Ultra-marathon running refers to any running event over marathon distances and can be performed in either a single stage or multi-stage setting (Scheer et al., 2020). In recent years, such competition have become increasingly popular in mountain areas (Stuempfle et al., 2011; Gatterer et al., 2013b, 2020; Mrakic-Sposta et al., 2015; Zanchi et al., 2016; Belinchon-deMiguel et al., 2019, 2021; Hoppel et al., 2019; Nguyen et al., 2021). The most striking features of ultra-mountain marathon running, apart from the long distance, is the difference in altitude that the participants have to overcome, as well as the possible reduced oxygen availability depending on the altitude level. The human body is challenged to the utmost during such races and health problems may occur. Common problems beside others are electrolyte disturbances, hydration imbalances, cardiocirculatory and muscular issues (Gatterer et al., 2013b; Krabak et al., 2017; Knechtle and Nikolaidis, 2018). In the literature, special attention is given to dysnatremia since both dehydration and exercise-associated hyponatremia (EAH) (serum sodium concentration<135mmol/L; Hew-Butler et al., 2017) can lead to serious health issues and even death (Knechtle et al., 2019). The literature on the prevalence of dysnatremia in ultra-endurance events is ample as recently reviewed (Knechtle et al., 2019). Conversely, less data is available on ultra-marathon races performed in the mountains, even though research is progressing fast. During mountain ultra-marathon races, asymptomatic hyponatremia was found in 4–8% of the ultraendurance mountain runners (Page et al., 2007; Arnaoutis et al., 2020), whereas 3% of the participants were hypernatremic (>145mmol/L; Page et al., 2007). Page et al. (2007) furthermore showed that hypernatremia was associated with weight loss whereas hyponatremia with weight gain.

The pathogenesis and clinical representation of EAH have been recently reviewed by Knechtle et al. (2019) and Hew-Butler et al. (2017). Overdrinking and fluid retention due to non-osmotic secretion of arginine vasopressin (AVP) have been identified as the main risk factors (Speedy et al., 2000; Hew-Butler et al., 2017; Knechtle et al., 2019). Overdrinking may stem from the fact that in 1996, the American College of Sports Medicine position stand recommended to consume the maximal amount of fluids that can be tolerated (Convertino et al., 1996). Only in 2007, new recommendation emphasized the risks and consequences of both insufficient and excessive fluid consumption (Sawka et al., 2007). Since then drinking to thirst was considered best practice (Noakes, 2010; Hew-Butler et al., 2017). In addition, factors such as inflammation (linked to AVP regulation; Hew-Butler et al., 2017; Knechtle et al., 2019) and oxidative stress have been linked to the development of dysnatremia (Siegel, 2006; Hew-Butler et al., 2017). Further risk factors include exercise duration >4h, event inexperience, pre-exercise overhydration, low body weight, weight gain during exercise, female sex, and nonsteroidal anti-inflammatory drug use (Almond et al., 2005; Knechtle et al., 2019). Moreover, intrinsic factors like physical strain, perceived exertion, sweat rate and the sensation of thirst and external factors such as race distance, temperature, altitude, humidity, availability and quality of drinks have been identified as contributing factors (O’Connor, 2006; Rosner and Kirven, 2007; Hew-Butler et al., 2008; Schenk et al., 2010; Chlibkova et al., 2018; Knechtle et al., 2019). In this regard, two special features of ultra-mountain marathon running need to be addressed. These events are held at altitude of different levels. Altitude exposure, especially when combined with exercise, induces hypoxemia with profound physiological effects. For instance hypoxia induces oxidative stress (Gatterer et al., 2013a) and inflammation (Pham et al., 2021) and affects hormones regulating fluid homeostasis (Schlittler et al., 2020; Gatterer et al., 2021). A further characteristic is the large proportion of eccentric exercise (downhill running). Similar to hypoxia eccentric exercise induces oxidative stress and inflammation (Bruunsgaard et al., 1997; Lee et al., 2002; Goldfarb et al., 2005). As outlined before inflammation, oxidative stress and dysregulation of hormones involved in fluid homeostasis have been postulated to be involved in the development of dysnatremia. Therefore, these factors must be given special attention, especially in ultra-mountain marathon competitions of varying duration held at altitude (which means different hypoxia dose and eccentric exercise load).

The present work thus aims to describe changes in hormones (e.g., copeptin as a stable marker of AVP), cytokines (e.g., IL-6), oxidative stress and metabolic byproducts (e.g., lactate, ketones) postulated to be linked to sodium regulation during a short (i.e., 69km, 4,260m elevation-gain) and a long (i.e., 121km, 7,554m elevation-gain) mountain ultra-marathon running race with different high altitude exposure levels and eccentric exercise volumes.

