The effect of physical activity on mediators of inflammation

Skeletal Muscle

Within recent years it has been recognized that skeletal muscle is a highly metabolically active organ, producing and secreting an array of molecules in response to contraction 5. The products of this tissue have been termed ‘myokines’ and have been shown to exert many diverse effects, some of which occur locally within the skeletal muscle itself whilst others act distally within other organs and tissues including the liver, pancreas, adipose tissue and cardiovascular system 6. Although at present our understanding of skeletal muscle as an endocrine organ is far from complete, available evidence suggests that contraction‐induced myokine production may mediate many beneficial physiological effects of regular exercise which favourably impact upon metabolic health 7. Currently, several myokines have been characterized; however, undoubtedly the most well studied is the cytokine interleukin‐6 (IL‐6). This interest stems from the fact that IL‐6 displays the most marked response to acute exercise stimuli while also being thought to orchestrate the anti‐inflammatory effect of acute exercise 8.

Resting concentrations of IL‐6 in young, healthy individuals are typically very low with two‐ to threefold higher levels being apparent in older adults or those with metabolic disease such as T2D and CVD. At rest, the secretion of IL‐6 from skeletal muscle is minimal, with the majority being produced from leucocytes and adipose tissue 9. During acute exercise, muscle contraction stimulates the release of IL‐6 from the skeletal muscle, with circulating levels increasing over 100‐fold with intense and prolonged exercise, for example, marathon and ultra‐endurance events 10. Typically, however, more moderate responses are observed which are proportional to the exercise intensity and duration, individual fitness level (inversely) and muscle mass utilized by the specific exercise modality 9. Thus, when exercise duration and external workload are matched, HIIT induces a greater IL‐6 response than moderate‐intensity continuous exercise (figure 2, upper panel) 11.

For many years it had been postulated that a so‐called exercise factor was responsible for mediating several metabolic responses that occur during exercise. However, until recently the candidate remained elusive. Skeletal muscle‐derived IL‐6 has now been identified as a key metabolic intermediary and research has defined IL‐6 as an energy sensor, functioning to preserve fuel availability during exertion 12. Specifically, during exercise IL‐6 serves to augment hepatic glucose and adipose tissue fatty acid release to provide sufficient fuel to meet the extra metabolic demand. Accordingly, exercise performed with depleted muscle glycogen or with exogenous glucose ingestion has been shown to augment and suppress the exercise‐induced IL‐6 response, respectively 12. These effects of IL‐6 on the liver and adipose tissue are mediated through the activation of AMP‐activated protein kinase (AMPK).

During exercise the release of IL‐6 can be stimulated by increased intracellular calcium content, a result of muscle contraction, with subsequent downstream activation of transcription factors that induce an increase in IL‐6 production 6. This makes human skeletal muscle unique in being able to produce IL‐6 independent of the pro‐inflammatory cytokine tumour necrosis factor‐α (TNF‐α) 13. Furthermore, in contrast to adipose‐derived IL‐6, which is commonly thought of as pro‐inflammatory, episodic elevations in skeletal muscle‐derived IL‐6 trigger an anti‐inflammatory cascade by inhibiting the release of pro‐inflammatory cytokines (TNF‐α and IL‐1β]) via the stimulation of their antagonistic receptors 8. The capacity of IL‐6 to blunt TNF‐α activity has been demonstrated in several murine models and in humans following exogenous IL‐6 administration, showing a key regulatory role of IL‐6 over TNF‐α 14. In addition to suppressing pro‐inflammatory factors, exercise‐related IL‐6 also triggers the release of IL‐10 8, a potent anti‐inflammatory molecule, which itself directly inhibits the synthesis of several pro‐inflammatory mediators, particularly of the monocytic lineage such as IL‐1α, IL‐1β, TNF‐α, the chemokines IL‐8 and macrophage inflammatory protein‐1α (MIP‐1α). These inflammatory mediators play a key role in propagating inflammatory responses and recruiting immune cells to the site of inflammation. Prolonged exercise can also lead to the release of C‐reactive protein (CRP) on the following day 8. This may confer an additional anti‐inflammatory effect by promoting the induction of anti‐inflammatory cytokines from circulating monocytes and by suppressing the synthesis of pro‐inflammatory cytokines in tissue macrophages. Additionally, exercise may also induce an acute anti‐inflammatory effect by stimulating the production of cortisol and adrenaline, two hormones which exert potent anti‐inflammatory effects. Muscle‐derived IL‐6 may partly be responsible for the cortisol response to exercise 9.

