Eight randomised trials suggest a favourable effect on immunity, especially in immunocompromised people — but biofeedback was always combined with other approaches. One separate randomised trial reports higher CD4+ counts after alpha training.
Updated :
What the research shows
EEG neurofeedback — standard amplitude training
Alpha-amplitude EEG neurofeedback — CD4+ counts in HIV-positive adults
Ages
Adult (18+)
Techniques
Alpha training
AAPB3Probably efficaciousNeuroLogic3Probably efficacioussame vs AAPB
Level 3 maintained. The AAPB's neurofeedback rating rests on a single trial: Schummer 2013 randomised 40 HIV-positive, AIDS-negative volunteers to neurofeedback, cranial electrotherapy, both, or a waitlist control (n = 10 per arm), gave the neurofeedback arms 20-minute twice-weekly sessions to raise eyes-closed alpha amplitude over 16 weeks, and found significant CD4+ increases in the neurofeedback and combined groups only, with the trained groups meeting their alpha amplitude target by the end of training. That is a randomised four-arm design with a documented learning check, but a single small study in one setting: Level 3, not Level 4. 2022-2026 window: Lubianiker 2026 randomised 85 healthy adults to fMRI neurofeedback upregulating reward mesolimbic activity, non-mesolimbic control-region upregulation, or no neurofeedback before hepatitis B vaccination; post-vaccination antibody levels did not differ between groups. That is a different technique in healthy volunteers, rated on its own row, and does not bear on the alpha-amplitude rating.
Other neurofeedback methods
fMRI neurofeedback — modulation of the immune response (adult)
Level 1. New row: AAPB does not rate fMRI neurofeedback for immune outcomes. Lubianiker 2026 is the only study — a preregistered, double-blind trial randomising 85 healthy adults to reward mesolimbic upregulation (n = 34), non-mesolimbic control-region upregulation (n = 34) or no neurofeedback (n = 17) before hepatitis B vaccination. Both neurofeedback groups increased activation in their target region, and ventral tegmental area upregulation correlated with the post-vaccination rise in hepatitis B antibody (r = 0.31, P = 0.018); but the prespecified group comparison of antibody change was null, and no adverse effects occurred. A solid mechanistic demonstration in healthy volunteers, with a null outcome and no replication, supports nothing above Level 1.
Biofeedback
EMG and thermal biofeedback, HRV, within combined interventions — immune and inflammatory markers (adult)
AAPB3Probably efficaciousNeuroLogic3Probably efficacioussame vs AAPB
Level 3 maintained. AAPB base: eight randomised trials in which biofeedback was one component of a relaxation package. Peavey 1985 (n = 16, EMG and thermal training) increased phagocytic capacity versus control; McGrady 1992 (n = 31 healthy adults, four sessions) increased T-lymphocyte blastogenesis and lowered white cell count versus a no-treatment group, with cortisol unchanged; Gruber 1993 (n = 13 post-mastectomy) increased natural killer cell activity, mixed lymphocyte responsiveness and peripheral blood lymphocytes versus delayed treatment; Taylor 1995 (n = 10 HIV-positive men) increased T-cell count, maintained at 1 month; Coen 1996 and Kern-Buell 2000 (mild asthma, n = 20 and n = 16) reduced asthma severity and medication use, with higher CD4+ and CD8+ counts in the first and a lower neutrophil percentage in the second; Birk 2000 (n = 31 with HIV) showed no group gains in CD4+, CD8+ or natural killer counts. Lehrer 2010 (n = 11) found less HRV reduction after an endotoxin injection but unchanged cytokines, and Nolan 2012 was inconclusive. 2022-2026 window: Herhaus 2023 randomised 55 people with panic disorder to four weeks of slow-paced breathing with HRV biofeedback or sham HRV biofeedback and reduced TNF-alpha (η² = 0.077), with no effect in the sham arm. Not Level 4: no trial isolates the biofeedback contribution.
In short
Clinical reading
AAPB Level 3; NeuroLogic Level 3 for biofeedback and for alpha neurofeedback alike. All eight biofeedback studies combined it with other components without isolating its contribution; several failed to demonstrate training success (lower EMG, higher hand temperature). The neurofeedback rating rests on Schummer 2013 alone (n = 40, alpha training, higher CD4+ counts). fMRI neurofeedback is rated separately at Level 1 on a single randomised trial with a null result.
Protocols
EMG- and temperature-assisted relaxation, resonance-frequency HRV; 6-10 session protocols within multimodal programmes. Neurofeedback: eyes-closed alpha amplitude, 20 minutes twice weekly over 16 weeks.
Limits
Clinical improvement (medication use, infections) is rarely measured; intermediate immunological markers dominate. The one sham-controlled result (Herhaus 2023) concerns a single cytokine in one setting; the fMRI trial was run in healthy volunteers with a null outcome. No paediatric data.
Study base
Eight RCTs plus two specific studies (HRV and endotoxin, neurofeedback and CD4+). 2022-2026 base: 2 publications indexed in the archive (1 HRV biofeedback, 1 fMRI neurofeedback).
Brendan's perspective
The same level as the AAPB on both rows, and I am comfortable there. Eight randomised trials, biofeedback bundled every time with relaxation, imagery or autogenic training, and several with no evidence that the training took at all — hand temperature did not rise in Coen 1996 or Kern-Buell 2000. When the manipulation check fails, the trial tells you about the package, not about the method. Schummer 2013 is better than its reputation: four arms, randomised, a documented alpha learning check, higher CD4+ counts in the trained groups. It is also still alone thirteen years later, with nothing replicating it. Lubianiker 2026 is a beautifully built fMRI study with a null primary outcome in healthy volunteers, which is science rather than a clinic option. In practice I approach immune physiology only through the front door — HRV biofeedback for autonomic regulation, sleep and stress — and I would not offer neurofeedback as training aimed at immune function. Nobody has yet shown a change in infections, medication use, or anything the patient would notice.
McGrady et al. (1992) The effects of biofeedback-assisted relaxation on cell-mediated immunity, cortisol, and white blood cell count in healthy adult subjects doi:10.1007/BF00844727
Lehrer et al. (2010) Voluntarily produced increases in heart rate variability modulate autonomic effects of endotoxin induced systemic inflammation: An exploratory study doi:10.1007/s10484-010-9139-5
Schummer et al. (2013) The effect of neurofeedback and cranial electrotherapy on immune function within a group of HIV+ subjects: A controlled study doi:10.1080/10874208.2013.813168
Herhaus et al. (2023) Effect of a slow-paced breathing with heart rate variability biofeedback intervention on pro-inflammatory cytokines in individuals with panic disorder - A randomized controlled trial doi:10.1016/j.jad.2023.01.091
Lubianiker et al. (2026) Upregulation of reward mesolimbic activity and immune response to vaccination: a randomized controlled trial doi:10.1038/s41591-025-04140-5
