Clin Physiol Funct Imaging (2011)

doi: 10.1111/j.1475-097X.2011.01027.x

The measurement of lactate threshold in resistance exercise: a comparison of methods Nuno Manuel Frade de Sousa1, Rodrigo Ferro Magosso1, Guilherme Borges Pereira2, Richard Diego Leite2, Vivian Maria Arakelian1, Arlindo Neto Montagnolli3, Se´rgio Andrade Perez2 and Vilmar Baldissera1 1

Programa de Po´s-Graduac¸a˜o Interunidades Bioengenharia, EESC ⁄ FMRP ⁄ IQSC, USP, 2Laborato´rio de Fisiologia do Exercı´cio, Departamento de Cieˆncias Fisiolo´gicas da Universidade Federal de Sa˜o Carlos, and 3Departamento de Engenharia Ele´trica, Universidade Federal de Sa˜o Carlos, Sa˜o Carlos, SP, Brasil

Summary Correspondence Nuno Manuel Frade de Sousa, USP, Departamento de Bioengenharia, Av. Trabalhador Sa˜o-carlense, 400, Parque Arnold Schimidt, 13566-590 Sa˜o Carlos, SP, Brasil E-mail: [email protected]

Accepted for publication Received 4 February 2011; accepted 28 March 2011

Key words anaerobic threshold; blood lactate concentration; mathematical models; onset of metabolic acidosis; resistance training

Resistance incremental tests (IT) make it possible to determine critical metabolic and cardiovascular changes, such as the lactate threshold (LT). Different methods are frequently used to improve the exactness of LT identification. The objective of the study was to identify LT by four different methods (visual inspection, log–log, algorithmic adjustment and QLac) during resistance exercise and to evaluate which methods present more precision. Twelve men performed a maximal IT on the leg press at relative intensities of 10%, 20%, 25%, 30%, 35%, 40%, 50%, 60%, 70%, 80% and 90% of 1RM with 1-min stages. During the 2-min interval between stages, capillary blood was collected for blood lactate analysis. LT was detected using each of the four methods. The intensity of LT by visual inspection method was 26Æ9 (5Æ2)% of 1RM, adjustment algorithmic method was 27Æ8 (3Æ6)% of 1RM, log–log method was 23Æ3 (3Æ5)% of 1RM and QLac method was 31Æ6 (9Æ8)% of 1RM, with significant difference only between log–log and QLac methods. Bland and Altman analysis shows better concordance for visual inspection versus adjustment algorithmic methods. The visual inspection, algorithmic and log–log methods detected the LT at the same intensity. The mathematical models, specially the algorithmic method, provide more precision.

Introduction The anaerobic threshold (AT), initially proposed by Wasserman (Wasserman & McIlroy, 1964), has been the subject of innumerous studies since the 1960, especially through the study of the lactate threshold (LT), an exercise intensity in incremental exercise where blood lactate concentration (BLC) begins to increase in an exponential manner (Green et al., 1983; Beaver et al., 1985; Bishop et al., 1998). This intensity is also used as a marker for the onset of a metabolic acidosis (Davis, 1985). This physiological delimitation provides important information about the major physiological systems, supporting the prescription of resistance training and to monitor training progress for healthy population, athletes and rehabilitation (Svedahl & MacIntosh, 2003). The continuous change in BLC during an incremental test (IT) makes that the visual identification of a threshold lacks precision (Gladden et al., 1985). For that reason, various methods have been proposed to improve LT detection. To determine the breakpoint in BLC, beyond the visual inspection method (Lucia et al., 2002; Riedl et al., 2010), the log–log (Burke et al., 1994; Candotti et al., 2008) and algorithmic computation (Green et al., 1983; Podolin et al., 1991) methods were frequently used.

