Respiratory Muscle Training and endurance performance in healthy individuals: systematic review (46 studies) and meta-analysis (8 controlled trials)
Methods (abstract)
Two RMT types: respiratory muscle strength training (resistive/threshold, inspiratory or combined inspiratory/expiratory) and respiratory muscle endurance training (hyperpnoea).
Search in MEDLINE, EMBASE and CINAHL without language restriction; studies in healthy individuals with endurance performance measured by maximal tests (constant load, time trial, intermittent incremental, conventional incremental).
Multiple linear regression over 46 original studies for moderators (fitness, RMT type, test type, test duration, sport: rowing, running, swimming, cycling); meta-analysis on the 8 controlled trials with usable data.
Key Findings
RMT improves endurance exercise performance in healthy individuals (meta-analysis of 8 controlled trials: significant improvement).
Less fit subjects benefit more than highly trained athletes: +6.0% additional improvement per 10 mL·kg⁻¹·min⁻¹ lower VO2max (95% CI 1.8-10.2%; p = 0.005).
Inspiratory muscle strength training and respiratory muscle endurance training do not differ in effect (p = 0.208).
Combined inspiratory + expiratory strength training seems superior: +12.8% vs inspiratory strength training alone (95% CI 3.6-22.0%; p = 0.006) — but based on only 6 studies.
Test type matters: constant load tests (+16%, 95% CI 10.2-22.9%) and intermittent incremental tests (+18.5%, 95% CI 10.8-26.3%) detect the change better than conventional incremental tests (both p < 0.001); time trials did not differ from conventional incremental tests (p = 0.286). The meta-analysis found the improvement in constant load tests, time trials and intermittent incremental tests, but not in conventional incremental tests.
Longer tests show larger improvements (+0.4% per minute of test duration, 95% CI 0.1-0.6%; p = 0.011).
Type of sport does not influence the magnitude of improvement (all p > 0.05).
Practical Application
Meta-analysis of 8 controlled trials plus regression over 46 studies: respiratory muscle training improves endurance performance in healthy people, more so in less fit subjects (+6.0% per 10 mL/kg/min lower VO2max) and in longer tests (+0.4% per minute); inspiratory-strength and hyperpnoea training work equally, combined inspiratory/expiratory strength training may work better (+12.8%, 6 studies); the sport does not matter. For a highly trained athlete the expected gain is small; for a less fit or returning athlete it is worth the ten minutes a day.
Not from this paper; the base's own practice: a threshold-loading device (POWERbreathe or equivalent) at a load that allows about 30 breaths per set, twice daily; progress the load as it gets easier. The abstract gives no protocol, load, set count or duration.
The performance effect is measured in constant-load and intermittent tests and time trials; a conventional ramp test will not show it (abstract) — do not judge the intervention by a ramp-test result.
Limitations
Regression pools heterogeneous protocols, populations and tests; the meta-analysis proper covers only 8 controlled trials.
The combined inspiratory/expiratory advantage rests on 6 studies.
Absolute percent gains depend strongly on the test used (abstract), so effect sizes are not directly transferable to race performance.
How this entry was checked and changed
Content correction 2026-09-27
"Effect: +3-5% time trial power" and "Expected gain: +3-5% sustained power output" -> not in the abstract; deleted. Replaced by the abstract's regression and meta-analysis figures (+6.0% per 10 mL/kg/min lower VO2max; +12.8% combined vs inspiratory; test-type and duration effects).
"Weeks 1-14: 2 × 30 breaths/day at 50% MIP" -> relabelled as the base's own practice without a load percentage or duration (not in the abstract).
Type "Meta-analysis (46 studies)" -> regression over 46 studies, meta-analysis over 8 controlled trials (abstract).
Sport "cycling" -> rowing, running, swimming, cycling (the sports covered by the review; abstract: sport does not change the effect).