Jacob Schor, ND
A young man in his late twenties enters my office. He is distraught, pacing back and forth, although there is hardly room for doing so between my desk, filing cabinets, and exam table. I should have been eating lunch, but his urgency on the phone earlier in the day prompted our receptionist to slip him in and so here he was. He’d been a patient for years, really since childhood, but he’d never acted like this. I sit him down and gradually pull out the story. It turns out to be simple. Apparently, he was combing his hair that morning and noticed several gray hairs. It dawned on him that “I’m getting older.”
He went on: “I thought this wasn’t supposed to happen to me. I’ve been taking all those vitamins, all those antioxidant things forever, like since I was a little kid. They were supposed to prevent this.”
Really, this is it; this is his chief complaint. He’s getting older and he had been certain that antioxidant vitamins would prevent hm from aging; he had just realized that someone had been lying to him, most notably me.
Once upon a time a bit more than half a century back, starting in the mid-1950s, so really about seventy years ago, there was a time in which scientists theorized that antioxidants might increase human lifespan.
The idea that my young patient had apparently fixated on is the free radical theory of aging first proposed by Denham Harman in 1956. Harman suggested that “aging and the degenerative diseases associated with it are attributed basically to the deleterious side attacks of free radicals on cell constituents and on the connected tissues.” He expanded this theory in the 1970s to suggest that the mitochondria are the source of these free radicals. As most chronic disease, including heart disease, cancer, and diabetes, increase with advancing age, finding ways to impede the aging process certainly seemed as if it might also have significant implications on other forms of morbidity.1 That seems to have been the start, the reason we put antioxidants on the panacea pedestal.
As I think about this situation, it was more likely his mother who fixated on this idea and then conveyed a message something to the effect, “Take your vitamins so you don’t get old.”
Harman’s theory was interpreted to suggest that if free radicals are responsible for aging and antioxidants can quench free radicals then consuming more antioxidants, as nutritional supplements or as foods, would interrupt the aging process and prevent chronic disease. The idea has proven to have immense sticking power and persists to this day. My patient apparently was among those people who still cling to this belief, yet this idea is no longer considered true. See https://www.hsph.harvard.edu/nutritionsource/antioxidants/.
Somewhere I should mention that “Beliefs are not necessarily correlated with the truth.” I can’t recall though if that is a direct quote, my paraphrasing of something I read, or just an idea that drifted to the surface of my own mind.
If we want to explore the history of this free radical theory, we should go back further than Harman. The beginnings of this thinking probably date back to Paul Bert, who in 1878 described that oxygen at high concentration was toxic, particularly to warm-blooded animals. In cold-blooded animals, oxygen’s toxicity varies with ambient temperature.2 Then in 1908, Max Rubner reported that longevity of mammals increases with body size and that the rate of metabolism of mammals decreases with increases in body size. He combined these two measurements for five mammal species (guinea pigs, cats, dogs, cattle, and horses) and described it as their “lifetime energy potential.” This calculated potential turned out to be constant across these varied species.3
In 1928 Raymond Pearl reported, on an experiment reminiscent of Bert’s, that the lifespan of fruit flies can be shifted simply by changing the temperature of their cages and thus shifting their metabolic rates or, to use the term he originated, their “rate of living.” 4
Over the past century there have been something like 300 theories put forth to explain why aging occurs. Although Harman’s free radical theory is just one of these 300, it has proven to have greater staying power.
Any believable theory to explain aging must explain why lifespans vary so much between species. There is a 200-fold difference in lifespan when comparing shrews to whales and a 30-fold difference between mice and men.5,6 Longevity is tightly regulated within a species but widely variable between species.
A thorough discussion of Harman’s theory of aging can be found in Gustavo Barja’s 2013 comprehensive review .7
Over the years only two factors relevant to Harman’s theory have been clearly correlated with longevity in vertebrates:
1. The rate of production of reactive oxygen species (ROS) in the mitochondria.
2. The degree of fatty acid unsaturation in the cellular membranes.
The difficulty is that the longer a species lives, the lower both these values are. This is the opposite of what Harman’s theory suggested and runs counter to what most of us might guess on a multiple-choice exam.
For years after Harman’s proposal, researchers focused on the effects of antioxidants in search of evidence suggesting a slowing down of the aging process. Antioxidant levels are relatively easy to measure, and the hypothesis was quickly and thoroughly tested. Lopez-Torres reported in 1993 that tissue antioxidants correlated across a wide range of vertebrate species with longevity, but the correlation was inverse; the more antioxidants in their tissue, the shorter an animal’s lifespan.8
By the late 1990s, evidence showed that long-lived animals make way less antioxidants than short-lived animals.9 Hamsters, for example, make 20 times the glutathione as people do.10
Pérez-Campo et al examined 27 correlations between antioxidant production and longevity; of these, 21 showed an inverse association (that is, longer the lifespan, the lower the endogenous antioxidant production), and 6 did not show significant association. There was no example of a positive association between higher endogenous antioxidants and a longer life.
