Neurology and neurodevelopment · Not in the 2023 AAPB book
Healthy aging
Healthy cognitive aging (adults aged 60 and over)
In healthy older adults, neurofeedback aims to maintain memory and attention, not to treat a disease. Around fifteen small trials, pooled in a recent meta-analysis, show moderate gains in working and episodic memory when training is long enough; but the most rigorous trials, in which the brain learns to modulate its signal well, do not always find a benefit on the tests. Optical (fNIRS) neurofeedback embedded in app games and heart-rate-variability biofeedback remain at a preliminary stage.
Alpha training, Sensorimotor rhythm (SMR), Theta/beta ratio
AAPB0Not ratedNeuroLogic3Probably efficacious
Level 3. Lin 2024 meta-analysis: 14 clinical trials (n = 284) of EEG neurofeedback in healthy elderly or MCI — working memory g = 0.665 (95 % CI 0.473-0.858) and episodic memory g = 0.595 (0.333-0.856), significant in the healthy subgroup (working memory g = 0.495, episodic memory g = 0.729) and only when total training exceeded 300 minutes. Randomised superiority in small samples: Alatorre-Cruz 2022 (20 seniors with excess theta, contingent theta reduction vs random reward — global EEG reorganisation only in the trained group, executive-function gains significant at one year); Rogala 2026 (57 adults aged 41-64, DNN-personalised task-oriented neurofeedback vs sham, 10-11 sessions, training group superior on all reasoning conditions at post-test, each p < .03 — a middle-aged sample). Kim 2024 (quasi-experimental, n = 59, 16 alpha sessions vs routine daily life in presbycusis): left-frontal alpha, MMSE-K, digit span forward and speech recognition improved vs control. Against this, the best-controlled recent trials show learning without cognitive gain (Andrade 2024, double-blind sham, gamma synchrony self-modulated, no neuropsychological change) or no frequency-specific learning at all (Paban 2024, subjective cognitive decline). Multiple controlled studies with moderate pooled effects, but no adequately powered randomised superiority replicated in independent settings.
Other neurofeedback methods
HEG / fNIRS — cognitive training with prefrontal optical neurofeedback
Ages
Adult (18+)
Techniques
HEG / fNIRS
AAPB0Not ratedNeuroLogic2Possibly efficacious
Level 2. Acevedo 2022: 86 healthy adults aged 54-84 (mean 66) randomised to four weeks of app-based cognitive training with fNIRS neurofeedback (ABC games) or Tetris; the treatment group improved significantly pre/post on memory, verbal memory and composite cognition while the control group did not, and both groups improved on processing speed and executive function (within-group tests; the abstract reports no between-group contrast). Acevedo 2023 (same trial): higher sensory-processing sensitivity predicted larger memory gains in the treatment arm. Matsuzaki 2023: multilevel meta-analysis of three NIRS cognitive-training-with-neurofeedback studies (n = 166, mean ages 21-66) — benefit on episodic, long-term and working memory. Tinello 2023: 34 healthy older adults, ten weeks of HRV biofeedback plus nirHEG neurofeedback vs the same protocol without real-time feedback — gain on one of two interference tasks, none on response inhibition. One adequately powered randomised trial with an active comparator, a small feedback-vs-no-feedback study and a three-study meta-analysis, all on consumer-grade optical sensors embedded in games: enough for possibly efficacious, not for probably efficacious.
Level 2. Lin 2024: real-time fMRI neurofeedback training of bilateral dorsal anterior cingulate cortex during a selective-attention task in 27 older and 19 young adults. Older participants in the up-regulation condition showed increasing reward points and dACC BOLD signal across sessions, with faster reaction times and better accuracy; these effects were absent in the older down-regulation group (inverse-condition control) and, unexpectedly, in young up-regulators. A single proof-of-concept study with small groups and an outcome measured within the trained task rather than a transfer measure.
