Physiological and Performance Effects of Low- versus Mixed-Intensity Rowing Training : Medicine & Science in Sports & Exercise

Journal Logo

APPLIED SCIENCES: Physical Fitness and Performance

Physiological and Performance Effects of Low- versus Mixed-Intensity Rowing Training


Author Information
Medicine & Science in Sports & Exercise 40(3):p 579-584, March 2008. | DOI: 10.1249/MSS.0b013e31815ecc6a
  • Free


Elite rowers possess large physiological capabilities, such as peak oxygen uptake (V˙O2peak), power associated with V˙O2peak (WV˙O2peak), power at lactate threshold (WLT), economical movement, and a fast primary component time constant of V˙O2 (7,11,12,26). It is likely that these characteristics are dependent on preexisting, genetically determined function and a product of training (11,16).

International standard rowers are known to perform about 90% of training at a low intensity (below LT; < 1.5-2 mM [blood lactate]) during the preparation period. During the precompetition phase, the proportion of low-intensity training accounted for 75% of the total volume (29,30). Steinacker (29) suggests that intense training (> 6.5 mM) should not exceed 10% of total training during preparatory periods, increasing to 25% of training in precompetition phases. A 14-20% increase in V˙O2peak and an 8-10% increase in the power associated with a blood lactate concentration of 4 mM (W4 mM) following 10-45 wk of rowing training has been associated with performance improvement (19,33). Improvements in mechanical efficiency are also associated with the number of years spent training or total training volume (10).

High-performance rowing programs, therefore, place little overall emphasis on the apparent requisite training intensity (80-100% V˙O2peak) considered necessary to develop maximal aerobic power (31). Further, training above the lactate threshold (intensity domain LT to maximum lactate steady state) is necessary to stimulate the development of the lactate response for conditioned athletes. However, observations of high-performance athletes show inclusion of large volumes of training below the LT and a smaller yet significant proportion of time between maximum lactate steady state and V˙O2peak (9,27,30). Improvements in mechanical efficiency are associated with the number of years spent training or total training volume (10) and, in particular, low-intensity training (21). It is not clear whether the changes in rowing movement economy would be related to the reduction in the V˙O2 slow component observed in response to a variety of endurance training studies (5,6).

The faster V˙O2 transition from rest to exercise observed in elite rowers (11) is in agreement with observations on "fit" individuals (17). The effect of training on the V˙O2 time constant may depend on analysis method, training status, duration of observation, and on predetermined characteristics such as muscle fiber composition (14,15). Consequently, it remains unclear whether the V˙O2 time constant response is influenced by training intensity.

Given the importance of a high V˙O2peak and LT for rowing performance, and the discrepancy between the perceived knowledge on parameter development versus training practice, a study investigating the effects of rowing-specific training models is warranted. This study aimed to quantify the physiological and performance effects of low-intensity (LOW, 100% below LT) versus mixed-intensity (MIX, 70% below LT with 30% at an intensity halfway between LT and V˙O2peak, 50%Δ) rowing training programs, with an additional emphasis on V˙O2 kinetics.


Following ethics approval from the regional committee, and the collection of written informed consent, 18 experienced (5.3 ± 3 yr competing, mean ± SD) national standard male rowers were randomly assigned to 12 wk of either low-intensity (LOW, N = 9, age 24.2 ± 7.1 yr; stature 181.8 ± 4.4 cm; mass 78.1 ± 5.0 kg) or a mix of low- and high-intensity (MIX, N = 9; age 23.8 ± 1.4 yr; stature 182.0 ± 4.9 cm; mass 74.5 ± 5.2 kg) endurance training.

Experimental Design

Subjects reported to the laboratory immediately following an off-season period of 25 ± 3 d of less than 10 ± 1% full training. Subjects were rested and fully hydrated, at least 2 h postprandial and having avoided strenuous exercise in the 24 h preceding a test session. All exercise tests were performed on an air-braked rowing ergometer (Concept II C) with a drag factor of 140. Subjects performed a progressive stepwise exercise test to volitional exhaustion, two "square-wave" transitions from rest to exercise, and a 2000-m time trial. All testing procedures were performed within a 7-d period before and after 12 wk of training.

