2.5. Statistical analysis

The data were conducted using the SPSS Version 21.0 (SPSS, IBM Corporation). A nominal P value < 0.05 was regarded as significant using the 2-tailed test. Using the Shapiro–Wilk test, we determined that age, metabolite concentrations, brain volumes, and DTI measures of cortical and subcortical structures were normally distributed (Shapiro–Wilk test, P > 0.05 for all of them), whereas white matter lesion volume was not normally distributed (Shapiro–Wilk test, P < 0.001).

Differences between the groups were then tested using 2 independent sample t-tests for continuous and normally distributed data and χ2 test or Fisher's exact test to analyze categorical variables. The relationships between metabolite measures and clinical parameters (cognitive and psychiatric measures) and brain volumes were assessed using Pearson's correlation. Correlations of 1H-MRS findings with white matter lesion volume were assessed using the nonparametric Spearman's rank correlation.

Differences in fractional anisotropy and mean diffusivity between groups were obtained by performing a voxelwise morphometry analysis by means of the T-test, with age and gender as covariates. The result of the statistical analysis was a T map that showed the differences between the control and patient group. Correction for multiple comparisons was applied to test these differences between groups. A threshold of P < 0.001 was considered statistically significant.