Antioxidants block cell repair —
New information and what it may mean

by Eric Drexler on 2010/09/26

Antioxidants inhibit basal autophagy and block the induction of autophagy by calorie restriction and other means. Because this effect inhibits the central mechanism of cell repair, it helps explain why dietary antioxidants have failed to deliver their expected benefits to health and longevity. The nature of the effect suggests prudent modifications to popular supplementation regimens.

Promising, but...
“Imagine this program could keep you young…”
(© 1999)
… but it won’t …

The puzzle:
Dietary antioxidant trials
show little benefit

A paper published a few weeks ago may help answer a long-standing question in aging research: Why have treatments that protect cells from oxidative damage produced so little net benefit?

It seems reasonable to expect benefits from increasing antioxidant protection in cells. Reactive oxygen species (ROS) damage cell components — especially mitochondria — and elevated ROS levels can shorten life, causing cancer, heart disease, neurological damage, etc. Dietary interventions successfully reduce ROS levels, hence the damage caused by oxidative stress, by increasing antioxidant activity in cells (for example, by elevating levels of antioxidants like vitamin E, superoxide dismutase, and α-lipoic acid).

Surprisingly, though, the well-studied dietary antioxidants have delivered little, no, or negative benefit in reducing age-related disease and mortality. Consider this review, “Mortality in Randomized Trials of Antioxidant Supplements for Primary and Secondary Prevention: Systematic Review and Meta-analysis” (JAMA, 2007):

In 47 low-bias trials with 180 938 participants, the antioxidant supplements significantly increased mortality….beta carotene…, vitamin A…, and vitamin E…, singly or combined, significantly increased mortality. Vitamin C and selenium had no significant effect on mortality.

In each case, something is offsetting the benefits of ROS protection, yet these interventions have nothing in common but ROS protection itself. Stranger still, that something seems to result in the accumulation of ROS damage. But how can reducing the rate of damage fail to reduce the accumulation of damage?

There’s a relevant mathematical truism:

(damage accumulated) = (damage caused) – (damage removed).

The bad news:
Antioxidants down-regulate autophagy (damn.)

Autophagy removes cell components — including ROS-damaged proteins and organelles — by engulfing and digesting them, producing wastes and recycled nutrients. It’s ongoing, tightly regulated, and as I discussed in “Autophagy: Why you should eat yourself”. it’s essential to life and health,

Upregulating autophagy has metabolic costs (burns scarce calories) but has extraordinarily wide-ranging benefits. Interventions that extend healthy lifespan in animal models include calorie restriction, resveratrol, spermidine, and rapamycin, and in each operates, at least in part, through autophagy. Upregulating autophagy has positive effects in models of several specific neurodegenerative diseases, too, as I discussed in “Trehalose, autophagy, and brain repair”.

Interventions that increase production of ROS induce increased autophagy in response. As one might expect, antioxidants that reduce ROS have the opposite effect, because they reduce this inducer. There are, however, other inducers that don’t work by increasing ROS, including rapamycin and the inexpensive wonder-sugar, trehalose.

The surprising, nasty part of this story is reported in a big new paper, “Antioxidants can inhibit basal autophagy and enhance neurodegeneration in models of polyglutamine disease” (in Human Molecular Genetics, 1 September 2010). The paper has a lot of content, but in outline:

All tested antioxidants blocked autophagy induction, and all autophagy inducers were subject to being blocked. This included rapamycin and trehalose, which have nothing to do with oxidation and ROS levels, and it includes calorie restriction, too.

With caveats regarding the antioxidant doses used and the limitations of both in-vitro experiments and animal models, it seems that:

  • Life-extending interventions induce autophagy.
  • Strong antioxidant interventions block this effect.

In short: Slowing damage interferes with repairing damage.

Possible directions

This result suggests lines of research, but perhaps also directions for prudent modification of (still) popular antioxidant supplementation regimens.

Levels of antioxidants

One response is to give more credence to the negative results of antioxidant trials, because evidence for a mechanism always reinforces evidence for a result, and in this instance, the nature of the mechanism also suggests that the negative results can be generalized. This shift in the balance of the evidence tends to discourage antioxidant consumption, at the margin.

But which antioxidants, and to what extent? Although “antioxidants” may share a name, their effects differ, as do people, their aims, and their states of health. Vitamin E is not interchangeable with cocoa powder (in fact, nothing is interchangeable with cocoa powder…[slurp, return cup to table]…). Context matters, too: for example, antioxidants can have a strong positive effect on the immediate, postprandial response of the vascular system when consumed with a meal.

