*2.5. Statistical Analysis*

Statistical analyses were carried out using the Statistical Package for the Social Sciences software (IBM SPSS Statistics for Windows, version 25.0, IBM Corp., Armonk, NY). Normal distribution of the variables was verified through the Shapiro-Wilk test (*p* > 0.05) [34].

A Pearson or Rho Spearman correlation analysis was used to assess whether the concentration of oxidative stress biomarkers (GR, GPX, MDA and CG) was interrelated or related to the loss of upper (HG) and lower limb (SJ) strength, hematologic variables of systemic inflammation (CRP) and muscle damage (CK, LDH), as well as cardiopulmonary exercise test results. Post-race and at 24 and 48 h values for these variables (GR, GPX, MDA, CG, HG, SJ, CRP, CK and LDH) for each participant were related to the individual pre-race level to define the delta scores (Δ): Δ (fold increase) = (post-race value − Pre-race value)/Pre-race value [3].

On the other hand, the quantitative variables of oxidative stress were compared using the Student method Tests or U Mann Whitney in each of the sectors where measurements were taken (pre-race, finish line, 24 and 48 h after the race) when they existed two categories and ANOVA test or Kruskall Wallis when there were more categories. Post-hoc comparisons were performed using Bonferroni adjustment for multiple comparisons.

The meaningfulness of the outcomes was estimated through the partial estimated effect size (η2 partial) for ANOVA and Cohen's d effect size for pair wise comparisons. In the latter case, a Cohen's d < 0.5 was considered small; between 0.5–0.8, moderate; and greater than 0.8, large [35]. Likewise, correlations > 0.5 were considered strong, 0.3–0.5, moderate and <0.3, small. The significance level was set at *p*-value < 0.05 and data are presented as means and standard error of the means (±SEM).

Finally, the multiple regression analysis was performed using the forward stepwise method. Only normally distributed variables were used as dependent variables. Among the different models obtained, the parsimony principle was applied [36]. Given our limited sample size and the non-normal distribution of independent variables, residual errors from the resulting models were inspected to ensure their normal distribution and thus the reliability of our regression models [37]. To identify the predictive value of the model, the Cohen criterion [38] was applied to one-way ANOVA models. This criterion indicates that R<sup>2</sup> values less than 0.10 do not present a relevant explanatory value; an R2 between 0.10 and 0.25 indicates a dependency of the analyzed variables variance explanation for the identified factors; and R<sup>2</sup> values above 0.25 is possible to affirm that the explanatory model clinically relevant.

#### **3. Results**

#### *3.1. Demographic Characteristics of The Participants*

Thirty-two runners reached the finish line. Nineteen were male and thirteen females, with an average finish time of 21 h 21 min ± 3 h 28 min. All levels of performance were represented in our sample, as shown by their rank, ranging from 7th to 32nd. The main characteristics of these runners are described in Table 1, including sex and age differences. As expected, males showed significant higher pre-race values in weight, BMR, BMI and percentage of muscular mass when compared with female runners. No differences were found in training characteristics or experience between male and female runners. The runner's age did not affect any of the parameters measured except for the weekly running volume that was smaller in the senior runner group.

**Table 1.** Baseline characteristics of the runners which completed the race by sex and age (Average ± SE).


Data partially published previously by our group (Martinez-Navarro et al., 2020) [3]. Abbreviations: BMR: Basal Metabolic Rate; BMI: Body Mass Index. \* *p* < 0.05 vs. Males; # *p* < 0.05 vs. Young and Medium.

#### *3.2. Analysis of Plasma Markers of Oxidative Stress*

Descriptive data of oxidative stress biomarkers pre-race (baseline), finish line and after 24 and 48 h post-race are depicted in Table 2. Regarding the antioxidant defenses, no significant changes were observed in GPx activity. The GR activity was significantly enhanced in the finish line. The GR enzymatic activity reached the highest value 24 h postrace and returned to normal values after 48 h. Lipid peroxidation (MDA concentration) was also increased in the finish line, declined 24 h post-race, and was significantly increased after 48 h. Oxidative damage to proteins (CG content) also increased immediately after the race and remained elevated 48 h later.

**Table 2.** Changes in plasma markers of oxidative stress throughout the study period (Average ± SE).


\* *p* < 0.05 vs. preceding time point; # *p* < 0.05 vs. baseline value.
