Antioxidants are more complicated than we thought


As of September 2017, new Sweat Science columns are being published at Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


I wrote last month about emerging evidence that taking antioxidant supplements may actually be counterproductive, especially for athletes. There’s an interesting paper coming up in Cell Metabolism that casts further doubt on the whole basis of our belief that (a) oxidative stress is what causes aging and other cellular damage, and (b) antioxidants can counteract these effects (press release here; abstract here).

What’s new about the study is that they didn’t use the typical indirect methods to guess about levels of reactive oxygen species (ROS – the bad guys) and antioxidants, averaged over the entire body. Using transgenic fruit flies, they were able to measure ROS activity directly and in real time, in different parts of the body. There are a whole bunch of surprising results in this study that fly in the face of the conventional antioxidant theory. A few of the key ones:

  • If oxidation is linked to aging, we’d expect to see more ROS as the flies get older. In fact, ROS levels stay constant throughout the aging process except in the intestine. Only the intestine shows gradually increasing levels of oxidation.
  • Even in the intestine, the flies who lived longest had ROS levels that increased more quickly than the flies who died earliest — exactly the opposite of what we’d expect if oxidative stress is linked to aging and eventual death.
  • When the flies were fed an antioxidant, their production of ROS increased in response.

This last point, of course, is most interesting from the perspective of supplementation. It suggests that there’s a homeostatic response that attempts to preserve the balance between oxidants and antioxidants at a roughly constant level. So if you take a lot of antioxidants, your body simply ramps up its level of oxidants in response — with who knows what effect!

Let me emphasize: this is a study of fruit flies. Fruit flies are not humans. Still, the results suggest that our basic understanding of the real-time dynamics of oxidant-antioxidant interactions is far from complete. Now, I’d happily ignore all this cell biology stuff if there were lots of controlled trials showing increased health or performance from antioxidants. If it works, who cares about the mechanism? But since (a) we have no consistent evidence that antioxidants work, and (b) it’s increasingly clear that we don’t even understand how or why they should work, or what supposed problem they’re supposed to fix, I still think a balanced diet with lots of brightly coloured fruits and veggies in the way to go. Not powders, and not pills.

13 Replies to “Antioxidants are more complicated than we thought”

  1. Very interesting, and I fully subscribe to your conclusion. Yet, as you point out, the question is how good a model fruit flies are for human biology in this particular respect. I can think of at least two reasons why this is not necessarily the case.

    First, the life cycle of a fruit flie is measured in days. Though I am anything but an expert on fruit flies (though sometimes, I seem to be an expert breeder), I doubt that oxidative stress is a such a big issue in fruit flies. They can afford a homeaostatic reaction, (perhaps) unlike humans without pills.

    Secondly, Fruit flies live their larva stadium in fruit. If the balance between ROS and antioxidants is an issue, they have plenty of vitamin C to draw on.

    Nonetheless, a very useful reminder that modeling/marketing every health issue as a deficiency disease is silly. Many thanks for that!

  2. Thanks for this interesting perspective. I’ve long recommended against exogenous antioxidant supplementation (especially megadosing of specific “antioxidants”), especially for athletes, given that supplementation seems to block the adaptive training effect. But I had to think when reading this post, “is upregulation of ROS necessarily a bad thing?” Considering that ROS are a potent weapon that immune cells use to selectively target and destroy bacterial invaders, might the increased ROS production indicate enhanced immune function, and not indicate any sort of (assumed) harmful oxidative effect? In light of the positive correlation between ROS levels and lifespan, I’d hypothesize that supplementation might be good (for fruit flies). I guess the long story short is (and has been for a long time): eat nutrient-dense, non-toxic foods and call it good. And live a long and happy life.

  3. I too advise against exogenous antioxidant supplemetation based on the adaptation inhibition. However, we still have to put things in context I think. The average person going to the gym 3 times a week certainly does not need antioxidant help but would the elite triathlete training 3 times a day possibly benefit ? It comes down to whether athlete is suffering from acute versus chronic oxidative stress. Acute – good, chronic – bad. There doesn’t appear to be a direct measure of this yet, there are things like TAC and other inflammatory biomarkers, but each have there own variables.

