Entries in Lactate (4)


Estimating the Maximal Lactate Steady State Power from an Incremental Test Using Lactate Pro LP1710

International Journal of Applied Sports Sciences 2009, Vol. 21, No. 1, 74-85.
Korea Institute of Sport Science
Received : 1 December 2008, Accepted : 10 February 2009

By: Asgeir Mamen, & Roland van den Tillaar Sogn og Fjordane
University College, Norway

The aim of the study was to explore how the Lactate Pro® LP1710 (LP1710) blood lactate analyser can be used to find the Maximal Lactate Steady State (MLSS) power from an incremental cycle test. Methods: Nine cyclists were tested.

They performed an incremental test to establish a power vs. blood lactate concentration (BLC) curve and find two threshold definitions: Lactate Breakpoint (LB) and Onset of Blood Lactate Accumulation (OBLA). Then several continuous load tests were performed to establish the power output that elicited MLSS power (WMLSS). Results: From the blood lactate curve of the incremental test a BLC of 2.7(0.6) mmol․ L-1 (or 1.7(0.6) mmo․ L-1 above resting BLC) equalled the WMLSS. The W, HR and VO2 from the LB and OBLA tests had a range of 91-93% (LB) and 103-111% (OBLA) of MLSS values. The LB produced the lowest power, 228(46) W, 18 W below WMLSS (p=0.005), OBLA the highest, 273(47) W, 27 W above WMLSS (p=0.007).

The oxygen uptake at LB was 70(5)% of VO2max, at MLSS 77(5)% and OBLA 82(3)%. Conclusion: From the results of an incremental test the WMLSS can be estimated with LP1710, if a fixed BLC of 2.7 ± 0.7 mmol․ L-1 is used.

key words: MLSS, OBLA, Lactate Breakpoint, blood lactate concentration, training 


The definition of the lactate threshold is the highest load a subject can endure over time without a rise in blood lactate concentration (BLC), often called Maximal Lactate Steady State (MLSS) (Weltman, 1995). This load can be found by performing several constant load tests, usually of 30 minutes duration (Beneke, 2003; Harnish et al., 2001). Such a method is time-consuming. Therefore, several indirect methods have been developed, mostly using an incremental protocol with step durations of 3-8 minutes (Weltman, 1995). These tests produce a power vs.BLC curve that can be used to define the point of threshold using some criteria.

Unfortunately, no agreement on criteria for threshold determination exists (Weltman, 1995). Two common ways of defining a threshold are the Lactate breakpoint (LB), which is the first increase in BLC > 1.0 mmol․L-1 seen in an incremental protocol (Davis et al., 1983) and the Onset of Blood Lactate Accumulation (OBLA), which Sjödin defined as a fixed BLC level of 4 mmol․L-1 (Sjödin & Jacobs, 1981b).

When comparing results of threshold tests, several factors need to be taken into consideration. point to stage duration and size of load increment as factors that may affect the test result, and recommend the use of 3-minute stages Bentley (Bentley, Newell, & Bishop, 2007). Foxdal have pointed out that if BLCs from incremental and continuous tests are to be compared, the stage duration of the incremental test should be eight minutes, to assure a steady state of lactate (Foxdal et al., 1996).

The measured BLC is dependent on sampling site and type of blood (El-Sayed et al., 1993; Feliu et al., 1999; Foxdal et al., 1990). Finger sampling produces higher BLC than ear lobe sampling (Feliu et al., 1999). El-Sayed found that the OBLA load calculated from venous blood was too high for the subjects to sustain with stable BLC, whereas the OBLA load from capillary blood could be endured in a BLC steady state (El-Sayed et al., 1993). Results from tests using different sampling sites should thus be compared with caution. When using resting BLC as a baseline for threshold determination, sampling site might seriously influence the result, as the difference between ear and finger sampling is greatest at low BLCs (Feliu et al., 1999).

Foxdal concluded that direct comparisons between BLC in capillary finger blood, venous whole blood and plasma could not be made (Foxdal et al., 1990). Even more caution should be exercised when different lactate analysers have been used as the BLC result is analyser specific. Medbø et al.(Medbø et al., 2000) and Buckley and co-workers (Buckley et al., 2003) have shown that the BLC will differ between lactate analysers, so that results from one brand of analyser are not always interchangeable with another analyser. Medbø et al. (2000) compared the LP1710 with several YSI 1500 sport analysers and found that the LP1710 blood lactate results was ~40% higher than the YSI results (Y(LP1710)=-0.21+(1.50․X(YSI)). A difference between analysers is especially important to be aware of for threshold determinations that use a fixed BLC (as OBLA). A difference between two analysers of 10% might satisfy the OBLA criterion in one analyser, but gives only 3.6 mmol․L-1 in the other. Threshold determinations based on a relative change in BLC (as LB), seem not to be as prone to analyser specificity (Buckley et al., 2003), neither should it be so sensitive to variation in sampling site as the calculations are not dependent on absolute BLC values.

The fitness level can also influence the result of an incremental test. Bentley found that well-trained and lesser-trained cyclists responded differently to incremental protocols with short (3 min) and long (8 min) step duration (Bentley et al., 2001). Results from a low BLC threshold (Lactate Threshold; first increment above resting BLC) differed between the two-step durations in the well-trained group, but not in the lesser-trained group. OBLA results were not affected by step duration in either fitness groups.

According to Bentley no threshold definition can be termed “best” (Bentley et al., 2007). The selection of a diagnostic test must be made according to specific needs, but coaches and athletes should be aware of how changes in the test protocol, analysers included, can influence the result.

Thus there is a need to investigate how a specific lactate analyser behaves with respect to specific lactate threshold tests and MLSS. The lactate analyser Lactate Pro® LP1710 (Arkray inc, Japan) (LP1710) has become popular due to both size and pricing. It is easy to use and requires little blood and consequently is well suited for both laboratory and field measurements. Several investigators have examined this analyser and found it accurate and reliable (Buckley et al., 2003; Mc Naughton et al., 2002; McLean et al., 2004; Medbø et al., 2000; Pyne et al., 2000; van Someren et al., 2005).

The aim of this study is therefore to investigate how the MLSS power can be estimated from the results of an incremental lactate profile test using capillary blood from the finger and the LP1710 in cycling.


Material and Methods


Nine male subjects participated in this experiment (Table 1). All were physically active with cycling as their main activity. This study complied with the requirements of the Helsinki declaration and with current Norwegian law and regulations. The subjects signed an informed consent that stated that they participated of free will, and could leave the project at any time without explaining why.



A resting blood lactate sample was taken from the fingertip before warming up.

The subjects performed an incremental lactate profile test with 5 min step duration and an increase in load of 30 W per step from a starting load of ~1.5 W kg-1 body mass, giving a slope of 6 W․min-1. BLC was measured at the end of each step, again by pricking a finger. There were no resting periods during the blood sampling. When BLC exceeded 4 mmol․L-1, the test was terminated and after a resting period of ~20 min, a maximal oxygen uptake test started. In this test load was increased by 30 W each min from a load of approximately 2 W kg-1 body mass until voluntary exhaustion.

The load at LB (WLB) was defined in each individual as the load preceding an increase in BLC of >1.0 mmol․L-1 (Davis et al., 1983). The load at a BLC of 4.0 mmol․L-1 was defined as the OBLA load (WOBLA) (Sjödin & Jacobs, 1981).

Within a week after the incremental test, the MLSS testing was started. The first 30 min constant load test had a load ~12% below WOBLA. This load would in most cases be below the MLSS power (WMLSS), but so close that only two or three constant load tests would be required to reach the WMLSS. Blood samples were taken at 0, 5, 10, 20 and 30 min of exercise. According to the development of the BLC, the load was either raised or lowered for the next constant load test by 10 W, continuing until a steady increase in BLC was obtained during the test or the BLC did not increase more than the MLSS definition allowed. The highest load that resulted in a BLC increase less than 1 mmol․L-1 during the last 20 min was considered the WMLSS (Harnish et al., 2001). The heart rate of the MLSS test (HRMLSS) was the mean HR from the 15th to the 20th min of exercise. Oxygen consumption (VO2MLSS) was the mean VO2 from the 15th to the 20th min. This sampling interval was chosen to avoid possible drift in HR and VO2.

BLC was measured with a Lactate Pro LP1710® (Arkray Inc, Kyoto, Japan). It is a small size analyser (84x55x14.5 mm, weight 50g) that uses lactate oxidase and K ferricyanide to measure the lactate content of whole blood. Values are displayed as haemolysed values. Only a small amount of blood, 5 μL, is necessary. The manufacturer claims a coefficient of variation (CV) of ~3%.

The cyclists used their own bikes mounted on a Computrainer PC1 electromagnetic roller (RacerMate Inc, Seattle, USA). As load is independent of cadence, they were free to choose pedalling frequency. For one person, a Monark mechanical ergometer cycle (Monark Exercise AB, Varberg, Sweden) was used for all testing, as his own bike was not available. In his case, a cadence of 75 RPM was used throughout. Oxygen consumption was measured with a MetaMax CBS metabolic cart (Cortex Biophysik, Leipzig, Germany) at 10 s intervals throughout the test. Heart rate was assessed with a Polar heart rate monitor (Polar Electro OY, Kempele, Finland) every 5th second 

Statistical analysis

Results are presented as mean (SD) unless otherwise stated. An ANOVA for repeated measurements was used to compare power, heart rate, oxygen uptake and percent of HRmax and VO2max between the three definitions. When a significant difference was found, a Holm-Sidak post hoc test was performed. Data normality was investigated both with Kolmogorov-Smirnov (Lilliefors modification) and Shapiro-Wilk's tests due to the low number of subjects. If any of the normality tests failed, a Kruskal-Walis ANOVA on ranks was performed with Tukey post hoc test.

