Bolte, L.; Gifford, G.; Serpa, M.; Cottaar, M.; Dazzan, P.; Fusar-Poli, P.; Tognin, S.; Kempton, M.; Winter-von Rossum, I.; Slot, M. I. E.; van Hell, H.; Maat, A.; de Haan, L.; Crespo Facorro, B.; Glenthoej, B.; Lawrie, S.; McDonald, C.; van Amelsvoort, T.; Arango, C.; Falkenberg, I.; Nelson, B.; Galderisi, S.; Bressan, R. A.; Kwon, J. S.; Cho, K. I. K.; Weiser, M.; Mizrahi, R.; Sachs, G.; Kirschner, M.; Taquet, M.; Oliver, D.; PSYSCAN Consortium, ; Kahn, R.; McGuire, P.

Psychosis has been linked to changes in diffusion-derived fractional anisotropy (FA) across multiple white matter tracts, yet most evidence is cross-sectional. Longitudinal findings are inconsistent, and the clinical relevance of white matter changes over time remains unclear. To address this gap, we investigated longitudinal white matter changes and their clinical correlates in early psychosis in a large, multi-centre diffusion-tensor imaging study. Across 18 sites, 407 participants (healthy controls: n = 97, 59.8% men, mean age 23.9 years; clinical high risk of psychosis: n = 158, 53.5% men, mean age 23.0 years; first-episode psychosis: n = 152, 71.1% men, mean age 25.1 years) were scanned at up to three time points over 12 months. FA was assessed in the cingulum bundle, the superior longitudinal fasciculus, the inferior fronto-occipital fasciculus, and at the whole brain level. FA trajectories were analysed using linear mixed-effects models to test effects of group, social and occupational functioning, and (attenuated) psychotic symptoms, while controlling for demographic and socioeconomic covariates. No significant group differences in FA were observed, either globally or within tracts (p > .05). Longitudinal FA trajectories were not associated with changes in (attenuated) psychotic symptoms or functioning (p > .05). In contrast, higher baseline antipsychotic medication dose was significantly associated with lower FA (pcorr < .05). Overall, these findings indicate that white matter microstructure is relatively stable during the clinical-high-risk and first-episode phases of psychosis, and that it is not clearly associated with variation in symptom severity or functional outcomes over time.