Real-time feedback on breathing often helps patients breathe more efficiently and tolerate exercise better. COPD is not curable; the aim is comfort and functional capacity. The feedback itself seems to be what matters: when one trial merely paced breathing without feeding anything back, the benefit disappeared.
Updated :
What the research shows
Biofeedback
Real-time feedback of a breathing variable (ventilation, capnometry, incentive spirometry, flutter device)
Ages
Adult (18+)
Techniques
Respiratory / capnometry, Other biofeedback
AAPB3Probably efficaciousNeuroLogic4Efficacioushigher vs AAPB
Level 4 (the AAPB Level 3 pools every kind of biofeedback tried in COPD; the AAPB's stated reason for not going higher is that the modalities are too diverse to be equivalent). Narrowed to real-time feedback of a breathing variable, the result replicates across independent settings. Collins 2008: 64 participants randomised to 36 sessions of treadmill and cycle training with or without visual feedback of inhalation and exhalation times — greater breathing efficiency, more exercise tolerance, less hyperinflation and longer exhalation with feedback; the same group's later metronome trial, which paced breathing without feeding anything back, found no difference (Collins 2019). Estève 1996: n = 20, 30-35 sessions matching breath traces to a screen template — FEV1 up 22 % and FVC up 19 %, with no significant change in controls. Kaja 2020: n = 168 randomised to a flutter device with visual, auditory, both or no feedback — better oxygen saturation, dyspnoea scores and secretion clearance with feedback. Rashid 2025: double-blind randomised trial, n = 70, incentive-spirometry visual feedback added to home-based rehabilitation — better FEV1, FVC and St George's scores at four weeks. Norweg 2025 (n = 42, capnography biofeedback vs usual care) improved respiratory symptoms, activity avoidance, SpO2, end-tidal CO2 and interoception. Two trials against active rehabilitation were null (van Gestel 2012, n = 40; Lagravinese 2026, n = 30).
Heart rate variability biofeedback as an adjunct to standard care
Ages
Adult (18+)
Techniques
HRV — resonance-frequency breathing
AAPB3Probably efficaciousNeuroLogic3Probably efficacioussame vs AAPB
Level 3 (a disaggregation of the pooled AAPB rating for biofeedback in COPD). Wu 2025 and Yang 2025 report the same cohort of 53 patients allocated to six weekly one-hour HRV biofeedback sessions plus standard medical care (n = 26) or standard care alone (n = 27). Group by time interactions favoured biofeedback: six-minute walk distance increased and mMRC dyspnoea decreased relative to control, the BODE index fell and HRV indices rose, and COPD self-efficacy and St George's Respiratory Questionnaire quality of life improved; HRV reactivity during the walk test fell and recovery improved. Depression and anxiety did not differ between groups. Giardino 2004, the study AAPB reviews, was uncontrolled: 20 patients given five HRV biofeedback sessions and four weeks of walking with pulse-oximetry feedback improved six-minute walk distance, St George's quality of life, dyspnoea, disability and resting HRV. One controlled trial whose allocation method is not described, plus an uncontrolled series, is Level 3; a randomised trial against an attention-matched comparator would be needed for Level 4.
Physical-activity (pedometer) feedback and informational feedback on inhaler adherence
Ages
Adult (18+)
Techniques
Other biofeedback
AAPB3Probably efficaciousNeuroLogic3Probably efficacioussame vs AAPB
Level 3 (a disaggregation of the pooled AAPB rating). Feedback aimed at behaviour rather than at a breathing signal has one clear positive trial and two null ones. O'Dwyer 2020: 153 participants with obstructive airway disease randomised to standard inhaler instruction including a single demonstration of the device data, to live biofeedback of the recorded data delivered in person by a pharmacist, or to inhaler use as usual; at six months the biofeedback group had the highest adherence, the greatest change on the St George's Respiratory Questionnaire and significantly improved daily respiratory symptoms. Wootton 2019: twelve months of pedometer feedback with telephone calls and progressive goal setting against usual care with no feedback, activity measured by armband — no difference in daily physical activity, and no COPD symptom difference was reported; the same group's earlier trial of the same design found no difference in health-related quality of life at two months. Held at the pooled AAPB level: one adequately sized positive randomised trial with a device-recorded primary outcome, set against two null activity-feedback trials, supports Level 3 and no more.