An early indication of the fact that IL‐6 alone could not be the sole mediator of these anti‐inflammatory effects was identified by Febbraio et al. who showed that recombinant IL‐6 infusion only elevated circulating glucose when skeletal muscle was active. This evidence suggested that the exercising muscle provided a co‐factor for IL‐6 to function fully 15. Specifically, research has shown that both IL‐6 and the IL‐6 receptor (IL‐6R) are required to be present in order to initiate glucose uptake at physiological levels 16. At rest, expression of this membrane‐bound receptor in skeletal muscle is limited, with significantly greater expression found in hepatocytes and leucocytes 17. IL‐6 can, however, bind with the circulating soluble receptor (sIL‐6R) which increases after endurance exercise and remains elevated for prolonged periods after the exercise has ceased 11 (figure 2, lower panel). Through a process of trans‐signalling involving the ubiquitously expressed glycoprotein‐130 (gp130) IL‐6 is able to trigger similar cellular pathways to the membrane‐bound form (which also increases in skeletal muscle after exercise) in tissues that express little or no membrane‐bound IL‐6R (figure 3). Of note, this signalling mechanism is subject to regulatory control via a soluble (circulating) form of GP130 (sGP130).

The sIL‐6R receptor appears in at least two identifiable isoforms—proteolytically cleaved and differentially spliced (mRNA). The predominant isoform in plasma of healthy individuals is derived from proteolytic cleavage (>99%) and although both forms increase after exercise, the cleaved form remains dominant 11. The formation of the binary complex (IL‐6/sIL6R) substantially extends the plasma half‐life of IL‐6, and mathematical modelling suggests that 70% of sIL‐6R is in this form at rest and that the complex increases twofold after endurance exercise 18. When thinking about the inflammatory capacity of IL‐6 it is therefore essential to also consider the effect of the soluble receptor.

Other notable skeletal muscle‐derived factors which may exert indirect influences (by affecting adiposity) on systemic inflammation are irisin and IL‐15 5. Irisin is a recently identified myokine which is released into the circulation at elevated levels after exercise and potentially has significance for inflammatory meditators because of its capability to trigger the change of white adipocytes to ‘brite’ cells (brown‐in‐white). The significance of this is that ‘brite’ cells display a phenotype similar to that of brown adipocytes which are metabolically inefficient cells that augment thermogenesis and energy expenditure. This contrasts with white adipocytes whose primary function is for the deposition and storage of lipid. Thus, irisin may be an intermediary through which skeletal muscle communicates with adipose tissue to regulate thermogenic programming. Theoretically, this has favourable implications for energy homeostasis and metabolic regulation.

IL‐15 is an additional factor produced in skeletal muscle in response to contraction which also appears to exert favourable effects on adiposity. In particular, IL‐15 seems to exert a beneficial influence on abdominal fat with transgenic overexpression in murine models leading to a reduction in central adiposity 19. In humans, circulating IL‐15 has been found to be inversely associated with trunk adipose tissue mass 20. Additional research is necessary to determine the mechanism by which IL‐15 affects central adiposity.

Leucocytes

The effects of acute exercise on leucocyte trafficking and function have been investigated extensively. It is widely accepted that acute exercise increases the number of circulating leucocytes yet causes a temporary depression on several aspects of immune cell function. These effects are related to both the duration and intensity of exertion, with strenuous exercise (prolonged and/or high‐intensity) yielding the most profound effects 21.

