Quantitative magnetization transfer and g-ratio imaging of white matter myelin in early psychotic spectrum disorders

Abstract

Myelin abnormalities in white matter have been implicated in the pathophysiology of psychotic spectrum disorders (PSD), which are characterized by brain dysconnectivity as a core feature. Among evidence from in vivo MRI studies, diffusion imaging findings have largely supported disrupted white matter integrity in PSD; however, they are not specific to myelin changes. Using a multimodal imaging approach, the current study aimed to further delineate myelin and microstructural changes in the white matter of a young PSD cohort. We utilized quantitative magnetization transfer (qMT) imaging combined with advanced diffusion imaging to estimate specific myelin-related biophysical properties in 51 young adult PSD patients compared with 38 age-matched healthy controls. The macromolecular proton fraction (MPF) obtained from qMT was used as a specific marker of myelin content. Additionally, MPF was employed along with diffusion metrics of axonal density (v_ic) and extra-cellular volume fraction to derive the g-ratio, a measure of relative myelin sheath thickness defined as the ratio of inner to outer axonal diameter. Compared to controls, we observed a widespread MPF reduction and localized g-ratio increase in patients, primarily those with a diagnosis of schizophrenia or depressive schizoaffective disorder. No between-group differences were noted in v_ic, suggesting similar axonal densities across groups. Correlation analysis revealed that lower MPF was significantly related to poorer working memory performance in PSD, while the HC group showed a positive association for working memory with both g-ratio and v_ic. The pattern of changes observed in our multimodal imaging markers suggests that PSD, depending on symptomatology, is characterized by specific alterations in white matter integrity and myelin-axonal geometry of major white matter tracts, which may impact working memory function. These findings provide a more detailed view of myelin-related white matter changes in early stages of PSD.

Figure 1.

Multimodal imaging processing pipeline. Greyscale display ranges for quantitative MRI and diffusion maps: [0 – 0.12] for MPF, [0 – 1] for FA, v_iso, v_ic, and g-ratio maps. Red and blue boxes denote acquired images and fitted maps. Green box shows TBSS-related images (from top to bottom: individual FA map registered to MNI space; white matter skeleton superimposed on MNI space T1w image, colored by FA values; JHU atlas used for anatomical reference in result interpretation).

Figure 2.

Differences in MRI metrics between patient groups and healthy controls (HC) overlaid on white matter skeleton (green), colored by mean FA value (range 0.3–1). Significant differences (FWER<5%) are shown in red, except for significant MPF decreases in the PSD-SZ group that overlap with trend-level MPF decreases in the entire PSD patient group (yellow in the first row). Voxels with decreased MPF in PSD but not PSD-SZ are very limited (3 voxels only) and are therefore not shown here. Confirmatory ROI analyses of group comparisons were performed with age and sex included as covariates. Results from an example region (right superior longitudinal fasciculus, approximate location indicated by the cyan box) are shown in boxplots on the right. White dots inside the box plots indicate group medians.

Figure 3.

Correlations between MRI metrics and behavioral measurements in the PSD and HC groups. Red and yellow voxels overlaid on the white matter skeleton indicate correlations at statistically significant or trend levels, respectively. Scatter plots with linear fitting lines and 95% confidence intervals are shown for example regions, with blue and grey representing the HC and PSD subjects. Data plotted in scatter plots are residuals controlling for age and sex. Additional skeleton figures are included in the supplementary materials (Figure S2).