Measurement of Lactate Threshold and V˙O2peak

Subjects self-paced a 10-min warm-up followed by an incremental test with five to six 4-min stages, interspersed by 30 s of rest for blood collection. Each stage saw an increased intensity of 25 W and two strokes per minute. Subjects rested for 150 s after the final incremental stage before performing a 4-min maximal effort. Earlobe capillary blood samples were assayed for [blood lactate] using a GM7 Analox analyzer (London, UK). Power at lactate threshold (WLT) was determined by two experienced, independent reviewers as the first lactate turnpoint in the profile of [blood lactate] against V˙O2 as a marked and sustained increase in [blood lactate] of 1 mM from baseline. Pulmonary gas exchange was measured throughout exercise. The highest 30-s V˙O2 value during maximum exercise was recorded as V˙O2peak. Solving the regression equation describing V˙O2 and exercise intensities calculated the power associated with V˙O2peak (WV˙O2peak).

Square-Wave Transitions

Following 2 min of seated rest, subjects performed two 6-min "square-wave" transitions from rest to heavy-intensity rowing (50% of the difference in V˙O2 between LT and V˙O2peak), interspersed with 60 min of rest. Stroke rate and power output were recorded stroke-by-stroke (PCi; Nottingham, UK), where acceleration of the ergometer flywheel to the desired power was achieved within 4.7 ± 1.1 s, or two to three strokes. After 12 wk of training, transitions were performed at the "new," relative 50%Δ intensity.

Pulmonary Gas Exchange

Gas exchange and minute ventilation were monitored breath-by-breath. Subjects wore a nose clip and breathed through a low-dead space (90 mL), low-resistance (0.1 kPa·L−1·s−1 at 15 L·s−1) mouthpiece. Air was sampled through a 2-m small-bore (0.5 mm) capillary line at a rate of 60 mL·min−1, where differential paramagnetism analyzed for V˙O2 and a sidestream infrared analyzer for V˙CO2 (Oxycon Alpha, Viasys, UK). Both instruments were precalibrated with reference concentration gases. Expiratory volumes were determined using a turbine volume transducer (Viasys, UK), precalibrated with a 3-L syringe. A computer integrated the volume and concentration signals, allowing for capillary-line transit delay. Respiratory gas-exchange variables (V˙O2, V˙CO2) were calculated for every breath.

Oxygen Uptake Kinetic Data

Data were interpolated for second-by-second values. Transitions were time-aligned to exercise onset and averaged to enhance the response characteristics. Nonlinear regression techniques were used to fit V˙O2-onset data with an exponential function, using a reiterative process-minimized sum of squared error. The mathematical model consisted of three (heavy-intensity exercise) exponential terms, each representing a phase of the response (2). The first exponential term started with the exercise onset, and each other term began after independent time delays based on the asymptotic value of the previous component.

where V˙O2(b) is the resting baseline value; Ac, APC, and ASC are the asymptotic amplitudes for the exponential term; τc, τPC, and τSC are the time constants; and TDPC and TDSC are the time delays. Amplitudes expressed per unit of work are expressed as a gain above baseline; gainC, gainPC, and gainSC. The cardiodynamic component term was terminated at the start of the primary component and was assigned the value for that time (AC′):

The V˙O2 at the end of phase 1 (Ac′) and the amplitude of primary component (APC) were summed to calculate the amplitude of the primary component (APC′). The V˙O2 slow component was calculated at the end of exercise (ASC′).

Ergometer 2000-m Performance Trial


A program of weekly training was prescribed with an allowance of ± 10% of training volume suggested to accommodate for varying standard of athlete. Training zones were presented as power output, 500-m split time, and heart rate, to enable equivalent ergometer and "on-water" training. A diary record of all training ascertained the balance of training and program adherence.

Statistical Analysis

A mixed-factorial, two-way repeated-measures ANOVA was used to analyze data by time and group, and independent t-tests were used to compare groups. Pearson's product-moment correlation was used to examine the relationship between physiological variables and performance. The alpha level was 5%, and data are presented as means ± SD.


No physiological or performance differences existed between groups prior to training. Training was translated from power output (ergometer) and 500-m split time (on water) from session-by-session records. LOW training equated to the required level (98%), and MIX involved 28% training above LT compared with the planned 30% (Table 1).

Prescribed and executed training for rowing in LOW and MIX regimes.

Performance and Physiological Response

Sixteen out of 18 athletes achieved a new personal record in the performance trial at the end of the training period (LOW, 5.04 ± 0.11 m.s−1; MIX, 4.99 ± 0.12 m.s−1, prior to the week 1 performance trial). Ergometer 2000-m time trial performance improved (P < 0.001) for LOW (2.0%) and MIX (1.4%), but independently of group (P = 0.42). The V˙O2peak (P < 0.001) and WV˙O2peak (P = 0.005) increased with training, but not differently between groups (P = 0.8 and 0.14, respectively). With training, WLT and W4mM improved (P < 0.001); however, LOW improved more than MIX (WLT, P = 0.013; W4mM, P = 0.03) (Table 2).