Leucocytes secrete a variety of cytokines (including IL‐6), which are involved in the activation and regulation of the immune response. As monocytes are one of the main sources of IL‐6 (as well as other cytokines) in vivo, several studies have assessed whether monocytes are the source of the exercise‐induced increase in systemic cytokine concentrations. However, research examining cytokine mRNA expression (IL‐6, IL‐1α, IL‐1β and TNF‐α) in peripheral blood mononuclear cells (monocytes and lymphocytes) in response to acute moderate‐vigorous cycling indicated that this was unlikely to be the case 22. These findings were later confirmed by studies investigating spontaneous (i.e. unstimulated) intracellular cytokine production from monocytes in response to acute exercise 23, 24. Other research has also documented no change in the spontaneous production of a variety of cytokines (IL‐1α, IL‐2, IL‐4, IL‐6, IL‐10, IFN‐γ and TNF‐α) from several leucocyte subsets (granulocytes, lymphocyte subsets and monocytes) after acute high‐intensity cycling 25. Thus, although exercise might mobilize distinct populations of leucocytes that are already expressing cytokines from marginal pools, unlike skeletal muscle, exercise does not directly stimulate production of cytokines from leucocytes.

On the other hand, the influence of acute exercise on stimulated cytokine production from leucocytes has also received considerable attention. Studies have mainly focused on monocytes and T lymphocytes because of the important role that cytokine production from these cells plays in the orchestration of the immune response. Several studies have shown that stimulated cytokine production by monocytes (IL‐6, TNF‐α and IL‐1α) and type 1 T lymphocytes (IFN‐γ), but not type 2 T lymphocytes (IL‐4), is inhibited following strenuous exercise 23, 24, 26, suggesting a weakening of cell‐mediated immune responses. In agreement with this, is the emerging evidence that acute exercise downregulates monocyte toll‐like receptor (TLR) expression 27. TLRs are triggered by microbial pathogens (e.g. lipopolysaccharide) as well as endogenous danger signals (e.g. heat shock proteins), leading to activation of the antigen‐presenting cell (APC), such as the monocyte. This subsequently leads to activation of the T lymphocyte through the interactions between major histocompatibility complex (MHC) class II molecules and co‐stimulatory cell surface molecules (CD80/86) on the APC with the T‐cell receptor and CD28 on the T‐cell membrane, respectively. Activation of this signalling cascade results in the secretion of inflammatory cytokines. Notably, the upregulation of CD80, CD86, MHC class II and IL‐6 by CD14⁺ monocytes following activation with specific TLR ligands is also reduced following strenuous exercise 28. Thus, although exercise may impair immune function and possibly increase the susceptibility to infection, this appears to be a small price to pay for a reduction in the inflammatory capacity of the leucocytes.

Mobilization of regulatory T (T_reg) cells (which are a major source of IL‐10) may also be involved in the anti‐inflammatory effects of exercise. Studies addressing the effects of acute exercise on T_reg cells are scarce, however, with the available evidence documenting contradictory findings 29, 30. Interestingly, a recent study in a running mice model showed that high but not moderate‐intensity exercise training resulted in increased circulating T_reg cell numbers 31. This data may suggest that high‐intensity exercise provides a greater anti‐inflammatory stimulus.

Adipose Tissue

Mounting evidence shows an integral role of adipose tissue in the development of cardio‐metabolic disease. Chronic low‐grade inflammation is now understood to be the key link between adipose tissue, metabolic dysfunction and disease outcomes with adverse changes in the adipose tissue proteome underpinning these events 32. Notably, adipose tissue has the capacity to produce at least 75 inflammatory proteins, and with increasing obesity, the infiltration of macrophages and a change in the macrophage phenotype (alternatively to classically activated) prompt adverse changes in the adipose tissue metabolic profile. These changes lead to a pro‐inflammatory and metabolically unhealthy environment 33 (figure 4). If in abundance this may spill over into the circulation, resulting in systemic inflammation.