Level 1. Dobrushina 2022: uncontrolled pilot, nine women aged 52 ± 7 with subclinical signs of emotional dysregulation, 15 sessions of infra-low frequency neurofeedback; resting-state fMRI showed increased connectivity in an extended temporal-occipital-amygdala network, with decreases in alexithymia, depression and anxiety scores. No control group, no cognitive outcome, no replication.
EEG brain-computer-interface (BCI) cognitive training — attention or motor-imagery classifier feedback
Ages
Adult (18+)
Techniques
Other neurofeedback
AAPB0Not ratedNeuroLogic2Possibly efficacious
Level 2. Tsai 2025: PRISMA review of 16 EEG brain-computer-interface / neurofeedback trials in healthy older adults or MCI — promising cognitive-enhancement results with research-grade systems, but high risk of bias from randomisation and outcome measurement. The two best-controlled recent trials are null on cognition: Marcos-Martínez 2026 (92 healthy adults aged 65-75, double-blind, motor-imagery neurofeedback vs placebo feedback vs classical cognitive stimulation, ten sessions — no training-specific cognitive improvement despite good motor-imagery accuracy in many trainees, with responders showing distinct spectral and network modulation); Qian 2025 (resting-state fMRI sub-study of an EEG-BCI attention-driven cognitive-training trial vs waitlist — no significant behavioural improvement, but increased functional-network segregation that correlated with language and everyday-memory change). The positive uncontrolled and waitlist-controlled trials in the review support possibly efficacious; the controlled evidence does not go beyond that.
Biofeedback
HRV biofeedback (resonance-frequency breathing) — autonomic regulation, mood, sleep, cognition and biomarkers in healthy older adults
Ages
Adult (18+)
Techniques
HRV — resonance-frequency breathing
AAPB0Not ratedNeuroLogic2Possibly efficacious
Level 2. Allen 2026: 53 healthy older adults randomised to five weeks of HRV biofeedback or sham — no group difference in resting HRV, emotional functioning, sleep or cognitive performance (both arms improved sleep quality and HF, RMSSD and SDNN); only low-frequency power during recovery from a laboratory stressor rose more with biofeedback. Min 2023: 108 healthy younger and older adults randomised to four weeks of daily slow-paced breathing with HRV biofeedback to increase heart-rate oscillations (Osc+) or to decrease them (Osc−) — large between-condition differences in change in plasma Aβ40 and Aβ42, decreasing under Osc+, with opposing effects on pTau-181 in older adults; a surrogate biomarker outcome in a healthy sample. Tinello 2023 (combined HRV biofeedback and nirHEG, n = 34): HRV rose only in participants with the lowest baseline HRV. One adequately powered randomised trial with a mechanistic outcome and one sham-controlled trial null on clinical outcomes: possibly efficacious for physiological targets, unproven for mood, sleep or cognition in this age group.
In short
Clinical reading
Indication outside the AAPB 2023 book (no AAPB level). NeuroLogic Level 3 for amplitude neurofeedback in older adults: Lin 2024 meta-analysis (14 trials, n = 284, working memory g = 0.67, episodic memory g = 0.60, benefit above 300 minutes of training), randomised superiority in small samples (Alatorre-Cruz 2022 at one year; Rogala 2026, ages 41-64), but double-blind trials null on cognition (Andrade 2024, Marcos-Martínez 2026). fNIRS/HEG 2 (Acevedo 2022, Matsuzaki 2023, consumer sensors), fMRI 2 (Lin 2024), BCI 2, ILF 1; HRV biofeedback 2 (Allen 2026 null vs sham; Min 2023 on plasma amyloid).
Protocols
Alpha or individual peak-alpha enhancement, theta reduction in subjects with excess theta, SMR or gamma synchrony, at frontal or central sites; 16-24 supervised sessions, more than 300 minutes of total training; resonance-frequency HRV with daily home practice.