Pamplona and Constantini made a similar survey in 2011, comparing endogenous antioxidant levels with life expectancy in 78 different species; the associations were negative in 72, no different in 6 and positive in just 1 species.11 This all suggests that animals that live longer do not employ a strategy of making more antioxidants to do so.
Supplementation with exogenous antioxidants, that is taking oral antioxidant vitamins, does not appear in most cases to increase maximum lifespan. Antioxidants, when they have any effect, increase mean lifespan in only relatively short-lived animals (lifespan <3 years). This has been interpreted to say that when living conditions were not ideal, the antioxidants protected the animal from early or premature death. At best, taking antioxidants may protect against the chemical and oxidative insults brought on by life, but they do not slow the clock of aging.12
Faith that increasing antioxidant levels is beneficial remains an underlying assumption that has influenced our interpretation of decades of published data.
For example, consider the CARET Trial on lung cancer. Our older practitioners will remember how we once were so sure that beta-carotene could retard cancer growth that our goal was often to dose beta-carotene in doses high enough to turn our cancer patients’ skin orange.
This wasn’t so outlandish an idea. The evidence from observational studies showed people who ate more fruits and vegetables, have a lower risk of cancer, in particular lung cancer. Fruits and vegetables were, and still are, rich in beta-carotene, which can be converted to vitamin A. People with higher serum beta-carotene concentrations had lower rates of lung cancer. Vitamin A was an antioxidant, part of the list of antioxidants we had all committed to memory (A, C, E, Zinc, and selenium). The assumption was that dietary antioxidants conveyed cancer protection. These days given the same association between diet and cancer risk, we might think of all sorts of possible explanation: more phytonutrients, lower BMI with more vegetables, better insulin control, shifts in gut biome, etc. We would not assume it was a single antioxidant that prevented cancer.
The Caret Trial was set up in the early 1980s and was ongoing when I studied naturopathy at NCNM. It was a randomized, double-blinded, placebo-controlled cancer prevention trial with two groups of adults, 18,314 men and women at high risk for lung cancer (think smokers) who received either 30 mg beta-carotene and 25,000 IU vitamin A daily or a second group that received placebo.13 Thus, it seemed reasonable for us to encourage smokers to take these antioxidants while we waited for the study to prove how much they helped. In the worst case, our advice would be a Type I error and not help. The results were worse than that.
CARET was halted ahead of schedule, in 1996, because one group was doing so much worse than the other, so the code was broken. It turned out to be the experimental group that was doing poorly–those taking the vitamins–had a 28% increase in lung cancer, a 17% increase in death, and a higher rate of cardiovascular disease death compared to the placebo group. These increases in risk remained high even years after the participants stopped taking the vitamins.
A follow up report five years after the study was ended showed participants were still at 12% greater risk of lung cancer.14
CARET isn’t the only example of how we mistakenly assumed that antioxidants prevent cancer. I’m thinking of the SELECT Trial.
Before talking about SELECT we should back up and mention a few earlier studies. First, recall Clark’s 1996 selenium yeast and skin cancer paper that inspired us to exhort everyone to take yeast-based selenium. Clark had recruited people with a history of skin cancer to see if selenium lowered risk of recurrence. It didn’t change skin cancer risk (in longer follow ups it maybe increased it) but general cancer risk decreased by half, in particular,the risk of prostate cancer was one third of men receiving placebo.15
At nearly the same time, in 1998, Yoshizawa’s toenail study results showed up. More than a decade earlier, in 1987, toenail clippings were collected from 33,737 men. These men were then monitored until 1994 to see which of them got prostate cancer. Higher toenail selenium levels were associated with a reduced risk of advanced prostate cancer (odds ratio [OR] for comparison of highest to lowest quintile = 0.49). (After controlling for family history, BMI, calcium, lycopene, and saturated fat intakes, vasectomy, and geographical region, the OR was 0.35.)16 Having more selenium in one’s toenails seemed like a smart idea for men.
So, when the SELECT Trial was set up in 1998, we were rather certain that mimicking the study protocol would be a good bet. There was no need to wait for results; the supplements would either help or they wouldn’t. And while we waited, we encouraged our male patients to take selenium and vitamin E following the SELECT Trial’s protocol, believing that doing so would lower prostate cancer risk by half to two thirds. Nearly 36,000 men, aged 50 to 55 years old, were enrolled in the study between 2001 and 2004 and divided into four groups of about 8,700 each; one group took selenium, another vitamin E, the third both supplements, while the last group took only placebo. Two reports were published on these men, the first in 2008 and a second in 2011.