A key trigger of inflammatory events in adipose tissue in obesity is thought to be adipocyte hypertrophy with cell volume expansion leading to tissue hypoxia, cellular stress, necrosis and ultimately tissue macrophage invasion 32. Large adipocytes are insulin resistant and hyperlipolytic whilst small adipocytes are insulin sensitive and metabolically healthy. The location of adipose tissue also has a key bearing on its metabolic function with growing evidence showing that the dominant source of the inflammatory proteins is the visceral site which accounts for the strong link between central obesity and cardio‐metabolic disease. A large body of evidence has described enhanced gene expression and protein content of some pro‐inflammatory molecules in visceral adipose sites compared with subcutaneous depots 34. The adverse consequence of this may be related to the anatomical location of visceral adipose tissue which directly exposes the liver to a continuous abundance of fatty acids and pro‐inflammatory adipokines. In healthy weight humans visceral fat accounts for <20% of total fat in males and significantly less in females 35, therefore its overall contribution to the chronic inflammatory state must be aggregated. Usefully, exercise is a potent stimulus for visceral adipose tissue lipolysis and reductions in central adiposity have been shown to occur with exercise training independent of overall weight change 36. This indirect influence may represent one important mechanism by which exercise favourably impacts systemic inflammation.

Information concerning the direct effects of acute exercise on adipokine secretions is sparse; however, some information is available regarding the effects on IL‐6, leptin, adiponectin, resistin and TNF‐α 37. In contrast to skeletal muscle‐derived IL‐6 the literature is consistent in stating that the release of IL‐6 from adipose tissue is suppressed during low‐intensity prolonged exercise, although levels may increase after exercise 38. It is possible that this adipose‐derived IL‐6 may mediate lipolysis and fat metabolism in the recovery period after exertion. A recent study has highlighted the centrality of IL‐6 in mediating an anti‐inflammatory influence within subcutaneous adipose tissue in response to acute exercise 39. Specifically, the capacity of exercise to suppress leptin and TNF‐α mRNA within subcutaneous adipose tissue (as seen in wild‐type rodents) was absent in a rodent model with global IL‐6 knockout.

Resistin and TNF‐α are proteins produced within the non‐adipocyte fraction within adipose tissue that are known to exert potent pro‐inflammatory actions. Resistin was identified as an adipokine directly linked to the induction of insulin resistance in several animal models 40 with research describing negative effects on glycaemic control and cardiovascular function via adverse effects within the liver, adipose tissue and vasculature. Similarly, the primary characteristic of TNF‐α is its negative impact on insulin action and glucose metabolism through impairing insulin signalling within skeletal muscle and adipose tissue 32. A handful of studies have assessed the acute impact of exercise on circulating levels of resistin and TNF‐α. The consensus of evidence suggests that acute exercise has no impact on circulating TNF‐α 37. Similarly, despite notable improvements in insulin sensitivity occurring in response to single bouts of continuous‐moderate‐intensity cycling, circulating levels of resistin remain unchanged 41. These data suggest that short‐term improvements in insulin action after single episodes of exercise are not mediated by TNF‐α or resistin.

Leptin and adiponectin are proteins produced primarily by adipocytes with circulating levels, correlating positively and negatively with adipose tissue mass, respectively. Leptin was identified as a key regulator of energy balance but is now also recognized as a key mediator of immune function. Leptin has potent pro‐inflammatory actions influencing both the innate and adaptive immune systems, and heightened levels seen in obesity contribute to systemic inflammation 42. The impact of exercise on circulating leptin has received significant attention with the consensus of evidence, suggesting that circulating levels remain stable during acute exercise. An exception to this may occur when the exercise volume and associated energy expenditure are substantial (>3348 kJ) whereby reduced circulating levels may be seen 43. This change may represent a compensatory response to the energy deficit but may have consequences for the inflammatory response.

