1551

Braz J Med Biol Res 37(10) 2004

Anaerobic threshold and maximal lactate steady state

Effect of the aerobic capacity on thevalidity of the anaerobic threshold fordetermination of the maximal lactatesteady state in cycling

Laboratrios de 1Avaliao da Performance Humana, and2Biomecnica, Instituto de Biocincias, Universidade Estadual Paulista,Rio Claro, SP, Brasil

B.S. Denadai1,T.R. Figuera1,

O.R.P. Favaro2 andM. Gonalves2

Abstract

The maximal lactate steady state (MLSS) is the highest blood lactateconcentration that can be identified as maintaining a steady stateduring a prolonged submaximal constant workload. The objective ofthe present study was to analyze the influence of the aerobic capacityon the validity of anaerobic threshold (AT) to estimate the exerciseintensity at MLSS (MLSS intensity) during cycling. Ten untrainedmales (UC) and 9 male endurance cyclists (EC) matched for age,weight and height performed one incremental maximal load test todetermine AT and two to four 30-min constant submaximal load testson a mechanically braked cycle ergometer to determine MLSS andMLSS intensity. AT was determined as the intensity corresponding to3.5 mM blood lactate. MLSS intensity was defined as the highestworkload at which blood lactate concentration did not increase bymore than 1 mM between minutes 10 and 30 of the constant workload.MLSS intensity (EC = 282.1 23.8 W; UC = 180.2 24.5 W) and AT(EC = 274.8 24.9 W; UC = 187.2 28.0 W) were significantly higherin trained group. However, there was no significant difference inMLSS between EC (5.0 1.2 mM) and UC (4.9 1.7 mM). The MLSSintensity and AT were not different and significantly correlated in bothgroups (EC: r = 0.77; UC: r = 0.81). We conclude that MLSS and thevalidity of AT to estimate MLSS intensity during cycling, analyzed ina cross-sectional design (trained x sedentary), do not depend on theaerobic capacity.

CorrespondenceB.S. Denadai

Laboratrio de Avaliao da

Performance Humana

IB, UNESP

Av. 24A, 1515

13506-900 Rio Claro, SP

Brasil

E-mail: bdenadai@rc.unesp.br

Research supported by CNPq and

FAPESP.

Received September 29, 2003

Accepted June 14, 2004

Key words Anaerobic threshold Lactate steady state Aerobic capacity Cycling

The maximal lactate steady state (MLSS)is the highest blood lactate concentration(BLC) that can be identified as maintaining asteady state during prolonged submaximalconstant workload. The MLSS has been usedto determine the highest intensity (MLSSintensity) that can be maintained over timewithout continuous blood lactate accumula-

tion (1,2). There are many reasons for tryingto quantify this intensity of exercise, includ-ing assessment of cardiovascular or pulmo-nary health, evaluation of training programs,prescription of appropriate exercise inten-sity, and categorization of the intensity ofexercise as moderate, heavy or severe (3-5).However, the determination of MLSS inten-

Brazilian Journal of Medical and Biological Research (2004) 37: 1551-1556ISSN 0100-879X Short Communication

1552

Braz J Med Biol Res 37(10) 2004

B.S. Denadai et al.

sity requires the subject to perform 4-5 exer-cise bouts of 30-min duration, preferably onseparate days. This procedure is time-con-suming and can interfere with the training ofthe athlete.

In contrast to the latter procedure, somestudies have proposed to predict the MLSSintensity within a single incremental loadtest. Heck et al. (6) showed in a heteroge-neous group (endurance runners and activeindividuals) that the intensity obtained dur-ing an incremental test corresponding to 4mM blood lactate (anaerobic threshold) isvalid to indirectly determine MLSS intensityduring running. Subsequently, other studiesconducted on endurance runners (7) andsoccer players (8) confirmed the validity ofanaerobic threshold to estimate MLSS inten-sity during running. However, Beneke (9)showed that the use of the anaerobic thresh-old overestimated MLSS intensity in rowerswith different performance levels. Differ-ences in muscle mass involved in these typesof exercise (running x rowing) might ex-plain, in part, these contradictory data.Beneke (10) has proposed that the MLSSdepends upon the motor pattern of exercisecaused by task-specific interrelationshipsbetween power output per unit muscle massand the specific masses of the primarily en-gaged muscles. To our knowledge, no stud-ies have compared anaerobic threshold withthe directly and independently determinedMLSS intensity during cycling.