Because of the large number of methods for LT identification, the knowledge concerning those methods is essential, either for their application or for their interpretation in studies where they are employed. Although studies have demonstrated the application of LT for functional evaluation on cycling (MacIntosh et al., 2002), swimming (Ribeiro et al., 1990) or running (Weltman et al., 1990), only a few studies investigating the identification of LT in resistance exercise (Barros et al., 2004; Moreira et al., 2008; Rocha et al., 2010; Simoes et al., 2010). Consequently, many studies make a comparison between different methods during aerobic exercise, in an attempt to evaluate the most appropriate method for prescription of exercise programs (McMorris et al., 2000; Davis et al., 2007; Higa et al., 2007; Thomas et al., 2008). However, when LT was identified in resistance exercise, only two methods were used. The first was the visual inspection method (Barros et al., 2004; Simoes et al., 2010) and the second was a mathematical model of a second-order polynomial function (Moreira et al., 2008), obtained by the ratio between BLC and the relative exercise intensity (QLac = BLC ⁄ %1RM). In this context, it is essential to submit more data in the literature demonstrating LT in resistance exercise. Furthermore,

 2011 The Authors Clinical Physiology and Functional Imaging  2011 Scandinavian Society of Clinical Physiology and Nuclear Medicine

1

2 Lactate threshold in resistance training, N. M. F. de Sousa et al.

resistance IT at progressive percentages of 1 repetition maximum (1RM) increments make it possible to determine critical metabolic and cardiovascular changes during this type of exercise. Moreover, the mathematical models for LT identification need more investigation. The recent technological advances in equipment for the detection of LT and the use of mathematical and statistical algorithms have attended to facilitate, automate and ⁄ or semi-automate this procedure. The objectives of this study were (i) to identify LT by four different methods (visual inspection, log–log, algorithmic adjustment and QLac) during an incremental leg press protocol at progressive percentages of 1RM and (ii) to evaluate which methods present more precision. We hypothesized that is possible to determine the LT through these different methods in resistance exercise and the mathematical methods would be providing more precision for LT measurement.

Methods Subjects

After a 1-min rest, three repetitions at 70% of the expected 1RM were performed. The following attempts were made to identify 1RM, with a 5-min rest between attempts. Incremental test The IT was performed 48 h after the 1RM test. The selected intensities during IT were 10%, 20%, 25%, 30%, 35%, 40%, 50%, 60%, 70%, 80% and 90% of 1RM. This division of intensities was chosen because of previous studies demonstrating that the AT intensity in resistance exercise is around 30% of 1RM (Barros et al., 2004; Moreira et al., 2008; Simoes et al., 2010). Each stage lasted 1 min, and the subjects performed 20 repetitions, with each repetition lasting about 3 s, controlled by verbal commands. Passive recovery between the stages lasted 2 min to increase the intensity (%1RM) and for blood sampling collect. The end of the test was determined by the subjects incapacity to perform the movement within the previously established correct biomechanics or by the incapacity to perform the number of repetitions established for the stage. The blood samples (25 ll) were collected from the earlobe 30 s after the end of each stage using heparinized capillaries previously calibrated (Goodwin et al., 2007). BLC was immediately determined by electroenzymatic method with a lactate analyzer (1500 Sport; Yellow Springs Instruments Inc., Yellow Springs, OH, USA).

Twelve healthy men [mean (SD) age 26Æ0 (2Æ9) years; height 1Æ8 (0Æ1) m; body mass 83Æ0 (8Æ5) kg] participated in the study. Subjects had a minimum experience of 6 months in resistance training. The participants also answered the Physical Activity Readiness Questionnaire (Shephard, 1988). The following additional exclusion criteria were adopted: (i) use of any kind of medication or anabolic steroids; (ii) cardiovascular, neurological or orthopaedic complications that could limit the performance of the exercises. This study was approved by the local Institutional Research Ethics Committee and has been conducted according to the principles expressed in the Declaration of Helsinki. The procedure was explained to each subject separately, and they signed informed consent forms.

The LT detection was performed by two different methods that were previously used in the literature for AT identification in resistance exercise (Moreira et al., 2008; Simoes et al., 2010) and by two other methods that were only used in aerobic exercises (Green et al., 1983; Beaver et al., 1985).

Experimental design

Visual inspection method

The subjects accomplished two exercise sessions in the morning on different days. In the first session, 1RM was determined. After 48 h, each subject performed an IT on leg press (LP) where BLC was measured at each stage. Finally, the LT was determined using four different methods.

The visual inspection method, proposed by Wasserman et al. (1973), was performed by the visual inspection of the plotted graphs between BLC and work load. The exercise intensity at the onset of a systematic increase in BLC was considered the LTVI. The visual inspection method was made by three independent people who had experience with this method.

