A prevailing view in medicine is that free radicalsi, including reactive oxygen species (ROS), are unwanted and potentially harmful by products from numerous sources including normal cellular metabolism (eg. mitochondrial respiration, NADPH oxidases, CYP p450 transformation), industrialization (eg. air pollution), ultraviolet radiation and smoking1. Strong evidence indicates that prolonged exposure to excess ROS is a common denominator in many important chronic diseases, including cancer, cardiovascular disease, neurodegenerative diseases, macular degeneration, non-alcoholic fatty liver disease as well as aging2. This evidence has fostered a common misconception that all levels of ROS are “bad”. It also has contributed to widespread, regular use of supplemental antioxidants for prevention and/or treatment of chronic disease. Current scientific evidence, however, supports ROS as important physiological signals as well as suggests there are times when supplemental antioxidants might not provide benefit and could even cause harm.
It is well documented that ROS, either in constant low levels or in transient bursts, are important signals in numerous physiological processes including inflammation, cellular growth and differentiation, activation of the immune system, stem cell differentiation and renewal, autophagy, mitochondrial biogenesis, vasodilatation and insulin sensitivity.1, 3, 4 Redox biologyii involving ROS also can support tissue repair and regeneration5; this discovery, as well as others, challenges the mitochondrial free radical theory of aging.6 ROS-associated mechanisms also help maintain a healthful microbiome, promote intestinal stem cell proliferation and support intestinal homeostasis.3 Diurnal regulation of glucocorticoid synthesis in the adrenal cortex also requires a ROS-dependent signal.7 Signaling by ROS is a common and important part of normal physiology.
Exposure to ROS and risk of pathology in many biological systems does not follow a linear relationship but rather a “U”-shaped curve, demonstrating different biological effects at low concentrations compared to higher concentrations. This type of phenomenon is named hormesis and is common to many biological systems, and with many different stressors.8 A number of in vitro as well as animal studies demonstrate that low levels of ROS do not cause harm but provide benefits, including promoting longevity, through a number of different mechanisms.9 For example, low levels of ROS upregulate mitochondrial enzymes such as super oxide dismutase and catalase, preconditioning them to act more efficiently during periods of eustress.iii, iv More efficient functioning of these enzymes augments their capability to respond appropriately to subsequent oxidative stressors; ultimately protecting cells and reducing the overall risk of pathology in associated tissues. The preconditioning effect of low levels of ROS on mitochondrial health is termed mitohormesis;10 it may “precondition the organism, including cellular defense mechanisms that ultimately serve as a long-term protective shield.”11
Different preclinical studies demonstrate that supplementation with antioxidants interferes with the beneficial adaptive effects of small, transient increases in ROS, resulting in increased tissue damage by subsequent oxidative stress.8 In a study using Caenorhabitis elegans, antioxidants such as Vitamin C, N-acetyl-cysteine and resveratrol “both lengthen and shorten lifespan, dependent on concentration, genotypes and conditions.”12 Moreover, increasing reports suggest that antioxidant supplements provide no benefit, or could even be detrimental, to the performance of healthy athletes.13 Exercise-induced ROS contribute to the health benefits of acute exercise by increasing antioxidant and cellular defense mechanisms, augmenting angiogenesis and triggering positive adaptations in skeletal muscle including mitochondrial biogenesis, increased insulin sensitivity and muscle fiber hypertrophy12; antioxidant supplements interfere with these beneficial ROS signals. The potentially negative effects of antioxidant supplementation depends on the intensity, duration and type of exercise as well as longer term training effects;12 suggesting that different types and doses of antioxidants might benefit some athletes, some of the time. Most interventional studies using antioxidants to prevent chronic diseases have failed to show benefit compared to placebo,8 also challenging the usefulness of recommending antioxidant supplements to all people.
In conclusion, it is important to recognize that ROS signaling only becomes potentially harmful, and “bad” if it’s too strong, lasts too long or arises in the wrong place at the wrong time. Recognizing potential benefits of ROS-associated signaling supports more cautionary usage of antioxidant supplements. Current evidence suggests that the most optimal way to recommend antioxidants is through personalized methods, individualizing the type and dose according to the patient needs, which are ideally identified using laboratory or point-of-care tests for oxidative stress.14 Measuring antioxidant levels in the body is currently an inexact science, with many variables and an as yet unestablished consensus baseline level;15 although it still provides value in identifying those patients where antioxidant supplementation would provide the most likely benefit. Numerous misconceptions about ROS and antioxidants still exist.16 It is important for health care practitioners to be aware of these misconceptions as well as the current science regarding the double-edged sword of ROS, this information will be helpful in making the most beneficial recommendations to their patients.
i Free radicals are defined as small diffusible molecules that are highly reactive because of an unpaired electron. The two more common classes of free radicals are reactive oxygen species (ROS) (i.e. super-oxide, hydrogen peroxide, singlet oxygen, hypo-halous acids and organic peroxides) and reactive nitrogen species (RNS) (i.e. nitric oxide, nitrogen dioxide). ROS and RNS interact and have similar targets. However, for the purpose of clarity, this text will only refer to ROS.
ii Some authors suggest using the term using the term “redox biology” when referring to low levels of ROS that activate signaling pathways to initiate biological processes and “oxidative stress” when referring to high levels of ROS that incur damage to DNA, protein or lipids. Scheiber, M., et al. Curr Biol. 2014;24(10):R453-R462
iii Eustress was initially defined by Hans Seyle as beneficial stress, either psychological, physical, or biochemical/ radiological. This definition was introduced to support the concept that not all levels and types of stress are harmful; biological outcomes from different stressors range from adaptive (usually beneficial) to maladaptive (usually harmful). Stress initiating a beneficial outcome was termed “eustress”, whereas stress provoking a harmful response was termed “distress”. Szabo, S., et al. “Letter” to the Editor of Nature. Stress. 2012:15(5):472-478
iv Some authors suggest using the term “oxidative eustress” to denote levels of ROS that promote physiological signaling (i.e. 1 to 10 nM) and “oxidative distress” to denote levels of ROS that lead to pathological signaling (i.e. >100 nM). Sies, H., Oxidative eustress. Redox Biol. 2017;11:613-619
- Nathan, C., and Cunningham-Bussel A. Nat Rev Immunol. 2013;13(5):349-361.
- Matschke, V. et al. Neural Regen Res. 2019;14(2):238-241
- Finkel, T. J Cell Biol. 2011; 194(1):7-15
- Holmstrom, K.M., and Finkel, T. Nat Rev Mol Cell Biol. 2014;15(6):411-421
- Serras, F. Fly (Austin). 2016;10(3):128-133
- Hekimi, S., et al. Trends Cell Biol. 2011;21(10):569-576
- Kil, I.S., et al. Mol. Cell 2012; 46, 584-594.
- Calabrese, E.J. Microb Cell. 2014;1(5):145-149
- Ristow, M., et al. Dose-Response. 2014;13:288-341
- Yun, J., and Finkel T. Cell Metabolism. 2014:19:757-766
- Kawagishi, H, and Finkel, T. Nat Med. 2014;20(7):711-713
- Desjardins, D., et al. Aging Cell. 2017;16(1):104-112
- Merry, T.L., and Ristow, M. J Physiol. 2016;594:5135-5147.
- Margaritelis, N.V., et al. Adv Nutr. 2018;9(6):813-823
- Gutteridge, J.M.C., and Halliwell, B. Biochem Biophys Res Comm. 2010;393:561-564
- Bast A, and Haenen, G.R.M.M. Trends Pharm Sci. 2013; 34(8):430-436