In 2008, after five and a half years taking supplements the risk for prostate cancer was 13% higher for those taking vitamin E, 4 % higher in those taking selenium, and 5% higher for those taking both supplements when compared to the men who took only placebo. The men stopped taking the supplements but were still followed. The supplements continued to have an impact.
In the analysis published in 2014, men who had high levels of selenium at the start of the trial, had almost double the chance of developing a high-grade prostate cancer if they took the selenium supplement when compared to men with low levels of selenium at the start of the trial. This finding was unexpected. Additionally, men with low levels of selenium at the start of the trial had double the chance of developing a high-grade prostate cancer if they took vitamin E. 17
We could list other studies in which early epidemiologic data on diet and health were misinterpreted to suggest that dietary antioxidants were the factor that had lowered disease risk and then subsequently shown to be ineffective when supplemented in randomized clinical trials.
Rather then attempt to create such a list from memory, let’s instead consider Denish Moorthy’s 2013 review that nails this very question. Moorthy and colleagues identified and compared the results of large clinical trials on nutritional interventions, such as those mentioned, with the results that had been predicted based on earlier epidemiological studies. In other words, they asked how often the initial interpretation of epidemiological data eventually been proven true in eventual human trials?18 Moorthy’s analysts identified 34 pairs in which meta-analyses of “well studied” epidemiological findings were matched with randomized clinical trials.
Only a dozen associations between the 34 pairs of epidemiological observations and RCTs were statistically significant. But in that dozen, only half were in same direction as predicted by the epidemiologic observations. The other six, the “discordant outcomes,” pointed in the opposite direction from what was predicted. So much for our rationale of “This might or might not help but it won’t hurt.”19 Half the time apparently, they did. Five out of these six discordant pairs were about antioxidants and disease prevention: vitamin C did not prevent heart disease as predicted, beta-carotene and vitamin A did not prevent esophageal cancer. And so on. The only intervention that backfired, which wasn’t about antioxidants, was whether dietary fiber prevented colon cancer. Early studies suggested fiber would lower risk of colorectal cancer by 27% yet eventual clinical trials showed no benefit. We are left to wonder why nutritional science is so prone to error?
Of course, many of us jump to conclusions whenever we see statistically significant associations in a publication and assume a low p value proves causation. But aren’t the researchers who design these large clinical trials wise enough not to fall for these simple conclusions? I kind of think so.
One explanation for these repeated errors that makes sense is that the researchers, like everyone else, were so entrenched in the paradigm created by Harman’s theory of oxidative stress that they jumped to what seemed to be obvious conclusions and assumed antioxidants were the key factor.
We humans are strongly attracted to dichotomies to process information. Good vs. bad, healthy vs. unhealthy, and so on. Knowing that diets high in fruits and vegetables lower cancer risk, might just as easily have been explained by the old theory of alkaline ash residues—or these days, whether the foods were GMO or not. Back then in our minds we were sure the explanation involved antioxidants.
In hindsight, the failure of antioxidants in cancer prevention seems simple enough. One of the primary strategies the body uses to limit cancer is production of reactive oxygen species both to trigger and power apoptosis. We now often consider the balance between reactive oxygen species and antioxidants as a balance between controlling cancer and protecting healthy tissues.
Why do diets that are high in fruits and vegetables promote better health? If a patient asked me this today my answer would be simple, “What doesn’t kill you, makes you stronger. Plants synthesize a wide range of chemicals to protect themselves from being infected by viruses, funguses, and parasites, or being eaten or damaged by environmental insults. We consume plants in which these antiviral, antifungal, antibacterial poisons are in dilute enough form so that the plant is considered edible, yet these chemicals are still present. These diluted ‘poisons’ are enough to stimulate a reactionary protective response.” Or at least that’s my thinking today. Tomorrow, though, it may be that thoughts about how food choices impact the gut biome that regulates immune function may dominate my views. Our understanding of human biology and health changes continually over time. In case you haven’t been keeping up, the Golden Days of antioxidants is over.
Reading this over, I am tempted to temper my words as I can already visualize the list of examples in which antioxidants have proven helpful that our eminent Dr. Gaby may send to our editor. Perhaps my message is a bit too strong; I should admit antioxidants are helpful sometimes in some situations. Yet the bottom line, that antioxidants will not increase lifespan, remains. The biologic balance between oxidation and antioxidants is more intricate and nuanced than we thought, and we should be hesitant to make assumptions that one is always better.
Be that all as it may, my patient had apparently not received any updated memo that antioxidants were not going to stop him from aging, and it was left to me to break the bad news.
