Visually identified breakpoints in muscle oxygen saturation are not equivalent to ventilatory thresholds in national-level triathletes

Purpose Portable near-infrared spectroscopy (NIRS) enables cost-effective monitoring of muscle oxygen saturation (SmO2) during exercise, presenting a promising alternative for delineating exercise intensity domains. This study investigated whether SmO2 breakpoints (BP1 and BP2) can provide accurate estimates of ventilatory thresholds (GET and RCP). Methods Twelve male Tier 3 triathletes completed an outdoor incremental TestVAM. Vastus lateralis SmO2 was recorded using NIRS, alongside pulmonary gas exchange measurements. BP1 and BP2 were identified visually. Heart rate (HR) and running speed at BP1, BP2, GET, and RCP were extracted. The intraclass-correlation coefficient (ICC) was calculated and equivalence between SmO2 breakpoints and ventilatory thresholds was assessed with equivalence bounds set at 5 bpm for HR and 0.5 km h−1 for speed. Results The ICC(3,1)abs was poor to moderate for all comparisons (range: 0.18–0.69). HR and speed between BP1 and GET were not equivalent (p>0.91), with GET occuring ealier in both HR (p= 0.014, Δ=− 9.92 bpm, 90% CI [ −16,−3.83]) and speed (p=0.007, Δ=−1.29 km h−1, 90% CI [ − 1.99,−0.59]). Similarly, BP2 and RCP were not equivalent (p>0.98), and both HR (Δ=−9.3 bpm, 90% CI [− 12.18,−6.42]) and speed (Δ=−1.75 km·h−1, 90% CI [−2.33,− 1.17]) occuring earlier at RCP (p<0.001). Conclusion In this sample, SmO2 breakpoints did not fullfill criterion validity against ventilatory thresholds and therefore should not be used for determining exercise intensity domains. Further research should evaluate their equivalence using robust detection methodologies in larger samples.