In short
Clinical reading
The AAPB gives a single Level 3 for all biofeedback in COPD, on the ground that such diverse modalities cannot be treated as equivalent. NeuroLogic separates three rows: real-time feedback of a breathing variable Level 4 (Collins 2008; Estève 1996; Kaja 2020, n = 168; Rashid 2025, double-blind; Norweg 2025), HRV biofeedback Level 3 (Wu 2025 and Yang 2025, same cohort of 53; Giardino 2004), activity and adherence feedback Level 3 (O'Dwyer 2020 positive; Wootton 2019 null).
Protocols
Ventilation feedback during exercise, slow-breathing training, end-tidal CO2 capnography, incentive spirometry, instrumented flutter devices, accessory-muscle EMG; six weekly hour-long sessions for HRV; embedded in pulmonary rehabilitation.
Limits
Major heterogeneity of techniques and no data on whether habitual breathing-pattern changes persist; recent trials remain small and often designed as feasibility studies; two trials against active rehabilitation were null on dyspnoea and six-minute walk distance; the HRV trials have no attention-matched control; there are no paediatric data, COPD being an adult disease.
Study base
Seven controlled studies in the AAPB base, with non-comparable modalities; 2022-2026 base: 5 publications indexed in the archive (3 respiratory biofeedback, 2 HRV), with the sweep additionally returning the Rashid 2025 double-blind trial and the de Souto Barbosa 2023 uncontrolled pilot.
Brendan's perspective
Splitting the AAPB's pooled 3 was right, and one row earns a 4. Real-time feedback of a breathing variable replicates across independent settings, and the cleanest argument for it is a null: when Collins' own group paced breathing with a metronome in 2019 and fed nothing back, the benefit disappeared. Feedback is not wallpaper, it is the learning signal. HRV stays at 3: Wu 2025 and Yang 2025 are one cohort of 53 reported twice, against standard care with no attention-matched control. Two trials against active rehabilitation were null, a fair measure of what the feedback adds and what it does not. COPD is not a condition I treat; it belongs in pulmonary rehabilitation, and biofeedback is a layer inside that programme. What I tell a patient: this aims at breathing efficiency and what you can do in a day, not at the disease. And no trial has yet shown the trained breathing pattern survives the end of the programme.
Collins et al. (2008) Can ventilation-feedback training augment exercise tolerance in patients with chronic obstructive pulmonary disease? doi:10.1164/rccm.200703-477OC
Estève et al. (1996) The effects of breathing pattern training on ventilatory function in patients with COPD doi:10.1007/BF02214431
Kaja et al. (2020) Biofeedback flutter device for managing the symptoms of patients with COPD doi:10.3233/THC-202222
O'Dwyer et al. (2020) Personalized biofeedback on inhaler adherence and technique by community pharmacists: a cluster randomized clinical trial doi:10.1016/j.jaip.2019.09.008
Wu et al. (2025) Effects of Heart Rate Variability (HRV) Biofeedback in Pulmonary Indicators and HRV Indices Among Patients with Chronic Obstructive Pulmonary Disease doi:10.1007/s10484-024-09664-z
Norweg et al. (2025) Mind the Breath: Feasibility of Capnography-Assisted Learned Monitored (CALM) Breathing for Dyspnea Treatment doi:10.1097/HCR.0000000000000939
Lagravinese et al. (2026) Respiratory Biofeedback Training as an Adjunct Intervention in Pulmonary Rehabilitation for Late-Stage COPD: A Pilot Trial doi:10.1007/s10484-025-09763-5