Limits
Small samples (n = 20 to 92), ceiling effects in cognitively intact subjects, EEG learning shown without cognitive gain in the double-blind trials; a single one-year follow-up; fNIRS neurofeedback rests on consumer sensors embedded in games; no paediatric literature by definition; trials in patients with mild cognitive impairment belong to the neurodegenerative indication.
Study base
One meta-analysis of 14 trials (2024) and one BCI systematic review of 16 trials (2025); 2022-2026 base: 17 publications indexed in the archive (15 neurofeedback, 2 HRV biofeedback).
Brendan's perspective
There is no AAPB level here, so the 3 is ours. It rests on Lin 2024 — working memory g = 0.67, episodic memory g = 0.60 — with one detail that matters more than the effect sizes: the benefit appears only above 300 minutes of training, roughly the 20 to 30 session course a real clinic delivers. Alatorre-Cruz 2022 is the trial I like best: seniors selected for excess theta and trained on it, executive gains still there at one year. Older than either, and worth not losing: Angelakis 2007 trained peak alpha frequency in the elderly under a double-blind controlled design and reported gains in processing speed and executive function, with no clear effect on memory. Peak alpha frequency correlates with cognitive performance and declines with age, so training it is a protocol that follows a physiological marker rather than a menu — the same logic as Alatorre-Cruz, arrived at fifteen years earlier. It is a small pilot and I do not lean on it, but it is part of why this row is a 3 and not a 2, and it predates the AAPB window, which is why the book has nothing to say here. Protocol mirroring physiology, not a menu. Against it, Andrade 2024 shows people learning to modulate gamma synchrony with no cognitive change — the cleanest reminder that the trained metric and the clinical outcome are two different columns. I would train ageing adults, qEEG first, with transfer tasks that look like their actual week. What I would not do is sell it as prevention: nothing here shows a changed trajectory, ceiling effects in intact people are real, and one twelve-month follow-up is not a maintenance literature.
Lin et al. (2024) Effectiveness of Electroencephalography Neurofeedback for Improving Working Memory and Episodic Memory in the Elderly: A Meta-Analysis doi:10.3390/medicina60030369
Alatorre-Cruz et al. (2022) One-Year Follow-Up of Healthy Older Adults with Electroencephalographic Risk for Neurocognitive Disorder After Neurofeedback Training doi:10.3233/JAD-215538
Rogala et al. (2026) Cognitive and brain function enhancement in Gen X group after personalized, AI supervised EEG-neurofeedback training doi:10.1088/1741-2552/ae8577
Marcos-Martínez et al. (2026) Motor imagery-based neurofeedback in older adults: neural signatures and feasibility in a randomized controlled trial targeting age-related cognitive decline doi:10.1186/s12984-026-01912-z
Andrade et al. (2024) Self-Modulation of Gamma-Band Synchronization through EEG-Neurofeedback Training in the Elderly doi:10.31083/j.jin2303067
Acevedo et al. (2022) Cognitive Training with Neurofeedback Using fNIRS Improves Cognitive Function in Older Adults doi:10.3390/ijerph19095531
Matsuzaki et al. (2023) The Effect of Cognitive Training with Neurofeedback on Cognitive Function in Healthy Adults: A Systematic Review and Meta-Analysis doi:10.3390/healthcare11060843
Tsai et al. (2025) Brain computer interfaces for cognitive enhancement in older people - challenges and applications: a systematic review doi:10.1186/s12877-025-05676-4
Allen et al. (2026) A Randomized Sham-Controlled Study of Heart Rate Variability Biofeedback in an Older Adult Sample doi:10.1007/s10484-026-09803-8
Min et al. (2023) Modulating heart rate oscillation affects plasma amyloid beta and tau levels in younger and older adults doi:10.1038/s41598-023-30167-0
Angelakis et al. 2007 (2007) EEG neurofeedback: a brief overview and an example of peak alpha frequency training for cognitive enhancement in the elderly doi:10.1080/13854040600744839