Some studies have developed the con-cept of individual anaerobic threshold, whichis based on the hypothesis that the BLC atanaerobic threshold may decrease with in-creasing performance capacity (11,12). How-ever, the physiological mechanisms and ex-perimental proof of this hypothesis are miss-ing. Possible effects of aerobic capacity onthe validity of anaerobic threshold for esti-mating MLSS intensity can only be evalu-ated by using constant load tests and thedirect determination of the MLSS.

The objective of the present study was to

analyze the influence of the aerobic capacityon the validity of anaerobic threshold as anestimate of MLSS intensity during cycling.

Ten untrained males (UC) (22.6 4.1years, 71.5 12.9 kg, 175.1 4.3 cm), and 9endurance cyclists (EC) (20.6 2.3 years,69.1 9.9 kg, 177.5 5.0 cm) volunteered toparticipate in this study. The subjects gaveinformed consent and the protocol was ap-proved by the Biosciences Institute EthicsCommittee, UNESP, Rio Claro, SP, Brazil.

All subjects performed one incrementalmaximal load test to determine anaerobicthreshold and 2 to 4 constant submaximalload tests on a mechanically braked cycleergometer (Monark, So Paulo, SP, Brazil)to determine MLSS. The pedaling rate wasset at 70 rpm. The interval between the twotests was at least 48 h. The subjects wereinstructed to arrive at the laboratory in arested and fully hydrated state, at least 3 hpost-prandial, and to avoid strenuous exer-cise in the 48-h preceding a test session.Each subject was tested at the same time ofday (9:30 1:00 h) to minimize the effects ofdiurnal biological variation.

The incremental load test started with 70W for the UC and 105 W for the EC and wasincreased to exhaustion by 35 W by every3rd minute. The peak workload (PW) wasdefined as the highest workload maintainedat least 1 min. The anaerobic threshold wasdetermined as the intensity corresponding to3.5 mM blood lactate (6). The constant loadtests lasted 30 min. Workload of the firstconstant load test corresponded to a BLC of3.5 mM measured during the incrementalload test. If during the first constant load testa steady state or a decrease in lactate wasobserved, further subsequent 30-min con-stant load tests with 3-7% higher workloadintensities were performed on separate daysuntil no BLC steady state could be main-tained. If the first constant load test resultedin a clearly identifiable increase of the BLCand/or could not be completed due to ex-haustion, further constant workloads were

1553

Braz J Med Biol Res 37(10) 2004

Anaerobic threshold and maximal lactate steady state

conducted with subsequently reduced work-load intensities. The MLSS intensity wasdefined as the highest workload at whichBLC did not increase by more than 1 mMbetween minutes 10 and 30 of the constantworkload (1). MLSS was calculated as aver-age value of the BLC measured at minutes10 and 30 of the MLSS intensity (1).

Blood samples (25 l) were collectedfrom the ear lobe into microcentrifuge tubescontaining 50 l NaF (1%) during the final15 s of every 3rd (incremental load test) orevery 5th (constant load test) minute. BLCwas determined by an electrochemicalmethod (YSL 2300 STAT, Yellow Springs,OH, USA).

Data are reported as means SD. Thepaired t-test was applied to compare anaero-bic threshold and MLSS intensities. The cor-relations between anaerobic threshold andMLSS intensities were calculated usingPearson Product Moment correlation coeffi-cients. In addition, the bias and limits ofagreement between anaerobic threshold andMLSS intensities were calculated (13). Dif-ferences between the trained and untrainedgroups were tested by non-paired t-test. Sig-nificance was set at P 0.05.

