Cognitive Impairment Characteristics and Biological Mechanisms in Untreated Severe Mental Illness

Cognitive function will be assessed using the MATRICS Consensus Cognitive Battery (MATRICS-CN). The following domains will be measured: processing speed, verbal learning, working memory, reasoning and problem-solving, visual learning, and attention/alertness. It possesses excellent psychometric properties, correlates with functional outcomes, and is brief, easy to administer, can be used for home-based assessments, is acceptable to participants, and has domain "robustness." The MATRICS-CN has been translated into Chinese and is available for purchase from PAR, Inc.

The psychiatric symptoms of schizophrenia were assessed in all enrolled patients using the Positive and Negative Syndrome Scale (PANSS). The PANSS is a 30-item clinician-rated scale yielding a total score ranging from 30 (least symptomatic) to 210 (symptomatic), where higher scores indicate more severe psychopathology.

This is an 11-item clinician-rated scale (scored from 0-4 or 0-8, depending on the items) that assesses the severity of mania and the response to treatment in patients with bipolar disorder. The assessment is based on clinical observation and patient subjective reports regarding symptoms over the past 48 hours.

This is a 10-item clinician-rated questionnaire (scored 0-6 per item) that assesses the severity of depressive episodes in patients with affective disorders. The assessment is based on the patient's subjective report of their symptoms over the past week.

Primarily used to assess the subject's psychological, social, and occupational functioning. The GAF consists of only one dimension: the overall clinical severity level. It is divided into 100 levels (scored from 1-100), where a higher score indicates less severe illness.

The antisaccade task consists of 4 blocks of 20 trials each. Each trial begins with a central fixation cross. A peripheral visual target appears at 10° or 15° visual angle from fixation, and the fixation cross disappears 200ms after target onset. Participants are instructed to inhibit the reflexive saccade toward the peripheral target and instead generate a saccade to the mirror image location in the opposite visual field. Eye movements are recorded using an eye tracker integrated with the EEG system. The primary outcome is the error rate, calculated as the percentage of trials in which the participant fails to inhibit the reflexive saccade and looks toward the peripheral target rather than the correct mirror location.

Auditory paired stimuli are presented binaurally as 150 pairs of broadband auditory stimuli (duration 4s, 75dB). The inter-pair interval averages 9.5s (range 9-10s), and the inter-stimulus interval within each pair is 500ms. EEG is recorded continuously during the task. The P50 evoked potential is identified as the positive deflection occurring 40-80ms post-stimulus at the Cz electrode. The P50 suppression ratio is calculated as the amplitude of the response to the second stimulus (S2) divided by the amplitude of the response to the first stimulus (S1). A lower ratio indicates stronger sensory gating.

Participants complete an auditory oddball task consisting of 567 standard tones (1000Hz) and 100 target tones (1500Hz) presented in a pseudo-random order, with an inter-stimulus interval of 1300ms. Participants are instructed to press a button upon hearing the target tone. EEG is recorded continuously. The P300 component is measured at the Pz electrode site as the largest positive deflection between 250-500ms post-stimulus. Primary outcomes are: 1) P300 amplitude (µV), defined as the peak voltage relative to a 200ms pre-stimulus baseline; and 2) P300 latency (ms), defined as the time from stimulus onset to the peak amplitude.

Resting-state EEG is recorded for 5 minutes with eyes closed using a [具体设备] EEG system with electrodes placed according to the 10-20 system. Data are segmented into 2-second epochs and artifacts are removed via independent component analysis. Power spectral density (µV²/Hz) is calculated using fast Fourier transform. The primary outcomes are mean power in the theta band (4-8 Hz) and alpha band (8-13 Hz) at frontal and parietal electrode sites.

High-resolution T1-weighted structural MRI scans are acquired using a [3T Siemens Prisma] MRI scanner. Images are preprocessed using voxel-based morphometry (VBM) with SPM12 software. Gray matter is segmented from white matter and cerebrospinal fluid. The primary outcome is the regional gray matter volume (mm³) in regions of interest, normalized for total intracranial volume.

Diffusion tensor imaging (DTI) data are acquired using a single-shot echo-planar imaging sequence with 30 diffusion directions (b-value = 1000 s/mm²) and one non-diffusion-weighted image (b=0). Data are preprocessed for eddy current and motion correction using FSL software. Fractional anisotropy (FA) maps are calculated, representing the directionality of water diffusion in white matter tracts. The primary outcome is the mean FA value in major white matter tracts (e.g., corpus callosum, cingulum bundle), with FA values ranging from 0 (isotropic diffusion) to 1 (completely anisotropic diffusion).

Resting-state functional MRI (fMRI) data are acquired using a gradient-echo echo-planar imaging sequence while participants are instructed to remain awake with eyes closed. Data are preprocessed including slice-timing correction, head motion correction, and spatial normalization to standard MNI space. The amplitude of low-frequency fluctuations (ALFF) is calculated as the square root of the power spectrum in the low-frequency range (0.01-0.08 Hz). The primary outcome is the mean ALFF value in predefined regions of interest, reflecting the intensity of spontaneous brain activity.

Using the same resting-state fMRI data described above, functional connectivity is assessed using a seed-based correlation approach. The mean time series is extracted from a seed region of interest (e.g., posterior cingulate cortex for default mode network). Pearson's correlation coefficients are calculated between the seed time series and all other voxels in the brain. Correlation coefficients are then transformed to Fisher's z-scores for normalization. The primary outcome is the strength of functional connectivity (Fisher's z) between the seed region and target brain regions.