Pearson's r was used for correlations. Level of statistical significance was set to p<0.05. Statistical software: SigmaPlot 10/SigmaStat 3.5 (Systat Software Inc, San Jose, CA, USA), Winks 4.80 (TexaSoft, Cedar Hill, TX, USA) and Mystat 12 (Systat software, Inc, Chicago, IL, USA). 


The power that equalled the WMLSS at the incremental test corresponded to a BLC of 2.7(0.6) mmol․L-1. When exercising at WMLSS at the constant load test, the BLC was 3.4(0.7) mmol․L-1. The mean resting BLC was 1.0(0.2) mmol․L-1. A low BLC on the first load of the incremental test (1.1(0.3) mmol․L-1) indicated that the starting load was suitable for the subjects.

An ANOVA on repeated measurements showed significant differences for BLC (F2,8=58.8, p=0.001), W (F2,8=51.1, p=0.001), HR (F2,8=30.1, p=0.001) and VO2 (F2,8=28.3, p= 0.001) between the three definitions. The LB definition gave the lowest results; the MLSS results were intermediate and the OBLA results highest (Table 2).


The LB threshold occurred after the second to fourth load, and had an average BLC of 1.8(0.5) mmol․L-1. The 4.0 mmol․L-1 BLCOBLA was reached after the third to sixth load. LB results were 91-93% of the MLSS values, whereas OBLA values were 103-111% of them (see Figure 1). The different conditions correlated significantly for absolute values, but not for relative values, as expected. The three conditions did correlate highly with VO2max, but not with %VO2max. See Table 3a and b.

Figure 1. Box-plot of threshold results. Boxes represents 25 to 75 percentile, whiskers are 5th and 95th percentile. Horizontal line is median. LB=lactate breakpoint, OBLA=onset of blood lactate accumulation, W=watt, HR=heart rate, VO2 =oxygen uptake.




Our main finding is that the load corresponding to WMLSS can be estimated from the results of an incremental test with the LP1710 by using a fixed BLC of 2.7(0.6) mmol․L-1 or alternatively, using a delta value of 1.7(0.6) to the resting BLC.

When using the threshold definitions of LB and OBLA, the load has to be heightened by 8% (WLB) or lowered by 11% (WOBLA) to equal WMLSS. Equally, HRLB and HROBLA, were 8% and 3% to low/high respectively compared with the MLSS results.

From incremental tests the MLSS results can be found by adding or subtracting from LB or OBLA data. The workload, heart rate and oxygen uptake from the LB definition had values of 91% to 93% of the MLSS values. As LB is a threshold definition that uses a relative BLC change, the results are probably less dependent of the analyser used (Buckley et al., 2003). Such a threshold definition is on the other hand sensitive to sampling site (Feliu et al., 1999), so finger sampling must be used for comparison with our data. The OBLA W, HR and VO2 were from 3% to 11% higher than MLSS values. OBLA is thought to equal MLSS (Heck et al., 1985), but the finding that OBLA exceeds MLSS is not unique. In an investigation by Laursen et al. (Laursen et al., 2002) on ultra-endurance athletes, the power output during a 5 h triathlon was 69% of the secondary ventilatory threshold (VT2). This threshold is regarded as being close to the OBLA threshold (Lucia et al., 1999), indicating that OBLA would over-estimate ultra-marathon performance.

Welde and co-workers (Welde et al., 2003) used a LP1710 for OBLA determination in running and skiing, and found that well-trained female skiers had an average of 95% of VO2OBLA during a six km simulated ski competition lasting less than 25 min, indicating that the OBLA values found by the LP1710 would over-estimate the athlete. Higher OBLA results compared with MLSS results were also found by Stegmann (Stegmann & Kindermann, 1982) in rowers, and Aunola, using cycling as the form of exercise (Aunola & Rusko, 1992). HROBLA was 89% of HRmax. That value is significantly lower than what Impellizzeri (Impellizzeri & Marcora, 2007) found (p=0.009) using highly trained MTB cyclists. They reported a HROBLA of 93% of HRmax. It is known that highly trained endurance athletes can utilise a larger proportion of their aerobic power over time than less trained individuals (Wilmore & Costill, 2004), so training status may explain the difference we see in our data. This is further highlighted by the findings of Chicharro et al. who compared professional and amateur cyclists on HR, VO2 and W at OBLA and at a heart rate of 175 (HR175), (Chicharro et al., 1999). For professional cyclists OBLA results were significantly higher than at HR175, p>0.01, but no difference was found for the lesser trained amateur cyclists, documenting the effect of training status on threshold performance as HRmax did not differ significantly between the groups.

Foxdal, compared the BLC of 50 min continuous runs at the OBLA speed (4 mmol․L-1) determined by steps protocols of different durations (4 to 8 min step duration) (Foxdal et al., 1996). They found that the continuous runs gave higher BLC than the incremental tests if step duration was shorter than eight minutes and warned that OBLA results from incremental tests with step durations shorter than 8 min would over-estimate MLSS performance. This is compatible with our finding; the BLC during the MLSS test was 26% higher than the BLC from the incremental test that we found could reproduce the WMLSS. This difference may be due to how the lactate is distributed in blood during exercise (Medbø & Toska, 2001).

Our subjects must be classified as fit with a mean VO2max of nearly 60 ml․kg-1․min-1 and Bentley found that well-trained cyclists had different “low BLC threshold” (Lactate Threshold; first increment above resting BLC) in incremental tests of 3 and 8 min step duration (Bentley et al., 2001). Recreational cyclists, on the other hand, performed equally on both tests. Dividing our group into “high level” (VO2max >64 ml․kg-1․min-1) and ”low level” fitness (VO2max <65 ml․kg-1․min-1), no difference was found in LB or OBLA results. This discrepancy with the results of (Bentley et al., 2001) may be caused by a higher fitness level of their trained subjects, or may be due to the fact that the duration of our incremental test, five minutes, was long enough to eradicate any differences in response. It’s also important to note that the two threshold definitions are not equal; the BLC from the LT definition is probably lower than our LB definition, thereby making exact comparison difficult.

Given the large variability of resting BLC in the small sample of the current study (range 0.8 mmol․L-1), and the fact that a threshold definition based on relative changes in BLC are less analyser sensitive, an approach that uses a fixed level of BLC does not seem to be preferable. By adding 1.7 ± 0.6 mmol․L-1 to resting BLC, the power corresponded to WMLSS. It is important to note that the spread of this delta value was large in our group, from 0.8 to 3.0 mmol․L-1, so the use of a fixed delta value will therefore lead to underestimation of some, and overestimation of others.

All conditions correlated highly with aerobic power, (table 3b, p<0.02). The highest correlation was seen in the WOBLA, r=0.89, which has the highest BLC. WLB had r=0.78 and the lowest BLC. Relating power with %VO2max did change the situation, and none of the relations were statistically significant (r<0.35). Our results are in line with Tokmakidis and co-workers (Tokmakidis et al., 1998) who were unable to find a unique BLC that had a superior correlation with performance compared to other levels of BLC, indicating that it is the profile of the whole curve that matters (Bentley et al., 2007).

Despite several attempts to develop a simple method to estimate the MLSS power (Billat et al., 1994; Harnish et al., 2001; Van et al., 2004), if the coach and athlete need high accuracy in their testing, the time consuming MLSS test protocol has to be applied.


It is possible to estimate the WMLSS from an incremental test with the LP1710 analyser. A fixed BLC of 2.7(0.6) mmol․L-1 or a delta value of 1.7(0.6) mmol․L-1 added to the resting BLC gives the WMLSS. These results are valid for incremental tests with step durations of five minutes, and finger sampling. MLSS power can also be derived from LB and OBLA test results. The values here found are analyser specific and may induce errors in the diagnostics of athletes if applied to other brands of analysers.


Aunola, S., & Rusko, H. K. (1992). Does anaerobic threshold correlate with maximal lactate steady-state? Journal of Sports Science, 10, 309-323.

Beneke, R. (2003). Methodological aspects of maximal lactate steady state-implications for performance testing. European Journal of Applied Physiology, 89, 95-99.

Bentley, D. J., McNaughton, L. R., & Batterham, A. M. (2001). Prolonged stage duration during incremental cycle exercise: effects on the lactate threshold and onset of blood lactate accumulation. European Journal of Applied Physiology, 85, 351-357.

Bentley, D. J., Newell, J., & Bishop, D. (2007). Incremental exercise test design and analysis: implications for performance diagnostics in endurance athletes. Sports Medicine, 37, 575-586.

Billat, V., Dalmay, F., Antonini, M. T., & Chassain, A. P. (1994). A method for determining the maximal steady state of blood lactate concentration from two levels of submaximal exercise.

European Journal of Applied Physiology and Occupational Physiology, 69, 196-202.

Buckley, J. D., Bourdon, P. C., & Woolford, S. M. (2003). Effect of measuring blood lactate concentrations using different automated lactate analysers on blood lactate transition thresholds. Journal of science and medicine in sport / Sports Medicine Australia, 6, 408-421.

Chicharro, J. L., Carvajal, A., Pardo, J., Perez, M., & Lucia, A. (1999). Physiological parameters determined at OBLA vs. a fixed heart rate of 175 beats x min-1 in an incremental test performed by amateur and professional cyclists. Japan Journal of Physiology, 49, 63-69.

