Abdominal pain in children is the best-studied: two small trials and several series show a clear fall in pain with heart-rate-variability biofeedback or assisted relaxation. For other pain (joint, chest, sickle cell, mixed chronic pain, medical procedures), each site has only one or two studies, but the recent randomised trials are mostly positive. Trials remain small. All of this concerns biofeedback. Neurofeedback has never been tested for these pains in children, so it carries the lowest level here — which means untested rather than shown not to work.
Updated :
What the research shows
EEG neurofeedback — standard amplitude training
Chronic and recurrent pain other than headache in children and adolescents
Level 1 (the AAPB rates biofeedback for paediatric pain other than headache but does not rate neurofeedback for it at all; this row is entirely NeuroLogic's). No controlled trial of EEG amplitude training exists for chronic or recurrent pain in children or adolescents outside headache. Level 1 records that absence and nothing more — it is not a finding that the method fails here. The adult chronic-pain row sits at 3 on Hesam-Shariati 2022, whose post-hoc analysis of conventional single-region protocols is the part that carries the level, and there is no physiological argument for alpha and SMR training working in adults and not below eighteen. That is a reason to run the studies, not a substitute for them: extrapolation from adults is not evidence. This is one of the indications where the gap between reported clinical practice and published work is widest, and published case series would be enough to begin closing it.
Biofeedback
(HRV, multimodal relaxation, anorectal EMG) — abdominal and defecation pain in children and adolescents
AAPB3Probably efficaciousNeuroLogic3Probably efficacioussame vs AAPB
Level 3 maintained. Schurman 2010: randomised pilot, n = 20 (ages 8-17, functional dyspepsia), 10 sessions of multimodal biofeedback-assisted relaxation added to standard medical care — pain intensity and duration significantly reduced vs standard care alone. Sowder 2010: manualised HRV biofeedback over six sessions in 20 children with functional abdominal pain (10 matched healthy controls, not randomised) — reduced pain frequency (eta² = 0.67) and intensity (eta² = 0.73), 75 % improved by more than 50 %. Stern 2014: clinical series, 63.6 % full remission. Humphreys 2000 (AAPB, paediatric pain): n = 47, three treatment arms superior to fibre alone. For defecation pain, Cox 1998 (n = 87, encopresis) and Ladi-Seyedian 2022 (n = 40, electrical stimulation with or without EMG): pain reduced in every arm with no between-group difference. 2022-2026 window: nothing specific to abdominal pain; the pelvic-floor physiotherapy meta-analyses for paediatric functional constipation (Hao 2025, 13 RCTs, n = 997: biofeedback RR 0.91, 95 % CI 0.50-1.68 vs control; Medrano-Sánchez 2026, 7 RCTs, n = 775, painful defecation improved by pooled physiotherapy) do not support anorectal biofeedback on its own. One population per trial and samples of 20: not Level 4.
AAPB2Possibly efficaciousNeuroLogic3Probably efficacioushigher vs AAPB
Level 3. AAPB rates 2 because each pain site rests on a single study; taken together, randomised trials of clinician-delivered biofeedback are now several and in independent groups. Eid 2016: n = 36 (ages 8-13, juvenile arthritis), EMG biofeedback three times weekly for 12 weeks added to physiotherapy — greater and faster pain reduction than physiotherapy alone. Yetwin 2022: n = 21 (ages 10-17, chronic pain), brief HRV biofeedback with home breathing practice vs a 4-week waitlist — reduced pain intensity, better school functioning, increased BVP amplitude, while waitlist pain rose. Smith 1989 (n = 10, chest pain): less pain at 6 months than attention placebo. Procedural pain: Han 2025 (n = 80 dyads, leukaemia, four-unit biofeedback plus education vs education alone; pain F = 6.06, p = 0.015, and fear reduced at 4 weeks) and Savaş 2024 (n = 62, respiratory-sensor-driven VR game, pain reduced, p < 0.001), but Ostojic 2022 (n = 38, crossover) finds no difference between app-based assisted relaxation and distraction. Sickle cell: two studies reduce pain frequency or intensity and analgesic use (van Veelen 2023). Populations and protocols too heterogeneous, samples too small, for Level 4.