Timing of antioxidants

A second, compatible, direction of response might be to try shift the balance of cost and benefit associated with dietary antioxidants.

The key is that a brief interval with intensified autophagy could potentially induce a lot of repair, while forgoing supplementary antioxidants during that interval would sacrifice at most a small increment of benefit (a fraction of whatever the cumulative benefit might be).

In other words, sometimes remove the blocking agent.

How much good might upregulated autophagy do, and how quickly? Research noted in this review — “Towards an Understanding of the Anti-Aging Mechanism of Caloric Restriction” [pdf] (in Current Aging Science, 2008) — suggests surprising potential:

Recent data show that the acute stimulation of autophagy by the injection of an antilipolytic* drug can rescue older liver cells [in rats fasted for a day] from the age-related accumulation of oxidative damage of mtDNA in less than 6 hours [as indicated by levels of the marker 8-hydroxy-2-deoxyguanosine]

A strong stimulus evidently changes not just the rate of autophagy, but also the targets.

[*Addendum: Antilipolytics (for example, Acipimox high-dose niacin) inhibit fat metabolism.]

Research in autophagy is exploding. I’ve seldom explored a literature where so many of the important papers are less than a year old.

What I’ve written here is just a sample of some of the recent information, together with a few ideas about what some new information may mean for an old puzzle and some practical questions. I’m sure that there are further insights (and corrections) that can be extracted from the literature in place today, and I look forward to seeing that literature itself become half-obsolete next year.

Meanwhile, please don’t inhibit autophagy in your cells (at least not all the time), and consider giving them a healthful autophagy-inducing kick from time to time. The current state-of-the-art advice (Nature Cell Biology, September 2010) is basically simple: Fast.

See also:

{ 18 comments… read them below or add one }

bbot September 26, 2010 at 9:49 am UTC

Unfinished sentence:

>The bad news:
>Antioxidants down-regulate autophagy (damn.)
>Autophagy is
>Autophagy removes cell components — including ROS-damaged proteins and organelles — by engulfing and digesting them, producing wastes and recycled nutrients.

Thanks! (fixed)
— Eric

Fabio September 26, 2010 at 6:01 pm UTC

From a practical point 0f view, I wonder how long it might take for the autophagy-inhibiting effect of antioxidants to wear off. Is it just a case of not taking supplements on fasting days? That should work for something like Vit. C, but I suspect that fat-soluble vitamins might present a problem…

Btw, thank you for these posts on autophagy – fascinating and useful stuff! I’m reporting your findings on my Italian-language blog and you might like the rather graphic image I used to illustrate the concept of autophagy (taken from an old Italian comic):


Hi Fabio — Yes, the persistence of effects must vary widely among different antioxidants. Because the effects are approximately additive, rapid depletion of a subset might be sufficient for the purpose. I should add that there’s also a question of adaptation, that is, the extent to which homeostasis of antioxidant activity counters the effect of relatively long-term supplementation. I’m sure there’s a literature on this, but I’m not familiar with this in a quantitative sense (for example, what is “relatively long-term”, and how thorough is the homeostatic response?). This is presumably a contributor to the attenuation of hoped-for benefits of antioxidant supplementation, and a related point is that general system-dynamic considerations suggest that an overall homeostatic response would result in a shift in the pattern of antioxidant protection (because the effects are only approximately additive): With what might be called low-resolution homeostasis, artificially increasing protection of one sort (e.g., of aqueous cytoplasm) would likely lead to a decrement in protection of another sort (e.g., of components in one or more membranes). There is surely a literature on this, too.
— Eric

Brian September 27, 2010 at 12:02 am UTC

Fascinating. I had heard that anti-oxidant consumption was correlated with increased morbidity, but I figured it was probably just a correlation due to the self-selected nature of the those who take anti-oxidant supplements. Perhaps they took supplements in lieu of mainstream medicine or something. It is interesting to learn that there may be a causal relationship.

Someone should forward this article to Kurzweil; maybe he’ll re-think his nutty ideas on health and anti-oxidant supplementation.

