UCLA

Accumulating evidence suggests a neuroinflammatory subtype of psychotic illness. Magnetic resonance spectroscopy (MRS) can indirectly assess neuroinflammation via measurements of neurometabolite biomarkers: myo-inositol (mI) for glia; choline (Cho) for membrane turnover; creatine (Cr) for energy metabolism; N-acetylaspartate (NAA) for neural integrity; and glutamate and glutamine (Glx) for glutaminergic neurotransmission. Previous MRS studies of psychoses show mixed findings, which could reflect most: 1) only comparing metabolite levels between cases and controls rather than also examining associations with symptoms and cognitive deficits; 2) intermixing different psychoses (e.g., schizophrenia vs. affective psychoses) rather than stratifying them, a common research practice that echoes transdiagnostic approaches like NIMH’s Research Domain Criteria (RDoC), which posits similar cellular pathologies across diagnostic categories for deficits with neurobehavioral functions. The present study used hippocampal proton magnetic resonance spectroscopic imaging (¹H-MRSI) and measured correlations of neurometabolite concentrations with MATRICS cognitive scores and symptom levels (five-factor PANSS and Hamilton Depression and Young Mania Scales) in 26 healthy, 22 non-psychotic affective, and 33 psychotic participants (including 20 schizophrenia and 13 affective psychosis). Despite similar metabolite levels across groups, affective psychosis showed a strong association between increased mI (glia) and worse attention (ρ=-.847, p=.016), and a moderate association between increased Cho (membrane turnover) and mania (ρ=.654, p=.015). In schizophrenia, elevated levels of PANSS activation (behavioral dyscontrol) factor were related to reductions in all neurometabolites (NAA, Cr, Cho, mI, Glx) (p’s<.05), robustly for mI (ρ=-.842, p=.0002). Between-group comparisons suggested distinct cellular abnormalities across diagnostic categories for social cognition and reasoning deficits, activation, and mood symptoms, but similar abnormalities for negative symptoms and PANSS autistic preoccupation (disorganized) factor. These findings suggest neuroinflammatory underpinnings for attention deficits and manic symptoms in affective psychosis patients with greater severity of these clinical features. Results in schizophrenia indicate a process causing widespread multisystem cellular deficits and substantial glial loss for patients with higher levels of activation. Both distinct and similar cellular pathologies were suggested for cognitive and symptom domains across psychotic vs. non-psychotic disorders and affective vs. non-affective psychoses. The differences support categorizing disorders and stratifying different psychoses in research rather than fully abandoning diagnostic labels for transdiagnostic approaches.