In short
Clinical reading
AAPB Levels 3 (abdominal pain) and 2 (other pain); NeuroLogic Levels 3 and 3. A NeuroLogic neurofeedback row is added at Level 1: the AAPB does not rate neurofeedback for paediatric pain other than headache, and no controlled trial of amplitude training exists in this population — the level records that absence, not a demonstrated failure, and the adult chronic-pain row sits at 3. Abdominal pain: Schurman 2010 (randomised vs standard care) and Sowder 2010 (HRV, very large effects) hold the 3. Other pain: the 3 rests on randomised trials of clinician-delivered biofeedback in independent groups (Eid 2016, juvenile arthritis; Yetwin 2022, chronic pain; Han 2025, procedural pain), each in a different population. The paediatric literature predominantly targets abdominal and defecation pain.
Protocols
Resonance-frequency HRV (manualised protocol, 6-10 sessions), multimodal assisted relaxation (bifrontal EMG, temperature, EDA, respiration), muscle EMG for joint pain; short age-adapted sessions, daily home practice, parental involvement.
Limits
Few studies per pain site, samples mostly of 10 to 40, outcomes often secondary, unblinded ratings, no long-term follow-up; the recent procedural-pain studies rest on single-session apps or games; biofeedback is often one component of a multimodal programme whose specific contribution is not isolated.
Study base
Modest literature, concentrated on abdominal pain; paediatric HRV systematic review (Dormal 2021). 2022-2026 base: 9 publications indexed in the archive (6 trials, 3 reviews), none specific to abdominal pain.
Brendan's perspective
We move the non-headache row from the AAPB's 2 to a 3. AAPB rated each pain site on its own single study; read together, Eid 2016 in juvenile arthritis, Yetwin 2022 in mixed chronic pain and Han 2025 in procedural pain are randomised, clinician-delivered and from independent groups. That is what a 3 describes. Abdominal pain stays at 3: Schurman 2010 and Sowder 2010 report large effects on twenty children each, which is encouraging arithmetic and thin evidence. Note what I am not counting — the single-session app and VR-game studies are a different intervention wearing the same word. In clinic this is resonance-frequency HRV work, six to ten short sessions, parents involved, daily practice, because breathing at resonance is a skill the child keeps. And a raised level is still a 3: samples of twenty to forty, unblinded ratings, no long follow-up. There is now a neurofeedback row here too, and it sits at 1. The AAPB does not rate neurofeedback for paediatric pain other than headache at all, so this one is entirely ours, and a 1 is the accurate reading: not a single controlled trial of amplitude training exists for chronic or recurrent pain in children and adolescents outside headache. I want to be clear about what that 1 means, because it will be misread. It records an absence of evidence, not a demonstration of failure. The adult chronic pain row is at 3 on evidence I find substantive, and there is no physiological argument I know of for why alpha and SMR training should work at nineteen and not at fifteen. It stands to reason that it transfers. Reason is not evidence, which is why the row is a 1 and not a 2 — but this is one of the places the field most needs to do the work. I know a good number of clinicians getting results with these children. What we do not have is a literature, and we will not get to controlled trials without going through published case series first. Write them up.
Dormal et al. (2021) Is heart rate variability biofeedback useful in children and adolescents? A systematic review doi:10.1111/jcpp.13463
Schurman et al. (2010) A pilot study to assess the efficacy of biofeedback-assisted relaxation training as an adjunct treatment for pediatric functional dyspepsia associated with duodenal eosinophilia doi:10.1093/jpepsy/jsq010
Sowder et al. (2010) Restoration of vagal tone: A possible mechanism for functional abdominal pain doi:10.1007/s10484-010-9128-8
Eid, Aly & El-Shamy (2016) Effect of electromyographic biofeedback training on pain, quadriceps muscle strength, and functional ability in juvenile rheumatoid arthritis doi:10.1097/PHM.0000000000000524
Yetwin et al. (2022) Heart Rate Variability biofeedback therapy for children and adolescents with chronic pain: A pilot study doi:10.1016/j.pedn.2022.06.008
Darling, Benore & Webster (2020) Biofeedback in pediatric populations: A systematic review and meta-analysis of treatment outcomes doi:10.1093/tbm/ibz124
Hesam-Shariati et al. (2022) The analgesic effect of electroencephalographic neurofeedback for people with chronic pain: A systematic review and meta-analysis doi:10.1111/ene.15189