During the short competition two out of the 11 finishers (18.2%) showed serum sodium concentrations >145mmol/L (148 and 147mmol/L). The single non-finisher also demonstrated a serum sodium concentration of >145mmol/L (146mmol/L). No cases of hyponatremia were registered. During the long competition, one athlete out of the seven finishers (14.3%) was hyponatremic (134mmol/L) and one hypernatremic (147mmol/L). All eight non-finishers showed serum sodium concentration in the normal range. Fluid intake per hour was 612.4±187.6ml/h during the short and 422.3±197.8ml/h during the long competition (p=0.057). Table 2 displays changes in putative factors regulating serum sodium homeostasis as well as standard clinical parameters during the short and the long competition. Whereas marked changes were registered in most of the parameters from before to after both races, different changes between the short and long competition were only found for reactance and phase angle of the bioimpedance analyses, red blood cell count, Hb and La concentration and CK (Table 2). The four cases of hypernatremia showed a loss in body mass from 76.5±3.9 to 72.4±3.3kg (p=0.002) and increases in ROS (0.179±0.004 to 0.200±0.142μmol/min; p=0.020), IL-6 (2.25±0.25 to 33.10±10.56pg./ml; p=0.010), copeptin (8.4±2.5 to 24.7±9.3pmol/L; p=0.040) and WBC count (6.9±3.0 to 15.9±5.4 10³/μl; p=0.005).

During the short competition percent changes in serum sodium concentrations were negatively related to percent changes in body mass (r=−0.812, p=0.002), total time (r=−0.608, p=0.047) and TRIMP (r=−0.653, p=0.030) and positively related to UREA (r=0.612, p=0.046). Correlations between percent changes in serum sodium concentrations and percent changes in bicarbonate (r=−0.572, p=0.066) and osmolality (r=0.582, p=0.060) did not reach statistical significance even though effect size was large. During the long race, no significant correlation between percent changes in serum sodium concentration and any other measured variable was detected. When combining data of the long and the short run, significant correlations between percent changes in serum sodium concentration and changes in body mass (r=−0.702, p=0.001) and osmolality (r=0.600, p=0.008) were recorded. Table 3 displays differences in changes between finishers and non-finishers of the long competition. Data show that body mass, UREA, CK and osmolality differed between finishers and non-finishers. When including all runners (short, long, finishers and non-finishers) percent changes in serum sodium concentration and body mass showed a more curvilinear (quadratic) rather than linear relationship with race time (Figure 2).

The main findings of this investigation were that during the short ultra-endurance mountain race no case of hyponatremia was observed whereas during the long race one out of seven was hyponatremic, yet without showing any symptoms potentially related to serum sodium aberrations (e.g., malaise, headache; Knechtle et al., 2019). Conversely, during both races cases of hypernatremia were recorded. Serum sodium concentration changes were related to race time, race intensity (i.e., TRIMP) and body mass changes during the short race whereas none of the assessed parameters where related to the long race indicating a threshold where other factors may determine serum sodium aberrations.

It is well established that drinking habits in conjunction with hormonal dysregulation, race duration and race intensity are the main factors linked to serum sodium alterations (Rosner, 2019). Present data support these findings during the short race showing a relationship between serum sodium changes and race time, race intensity and body mass changes. Interestingly no such relationship was detected for the long race. It is important to mention that hyponatremia is a condition linked to death during endurance races (Hew-Butler et al., 2017; Rosner, 2019). There was only one single case of asymptomatic hyponatremia occurring in an athlete competing in the long race. When having a closer look at this athlete, data indicate similar mechanisms as reported in the literature (Hew-Butler et al., 2017; Rosner, 2019). The athlete ran for 30h and showed only small changes in body mass (~−0.5kg), whereas Hb concentration and Hct indicated hemodilution (14.9g/dl and 46.7% before the race and 13.8g/dl and 41.8% after the race, respectively). This athlete also showed the highest values for ROS production corrected for plasma volume changes (0.258 vs. 0.203±0.017) and one of the highest post exercise copeptin values (37.2 vs. 26.7±12.3). High levels of copeptin and AVP during and at the end of ultra-marathon race (160km) despite decreases in [Na⁺] (confirming non-osmotic simulation of AVP secretion) are in agreement with Hew Butler et al. (2011). Obviously, this is a single case observation prone to errors, nonetheless it supports both well-established research findings (Page et al., 2007; Rosner, 2019), as well as recent postulations stating that oxidative stress is involved in hyponatremia development (Hew-Butler et al., 2017). Changes in values of the four hypernatremia cases were similar to the ones recorded for the normonatremic competitors (finishers) of both races. This was to be expected because the serum sodium values of the hypernatremic cases were only slightly above the threshold for hypernatremia of >145mmol/L.