Effects of 12 wk of LOW and MIX training on physiological and performance response and intergroup correlation coefficient of the change against that in 2000-m rowing ergometer performance speed.

Oxygen Uptake Kinetics

The time constants and time delays of the primary (P = 0.25 and 0.35, respectively) and slow (P = 0.30 and 0.82, respectively) components were no different after training. The amplitude of the V˙O2 primary component expressed in relation to exercise intensity (gainPC′) showed no differences with training (P = 0.13), whereas the V˙O2 slow component expressed relative to end-exercise V˙O2 (relative ASC′, P = 0.045) and per watt (gainSC′, P = 0.045) were reduced with training, but independently of group.

Correlations with Performance

The changes in each parameter were correlated (intergroup) with the change in 2000-m speed (Table 1). The strongest relationship observed was between ΔWV˙O2peak and Δ2000-m speed (r = −0.83).


This study examined the effects of low-intensity sub-LT training (LOW) versus a mix of low- and high-intensity (MIX) training and found that both groups improved ergometer rowing performance, V˙O2peak, and WV˙O2peak, whereas LOW improved the work performed at a given [blood lactate] more so than MIX.

Effect of Training

A meta-analysis indicated that greater improvements in LT are associated with higher-intensity training, but from mainly sedentary populations (18). The highly trained rower aims to advance the power at which muscle fatigue occurs, of which LT may be a surrogate marker (12,20,25). The optimal training intensity for LT progression, from intervention-based studies versus observations of high-performance rowing training, are contradictory (9,18,25,30).

Frequent high-intensity training is linked with increased risk of overtraining (4). Thus, high-performance athletes might naturally self-select low-intensity training as a method of maintaining progression in total training volume or reducing the demand for recovery (28,30). The training plan for MIX replicated the precompetitive phase of rowing training, providing a valid stimulus (but similar to LOW) contrasting with studies comparing training responses to more divergent programs (8). MIX involved higher-intensity bouts on approximately 3 d·wk−1. No fatigue symptoms were recorded, indicating that excessive sympathetic stress was unlikely to have been experienced by MIX (17). If a reduced peak lactate is a marker of sympathetic nervous system suppression (13), its stability before and after training (9.0 mM for week 1 vs 9.6 mM for week 12) would hint toward adequate recovery for MIX. Indeed, although peak lactate seemed to respond divergently between groups, this was not significant (P = 0.12). If the anaerobic capacity was preserved/developed more in MIX than LOW, this is likely to have compensated for the smaller change in aerobic capacity where performance improvements are similar.

It is not clear whether the improved rate of LT adaptation for LOW translates to superior exercise tolerance through reduced acidosis, improved cell-to-cell or within-cell lactate shuttling, or oxidation; of the latter, a major component is monocarboxylic transporters (MCT). The MCT1 isoform is predominantly expressed in slow oxidative fibers and MCT4 in fast-glycolytic fibers (22), and expression and concentration of MCT increases with both chronic endurance and intensive training and is related to submaximal and maximal [blood lactate] (32). It is not obvious how sensitive MCT isoform expression or its measurement is to divergent training. Previously, responses to high- and low-intensity cross-country skiing regimes have found that MCT concentration mirrored the change in LT, but, in contrast to our study, this was primarily attributable to a reduced MCT and LT in the low-intensity group (8).

High-intensity training is predicated as particularly important for V˙O2peak development, for athletes already undertaking a minimum amount of training (31). However, improvements observed in aerobic power (~10% increase in V˙O2peak for both groups) occur independently of training intensity. At the onset of a training program, particularly after a period of reduced training, changes in V˙O2peak have been attributed to hypervolemia (24). This effect may be consistent with the current findings because vascular volume increases independently of training intensity (3).

Improvements in WV˙O2peak were no different between groups. As V˙O2peak increased in response to training, WV˙O2peak increased concomitantly. Changes in WV˙O2peak can also result from a reduction in submaximal V˙O2, although the V˙O2 relative to power output was not reduced for either group. Power at V˙O2peak was identified as a strong correlate of rowing performance (12). The correlation between ΔWV˙O2peak and Δperformance speed (r = −0.83) supports the importance of this parameter to rowing performance regardless of the training method used.

