Swimming performance relies on the coordinated function of multiple physiological systems, with respiratory efficiency being a crucial yet often overlooked limiting factor. Swimmers face several unique physiological challenges in the aquatic environment. First, hydrostatic pressure causes elastic loading of the thoracic cavity, increasing the work and energetic cost of breathing, thereby elevating the burden on the inspiratory muscles (Pendergast et al., 2015). Second, the biomechanics of the stroke constrain the rhythm and timing of breathing, forcing it to adapt to the stroke cycle (Seifert et al., 2010). Additionally, during submersion, the diaphragm and respiratory muscles generate less force in the supine position than in the upright posture (Lomax and McConnell, 2003). These factors increase the likelihood of respiratory muscle fatigue, which can trigger the respiratory metaboreflex—leading to vasoconstriction in inactive limbs (St Croix et al., 2000), a 23%–30% reduction in limb blood flow, and increased vascular resistance (Sheel et al., 2001). In swimmers, these physiological constraints can impair stroke efficiency, elevate energy expenditure, and reduce overall training capacity. For instance, inspiratory muscle fatigue may lead swimmers to adopt suboptimal breathing patterns or shorten their stroke cycles to accommodate respiratory demands (Seifert et al., 2010; Lomax and Castle, 2011), both of which can negatively impact performance metrics such as sprint time and distance per stroke.

Respiratory muscle training (RMT), including inspiratory muscle training (IMT), expiratory muscle training (EMT), combined, or breath-holding protocols, aims to improve respiratory endurance and/or strength through resistance-based (e.g., pressure-threshold loading) or endurance-based strategies (e.g., voluntary isocapnic hyperpnea using devices like Spirotiger) (McConnell and Romer, 2004; McConnell, 2013). A typical RMT protocol involves training using pressure-threshold devices at 50%–70% of maximal inspiratory pressure (MIP), performed twice daily, with 30 effective breaths per session, 6–7 days per week, for at least 4–6 weeks (McConnell, 2013). Over the past decades, the role of RMT in enhancing athletic performance has been extensively investigated. For instance, in land-based endurance sports such as cycling and running, multiple studies have shown that RMT significantly improves time trial performance and delays fatigue (Holm et al., 2004; Romer et al., 2002; Bailey et al., 2010; Archiza et al., 2018). In combat sports, Tosun et al. reported that RMT significantly improved respiratory muscle strength and aerobic endurance in young wrestlers (Tosun et al., 2025).

Swimming is no exception. Several studies have reported that RMT can significantly improve swimming performance (Jakubovskis et al., 2024; Ohya et al., 2021). The proposed physiological mechanisms may include: enhanced inspiratory muscle strength, enabling swimmers to overcome hydrostatic pressure on the thoracic cavity, reduce the energy cost of breathing, and redirect metabolic resources toward propulsion (Pendergast et al., 2015; Gómez-Albareda et al., 2023); improved respiratory muscle function, allowing for more efficient breathing within constrained stroke cycles, minimizing movement disruption, and optimizing stroke frequency and length (Kilding et al., 2010); targeted training that increases diaphragmatic force output efficiency in a supine aquatic position (Lomax and McConnell, 2003); and greater respiratory muscle fatigue resistance, which may attenuate sympathetically mediated respiratory metaboreflexes, preserve limb blood flow and vascular function, and delay peripheral fatigue during exercise (St Croix et al., 2000; Sheel et al., 2001; Kowalski et al., 2022). However, other studies have failed to observe significant improvements in swimming performance following RMT (Cunha et al., 2019; Wells et al., 2005). These inconsistencies may be attributed to methodological limitations, including lack of randomisation, small sample sizes, and variations in RMT protocols, all of which limit cross-study comparability. For example, Cunha et al. implemented a 12-week program involving 30 inspiratory efforts twice daily, 5 days per week, at 50% of MIP in 32 elite swimmers (Cunha et al., 2019). In contrast, Jakubovskis et al. conducted a 4-week intervention consisting of modified respiratory exercises three times per week with 31 swimmers (Jakubovskis et al., 2024). Given these methodological differences, a meta-analytic synthesis is warranted to enhance the scientific rigor of RMT-based training recommendations for swimmers.

Previous reviews, including that by HajGhanbari et al., have focused on endurance sports such as football, cycling, and rowing, and confirmed RMT’s positive effects on endurance capacity, pulmonary function, and muscle strength (HajGhanbari et al., 2013; Xavier et al., 2025; León-Morillas et al., 2021; Illi et al., 2012). However, to date, no meta-analysis has systematically evaluated the effects of RMT on swimming performance while assessing the quality and risk of bias of included studies. One study by Tello et al. showed improvements in MIP in both elite and non-elite swimmers but did not assess direct performance outcomes (Carvajal-Tello et al., 2024). Therefore, this meta-analysis included randomized controlled trials (RCTs) or controlled clinical trials (CCTs) to investigate the effects of RMT on swimming performance in healthy individuals. Based on the PICO framework, the review focused on: Population: Healthy individuals or competitive swimmers; Intervention: RMT, IMT and/or EMT; Comparison: No RMT or placebo intervention; Outcome: Swimming performance outcomes.

