Novel Derivative-Based Approach for Ventilatory Threshold Detection in Recreationally Active Female and Male College Students

Background: Ventilatory threshold (VT) delineates moderate from heavy intensity exercise and serves as a key training marker, yet existing determination methods from VO 2 max tests are imprecise, subjective, or validated predominantly in elite athletes. have significant. V-Slope requires subjective visual inspection, fixed percentages lack individual precision, and Strucchange propagates regression error—both validated predominantly in elite athletes. Purpose: To develop and validate an objective, derivative-based method for determining ventilatory threshold (VT) from VO 2 max data. Hypothesis: A first-derivative approach analyzing breath-by-breath metabolic data will more accurately determine VT in recreational athletes compared to established methods (V-Slope and Strucchange). Methods: We developed a novel method that identifies VT by calculating the maximal rate of change in the ventilation-VO 2 relationship from breath-by-breath data without requiring regression equations or subjective visual inspection. We analyzed VO 2 max data in 3 recreational athletes (1 female, 2 males; age: 20.7 ± 0.6 years; VO 2 max: 49.6 ± 3.1 ml/kg/min) who completed a graded ramp exercise test to volitional exhaustion on a cycle ergometer. VO 2 max was confirmed by RER ≥1.10 and HR ≥90% age-predicted maximum. Validation: (1) laboratory VO 2 max data comparing all 3 methods, and (2) synthetic data with systematically varied noise levels (σ = 1 to 5) simulating elite (low noise) versus recreational (high noise) athlete data quality. Accuracy was quantified by mean squared error (MSE) and cross-checking the number of true positives and true negatives. Results: In preliminary laboratory VO 2 max data our novel method identified VT at 60.9 ± 0.2 % VO2max (p < 0.001), V-Slope at 68.8 ± 0.03% VO2max (p < 0.001), and Strucchange at 46.8 ± 0.1% VO2max (p = 0.0496). In synthetic data simulating elite athletes (low noise, σ=1.5), all methods demonstrated similar accuracy assessed by MSE: novel method = 2.2 ± 0.17; V-Slope = 2.3 ± 1.54; and Strucchange = 2.17 ± 1.08). In synthetic data simulating recreational athletes (high noise, σ = 5), our novel method demonstrated significantly lower prediction error (MSE = 2.3 ± 0.05) compared to V-Slope (MSE = 5.57 ± 1.97) and Strucchange (MSE = 2.75±1.92), representing a 58% improvement in accuracy compared to V-Slope and 15% improvement compared to Strucchange. Conclusions: This novel derivative-based method is equally effective in determining VT as conventional methods in elite athletes but more effective in finding VT in recreational athletes and untrained individuals. Validation in a larger sample (n=11) is ongoing. This novel approach addresses a critical exercise physiology methodological gap, providing a more reliable tool for accurately determining VT across diverse populations. This abstract was presented at the American Physiology Summit 2026 and is only available in HTML format. There is no downloadable file or PDF version. The Physiology editorial board was not involved in the peer review process.