1. Introduction
Humans and other aerobic organisms constantly produce free radicals as part of normal metabolic processes. Free radicals are defined as molecules or molecular fragments with one or multiple unpaired electrons in the atomic or molecular orbital. Free radicals derived from oxygen are called reactive oxygen species (ROS) [
1]. However, ROS includes not only free radicals but also non-radicals, and are closely related to other free radical families, such as reactive nitrogen species [
1]. Superoxides (O
2−) and nitrogen monoxide (NO) are the primary free radicals that trigger a chain reaction of hydrogen peroxide (H
2O
2), hydroxyl radicals (OH
−), peroxynitrite (ONOO
−), and hypochlorous acid (HOCl) [
1]. Although these free radicals have positive effects in immune reactions and cellular signaling, they are also known to have negative effects, such as oxidative damage of lipids, proteins, and nucleic acids [
1]. As has been reported in previous studies, free radicals are constantly produced in various tissues (liver, kidneys, heart, skeletal muscle, etc.) at rest, inducing a certain level of oxidative damage in these tissues [
2,
3,
4,
5].
Organisms are equipped with antioxidant defense systems that protect cells from the toxic effects of free radicals. Antioxidant defense systems are divided into enzymatic antioxidants, such as superoxide dismutase (SOD), catalase, and glutathione peroxidase, and non-enzymatic antioxidants, such as vitamin C, vitamin E, glutathione, and bilirubin. These antioxidants play important roles in delaying or preventing oxidation of intracellular and extracellular biomolecules [
6,
7]. While the capacity of antioxidant systems is affected by intake of nutrients, such as vitamins and minerals, it is important to note that many mammals, with the exception of humans, can synthesize vitamin C in vivo [
8]. This knowledge may be helpful in correctly evaluating the influence of supplementary intake and restriction of vitamin C on vitamin C concentration in tissues at rest and during exercise.
As mentioned above, the oxidation-reduction (redox) balance is maintained in vivo by a complex regulatory mechanism; however, various physiological stimuli (i.e., radiation, alcohol use, smoking, and exercise) disturb this balance toward oxidation, thus inducing oxidative stress (
Figure 1). Oxidative stress was initially defined as “a disturbance in the prooxidant-antioxidant balance in favor of prooxidants” [
9], but was later redefined as “disturbance of the oxidation-reduction balance in favor of oxidants, leading to a disturbance in redox signaling and control and/or molecular damage” [
10]. Previous studies have suggested that chronic oxidative stress is deeply involved in the onset or progression of various diseases such as diabetes, cancer, cardiovascular diseases, and neurological disorders, and may also play a role in the mechanisms of aging [
11,
12]. Thus, maintaining redox balance is pivotal for the survival and health of organisms.
Powers and Jackson, in their own review in 2008, classified biomarkers that reflect in vivo oxidative stress into four broad categories [
7]. The first method is to detect oxidants such as free radicals. Although most ROS and free radicals are highly reactive and have an extremely short half-life, which makes it difficult to measure them directly, they can be measured after first using exogenous molecules such as fluorescent probes or spin traps (i.e., 5,5-dimethyl-1-pyrroline-
N-oxide and
N-
tert-butyl-α-phenylnitrone) to produce luminescence or to stabilize oxidants. However, as mentioned above, oxidative stress is determined by the redox balance, meaning that measurement of oxidants alone is not sufficient to accurately assess oxidative stress [
7].
The second method is to measure antioxidant levels in tissues. Antioxidant levels (concentration or activity) vary according to tissue type, but can increase or decrease depending on the degree of exposure to oxidative stress. Antioxidant levels are useful markers of oxidative stress; however, nutritional status may affect the measurement results.
The third method is to measure oxidation products. Oxidation products include protein carbonyl (PC), which is a marker of protein oxidation; F
2-isoprostanes and malondialdehyde (MDA), which are markers of lipid peroxidation; and 8-oxo-2′-deoxyguanosine (8-OHdG), which is a marker of deoxyribonucleic acid (DNA) oxidation. Measurement of these oxidation products is considered to be the most important aspect of oxidative stress assessment [
7]. However, as these products exist only in trace amounts even during oxidative stress, they are often difficult to measure.
The fourth method is to measure the redox balance. The most frequently used redox marker is the reduced glutathione/oxidized glutathione (GSH/GSSG) ratio. Redox markers can be used to assess the redox balance with respect to both reduction and oxidation, thus making them extremely useful markers of oxidative stress. On the other hand, as with other methods, results can be affected by the techniques used to sample tissues or process tissue specimens [
7].
