The beneficial effects of regular moderate exercise on the cardiovascular system are well established. Regular exercise improves oxidative capacity (Irving et al., 2015), microvascular collateral formation (Möbius-Winkler et al., 2016) and contractile myocardial function (Nystoriak and Bhatnagar, 2018). Among endurance athletes, there is evidence that prolonged and strenuous exercise is associated with oxidative stress, increased cardiac volume load (Małek Ł et al., 2019) and oxidative DNA damage (Borghini et al., 2015) and may exert exercise-induced myocardial damage and even injury in terms of reversible fatigue or irreversible heart failure (Douglas et al., 1990). Furthermore, an increased risk of exercise-related sudden cardiac death (SCD) has been reported among endurance runners with pre-existing cardiac abnormalities, most likely caused by prolonged strenuous exercise (Albert et al., 2000).

This assumption is supported by studies from the 1990s which reported increased markers of cardiac and muscular damage like troponins (Tn) and creatine kinase (CK) in otherwise healthy individuals after heavy exercise and extreme endurance events such as the Berlin Marathon (Artner-Dworzak et al., 1990; Koller et al., 1995). Additionally, systematic reviews and meta-analyses within the last decade (Regwan et al., 2010; Sedaghat-Hamedani et al., 2015) showed increased Tn values above the 99th percentile (14 ng/L) in 83% of endurance athletes immediately after prolonged and strenuous exercise (Sedaghat-Hamedani et al., 2015). However, our group was able to demonstrate a return of Tn levels to baseline levels within 72 h after a marathon race (Scherr et al., 2011). The time of appearance of the absolute peak value and downslope differ from e.g. myocardial infarction with much faster kinetics pointing to a revesible effect, which leads to the conclusion that these are benign phenomenon. However, recent imaging studies suggest a compromised cardiomyocyte integrity (Aengevaeren et al., 2020).

One potential key to unmask the discussed underlying mechanisms of this subclinical cardiac damage could be markers of immunogenic cell damage (ICD) such as circulating nucleosomes, high-mobility group box 1 protein (HMGB1), and soluble receptor for advanced glycation end-products (sRAGE). Nucleosomes are basic components of nuclear chromatin and are formed by a complex of core histone proteins and DNA that is twisted around them. Linker DNA connects them to a nucleosomal chain (Koyama and Kurumizaka, 2017). After cleavage by specific endonucleases, they are released by dying and stressed cells into the blood circulation. HMGB1 is a nuclear non-histone-DNA-binding protein that stabilizes the nucleosome structure, enables the bending of DNA and facilitates transcription (Balliano et al., 2017). Upon cell stress, HMGB1 can be released in association with nucleosomes or dissociated from them (Bell et al., 2006). RAGE is an important cellular receptor and binding partner for HMGB1, S100 and other danger associated molecular pattern (DAMP) markers transmitting proinflammatory signals on the surface of antigen presenting cells, dendritic cells and macrophages (Pilzweger and Holdenrieder, 2015). However, RAGE can also be shed from these cells and act as a soluble anti-inflammatory decoy receptor by catching HMGB1 and other DAMPs (Bell et al., 2006). These immunogenic markers are released by stressed/damaged cells like muscle, immune, endothelial or liver cells by necrotic, apoptotic or NETosis pathways and therefore, are associated with cardiac cell stress, acute and chronic diseases like diabetes (Volz et al., 2010), arteriosclerosis (Andrassy et al., 2012), cancer (Stoetzer et al., 2012), trauma (Stahl et al., 2016), sepsis (Matsumoto et al., 2015), myocardial infarction (MI) (Sorensen et al., 2011), and heart failure (Volz et al., 2010; Pilzweger and Holdenrieder, 2015).