The PW, MLSS intensity, and anaerobicthreshold reported as absolute as well asrelative values related to body mass, and alsoMLSS intensity as percent of PW (%MLSSintensity) were significantly higher in ECthan in UC. However, there was no signifi-cant difference in MLSS between EC andUC (Table 1). There were no significantdifference between MLSS intensity andanaerobic threshold in both groups (Table1). MLSS intensity was significantly corre-lated with anaerobic threshold in both groups(EC: r = 0.77; UC: r = 0.81). Moreover, thebias 95% limits of agreement for compari-sons between MLSS intensity and anaerobicthreshold in EC [7.2 (16.4) W], UC [-7.0(16.4) W] and EC + UC [-0.3 (17.5) W],suggest the validity of anaerobic threshold toestimate MLSS intensity during cycling in

both groups (Figure 1).The main objective of the present study

was to analyze the influence of the aerobiccapacity on the validity of anaerobic thresh-old to estimate MLSS intensity during cy-cling. In contrast to previously publishedspeculations that higher aerobic capacity re-duces the BLC at anaerobic threshold (11,12,14), we showed here that the anaerobic

Diff

eren

ce b

etw

een

MLS

S in

tens

ity a

nd A

T (W

)

50

40

30

20

10

0

-10

-20

-30

-40

-50130 150 170 190 210 230 250 270 290 310 330

Mean MLSS intensity and AT (W)

Figure 1. Bland-Altman diagram comparing exercise intensity at anaerobic threshold (AT)and maximal lactate steady state (MLSS intensity) for all subjects (untrained and trained).The solid horizontal line represents the bias between the two measures (i.e., the meandifference between the group means for the two variables). The dashed lines representthe 95% limits of agreement between the two variables, and reflect the extend to whichone variable might be expected to differ from another in individual subjects.

Table 1. Maximal and submaximal responses obtained in endurance cyclists anduntrained subjects.

Endurance cyclists Untrained cyclists(N = 9) (N = 10)

Peak workload (W) 355.1 27.7 267.3 39.5*Peak workload (W/kg) - relative 5.2 0.7 3.7 0.5*MLSS (mM) 5.0 1.2 4.9 1.7MLSS intensity (W) 282.1 23.8 180.2 24.5*MLSS intensity (W/kg) - relative 4.1 0.7 2.5 0.4*AT (W) 274.8 24.9 187.2 28.0*AT (W/kg) - relative 4.0 0.6 2.6 0.5*%MLSS intensity (%) 79.5 4.1 68.0 9.5*

W, Watts; MLSS, maximal lactate steady state; MLSS intensity, intensities at MLSS;AT, anaerobic threshold; %MLSS intensity, MLSS intensity as percent of peak work-load; relative, values related to body mass. Data are reported as means SD.*P 0.05 compared to endurance cyclists (non-paired t-test).

1554

Braz J Med Biol Res 37(10) 2004

B.S. Denadai et al.

threshold presents a good validity for theestimate of MLSS intensity irrespective ofaerobic capacity. In agreement with Benekeet al. (15), we also found that the MLSS doesnot depend on aerobic capacity during cy-cling.

The PW (W/kg), MLSS intensity and%MLSS intensity values obtained for the ECgroup were compatible with those reportedin the literature for individuals classified aswell trained (16). In addition, these valueswere significantly higher in the EC than inthe UC group. As the MLSS intensity isconsidered the gold standard for the as-sessment of aerobic capacity (4), our objec-tive of comparing groups with different lev-els of aerobic training (trained x sedentary)was achieved.

Different methods for the identificationof MLSS and, consequently, MLSS intensityhave been reported in the literature. Thesemethods basically differ with respect to testduration and the period of constant work-load selected for the interpretation of theBLC response and the maximally acceptedincrease of BLC. In the present study, weused a method employed by different re-search groups (2,10), whose validity has beenrecently confirmed by Beneke (1). The MLSSvalues obtained here are similar to thosereported in other studies on cycling (1,10,15).