The necessity for high volumes of low-intensity training for rowing performance may be related to factors not measured in the current study, such as neural entrainment. The low-intensity approach would be most applicable to noninnate locomotory neuromuscular pattern development of rowing. Rowing skill autonomy would require disproportionately greater iterative repetition than innate action such as running. There is also interplay between contraction frequency and intensity, which is common to kayak paddling and swimming, where speed is more dependent on contraction frequency than running. Thus, where a higher duty of contraction can be applied with lower contraction rates, greater "performance" specificity can be achieved in the motor unit pool enlisted, thus negating the need to perform larger amounts of high-intensity work.

Oxygen Uptake Kinetics

The V˙O2 primary component time constant did not change in response to the same relative load (50%Δ) following training. Carter et al. (5) note that for less-fit subjects, τPC became quicker with training. In the current study, LOW appeared to increase the speed of τPC (−21%); compared with MIX (+1.4%), the difference was not significant. Whipp et al. (34) comment that if the time constant was a determining factor of endurance performance ability, then it was a "frail" one. Although a moderate significant association (r = −0.59) between τPC and 2000-m ergometer performance has been noted (12), in relation to training-based changes, our data support Whipp's assertion.

The significant reduction in ASC′ (11% for LOW; 36% for MIX) suggests an improvement in "respiratory efficiency" at the same relative intensity, and, by inference, the same absolute intensity. The size (%) of the V˙O2 slow-component reduction shown was similar to reports of continuous and interval training-induced changes (5,6). Poole et al. (23) found that 86% of the V˙O2 slow component arises from the exercising limbs. Whereas no measure of motor unit recruitment was recorded, it may be reasonable to purport that the reduction in the V˙O2 slow component was attributable to improvements in the oxidative capacity of the motor unit pool recruited during heavy-intensity exercise. Reponses might indicate that LOW and MIX were too similar for differential adaptation to be noted in slow component size.


LOW and MIX training regimes have similar effects on both performance and physiological training markers. The response of [blood lactate] to submaximal exercise did react differently to training intervention, improving by a greater margin in LOW. Characteristics of the V˙O2 kinetics appear unresponsive to 12 wk of preparatory and precompetition training programs in rowers.

The authors are thankful for the support of the British Olympic Medical Trust.