Lactate threshold detection

One-repetition maximum test The participants performed one or two sessions of adaptation on LP (Leg Press 45; Reforce, Jau´, Sao Paulo, Brazil) to establish the correct biomechanics of the movement, which includes the following: (i) each repetition cycle lasted about 3 s, with 1Æ5 s for the concentric phase and 1Æ5 s for the eccentric phase; (ii) motion angle of knee between 90 and 180, controlled by an eletrogoniometer (EMG system, Brazil). After adaptation, the subjects performed 1RM test on LP as suggested by Kraemer & Fry (1995). Briefly, in the warm up they performed eight repetitions at 50% of the expected 1RM.

Algorithmic adjustment method The algorithmic adjustment method was based on Orr et al. (1982), by a double linear regression model. The calculi were developed in MATLAB version 7.4 (MathWorks, Natick, MA, USA). We performed a computerized 2-segment regression analysis to locate the intersection point of the segments in the BLC versus work load. The best-fit regression model is chosen by minimizing the pooled residual sum of squares. The two linear regression equations (one for the lower component and one for the upper component) were set equal to each other and solved

 2011 The Authors Clinical Physiology and Functional Imaging  2011 Scandinavian Society of Clinical Physiology and Nuclear Medicine

Lactate threshold in resistance training, N. M. F. de Sousa et al. 3

for x, which represents the work rate at the LT, called for this method as LTAA.

log work rate at the intersection of the two linear lines. The antilog of that log work rate is the work rate at the LTLL.

Log–log method

QLac method

The log–log method was based on the traditional research of Beaver et al. (1985). The BLC and exercise intensity data were transformed to their logarithmic values. The effect of a logarithmic transformation was the linearization of the curvilinear manner observed between the BLC and exercise intensity in a non-logarithmic scale. The next step was to determine the log work rate at the LT from the log–log plot (ATLL). This was carried out by dividing the log–log data points into two segments that have a common data point as there is no statistical rationale for assigning that data point to one segment versus the other. For this study, the first choice for the common data point was the point found by the visual inspection method. Linear regression was then performed on both segments, and the residual sum of squares was computed for each segment. The residual sum of squares for both segments was then added together as it is an index of the goodness of fit for the two segments taken together. The analysis was then repeated by advancing one data point above the initial choice for the common point. It was then repeated again by retreating one data point below the initial choice for the common point. Once a minimum value was found, the two linear regression equations (one for the lower component and one for the upper component) were set equal to each other and solved for x, which represents the

For the QLac method, a polynomial function was applied, producing a U-shape curve similar to a lactate minimum test (Moreira et al., 2008). The ratio between BLC and the relative exercise intensity (QLac = BLC ⁄ %1RM) was plotted against the relative exercise intensity (%1RM) for each stage of IT to make possible, through a mathematical adjustment, the use of a second-order polynomial function. The resulting second-order polynomial equation was derived and then solved to accurately identify the vertices of the curve, being its corresponding workload considered as LTQLac.

Table 1 Means (SD) of intensity and blood lactate concentration (BLC) at lactate threshold calculated by visual inspection (LTVI), algorithmic adjustment (LTAA), log–log (LTLL) and QLac (LTQLac) methods. LTVI

LTAA

LTLL

Results are expressed as means (SD). The statistical analysis was performed initially by the Kolmogorov–Smirnov normality test and by the homocedasticity test (Bartlett criterion). All variables analysed in the study presented normal distribution and homocedasticity. One-way analysis of variance (ANOVA) test was used to compare the variables LT intensity (percentage of 1RM) and BLC. Tuckeys post-hoc test was applied in the event of significance. The intraclass correlation coefficient was used to examine the strength of relationships among the four LT detection methods. The agreement between the protocols was confirmed using the Bland and Altman method (Bland & Altman, 1986), which were expressed as means ± 2SD. The level of significance was P<0Æ05, and MEDCALC version 9.6 (Bvba, Mariakerke, Belgium) software was used.

LTQLac

Intensity (%1RM) 26Æ9 (5Æ2) 27Æ8 (3Æ6) 23Æ3* (3Æ5) 31Æ6 (9Æ8) BLC (mmol L)1) 1Æ91 (0Æ78) 1Æ84 (0Æ61) 1Æ58* (0Æ63) 2Æ6 (1Æ4) *Significantly different to LTQLac (P<0Æ05).

