Research Appraisals

Journal of neurology

Effects of respiratory muscle training on respiratory function in patients with amyotrophic lateral sclerosis: a systematic review and meta-analysis

BACKGROUND: Respiratory decline is prognostically important in amyotrophic lateral sclerosis (ALS), but the efficacy of respiratory muscle training (RMT) remains uncertain. We synthesized randomized evidence on respiratory and related outcomes. METHODS: We searched databases, specialized registers, and trial registries through June 19, 2025, for randomized RMT trials in ALS. Risk of bias was assessed with the Cochrane tool. Random-effects meta-analyses reported standardized mean differences (SMDs) with 95% confidence intervals (CIs), and certainty was rated with GRADE. RESULTS: Six studies were included; five concurrently randomized trials contributed quantitative data, whereas Pinto 2013 was retained for qualitative context. RMT improved maximal expiratory pressure (MEP; SMD 0.387, 95% CI 0.192-0.581; P = 0.008) and showed small favorable effects on maximal inspiratory pressure (MIP; SMD 0.156, 95% CI 0.089-0.224; P = 0.005), sniff nasal inspiratory pressure (SNIP; SMD 0.216, 95% CI 0.039-0.392; P = 0.034), peak expiratory flow (PEF; SMD 0.205, 95% CI 0.088-0.323; P = 0.017), and ALSFRS-R (SMD 0.214, 95% CI 0.039-0.388; P = 0.030). Forced vital capacity showed a borderline favorable trend (SMD 0.129, 95% CI - 0.003 to 0.262; P = 0.053). Low heterogeneity estimates were imprecise because most endpoints included only three to four small studies. Certainty was moderate for MIP, low for MEP, FVC, and PEF, and very low for SNIP and ALSFRS-R. Exploratory analyses identified no reliable effect modifiers. CONCLUSIONS: RMT was associated with small, directionally consistent improvements mainly in pressure- and flow-based outcomes, whereas effects on FVC and longer-term clinical outcomes remain uncertain. RMT may be considered an individualized adjunct to ALS respiratory care, not a disease-modifying therapy. Larger, longer, standardized trials with patient-important endpoints are needed.