This study is a confirmatory, double-blind, placebo-controlled randomized clinical trial (RCT) testing whether gamma EEG neurofeedback (EEG-NFB) improves working memory in adults with schizophrenia or schizoaffective disorder. Participants are randomly assigned to receive either active gamma EEG-NFB (real-time feedback of frontal gamma brain activity) or sham EEG-NFB (false pre-recorded feedback), twice weekly for 12 weeks. Working memory (N-back task), brain gamma coherence, and everyday community functioning are assessed at baseline, mid-treatment, end of treatment, and follow-up.

Schizophrenia is associated with deficits in gamma frequency (30-45 Hz) neural synchrony and working memory that are resistant to antipsychotic treatment. EEG neurofeedback (EEG-NFB) targeting frontal gamma coherence is a non-invasive approach that may restore gamma synchrony and improve working memory by reinforcing endogenous neural oscillations in real time.

This confirmatory RCT builds on a successful pilot trial (R33) demonstrating feasibility and preliminary efficacy of gamma EEG-NFB in schizophrenia. Participants (N=104) are randomized 1:1 to active gamma EEG-NFB or sham EEG-NFB. Active EEG-NFB delivers real-time feedback of frontal gamma coherence (F3/F4, 30-50 Hz) via a CGX wireless headset. Sham participants undergo identical procedures but receive pre-recorded feedback unrelated to their own brain activity. Both groups complete 30 sessions over 12 weeks (two sessions per week).

Co-primary outcomes are working memory (N-back d-prime) and frontal gamma coherence, assessed at baseline and Week 12. Secondary outcomes include global cognition, symptoms (PANSS), negative symptoms (CAINS), and community functioning (ILSS, UPSA-Brief, SFS). Follow-up assessments are conducted at 3 months (Week 24) and 6 months (Week 36) post-treatment. Primary analysis uses mixed model repeated measures (MMRM).

Inclusion Criteria:

• 1. DSM-5 diagnosis of schizophrenia or schizoaffective disorder, confirmed by the Structured Clinical Interview for DSM-5 (SCID-5) 2. Age 18-55 years 3. Clinically stable: no psychiatric hospitalization in the 3 months prior to enrollment 4. No change in antipsychotic medication type or dosage within 4 weeks prior to baseline assessment 5. Ascertained to be clinically and medically stable by a study investigator 6. Does not meet DSM-5 diagnostic criteria for bipolar disorder or current major depressive episode 7. No electroconvulsive therapy (ECT) within 6 months of baseline assessment 8. Able to read and speak English (corrected vision or hearing aids acceptable) 9. Able and willing to provide written informed consent

Exclusion Criteria:

• 1. Self-reported history of seizure disorder 2. Diagnosed with multiple sclerosis 3. History of stroke or major vascular disease, including insulin-dependent diabetes mellitus 4. HIV/AIDS diagnosis 5. Current (not past) major depressive episode 6. Substance use disorder other than nicotine use disorder or caffeine use disorder in the past year 7. Brain cancer (primary or metastatic) 8. Prior head injury involving loss of consciousness 9. Inability to read or speak English 10. Color blindness that interferes with administration of study assessments 11. Neuropsychological or cognitive testing in the past 6 months using the same measures as this study 12. Score on Letter-Number Sequencing greater than 1 standard deviation above the age-normed mean

De-identified individual participant data will be shared via the NIMH Data Archive (NDA) in accordance with the NIH Data Management and Sharing Plan submitted with R01MH139831. Data will be made available no later than the time of primary results publication.

Within 12 months of primary results publication (estimated 2031)

De-identified individual participant data and supporting documents will be available to qualified researchers who submit a data access request through the NIMH Data Archive (NDA) at https://nda.nih.gov. Requestors must complete a Data Use Certification (DUC) agreeing to conditions of use including protecting participant confidentiality, not attempting to re-identify participants, and not sharing data outside the approved research team. Access is reviewed and approved by the NDA Data Access Committee.