A 10-Year Longitudinal Study of Brain Cortical Thickness in People with First-Episode Psychosis Using Normative Models

Abstract

Background: Clinical forecasting models have potential to optimize treatment and improve outcomes in psychosis, but predicting long-term outcomes is challenging and long-term follow-up data are scarce. In this 10-year longitudinal study, we aimed to characterize the temporal evolution of cortical correlates of psychosis and their associations with symptoms.

Design: Structural magnetic resonance imaging (MRI) from people with first-episode psychosis and controls (n = 79 and 218) were obtained at enrollment, after 12 months (n = 67 and 197), and 10 years (n = 23 and 77), within the Thematically Organized Psychosis (TOP) study. Normative models for cortical thickness estimated on public MRI datasets (n = 42 983) were applied to TOP data to obtain deviation scores for each region and timepoint. Positive and Negative Syndrome Scale (PANSS) scores were acquired at each timepoint along with registry data. Linear mixed effects models assessed effects of diagnosis, time, and their interactions on cortical deviations plus associations with symptoms.

Results: LMEs revealed conditional main effects of diagnosis and time × diagnosis interactions in a distributed cortical network, where negative deviations in patients attenuate over time. In patients, symptoms also attenuate over time. LMEs revealed effects of anterior cingulate on PANSS total, and insular and orbitofrontal regions on PANSS negative scores.

Conclusions: This long-term longitudinal study revealed a distributed pattern of cortical differences which attenuated over time together with a reduction in symptoms. These findings are not in line with a simple neurodegenerative account of schizophrenia, and deviations from normative models offer a promising avenue to develop biomarkers to track clinical trajectories over time.

Keywords: cortical thickness; long-term follow up; normative modeling; schizophrenia.

Fig. 1.

Joint-plot and marginal distributions of the median CT measures (y-axis) and associated deviation scores (x-axis), color-coded by a scanner, from the test samples of the adaptation set (cross-sectional TOP samples).

Fig. 2.

(A) Conditional main effect of diagnosis from the linear mixed model, showing that patients have reduced CT deviations overall (i.e., across all timepoints), color-coded by effect size; (B) time × diagnosis interaction effect, showing ROIs where the effect of diagnosis changes over time; (C) regression plot for mean CT deviation across all ROIs, showing that the differences evident in the deviations at the first timepoint attenuate at later timepoints. The color bar shows the effect size for each effect multiplied by the sign of the coefficient.

Fig. 3.

(A) Raw mean CT scores for individuals with schizophrenia and healthy CTRL, where the line segments connect the successive time points of each individual. (B) Deviations from the normative model for mean CT.

Fig. 4.

(A–C) ROIs showing a significantly higher proportion of negative extreme deviations among patients with schizophrenia (SCZ) compared with healthy CTRL at each timepoint. (d) χ test significant differences in negative extreme deviation distributions between people with SCZ and CTRL at each time point (see supplementary table 4 for a detailed summary of each ROI).

Fig. 5.

PANSS domain scores at the 3 timepoints. Most follow-up scores are significantly lower than baseline scores indicating a decrease in symptom severity over time (*P < .05, **P < .01, ***P < .001).

Fig. 6.

Results from the LME model testing for associations between symptom scores and cortical deviations (equation 2). We report man effects for the PANSS total domain (A), the PANSS general domain (B), PANSS negative symptoms (C), and a significant interaction in a single region between the PANSS negative scores and follow-up time (D). The color bar shows the effects size of each effect.