This study adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021).

This meta-analysis is the first to systematically evaluate the effect of RMT on swimming performance. The pooled results revealed a small but significant improvement in performance among swimmers receiving RMT, with low heterogeneity and no publication bias. These findings provide quantitative support for the inclusion of RMT in swim training programs, reinforcing its potential as an effective ergogenic strategy in aquatic sports.

The results of the study showed a small effect size improvement of RMT on swimming performance compared to the control group, with a combined effect size of SMD = −0.49 (95% CI [-0.70, −0.28], p < 0.00001), low heterogeneity (Chi² = 20.63, p = 0.42; I² = 3%) and no significant publication bias (Egger’s test: p = 0.408 > 0.05) further supported this result. In addition, a leave-one-out sensitivity analysis was conducted to evaluate the robustness of the findings (Figure 6). The analysis showed that only the study by Tan et al. exhibited substantial skewness. Excluding this study led to a reduced overall effect size, although it remained within the small effect range (0.2–0.49). This suggests that the exclusion of any single study did not materially alter the overall effect estimate—that is, the beneficial effect of RMT on swimming performance was not driven by any one study. Nevertheless, risk-of-bias considerations temper the confidence in these results. According to the RoB 2 assessment, four included (Tan et al., 2023; Lomax et al., 2019; Yañez-Sepulveda et al., 2021; Bernhardt, 2025) studies were rated as having a high overall risk of bias due to concerns related to deviations from intended interventions and missing outcome data (Figure 2). Although their exclusion did not significantly change the pooled estimate, the potential for bias in these studies cannot be fully dismissed. Therefore, while the meta-analysis supports the positive effects of RMT, the strength of this conclusion should be interpreted with caution given the methodological limitations present across several studies. Future trials should adopt more rigorous designs, including improved randomization, adherence monitoring, and complete outcome reporting, to increase the confidence in effect estimates.

This finding is consistent with the majority of previously published empirical studies. For example, Ohya et al. conducted a 6-week MIP intervention in 30 male swimmers and reported significant improvements in 100 m freestyle performance compared to the control group (HI: p = 0.046; MOD: p = 0.042) (Ohya et al., 2021). However, some studies reported no significant effects of RMT on swimming performance. For instance, Cunha et al. conducted a 12-week intervention involving 32 participants and found no significant changes in swimming performance (p = 0.271), inspiratory muscle strength (p = 0.914), exercise-related lung capacity (p = 0.262), forced expiratory volume in one second (p = 0.265), peak expiratory flow rate (p = 0.270), or perceived dyspnoea (p = 0.568) (Cunha et al., 2019). Despite the overall positive trend, inconsistencies across studies warrant further investigation. Variations in RMT outcomes may be attributed to differences in training load (e.g., 30% vs. 80% of MIP), training frequency (5–14 sessions per week), and intervention duration (4–12 weeks), suggesting the existence of a potential dose–effect relationship. However, only two studies reported detailed training duration (Lemaitre et al., 2013; Tan et al., 2023), limiting our understanding of the optimal training dose. Comparisons were further complicated by the absence of standardised progression protocols (e.g., criteria for increasing training intensity). Future research should clearly define and standardise these parameters. Regarding dose–effect, Mickleborough et al. emphasised that exceeding certain thresholds and maintaining sustained loads are critical for inducing adaptations in pulmonary function, a principle that may also apply to RMT’s performance effects (Mickleborough et al., 2008). Nevertheless, optimal RMT parameters for swimmers remain unclear. With respect to gender, no included study enrolled exclusively female participants, precluding conclusions about potential gender-specific responses to RMT. Variables such as baseline respiratory muscle strength, athletic level (elite vs. sub-elite), age, and intrinsic pulmonary capacity may influence the response to RMT. Yet, the study lacked subgroup stratification or sufficient statistical power to examine these interactions. Lastly, although RMT improves respiratory muscle strength, allocating excessive time to RMT may detract from swimming-specific technical training. Consequently, RMT must be carefully integrated within an athlete’s overall training plan. McConnell emphasized that excessive RMT may compromise sport-specific adaptations if improperly timed or dosed, particularly during periods of high-volume technical training (McConnell, 2013). The concept of a ‘training interference’ underscores the importance of applying the minimum effective dose and aligning RMT with the athlete’s periodized program. This highlights the importance of balancing general physiological adaptation with sport-specific skill acquisition. Given that RMT constitutes an additional training component, its potential to contribute to accumulated fatigue and interfere with sport-specific training adaptations should not be overlooked. Future studies should examine whether integrated periodised programmes combining RMT and swimming training yield superior performance outcomes.