Statistical analysis

Results During IT, the mean maximum intensity achieved was 68Æ5 (9Æ9) % of 1RM. Mean lactate concentrations at rest and at

Figure 1 Lines of best fit and lactate threshold for subject no. 6, using the visual inspection (a), algorithmic adjustment (b), log–log (c) and QLac (d) methods.  2011 The Authors Clinical Physiology and Functional Imaging  2011 Scandinavian Society of Clinical Physiology and Nuclear Medicine

4 Lactate threshold in resistance training, N. M. F. de Sousa et al.

maximum intensity were 1Æ01 (0Æ50) mmol L)1 and 8Æ52 (2Æ27) mmol L)1, respectively. Lactate threshold was identified in all four methods (Table 1). However, for three subjects (nos. 5, 8 and 11) it was not possible to identify LT by the QLac method. The derived secondorder polynomial equation that identifies the vertices of the curve outside the workload performed did not allow the identification of LT. The mean intensity of LTVI was 26Æ9 (5Æ2)% of 1RM, the LTAA was 27Æ8 (3Æ6)% of 1RM, the LTLL was 23Æ3 (3Æ5)% of 1RM and the LTQLac was 31Æ6 (9Æ8)% of 1RM. Repeated measures ANOVA

shows that only the intensity of LTLL was significantly lower than LTQLac (P<0Æ05), expressed in percentage of 1RM. Figure 1 shows the line best fit for lactate for subject 6, whose intensity values of the LT were closely related using three methods (visual inspection, algorithmic adjustment and log–log) and overeated for QLac method. The intraclass correlation coefficients was 0Æ89 for LTVI versus LTAA, 0Æ91 for LTVI versus LTLL, 0Æ73 for LTVI versus LTQLac, 0Æ96 for LTAA versus LTLL, 0Æ77 for LTAA versus LTQLac and 0Æ88 for LTLL versus LTQLac. The statistical procedure suggested by Bland and Altman was used to express the degree of agreement between the different methods, expressed in percentage of 1RM (Fig. 2). The QLac method was excluded from this analysis, once it was not possible to determine the LT with this method in three subjects. For LTVI versus LTAA and LTVI versus LTLL, 100% of the values were within the limits of agreement (Fig. 2a,b). However, the bias for LTVI versus LTAA ()0Æ9%1RM) was lower than LTVI versus LTLL (3Æ6%1RM). The Bland and Altman analysis for LTAA versus LTLL shows that one subject remained outside the limits of agreement and the bias was 4Æ5% of 1RM (Fig. 2c). The mean values of blood lactate concentration at the intensity of the LT for the different methods are shown in Table 1. Mean blood lactate concentration was 1Æ91 (0Æ78) mmol L)1 in LTVI, 1Æ84 (0Æ61) mmol L)1 in LTAA, 1Æ58 (0Æ63) mmol L)1 in LTLL and 2Æ6 (1Æ4) mmol L)1 in LTQLac. Repeated measures ANOVA shows that BLC in ATLL was significantly different from the LTQLac (P<0Æ05). In addition, BLC in LTVI, LTLL and LTAA was not significantly different.

Discussion

Figure 2 Graphic analysis of intensity (%1RM) data corresponding to the LTVI versus LTAA (a), LTVI versus LTLL (b) and LTAA versus LTLL (c). LT, lactate threshold.

This is the first study to our knowledge that suggests and compares different methods for LT identification in resistance exercise with support of algorithmic and log–log alternative. This cohort confirmed that it is possible to determine the LT through an IT protocol in resistance exercise using visual inspection and mathematical models. In this sense, the algorithmic method showed more quality and precision for LT identification. However, some precautions must be taken, because were also found differences between methods. Among all of these methods for LT detection, only visual inspection (Barros et al., 2004; Simoes et al., 2010) and QLac methods (Moreira et al., 2008) were used in previous studies, showing similar results to those found in this study, with LT intensity around 30% of 1RM. It is speculated that the LT in resistance exercise should be mainly marked by a hemodynamic factor (Petrofsky et al., 1981; Williams et al., 2007; Moreira et al., 2008), in which the higher percentage of 1RM (above 30% of 1RM) promotes an intramuscular pressure that increases muscular tension and becomes higher than capillary pressure, causing their blockade (Williams et al., 2007). This blockade increases anaerobic glycolysis activity and allows for the identification of a metabolic transition of the energetic pathways.

