Findings are preliminary but promising: a case series replicated by a randomised trial suggests better control of syncopal episodes. Both studies come from the same group.
Updated :
What the research shows
Biofeedback
Assisted relaxation (facial EMG and finger temperature), diaphragmatic breathing and stress management
AAPB2Possibly efficaciousNeuroLogic2Possibly efficacioussame vs AAPB
Level 2 maintained. The AAPB rating rests on two studies from a single group. McGrady 2003 randomised 22 patients with vasovagal syncope to ten 50-minute sessions of biofeedback-assisted relaxation (facial EMG and finger temperature feedback, autogenic training, progressive relaxation, stress management, twice-daily taped home practice) or a waitlist: the trained group lowered EMG and raised hand temperature while controls did not, and reduced loss-of-consciousness episodes from pre-test to post-test; headache pain did not improve significantly relative to control (p = .05) and both groups improved on state anxiety and depression. The trial replicated McGrady 1997, a series of 10 consecutive patients trained over an average of 8.5 sessions, in which six of seven patients diagnosed with syncope reported at least a 50 % reduction in episodes. 2022-2026 window: Han 2025 and Kim 2024 report the same non-randomised controlled pilot (n = 27 versus 24) of a 12-week HRV-biofeedback-based programme in adults aged 65 and over with orthostatic hypotension — a related but distinct diagnosis — with reduced orthostatic symptom severity (P < .001) and smaller systolic (P = .003) and diastolic (P = .012) postural blood pressure drops. One small waitlist-controlled pilot trial replicating a case series from the same group supports Level 2 and no more.
In short
Clinical reading
AAPB Level 2; NeuroLogic Level 2. Replication of a 10-patient case series (McGrady 1997) by a small waitlist-controlled pilot trial (McGrady 2003, n = 22) justifies this level, without supporting anything higher: both studies come from the same group and the outcomes are self-reported. The trained group did, however, demonstrate learning in both channels (lower facial EMG, higher finger temperature), which the control group did not.
Protocols
Diaphragmatic breathing, relaxation assisted by facial EMG and finger temperature, autogenic training and progressive relaxation, stress management; training to recognise prodromal signs and to contract muscles when syncope threatens.
Limits
A single small RCT, no independent replication, self-reported outcome measures, and no follow-up beyond post-intervention. The evidence is in adults; nothing is specific to adolescence. The 2022-2026 publications are non-randomised and concern orthostatic hypotension in older adults, not vasovagal syncope.
Study base
One RCT and a 10-patient case series, both from the same group. 2022-2026 base: 2 publications indexed in the archive, both reporting one non-randomised pilot in orthostatic hypotension.
Brendan's perspective
Level 2, unchanged, and I have no argument with that. Two studies, one group: McGrady 1997 replicated by a waitlist-controlled pilot of 22, in which the subgroup with frequent loss of consciousness was three treated patients and four controls. The syncope counts were self-reported and unblinded. So the number is small, and it should be. What makes this a level 2 worth knowing about rather than one to discard is that the trained group actually learned — facial EMG down, finger temperature up, in both channels, while the controls did not. That is more than most level 2s can show. The 2022-2026 papers do not help: two reports of one non-randomised pilot, in orthostatic hypotension, which is a different diagnosis. I do not see this presentation, and the first step is cardiological rather than behavioural anyway. Once that is settled, what is on offer is an autonomic skill and prodrome recognition, not a demonstrated reduction in syncope. Say it that way.
Han et al. (2025) The Effect of a Biofeedback-Based Integrated Program on Improving Orthostatic Hypotension in Community-Dwelling Older Adults: A Pilot Study doi:10.1097/JCN.0000000000001026