Adiponectin is one of the most highly abundant plasma proteins 44. Plasma adiponectin is positively related to insulin sensitivity and glucose tolerance which is mediated via favourable effects on skeletal muscle glucose uptake, fatty acid oxidation 45 and hepatic gluconeogenesis 46. Circulating levels of adiponectin have been measured in response to acute exercise to determine whether changes in adiponectin are related to transient exercise‐induced improvements in insulin sensitivity. Data suggest that changes in adiponectin are not related to improved insulin action in response to acute exercise as neither moderate or high intensity continuous aerobic exercise affect circulating levels of adiponectin 47, 48.

Sedentary Behaviour

Recently, interest has ignited around sedentary behaviour, a new physical activity paradigm which stresses the negative health implications of too much time spent sitting or lying 52. The major stimulus for this research derives from animal studies which have shown that experimental muscle unloading negatively impacts lipoprotein lipase activity and lipid metabolism 53. The important distinction in this field is that the adverse health outcomes of sedentary behaviour are independent of moderate–vigorous physical activity 54. Emerging research has indicated that sedentary time is positively associated with elevated inflammatory proteins 55, 56 with one report suggesting that the association may be stronger in women than in men 57. At present, no interventions have been implemented to assess the impact of reducing sedentary time on systemic inflammation. Our research group is currently conducting a randomized behavioural intervention (project STAND—Sedentary Time ANd Diabetes) examining the health impact of reducing sedentary time in young adults at risk of T2D 58. This study will provide detailed information regarding the effects of reducing sedentary time on candidate inflammatory markers in an ‘at risk’ population.

Low to Moderate Intensity Continuous Exercise

Walking has been suggested to be the most likely form of exercise to increase physical activity as it is accessible to all, poses little risk of injury, and can be accumulated into daily routine 59. A study conducted by our own group investigated the effectiveness of a community‐based walking intervention on systemic inflammatory proteins in overweight middle‐aged men and women in Scotland 60. In this study, there was no attempt to change any other lifestyle factors and the control group was given exactly the same support but was delayed in starting their walking intervention. The study concluded that although all participants achieved the physical activity targets (an increase of 3000 steps/day, at least 5 days/week), which were the physical activity guidelines at the time, the intervention brought no improvements in CRP, IL‐6, sIL‐6R, TNF‐α or soluble TNF receptors I and II (sTNFR1 and sTNFRII). This was the first study to undertake an exercise intervention in a community setting using the public health guidance.

This finding of no change to inflammatory markers after a low‐intensity exercise intervention was reproduced by Polak et al. 61 who observed no change in circulating inflammatory proteins (IL‐6, TNF‐α and adiponectin) following a period of low‐intensity cycling. Additionally, after a 12‐week moderate‐intensity aerobic exercise intervention Christiansen et al. 62 did not see any changes in circulating inflammatory proteins (IL‐6, IL‐15, IL‐18, MCP‐1, MIP1α and adiponectin). This was despite witnessing favourable changes in the adipose tissue gene expression of some inflammatory markers.

Thompson et al. 63 examined the impact of a higher intensity exercise intervention in a group of middle‐aged sedentary men over the course of 6 months. In this study, aerobic exercise at 50–70% of maximum aerobic capacity (30–60 min/day) induced a reduction in circulating IL‐6 but had no influence on CRP or soluble intercellular adhesion molecule 1 (ICAM‐1)–a vascular adhesion molecule used as a biomarker for endothelial dysfunction. Uniquely this study also followed participants through a 2‐week detraining period and documented a reversal of circulating IL‐6 to pretraining levels, pointing to a direct influence of training.

