Accuracy Of K4b2 Portable Breath-By-Breath System During Treadmill And Outdoor Running

The portable K4b2 (K4b2, COSMED s.r.l, Rome, Italy) was designed to measure breath-by-breath pulmonary gas exchange in the field. Its validity has been confirmed during treadmill running, but its accuracy has never been assessed during actual outdoor running. The main difference between treadmill and outdoor running is the facing air velocity that increases as a function of running speed. PURPOSE: To determine the accuracy of the K4b2 as compared to the criterion Douglas bag method during treadmill and outdoor running. METHODS: Twelve subjects completed three discontinuous maximal incremental tests running for 3 min at speeds between 8 to 20 km·hr−1 in 2 km·hr−1 steps, with 3 min recovery between bouts. One treadmill test used the criterion method; two tests used the K4b2, one on the treadmill and the other outdoors. Treadmill tests were done indoors with a gradient of 1%, the outdoor tests were done on a running track. Gas exchange parameters were recorded throughout. RESULTS: Oxygen uptake (VO2) with K4b2 did not differ significantly from the criterion method during submaximal (8-16 km·h−1) treadmill running although there was a tendency for VO2 measured with K4b2 to be lower (mean difference of 0.73±3.18 mL·kg−1·min−1, p=0.056; equivalent to 95% limits of agreement of ±17%). Submaximal VO2 with K4b2 outdoors was 6.10±5.39 mL·kg−1·min−1 (p<0.001) lower than the criterion method and 6.83±5.19 mL·kg−1·min−1 (p<0.001) lower than with K4b2 on the treadmill. VO2 peak was lower with K4b2 outdoors (VO2 peak: 48.4±8.6 mL·kg−1·min−1) vs. criterion method (59.1±9.2 mL·kg−1·min−1; p=0.001) and vs. K4b2 on the treadmill (59.2±10.0 mL·kg−1·min−1; p=0.001) but there was no difference between K4b2 on the treadmill vs. criterion method (p=0.897) corresponding to a mean bias of 0.1 mL·kg−1·min−1, with limits of agreement-5.2 to 5.4 mL·kg−1·min−1. CONCLUSION: The K4b2 tends to systematically underestimate VO2 during submaximal treadmill running indoors but satisfactorily measures VO2 peak. During actual outdoor running, the K4b2 underestimates VO2 during submaximal and maximal running. We hypothesise that this may due to the facing air velocity affecting sample collection and/or measurement of airflow by the K4b2 setup on the facemask.