1. Altenburg D. Planning and organising general fitness training for rowers. FISA Coach. 1992;12:5-12.
    2. Barstow TJ, Mole PA. Linear and nonlinear characteristics of oxygen uptake kinetics during heavy exercise. J Appl Physiol. 1991;71:2099-6.
    3. Branch JD 3rd, Pate RR, Bourque SP, Convertino VA, Durstine JL, Ward DS. Exercise training and intensity does not alter vascular volume responses in women. Aviat Space Environ Med. 1999;70:1070-6.
    4. Bruin G, Kuipers H, Keizer HA, Vander Vusse GJ. Adaptation and overtraining in horses subjected to increasing training loads. J Appl Physiol. 1994;76:1908-3.
    5. Carter H, Jones AM, Barstow TJ, Burnley M, Williams C, Doust JH. Effect of endurance training on oxygen uptake kinetics during treadmill running. J Appl Physiol. 2000;89:1744-2.
    6. Casaburi R, Storer TW, Ben Dov I, Wasserman K. Effect of endurance training on possible determinants of V˙O2 during heavy exercise. J Appl Physiol. 1987;62:199-207.
    7. Cosgrove MJ, Wilson J, Watt D, Grant SF. The relationship between selected physiological variables of rowers and rowing performance as determined by a 2000 m ergometer test. J Sports Sci. 1999;17:845-52.
    8. Evertsen F, Medbo JI, Bonen A. Effect of training intensity on muscle lactate transporters and lactate threshold of cross-country skiers. Acta Physiol Scand. 2001;173:195-205.
    9. Fiskerstrand A, Seiler KS. Training and performance characteristics among Norwegian international rowers 1970-2001. Scand J Med Sci Sports. 2004;14:303-10.
    10. Hagerman FC. Applied physiology of rowing. Sports Med. 1984;1:303-26.
    11. Ingham SA, Carter H, Whyte GP, Doust JH. Comparison of the oxygen uptake kinetics of club and olympic champion rowers. Med Sci Sports Exerc. 2007;39(5):865-71.
    12. Ingham SA, Whyte GP, Jones K, Nevill AM. Determinants of 2,000 m rowing ergometer performance in elite rowers. Eur J Appl Physiol. 2002;88:243-6.
    13. Jeukendrup AE, Hesselink MK. Overtraining-what do lactate curves tell us? Br J Sports Med. 1994;28:239-40.
    14. Jones AM, Koppo K. Effect of training on V˙O2 kinetics and performance. In: Jones AM, Poole DC, editors. Oxygen Uptake Kinetics in Sport, Exercise and Medicine. London (UK): Routledge; 2005. p. 373-98.
    15. Jones AM, Pringle JS, Carter H. Influence of muscle fibre type and motor unit recruitment on V˙O2 kinetics. In: Jones AM, Poole DC, editors. Oxygen Uptake Kinetics in Sport, Exercise and Medicine. London (UK): Routledge; 2005. p. 261-93.
    16. Larsson L, Forsberg A. Morphological muscle characteristics in rowers. Can J Appl Sport Sci. 1980;5:239-44.
    17. Lehmann M, Baumgartl P, Wiesenack C, et al. Training-overtraining: influence of a defined increase in training volume vs training intensity on performance, catecholamines and some metabolic parameters in experienced middle- and long-distance runners. Eur J Appl Physiol Occup Physiol. 1992;64:169-77.
    18. Londeree BR. Effect of training on lactate/ventilatory thresholds: a meta-analysis. Med Sci Sports Exerc. 1997;29(6):837-43.
    19. Mahler DA, Parker HW, Andresen DC. Physiologic changes in rowing performance associated with training in collegiate women rowers. Int J Sports Med. 1985;6:229-33.
    20. Nielsen OB, Overgaard K. Point:counterpoint authors respond to commentaries on "Lactic acid accumulation is an advantage/disadvantage during muscle activity". J Appl Physiol. 2006;101:367.
    21. Pate RR, Macera CA, Bailey SP, Bartoli WP, Powell KE. Physiological, anthropometric, and training correlates of running economy. Med Sci Sports Exerc. 1992;24(10):1128-33.
    22. Pilegaard H, Terzis G, Halestrap A, Juel C. Distribution of the lactate/H+ transporter isoforms MCT1 and MCT4 in human skeletal muscle. Am J Physiol. 1999;276:E843-8.
    23. Poole DC, Schaffartzik W, Knight DR, et al. Contribution of excising legs to the slow component of oxygen uptake kinetics in humans. J Appl Physiol. 1991;71:1245-60.
    24. Saltin B, Blomqvist G, Mitchell JH, Johnson RL Jr., Wildenthal K, Chapman CB. Response to exercise after bed rest and after training. Circulation. 1968;38:VII1-78.
    25. Secher NH. Physiological and biomechanical aspects of rowing. Implications for training. Sports Med. 1993;15:24-42.
    26. Secher NH, Vaage O, Jensen K, Jackson RC. Rowing performance and maximal aerobic power of oarsmen. Scand J Sports Sci. 1982;4:9-11.
    27. Seiler KS, Kjerland GO. Quantifying training intensity distribution in elite endurance athletes: is there evidence for an "optimal" distribution? Scand J Med Sci Sports. 2006;16:49-56.
    28. Seiler S, Haugen O, Kuffel E. Autonomic recovery after exercise in trained athletes: intensity and duration effects. Med Sci Sports Exerc. 2007;39(8):1366-73.
    29. Steinacker JM. Physiological aspects of training in rowing. Int J Sports Med. 1993;14(Suppl. 1):S3-10.
    30. Steinacker JM, Lormes W, Lehmann M, Altenburg D. Training of rowers before world championships. Med Sci Sports Exerc. 1998;30(7):1158-63.
    31. Tabata I, Irisawa K, Kouzaki M, Nishimura K, Ogita F, Miyachi M. Metabolic profile of high intensity intermittent exercises. Med Sci Sports Exerc. 1997;29(3):390-5.
    32. Thomas C, Perrey S, Lambert K, Hugon G, Mornet D, Mercier J. Monocarboxylate transporters, blood lactate removal after supramaximal exercise, and fatigue indexes in humans. J Appl Physiol. 2005;98:804-9.
    33. Vermulst LJ, Vervoorn C, Boelens-Quist AM, et al. Analysis of seasonal training volume and working capacity in elite female rowers. Int J Sports Med. 1991;12:567-72.
    34. Whipp BJ, Rossiter HB, Ward SA. Exertional oxygen uptake kinetics: a stamen of stamina? Biochem Soc Trans. 2002;30:237-47.


    ©2008The American College of Sports Medicine