 2011 The Authors Clinical Physiology and Functional Imaging  2011 Scandinavian Society of Clinical Physiology and Nuclear Medicine

Lactate threshold in resistance training, N. M. F. de Sousa et al. 5

However, the LT in resistance exercise may be explained by other mechanisms. Simoes et al. (2010) showed a significant augmentation in sympathetic activity at 30% of 1RM. This activity may increase the adrenal secretion of catecholamines which stimulates adrenergic receptors in muscle cells further leading to heightened glycolysis. Another mechanism that may explain the LT in resistance exercise is the progressively increasing recruitment of fast-twitch glycolytic muscle fibres along with increased load. The electromyography analysis of the active muscles may give answers for this possible mechanism. The results showed that the QLac method [31Æ6 (9Æ8) % of 1RM] calculates the LT at a significantly higher intensity than the log–log method [23Æ3 (3Æ5) % of 1RM]. In addition, QLac and log–log methods did not differ significantly from the other methods [LTVI = 26Æ9 (5Æ2)% of 1RM; LTAA = 27Æ8 (3Æ6)% of 1RM]. Moreira et al. (2008) also did not identify differences between QLac and visual inspection methods, despite the trend to higher intensities for the QLac method. The QLac method, that uses a mathematical adjustment of a second-order polynomial function, has some disadvantages compared with other methods. It was unable to identify LT in 25% of the subjects, and the standard deviation was far superior to the others. In this sense, the QLac method was excluded from the analysis of agreement between methods due the inability to identify the LT in three subjects and could be considered a weaker method. As observed, visual inspection, algorithmic adjustment and log–log methods had no significantly different results for LT detection in resistance exercise. Furthermore, the intraclass correlation coefficient between the three methods was high, showing consistency and conformity of methods. Previous researches with aerobic exercises have already shown that these three methods can be used to determine LT (Bishop et al., 1998; McMorris et al., 2000; Davis et al., 2007; Higa et al., 2007). However, although the three methods can be used for LT identification in aerobic exercises, there are some uncertainties for the algorithmic adjustment method (McMorris et al., 2000). McMorris et al. (2000) showed that the algorithmic adjustment calculated the LT at a significantly higher power output than the log–log method and that the fit provided by the algorithmic method was not as good as those provided by the log–log method. This proposition may be in accordance with the results of this study, because Bland and Altman analysis between algorithmic adjustment and log–log methods did not show concordance. Not all subjects were within the limits suggested (one subject out of limits), and the bias was about 4Æ5% of 1RM. Consequently, LTAA had higher values than the LTLL, although no statistical differences were observed. When these two methods are separately evaluated with visual inspection technique, Bland and Altman analysis showed that all subjects are within the limits of agreement. However, the mean difference (bias) between LTVI versus LTAA was lower ()0Æ9%1RM) than between LTVI versus LTLL (3Æ6%1RM). In this manner, different to what was proposed by McMorris et al. (2000) for aerobic exercises, the algorithmic adjustment method seems to be a

good mathematic model to determine the LT in resistance exercise. The search for a mathematical model attempts to reduce the disadvantage of the visual inspection method that it is almost certainly experience dependent and clearly subjective (Davis et al., 2007). The log–log method had some advantages over the visual inspection method, as its objectivity and can occur at any value on the work rate axis. Note that in Fig. 1a, the LT determined for subject no. 6 by the visual inspection method at the work rate corresponds to the third blood lactate value. However, the LT determined by the log–log method for this subject is at a work rate between the second and third blood lactate values (see Fig. 1c). The algorithmic adjustment method has also all the advantages presented for the log–log method, and it is less likely to be fooled by sample-to-sample variations in blood lactate concentration during the early stages of graded exercise testing. In this sense, the algorithmic model used in the present investigation allowed the determination of the break point in the dynamic behaviour pattern of BLC data that occurs during resistance exercise. The disadvantage of the algorithmic method is that it is more labour-intensive than the other three methods and requires competence in working with mathematical models. However, QLac and log–log methods also require some mathematical experience. The lowest values of BLC at the LT intensity found with the log–log method compared with QLac are explained by the lowest LT intensity identified by the log–log technique. Furthermore, the BLC at the LT intensity was about 2 mmol L)1, well below the 4 mmol L)1 proposed by Heck et al. (1985) for aerobic exercises, despite its known limitations (Svedahl & MacIntosh, 2003) because of high interindividual differences (Strupler et al., 2009). More investigations are necessary to analyse if the 2 mmol L)1 range is acceptable for resistance exercises or if the variability between subjects does not suggest these results. In conclusion, the inspection, algorithmic and log–log methods detected the LT in the same intensity of resistance exercise. We recommend mathematical models, specially the algorithmic method, because it provides more precision in the LT identification. Furthermore, the QLac method is less effective and precision in healthy young men. The methods that detected the LT in the same intensity of resistance exercise are apparently simple, inexpensive and reliable.