[1]. Viña J, Borras C, Abdelaziz KM, Garcia-Valles R, Gomez-Cabrera MC. The free radical theory of aging revisited: the cell signaling disruption theory of aging. Antioxid Redox Signal. 2013 Sep 10;19(8):779-87.
[2]. Bert P. La Pression Barométrique: Recherches de Physiologie Experimentale. Paris: Masson; 1878
[3]. Rubner M. Das Problem der Lebensdauer. Munich: Oldenburg; 1908
https://www.openagrar.de/receive/openagrar_mods_00000255
[4]. Pearl R. The Rate of Living. New York: Knopf; 1928
[5]. Ma, Siming, and Vadim N Gladyshev. “Molecular signatures of longevity: Insights from cross-species comparative studies.” Seminars in cell & developmental biology vol. 70 (2017): 190-203.
[6]. Hulbert AJ, Pamplona R, Buffenstein R, Buttemer WA. Life and death: metabolic rate, membrane composition, and life span of animals. Physiol Rev. 2007 Oct;87(4):1175-213.
[7]. Barja G. Updating the mitochondrial free radical theory of aging: an integrated view, key aspects, and confounding concepts. Antioxid Redox Signal. 2013 Oct 20;19(12):1420-45.
[8]. Lopez-Torres M, Perez-Campo R, Rojas C, Cadenas S, Barja G. Maximum life span in vertebrates: relationship with liver antioxidant enzymes, glutathione system, ascorbate, urate, sensitivity to peroxidation, true malondialdehyde, in vivo H2O2, and basal and maximum aerobic capacity. Mech Ageing Dev. 1993 Aug 15;70(3):177-99.
[9]. Perez-Campo R, López-Torres M, Cadenas S, Rojas C, Barja G. The rate of free radical production as a determinant of the rate of aging: evidence from the comparative approach. J Comp Physiol B. 1998 Apr;168(3):149-58.
[10]. Lawrence RA, Burk RF. Species, tissue and subcellular distribution of non Se-dependent glutathione peroxidase activity. J Nutr. 1978 Feb;108(2):211-5.
[11]. Pamplona R, Costantini D. Molecular and structural antioxidant defenses against oxidative stress in animals. Am J Physiol Regul Integr Comp Physiol. 2011 Oct;301(4):R843-63.
[12]. Pérez VI, Bokov A, Van Remmen H, et al. Is the oxidative stress theory of aging dead? Biochim Biophys Acta. 2009 Oct;1790(10):1005-14.
[13]. Omenn GS, Goodman G, Thornquist M, et al. The beta-carotene and retinol efficacy trial (CARET) for chemoprevention of lung cancer in high risk populations: smokers and asbestos-exposed workers. Cancer Res. 1994 Apr 1;54(7 Suppl):2038s-2043s. PMID: 8137335.
[14]. Goodman GE, Thornquist MD, Balmes J, et al. The Beta-Carotene and Retinol Efficacy Trial: incidence of lung cancer and cardiovascular disease mortality during 6-year follow-up after stopping beta-carotene and retinol supplements. J Natl Cancer Inst. 2004 Dec 1;96(23):1743-50.
[15]. Clark LC, Combs GF Jr, Turnbull BW, et al. Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. A randomized controlled trial. Nutritional Prevention of Cancer Study Group. JAMA. 1996 Dec 25;276(24):1957-63.
[16]. Yoshizawa K, Willett WC, Morris SJ, et al. Study of prediagnostic selenium level in toenails and the risk of advanced prostate cancer. J Natl Cancer Inst. 1998 Aug 19;90(16):1219-24. doi: 10.1093/jnci/90.16.1219. PMID: 9719083
[17]. Klein EA, Thompson IM Jr, Tangen CM, et al. Vitamin E and the risk of prostate cancer: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA. 2011 Oct 12;306(14):1549-56.
[18]. Moorthy D, Chung M, Lee j, PhD, et al.. Concordance Between the Findings of Epidemiological Studies and Randomized Trials in Nutrition: An Empirical Evaluation and Citation Analysis. Nutritional Research Series, Vol. 6. Technical Reviews, No. 17.6
[19]. Ibid Moorthy: Page 13 Table 2. Qualitative and quantitative concordance of effects in epidemiological studies and RCTs
Published August 26, 2023
About the Author
Jacob Schor, ND, now retired, had a general practice with a focus on naturopathic oncology in Denver, Colorado. He served as Abstract & Commentary Editor for the Natural Medicine Journal for several years (https://www.naturalmedicinejournal.com/) and posts blog articles on natural therapies, nutrition, and cancer (https://drjacobschor.wordpress.com/). He is a board member of CoAND and past president of OncANP, and someone who is happier outdoors than inside.