SciELO - Scientific Electronic Library Online

vol.14 issue4Energy demand and heart rate evaluation at different phases during a match along an official soccer competitionJudging criteria in international professional surfing championships author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Revista Brasileira de Cineantropometria & Desempenho Humano

On-line version ISSN 1980-0037

Rev. bras. cineantropom. desempenho hum. vol.14 no.4 Florianópolis  2012 



Relationship between running kinematic changes and time limit at vVO2max



Leonardo De Lucca; Sebastião Iberes Lopes Melo

Universidade do Estado de Santa Catarina. Centro de Ciências da Saúde e do Esporte. Laboratório de Biomecânica. Florianópolis, SC. Brasil

Corresponding author




Exhaustive running at maximal oxygen uptake velocity (vVO2max) can alter running kinematic parameters and increase energy cost along the time. The aims of the present study were to compare characteristics of ankle and knee kinematics during running at vVO2max and to verify the relationship between changes in kinematic variables and time limit (Tlim). Eleven male volunteers, recreational players of team sports, performed an incremental running test until volitional exhaustion to determine vVO2max and a constant velocity test at vVO2max. Subjects were filmed continuously from the left sagittal plane at 210 Hz for further kinematic analysis. The maximal plantar flexion during swing (p<0.01) was the only variable that increased significantly from beginning to end of the run. Increase in ankle angle at contact was the only variable related to Tlim (r=0.64; p=0.035) and explained 34% of the performance in the test. These findings suggest that the individuals under study maintained a stable running style at vVO2max and that increase in plantar flexion explained the performance in this test when it was applied in non-runners.

Key words: Kinematics; Running; vVO2max




Kinematic analysis of exhaustive running (prolonged submaximal running and short and medium-duration running at high intensity) becomes important due to possible changes in stride pattern1. This fact is related to the continuous increase in blood lactate [La] and to metabolite accumulation in muscle fibers at intensities corresponding to heavy and severe domains2, which would cause kinematic changes in the stride cycle3.

The lower limbs have the role of absorbing energy by means of eccentrically controlled dorsiflexion4 and knee flexion3 at landing. Elliot and Ackland5 highlighted the importance of the ankle joint by reporting that foot biomechanical parameters may have the major influence on running biomechanical characteristics and on the performance in a 10-km race. Fatigue could cause an imbalance in dorsiflexor and plantar flexor muscles, increasing leg impact acceleration and becoming a risk factor for injuries6. Other authors reported that ankle muscle fatigue caused decreased dorsiflexion at foot contact with the ground7 and that changes in ankle angle explained 67% of the variance in oxygen uptake (VO2) during running performed after long-duration cycling exercise8. Similarly, the knee joint would have a key role both in shock absorption3 and running energy cost9.

It is well-known that training sessions to increase VO2max should be performed at velocities close or corresponding (vVO2max) to this physiological index10. Therefore, it is interesting for sport science researchers and for trainers for some running competitions to understand the physiological and biomechanical mechanisms that may influence the time during which vVO2max can be sustained. For instance, in running competitions such as 800- and 1500-m races, the required VO2 is close to or above VO2max11; furthermore, Tlim and vVO2max were associated with performance at distances from 1500 meters to Marathon12,13.

In order to study the great variability in Tlim among individuals, Ribeiro et al.14 analyzed cardiorespiratory (VO2max, vVO2max, running economy, and ventilatory threshold) and neuromuscular (isotonic strength, velocity at maximal anaerobic running test, and vertical jump) variables and demonstrated that none of these variables could explain such variability. Gazeau et al.15 investigated biomechanical variables and, based on the results of their research, stated that runners who maintained more stable running styles were able to increase Tlim. Hayes et al.16 showed strong negative correlations between local muscle resistance of flexors and extensors of the hip and knee and the kinematic changes () in vVO2max that occur between the beginning and the end of the run, which reveals that runners with higher local muscle resistance had less changes in kinematic variables during the run.

Thus, a void in the literature is noted, because the kinematic parameters of the ankle joint were not analyzed in the few studies that aimed to investigate kinematic changes and Tlim, which focused on the hip and the knee. Assuming that kinematic changes in the entire lower extremity are caused by muscle fatigue and that they can influence running performance, the aims of the present study were to compare characteristics of ankle and knee kinematics during running at vVO2max and to investigate the relationship between kinematic changes and Tlim.




The study included eleven physically active male volunteers, recreational athletes who played team sports such as soccer, indoor soccer, and handball for 12 years, on average, and who had no previous experience in running on a treadmill. Mean and standard deviation for age, body mass, height, and body fat percentage were 23.17 (3.17) years, 71.7 (6.2) kg, 173.9 (5.5) cm, and 9.4 (2.9)% respectively. Before the beginning of the procedures, all volunteers signed a free and informed consent form providing information on the study, which was approved by the Ethics Committee for Research involving Human Beings at Universidade do Estado de Santa Catarina, protocol no. 27/2010.

Determination of vVO2max

VVO2max was determined by an incremental protocol on a treadmill, with an initial velocity of 8.3 km.h-1 (1% slope) and increments in velocity every 3 minutes (9.3km.h-1; 10.4 km.h-1; 11.6 km.h-1; 12.6 km.h-1; 13.7 km.h-1; 14.7 km.h-1; 15.7 km.h-1;16.8 km.h-1; 17.8 km.h-1) until volitional exhaustion. VO2max was considered the highest recorded value from means calculated every 15 seconds17. VVO2max was considered the minimum velocity in which VO2max was reached and sustained for at least 1 minute. If VO2max was not maintained for 1 minute, the velocity from the preceding stage was considered the vVO2max17.

Determination of Tlim

After the incremental test, new visits took place 48 hours after the first one, for running tests at 100% of vVO2max. Subjects performed a warm-up consisting of a 5-minute run on the treadmill at 60% of vVO2max17. At the end of the warm-up, they rested their feet outside the treadmill belt and were verbally informed about the test procedures. Subjects began running at lower velocities until reaching vVO2max, for a period of less than 30 seconds on average. The Tlim was considered the time interval between the moment when the treadmill reached vVO2max and the moment when the individual volitionally stopped the test. During both tests, subjects were verbally encouraged by the evaluators until volitional exhaustion.

Acquisition of kinematic data

Kinematic variables were obtained by two-dimensional filming from the moment the participant reached vVO2max until exhaustion. A filming camera (CASIO® High SpeedExlim Model EX-FH20) with acquisition frequency of 210 Hz was used at 2.30 m of distance from the treadmill and 1.0 m above the ground level. Eight joint reflective markers (5th metatarsal, lateral border of the calcaneus, lateral malleolus, lateral epicondyle of the knee, greater trochanter of the femur, greater tubercle of the humerus, lateral epicondyle of the humerus, and styloid process of the ulna) were fixed at the left side of the body directly onto the skin after shaving and alcohol cleaning.

For the purpose of comparative analysis, data were collected at two time-points: a) beginning: first foot contact with the ground 20 seconds after the subject reached vVO2max; b) final: the moment when the 5 strides of interest precede 10 seconds to the end of the test. In each of the moments, five full stride cycles were analyzed. A stride cycle was considered the interval between two successive contacts of the calcaneus of a same foot with the ground.

Studies with a similar design used cycles of 1 stride18, 3 strides16, and 5 strides7. Authors showed that there is great validity (r=0.96) between strides in relation to temporal and kinematic variables for male runners19 and that, even among not highly trained runners, the reproducibility of kinematic parameters (hip and knee angles) was high20.

To determine ankle angle in neutral position, video imaging of the subject in upright position with no treadmill slope was obtained. We considered the angle formed between the segment feet (defined from 5th metatarsal and lateral malleolus markers) and the segment leg (defined from lateral malleolus and lateral epicondyle of the femur markers). For reference, ankle angle in neutral position was considered to be 0º (angles greater than 0º indicate dorsiflexion and angles less than 0º indicates plantar flexion)7. Knee values were estimated on the basis of Knee Supplementary Angle: 180º minus knee internal angle7.

As for the filming area, a square-shaped calibrator with 4 m2 was used, manufactured with white pipes and tubes of polyvinyl chloride (PVC). The vertices of this square were marked with black insulating tape to contrast with the material from the pipes and tubes and to be used as a reference for measuring the sides of the calibrator.

The videos were exported to the Ariel Performance Analysis System (APAS) software and then the process of semi-automated digitalization began. A model for digitalizing the 8 marker points was created, followed by the formulation of a two-dimensional calibration model with X and Y coordinates. After the formulation of the model and the digitalization, the marker points were transformed by the APAS software using the DLT (Direct Linear Transformation) method and filtered with a 4th order Butterworth filter with a cutoff frequency of 6 Hz, in an attempt of eliminating any possible noise, instrument failure or digitalization errors.

Angular displacement values were calculated by the APAS software and, for analysis purposes, mean kinematic values for the 5 strides of interest from each running stage were used. The kinematic variables of interest were: ankle angle at initial contact (AAC), maximal dorsiflexion during stance (MDS), ankle angle at toe-off (AAT), maximal plantar flexion during swing (MPFS), ankle range of motion (ARM), knee range of motion (KRM) - both ranges defined as the difference between the higher and the lower angle recorded during the stride cycle -, knee angle at initial contact (KAC), maximal knee flexion angle during stance (MKFSt), maximal knee flexion angle during swing (MKFSw), knee extension angle at toe-off (KET). The kinematic changes between the beginning and the end of the running exercise were computed as the difference between initial and final values and expressed in absolute values.

Statistical Treatment

After the normal distribution of data (Shapiro-Wilk test) was confirmed, the t-Student test was applied to compare angular variables from the beginning and the end of the run. Cohen's d effect size was used for a better practical description. Effect size values (d) were classified as: 0.0 to 0.19 = trivial; 0.20 to 0.59 = small; 0.60 to 1.19 = moderate; 1.20 to 1.99 = large; 2.00 to 4.00 = very large21. The relationship between kinematic changes and Tlim was tested using the Pearson correlation test. A regression analysis (enter method) was applied to ascertain the contribution of kinematic changes to Tlim. To do so, variables that presented significance level for the correlation between kinetic changes and Tlim below 0.25 were selected, because this condition can result in variables with high relationship power22. The significance level was set at p<0.05.



Mean results for the group that performed the progressive and continuous tests are shown in table 1.



The comparison of mean values for the kinematic variables between initial (I) and final (F) running stages at Tlim is shown in table 2.



Based on the comparisons presented in table 2, it was possible to observe that maximal plantar flexion during swing was the only variable that increased significantly (p<0.01), and that none of the knee angles showed significant difference between the beginning and the end of the run. Angular values (mean for the group) for ankle and knee along the stride cycle are expressed in figure 1A and 1B respectively.

It is possible to notice a superposition of values from the initial contact to the end of the stance phase and toe-off (between 1 and 40% of the cycle) with higher plantar flexion during swing at the end of the run (between 40 and 100% of the cycle). Knee values were superposed from initial contact to near toe-off (~28% of the cycle), and also at knee extension - propulsion phase - (between 28 and 35% of the cycle), at maximal flexion during swing (between 64 and 71% of the cycle), and during the preparation for the next contact with the ground (71 to 100% of the cycle).

The results for the correlation tests between kinematic changes and Tlim are shown in table 3.



Based on the results for the correlation test presented in table 3, it is noticed that the change in ankle angle at contact was the only variable that showed a significant positive correlation with Tlim.

Table 4 presents the variables that had a prediction ability with Tlim. It can be observed that the increase in ankle angle at contact explains 34% of the performance in the test. However, when it was analyzed together with the changes in knee angle at contact and maximal knee flexion during swing, its predictive power was not significant.




The aims of present study were to compare kinematic characteristics between the beginning and the end of the run at vVO2max and to investigate the relationship between kinematic changes and Tlim. It was demonstrated that a) maximal plantar flexion during swing was the only variable to increase significantly; and b) the increase in ankle angle at contact was related to Tlim and explained 34% of the performance in the test.

The increase in maximal plantar flexion during swing corroborates the findings of Kellis and Liassou7. In fact, dorsiflexor and plantar flexor muscles affect foot position not only at contact but also during toe-off and swing4,22. Accordingly, an increase in gastrocnemius activity was observed during swing with plantar flexor fatigue7, which may explain the higher plantar flexion during swing found also in the present study. However, these findings warrant caution, because the percentage difference was small (7%), as well as that for all other study variables. Additionally, the effect size, which measures the difference between means in terms of standard deviation units (d=0.24), was also considered small. The use of this measuring tool is an attempt of replacing the concept of statistical significance with more useful notions of practical significance. Thus, if we link the results of the present study with those of other authors who consider a difference below two degrees in ankle and knee angles during the run to be insignificant8, it can be inferred that these changes are not relevant.

The ankle angle at initial contact, stance and toe-off did not change between the beginning and the end of the run. On the other hand, Christina et al.4 and Kellis and Liassou7 showed that dorsiflexion at contact decreased after a fatigue protocol of dorsiflexor and plantar flexor muscles. Increased dorsiflexion during contact reduces energy conversion from translational to rotational, because most of the energy is lost at collision with the ground. Therefore, landing with less dorsiflexion can improve storage and conversion of elastic energy23.

Based on these assumptions, one have an explanation for the fact that the change in ankle angle at contact was the only variable that showed a significant positive correlation with Tlim (r=0.64; p<0.05) and explained 34% of Tlim. Thus, it is inferred that increasing plantar flexion at foot strike with the ground can improve the performance in the test. Similarly, Bonacci et al.8 showed that changes in ankle angle at contact explained 67% of the variance in VO2 when triathletes performed a submaximal running after a 45-minute cycling exercise. In fact, plantar flexion position at initial contact can be more effective, reducing stance time, because rotational energy is transferred more effectively23, which improves running economy8. Although the difference in ankle angle at contact between the beginning and the end of the run was not significant in the present study, it was the variable that showed the highest percentage variation (7%) and the highest effect size (d=0.32).

The knee angle did not show significant differences in any phase of the cycle, which corroborates with Heyes et al.15, who analyzed sub-elite runners at vVO2max, but is opposed to several studies that investigated the effects of fatigue on knee angle at contact and maximal knee flexion during stance in submaximal running1,7,24, in which subjects increased flexion. The higher flexion at initial contact reduces the likelihood of injuries due to the lower reaction force of the ground and the better shock absorption18,24. Moreover, the stretching-shortening cycle has the role of improving the ability of producing force during the final phase (concentric action). It is inferred that the less economic running is related to a more complacent running style (less vertical stiffness), which can be represented by a higher knee flexion and a delay in the transition from stretching to subsequent shortening25. In the fatigue state, changes in the ground reaction force are associated with difficulties in maintaining angular displacements constant, and the reduction in force after the impact is probably related to a higher knee flexion26. A consequence of this process would be that, in order to keep the same performance for the stretching-shortening cycle at a given running velocity, the individual must perform a higher muscle workload during the propulsion phase, causing a higher fatigue progression26. Valiant9 estimated an increase of 25% in VO2 for every 5º of increase in maximal knee flexion during stance, which leads one to believe that this angle determines the metabolic cost associated with shock attenuation. In this case, as with knee angle at contact, the subjects of the present study may have maintained maximal knee flexion during stance as an attempt to avoid an increase in metabolic cost during the final stages, even with impaired shock attenuation

The study presented some technical limitations, such as the recruitment of non-runners individuals with no experience in running on a treadmill and the fact that motion analysis was limited to the sagittal plane. In this case, differences in running economy between treadmill and track seems to exist, due to runner's inexperience on the treadmill, which could lead to imbalances and changes in running technique and a possible variation in velocity on the treadmill from foot contact with the walking belt27.

Furthermore, the ankle moves in the three planes of motion, and it is well-known that the rotation axis of this joint is not perpendicular to the sagittal plane. Thus, it bears stressing that the two-dimensional measure can be limited in comparison with the three-dimensional measure. Foot pronation, dorsiflexion and abduction occur in the frontal, sagittal and transverse planes respectively. There is a causal relationship between hyperpronation and excessive use injuries, because pronation is necessary to attenuate impact forces28. Therefore, in situations of muscle fatigue, it would be interesting to analyze foot pronation angle as well, for a better discussion on motion in the sagittal plane and on the relationship with the performance in the test.



It can be concluded that, during running at vVO2max, subjects maintain a relative stable running style, because no difference with practical significance was observed between the beginning and the end of the run. Increasing ankle plantar flexion at contact during the test might have some beneficial effect on prolonging Tlim and could thus explain the performance in the test among non-runners subjects.



We thank CAPES for providing scholarships, which was essential for carrying out this research.



1. Mizrahi J, Verbitsky O, Isakov E, Daily D. Effect of fatigue on leg kinematics and impact acceleration in long distance running. Human MovSci 2000;19(2):139-51        [ Links ]

2. Smith, CGM, Jones AM. The relationship between critical velocity, maximal lactate steady-velocity and lactate turnpoint velocity in runners. Eur J Appl Physiol 2001;85(1-2):19-26.         [ Links ]

3. Derrick TR, Dereu D, Mclean SP. Impacts and kinematic adjustments during an exhaustive run. Med Sci Sports Exerc 2002;34(6):998-1002.         [ Links ]

4. Christina KA, White SC, Gilchrist LA. Effect of localized muscle fatigue on vertical ground reaction forces and ankle joint motion during running. Hum Mov Sci 2001;20(3):257-76.         [ Links ]

5. Elliot B, Ackland T. Biomechanical effects of fatigue on 10.000 meter running technique. Res Q Exerc Sport 1981;52(2):160-6.         [ Links ]

6. Mizrahi J, Verbitsky O, Isakov E. Shock accelerations and attenuation in downhill and level running. Clin Biomech 2000;15(1)15-20.         [ Links ]

7. Kellis E, Liassou C. The effects of selective muscle fatigue on Sagittal Lower Limb Kinematics and Muscle Activity during Level Running. J Orth Sports Phys Ther 2009;39(3):210-20.         [ Links ]

8. Bonacci J, Green D, Saunders PU, Blanch P, Franettovich M, Chapman AR, et al. Change in running kinematics after cycling are related to alterations in running economy in triathletes. J Sci Med Sport 2010;13(4):460-4.         [ Links ]

9. Valiant GA. Transmission and attenuation of heelstrike accelerations. In: Biomechanics of Distance Running, P. R. Cavanagh PR, editor. Champaign, IL: Human Kinetics;1990.p.225-247,         [ Links ]

10. Bassett DRJ, Howley ET. Limiting factors for maximum oxygen uptake and determinants of endurance performance. Med Sci Sports Exerc 2000;32(1):70-84.         [ Links ]

11. Hill DW. Energy systems contributions in middle-distance running events. J Sports Sci 1999;17(6):477-83.         [ Links ]

12. Billat VL, Hill DW, Pinoteau J, Petit B, Koralsztein J. Effect of protocol on determination of velocity at VO2max and on its time to Exhaustion. Arch Physiol Biochem 1996;104(3):313-21.         [ Links ]

13. Di Prampero PE, Atchou G, Bruckner JC, Moia C. The energetics of endurance running. Eur J Appl Physiol1986;55(1):259-66.         [ Links ]

14. Ribeiro LG, Santos TM, Lima JRP, Novaes JS. Determinantes do Tempo Limite na velocidade correspondente ao VO2máx em indivíduos fisicamente ativos.vRev Bras Cineantropom Desempenho Hum 2008;10(1):69-75.         [ Links ]

15. Gazeau F, Koralsztein JP, Billat V. Biomechanical Events in the Time to Exhaustion at Maximum Aerobic Speed. Arch Physiol Biochem 1997;105(6):583-90.         [ Links ]

16. Hayes PR, Bowen SJ, Davies EJ. The relationships between local muscular endurance and kinematic changes during a run to exhaustion at vVO2max. J Strength Cond Res 2004;18(4):898-903.         [ Links ]

17. Billat VL, Blondel AN, Berthoin AS. Determination of the velocity associated with the longest time to exhaustion at maximal oxygen uptake. Eur J Appl Physiol 1999;80(2):159-61.         [ Links ]

18. Tartaruga LAP, Coertjens M, Black GB, Tartaruga MP, Ribas LR, Kruel LFM. Efeitos da fadiga na cinemática de corredores.vRev Bras Biomec 2003;4(6):39-44.         [ Links ]

19. Morgan DW, Martin PE, Krahenbuhl GS, Baldini FD. Variability in running economy and mechanics among trained male runners. Med Sci Sports Exerc 1991;23(3):378-83.         [ Links ]

20. Hershler C, Milner M. Angle-angle diagrams in the assessment of locomotion. Am J Phys Med 1980;59(3):109-125.         [ Links ]

21. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 2009;41(1):3-13.         [ Links ]

22. Hosmer DW, Lemeshow S. Applied logistic regression. New York: Wiley, 1989.         [ Links ]

23. Lieberman DE, Venkadesan M, Werbel WA, Daoud AI, D'Andrea S, Davis IS, et al. Foot strike patterns and collision forces in habitually barefoot versus shod runners. Nature 2010;463(7280):531-535.         [ Links ]

24. Derrick TR, Dereu D, Mclean, S. P. Impacts and kinematic adjustments during an exhaustive run. Med Sci Sports Exerc 2002;34(6):998-1002.         [ Links ]

25. Morin JB, Jeannin T, Chevallier B, Belli A. Spring-mass model characteristics during sprint running: correlation with performance and fatigue-induced changes. Int J Sports Med 2006;27(2):158-65.         [ Links ]

26. Horita T, Komi PV, Nicol C, Kyrolainen H. Effect of exhausting stretch-shortening cycle exercise on the time course of mechanical behaviour in the drop jump: possible role of muscle damage. Eur J Appl Physiol 1999;79(2):160-7.         [ Links ]

27. Davies CT. Effects of wind assistance and resistance of the forward motion of a runner. J Appl Physiol 1980;48(4):702-9.         [ Links ]

28. Tartaruga LAP, Tartaruga MP, Black GL, Coertjens M, Ribas LR, Kruel LF. Comparação do ângulo da articulação subtalar durante velocidades submáximas de corrida. Acta Ortop Bras 2005;13(2):57-60.         [ Links ]



Corresponding author
Leonardo De Lucca
R. Pascoal Simone, 358
Centro de Ciências da Saúde e do Esporte
Universidade do Estado de Santa Catarina
Laboratório de Biomecânica

Received: 08 February 2011
Accepted: 12 February 2012

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License