Acknowledgments The lead author was funded by the Portuguese Foundation for Science and Technology (FCT)—Ministry of Science, Technology and Higher Education (MCTES; Ref. SFRH ⁄ BD ⁄ 46898 ⁄ 2008).

Conflicts of interest The authors have no potential conflicts of interest.

 2011 The Authors Clinical Physiology and Functional Imaging  2011 Scandinavian Society of Clinical Physiology and Nuclear Medicine

6 Lactate threshold in resistance training, N. M. F. de Sousa et al.

References Barros CLM, Agostini GG, Garcia ES, Baldissera V. Limiar de lactato em exercı´cio resistido. Revista Motriz (2004); 10: 31–36. Beaver WL, Wasserman K, Whipp BJ. Improved detection of lactate threshold during exercise using a log-log transformation. J Appl Physiol (1985); 59: 1936– 1940. Bishop D, Jenkins DG, Mackinnon LT. The relationship between plasma lactate parameters, Wpeak and 1-h cycling performance in women. Med Sci Sports Exerc (1998); 30: 1270–1275. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet (1986); 1: 307–310. Burke J, Thayer R, Belcamino M. Comparison of effects of two interval-training programmes on lactate and ventilatory thresholds. Br J Sports Med (1994); 28: 18–21. Candotti CT, Loss JF, Melo Mde O, La Torre M, Pasini M, Dutra LA, de Oliveira JL, de Oliveira LP. Comparing the lactate and EMG thresholds of recreational cyclists during incremental pedaling exercise. Can J Physiol Pharmacol (2008); 86: 272–278. Davis JA. Anaerobic threshold: review of the concept and directions for future research. Med Sci Sports Exerc (1985); 17: 6–21. Davis JA, Rozenek R, DeCicco DM, Carizzi MT, Pham PH. Comparison of three methods for detection of the lactate threshold. Clin Physiol Funct Imaging (2007); 27: 381–384. Gladden LB, Yates JW, Stremel RW, Stamford BA. Gas exchange and lactate anaerobic thresholds: inter- and intraevaluator agreement. J Appl Physiol (1985); 58: 2082–2089. Goodwin ML, Harris JE, Hernandez A, Gladden LB. Blood lactate measurements and analysis during exercise: a guide for clinicians. J Diabetes Sci Technol (2007); 1: 558–569. Green HJ, Hughson RL, Orr GW, Ranney DA. Anaerobic threshold, blood lactate, and muscle metabolites in progressive exercise. J Appl Physiol (1983); 54: 1032–1038. Heck H, Mader A, Hess G, Mucke S, Muller R, Hollmann W. Justification of the 4-mmol ⁄ l lactate threshold. Int J Sports Med (1985); 6: 117–130.

