.

But take heart:  exercise may not be nearly as great as we've been told.  It definitely helps ward off diabetes, but most of that benefit comes at very modest levels.  It might have modest effects on depression, heart disease, and cancer, but it's hard to tell because of selection effects:  if you stay depressed, you probably stop exercising.  And it's hard to tease out the confounding factors in the other two:  the exercisers are also thinner, more educated, and less likely smoke than the others.  I was shocked to find out how much of the difference between women's and men's life expectancy was accounted for by their different rates of smoking, and it seems the same sort of thing may be operating here.

Oh, and exercise probably won't make you thin, either, particularly if you're the sort of person who finds it hard to lose weight.  Your appetite eventually seems to increase enough to compensate.

Basically, unless you're at risk for diabetes, Kolata says there's no solid evidence that exercise will do much besides make you sweat.

Of course, my idea of exercise is biking to work, so you'd expect me to say that.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.