After completing the first "marathon-run" ever from Marathon to Athens 2500 years ago, Pheidippides died [25, 26]. Today morbidity and mortality during mara thon running are still considerably. Mortality varies with age from 0.3/100.000 in young athletes [27, 28] to 7/100.000 in elderly runners . An analysis of the London marathon (650.000 completed runs) indicated an average cardiac death rate of 1.25/100.000 . Although the most frequent disease associated with deaths during marathon is previously unrecognized coronary artery disease [30, 31] other thrombembolic incidents have been reported as well 1-7. Our data further emphasize the clinical significance of previously demonstrated activation of the coagulation system during excessive physical exercise [7–13, 15–19, 21–23, 32].
The results of our study demonstrate that whole blood coagulation is markedly altered during heavy endurance sports irrespective of the running fraction. In contrast, elevated platelet aggregation is observed particularly during marathon and to a lesser extent also during triathlon. We speculate that mechanical stress during running may trigger platelet activation. Thus, the higher incidents of severe thrombembolic incidents in otherwise healthy athletes during marathon compared to triathlon or long distance cycling may be explained by a marked activation of coagulation and platelets at the same time. Influences like venous insufficiency  may also contribute to a greater risk in runners.
Changes in hemostatic and fibrinolytic markers during marathon running were discovered more than 30 years ago . At that time the authors demonstrated a significant shortening of partial thromboplastin time while plasma fibrinogen levels and prothrombin times were not altered. By finding shortened euglobulin lysis times and increased levels of fibrin degradation products the conclusion was made that during marathon coagulation and fibrinolysis are globally activated. In addition, the concentration of prothrombin fragment 1+2 concentration as well as thrombin-antithrombin complexes in plasma increased indicating activated coagulation. Finally, the plasma concentrations of tissue plasminogen activator antigen, plasminogen activator inhibitor type 1 antigen, d-dimer and plasma fibrinogen degradation products increased in parallel confirming activated fibrinolysis . Accordingly, both coagulation and fibrinolytic activation occurs during a marathon run at the same time.
In contrast, a two hour treadmill exercise at an individual anaerobic threshold of 90%  showed only a small increase in thrombin generation markers, but a distinct increase in fibrinolysis. While moderate exercise led to an increase in plasmin formation only, very heavy exercise also inducted thrombin and fibrin generation balanced by an increase in plasmin generation . Similar results were reported for two hour triathlon demonstrating a moderate activation of coagulation as indicated by thrombin and fibrin formation accompanied by markedly increased plasmin formation . Thus, during exercise free of heavy running periods activation of coagulation and fibrinolysis does not occur in favour of increased fibrinolysis.
The effect of endurance sports on platelets and their activation was assessed during marathon by Rock  and Dimitradou , during triathlon by Mockel  and during a cycle ergometer test by Li . All studies report on platelet activation during physical activity, but a comparison of these sports using identical methods had not been performed previously. While all these studies were able to demonstrate platelet activation during heavy exercise they were not designed to distinguish between various disciplines and did not relate these findings to variables of coagulation.
While most studies on coagulation changes during physical activity assessed more or less isolated variables describing coagulation by measuring single clotting factors or decomposition products, rotational thrombelastometry analyzes both, clotting times and the stability of the clot over the time. Since rotational thrombelastometry is performed in whole blood, its results are affected by interaction of platelets, clotting factors, fibrin polymerization and fibrinolysis, giving information about the end product of coagulation, the clot itself, displaying overall coagulation activity. This method was validated in a model of systemic activation of fibrinolysis and coagulation . Whole blood coagulation in marathon runners demonstrated decreased clotting times and increased clot stability in runners after participating a downhill marathon . Our results are consistent with these alterations in coagulation after intrinsically activation. However, fibrin polymerization was not tested and whole blood coagulation in other endurance sports has not been analyzed to date. The INTEM assay was chosen with regard to the previously published study to confirm their findings during a marathon and to compare these findings to other endurance sports. The FIBTEM assay allows excluding platelet function from the whole blood coagulation assay and displays fibrin polymerization only. Since fibrinogen is an acute phase protein and inflammation during heavy exercise is a matter of discussion we decided to include this particular test into our study. Platelet function itself is not accurately mirrored during ROTEM analysis. Accordingly, we evaluated platelet function by multiple platelet function analysis (Multiplate, Dynabyte, Munich, Germany). For reproduction of results from platelet activation during a marathon after ADP activation which were previously described [9, 16, 21, 35] we activated platelets with ADP. Since TRAP-6 is recognized as an even stronger platelet activator we also decided to include TRAP-6 activation.
Clotting times in rotational thrombelastometry mirror activity of coagulation factors and were shortened throughout the tested groups indicating an increase of coagulation factor activity during heavy physical exercise. In contrast, maximum clot firmness is mostly influenced by fibrin polymerization and platelet function. In a perioperative setting it is shown that elevated results of maximum clot firmness in thrombelastometry strongly correlate with the incidence of postoperative thromboembolic events and myocardial infarction . Furthermore, increased platelet activation also elevates the thrombembolic risk [37, 38]. We found whole blood coagulation parameters increased in all tested groups, while platelet aggregation was altered during marathon and to a lesser extent during triathlon but not during cycling, conclude that the running fraction is responsible for platelet activation, while an activation of the plasmatic coagulation system is triggered by physical activity itself. Strenuous exercise activates coagulation probably due to inflammatory effects as shown by previously published studies [21, 39]. We found increasing leukocyte levels in all groups matching these results. Greater increases of leukocyte levels in marathon and triathlon might indicate greater inflammatory effects of these endurance sports. Moreover, we speculate that comparable to effects of mechanical stress during prolonged running on red blood cells leading to hemolysis [40–42], direct mechanical stress on circulating platelets during running contributes to the observed platelet activation.
Participants in the cycling group were slightly older compared to the marathon and triathlon group. However, there are no indicators that coagulation and fibrinolysis may be different at baseline when comparing subjects at age 41 and age 35. Training status may flaw baseline coagulation and fibrinolysis. However, this influence could be excluded since the duration of weekly training periods did not differ between the groups studied. Since we did not evaluate fibrinolysis further studies may clarify whether the enhanced whole blood coagulation and platelet aggregation during ex vivo test may be balanced by increased fibrinolysis.
The use of antithrombotic strategies has been shown to be effective in other not desease related situations of prothrombotic stages like thrombembolic events during air travel . Whether this also accounts for marathon running can not be assessed by our study and should be investigated in further studies.