As described above, as each oxidative stress marker has both advantages and disadvantages, there is currently no optimal marker for evaluating oxidative stress. Therefore, it was widely recommended to assess oxidative stress by measuring multiple oxidative stress markers [
1,
13].
During exercise, oxygen demand increases, particularly in skeletal muscle, causing a dramatic change in the blood flow to various organs. Furthermore, exercise-induced muscle damage promotes infiltration of phagocytes (i.e., neutrophils and macrophages) at the site of injury. These physiological changes that occur during acute exercise increase free radical production, leading to oxidative damage to biomolecules. Exercise-induced oxidative stress associated with increased free radical production has been studied for 40 years, since it was first reported by Dillard et al., in 1978 [
14]. Recent developments in biochemical and molecular biological techniques have enabled observation of events at the cellular level, and have increasingly demonstrated that free radicals play at least some role in physiological adaptations after exercise training [
15,
16,
17]. Therefore, free radicals generated by exercise are considered to have both positive and negative physiological effects.
The present review focuses on oxidative stress induced by acute exercise and outlines the effects of aerobic and anaerobic exercise on oxidative stress in organisms. This review also summarizes findings on the determinants of exercise-induced oxidative stress and sources of free radical production, about which much is still unknown. Finally, the present review briefly summarizes the effects of antioxidant supplementation on exercise-induced oxidative stress, which have been studied extensively. In principle, this review covers findings for the whole body, and describes human trials and animal experiments separately.
5. Conclusions and Future Trends
The present review summarized the results of studies on the effects of acute exercise on oxidative stress in organisms, and the effects of antioxidant intake on exercise-induced oxidative stress obtained in human trials and animal experiments. The present review also summarized the determinants of exercise-induced oxidative stress, sources of free radical production during exercise, and time-course change in oxidative stress responses according to exercise type.
There is already a consensus that exercise increases the production of free radicals. However, free radicals are highly reactive and have extremely short half-lives, which makes them difficult to measure directly [
1]. For this reason, previous studies have used numerous oxidative stress markers, including oxidation products, as indicators that reflect increases in free radicals. These markers of oxidative stress all have advantages and disadvantages, which is why oxidative stress should be assessed by measuring multiple markers [
1,
13]. Ultimately, this makes exercise-induced oxidative stress more difficult to understand. Furthermore, it is currently unknown whether changes in oxidative stress markers associated with exercise actually represent major deviations from optimal ranges, from a physiological point of view. Therefore, a future study identifying biomarkers with higher sensitivity and validity is warranted. In addition, it is necessary to examine the physiological significance of exercise-induced oxidative stress with established standard values of oxidative stress markers.
Moreover, for many years, exercise-induced oxidative stress has been interpreted with a focus on skeletal muscle based on the assumption that the production of ROS and free radicals during exercise occurs primarily in skeletal muscle. For example, several sites of ROS and free radical production in skeletal muscle have been identified with inhibitors of specific enzymes (e.g., apocynin and allopurinol) and in in vitro experiments mimicking exercise [
7,
22]. However, in recent years, there has been an increasing understanding that blood itself, in addition to organs including skeletal muscle, is also a source of free radical production [
140]. Accordingly, it has been reported that free radicals may be produced in erythrocytes and leukocytes [
141]. Therefore, the site of free radical production during exercise must be investigated through an integrative approach. In addition, sites of free radical production during exercise are speculated to be affected by experimental conditions such as exercise type (i.e., muscle contraction type), intensity, and duration. Therefore, findings on the determinants of exercise-induced oxidative stress should be accumulated step by step for each exercise condition, starting with basic research.
As a final note, it must be mentioned that exercise-induced oxidative stress has wide inter- individual variability. In a previous study, exercise-induced oxidative stress was unexpected or negligible in one in three individuals who performed high-intensity eccentric exercise (i.e., eccentric knee extension exercise) [
51]. Such phenomena are also likely to be observed in treadmill or cycle ergometer exercises on a flat ground. Therefore, the large inter-individual variability in oxidative stress responses to acute exercise may provide important clues about why the study results are inconsistent. As mentioned earlier, the effects of antioxidant intake on exercise-induced oxidative stress are also reportedly affected by individual antioxidant levels at rest [
124,
142]. These findings indicate that when examining exercise-induced oxidative stress and how it is affected by antioxidant intake, individual redox status should be screened in advance. Moreover, further studies about the causes of wide inter-individual variability in redox status at rest are warranted.