Davis, J. A., Caiozzo, V. J., Lamarra, N., Ellis, J. F., Vandagriff, R., Prietto, C. A. et al. (1983). Does the gas exchange anaerobic threshold occur at a fixed blood lactate concentration of 2 or 4 mM? International Journal of Sports Medicine, 4, 89-93.

el-Sayed, M. S., George, K. P., & Dyson, K. (1993). The influence of blood sampling site on lactate concentration during submaximal exercise at 4 mmol.l-1 lactate level. European Journal of Applied Physiology and Occupational Physiology, 67, 518-522.

Feliu, J., Ventura, J. L., Segura, R., Rodas, G., Riera, J., Estruch, A. et al. (1999). Differences between lactate concentration of samples from ear lobe and the finger tip. Journal of Physiological Biochemistry, 55, 333-339.

Foxdal, P., Sjödin, A., & Sjödin, B. (1996). Comparison of blood lactate concentrations obtained during incremental and constant intensity exercise. International Journal of Sports Medicine, 17, 360-365.

Foxdal, P., Sjödin, B., Rudstam, H., Ostman, C., Ostman, B., & Hedenstierna, G. C. (1990). Lactate concentration differences in plasma, whole blood, capillary finger blood and erythrocytes during submaximal graded exercise in humans. European Journal of Applied Physiology and Occupational Physiology, 61, 218-222.

Harnish, C. R., Swensen, T. C., & Pate, R. R. (2001). Methods for estimating the maximal lactate steady state in trained cyclists. Medicine and Science in Sports and Exercise, 33, 1052-1055.

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

Impellizzeri, F. M., & Marcora, S. M. (2007). The physiology of mountain biking. Sports Medicine, 37, 59-71.

Laursen, P. B., Rhodes, E. C., Langill, R. H., McKenzie, D. C., & Taunton, J. E. (2002). Relationship of exercise test variables to cycling performance in an Ironman triathlon. European Journal of Applied Physiology, 87, 433-440.

Lucia, A., Sanchez, O., Carvajal, A., & Chicharro, J. L. (1999). Analysis of the aerobic-anaerobic transition in elite cyclists during incremental exercise with the use of electromyography. British Journal of Sports Medicine, 33, 178-185.

Mc Naughton, L. R., Thompson, D., Philips, G., Backx, K., & Crickmore, L. (2002). A comparison of the lactate Pro, Accusport, Analox GM7 and Kodak Ektachem lactate analysers in normal, hot and humid conditions. International Journal of Sports Medicine, 23, 130-135.

McLean, S. R., Norris, S. R., & Smith, D. J. (2004). Comparison of the Lactate Pro and the YSI 1500 Sport Blood Lactate Analyzers. International Journal of Applied Sports Sciences, 16, 22-31.

Medbø, J. I., Mamen, A., Holt Olsen, O., & Evertsen, F. (2000). Examination of four different instruments for measuring blood lactate concentration. Scandinavian Journal of Clinical and Laboratory Investigation, 60, 367-380.

Medbø, J. I., & Toska, K. (2001). Lactate release, concentration in blood, and apparent distribution volume after intense bicycling. Japan Journal of Physiology, 51, 303-312.

Pyne, D. B., Boston, T., Martin, D. T., & Logan, A. (2000). Evaluation of the Lactate Pro blood lactate analyser. European Journal of Applied Physiology, 82, 112-116.

Sjödin, B., & Jacobs, I. (1981). Onset of blood lactate accumulation and marathon running performance. International Journal of Sports Medicine, 2, 23-26.

Stegmann, H., & Kindermann, W. (1982). Comparison of prolonged exercise tests at the individual anaerobic threshold and the fixed anaerobic threshold of 4 mmol.l(-1) lactate. International Journal of Sports Medicine, 3, 105-110.

Tokmakidis, S. P., Leger, L. A., & Pilianidis, T. C. (1998). Failure to obtain a unique threshold on the blood lactate concentration curve during exercise. European Journal of Applied Physiology and Occupational Physiology, 77, 333-342.

van Someren, K. A., Howatson, G., Nunan, D., Thatcher, R., & Shave, R. (2005). Comparison of the Lactate Pro and Analox GM7 blood lactate analysers. International Journal of Sports Medicine, 26, 657-661.

Van, S. R., Vanden, E. B., & Hespel, P. (2004). Correlations between lactate and ventilator thresholds and the maximal lactate steady state in elite cyclists. International Journal of Sports Medicine, 25, 403-408.

Welde, B., Evertsen, F., von Heimburg, E., & Medbø, J. I. (2003). Energy cost of free technique and classical cross-country skiing at racing speeds. Medicine and Science in Sports and Exercise, 35, 818-825.

Weltman, A. (1995). The blood lactate response to exercise. Champaign, Ill: Human Kinetics. Wilmore, J. H., & Costill, D. L. (2004). Physiology of sport and exercise. (3 ed.) Champaign, IL: Human Kinetics.


7 x 4 min step test protocol

Information for Athletes and NTC Scientists

(To be used from October 2006 - Updated 27/12/08)
Compiled by: Tony Rice
Dept of Physiology: Australian Institute of Sport and Australian Rowing
Office: (02) 6214-7891
Email: Tony.Rice@ausport.gov.au


The laboratory test protocol adopted by Rowing Australia aims to provide detailed physiological information of the rower’s submaximal capacity and efficiency and to measure maximal performance parameters in a time efficient manner. To do this a 7 x 4 min protocol has been implemented across the country since 2006. For the 2009-2012 Olympic cycle the standard protocol will change to reflect new information that has been presented by Ivan Hooper and others on the use of sliders and lower drag factors. The main aims of altering the protocol are to 1. more accurately reflect the stroke rate and drive:recovery ratio of on-water rowing and 2. minimise the risk of injury that may result from considerable time being spent on the ergometer with significant lower back load. Implementation of the new drag factor settings and sliders will be mandatory from Nov 2009.

For the 2008-2009 season each SIS/SAS will be able to asked to do complete the laboratory test in one of two ways:
1. If sliders are available then it is asked the laboratory completes the test protocol using sliders and the updated drag factor settings
2. If sliders are not available then it is asked the laboratory completes the test protocol using the old drag factor settings as per the previous Olympic cycle

The standard laboratory test will be completed at least two times within each season (all dates to be communicated early 2009 and confirmed at the beginning of each season). Rowing Australia and the
National Rowing Centre of Excellence (NRCE) require a standard summary of data on all SIS/SAS athletes who are aspiring for National Selection in the current season to be returned to the Sports
Science Coordinator shortly after the completion of each testing period.

NB Testing on Concept IID ergometers - there are no significant differences in the physiological responses to either the IIC or IID ergo and thus tests can be completed on either one with the proviso that all tests for an individual athlete are carried out on the same ergometer throughout the rowing season and every is made to continue using the same ergometer for all subsequent seasons.
The following information is designed as a detailed guide to the testing methods.

7 x 4 min Step Test Protocol Laboratory Environment and Subject Preparation


The athlete must not train at all in the 12 hours preceding the test. On the day before the test, the afternoon training session should consist of no more than 12 km on the water, and should be of low intensity (T3-T2 range). There should be no heavy weight training, or exercise to which the athlete is not accustomed. It is suggested that the athlete replicate as closely as possible similar training loads in the 24 hours leading into each testing block.


A normal meal (incorporating a high carbohydrate component) should be eaten on the evening preceding the test and, if scheduling allows, also on the day of the test. No alcohol should be consumed in the 24 hours preceding the test. The athlete should give special attention to ensuring good hydration in the lead-up to the test.

Test Preparation

Each laboratory may have information and consent forms that may need to be provided prior to the

Equipment Checklist
• Concept IIC or IID rowing ergometer
• Heart rate monitoring system
• Expired gas analysis system (as per general recommendations)
• Stopwatch
• Lactate analyser (Lactate Pro recommended) and blood gas analyser if possible.
Blood sampling equipment:
• Finalgon ointment or cream
• Autolet, lancets and platforms
• Sterile alcohol swab
• Tissues
• Heparinised capillary tubes
• Pipette (if required)
• Disposable rubber gloves
• Sharps container
• Biohazard bag.

Ergometer Settings
Table 1: Old Ergometer Drag Factor Settings
Category Drag Factor
Junior Female 110
Lightweight Female 110
Heavyweight Female 120
Junior Male 120
Lightweight Male 120
Heavyweight Male 130

New Ergometer Drag Factor Settings
(only to be used if tested on sliders – see Introduction)
Category Drag Factor
Junior Female 95
Lightweight Female 95
Heavyweight Female 105
Junior Male 105
Lightweight Male 105
Heavyweight Male 115

Step - Test Administration:

Athletes will start the 7 step test protocol with a work load based and increment based purely on their previous year’s best 2000m time. The range of times for the 2000m ergometer tests have been
divided on 10 sec increments with the fastest 2000m times having the highest starting work load and increment (see Table 2).


See resource list below for a print copy of table 2.  

In order to ensure that individual athlete’s complete the identical amount of work prior to beginning the 7th step (4 min at maximal pace) every 7 x 4 min step test undertaken by the athlete for that seasonal year must use the same starting work load and increment. In other words there will be no increment of starting work load during the season as has been in previous years. The only way an athlete will be able to change their starting work load or increment will be to perform a 2000m test that has a time that places them into a different time bracket.