Many investigations have shown that at agiven submaximal exercise intensity, ex-pressed as absolute (i.e., km or W) or relativevalues (i.e., %VO2max or %PW), the BLC islower in trained than in sedentary individu-als (17). This decline in BLC after a periodof training can be observed even in highlytrained athletes, and in the absence of anychanges in VO2max (18). The BLC dependson the dynamic equilibrium between lactateappearance in and disappearance from theblood compartment. Bergman et al. (17)showed that mechanisms for dampened arte-rial lactate concentration after endurancetraining vary depending on exercise inten-sity. At the same absolute workload, endur-

ance training decreases whole body and work-ing muscle lactate production and increasesclearance by active muscle. However, at simi-larly high relative exercise intensities, en-durance training increases whole body andactive muscle lactate clearance, but does notinfluence whole body or muscle production.

Although these modifications in the lac-tate response to exercise might exist, theMLSS does not seem to depend on aerobiccapacity. Similar to the results of the presentstudy in which the MLSS did not differbetween groups (EC = 5.0 1.2 mM vs UC =4.9 1.7 mM), Beneke et al. (15) reportedthat MLSS (4.9 1.4 mM) was independentof MLSS intensity (3.4 0.6 W/kg). Thus,factors other than the training status (e.g.,exercise type) (10) seem to influence MLSS.

To our knowledge, this paper reports thefirst study analyzing the influence of theaerobic capacity on the validity of anaerobicthreshold to estimate MLSS intensity duringcycling. Investigations involving a heteroge-neous group (endurance runners and activeindividuals; 6), endurance athletes (7) andsoccer players (8) showed that anaerobicthreshold presents good validity in estimat-ing MLSS intensity during running. In thepresent study, we demonstrated that the va-lidity of anaerobic threshold to estimateMLSS intensity during cycling does not de-pend on the level of training. In contrast,Urhausen et al. (19) reported that anaerobicthreshold is not a valid method to estimateMLSS intensity during cycling, although thecited authors did not directly measure MLSS.In that study, 7 (43%) of 16 athletes did notpresent a lactate steady state during constantload exercise performed at anaerobic thresh-old. In contrast, in the present study only 3(30%) subjects of the UC group and onesubject (11%) of the EC group showed in-tensities corresponding to anaerobic thresh-old, which were 5% higher than the MLSSintensity. This percentage (5%) was used inthe present study, and also by Urhausen et al.(19), to determine differences in intensities

1555

Braz J Med Biol Res 37(10) 2004

Anaerobic threshold and maximal lactate steady state

between constant load exercises, i.e., thelevel of precision that identified the pres-ence or absence of lactate steady state. Dif-ferences in the increment rate of the incre-mental protocol (11.6 x 16.6 W/min) and inthe lactate concentration corresponding toanaerobic threshold (3.5 x 4.0 mM) usedhere and by Urhausen et al. (19), respec-tively, might explain, in part, these appar-ently contradicting data. Heck (20) demon-strated an increase of approximately 1.4 Win anaerobic threshold when the increase inworkload during incremental load test wasincreased by 1.0 W/min. In addition, proto-cols with lower stages (e.g., 3 min) shoulduse 3.5 instead of 4.0 mM to identify anaero-bic threshold (6).

Based on the data reported in some stud-ies (11,12,14), it has been proposed (4) thatthe validity of anaerobic threshold dependson the athletes level of aerobic training. Forexample, Mognoni et al. (14) showed thatanaerobic threshold (i.e., aerobic fitness) wasnegatively correlated with the maximum du-

ration of exercise performed at anaerobicthreshold intensity (4 mM) during cycling.However, we believe that these data do notdemonstrate that the validity of anaerobicthreshold to estimate MLSS intensity is com-promised in endurance athletes. In fact, theconcepts and mechanisms that can influencethe MLSS and maximum exercise durationat a given exercise intensity (e.g., at MLSSintensity) may be different, and have not yetbeen completely established. Thus, our datado not support previously published specu-lations that higher aerobic capacity reducesthe BLC at anaerobic threshold (11,12,14).

We conclude that MLSS and the validityof anaerobic threshold to estimate MLSSintensity during cycling, analyzed in a cross-sectional design (trained x sedentary), do notdepend on the aerobic capacity. However,longitudinal studies are necessary to analyzethe effects of different training programs(i.e., different intensities and volumes) onMLSS and on the validity of anaerobic thresh-old to estimate MLSS intensity.