Higa MN, Silva E, Neves VF, Catai AM, Gallo L Jr, Silva de Sa MF. Comparison of anaerobic threshold determined by visual and mathematical methods in healthy women. Braz J Med Biol Res (2007); 40: 501–508. Kraemer WJ, Fry AC. Strength testing: development and evaluation of methodology. In: Physiological Assessment of Human Fitness (eds Maud, PJ, Foster, C) (1995), pp. 115–138. Human Kinetics, Champaign, IL. Lucia A, Hoyos J, Santalla A, Perez M, Carvajal A, Chicharro JL. Lactic acidosis, potassium, and the heart rate deflection point in professional road cyclists. Br J Sports Med (2002); 36: 113–117. MacIntosh BR, Esau S, Svedahl K. The lactate minimum test for cycling: estimation of the maximal lactate steady state. Can J Appl Physiol (2002); 27: 232–249. McMorris T, Sproule J, Draper S, Child R, Sexsmith JR, Forster CD, Pattison J. The measurement of plasma catecholamine and lactate thresholds: a comparison of methods. Eur J Appl Physiol (2000); 82: 262–267. Moreira SR, Arsa G, Oliveira HB, Lima LC, Campbell CS, Simoes HG. Methods to identify the lactate and glucose thresholds during resistance exercise for individuals with type 2 diabetes. J Strength Cond Res (2008); 22: 1108–1115. Orr GW, Green HJ, Hughson RL, Bennett GW. A computer linear regression model to determine ventilatory anaerobic threshold. J Appl Physiol (1982); 52: 1349–1352. Petrofsky JS, Phillips CA, Sawka MN, Hanpeter D, Stafford D. Blood flow and metabolism during isometric contractions in cat skeletal muscle. J Appl Physiol (1981); 50: 493–502. Podolin DA, Munger PA, Mazzeo RS. Plasma catecholamine and lactate response during graded exercise with varied glycogen conditions. J Appl Physiol (1991); 71: 1427–1433. Ribeiro JP, Cadavid E, Baena J, Monsalvete E, Barna A, De Rose EH. Metabolic predictors of middle-distance swimming performance. Br J Sports Med (1990); 24: 196–200. Riedl I, Yoshioka M, Nishida Y, Tobina T, Paradis R, Shono N, Tanaka H, St-Amand J. Regulation of skeletal muscle transcriptome in elderly men after 6 weeks of endurance training at lactate threshold intensity. Exp Gerontol (2010); 45: 896–903.

Rocha RM, Bomfim DL, Boaventura T, Nascimento R, Moreira SR, Simo˜es HG. Variation in the incremental workload method does not change the lactate threshold determination in resistance exercise. Rev Bras Med Esporte (2010); 16: 282–285. Shephard RJ. PAR-Q, Canadian Home Fitness Test and exercise screening alternatives. Sports Med (1988); 5: 185–195. Simoes RP, Mendes RG, Castello V, Machado HG, Almeida LB, Baldissera V, Catai AM, Arena R, Borghi-Silva A. Heart-rate variability and blood-lactate threshold interaction during progressive resistance exercise in healthy older men. J Strength Cond Res (2010b); 24: 1313–1320. Strupler M, Mueller G, Perret C. Heart ratebased lactate minimum test: a reproducible method. Br J Sports Med (2009); 43: 432–436. Svedahl K, MacIntosh BR. Anaerobic threshold: the concept and methods of measurement. Can J Appl Physiol (2003); 28: 299–323. Thomas V, Costes F, Chatagnon M, Pouilly JP, Busso T. A comparison of lactate indices during ramp exercise using modelling techniques and conventional methods. J Sports Sci (2008); 26: 1387–1395. Wasserman K, McIlroy MB. Detecting the threshold of anaerobic metabolism in cardiac patients DURING exercise. Am J Cardiol (1964); 14: 844–852. Wasserman K, Whipp BJ, Koyl SN, Beaver WL. Anaerobic threshold and respiratory gas exchange during exercise. J Appl Physiol (1973); 35: 236–243. Weltman A, Snead D, Stein P, Seip R, Schurrer R, Rutt R, Weltman J. Reliability and validity of a continuous incremental treadmill protocol for the determination of lactate threshold, fixed blood lactate concentrations, and VO2max. Int J Sports Med (1990); 11: 26–32. Williams MA, Haskell WL, Ades PA, Amsterdam EA, Bittner V, Franklin BA, Gulanick M, Laing ST, Stewart KJ. Resistance exercise in individuals with and without cardiovascular disease: 2007 update: a scientific statement from the American Heart Association Council on Clinical Cardiology and Council on Nutrition, Physical Activity, and Metabolism. Circulation (2007); 116: 572–584.

 2011 The Authors Clinical Physiology and Functional Imaging  2011 Scandinavian Society of Clinical Physiology and Nuclear Medicine