2.2.3. Statistical Analyses

The baseline characteristics of the participants were analyzed using one-way analysis of variance (ANOVA). After accounting for normality, two-way repeated-measures ANOVAs were performed since the autonomic nervous system response indices, salivary cortisol, POMS2, SF-36, OSA, and Uchida–Kraepelin test (correct answer rate) were normally distributed. Sessions (weeks 0, 4, 8, and 12) were within-subject independent variables and the study group (placebo and *Euglena* intake groups) were between-group independent variables. If a significant main effect of group, or a session x group, the interaction was found, we then followed up with post hoc comparisons using Dunnett's test. Kruskal–Wallis test was performed for VAS and PSQI since no normal distribution was found. If the Kruskal–Wallis test was significant, the Steel's test was then performed. Dunnett's test or Steel's test was used to compare the measured values at each measurement time point and/or the degree of change from week 0 to after test among the four groups. Numerical values are shown as means ± standard deviation (SD). Significance levels were set as follows: <5% in two-sided tests indicated a significant difference. In addition, < 10% in the two-tailed test showed a significant trend. R version 3.5.1 (R development Core Team), and JMP (ver. 12) were used for statistical analyses.