To the authors’ knowledge, this is the first meta-analysis to evaluate the effect of RMT on swimming performance. In a recent meta-analysis focusing solely on lung function in swimmers, Tello et al. reported that RMT significantly improved MIP (MD = 29.35 cmH₂O; 95% CI [13.04, 45.65] cmH₂O; p < 0.01), whereas no significant effects were observed for maximal expiratory pressure, forced expiratory volume in one second, or forced vital capacity (Carvajal-Tello et al., 2024). Furthermore, the current findings are broadly consistent with previous meta-analysis investigating the effects of RMT in trained athletes (HajGhanbari et al., 2013; Xavier et al., 2025; León-Morillas et al., 2021; Illi et al., 2012). For example, HajGhanbari et al. reported that RMT had significant positive effects on time-trial performance (MD = 0.40), exercise endurance (MD = 5.17), and Yo-Yo test performance (MD = 4.86) (HajGhanbari et al., 2013). Similar to cycling, swimming imposes high ventilatory demands, and respiratory muscle fatigue may limit blood flow redistribution to working limb muscles, thereby impairing performance (Romer and Polkey, 2008; Harms et al., 2000). RMT may enhance swimming economy by increasing MIP and MEP, delaying respiratory muscle fatigue, and improving oxygen utilization efficiency (Kilding et al., 2010; McConnell, 2009). Notably, the swimming distances included in this study (50–400 m) corresponded to submaximal intensities. The effects of RMT on shorter sprint events (<30 s) may be limited due to their predominant reliance on anaerobic energy pathways (Shei, 2018).

Although the present study observed a consistent trend of improvement across different strokes, the majority of performance outcomes were derived from freestyle events (13 out of 21), which may obscure stroke-specific effects. The lateral breathing pattern in freestyle places high demands on diaphragmatic and intercostal muscle coordination (McLeod, 2009). RMT may improve breathing efficiency by enhancing stroke-breath synchronization, for example, by reducing peak inspiratory resistance during the breathing cycle (Lavin et al., 2015). However, other strokes such as butterfly or breaststroke involve different breathing mechanics, including forward head lifts and varying trunk rotations, which may engage distinct respiratory and postural muscle groups (Barbosa et al., 2006). These biomechanical and muscular differences suggest that the effectiveness of RMT could vary by stroke type. Therefore, future studies should consider evaluating RMT efficacy in stroke-specific contexts and reporting outcomes separately for each stroke. Tailored RMT protocols may be necessary to address the unique ventilatory demands of each swimming style.

Several limitations of this study should be acknowledged. The small number of included studies (n = 10) limited the statistical power of the subgroup analysis. Notably, none of the studies specifically examined female swimmers, precluding any meaningful exploration of gender-based differences. Furthermore, the lack of detailed information on training duration and the absence of standardised progression protocols hindered a robust dose–effect analysis. Although the overall methodological quality was rated as ‘moderate’, four studies were assessed as being at high risk of bias, and most failed to implement adequate blinding and rigorous randomisation procedures. These methodological weaknesses could potentially result in performance bias or an overestimation of the true effects. Additionally, the analysis were unable to adequately account for how RMT was integrated into the swimmers’ overall training programmes, leaving possible confounding effects unresolved. Since all participants had prior swimming experience, the generalisability of the findings to untrained or recreational populations remains limited. Lastly, four studies relied on image-based data (Ohya et al., 2021; Kilding et al., 2010; Tan et al., 2023; Bernhardt, 2025), which may have introduced measurement errors.

The findings of this meta-analysis suggest that RMT may confer moderate improvements in swimming performance compared to control conditions (SMD = −0.49). RMT serves as a supplementary modality to resistance training in swimmers. Among the included studies, RMT protocols were generally implemented at intensities ranging from 50% to 80% of MIP, over 6–8 weeks, with a training frequency of 3–14 sessions per week. However, due to considerable heterogeneity in programme design and the absence of direct comparisons across RMT modalities, no definitive conclusions can be drawn regarding the optimal training parameters. Moreover, the physiological mechanisms underlying the observed performance benefits remain speculative, as they were not directly assessed within the scope of this review. Future research should aim to clarify the dose-effect relationship, explore long-term physiological adaptations, and evaluate how individual swimmer characteristics—such as sex, stroke style, and competitive level—moderate the effects of RMT.