A recent study conducted by Beko and colleagues including 70 non-professional half marathon and marathon runners reported increased levels of HMGB1 and sRAGE immediately after the race. Both, HMGB1 and sRAGE returned to baseline values after two to 7 days of recovery (Bekos et al., 2016). Beko et al. were not able to detect an increase in sRAGE after a marathon race and they related this to the different training habits. However, they did not further evaluate the association with performance. Consistent with these observations of HMGB1 and sRAGE, circulating cell-free DNA (cfDNA), paralleled that of plasma HMGB1, increased in half marathon runners immediately after the race and decreased to baseline values within 2 h post-race (Atamaniuk et al., 2004). Although, the studies showed mostly an increase of these markers following exercise, many questions remain. To date, no study has evaluated the predictive factors e.g., performance, and laboratory markers associated with the increases of ICD and illuminated the sources of ICD. Therefore, we aimed to evaluate the kinetics of HMGB1, sRAGE, nucleosomes, hs-TnT and hs-CRP before, immediately after, and up to 12 weeks post-race after a marathon race. Furthermore, we identified routine markers of clinical chemistry (e.g., Creatinine and CK) and physiological covariates (exercise capacity, age, BMI and marathon finishing time) associated with post-race concentration of these biomarkers.

Out of the initial 100 participants, 51 were included in the present analysis. Reasons for exclusion from final analysis (exclusion reasons during the study period) were: time constraints (n = 18), internal sicknesses (n = 6), orthopedic disease (n = 7), intermediate visit missing for time reasons (n = 10), marathon not finished (n = 4), termination of the study due to personal reasons (n = 1), exclusion criterion for marathon at V-1 (n = 1) and missing blood samples (n = 2). Baseline characteristics of the study population [n = 51, aged 43 ± 9 years (82% male)] are shown in Table 1. A total of 27 (52%) of the included participants had a positive family history of cardiovascular diseases.

In this prospective observational study, we observed that sRAGE, HMGB1, nucleosomes and hs-TnT increased significantly directly after the marathon. Only hs-CRP increased- as expected- with a delay of 24 h after the marathon. Our findings concerning the kinetics of those markers after marathon complement and extend the work by Beiter et al. (2011); Scherr et al. (2011); Bekos et al. (2016). The three studies showed an increase of HMGB1, nucleosomes or TnT immediately after exhausting exercise in marathon or treadmill runners. Contrary to our results, Bekos et al. did not demonstrate an increase of RAGE immediately after a marathon (Bekos et al., 2016), but only after a half-marathon The author hypothesized that the missing increase of RAGE immediately after the marathon could be the result of training habits or exercise intensity. In this regard, we tested the hypothesis, that training had an influence on the expression of ICD markers, in our analysis. The fact that we found an association between RAGE and marathon finishing time—slower runners had a higher decrease in the RAGE units—suggests that training conditions influence expression of RAGE markers after a marathon. We could not demonstrate further associations between the expression of RAGE and performance parameters. When comparing studies described in the literature regarding training influences on RAGE expression, it shows inhomogeneous results, with studies indicating an inducible decrease of RAGE after training interventions (Kotani et al., 2011; Drosatos et al., 2021). Nevertheless, two studies indicated an increase of RAGE values within training interventions in cardiovascular risk patients (Choi et al., 2012; Sponder et al., 2018). In one randomized controlled trial with type 2 diabetes patients, RAGE levels increased after a 12-week aerobic exercise intervention (60 min at moderate intensity, 5 times/week) (Choi et al., 2012). In a second prospective study, the influence of long-term physical activity (8 months) on serum sRAGE levels in 98 participants was investigated. RAGE levels increased up to 22% in participants with a long-term performance gain of >5% compared to <4.9% (Sponder et al., 2018). Authors of the mentioned studies hypothesized that this variating pattern of decrease and increase in RAGE levels depending on exercise intensity, points towards the capacity of sRAGE to act as a marker for increased or decreased AGEs production (Choi et al., 2012; Sponder et al., 2018). In the present study, we could not demonstrate any association between HMGB1 or Nucleosomes and training intensity. However, previous studies reported that regular training could decrease HMGB1 levels (Gmiat et al., 2017) and that HMGB1 reacts depending on the intensity of the load. This can rather be attributed to an anti-inflammatory process and thus to the chronic long-term effects regarding training. In this case, HMGB1 and nucleosomes are low and sRAGE is rather elevated (as described above). This should be investigated in further studies.

Baseline values such as age, sex or body compositions may contribute to changes in biomarkers released into the circulation. Contrary to the literature, in which more studies indicate an inducible increase in RAGE and nucleosomes in older participants (Pinti et al., 2014; Recabarren-Leiva et al., 2021), we did not find an association between the kinetics of ICD markers and age or sex in our cohort. Only body fat correlated negatively with the increase in post-race sRAGE levels, consistent with existing literature (Dozio et al., 2017; Miranda et al., 2018). A possible explanation for this could be the homogeneous age level (43 ± 9 years) of the participants included in this study. Furthermore, it should be considered that our sample size consists only of 9 females and 42 males, which limits the statistical validity.

