Journal Logo

Applied Sciences: Biodynamics

Variability in energy cost and walking gait during race walking in competitive race walkers


Author Information
Medicine& Science in Sports & Exercise: September 1998 - Volume 30 - Issue 9 - p 1451-1455
  • Free


For exercise lasting more than 2 h (e.g., marathon, race walking), successful performances are dependent on various physiological, biomechanical, or psychological factors. In a previous study conducted on elite race walkers, we found a significant increase(mean 7.9%) in energy cost of walking after a 3-h walk (3). In accordance with the results of previous studies conducted on marathon runners(5) and triathletes (13,14,20), we observed a significant decrease in RER at the end of the walk, indicating the effect of substrate turnover in this energy cost increase. However, in this previous study, the decrease in RER from 0.93 to 0.83 corresponded only to a possible increase in energy cost of 1.8%. Therefore, it seems that this increase cannot be attributed to physiological factors only. It is well known that the energy cost of locomotion also reflects the biomechanical demand and a possible modification in the gait pattern may also affect the energy cost. To date, the relationships between energy cost and locomotion mechanics have not been fully clarified (e.g., 22,23). Experimental results seem inconsistent, and a wide range of individual variability is often observed (4,10,11,21,24,31,32). With regard to these studies conducted on runners, kinematic changes exist but are often individual and cannot be extended to a general alteration of running style. The existence of an individual variability is a characteristic of human movement and may be related to the number of functional degree of freedom within the system(1,2,15). Unlike running or normal walking that represent modes of freely chosen locomotion, some forms of human locomotion are strictly controlled, restraining the subject to certain movement patterns. This is the case in race walking, in which competition rules oblige the subjects to maintain a specific gait throughout the race whatever the walking speed or the subject's fatigue (e.g.,17,19). Thus, these rules prescribe a certain uniqueness to the race walkers gait adaptability (6,9,25,28,30). It is therefore possible that in this form of locomotion characterized by a more limited number of functional degree of freedom than running or normal walking, a change in metabolic energy demand could be related to a variation in walking technique.

The purpose of this study was to examine whether the physiological changes classically observed with exercise duration (i.e.; increase in energy cost of locomotion) could be related to a specific change in walking gait in a homogeneous group of race walkers.


Subjects. The subjects were nine well-trained race walkers regularly competing at national level. All subjects were experienced treadmill walkers and injury free. Physical characteristics of the subjects are given Table 1. Before participating in this study, subjects were fully informed about protocol and informed consent was obtained before all testing.

Physical characteristics and race walking performance of subjects (N= 9).

Experimental design. Each subject completed two testing sessions in the same week without any other training program. The first session was to determine V˙O2max. The second session was composed of two submaximal treadmill walks, before and after a 3-h overground walk (Fig. 1). All these submaximal tests were conducted at competition pace (12 ± 0.5 km·h−1; 76.5 ± 2% V˙O2max.). The second session was conducted 3 d after the maximal test.

Figure 1
Figure 1:
Experimental protocol.

Session 1: maximal oxygen uptake evaluation. V˙O2max was determined in a first session during an incremental protocol on a treadmill (Laufergostest) according to the criteria described by Howley et al.(18). After a treadmill accommodation at 9 km·h−1 and 3% grade, the speed of walking was increased by 1 km·h−1 every 1 min until the subject reached volitional exhaustion. V˙O2max was attained with a plateau in V˙O2, an expiratory ratio of 1.15 or greater, and a post blood lactate level above 8 mmol·L−1.