Resistance/Combined Aerobic and Resistance Training

Although the majority of studies have investigated aerobic exercise training, resistance exercise interventions have been shown to attenuate systemic inflammation by reducing circulating levels of CRP and increasing levels of adiponectin 64. Additional research suggests that a combination of aerobic and resistance exercise training may confer the greatest favourable impact on systemic inflammation 65, 66. Specifically, in a sample of patients with T2D or the metabolic syndrome, Balducci et al. compared the impact of aerobic exercise (low‐ and high‐intensity conditions) and combined aerobic and resistance exercise training on markers of inflammation and observed beneficial changes in pro‐inflammatory molecules with high‐intensity aerobic, and combination training (leptin, IL‐6 and resistin), independent of weight change 66. Notably, additional improvements (IL‐1b, TNF‐α and IFNγ) including changes in anti‐inflammatory factors (IL‐4 and IL‐10) were only witnessed in the combination training group. These data may suggest that the greatest benefits of exercise are realized when both forms of training are undertaken. Such inferences are congruent with other data which have demonstrated the utility of integrated lifestyle interventions (diet and exercise) that have included both aerobic and resistance training elements 67.

High‐intensity Intermittent Exercise

Striking health improvements, most notably enhanced insulin sensitivity, have been found with a novel form of exercise which involves intermittent bouts of high‐intensity exercise 68. This form of exercise training has been termed HIIT and several studies have shown that this form of training can yield rapid health improvements over the course of just 2 weeks (three training sessions per week) 68. Utilizing this form of exercise training within a sample of inactive overweight/obese men, our own group has been able to detect significant changes in the circulating and adipose tissue IL‐6 system, with decreased sIL6R observed after training in the former, and reduced IL‐6 but increased IL‐6R in the latter (figure 5) 69. Moreover, within the circulation, levels of MCP‐1 and adiponectin were also reduced following HIIT whilst the concentrations of TNF‐α, IL‐6, ICAM‐1 and IL‐10 remained unchanged. Within adipose tissue, concentrations of ICAM‐1 remained unchanged whilst TNF‐α, IL‐10 and MCP‐1 were undetectable. These data suggest that 2 weeks of HIIT is sufficient to induce some beneficial alterations in the resting inflammatory profile of an overweight and obese male cohort.

The necessity for exercise to be of high intensity to reduce the risk of chronic cardiovascular and metabolic diseases via its anti‐inflammatory effects is supported by a running mouse model, where the responses of circulating T_reg cells to moderate and HIIT were examined 31. Only the high‐intensity training resulted in increases in T_reg cell numbers with reduced pro‐inflammatory and increased anti‐inflammatory cytokine expression. Thus, if the total energy expenditure of exercise is held constant, exercise performed at a vigorous intensity appears to convey greater metabolic health benefits than exercise of a moderate intensity.

Overtraining

In contrast to the previous descriptions of exercise, an excessive volume of activity without sufficient opportunity for rest can induce deleterious health and performance effects, a phenomenon which has been termed the overtraining syndrome 70. This feature is typically witnessed only in athletes undergoing an intensified period of training without sufficient rest and is characterized by weight loss, malaise, mood/sleep disturbances, chronic fatigue and impaired athletic performance. The aetiology of this syndrome is not fully understood but is thought to be at least partly mediated by low‐grade systemic inflammation 71. Notable features of this are augmented circulating levels of pro‐inflammatory cytokines (TNF‐α, IL‐6 and IL‐1b) which act on the brain to induce sickness‐like behaviours including anorexia, depression and sleep disturbance 71.

The role of IL‐6 as a mediator of the overtraining syndrome has received greater attention than other cytokines. In various clinical settings IL‐6 has been found to relate to fatigue symptoms. Robson‐Ansley et al. have also described heightened circulating IL‐6 concentrations in a group of male triathletes subjected to an intensified period of exercise training which was associated with sensations of fatigue and malaise 72. Not all studies have reproduced this finding, however, and given the relatively sparse amount of data available on this issue the contribution of IL‐6 to the overtraining syndrome remains uncertain. Recently it has been suggested that the sIL‐6R may link IL‐6 with the overtraining syndrome through the potentiation of the brain's responsiveness to IL‐6 73 and circulating sIL6R levels have been found to increase significantly during a 6‐day mountain bike endurance event (468 km) with levels correlating positively with daily fatigue scores 74. Additional research is needed to fully characterize the role of cytokines and the IL‐6 system as orchestrators of the overtraining syndrome.