It is still unclear whether marathon running can lead to an irreversible cardiac damage or just a reversible cardiac fatigue. Comparing the current results with previous studies assessing the kinetics of TnT and ICM markers in patients, the patterns are significantly different (Apple et al., 1998). In patients after myocardial infarction, TnT and ICM kinetics are characterized by a steep peak increase and followed by prolonged elevation for at least 4–7 days (Apple et al., 1998; Kohno et al., 2009). Furthermore, we found a positive correlation between the increase in sRAGE and hs-TnT (rs = 0.352, p = 0.011), which is a different pattern to that reported in a previous study in patients with non-ST-segment elevation myocardial infarction. This observational study indicated that low levels of sRAGE are associated with high serum levels of cTnI and therefore, suggested that the severity of cardiac damage varies inversely with the levels of sRAGE (McNair et al., 2011). Finally, we plotted the increase in ICD markers from pre to post marathon in relation to the established cardiovascular risk factors (hypertension). We found a significant difference of 0.7 ± 1.0 vs. 2.0 ± 2.0, p = 0.031 in the change of HGMB1 in patients with hypertension compared to without hypertension. However, it should be noted that only two subjects were diagnosed with hypertension.

Since we could not detect any clear cardiac damage, the increases in the markers tend to originate from cells that play a greater role in terms of quantity (muscle cells, endothelia, leukocytes, CRP from liver cells, etc.) and less from perishing cardiac cells. The TnT increase can be explained by a release from muscle cells and the cytosol of the cardiomyocytes (10% of the TnT) (Bleier et al., 1998). The structurally bound fraction, however, is only released after the cardiomyocytes have perished, which can explain the days-long increases in TnT levels after an MI (Pilzweger and Holdenrieder, 2015). Regarding the ICM markers, a distinction must be made with regard to the reaction of the markers to acute and chronic effects. The acute effects only occur immediately after the marathon and are quickly compensated in a good training situation. In the acute phase, all three markers, nucleosomes, HMGB1 and sRAGE, are elevated. After the impact is absent again, the markers will decrease. At least for HMGB1 it is known that–after cessation of the damaging stimulus–it may contribute to regeneration of the tissue and preservation of the organ function on a local level. Therefore, HMGB1 after an acute marathon is hypothesized to be released from inflammatory cells and exerts protective functions for the heart preventing pathological remodeling (Germani et al., 2007). Recent studies suggest that the release of nucleosomes attributes either a rapid, active release mechanism from immune- or endothelial cells, or a passive shear stress-induced detachment mechanism from the cell surface area (Neuberger and Simon, 2022). The typical cell death mechanisms such as necrosis, apoptosis, and “suicidal NETosis (Brinkmann et al., 2004)”, in which chromatin is released from dying neutrophils after chromatin de-condensation and subsequent rupture of the nuclear and cell membrane, requires hours to complete (Neuberger and Simon, 2022). In contrast, the “vital” NETosis, in which detachment occurs passively from the cell surface can happen very rapidly (Pfister, 2022). As described in Pfister, in the vital NETosis, vesicles, containing nuclear DNA from neutrophils, merge with plasma membrane eventually resulting in an alive anuclear cytoplast (Pfister, 2022). Therefore, it seems to be well imaginable that “vital” NETosis is a major mechanism for the release of nucleosomes after marathon. A prolonged increase of nucleosomes following marathon however could originate either from reduced elimination and/or subsequent cell death originating from damaged tissues. In conclusion, the transient increase in these parameters may not reflect a cardiovascular strain. Further studies should investigate the mentioned other originating cell forms.

In our study, markers of immunogenic cellular damage increased significantly in the early period after prolonged and strenuous exercise and returned to baseline values after a recovery phase of 24–72 h. Based on our results, it seems as if a marathon event does not lead to necrotic alterations but rather points to transient ischemic alterations instead. Therefore, there seems to be no increased risk for persistent cardiac necrosis if no underlying pathology is present. Furthermore, the training condition of active patients should be observed, as more intensive training or a regular training rhythm can influence the expression of the evaluated biomarkers.