a. Significant effect of age, mean:(F(3,262) = 7.066; p = 0.0001) | Normal distribution (8/8 datasets) | Two-way ANOVA | 70% |
b. Significant effect of age, median:(F(3,262) = 7.027; p = 0.0001) | Mostly normal distribution (7/8 datasets) | Two-way ANOVA | 70% |
c. No significant effect of genotype in large-size study, (F(1,796) = 1.575; p = 0.21) | Mostly normal distribution (5/8 datasets) | Two-way ANOVA | 99.4% |
d. Significant effect of interaction of age and genotype, mean bouton number:(F(3,243) = 6.794; p = 0.0002)Uncorrected Fisher’s LSD post hoc for P120+ comparison: mean difference, 1.802; p < 0.0001; effect size, 1.07 | Mostly normal distribution (5/8 datasets) | Two-way ANOVA with Fisher’s LSD post hoc | Genotype: 68% |
e. Significant effect of interaction of age and genotype, mean bouton density:(F(3,236) = 3.346; p = 0.0199)Uncorrected Fisher’s LSD post hoc for P120+ comparison: mean difference, 0.9378; p = 0.0098; effect size, 0.66 | Normal distribution (7/8 datasets) | Two-way ANOVA with Fisher’s LSD post hoc | Genotype: 68% |
f. Per cell analysis shows C-bouton decrease by ∼8%; effect size, 0.51; p = 0.0337 | Normal distribution (2/2 datasets) | Mann–Whitney U test | 85% |
g. Per bouton analysis shows C-bouton decrease by ∼8%; effect size, 0.33; p = 0.0001 | Non-normal distribution (2/2 datasets) | Mann–Whitney U test | 99.9% |
h. Large-sampling strategy shows C-bouton decrease by ∼8%; effect size, 0.51; p = 0.0337 | Normal distribution (2/2 datasets) | Mann–Whitney U test | 85% |
i. Small-sampling strategy shows C-bouton increase by ∼16%; effect size, 0.36; p = 0.0039 | Non-normal distribution (one-half datasets) | Mann–Whitney U test | 99.9% |
j. Small-sampling strategy, reduced cell count, shows C-bouton increase by ∼28%; effect size, 0.89; p = 0.0038 | Normal distribution (2/2 datasets) | Mann–Whitney U test | 70% |
k. Unblinded analysis of C-bouton size shows decrease by ∼20%; effect size, 1.22; p = 0.0017 | Normal distribution (2/2 datasets) | Mann–Whitney U test | 94.4% |