The workloads and increments have been designed such that the 6th step (i.e. the step immediately preceding the maximal step) produces a blood lactate value in the range of 5-8 mmol/L. Obviously this will change depending on the time of year and the athlete’s current training status but importantly in a single season the athlete will complete an identical amount of work leading into the maximal performance component of the test.

The scientist in charge is given flexibility in choosing if moving to the next 10 sec increment is valuable or not. An example would be that when an athlete has 4 years of data all starting at the same work load and increment but finally betters their 2000m time from 5:50.6 to 5:49.8. Taking the protocol to its exact description would mean the athlete would change their starting work load and increment. However the gain in doing this is far outweighed by the inability to compare the athlete across the 4 years of data using the previous starting work load and increment. In cases such as this it is left to the discretion of the coach and scientist to decide what should be the appropriate starting work load and increment foir that athlete. Scientists can contact either the Head Coaches, HPD or Tony Rice for additional consultation if required.

The 2000m category times used in Table 2 are based on the athlete’s best 2000m time from the previous year and not their all-time personal best 2000m test. If the athlete had a medical exemption for the previous year or Thus it is possible that work loads would change slightly from year to year for a given individual. If this is the case, comparison between years for the same individual, or in the same year between individuals, must only be done using variables such as heart rate, blood lactate, VO2 and perceived exertion at LT1, LT2 and the maximal step.

Distance covered in the final maximal step can only be used as a comparison between athletes when those athletes have completed the same amount of work prior to the maximal step i.e. identical starting work load and increment.

7 x 4 min step test procedure:

1. Complete and give scientist your Informed Consent forms.
2. The scientist will measure the following physical parameters: height; weight; sitting height; arm span; sum of 7 skinfolds (only sum of 7 skinfolds is a requirement of RA but it is good practice if labs have the available resources to complete the additional measurements)
3. Attach a heart rate monitor and ensure it is working correctly.
4. The scientist will place a small quantity of Finalgon on one of your earlobes, or on some fingertips to ensure that the capillary blood is arterialised prior to sampling.
5. The scientist will adjust the ergometer drag factor to that appropriate to your competition category (see Table 1) and provide you with the work loads for your 6 increments (see Table 2).
6. The scientist will position the gas collection apparatus (respiratory valve etc) and ensure that the athlete is as comfortable as possible. Take several light strokes and the scientist will make any necessary adjustment to respiratory hoses or other apparatus to ensure that the hose is not pulling on the breathing apparatus at any stage during the stroke.
7. The scientist will collect a pre-exercise blood sample from the earlobe or fingertip using a Lactate Pro analyser.
8. The scientist set the ergometer output display to show Watts for each stroke as well as set the work load time and rest interval.
9. The scientist will attach the nose clip and prepare to start the test if the test requires gas analysis.
10. Start rowing when instructed.
11. Blood is collected from your earlobe or fingertip during each rest period and analysed. During the 1 min rest period, you are permitted to remove the gas collection apparatus to have a drink. However, it is important to ensure that the breathing apparatus is back in position well before the start of the next work bout (approximately 10–15 s).
12. You are required to complete 6 submaximal work loads prior to beginning the maximal step.
Your starting work load is based on your best 2000m time from the previous year and the work loads for all rowing categories are contained in Table 2.
13. There is only the standard 1 min break between the end of the final submaximal step (6th step) and beginning the 4 min maximal step.
14. Begin the test when instructed. Remember the aim of the test is to cover as many meters as possible in the 4 minutes. You should be exhausted at the completion of the 4 minutes of rowing. If possible try to even split the 4 minutes rather than starting conservatively and then coming home strong. You should be able to hold or better your average 2000m split for the entire 4 minutes. Coincident with the maximal performance assessment will be the attainment of maximal heart rate, blood lactate concentration and oxygen consumption.
15. At the end of the test, the scientist will help you remove the breathing apparatus as rapidly as possible.
16. An earlobe or fingertip blood sample should be collected and analysed at the completion and 4min post completion of the final maximal step. The highest value of these two readings should be used as the peak blood lactate.

Analysis of Test Results: Blood Lactate Profiles

Submaximal oxygen uptakes are calculated by averaging the readings recorded during the final 2 min of each submaximal workload. The maximum oxygen uptake is recorded as the highest value actually attained over a period of a full minute. Thus, if the gas analysis system is based on 30 s sampling periods, the maximum oxygen uptake is the sum, or if all results are expressed in L.min-1, the average of the highest two consecutive readings. If 15 s sampling periods are used, the maximum oxygen uptake is the highest value obtained on the basis of any four consecutive readings.

Submaximal heart rates are the values for the final 30 s of each submaximal workload. The maximum heart rate is the highest value recorded over a 5 s sampling period during the entire test. Computerised analysis allows for quite simple determination of the various blood lactate transition thresholds and associated measures. The ADAPT* (Automatic Data Analysis for Progressive Tests) software package is to be used to calculate these thresholds from the test data.

*(ADAPT software is available to the Australian sport science community from Sport Sciences, Australian Institute of Sport, PO Box 176, Belconnen ACT 2616.)
Data from the final maximal step are included as the final work load values (including peak lactate) and used in combination with values from the submaximal workloads for calculation of the blood lactate thresholds and related measures in ADAPT.


Gore CJ. (Editor) (2000) Physiological Tests For Elite Athletes / Australian Sports Commission. Human Kinetics Champaign IL. Chapters 2, 3, 8 and 22.
Medbø JI, Mohn A-C, Tabata I et al. (1988) Anaerobic capacity determined by maximal accumulated O2 deficit. Journal of Applied Physiology 64: 50–60.


Resources to conduct test:

For a print copy, see 7x4min step test protocol.

Table 2: Determination of Test Protocol.


Training Methods and Intensity Distribution of Young World Class Rowers 

By Arne Guellich (1), Stephen Seiler (2), and Eike Emrich (3)
1 Department of Sports Sciences, University of Kaiserslautern, GERMANY
2 Faculty of Health and Sport, University of Agder, Kristiansand, NORWAY
3 Institute of Sports Sciences, University of the Saarland, Saarbruecken, GERMANY

From: International Journal of Sports Physiology and Performance, 2009, 4, 448-460


Purpose: To describe the distribution of exercise types and rowing intensity in successful junior rowers and its relation to later senior success. Methods: 36 young German male rowers (31 international, 5 national junior finalists, 19.2 ± 1.4 yr, 10.9 ± 1.6 training sessions.wk-1) reported the volumes of defined exercise and intensity categories in a diary over 37 weeks. Training categories were analysed as aggregates over the whole season and also broken down to defined training periods. Training organisation was compared between juniors who attained national and international senior success three years later. Results: Total training time consisted of 52% rowing, 23% resistance exercise, 17% alternative training, and 8% warm-up programs. Based on heart rate control, 95% of total rowing was performed at intensities corresponding to <2 mmol.L-1, 2% at 2-4 mmol.L-1, and 3% at >4 mmol.L-1 blood lactate. Low-intensity work remained widely unchanged at ~95% throughout the season. In the competition period the athletes exhibited a shift within <2mmol-exercise towards lower intensity and within the remaining ~5% of total rowing towards more training near maximal oxygen consumption (VO2max) intensity. Retrospectively, among subjects going on to international success three years later had their training differed significantly from their peers only in slightly higher volumes at both margins of the intensity scope. Conclusion: The young world-class rowers monitored here exhibit a constant emphasis on low intensity steady-state rowing exercise, and a progressive polarization in the competition period. Possible mechanisms underlying a potential association between intensity polarization and later success require further investigation. 

Keywords: high performance, training analysis, intensity distribution, endurance, rowing


Elite endurance athletes subject themselves to very high training loads in pursuit of maximal performance. For example, world class senior rowers compete over a 2000m distance requiring ~6-7 minutes; yet invest a training volume in a season equivalent to many hours for each minute of an international competition. A key question that occupies the minds of athletes and coaches is how best to utilize this training investment. For numerous reasons, systematic intervention for research purposes is constrained in elite sport and experimental studies are lacking to identify any “optimal” training organization for maximizing both physiological and technical adaptations. We contend that the international competition environment is a quite effective experimental arena, in a Darwinistic sense. Extreme performance standards stimulate the emergence of self-organizing processes where athletes, national teams, and governing bodies pursue the training structure which gives the most consistent success. This process together with inter- and intra-individual method iteration, evaluation, and adjustments presumably drives changes in training over time that correspond with continued performance improvements. For example, one of the major changes in the training evolution of international medal winning Norwegian rowers over three decades was an increase in total training volume associated with a substantial shift in intensity distribution from higher to lower intensities.2 Accurate descriptions of the training characteristics of highly successful athlete groups have value in furthering our knowledge about performance improvement in endurance sport and catalyzing experimental studies.

Also, within an elite athlete generation fairly small improvements in performance may be critical to success. Top performers approximate the margin of individual load tolerance in training, and minor variation in the balance of beneficial adaptation and maladaptive load-related stress reaction may account for critical differences in performance development. The day-to-day and seasonal distribution of training intensity appears to be a crucial variable in training organization for endurance athletes. Following the 3-intensity zone structure representing exercise intensity below the first ventilatory threshold (VT1; where the ventilatory equivalent for O2 breaks from linearity, without an increase in the ventilatory equivalent for CO2; typically < 2 mmol.L-1 blood lactate), from VT1 to VT2 (where the ventilatory equivalent for CO2 also begins to increase; ~2-4 mmol.L-1), and above VT2 (> 4 mmol.L-1),3-5 it has previously been proposed that two basic intensity distribution patterns emerge from the research literature.5,6 The “threshold training model” emerges from some short-termed studies demonstrating that training at the lactate threshold intensity evokes significant physiological improvements among untrained subjects.7-10 A contrasting “polarized training model” has been proposed based on observations from a number of studies describing the distribution of work intensity among elite athletes in marathon running, rowing, track cycling, and cross-country skiing.2, 5,11-16 One consistent observation from these studies is that successful endurance athletes perform 75% or more of their training (sessions, distance, time) at intensities below VT1. In addition, about 10 to 20% of training volume is reported to be clearly above VT2, (i.e. 6-10 mmol.L-1 blood lactate).2,5,11,15,17 Consequently, remarkably little training is executed at the traditional lactate threshold. Thus, the training is apparently “polarized away” from the work range characterised by moderately hard intensity. If training intensity distribution is critical for optimal performance, we might expect to see quantifiable differences in organization between highly successful and less successful athletes with similar performance potential.

