A population based multimodal MRI study for Chinese youth with psychotic experiences in Hong Kong

Abstract

Psychotic experiences (PEs) refer to subclinical hallucinatory and delusional symptoms in community sample, and may represent early risk marker for psychotic disorders and other mental illnesses. Accumulating data have revealed that individuals with PEs, the early part of extended psychosis phenotype, exhibited similar but milder cognitive, reinforcement and effort-cost processing, and brain abnormalities than patients with psychosis. Notably, most studies were from Western countries, with some studies examining “non-pure” PE population and young children with modest sample size. Hence, in this thesis, a comprehensive study encompassing psychopathological, standard and hypothesis-driven cognitive functions, and multi-modal magnetic resonance imaging (MRI) measurement is conducted. In addition, meta-analysis on cognitive functioning in PEs was conducted because numerous studies with mixed findings have been recently published. PE individuals exhibited poorer general cognition, verbal fluency, visual and working memory. Meta-regression revealed that cognitive performance in PEs was moderated by age, sex and study design. In the current case-control study, 200 participants (100 with and without PEs) were recruited from the HK-YES cohort, a population-based epidemiological survey examining mental health condition in 3600 community-dwelling Chinese youths aged 15-24 years in Hong Kong. First, symptom severity, self-injurious behaviors, psychosocial functioning, subjective quality of life (QoL) and cognitive performance were examined in PE individuals. PE individuals exhibited more severe symptoms, more prevalent self-injurious behaviors, poorer psychosocial functioning and subjective QoL, and comparable cognitive performance than non-PE counterparts. Second, two computerized paradigms, effort-based decision making (EBDM) and reinforcement learning (RL), were examined in PE individuals. Subtle and circumscribed abnormalities in RL and EBDM were found. Specifically, PE individuals, in particular with high level of amotivation demonstrated effort allocation abnormalities, and abnormal EBDM was correlated with poor psychosocial functioning. In RL, PE individuals showed mild deficits in punishment-driven rapid RL and reward-driven overall RL. PE individuals exhibited preserved value-guided decision making. RL measures were also negatively correlated with negative symptoms in PE individuals. Third, brain abnormalities using multi-modal MRI were compared between PE and non-PE individuals. In structural MRI, PE individuals were associated with grey-matter volume reduction in cerebellum and cortical thinning in parietal and temporal regions. In white matter (WM) tracts using diffusion-tensor imaging, PE individuals showed approached significant reduction in fractional anisotropy clustering in right caudate WM, and significantly more widespread increase in mean, axial and radial diffusivity. When examining functional connectivity using resting-state functional MRI, within-network hypoconnectivity in salience network was observed in PE individuals. Overall, PEs are associated with more severe nonpsychotic psychopathologies, higher levels of suicidality and worse psychosocial functioning, underscoring the presence of PEs deserves psychiatric evaluation and appropriate interventions in its own right. In addition, current findings affirmed subtle neuroanatomical and functional brain abnormalities, and revealed mild RL and EBDM impairment in PE individuals, which may facilitate the identification of early markers for outcome prediction of PEs. Future study should develop gold standard on definition and assessment tools of PEs, and examine the persistence of PEs, and the conversion risk to psychosis and other mental illnesses using longitudinal design.