Neurofeedback in psychiatry: A decade of clinical and neuroimaging insights – a systematic review

Background: Neurofeedback (NF) provides real-time feedback on brain activity to enhance self-regulation and has been studied across psychiatric disorders. This systematic review (2015–2025) assessed NF efficacy in attention-deficit/hyperactivity disorder (ADHD), mood and anxiety disorders, post-traumatic stress disorder (PTSD), schizophrenia, and substance use disorders; compared electroencephalography (EEG) - and functional magnetic resonance imaging (fMRI)-based protocols; examined mechanistic and neuroimaging correlates; and evaluated methodological quality and risk of bias (RoB). Methods: Following Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines, PubMed, PsycINFO, Web of Science, and Scopus were searched to February 2025. Eligible studies included randomized and non-randomized trials, open-label designs, and case series comparing NF with sham, active, or waitlist controls, or using pre–post designs. Two reviewers independently screened and extracted data. RoB was assessed with Cochrane tools; adherence to Consensus on the Reporting and Experimental Design-NF guidelines was considered. Main outcomes were symptom change, neuroimaging/electrophysiological correlates, and adverse events. Results: Forty-five studies ( n ≈ 4,600) were reviewed: 28 randomized controlled trials and 17 non-randomized or open-label designs. In ADHD, several trials showed symptom improvements, though large sham-controlled studies gave mixed results. Depression studies using EEG frontal asymmetry or fMRI amygdala regulation reported significant reductions with neural target engagement. Anxiety studies showed decreases in trait anxiety and insomnia, while obsessive-compulsive disorder evidence was limited. PTSD trials consistently demonstrated meaningful benefits, with some patients losing their diagnosis. In schizophrenia, NF enhanced antipsychotic effects. Substance use studies suggested reduced craving, especially with EEG alpha-theta. NF was safe and well-tolerated. Conclusion: NF shows promise, particularly in ADHD and PTSD, with mechanistic support, but heterogeneous protocols and blinding challenges demand larger, rigorous sham-controlled RCTs to confirm efficacy.