Recent longitudinal observations in quasi-experimental post-hoc and experimental designs support the value of low intensity training in achieving desired physiological adaptations and performance enhancement.3,4,18,19

In the present study we extend these findings by reporting 1) a detailed description of the distribution of the exercise types and of the intensity distribution within the specific rowing workout, 2) their alteration over a complete training season from autumn until summer in a large group of internationally successful junior age rowers, and 3) a comparison of the training characteristics of the junior rowers who reached international senior finals three years later to those who did not attain this success level.


Study Design

The present study builds on the complete reported daily training data provided by 36 athletes from the men’s German junior national rowing squad. This study was approved by the German Federal Institute of Sports Science including the subjects’ informed, written consent for their training data to be used for research purposes.

All national squad members were requested to document their executed daily individual training in a standardized digital training diary and submit it to the national coach. Individual heart rate (HR) ranges for defined intensity categories in training were determined during a centralized rowing ergometer ramp protocol (Concept CIIC; 3 min stages, 20 W steps) in the first week of each training year. In addition, rowing power (watts) at 4 mmol.L-1 venous blood lactate (PLa4) was calculated from the blood lactate/rowing ergometer power relationship. The national rowing governing body did not perform standardized, centralized VO2 max testing on junior rowers. Therefore information regarding the maximal oxygen consumption of these athletes is not available. Individual heart-rate ranges were prescribed for each of the rowing intensity categories based on the stable blood lactate-heart rate relationship determined during ramp protocol rowing ergometry performed at the beginning of the training season.23 Heart rate was controlled during all rowing sessions via online HR-monitoring (Polar, Kempele, Finland).

Training monitoring

Prior to the training, athletes were briefed as to the desired training composition by the coach and provided a reporting scheme. Training was categorized as defined by the national federation (see Table 1). The training data reported here represents the executed, not the planned training for a complete training season (37 weeks, t1). In addition competitive senior success for the entire sample was followed up 3 years later (t2).

The intensity definitions used for training documentation closely correspond to the physiological 3-zone model used in a previous study of rowing intensity distribution.2 The categories Compensation and Extensive Endurance (<80% race pace; HR <160 b.min-1; [La-] <2 mmol.L-1; Table 1) correspond with work below VT1 (“zone 1”). Intensive Endurance (75-85% race pace; HR 156-168 b.min-1; [La-] 2-4 mmol.L-1) corresponds with work intensity between VT1 and VT2 (“zone 2”). The categories Highly Intensive Endurance, Race-Specific Velocity-Endurance, and Velocity Training (85-112% race pace; HR >180 b.min-1; [La-] >4 mmol.L-1) correspond with work intensity above VT2 (“zone 3”). This three intensity zone scheme has been previously described and used in both experimental and descriptive studies of endurance exercise intensity distribution.5,20-22

Training was recorded from the beginning of the training season (15th October) until the national trials for the junior world championships (30th June; 37 weeks in total). The 37 weeks were divided into 3 training periods: basic preparation period (BPP) 1st to 15th training week, specific preparation period (SPP) 16th to 25th week, and (early) competition period (CP) 26th to 37th training week. SPP culminated in the national small-boat championship regatta which is obligatory for all squad members. The CP recording concluded with the national trials.

To evaluate the reliability of athletes’ training documentation, diary figures reported to the national coach were compared to data reported directly to our research group (as “neutral” addressees) by 29 athletes participating in an anonymous postal survey after the referred season. Subjects were re-identified for this analysis based on birth date and success. The data from the training diary and from the postal survey correlated with r=0.88 (training frequency; p<0.01) and r=0.84 (training time; p<0.01). The diary figures deviated from those in the postal survey by 4.0 ± 8.5% in training frequency and –10.4 ± 12.3% in expended training time. No systematic relation between this deviation and performance achievement was observed (p>0.05, in each case).

Senior Success

The 36 athletes in this study continued rowing at a high level. We therefore retrospectively compared the junior training characteristics of 14 athletes who reached senior world championships and/or Olympic finals three years after the junior training registration period with the 22 who attained success ‘only’ at national level within the same period.

Statistical Analysis

All statistical analyses were performed using SPSS version 14.0. Physical, physiological, and training characteristics are presented as means and standard deviations. Training intensity distribution and other training characteristics were compared across the three defined training periods using Repeated Measures ANOVA. Comparison of the junior age training characteristics of internationally successful and less successful senior rowers was performed using Independent Samples T-tests. A p value of < 0.05 was considered statistically significant.


All 36 athletes remained members of the national junior squad throughout the observed training period (2001; t1) and became finalists at the national junior championships. Of these, 31 also reached the finals at the junior world championships, 27 attained a medal, and 15 became junior world champion. The athletes were 19.2 ± 1.4 years, 91.0 ± 6.0 kg and 193.3 ± 5.3 cm at t1. They performed a mean of 10.9 ± 1.6 training sessions and 12.8 ± 2.5 hours of (net) training time per week. Their PLa4 value during rowing ergometry ramp protocol at the beginning of the training season was 373 ± 29 W. Three years after the training registration season (t2), all 36 athletes were finalists at the national senior championships. Of these, 14 reached the finals at the Olympic Games (Athens 2004) and/or senior world championships, nine won medals.

Rowing specific activities made up 52% of total junior training time (Table 2). The remaining training time was devoted to resistance exercise (23%), general athletic training like jogging, strengthening gymnastics, and game play (17%), and warm-up programs (8%). Strength training was dominated by “power endurance” training with high repetitions performed with moderate loads (76%). The overall distribution of rowing specific training intensity is also provided in Table 2. Interestingly, ~95% of rowing specific training was performed at intensities corresponding to < 2 mmol.L-1 blood lactate (below VT1, zone 1; Compensation and Extensive Endurance range).

Weekly training frequency increased from 10.3 ± 2.5 in the Basic Preparation Period (BPP) to 11.3 ± 1.7 in the Specific Preparation Period (SPP; p<0.01) and was reduced again to 10.6 ± 1.8 sessions.wk-1 in the (early) Competition Period (CP; p<0.01). Figure 1 shows an approximate doubling in rowing specific training volume from BPP to CP. This doubling was achieved via both an increase in total training volume and a decrease in strength training and alternative training. The relative contribution of low intensity (zone 1) work to total rowing training remained almost constant throughout the entire season, but the higher intensity work became significantly more intense (Figure 2). Low intensity zone 1 endurance training made up 96% of all rowing volume during the BPP and only decreased to 94% in the CP. During the CP, there was also a small but significant shift within zone 1 rowing represented by a lowered volume of Extensive Endurance range (BPP 89%, SPP 88%; p>0.05; CP 84%; p<0.01) and an increase in the amount of very low intensity Compensation range rowing performed (BPP 7.1%, SPP 6.5%; p>0.05; CP 10.1%; p<0.01). The remaining 4-6% of rowing training distance shifted first towards a transiently higher volume of the Intensive Endurance range (“zone 2” lactate threshold intensity work) from BPP to SPP and then towards an enhancement of the highest intensity ranges (“zone 3”) of Race-Specific Velocity-Endurance and Velocity during CP (Figure 2). While high intensity training remained a small percentage of total training volume throughout the season, the 141% increase of the share of total rowing at race pace or higher intensity from BPP to CP (p<0.01) represents an enhancement of the absolute distance rowed at this intensity range by about 3-fold.



Among 14 athletes who reached the finals at the Olympics and/or senior world championships three years after the training registration period (t2), and 22 who did not, 12 (86%) and 19 (86%), respectively, had reached the finals at the junior world championships (t1), 10 (71%) and 17 (77%) had medalled, and 5 (36%) and 10 (45%) had been junior world champion. The respective groups did not differ systematically in age (19.0 ± 1.3 and 19.4 ± 1.4 years; M ± SD), weight (91 ± 6 and 91 ± 6 kg), height (193 ± 5 and 193 ± 6 cm), or PLa4 (368 ± 28 and 376 ± 30 W) at t1 (all p>0.05).

Table 3 compares the junior-age training characteristics between internationally and nationally successful athletes. The rowers reaching international success as seniors did not differ systematically (p>0.05) as juniors in total training frequency or volume, distribution of expended training time, or their time distribution among different training modes. However, small but statistically significant differences were exhibited within the intensity distribution of their specific rowing endurance training. The international finalists completed more distance in both the lowest intensity Compensation range and the highest intensity Race-Specific Velocity-Endurance range.