Charles Colenaty September 27, 2010 at 10:04 am UTC

In regard to antioxidants blocking autophagy — I wonder what time period is involved. Things move so quickly in our cells that a lot can happen in a second. Any delay between the deposit of misfolded protein or a dead mitochondrea provides opportunity for oxidation, and so when the protein is enfolded phagosome for delivery to a lysosome it is in a state of oxidation that does not allow for complete digestion. The undigested residue is lipofuscin, the stuff that liver spots are made of. The lysosome eventually fills with lipofuscin and leaks it out into the cytoplasm of the cell. This ongoing buildup of lipofuscin impairs cell function and is a major player in many age related diseases, such as adult macular degeneration, Alzheimer’s, and possibly the prostate hyperplasia that is the bane of elderly men (like me). Somehow, autophagy can be severely impaired in portions or all of certain organs. This is apparently what happens in some forms of autism, where the resulting buildup of lipofuscin results in extensive neuron death in the word processing area of the brain as well as in motor areas. Autophagy is negatively affected by H1N1 flu, where the virus prevents the autophagy phagosomes from fusing with lysosomes.

genoki September 27, 2010 at 5:27 pm UTC

Kurzweil & Grossman note this study (and other studies noting increased mortality with antioxidants) in their book ‘Transcend’ p 316. They point out two serious problems. Study involved alpha-tocopherol for vitamin E as opposed to mixed tocotrienols, and single dose vitamin A which is not recommended way to take vitamin A. Further, they note that out of 815 studies on vitamin A and E, only 68 were included. Finally, the study only looks at vitamin A and C, alpha-tocopherol (for vitamin E) and selenium, ignoring numerous other antioxidants.

In short, the jury is still out. However, while I will continue with antioxidants, I will be paying particular attention to their impact on autophagy, and autophagy in general as Drexler’s notes is rapidly growing area of knowledge.

Hi Genoki — Thanks for noting Ray’s recent (2009) assessment of the evidence. He and Terry Grossman have invested far more time and thought in this area than I have. I’m confident that they could propose better experimental interventions than most we’ve seen, and that the results of many reports folded into the meta-analysis that I cited should indeed be substantially discounted. Their critique should be seriously considered in forming one’s own assessment.

There are, of course, many results from animal experimentation, and these also influence my views. I’d be interested in seeing a collection of reports on the most positive results to date, if someone would like to link and summarize it here.

I’d also like to see trials of regimens along the lines that I discussed in the post: An example might be well-chosen antioxidant supplementation 6 days per week, together with autophagy-promoting conditions for one day. Suitably adapted pilot studies in some of the standard aging models (yeast, cell culture, C. elegans, Drosophila) could quickly give a good indication of what might be worth trying in mice.

Note that my post is about an unexpected negative effect that is a partial explanatory factor for results that are already known, which is that the widely anticipated broad, robust, beneficial effects of antioxidants haven’t materialized. It doesn’t provide new information about those results, and should not be regarded as an assessment of the best regimens known today.

— Eric

Sean Kiely September 28, 2010 at 2:51 am UTC

I follow posts like this with great interest.

As the amount of science and technology information available on the web rapidly increases, I find myself not only tracking accumulator sites (like Futurepundit and Nextbigfuture) but increasing following sites like Metamodern that analyze what developing trends are significant, well-studied and poorly understood. The flood of information is becoming too much to track. . . I need competent analysis!

genoki September 28, 2010 at 5:33 am UTC

@Sean – that is so very true. I have been getting good info from increasingly diverse sources.

Slashdot is still good (hint use mobile site, Lifehacker

Also sci-fi writers like Charlie Stross have techy-geeky followers with lots of news (incidentally links back to Drexler on this study)

Kurzweil’s news aggregator

Jordan Fisher September 28, 2010 at 7:13 am UTC

Any idea how metabolically expensive autophagy is? Cell upkeep in general? If we induced high levels of autophagy 24/7 would we need to stuff ourselves to get enough calories? And, really.. wouldn’t that just be an additional selling point of the drug? =P

Metabolically expensive enough for evolution to give us bodies that under-invest in maintenance, metabolically cheap enough to make little difference in a world of chronic abundance. Somewhere in that range.

— Eric

Bleyddyn September 30, 2010 at 9:45 pm UTC

The figure I see for Caloric Restriction is -%30 calorie intake. However, I have never seen a similar description of ‘fasting’, in terms of improving health and/or increasing autophagy. Have studies on fasting just not been done, or am I just not finding them?

Does fasting mean reducing calories, or not eating at all?
For 12 hours? 24 hours? More?
Once a week, once a month, once a year?

Hi, Bleyddyn — I wish that these weren’t such good questions. The term to search on is “intermittent fasting”, and benefits seem to start at intervals of less than 24 hours. Deep (80%?) reduction in calorie consumption — and especially protein consumption — seems to have some of the benefits of fasting. My reading has persuaded me that occasional (&gte; 1/year) fasts of several days are probably a good idea. I haven’t collected a set of relevant documents on this, however.
— Eric

Dr Peter Britton October 4, 2010 at 2:12 am UTC

Dear Sir
Please advise if you wish more direct information on SKQ1.
We have been involved with the project for the last 2 years.