Session 2: overground walk and submaximal tests. During the second session, an overground walk was undertaken on a 400-m track very close to the laboratory (20m). Every 400 m, the walking speed (12 ± 0.5 km·h−1) of each subject was checked and regulated so as to remain stable over 3 h. Heart rate was recorded continuously (Baumann 5000), and a blood sample was withdrawn from the earlobe at the end of the walk to determine lactate concentration. Before and after this overground test, each subject completed a submaximal test. Before the first submaximal test(test 1), the subjects performed a 15-min treadmill accommodation walk at 10 km·h−1. After a 4-min rest, the first submaximal walk 6 min in length and 0% in treadmill grade was carried out at 12 ± 0.5 km·h−1. The same test(test 2) was performed immediately after the 3-h overground walk. The treadmill speed was controlled by a photoelectric cell throughout the test. During the last 2 min, exhaled gases were collected. During the two tests, a pressure sensor was used to determine the occurrence of ground contact for each foot and therefore to detect a possible switch to a running pattern. From this signal, stride duration was continuously recorded, and stride length was calculated over 20 strides the last 2 min. Subjects were filmed between the 2nd and 4th min to be free from the mouthpiece connected to the breath-by-breath device. Before and after each test, a blood sample was collected to determine lactate concentration.

Energy cost of walking calculation. Physiological solicitation was assessed by calculating the energy cost of locomotion(4,8). To calculate this parameter, expired gases were collected using a breath-by-breath system (CPX Medical Graphics). The interval between measurement of respiratory parameters was set at 10 s. Analyzers were calibrated before and after each subject's session by using gases of known concentration. The analyzer's variation never exceeded 0.02% for CO2 and 0.01% for O2. Energy cost of walking (Cw) was calculated from the average of all the V˙O2 values recorded, according to the equation (8): Cw = ((V˙O2−V˙O2 rest)·v−1)0.60; Cw is expressed in mL·kg−1·km−1, V˙O2 in mL·kg−1·min−1, and v in km·h−1. V˙O2 rest was recorded during one minute before the first test, and the last 30-s values were used. Second, lactate concentration was analyzed by an enzymatic method (Bohringer enzymatic method). Blood lactate was withdrawn from the earlobe before and immediately after exercise and 3, 5, 7, and 9 min during recovery to determine the peak lactate concentration (e.g., 12).

Motion analysis. Subjects were filmed in a sagittal plane (right side of the body) using a camera operating at a nominal rate of 50 frames·s−1. Seven markers were positioned by palpation on the head of the fifth metatarsal, the midsole of the heel underneath the calcaneus, the lateral malleolus, the lateral femoral condyle, the greater trochanter, the acromion process, and the distal radius near the wrist. The static accuracy with which a landmark could be calibrated was systematically evaluated before the test using the method described for human walking by Capozzo (7). Three complete consecutive strides were analyzed. The co-ordinated data were filtered using a low-pass Butterworth filter. Variables were chosen according to previous studies relating economy of locomotion and gait variations (31), and race walking rules (17,19). Selected variables were: stride length (SL, cm), vertical oscillation during the walking cycle (VertOsc, cm), knee angle during support phase in the vertical position of each stride (KnVe, °), mean difference of trunk angle at toe-off and trunk angle at foot strike (TAmpl, °), maximal ankle flexion at toe-off (Pflex, °), maximal vertical and horizontal excursion of the heel (HeVExc, HeHExc, cm), and of the wrist (WrVExc, WrHExc, cm).

Statistical analysis. All data is expressed as mean ± SD. Analyses of variance with repeated measure (ANOVA) were performed using energy cost and kinematic variables as dependent variables. Newman-Keuls post-hoc tests were used to compare specific means. For all the statistical analyses, the level of significance was P < 0.05.


During the first session, the mean V˙O2max (70.6 ± 4.1 mL·kg−1·min−1) was attained at a walking speed of 16.1 ± 1.2 km·h−1. At the end of the incremental test, the mean maximal heart rate and blood lactate values were, respectively, 185.2 ± 4.1 beats·min−1 and 8.6 ± 1.8 mmol·L−1.

Data for physiological responses recorded during the second session are shown Table 2. At the end of this session a significant increase in energy cost of walking and heart rate were found, whereas no variations were observed for lactate concentration or minute ventilation (P > 0.05). Respiratory ratio values calculated after the overground walk were significantly lower than values calculated before the walk(0.81 vs 0.95, P < 0.05).

Physiological parameters recorded at the beginning (test 1) and at the end (test 2) of the 3-h walk.

Kinematic data are shown Table 3. After the 3-h walk, no significant difference was recorded for this group. Throughout the testing period, respect for the rules of race walking was observed. Knee angle during support phase in the vertical position of each stride remained stable (179.1 ± 0.8° vs 179.4 ± 0.3°, P > 0.05), and no failure to maintain contact of the foot with ground was detected.