An underlying premise for describing the training organization of highly successful athletes is that they are successful, in part, because of how they train. In this context, we believe the most important finding in this study is that, based on time-in-zone heart rate monitoring, internationally successful junior rowers performed 95% of all specific rowing training over a 37 week training period in “zone 1,” at a heart rate corresponding to blood lactate concentration under 2 mmol.L-1. In comparison, the same time-in-zone, 3-intensity zone method applied to a group of well-trained, non-elite distance runners showed that 71% of their training was in zone 1, 21% in zone 2 and 8% in zone 3.3 Accepting that this well established intensity quantification method tends to overestimate time spent at low training intensity,5 these findings still demonstrate a marked emphasis on basic endurance training throughout the training season. In the present group of rowers, the remaining 4-6% of rowing specific training volume quantified at higher intensities, progression from the basic preparation period to the competition period was associated with a shift from emphasis on moderately high intensity “lactate threshold” training, towards more race pace intensity training at near maximal oxygen consumption (VO2max) intensity. That is, the intensity distribution became more polarized in the competition period.

This was a quite homogenous group of talented rowers who had reached national elite level in a very strong rowing nation. Their physique (mean 91 kg, 193 cm) clearly exceeded previously published descriptions of rowing finalists at junior world championships,24 consistent with the fact that 15 of the subjects in the sample became junior world champions during the season training data was collected. The range in performance between “more successful” and “less successful” athletes as defined here is very small. Also, this study did not compare effects of different training programs on performance and thus does not establish a causal relationship between intensity distribution and performance. It was an ex post facto analysis of common variation between characteristics of junior training and later senior success. Based on a 3-year follow-up analysis, the only significant difference in training volume or organization observed between the most successful and less successful rowers in this sample was a modest but significant increase in the degree of intensity polarization observed in the most successful athletes. Athletes who went on to senior international success had, as juniors, tended to perform slightly more of their total rowing endurance exercise at very low intensity and at very high intensity compared to their peers. We can only speculate what advantages this increase in intensity polarization might have provided. It might be that the increased polarization observed merely demonstrates a form of intensity management discipline (keeping hard training hard and easy training easy) among the most successful athletes that could prove protective against overstress.

The present findings are consistent with previous studies demonstrating that low intensity (below LT1 or VT1) training dominates the total training volume of successful endurance athletes in a variety of sports.2,11-16 However, the extreme emphasis on low intensity, steady state training seen in these elite junior rowers has not been reported in the research literature previously. The 2000 meter rowing distance requires ~6 minutes to complete in a large team boat and is performed at 100-110% of VO2 max intensity.25 Clearly this distribution violates conventional wisdom regarding training intensity specificity; these athletes train relatively little at competition intensity. Recently, Ingham and colleagues compared 12 weeks of training at low intensity only (98% of total training performed at <75% of VO2 peak) with a regimen of 70% low intensity and 30% high intensity training (>84% VO2 peak). They found that in the British national standard rowers involved, both training regimes gave similar improvements in VO2 max and rowing test performance, but that low intensity only training actually improved blood lactate responses at sub-maximal intensity to a greater extent.19 No indicators of overreaching were present in the mixed intensity training group, making it unclear why the mixed training model failed to induce a greater performance improvement.

One of us has previously concluded that elite endurance athletes in running, cycling, cross-country skiing, and rowing often perform surprisingly little of their total training at intensities typically described as lactate threshold training, but instead tend to polarize their training away from this moderate intensity, training both a great deal at below VT1 and a significant amount above VT2 intensity.22 The present descriptive study of internationally successful rowers, and the recent experimental study on rowers by Ingham and colleagues,19 both suggest that marked performance adaptations and very high level performances in rowing can be elicited with a regiment of training that is dominated by low intensity, high volume work, with relatively little race pace, high intensity training. These findings run contrary to accepted theories that substantial high intensity training is critical for optimizing centrally limited oxygen delivery capacity in endurance athletes.26,27

Anecdotally, the training intensity distribution reported here is not unique to German rowers, but is also observed in other highly successful international programs. We propose that there are several unique characteristics of rowing specifically, as well as the elite training process in general that may explain the training distribution employed.

Expansion of total training volume generally is achieved at the expense of the high intensity work proportion. High performance athletes expose themselves to voluminous training load, mostly involving multiple daily sessions, and they approximate (at least temporarily) the margin of what is tolerable. Athletes attempt to balance loads evoking maximal positive adaptations (gene expression, synthesis of mitochondrial and other relevant proteins, cardiovascular performance, buffering capacity, technical efficiency at near race velocity) while avoiding excessive sympathetic stress leading to overtraining. Consistent with the evolution of training organisation among international rowing medallists over recent decades and with reports from elite athletes in other endurance sports,2,11-16 achieving this balance apparently favors the selection of a training intensity distribution characterised by voluminous low-intensity rowing below the lactate threshold with only intermittent highly intensive bouts.

Rowing power is a function of mean stroke force and stroke frequency. In trained rowers, peak forces during a rowing stroke remain quite consistent across rowing frequency,28 with rowing stroke rate (i.e. duty cycle) being the primary intensity control variable. We might speculate from this that extensive training at low intensities remains effective in recruiting a large proportion of available motor units, and achieves the specific muscular adaptations necessary to row at high power outputs as well. Elite rowers may therefore gravitate towards training below the first ventilatory threshold in order to stabilize technical aspects of the rowing stroke while still achieving desired physiological adaptations and perhaps avoiding excessive stress reactions.22 It is also important to point out that although the relative percentage of high intensity rowing was low, during the competition period, these athletes were still performing ~20 minutes of rowing weekly at heart rates corresponding to high intensity. Because a heart rate based “time-in-zone” approach to intensity classification will tend to underestimate the actual time (and physiological stress) of rowing at high work intensity due to delays in heart rate responses, the actual high intensity work duration each week is perhaps 30 minutes or more.

Rowing also differs from endurance sports like running and cycling in that substantial volumes of non-rowing training are performed. While almost all training time may be movement specific among road cyclists and distance runners, little more than 50% of the total training time of these rowers was rowing. Traditionally, rowing training often incorporates a significant strength training component, which reduced the relative time spent on rowing specific training. Rowing ranged from 40% during the basic preparation phase to 65% of total training time during the early competition phase. Previously, it has been suggested that 70% of total training volume of rowers should be specific.29 However, the lower value reported here is consistent with the fact that junior rowers often do not have the same access to good on-water conditions during the late fall and winter. In contrast, elite senior rowers are more likely to live in or travel to locations affording year-around access to on-water training.

Because a large proportion of total training volume in these elite junior rowers was not rowing, it is worthwhile to consider the impact of this training on the overall training intensity distribution of the rowers. Heart rate was not monitored during non-rowing activities such as stretching, game play, and jogging, but they were always performed at low intensity according to coaches and therefore would likely contributed almost exclusively to the low intensity volume. Strength training made up 23% of total training time. It is likely that some strength training sessions induced transient periods where local muscle metabolic rates and muscle and blood lactate values were consistent with training in zone 2 or zone 3. The potential contribution of this training to power output and fatigue resistance at race pace in rowing is unclear.

We contend that the reported observations prompt further research on training intensity distribution with particular attention to highly selective samples of extraordinary performers. Future goals are 1) to describe training load in more detail, including physiological responses and 2) to examine the effects of varying intensity distributions on physiological capabilities and performance.


The authors wish to express their sincere thanks to Michael Mueller, performance director of the German Rowing Federation, and Dieter Altenburg, national junior coach, for fruitful cooperation and helpful suggestions in this project.


1. Hopkins WG, Hawley JA, Burke LM. Design and analysis of research on sport performance enhancement. Med. Sci. Sports Exerc. 1999: 31: 472–485.

2. Fiskerstrand A and Seiler KS. Training and performance characteristics among Norwegian International Rowers 1970-2001. Scand. J. Med. Sci. Sports 2004: 14: 303-310.

3. Esteve-Lanao J, San Juan AF, Earnest CP, Foster C, Lucía A. How do endurance runners actually train? Relationship with competition performance. Med. Sci. Sports Exerc. 2005 (37): 3: 496-504.

4. Esteve-Lanao J, Foster C, Seiler S, Lucía A. Impact of training intensity distribution on performance in endurance athletes. Journal of Strength and Conditioning Research 2007, 21(3), 943-949.

5. Seiler KS, Kjerland GØ. Quantifying training intensity distribution in elite endurance athletes: is there evidence for an ‘‘optimal’’ distribution? Scand. J. Med. Sci. Sports 2006: 16: 49-56.

6. Seiler S, Hetlelid K. The impact of rest duration on work intensity and RPE during interval training. Med. Sci. Sports Exerc. 2005: 37: 1601-1607.

7. Denis C, Dormois D, Lacour JR. Endurance training, VO2 max, and OBLA: a longitudinal study of two different age groups. Int. J. Sports Med. 1984: 5: 167–173.

8. Gaskill SE, Walker AJ, Serfass RA, Bouchard C, Gagnon J, Rao DC, Skinner JS, Wilmore JH, Leon AS. Changes in ventilatory threshold with exercise training in a sedentary population: the HERITAGE Family Study. Int. J. Sports Med. 2001: 22 (8): 586-92.

9. Kindermann W, Simon G, Keul J. The significance of the aerobic-anaerobic determination of work load intensities during endurance training. Eur. J. Appl. Physiol. 1979: 42: 25-34.

10. Londeree BR. Effect of training on lactate/ventilatory thresholds: a meta analysis. Med. Sci. Sports Exerc. 1997: 29: 837-843.

11. Billat VL, Demarle A, Slawinski J, Paiva M, Koralsztein JP. Physical and training characteristics of top-class marathon runners. Med. Sci. Sports Exerc. 2001: 33: 2089-2097.