Bill LEstrange October 6, 2010 at 9:28 am UTC

maybe a study on 5day broad spectrum antioxidants alternating with 1 or 2 days hydrogen peroxide ?? just putting this in to stimulate some feedback on the practicality

Ron Burk October 14, 2010 at 10:34 pm UTC

If I think of autophagy, timing, and anti-oxidants, I think of only one word: melatonin. We already have a big timing system built into the body (though most people in the civilized world have wrecked it with artificial light, irregular bedtimes, poorly timed caffeine, etc.), and it runs on melatonin, whose downstream effects contain too many (timed) anti-oxidants to list (though it’s perhaps worth mentioning the upregulation of superoxide dismutase, as I think that will prove to be a key factor in “host response” diseases like periodontal disease and skin conditions such as acne). You can’t really dose yourself with exogenous melatonin to get the same effect, nor can you easily test a random patient to inspect their nocturnal melatonin surge (the very act of trying to make someone sleep in a lab with wires/tubes can effect the cycle being inspected). Thus, it’s so far easier to “look where the light is better” in problems that may well have the melatonin cycle as their answer.

JohnD November 14, 2010 at 9:56 pm UTC

Drexler writes: “The key is that a brief interval with intensified autophagy could potentially induce a lot of repair…”
This is the general principle of my fasting/autophagy regimen that I began about 15 months ago. I have no empirical data to support the hypothesis that autophagy can take place at a highly accelerated rate when conditions for autophagy are intentionally pushed to the extreme (aside: I can’t even find credible data on the half live of mitochondria). Just my intuitive sense of how chemical reactions work from my prior life experience as a Chemical Engineer. One can make the analogy between the collected cellular debris that one is trying to replace and a pot of water sitting on a stove top. As I see it, the conventional medical wisdom to maximize autophagy is simply to maximize the time the body is in an autophagic mode, without regard to the rate of autophagy taking place, and that this is analogous to setting the heat of the stove top to low in order to evaporate the pot of water. On low heat, the pot of water will heat up to maybe 160 Degrees F. before it comes to equilibrium with the surrounding environment. At 160 F. the pot of water will evaporate in 4 days (vs. 4 weeks for a pot of water with no heat source). But if one cranks the heat up high, and gets the water heated past its tipping point (boiling point), the pot of water will evaporate in two hours.
Applying that principle to autophagy: I fast 2 times a month (water only), 24 hours at a time. At the end of said 24 hours I engage in an hour of strenuous cardiovascular exercise. Theory being that by combining the two most effective means of up regulating autophagy (starvation and oxygen deprivation), I will achieve a synergistic effect which will push the rate of autophagy past the tipping point, and that more mitochondrial autophagy will occur during that one hour of exercise than 500 hours of CARN. I also believe that the body adapts a great deal to regular caloric restriction, and that fasting more than once every two weeks starts to produce diminished results, again I have no data to support that hypothesis.

Lara Pizzorno November 14, 2011 at 10:29 pm UTC

Terriffic blog THANK YOU! Great book on the mechanisms that drive and inhibit autophagy — Avoiding the First Cause of Death by Wulf Dröge (iUniverse, Inc: NY, 2009). I wrote a 3 part review of the research –focused on mitochondrial function — for Longevity Medicine Review — Part III looks at autophagy, and there are ways to stimulate it besides fasting. Here’s a link to this paper:
First section in it is: Restoring Mitophagy – the Key To Mitochondrial Rejuvenation

Andrew February 2, 2012 at 1:15 am UTC

My comment is somewhat late compared to the others, but isn’t resveratrol itself an antioxidant?

Wouldn’t its autophagy-inducing effects be inhibited by its antioxidant properties? Or are its antioxidant properties too weak to offset autophagy-inducing benefits?

“isn’t resveratrol itself an antioxidant” — not to any significant extent.

linkedin login September 5, 2014 at 5:06 pm UTC

Just wish to say your article is as surprising.
The clarity on your publish is simply nice and that
i can suppose you are knowledgeable in this subject. Fine along with your permission allow me to grab your RSS feed to stay up to date with drawing close post.

Thanks 1,000,000 and please continue the enjoyable work.

for more information visit comma September 29, 2014 at 10:41 am UTC

We’re a group of volunteers and starting a new scheme in our
community. Your site provided us with valuable information to work on. You have done an impressive job and
our whole community will be thankful to you.

Dan October 13, 2014 at 12:47 pm UTC

This explains the study about beta carotene supplementation increasing lung cancer

{ 3 trackbacks }

Leave a Comment

Previous post:

Next post: