Intermittent exercise alters the heart rate–blood lactate relationship used for calculating the training impulse (TRIMP) in team sport players

Available online at www.sciencedirect.com

Journal of Science and Medicine in Sport 14 (2011) 249–253

Original research

Intermittent exercise alters the heart rate–blood lactate relationship used for calculating the training impulse (TRIMP) in team sport players Ibrahim Akubat, Grant Abt ∗ Department of Sport, Health and Exercise Science, The University of Hull, UK Received 15 February 2010; received in revised form 30 November 2010; accepted 10 December 2010

Abstract Objectives: The training impulse (TRIMP) quantifies training dose by weighting heart rate according to the relationship between fractional elevation in heart rate (HR) and blood lactate concentration (BLa). This study compared the physiological responses to intermittent and continuous exercise and their influence on TRIMP weightings. Design: Repeated measures crossover. Method: Twelve team sport players undertook a vVO2max test and then a continuous trial (CT) and intermittent trial (IT) in a randomised order. Each trial consisted of 4 × 4 min stages corresponding to 25%, 50%, 75% and 100% of vVO2max . Trials were matched for distance and mean speed. Results: A repeated measures ANOVA revealed higher BLa for IT at 75% vVO2max (p = 0.023) and 100% vVO2max (p = 0.012); higher VO2 for IT at 25% vVO2max (p < 0.001); higher HR for IT at 25% vVO2max (p < 0.001), 75% vVO2max (p = 0.03) and 100% vVO2max (p = 0.018); higher TRIMP weightings for IT at 0.9 HR (p = 0.018) and 1.0 HR (p = 0.005). Conclusions: Intermittent exercise alters the HR–BLa relationship and TRIMP weightings at high exercise intensities. Determination of the training impulse from the HR–BLa relationship derived from a continuous exercise protocol may underestimate the exercise ‘dose’ of training and/or matches in team sport players. © 2010 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. Keywords: Intermittent exercise; Training impulse; Blood lactate; Heart rate

1. Introduction Banister1,2 introduced the concept of the training impulse (TRIMP), which integrates training intensity and training duration into a single overall measure of training ‘dose’. The TRIMP uses the exponential relationship between fractional elevation in heart rate (HR) and blood lactate concentration (BLa), as observed during incremental exercise, to ‘weight’ exercise at a particular intensity. This method provides an exponentially higher weighting for higher intensity sessions. The TRIMP was originally used with endurance athletes such as swimmers2 where the heart rate during a training session or competition is normally well within the aerobic range. However, the TRIMP has recently been examined in studies on team sport players3–5 where the high-intensity intermittent nature of games such as soccer might alter the HR–BLa relationship. ∗

Corresponding author. E-mail address: [email protected] (G. Abt).

The mean intensity in a soccer game has been reported to be around 87% of maximal heart rate (HRmax ),6 but the intermittent nature of the sport means the heart rate can also rise to its maximum.6 For example, Helgerud et al.7 reported youth players spend about 15–20% of match time at intensities >90% HRmax . Consequently, using mean heart rate for calculating the TRIMP in soccer and other team sports may lead to a loss of specific information and therefore not be reflective of the overall intensity or variation in the intensity of a match or training session. In an attempt to alleviate this problem Stagno et al.5 developed a modified ‘group’ TRIMP for use with hockey players where five zones were created around typical BLa thresholds (LT and OBLA) on a HR–BLa curve and each zone weighted according to the regression equation of the curve. The five zone weightings derived from this method were the same for each player, being 1.25, 1.71, 2.54, 3.61, and 5.16 for zones one to five, respectively. The time spent in each zone was multiplied by the relevant weighting to provide a TRIMP. Stagno et al.5 reported a significant relationship (r = 0.8) between the per-

1440-2440/$ – see front matter © 2010 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jsams.2010.12.003

250

I. Akubat, G. Abt / Journal of Science and Medicine in Sport 14 (2011) 249–253

centage change in maximal oxygen uptake (VO2max ) and the mean weekly modified TRIMP. However, the equation used to weight the time in each zone was derived from a continuous incremental test. There is evidence to suggest the blood lactate response to intermittent exercise is different when compared to continuous exercise, particularly at higher intensities.8–11 While these studies8–11 shed light on the differences between intermittent and continuous exercise, the results are limited by exercise mode (cycling) and low sample sizes (n = 3),10,11 lack of control over distance covered in each condition,8 and the use of only one treadmill speed (120% VO2max ).9 While measurement of BLa during continuous incremental tests for the determination of training adaptations or monitoring of training loads for intermittent team sports is advocated,5,12,13 no study has examined the potential effect of intermittent exercise on BLa and the subsequent calculation of weightings for the determination of the TRIMP. As team sport players engage in intermittent exercise during training and matches, then any effect of this mode of exercise on calculation of the TRIMP could lead to overor under-estimation of the training dose, which in turn has implications for player fitness and injury.14 The aim of the present study was therefore to investigate the effect of intermittent exercise on the HR–BLa relationship and its influence on the weightings used to generate the TRIMP.

2. Methods Twelve university-level male team sport players (age 19.4 ± 0.5 years; stature 1.75 ± 0.06 m; body mass 67.4 ± 11.6 kg; VO2max 55 ± 12 mL kg−1 min−1 ; vVO2max 16.0 ± 1.1 km h−1 ) volunteered for the study. They all competed regularly in their chosen team sport. Written informed consent was obtained prior to their participation. The study was approved by the Departmental Ethics Committee and conformed to the Declaration of Helsinki. The resting heart rate (HRrest ) of all participants was measured using a heart rate monitor (Polar FS1, Polar Electro, OY, Finland) sampling at 5 s intervals. Participants lay supine in a quiet room for 10 min and the lowest 5 s heart rate was recorded as the HRrest . HRmax , maximal oxygen uptake (VO2max ) and the velocity at VO2max (vVO2max ) were measured during an incremental protocol on a motorised treadmill (Woodway PPS 55sport, Woodway, Germany). Treadmill gradient was set at 1% for the entire test to reflect the energetic cost of outdoor running.15 Oxygen consumption (VO2 ) and heart rate were collected throughout the test using a breathby-breath system (Quark b2 , Cosmed Srl, Rome, Italy) which was calibrated before and after each test according to the manufacturer’s instructions. Cryospec calibration gases of 16.0% O2 , 4.5% CO2 , and N2 balance were used (Cryoservice Ltd, Worcester, UK). Participants performed a warm up by cycling (Monark 842E, Monark Exercise AB, Varberg, Sweden) at 75 W for 5 min. The treadmill protocol for the measurement

of HRmax , VO2max and vVO2max started at 8 km h−1 min−1 and increased by 1 km h−1 min−1 until volitional exhaustion. VO2max was recorded as the highest mean VO2 obtained for any 1 min period during the test.16 At least two of the following criteria were also required for the attainment of VO2max : a plateau in VO2 (defined as a change of less than 0.2 L min−1 ) despite increasing treadmill speed, respiratory exchange ratio >1.15, or the attainment of age predicted maximum heart rate.16 The lowest treadmill speed required to elicit VO2max was considered to be the vVO2max .17 Each participant undertook a continuous trial (CT) and an intermittent trial (IT) in a randomised order separated by a minimum of 48 h. During each trial VO2 and heart rate were collected using a breath-by-breath system that was calibrated prior to and after each trial following the same procedures as for the vVO2max test. The CT consisted of running on a motorised treadmill at speeds corresponding to 25%, 50%, 75% and 100% of vVO2max . Participants ran at each speed for 4 min with a 1 min rest between each stage, during which a fingertip blood sample was taken and subsequently analysed in duplicate for BLa (YSI 2300, YSI Inc., Yellow Spring, OH), with the mean value used for data analysis. The IT also consisted of running on a motorised treadmill for 4 min periods at 25%, 50%, 75% and 100% of vVO2max . However, in the IT participants ran intermittently at each intensity by alternating their speed between a higher and lower speed for 15 s each, so that the mean of those two speeds was the same as that used in the CT. For example, a participant with a vVO2max of 16 km h−1 , running at 25% vVO2max would be stationary for 15 s at 0 km h−1 and then run for 15 s at 8 km h−1 ; at 50% vVO2max they would run for 15 s at 4 km h−1 and then run for 15 s at 12 km h−1 ; at 75% vVO2max they would run for 15 s at 8 km h−1 and then run for 15 s at 16 km h−1 ; at 100% vVO2max they would run for 15 s at 12 km h−1 and then run for 15 s at 20 km h−1 , thereby matching the same average speeds as used in the CT. The distance covered for each pair of trials was matched so that the difference was as small as possible (25% vVO2max : CT 264 ± 19 m, IT 266 ± 19 m, difference = 0.6%; 50% vVO2max : CT 527 ± 37 m, IT 529 ± 39 m, difference = 0.5%; 75% vVO2max : CT 790 ± 56 m, IT 784 ± 54 m, difference = 0.8%; 100% vVO2max : CT 1053 ± 72 m, IT 1036 ± 70 m, difference = 1.6%). Treadmill gradient during CT and IT was set at 1%. The heart rate data from the CT and IT were expressed as HR (HRexercise − HRrest /HRmax − HRrest ) while the oxygen uptake data were expressed in mL kg−1 min−1 . The mean HR and VO2 during the last minute of each stage were used for subsequent analysis. Given that the method used by Stagno et al.5 uses a ‘group’ approach for the calculation of TRIMP weightings, it was not possible to conduct a statistical analysis of the effect of continuous and intermittent exercise on TRIMP weightings using their method as the weighting would be the same for all players. Consequently, an individualised TRIMP18 was used to determine weightings based on a player’s own

I. Akubat, G. Abt / Journal of Science and Medicine in Sport 14 (2011) 249–253

251

Table 1 Mean ± S.D. of BLa, VO2 and HR for both trials at each intensity. vVO2max (%)

BLa (mmol L−1 ) CT

25% 50% 75% 100% a b

1.5 2.4 4.5 10.8

VO2 (mL kg−1 min−1 ) IT

± ± ± ±

0.5b 0.4b 0.8b 2.1b

1.6 2.2 5.6 12.4

CT ± ± ± ±

0.3b 0.7b 1.7a,b 2.9a,b

13 31 44 54

± ± ± ±

HR

IT 3b 3b 3b 5b

20 31 47 55

CT ± ± ± ±

2a,b 3b 4b 5b

0.21 0.55 0.79 0.95

IT ± ± ± ±

0.08b 0.09b 0.06b 0.04b

0.35 0.55 0.82 0.97

± ± ± ±

0.08a,b 0.08b 0.05a,b 0.03a,b

Significantly different to CT at that intensity. Significantly different from all other intensities within that trial.

HR–BLa data. Each player’s HR–BLa relationship was plotted using SPSS Version 16 (SPSS Inc., Chicago, USA) and an individual exponential curve generated. From this individualised exponential curve, a TRIMP weighting could then be determined from any HR. To demonstrate the effect of intermittent exercise on the TRIMP weighting we chose four arbitrary HRs (0.7, 0.8, 0.9 and 1.0), which cover the range of HRs that would be observed in both training and match play. Data are expressed as mean ± SD. Prior to parametric analysis the assumptions of normality were verified by using the Shapiro Wilk test. A repeated measures analysis of variance with Sidak pairwise comparisons was used to identify significant differences within and between conditions. All statistical analyses were conducted using SPSS version 16. Statistical significance was set at p ≤ 0.05. When significant interactions were found the mean difference and 95% confidence interval for the mean difference are reported. Cohen effect sizes (ES) and their qualitative interpretation as defined by Hopkins19 (0–0.19 trivial; 0.2–0.59 small; 0.6–1.19 moderate; 1.2–1.99 large; ≥2.0 very large) are also reported for significant interactions.

3. Results Table 1 shows that the BLa in the IT was significantly higher than CT at 75% vVO2max (mean difference = 1.1 mmol L−1 ; p = 0.023; 95%CI = 0.2–2.0 mmol L−1 ; ES = 1.4; large effect) and 100% of vVO2max (mean difference = 1.6 mmol L−1 ; p = 0.012; 95%CI = 0.4–2.8 mmol L−1 ; ES = 0.8; moderate effect). The VO2 in the IT was significantly higher than CT only at 25% vVO2max (mean difference = 6 mL kg−1 min−1 ; p < 0.001; 95%CI = 4–8 mL kg−1 min−1 ; ES = 2.2; very large effect). The HRs in the IT were significantly higher than CT at 25% vVO2max (mean difference = 0.14; p < 0.001; 95%CI = 0.10–0.18; ES = 2.0; very large effect), 75% vVO2max (mean difference = 0.03; p = 0.03; 95%CI = 0.003–0.05; ES = 0.50; small effect) and 100% vVO2max (mean difference = 0.02; p = 0.018; 95%CI = 0.005–0.04; ES = 0.57; small effect). Table 2 shows that TRIMP weightings derived from the IT were significantly higher than those from the CT at 0.9 HR (mean difference = 1.03; p = 0.018; 95%CI = 0.22–1.85;

Table 2 Weightings derived from the individualised TRIMP method. HR

TRIMP weightings CT

0.7 0.8 0.9 1.0

4.26 5.46 7.04 9.20

Effect size IT

± ± ± ±

0.55 0.55 0.72 1.22

4.19 5.81 8.07 11.25

± ± ± ±

0.85 1.17 1.73a 2.65a

0.1 (trivial) 0.6 (moderate) 1.4 (large) 1.7 (large)

TRIMP CT, individualised TRIMP based on the weightings generated from a continuous trial; TRIMP IT, individualised TRIMP based on the weightings generated from an intermittent trial. CT, continuous trial; IT, intermittent trial. a Significantly different to CT.

ES = 1.4; large effect) and 1.0 HR (mean difference = 2.15; p = 0.005; 95%CI = 0.80–3.5; ES = 1.7; large effect).

4. Discussion The major finding of the present study is that the HR–BLa relationship is altered by intermittent exercise. This alteration results in large changes to the TRIMP weightings at high-intensities (0.9 HR and 1.0 HR) compared to TRIMP weightings generated from a continuous exercise test and could therefore result in substantial underestimation of the TRIMP in team sport players. Previous research8–11 has highlighted the higher BLa concentrations during intermittent exercise when compared to continuous exercise at the same mean power or speed. In the present study the moderate to large increases in BLa concentration in the IT at 75% vVO2max and 100% vVO2max compared to the CT support these previous findings. It is not surprising that higher BLa concentrations are observed with intermittent exercise when the more intense periods are above the lactate threshold,20 as would have been the case in IT at 75% vVO2max and 100% vVO2max . The higher BLa observed during IT in the present study has resulted in a change in the relationship between HR and BLa, resulting in large changes in the TRIMP weightings at the higher intensities (0.9 and 1.0 HR). Stagno et al.5 used a continuous incremental protocol to establish the HR–BLa relationship whereas the training and match play requirements for team sport players are very much intermittent. As can be seen in Table 2, the differences in the TRIMP weightings at the

252

I. Akubat, G. Abt / Journal of Science and Medicine in Sport 14 (2011) 249–253

high end of the intensity continuum where training has been suggested to be most important7 are substantially different between those generated from CT and those generated from IT. It should be noted that the HRs that we have found to be substantially altered by intermittent exercise (0.9 and 1.0) are very high intensities and that the time spent at these intensities during training and playing is relatively small (∼15–20%) compared to that spent at lower intensities.7 However, given that the weightings generated by any of the TRIMP methods increase in an exponential manner, any time spent at these very high intensities will disproportionately increase the total TRIMP. We can only speculate at this stage as to what effect this might have on the relationships between the TRIMP and changes in fitness and/or team sport performance, and therefore training studies are required to further explore these relationships. Although intermittent exercise caused a substantial increase in the TRIMP weightings at high exercise intensities, it appears that most of this increase can be explained by the moderate to large increases in BLa observed at those intensities, rather than changes in HR. This is because the changes in HR at 75% vVO2max and 100% vVO2max were small. Additionally, the VO2 at 75% vVO2max and 100% vVO2max were not significantly different between CT and IT. The similarity in VO2 between CT and IT largely supports previous research by Drust et al.21 who reported no significant difference in VO2 between continuous and intermittent protocols matched for mean speed over a 46 min period. The use of BLa to provide a ‘weighting’ to the intensity of exercise suggests that it is reflective of the overall physiological ‘stress’ imposed on the athlete at a given exercise intensity. Alternatively, as BLa is the net result of lactate production in the muscle, its release from the muscle, and uptake by other tissues,22 a change in BLa may reflect any of those factors independently and not necessarily the overall physiological stress of exercising at that intensity. Moreover, measurements of BLa do not necessarily reflect muscle lactate concentrations.23,24 There is conflicting evidence about the relationship between increasing exercise intensity and physiological stress. Some authors suggest this relationship is exponential, and therefore similar to the manner in which BLa increases with exercise intensity.25 This suggestion is based on the changes in hormones such as adrenaline, which has been reported to increase exponentially with increasing exercise intensity.26 However, studies in comparative physiology suggest a linear relationship.27

5. Conclusion In conclusion, we have found that intermittent exercise significantly and substantially alters the HR–BLa relationship and subsequent calculation of TRIMP weightings when based on an individualised HR–BLa relationship. This alteration could result in substantial underestimation of the TRIMP in team sport players. Our findings have implica-

tions for the use of the HR–BLa relationship derived from a continuous exercise protocol for the calculation of heart rate based TRIMP in team sport players.

Practical implications • Determination of the training impulse from the HR–BLa relationship derived from a continuous exercise protocol may underestimate the exercise ‘dose’ of training and/or matches in team sport players. • As this underestimation is most prominent at the highest intensities, the absolute exercise intensity and/or the time spent at these high intensities would have to be lowered if a given TRIMP is to be achieved. Alternatively, if the absolute exercise intensity and/or time spent at these high intensities are maintained, then the actual ‘dose’ will be higher than that which is recorded. This might have implications for the development of injury or overtraining and should therefore be considered when monitoring player loads.

Disclosures We have nothing to disclose.

Acknowledgements We would like to thank Mark Turner and Dr Ric Lovell for their assistance with this study.

References 1. Banister EW. Modeling elite athletic performance. In: Physiological testing of elite athletes. Champaign: Human Kinetics; 1991. p. 403–24. 2. Banister EW, Calvert TW, Savage MV, et al. A systems model of training for athletic performance. Aust J Sports Med 1975;7(3):57–61. 3. Alexiou H, Coutts A. A comparison of methods used for quantifying internal training load in women soccer players. Int J Sports Physiol Perform 2008;3(3):320–30. 4. Impellizzeri FM, Rampinini E, Coutts AJ, et al. Use of RPE-based training load in soccer. Med Sci Sports Exerc 2004;36(6):1042–7. 5. Stagno KM, Thatcher R, van Someren KA. A modified TRIMP to quantify the in-season training load of team sport players. J Sports Sci 2007;25(6):629–34. 6. Ascensão A, Rebelo A, Oliveira E, et al. Biochemical impact of a soccer match—analysis of oxidative stress and muscle damage markers throughout recovery. Clin Biochem 2008;41(10–11):841–51. 7. Helgerud J, Engen LC, Wisløff U, et al. Aerobic endurance training improves soccer performance. Med Sci Sports Exerc 2001;33(11):1925–31. 8. Bangsbo J. The physiology of soccer—with special reference to intense intermittent exercise. Acta Physiol Scand 1994;151(Suppl. 619):1–155. 9. Christmass MA, Dawson B, Passeretto P, et al. A comparison of skeletal muscle oxygenation and fuel use in sustained continuous and intermittent exercise. Eur J Appl Physiol 1999;80:423–35.

