Reliability of using D-max method to define physiological responses to incremental exercise testing (Presentation)

Document Type

Conference publication

Publication details

Zhou, S & Weston, SB 1996, 'Reliability of using D-max method to define physiological responses to incremental exercise testing', Australian Conference of Science and Medicine in Sport: abstract book, Canberra, ACT, 28-31 October, Sports Medicine Australia, Bruce, ACT, pp. 448-449.



The concept of anaerobic threshold has attracted a great attention in exercise sciences. Various methods have been developed in attempt to detect the exercise intensity at the threshold in the responses of either blood lactate concentration, pulmonary ventilation, or other physiological variables to exercise. Some commonly encountered problems in these methods include the lack of objectivity, individuality or supporting physiological mechanism. Recently, Cheng et al. (1) introduced a new method called D-max to identify anaerobic threshold in attempt to overcome some of these problems. However, no further publications have been found in the literature using this method. The purpose of the present investigation was to examine the reliability of using the D-max method to define blood lactate kinetics in response to incremental exercise test and compare the physiological responses corresponding to the workload at D-max with those determined by the traditional 4 mM lactate concentration and ventilatory threshold (VT).


Subjects and testing protocol. Ten male cyclists and triathletes, with a minimum of two years training in the sports, voluntarily participated in the study. Their age, body mass and height were (mean +/- SD) 25.6 +/- 8.2 years, 70.65 +/- 9.28 kg and 1.76 +/- 0.02 m, respectively. The subjects performed an incremental cycling exercise till exhaustion in two tests separated by four weeks. In attempt to control the pre-test conditions, the training program during four weeks, as well as the food consumption of 24 h and water consumption of 2 h, prior to the first test, were recorded and duplicated before the second test. The tests were performed in the morning, following approximately 12 h fast, on an air-braked pursuit cycle ergometer. The workload started at 150 watts for 3 min, then increased by 25 watts every 3 min until volitional exhaustion. The expired gas was analysed using an on-line gas analysing system, with the variable values averaged for each 15 s.

Determination of D-max. Two millilitre blood samples were obtained via an indwelling venous catheter, immediately before exercise, during the final 30 s of each workload and at exhaustion. The samples were then centrifuged and the plasma was used in the subsequent assay to determine the lactate concentration. Third order polynomial regression equation was established on the plasma lactate concentrations versus workloads. The D-max was identified as the point on the polynomial regression curve that yielded the maximal distance to the straight line formed by the two end data points.

Determination of ventilatory threshold. Ventilatory thresholds were determined by visually inspecting the non-linear changes in VE, FEO2, VCO2, VE/VO2 and VE/VCO2 in plotted graphs (2). Two break points were identified using this method, corresponding to VT1 and VT2 as defined by McLellan (3). The ventilatory threshold was also identified by a computerised analysis (VTc) on a two-compartment linear model (4).

Linear regression equations were established on the responses of VO2 and HR versus exercise time. The HR, lactate concentration, and VO2 (L/min and %VO2peak) at D-max, 4mM lactate, and VTc, VT1 and VT2 were derived from the regression equations. Paired student t-tests were performed to detect differences between mean values of the variables obtained in the two tests. Intraclass correlation was used to evaluate the test-retest reliability.


The results demonstrated a good test-retest reliability in the physiological measurements at both D-max point and exhaustion (Table 1). No significant differences (p>0.05) were found between the mean values of the two tests in these measurements.

Table 1. Mean values (+/-SE) and reliability coefficients of peak values of VO2, HR, plasma lactate concentration and exercise time, and the values of these parameters found at D-max. (* p<0.01) VO2 (L/min) HR (beats/min) Lactate (mM) Exerc. Time (min) Peak Values Test 1 4.52 +/-0.12 183.8 +/-3.1 13.34 +/-0.86 22.70 +/-0.73 Test 2 4.57 +/-0.11 184.6 +/-3.3 12.88 +/-0.83 22.59 +/-0.76 Reliability Coefficient 0.90* 0.93* 0.79* 0.93* D-max Test 1 3.66 +/-0.07 157.1 +/-3.3 3.38 +/-0.32 14.43 +/-0.46 Test 2 3.68 +/-0.07 158.6 +/-3.9 3.46 +/-0.11 14.52 +/-0.47 Reliability Coefficient 0.90* 0.93* 0.34 0.90*

In fact, the values of these measurements at the VTc, VT1, VT2 and 4mM lactate also showed similar reliability coefficients (range from 0.80 to 0.95, p<0.01). An exception was the lactate concentration, as the reliability coefficient was in a range of 0.34-0.42 (excluding 4mM lactate, all p>0.05). The lactate concentration at the thresholds was found to be the highest at VT2 (4.20 mM, p<0.05 compared with D-max, 3.37 mM) and the lowest at VT1 (1.65 mM, p<0.05 compared with all other thresholds).

In practice, D-max appears to be a reliable method to define the individual response to exercise test, with an advantage of objectivity. According to the physiological responses observed, the exercise intensity at the D-max appeared to be lower than that at VT2, but higher than that at VT1. However, there is no evidence to support that the exercise intensity defined by D-max method would be superior to that defined by other methods to prescribe training intensity or predict aerobic performance. Further investigations are needed to evaluate the validity of using D-max method in exercise prescription and consultation.


(1) Cheng B., Kuipers H., Snyder A.C., Keizer H.A., Jeukendrup A. & Hesselink M. (1992). Int. J. Sports Med. 13:518-522.

(2) Davis J.A., Vodak P., Wilmore J.H., Vodak J. & Kurtz P. (1976). J. Appl. Physiol. 41:544-550.

(3) McLellan T.M. (1987). Aust. J. Sci. Med. Sport. 19:3-8.

(4) Orr G.W., Green H.J. & Hughson R.L. (1982). J. Appl. Physiol. 55:1694-1700.