References

1. Beneke R (2003). Methodological aspects of maximal lactate steadystate - implications for performance testing. European Journal ofApplied Physiology, 89: 95-99.

2. Pringle JS & Jones AM (2002). Maximal lactate steady state, criticalpower and EMG during cycling. European Journal of Applied Physi-ology, 88: 214-226.

3. Londeree BR (1997). Effect of training on lactate/ventilatory thresh-olds: a meta-analysis. Medicine and Science in Sports and Exercise,29: 837-843.

4. Billat V, Sirvent P, Py G, Koralsztein JP & Mercier J (2003). Theconcept of maximal lactate steady state: a bridge between bio-chemistry, physiology and sport science. Sports Medicine, 33: 407-426.

5. Gaesser GA & Poole DC (1996). The slow component of oxygenuptake kinetics in humans. Exercise and Sport Sciences Reviews,24: 35-70.

6. Heck H, Mader A, Hess G, Mucke S, Muller R & Hollmann W (1985).Justification of the 4 mmol/l lactate threshold. International Journalof Sports Medicine, 6: 117-130.

7. Simes HG, Grubert Campbell CS, Kokubun E, Denadai BS &Baldissera V (1999). Blood glucose responses in humans mirrorlactate responses for individual anaerobic threshold and for lactateminimum in track tests. European Journal of Applied Physiology,80: 34-40.

8. Denadai BS, Gomide E & Greco CC (2003). Validity of critical ve-locity to determine the maximal lactate steady state in soccerplayers. 5th World Congress on Science and Football, Lisboa, Portu-gal, April 11-15.

9. Beneke R (1995). Anaerobic threshold, individual anaerobic thresh-old, and maximal lactate steady state in rowing. Medicine andScience in Sports and Exercise, 27: 863-867.

10. Beneke R (2003). Maximal lactate steady state concentration(MLSS): experimental and modelling approaches. European Journalof Applied Physiology, 88: 361-369.

11. Simon G, Berg A, Dicckhuth H, Simon-Alt A & Keul J (1981). Deter-mination of anaerobic threshold depending on age and performancecapacity. Deutsche Zeitschrift fr Sportmedizin, 32: 7-14.

12. Stegmann H, Kindermann W & Schnabel A (1981). Lactate kineticsand individual anaerobic threshold. International Journal of SportsMedicine, 2: 160-165.

13. Bland JM & Altman DG (1986). Statistical methods for assessingagreement between two methods of clinical measurement. Lancet,1: 307-310.

14. Mognoni P, Sirtori MD, Lorenzelli F & Cerretelli P (1990). Physiologi-cal responses during prolonged exercise at the power output corre-sponding to the blood lactate threshold. European Journal of Ap-plied Physiology, 60: 239-243.

15. Beneke R, Hutler M & Leithauser RM (2000). Maximal lactate-

1556

Braz J Med Biol Res 37(10) 2004

B.S. Denadai et al.

steady-state independent of performance. Medicine and Science inSports and Exercise, 32: 1135-1139.

16. Caputo F, Stella S, Mello MT & Denadai BS (2003). Indexes ofpower and aerobic capacity obtained in cycle ergometry and tread-mill running: Comparisons between sedentary, runners, cyclistsand triathletes. Revista Brasileira de Medicina do Esporte, 9: 1-8.

17. Bergman BC, Wolfel EE, Butterfield GE, Lopaschuk GD, CasazzaGA, Horning MA & Brooks GA (1999). Active muscle and wholebody lactate kinetics after endurance training in men. Journal of

Applied Physiology, 87: 1684-1696.18. Coyle EF (1995). Integration of the physiological factors determining

endurance performance ability. Exercise and Sport Sciences Re-views, 23: 25-63.

19. Urhausen A, Coen B, Weiler B & Kindermann W (1993). Individualanaerobic threshold and maximum lactate steady state. Interna-tional Journal of Sports Medicine, 14: 134-139.

20. Heck H (1990). Laktat in der Leistungsdiagnostik. Hofmann,Schorndorf, Germany.