I. Akubat, G. Abt / Journal of Science and Medicine in Sport 14 (2011) 249–253 10. Edwards RHT, Melcher A, Hesser CM, et al. Blood lactate concentrations during intermittent and continuous exercise with the same average power output. In: Muscle metabolism during exercise: advances in experimental medicine and biology. New York: Plenum Press; 1971. p. 425–8. 11. Edwards RHT, Melcher A, Hesser CM, et al. Physiological correlates of perceived exertion in continuous and intermittent exercise with the same average power output. Eur J Clin Invest 1972;2:108–14. 12. Impellizzeri FM, Rampinini E, Marcora SM. Physiological assessment of aerobic training in soccer. J Sports Sci 2005;23(6):583–92. 13. McMillan K, Helgerud J, Macdonald R, et al. Physiological adaptations to soccer specific endurance training in professional youth soccer players. Br J Sports Med 2005;39(5):273–7. 14. Gabbett TJ, Domrow N. Relationships between training load, injury, and fitness in sub-elite collision sport athletes. J Sports Sci 2007;25(13):1507–19. 15. Jones AM, Doust JH. A 1% treadmill grade most accurately reflects the energetic cost of outdoor running. J Sports Sci 1996;14(4):321–7. 16. Lucia A, Hoyos J, Perez M, et al. Which laboratory variable is related with time trial performance time in the Tour de France? Br J Sports Med 2004;38(5):636–40. 17. Billat VL, Blondel N, Berthoin S. Determination of the velocity associated with the longest time to exhaustion at maximal oxygen uptake. Eur J Appl Physiol 1999;80:159–61. 18. Manzi V, Iellamo F, Impellizzeri FM, et al. Relation between individualized training impulses and performance in distance runners. Med Sci Sports Exerc 2009;41(11):2090–6.

253

19. Hopkins WG. A new view of statistics. Available at: [accessed 26.01.10]. 20. Billat VL, Slawinski J, Bocquet V, et al. Intermittent runs at the velocity associated with maximal oxygen uptake enables subjects to remain at maximal oxygen uptake for a longer time than intense but submaximal runs. Eur J Appl Physiol 2000;81:188– 96. 21. Drust B, Reilly T, Cable NT. Physiological responses to laboratorybased soccer-specific intermittent and continuous exercise. J Sports Sci 2000;18:885–92. 22. Svedahl K, MacIntosh BR. Anaerobic threshold: the concept and methods of measurement. Can J Appl Physiol 2003;28(2):299– 323. 23. Medbo J. Glycogen breakdown and lactate accumulation during highintensity cycling. Acta Physiol Scand 1993;149(1):85–9. 24. Krustrup P, Mohr M, Steensberg A, et al. Muscle and blood metabolites during a soccer game: implications for sprint performance. Med Sci Sports Exerc 2006;38(6):1165–74. 25. Norton K, Norton L, Sadgrove D. Position statement on physical activity and exercise intensity terminology. J Sci Med Sport 2009, doi:10.1016/j.jsams.2009.09.008. 26. Richardson RS, Kennedy B, Knight DR, et al. High muscle blood flows are not attenuated by recruitment of additional muscle mass. Am J Physiol 1995;269(5 Pt 2):H1545–52. 27. Kram R, Taylor C. Energetics of running: a new perspective. Nature 1990;346:265–7.