12. Kellmann M, Altenburg D, Lormes W, Steinacker JM. Assessing stress and recovery during preparation for the World Championships in rowing. The Sports Psychologist 2001 (15): 151-167.

13. Schumacher YO, Müller P. The 4000-m team pursuit cycling world record: theoretical and practical aspects. Med. Sci. Sports Exerc. 2002: 34: 1029-1036.

14. Schumacher YO, Müller P, Keul J. Development of peak performance in track cycling. J. Sports Med. Phys. Fitness 2001 (41): 2: 139-146.

15. Steinacker JM, Lormes W, Lehmann M, Altenburg D. Training of rowers before world championships. Med. Sci. Sports Exerc. 1998: 30: 1158-1163.

16. Steinacker JM, Lormes W, Kellmann M, Liu Y, Reißnecker S, Opitz-Gress A, Baller B, Günther K, Petersen KG, Kallus KW, Lehmann M, Altenburg D. Training of junior rowers before world championships. Effects on performance, mood state, and selected hormonal and metabolic responses. J. Sports Med. Phys. Fitness 2000: 40: 327-335.

17. Billat V, Lepretre PM, Heugas AM, Laurence MH, Salim D, Koralsztein JP. Training and bioenergetic characteristics in elite male and female Kenyan runners. Med. Sci. Sports Exerc. 2003; 35(2):297-304.

18. Zapico AG, Calderón FJ, Benito PJ, Gonzáles CB, Parisi A, Pigozzi F, Di Salvo V. Evolution of physiological and haematological parameters with training load in elite male road cyclists: a longitudinal study. J. Sports Med. Phys. Fitness 2007 (47): 2: 191-196.

19. Ingham SA, Carter H, Whyte GP, Doust JH. Physiological and performance effects of low- versus mixed-intensity rowing training. Med. Sci. Sports Exerc. 2008; 40(3): 579-584.

20. Lucía A, Pardo J, Durantez A, Hoyos J, Chicharro JL. Physiological differences between professional and elite road cyclists. Int. J. Sports Med. 1998: 19: 342–348.

21. Lucía A, Hoyos J, Carvaljal A, Chicharro JL. Heart rate response to professional road racing: the Tour de France. Int. J. Sports Med. 1999: 20: 167–172.

22. Seiler S, Haugen O, Kuffel E. Autonomic recovery after exercise in trained athletes: Intensity and duration effects. Med. Sci. Sports Exerc. 2007 (39): 8: 1366-1373.

23. Foster C, Fitzgerald DJ, Spatz P. Stability of the blood lactate-heart rate relationship in competitive athletes. Med. Sci. Sports Exerc. 1999 (31): 4: 578-582.

24. Bourgois J, Claessens AL, Vrijens J, Philippaerts R, Renterghen BV, Thomis M, Janssens M, Loos R, Lefevre J. Anthropometric characteristics of elite male junior rowers. Br. J. Sports Med. 2000: 34: 213-216.

25. Hagerman FC. Applied physiology of rowing. Sports Med. 1984, 1 (4): 303-326.

26. Billat VL, Flechet B, Petit B, Muriaux G, Koralsztein JP. Interval training at VO2max: effects on aerobic performance and overtraining markers. Med. Sci. Sports Exerc. 1999: 31:156-163.

27. Laursen PB, Jenkins DG. The scientific basis for high-intensity interval training: optimising training programmes and maximising performance in highly trained endurance athletes. Sports Med. 2002; 32(1): 53-73.

28. McGregor AH, Bull AMJ, Byng-Maddick R. A comparison of rowing technique at different stroke rates: a description of sequencing, force production and kinematics. Int. J. Sports Med. 2004; 25: 465-470.

29. Mäestu J, Jürimäe J, Jürimäe T. Monitoring of performance and training in rowing. Sports Med. 2005; 35(7): 597-617.


VO2 Max, Aerobic Power & Maximal Oxygen Uptake

By: Sports Advisor.
Site link: Sports Advisor.

VO2 max has been defined as:

"the highest rate of oxygen consumption attainable during maximal or exhaustive exercise" (3).

As exercise intensity increases so does oxygen consumption. However, a point is reached where exercise intensity can continue to increase without the associated rise in oxygen consumption. To understand this in more practical terms, take a look at the diagram below:

The point at which oxygen consumption plateaus defines the VO2 max or an individual's maximal aerobic capacity. It is generally considered the best indicator of cardiorespiratory endurance and aerobic fitness. However, as well discuss in a moment, it is more useful as an indicator of a person's aerobic potential or upper limit than as a predictor of success in endurance events.

Aerobic power, aerobic capacity and maximal oxygen uptake are all terms used interchangeably with VO2 max.

VO2 max is usually expressed relative to bodyweight because oxygen and energy needs differ relative to size. It can also be expressed relative to body surface area and this may be a more accurate when comparing children and oxygen uptake between sexes.

One study followed a group of 12-year-old boys through to the age of 20 - half of which were trained, the other half untrained but active. Relative to bodyweight no differences in VO2 max were found between the groups suggesting that training had no influence on maximal oxygen uptake. However, when VO2 max was expressed relative to body surface area, there was a significant difference between groups and maximal oxygen uptake did indeed increase in proportion to training (4).

VO2 Max In Athletes and Non Athletes

VO2 max varies greatly between individuals and even between elite athletes that compete in the same sport.

The table below lists normative data for VO2 max in various population groups:

Genetics plays a major role in a persons VO2 max (11) and heredity can account for up to 25-50% of the variance seen between individuals. The highest ever recorded VO2 max is 94 ml/kg/min in men and 77 ml/kg/min in women. Both were cross-country skiers (16).

Untrained girls and women typically have a maximal oxygen uptake 20-25% lower than untrained men. However, when comparing elite athletes, the gap tends to close to about 10% (3). Taking it step further, if VO2 max is adjusted to account for fat free mass in elite male and female athletes, the differences disappear in some studies. Cureton and Collins (29) suggest that sex-specific essential fat stores account for the majority of metabolic differences in running between men and women.

Training & VO2 Max

In previously sedentary people, training at 75% of aerobic power, for 30 minutes, 3 times a week over 6 months increases VO2 max an average of 15-20% (6). However, this is an average and there are large individual variations with increases as wide ranging as 4% to 93% reported (6).

Amongst groups of people following the same training protocol there will be responders - those who make large gains, and non-responders - those who make little or no gains (14,9). This was originally put down to a simple issue of compliance but more recent research suggests that genetics plays a role in how well any one individual responds to an endurance training program (13).

The extent by which VO2 max can change with training also depends on the starting point. The fitter an individual is to begin with, the less potential there is for an increase and most elite athletes hit this peak early in their career. There also seems to be a genetic upper limit beyond which, further increases in either intensity or volume have no effect on aerobic power (5). This upper limit is thought to be reached within 8 to 18 months (3).

Crucially, once a plateau in VO2 max has been reached further improvements in performance are still seen with training. This is because the athlete is able to perform at a higher percentage of their VO2 max for prolonged periods (2). Two major reasons for this are improvements in anaerobic threshold and running economy.

Resistance training and intense 'burst-type' anaerobic training have little effect on VO2 max. Any improvements that do occur are usually small and in subjects who had a low level of fitness to begin with (17). Resistance training alone does not increase VO2 max (30,31,32) even when short rest intervals are used between sets and exercises (33).

Considerable training is required to reach the upper limit for VO2 max. However, much less is required to maintain it. In fact peak aerobic power can be maintained even when training is decreased by two thirds (18). Runners and swimmers have reduced training volume by 60% for a period of 15-21 days prior to competition (a technique known as tapering) with no loss in VO2 max (19,20,21).

VO2 Max as a Predictor of Performance

In elite athletes, VO2 max is not a good predictor of performance. The winner of a marathon race for example, cannot be predicted from maximal oxygen uptake (15).

Perhaps more significant than VO2 max is the speed at which an athlete can run, bike or swim at VO2 max. Two athletes may have the same level of aerobic power but one may reach their VO2 max at a running speed of 20 km/hr and the other at 22 km/hr.

While a high VO2 max may be a prerequisite for performance in endurance events at the highest level, other markers such as lactate threshold are more predictive of performance (3). Again, the speed at lactate threshold is more significant than the actual value itself.

Think of VO2 max as an athletes aerobic potential and the lactate threshold as the marker for how much of that potential they are tapping.

Factors Affecting VO2 Max

There are many physiological factors that combine to determine VO2 max but which of these are most important? Two theories have been proposed:

Utilization Theory

This theory maintains that aerobic capacity is limited by lack of sufficient oxidative enzymes within the cell's mitochondria (3). It is the body's ability to utilize the available oxygen that determines aerobic capacity. Proponents of this theory point to numerous studies that show oxidative enzymes and the number and size of mitochondria increase with training. This is coupled with increased differences between arterial and venous blood oxygen concentrations (a-vO2 difference) accounting for improved oxygen utilization and hence improved VO2max.

Presentation Theory

Presentation theory suggests that aerobic capacity is limited not predominantly by utilization, but by the ability of the cardiovascular system to deliver oxygen to active tissues. Proponents of this theory maintain that an increase in blood volume, maximal cardiac output (due to increased stroke volume) and better perfusion of blood into the muscles account for the changes in VO2max with training.

So what plays the greater role in determining an athlete's VO2 max - their body's ability to utilize oxygen or supply oxygen to the active tissues?

