Chemical analyses were performed at the end of the exercise (VO2max, and at the 5th min post-exercise).Statistical analysisBlood from an antecubital vein was collected on lithium heparinate at rest and at the end of the cycling exercise. An aliquot sample was used to measure TBARS and reduced ascorbic-acid (RAA) according to methods previously published [8, 10, 11, 36] and originally described by Uchiyama and Mihara  and Maickel , respectively.CD26 peptidase activityData are presented as mean ? standard error of means (SEM). A two-way ANOVA was performed to compare the baseline levels of the biochemical markers between ME/CFS patients and controls. The least square regression analysis was used to compare CD26-expression, TBARS (at rest and post-exercise), M-wave amplitude variations and LHS/MOS SF-36 data. The significance was considered when P < 0.05.ResultsBiochemical variables and muscle excitability at rest and at VO2maxPBMCs were isolated from the blood using the Vacutainer CPT system (Becton ickinson). As previouslyTable 1 shows the significant biochemical differences observed between ME/CFS patients and controls at rest:Fenouillet et al. J Transl Med (2016) 14:Page 4 ofthe RAA/TBARS ratio and the expression of surface CD26 per PBMC were lower in the patients. PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/28893839 Exercise-induced changes in M-wave amplitude (M-wave) were significantly higher in patients than in controls (Table 1). A significant increase in TBARS postexercise was found in patients only. Because the duration of the exercise test (10?2 min) is well below the time needed for the de novo synthesis and cell surface expression of CD26 , we did not examine in all patients whether the cycling exercise could affect CD26 expression (we addressed the situation in 10 patients and did not find any differences). Together, the data obtained at rest and VO2max show that the redox status, CD26-expression, and muscle excitability were altered in ME/CFS. When we examined whether these characteristics are associated, we found (1) a negative correlation between M-wave and TBARS (Fig. 1a), (2) a positive correlation between M-wave and order 4-Hydroxytamoxifen CD26-expression (Fig. 1b), and (3) a negative correlation between TBARS and CD26-expression (Fig. 1c). We found no correlation at rest between the TBARS level, RAA/TBARS ratio and CD26-expression.Relationship between biological markers and healthrelated quality of lifeaM-wave amplitude max,100 50 0 -50 -100 -150 -200 0 50 100 150 200 250 300M-wave = – 18.12 – (0.30 * TBARS)r = 0.475; p < 0.01 ME/CFS ControlsTBARS max, restbCD26 rest4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 100CD26 = 3.22 - (0.008 * M-wave) r = 0.592; p < 0.----M-wave amplitude max,cCD26 rest4,5 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0 50 100 150 200 250 300In the patients' population, the scores of health-related quality-of-life were plotted according to the level of TBARS or CD26-expression. The LHS score was negatively correlated with TBARS (Fig. 2a) and positively correlated with CD26-expression (Fig. 2b). The pain component of the SF-36 questionnaire was negatively correlated with CD26-expression (Fig. 2c).Relationship between biological markers and stressorsCD26 = 3.170 - (0.004 * TBARS) r = 0.456; p < 0.TBARS max, restFig. 1 M-wave, exercise-induced redox stress and CD26-expression. Correlation between the decrease in M-wave amplitude post-exercise (M-wave) and the maximal increase in TBARS level induced by exercise (TBARS) (percent of its resting level; a) Correlation between M-wave an.