1. To address the question of whether endotoxaemia could be involved in the inflammatory response induced by long-term strenuous exercise, 18 male marathon runners [mean age 41 +/- 2 (SEM) years] were studied. Their performance in the marathon ranged from 2 h 46 min to 4 h 42 min. 2. Four venous blood samples were drawn: at rest, just before the race (baseline); within 15 min following the completion of the marathon; after 1 h of recovery; and the morning after the race. 3. The following humoral markers of the inflammatory response to exercise were measured: polymorphonuclear myeloperoxidase (MPO), anaphylatoxin C5a (C5a), tumour necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6). Plasma endotoxin was measured by a sensitive and rapid chromogenic Limulus assay. All inflammatory markers were significantly increased (P < 0.001) after the race, reaching in most cases peak values in the first blood sample drawn following the completion of the marathon [MPO, 298 +/- 19 ng/ml (SEM); C5a, 1.45 +/- 0.32 ng/ml; TNF-alpha, 20 +/- 3 pg/ml; IL-6, 88 +/- 13 pg/ml] when compared with baseline [MPO, 146 +/- 16 ng/ml (SEM); C5a, 0.27 +/- 0.2 ng/ml; TNF-alpha, 12 +/- 1.5 pg/ml: IL-6, 1.0 +/- 0.5 pg/ml]. Traces of plasma endotoxin (ranging from 5 to 13 pg/ml, with one exceptionally high value of 72 pg/ml measured in one runner) were detected in seven subjects within the first hour of recovery. An ELISA method was used to determine the endogenous IgG antibodies toward a range of Gram-negative bacterial lipopolysaccharides (LPSs) of different sizes and structures. A transient decrease in certain anti-LPS activities, mainly against rough LPS, occurred in most cases in the first blood sample drawn after the race. There was no correlation between the magnitude of the inflammatory response to exercise, as assessed by the increase in blood levels of humoral markers of inflammation, and the changes in circulating endotoxin levels of anti-LPS IgG activity following the race. 4. From these results, we conclude that the mild, transient endotoxaemia detected in some of our subjects does not play a major role in the observed inflammatory response to a marathon competition.

To address the question of whether the increased plasma concentration of interleukin 6 (IL‐6) following strenuous muscular work could be related to exercise‐induced muscle damage, 5 moderately active male volunteers underwent two isokinetic exercise sessions in the eccentric mode, separated by a period of 3 weeks during which the subjects underwent five training sessions. Before training, exercise was followed by severe muscle pain (delayed‐onset muscle soreness; DOMS), and by significant increases in plasma IL‐6 level and serum myoglobin concentration (SMb) (P < 0.001). After training, postexercise DOMS and SMb values were significantly lower than those measured before training. There was no significant difference between plasma IL‐6 levels measured at the same time points before and after training. We conclude that the hypothetical relationship between exercise‐induced muscle damage and increased postexercise levels of circulating IL‐6 is not substantiated by the present results. © 1999 John Wiley & Sons, Inc. Muscle Nerve 22: 208–212, 1999

An increasing body of data suggest that strenuous exercise triggers an inflammatory response having some similarity with those occurring in sepsis. Indices of this inflammatory response to exercise (IRE) especially include leukocytosis, release of inflammatory mediators and acute phase reactants, tissue damage, priming of various white blood cell lines, production of free radicals; activation of complement, coagulation and fibrinolytic cascades. Inflammatory responses to strenuous exercise and sepsis could in part be due to the release of endotoxin in blood as common triggering factor, but it seems that tissue damage and/or contact system activation are more important triggering mechanisms in exercising subjects. While the magnitude and duration of cellular and humoral changes associated with IRE are quite different from those observed in sepsis, recent human studies suggested that chronic and/or excessive IRE could have adverse effects. Among the possible consequences of acute and chronic IRE are delayed onset muscular soreness and loss of force, cardiovascular complications, intravascular hemolysis, hypoferraemia and increased susceptibility to infection.

To examine whether endotoxaemia accompanying long-term, strenuous physical exercise is involved in exercise-induced increase in plasma tumour necrosis factor alpha (TNF-alpha) concentration and polymorphonuclear neutrophil (PMN) activation, 14 male recreational athletes [mean age 28 (SEM 1) years] were studied. Exercise consisted of a 1.5-km river swim, a 40-km bicycle race, and a 10-km road race. Mean time to complete the race was 149.8 (SEM 4.8) min. The plasma concentrations of granulocyte myeloperoxidase (MPO) and TNF-alpha were significantly higher than baseline values immediately and 1 h after exercise (P<0.001). Both variables returned to pre-race levels the day after exercise. Marked, transient decreases in plasma concentrations of anti-lipopolysaccharide (LPS) immunoglobulin G (IgG) and immunoglobulin M (IgM) antibodies directed against a panel of selected smooth gram-negative LPS were observed after the race, reaching in most cases minimal values in the blood sample drawn immediately following the completion of the triathlon. There was no significant correlation between the magnitude of PMN activation, as assessed by the increase in plasma concentrations of MPO, and the humoral markers of endotoxaemia and TNF-alpha. An inverse, highly significant relationship between the increase in plasma TNF-alpha concentrations and the changes in circulating anti-LPS IgM antibodies concentrations was observed (r = -0.7; P<0.01). These findings suggest that exercise-induced endotoxaemia was involved in the release of TNF-alpha, that the magnitude of the TNF-alpha response to exercise was down-regulated by anti-LPS antibodies of the IgM class, and that the production of TNF-alpha and endotoxaemia did not seem to play a role in the activation of circulating PMN in the exercising subjects.