In a review of the literature, Saltin and Rowell (7) concluded that it is oxygen supply that is the major limiter to endurance performance. Studies have shown only a weak relationship between an increase in oxidative enzymes and an increase in VO2 max (8,9,10). One of these studies measured the effects of a 6-month swim training program on aerobic function. While oxidative enzymes continued to increase until the end, there was no change in VO2 max in the final 6 weeks of the program (10). 

Determining VO2 Max

VO2 max can be determined through a number of physical evaluations. These tests can be direct or indirect. Direct testing requires sophisticated equipment to measure the volume and gas concentrations of inspired and expired air. There are many protocols used on treadmills, cycle ergometers and other exercise equipment to measure VO2 max directly.

One of the most common is the Bruce protocol often used for testing VO2 max in athletes or for signs of coronary heart disease in high risk individuals.

Indirect testing is much more widely used by coaches as it requires little or no expensive equipment. There are many indirect tests used to estimate VO2 max. Some are more reliable and accurate than others but none are as accurate as direct testing. Examples include the multistage shuttle run (bleep test), 12 minute walk test and 1.5 mile run.

VO2 Max at Altitude

VO2 max decreases as altitude increases above 1600m (5249ft) or about the altitude of Denver, Colorado. For every 1000m (3281ft) above that, maximal oxygen uptake decreases further by approximately 8-11% (3). Anyone with a VO2 max lower than 50 ml/kg/min would struggle to survive at the summit of Everest without supplemental oxygen.

The decrease is mainly due to a decrease in maximal cardiac output. Recall that cardiac output is the product of heart rate and stroke volume. Stoke volume decreases due to the immediate decrease in blood plasma volume. Maximal heart rate may also decrease and the net effect is that less oxygen is "pushed" from the blood into the muscles (2).

Effects of Aging on VO2 Max

VO2 max decreases with age. The average rate of decline is generally accepted to be about 1% per year or 10% per decade after the age of 25. One large cross sectional study found the average decrease was 0.46 ml/kg/min per year in men (1.2%) and 0.54 ml/kg/min in women (1.7%) (22,23).

However, this deterioration is not necessarily due to the aging process. In some cases the decease may be purely a reflection of increased body weight with no change in absolute values for ventilation of oxygen. Recall, that VO2 max is usually expressed relative to body weight. If this increases, as tends to happen with age, and aerobic fitness stays the same then VO2 max measured in ml/kg/min will decrease.

Usually, the decline in age-related VO2 max can be accounted for by a reduction in maximum heart rate, maximal stoke volume and maximal a-vO2 difference i.e. the difference between oxygen concentration arterial blood and venus blood (2).

Can training have an affect on this age-related decline?

Vigorous training at a younger age does not seem to prevent the fall in VO2 max if training is ceased altogether. Elite athletes have been shown to decline by 43% from ages 23 to 50 (from 70 ml/kg/min to 40 ml/kg/min) when they stop training after their careers are over (24). In some cases, the relative decline is greater than for the average population - as much as 15% per decade or 1.5% per year (27,28).

However in comparison, master athletes who continue to keep fit only show a decrease of 5-6% per decade or 0.5-0.6% per year (25,26,27,28). When they maintain the same relative intensity of training, a decrease of only 3.6% over 25 years has been reported (28) and most of that was attributable to a small increase in bodyweight.

It seems that training can slow the rate of decline in VO2 max but becomes less effective after the age of about 50 (3).


1) Baechle TR and Earle RW. (2000) Essentials of Strength Training and Conditioning: 2nd Edition. Champaign, IL: Human Kinetics

2) McArdle WD, Katch FI and Katch VL. (2000) Essentials of Exercise Physiology: 2nd Edition Philadelphia, PA: Lippincott Williams & Wilkins

3) Wilmore JH and Costill DL. (2005) Physiology of Sport and Exercise: 3rd Edition. Champaign, IL: Human Kinetics

4) Sjodin B, Svedenhag J. Oxygen uptake during running as related to body mass in circumpubertal boys: a longitudinal study. Eur J Appl Physiol Occup Physiol. 1992;65(2):150-7

5) Costill DL. (1986) Inside Running: Basics of Sports Physiology. Indianapolis: Benchmark Press

6) Pollock ML. (1973). Quantification of endurance training programs. Exercise and Sport Sciences Reviews. 1,155-188

7) Saltin B, Rowell LB. Functional adaptations to physical activity and inactivity. Federation Proceeding. 1980 Apr;39(5):1506-13

8) Gollnick PD, Armstrong RB, Saubert CW 4th, Piehl K, Saltin B. Enzyme activity and fiber composition in skeletal muscle of untrained and trained men. J Appl Physiol. 1972 Sep;33(3):312-9

9) Saltin B, Nazar K, Costill DL, Stein E, Jansson E, Essen B, Gollnick D. The nature of the training response; peripheral and central adaptations of one-legged exercise. Acta Physiol Scand. 1976 Mar;96(3):289-305

10) Costill DL, Thomas R, Robergs RA, Pascoe D, Lambert C, Barr S, Fink WJ. Adaptations to swimming training: influence of training volume. Med Sci Sports Exerc. 1991 Mar;23(3):371-7

11) Bouchard C, Dionne FT, Simoneau JA, Boulay MR. Genetics of aerobic and anaerobic performances. Exerc Sport Sci Rev. 1992;20:27-58

12) Pollock ML, Foster C, Knapp D, Rod JL, Schmidt DH. Effect of age and training on aerobic capacity and body composition of master athletes. J Appl Physiol. 1987 Feb;62(2):725-31

13) Bouchard C, Shephard RJ, Stephens T, Sutton JR and McPherson BD (Eds.), Exercise Fitness and Health (pp. 147-153). Champaign, IL: Human Kinetics

14) Green HJ, Jones S, Ball-Burnett M, Farrance B, Ranney D. Adaptations in muscle metabolism to prolonged voluntary exercise and training. J Appl Physiol. 1995 Jan;78(1):138-45

15) Energetics in marathon running. Medicine and Science in Sports. 1969 1(2):81-86

16) Astrand P-O and Rodahl K. (1986) The Textbook of Work Physiology: Physiological Bases of Exercise (3rd ed.). New York: McGraw-Hill

17) Kraemer WJ, Deschenes MR, Fleck SJ. Physiological adaptations to resistance exercise. Implications for athletic conditioning. Sports Med. 1988 Oct;6(4):246-56

18) Hickson RC, Foster C, Pollock ML, Galassi TM, Rich S. Reduced training intensities and loss of aerobic power, endurance, and cardiac growth. J Appl Physiol. 1985 Feb;58(2):492-9

19) Costill DL, King DS, Thomas R and Hargreaves M. Effects of reduced training on muscular power in swimmers. Physician and Sports Medicine. 1985; 13(2), 94-101

20) Houmard JA, Costill DL, Mitchell JB, Park SH, Hickner RC, Roemmich JN. Reduced training maintains performance in distance runners. Int J Sports Med. 1990 Feb;11(1):46-52

21) Houmard JA, Scott BK, Justice CL, Chenier TC. The effects of taper on performance in distance runners. Med Sci Sports Exerc. 1994 May;26(5):624-31

22) Jackson AS, Beard EF, Wier LT, Ross RM, Stuteville JE, Blair SN. Changes in aerobic power of men, ages 25-70 yr. Med Sci Sports Exerc. 1995 Jan;27(1):113-20

23) Jackson AS, Wier LT, Ayers GW, Beard EF, Stuteville JE, Blair SN. Changes in aerobic power of women, ages 20-64 yr. Med Sci Sports Exerc. 1996 Jul;28(7):884-91

24) Dill DB, Robinson S, Ross JC. A longitudinal study of 16 champion runners. J Sports Med Phys Fitness. 1967 Mar;7(1):4-27

25) Hagerman FC, Fielding RA, Fiatarone MA, Gault JA, Kirkendall DT, Ragg KE, Evans WJ. A 20-yr longitudinal study of Olympic oarsmen. Med Sci Sports Exerc. 1996 Sep;28(9):1150-6

26) Pollock ML, Mengelkoch LJ, Graves JE, Lowenthal DT, Limacher MC, Foster C, Wilmore JH. Twenty-year follow-up of aerobic power and body composition of older track athletes. J Appl Physiol. 1997 May;82(5):1508-16

27) Trappe SW, Costill DL, Goodpaster BH, Pearson DR. Calf muscle strength in former elite distance runners. Scand J Med Sci Sports. 1996 Aug;6(4):205-10

28) Trappe SW, Costill DL, Vukovich MD, Jones J, Melham T. Aging among elite distance runners: a 22-yr longitudinal study. J Appl Physiol. 1996 Jan;80(1):285-90

29) Cureton KJ, Sparling PB. Distance running performance and metabolic responses to running in men and women with excess weight experimentally equated. Med Sci Sports Exerc. 1980;12(4):288-94

30) Dudley GA, Fleck SJ. Strength and endurance training. Are they mutually exclusive? Sports Med. 1987 Mar-Apr;4(2):79-85

31) Fleck SJ. Cardiovascular adaptations to resistance training. Med Sci Sports Exerc. 1988 Oct;20(5 Suppl):S146-51

32) Kraemer WJ, Duncan ND, Volek JS. Resistance training and elite athletes: adaptations and program considerations. J Orthop Sports Phys Ther. 1998 Aug;28(2):110-9

33) Kraemer WJ, Patton JF, Gordon SE, Harman EA, Deschenes MR, Reynolds K, Newton RU, Triplett NT, Dziados JE. Compatibility of high-intensity strength and endurance training on hormonal and skeletal muscle adaptations. J Appl Physiol. 1995 Mar;78(3):976-89