Ventilatory and lactate threshold determinations in healthy normals and cardiac patients: methodological problems.

In healthy normal individuals (n = 69), coronary patients with myocardial ischaemia (n = 27) and patients with chronic heart failure (CHF, n = 33), four widely applied methods to determine ventilatory threshold (VT) were analysed: V-slope, ventilatory equivalent for O2 (EqO2), gas exchange ratio (R) and end-tidal partial pressure of oxygen. Lactate threshold [LAT, log lactate vs log oxygen uptake (VO2)] was also determined. Analysis focused on rate of success of threshold determination, comparability of threshold methods, reproducibility and interobserver variability. Cycle ergometry protocols with ramp-like mode and graded steady-state mode used in exercise testing were considered separately. In healthy normal individuals and coronary patients with myocardial ischaemia, at least three VT could be determined during ramp-like mode and two VT during graded steady-state mode, 82% of the time. For CHF patients, the rate of successful determination of VT was lower. Compared to LAT, VO2 at VT was significantly higher using R and EqO2 methods of VT determination in healthy normal subjects (P < 0.01), and significantly higher when using all four methods in coronary patients (P < 0.01 or P < 0.05, respectively). No difference was observed between VO2 at VT and LAT in CHF patients. In healthy normal individuals, day-to-day reproducibility of VT and LAT was high (error of a single determination from duplicate determinations was between 3.9% and 6.2% corresponding to a VO2 of 52.2 and 89.2 ml.min-1). Interobserver variability was low (error between 0.3% and 5% corresponding to a VO2 of 9.8 and 68 ml.min-1). In CHF patients, interobserver variability was moderately greater (error between 4.6% and 8.2%, corresponding to a VO2 of 35.1 and 62.4 ml.min-1). To optimize threshold determination, standardized procedures are suggested.