    Another important thing to consider is training versus competition. So maybe advise against supplementation during training but possibly use it during competition. For example, with Tour de France cyclists, during those 3 weeks, oxidative stress is obviously very high, therefore supplementation could help. It’s a similar case with something like ice-baths, again the jury is out as to whether they blunt the adaptative response or not but they do help reduce pain. So don’t use them for training but use them during competition…

    I am all for the nutrient dense diet in terms of antioxidants. However, I don’t think the door is fully closed on them yet. Lets see studies on elite high performance athletes with big training volumes and measure acute versus chronic stress with specific biomarkers. Then carry out double blind cross-over studies on elite cyclists during big stage tours. Then we can decide…. not sure how practical those studies would be though 😉

  4. Barry,

    I’m totally with you when it comes to context determining the appropriateness of a strategy, and that antioxidants *may* be right for elite athletes during competition. There’s a critical distinction between training and competition that plenty of folks totally miss. Good luck eliminating confounding factors with your double-blind crossover studies of elite athletes, though! 🙂

  5. @RH: Agreed – fruit flies can’t tell us how the human body works. The significance here is that the total failure of current theories to predict the results in this experiment suggests that, at a very fundamental level, we know a lot less than we think about antioxidants. That’s actually the final sentence of the article:

    “In conclusion, many of the observations we report could not have been predicted a priori and therefore highlight the importance of monitoring redox states in vivo.”

    In other words: more research needed, as usual. 🙂

  6. @Barry and @Dallas: Agreed about the potential for different strategies in training versus competition. That’s exactly what I wrote back in 2009 when I blogged here about one of the early studies that found impaired recovery in elite athletes after antioxidant supplementation:

    “When you consider that hypothesis, the conclusion you draw is that acute supplementation right before a competition or intense workout might help you, but chronic (i.e. daily) supplementation will make you recover more slowly.”

    That being said, the message I get from this study is that our incomplete understanding of the basics of redox signalling makes it essentially impossible to predict based on chemistry how the body will react to antioxidants under any given circumstances. They may well help during competition, pre-competition, post-competition, or during certain phases of training or after certain types of workouts. Or it may be that, in the right combination of the right doses of the right antioxidants, they help at all times. The point is, right now nobody knows with anything approaching certainty what those circumstances or doses actually are. So to me, taking antioxidants for performance is basically like flipping a coin — maybe it’ll help, maybe it’ll hurt, and you have no control over which one it is. But that coin toss is weighted by the fact that any negative effects are likely smaller than the potential placebo benefits. So it’s not a bad gamble to take; it’s just not one with any evidence behind it.

  7. Dallas – yeah, could be tricky ! But its the sort of study that is really needed.

    Alex – certainly the jury is out and the evidence is more in favour of not taking antioxidants regardless. However, its hard to really take anything from this fruit fly study… really need to see it in humans. Just like quercitin increased endurance in mice, but not so much in humans. Which reminds me of another possible benefit of certain antioxidant compounds, being a signal for mitochodrial biogensis. So maybe excluding the redox balance, there still could be a case for certain antioxidant compounds to promote cell signalling. I might be going off on a tangent here !?

  8. In my opinion, most supplementation outside of where a clear deficiency is present is useless or even counterproductive. An antioxidant in its “natural habitat” is just one of many physiologically active substances that probably depend on each other. An apple has about 2,000, IIRC, and if you use a multivitamin you may get 12 – where are the other 1,988?

    Beside the one mentioned above, this study,

    as well showed that supplementing vitamin C, perhaps *the* antioxidant in the public eye, actually decreased performance among endurance athletes.

    Yet the practice of supplementing it in large doses is still very much alive. There was virtually no market for vitamin C before the 1930s, then Hoffmann – La Roche managed to artifically synthesize it and started a marketing campaign about everyone having a vitamin C deficiency.

    It seems like it took us 80 years to have research catch up with a marketing idea that probably took a couple of weeks to come up with.

    That may be the problem: You can come up with bogus claims a lot faster than it takes to prove them wrong.

  9. I must say I am a huge believer in taking supplements and this study with fruit fly’s is not very convincing. What about the other studies that prove the other arguments. That supplements is a very strong antioxidant. It of course depends on what supplement you take. You should be taking something,

  10. @Michael: “What about the other studies that prove the other arguments.”

    Which studies are those?

  11. Interesting studying, especially considering that I just let someone talk me into buying a bottle of Protandim. I’ve started taking it, but I’ve also done some additional research on it that leads me to believe that it is pretty useless.

Comments are closed.