Swimming With the COSMED AquaTrainer and K5 Wearable Metabolic System in Breath-by-Breath Mode: Accuracy, Precision, and Repeatability

Purpose: To compare ventilatory and cardiorespiratory responses between the COSMED AquaTrainer coupled with the K4b ² and K5 wearable metabolic systems in breath-by-breath mode over a wide range of swimming speeds. Methods: Seventeen well-trained master swimmers performed 2 front-crawl 7 × 200-m incremental intermittent protocols (increments of 0.05 m·s ⁻¹ and 30-s rest intervals, with a visual pacer) with AquaTrainer coupled with either K4b ² or K5. Results: Post hoc tests showed that swimming speed was similar (mean diff.: −0.01 to 0.01 m·s ⁻¹; P = .73–.97), repeatable (intraclass correlation coefficient: .88–.99; P < .001), highly accurate, and precise (agreement; bias: −0.01 to 0.01 m·s ⁻¹; limits: −0.1 to 0.1 m·s ⁻¹) between all conditions. Ventilatory and cardiorespiratory responses were highly comparable between all conditions, despite a “small” effect size for fraction of expired carbon dioxide at the sixth 200-m step (0.5%; η ² _p = .12; P = .04) and carbon dioxide production at the fifth, sixth, and seventh 200-m steps (0.3–0.5 L·min ⁻¹; η ² _p = .11–.17; P = .01–.05). We also observed high accuracy, which was greater for tidal volume (0.0–0.1 L), minute ventilation (−3.7 to 5.1 L·min ⁻¹), respiratory frequency (bias: −2.1 to 1.9 breaths·min ⁻¹), and oxygen uptake (0.0–0.2 L·min ⁻¹). Bland–Altman plots showed that the distribution inside the limits of agreement and their respective 95% CIs were consistent for all ventilatory and cardiorespiratory data. The repeatability (intraclass correlation coefficient) of tidal volume (.93–.97), minute ventilation (.82–.97), respiratory frequency (.68–.96), fraction of expired carbon dioxide (.85–.95), carbon dioxide production (.77–.95), fraction of expired oxygen (.78–.92), and oxygen uptake (.94–.98) data ranged from moderate to excellent (P < .001–.05). Conclusions: Swimming with the AquaTrainer coupled with K5 (breath-by-breath mode) yields accurate, precise, and repeatable ventilatory and cardiorespiratory responses when compared with K4b ² (previous gold standard). Swimming support staff, exercise and health professionals, and researchers can now relate differences between physiological capacities measured with the AquaTrainer while coupled with either of these 2 devices.