ackground
Individuals at ultra-high-risk for psychosis (UHR) are characterized by decline in cognitive functions. As cognition are suggested to be structural dependent on cerebral white matter organization, we here examine, if patterns of cognitive functions are associated with alterations in white matter for UHR compared to healthy controls (HC).
Methods
116 UHR and 49 HCs underwent diffusion weighted imaging using a 3 Tesla magnetic resonance imaging scanner and were cognitively assessed. Group differences on whole brain fractional anisotropy were tested using tract-based spatial statistics. With univariate general linear modelling we tested group differences on cognition and white matter fractional anisotropy, voxel-wise and in regions of interest. Using multivariate testing, we examined associations between patterns of cognitive functions and regional fractional anisotropy. Finally, we tested covariance between the patterns of cognitive functions and additional white matter measures.
Results
As expected for UHR, we found significant impairments on 14 out of 17 outcomes for cognitive functions, and lower fractional anisotropy in one focal cluster comprising the superior longitudinal fasciculus R and cingulum (cingulate gyrus) R. Multivariate analyses indicated different associations between patterns of cognitive functions and white matter microstructure for UHR compared to HCs. Significant correlations between similar cognitive function and different regional fractional anisotropy patterns in UHR compared to HCs (omnibus test p=0.038) was revealed. Two strongly significant covariations were identified: LV5 (p=0.002) explaining 7.4% of the covariance; and LV6 (p=0.011) explaining 5.5% of the covariance. Patterns of cognitive functions were associated with interaction effect on fractional anisotropy in localized regions: fornix and medial lemniscus bilateral (LV5), and uncinate fasciculus L and superior Cerebellar Penducle L (LV6). Analyses of additional white matter measures suggested dysmyelination as partly underlying the unique covariation between cognition and white matter microstructure for UHR.
Discussion
The localization of the white matter abnormalities is in accordance with previous studies identifying superior longitudinal fasciculus and cingulum with altered white matter microstructure in patients with schizophrenia and affective disorders. The unique associations between specific patterns of cognitive functions and regional interaction-effects on fractional anisotropy, suggest that the underlying structural basis for cognitive function is different for UHR compared to HCs. The analysis of covariance on additional white matter measures suggested dysmyelination to be partly explanatory.