Comparison of the outcomes of the short and the long run (Table 2) shows that race distance did not affect ROS production, interleukin-6, copeptin and serum sodium concentrations. This indicates that different hypoxic doses and eccentric exercise volumes did not have a large effect on those changes, at least in these competitions. Since muscle damaging downhill running is associated with higher CK values (Gatterer et al., 2013b), the higher values found after the long run confirm a higher proportion of eccentric exercise compared to the short race. Concerning hypoxia dose, no information is available, as oxygen saturation was not measured. However, at altitude levels of around 2,000–2,500m (where much of the racing during the long run took place, Figure 1), oxygen saturation of around 90% may be expected during low intensity exercise, with individual values even around 80% (Wiseman et al., 2013). Nonetheless, it must be recognize that the physiological stress induced by this relatively low hypoxic dose may have been limited. The comparison between race distances furthermore shows that reactance and phase angle of the bioimpedance analysis are differently affected. According to recent literature, increases in phase angle indicate fluid shifts to the intracellular compartment (Francisco et al., 2020). Thus, data seem to indicate that during the short run fluid shifts to the intracellular compartment are more pronounced compared to the long run. Differences in exercise intensity between the races (Gatterer et al., 2020) may to some extent explain these findings, since race intensity was reported to be related to leg edema formation (Wilcock et al., 2006; Bracher et al., 2012). A further finding was that lactate concentration was lower after the long run. Single lactate measurements must be interpreted with caution, as they may only represent the activity prior to the measurement. However, they might also reflect a slower running speed (Gatterer et al., 2020) or muscle/liver glycogen depletion (Segal and Brooks, 1979) during the long run. No differences were observed for ketones, indicating similar free-fatty acid breakdown (Evans et al., 2017).

An interesting observation was that the percentage changes in serum sodium concentration and body mass showed a curvilinear (quadratic) rather than a linear relationship with race time when data from all athletes (short and long distance, finishers, and non-finishers) were included in the analysis (Figure 2). Such a curvilinear relationship was recently reported for cardiac biomarkers and race duration in marathon running (Traiperm et al., 2021). Additionally, cardiac biomarkers seemed less elevated the longer the running time in a 118km ultra-mountain race (Martinez-Navarro et al., 2019). It was speculated that short running times may prevent excessive and prolonged volume overload and myocyte stretch (Shave et al., 2007; Serrano-Ostariz et al., 2011) whereas slow running times reflect low running pace and thus less overall strain (Traiperm et al., 2021). In the present investigation, the fastest runners of the short run showed the highest serum sodium increases and body mass reductions. With increasing race time, serum sodium concentrations seem to decrease whereas less of a body mass loss was apparent. This is in accordance to reports showing that race duration and body mass gain (or less decrease) are risk factors for hyponatremia (Rosner, 2019). Furthermore it supports the notion that the fastest runners lose the most body mass (Noakes, 2010). Interestingly, during the long run the opposite pattern was found. There the fastest runners showed the highest decreases in serum sodium concentrations and the lowest decreases in body mass. The longer the performance time the more the serum sodium increased and the more the body mass decreased. The area around the nadir (Figure 2) is identified at about 20h of performance time and is mostly represented by runners that did not finish the race. In addition to showing that the shortest and the longest race times might be associated with both increases in serum sodium concentrations and losses of body mass, this graph indicates that too much of a serum sodium concentration decrease and too less of a body mass reduction are associated with a higher drop-out risk. Certainly, this observation has to be confirmed in further studies and it must be emphasized that effect sizes of the correlations were rather low, and no causal effect can be inferred. Furthermore, the reasons for a race being abandoned can be numerous (e.g., pain, fatigue, and gastrointestinal problems) as can the factors that determine race time.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethical Commission of the Bolzano Hospital, Italy. The patients/participants provided their written informed consent to participate in this study.

KS designed, planned, and implemented the study. KS, SR, EP, KG, SM-S, and HG performed the measurements. HG and KS drafted the manuscript and analyzed the data. SR, EP, KG, and SM-S contributed in writing the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.