Kinematics of race walking recorded at the beginning (test 1) and at the end (test 2) of the 3-h walk.

However, we observed a systematic and particular variation of the maximal vertical and horizontal excursion of the heel (respectively, for HeHExc and HeVExc, range −12 to 9 cm and −3.5 to 4 cm) of stride length (range −7 to 5 cm) and of maximal ankle flexion at toe-off (range −6.3 to 7.8°) (statistically significant when individual variations were expressed in absolute value, P < 0.05). Furthermore, no relations between individual variation in energy cost of walking and kinematic alterations were observed.


The purpose of this study was to identify alterations in walking pattern in relation to the increase in energy cost classically observed during long distance events. On the one hand, we have confirmed the increase in energy cost at the end of a 3-h walk previously observed with elite race walkers (3). On the other hand, the nonsignificant difference in the group mean values for kinematic variables indicated no general alteration in the walking style despite fatigue. Furthermore, no significant relation was found between energy cost and kinematic data changes. However, as is classically reported, a particular and individual variation in some kinematic parameters was found (22).

For mean group values, these results indicate that, despite an increase in energy cost of walking (i.e., in physiological constraints), no general alteration of the race walking gate was found after 3 h of exercise. In race walking, technique is strictly determined by competition rules. Walkers must adhere to these or face disqualification. Our subjects were well-trained race walkers and therefore particularly used to respecting these rules. First, race walk rules require that the leading foot make contact with the ground before the rear foot leaves it, this is called "lifting." In our study, no failure to maintain contact was observed. The result was expected because the mean walking speed (12 ± 0.5 km·h−1) in the study was a habitual one for these walkers. Furthermore, previous studies have shown that "lifting" was only observed in competitive race walkers from 13.5 km·h−1(19). The second rule called "creeping" requires the support leg to be extended at the knee so that the leg attains a vertical upright position in midstance. No significant change in knee extension at this moment was detected (Table 3). The mean values recorded during test 1 and test 2 were, respectively, 179.1 ± 0.8° vs 179.4 ± 0.3°(P > 0.05). These values, situated in the range defined by Knicker and Locke (19) (175-185°), allow us to consider throughout the session, despite the effect of fatigue, subjects succeeded in respecting the rules of race walking. Furthermore, in this study no general changes in race walking kinematics were associated with the poorer walking economy at the end of the test. Previous studies have shown a similar lack of relationship between running economy and running mechanics (21,22,24,27,32). Our result confirms the individuality of the adaptation process and the difficulty to link a change in energy cost of locomotion to kinematic changes even in a strictly controlled locomotion form.

However, when one consider individual adaptations, no difference in mean values does not necessarily imply small changes for individuals. Human bipedal locomotion has developed kinetically as well as structurally to become a metabolical and mechanical system of transportation (15,16,26). Therefore, optimization principles governing locomotion are numerous and the adoption of a specific gait pattern could be seen as a function of (a) the task constraints and (b) the constraints of the performer. On the one hand, in a race walking event, the task constraints were similar for all subjects. They were characterized first by respect for race walking rules and second by the increase in the metabolic energy demand described with exercise duration. In this study, among all subjects, we observed an increase in energy cost (range 3-16%), and no technical fault was detected. On the other hand, the individual's constraints are unique, and each race walker has to cope with the task constraints using restricted number of mechanical degrees of freedom. Like running or normal walking, biomechanical analysis of race walking have shown that the ankle plantarflexion provided the major energy necessary to propel the body forward and therefore to sustain the walking speed during the race (6,25,30). However, unlike freely chosen locomotion, the stride cycle of the race walker does not have a flight phase. Therefore, each race walker attempts to progress with the optimal stride length/stride frequency ratio attainable within the limits of these mechanical degrees of freedom (9,28,30). Within this framework, at the end of the 3-h race walk, a large interindividual variability in stride length, maximal vertical, and horizontal excursion of the heel and in maximal ankle flexion at toe-off were observed (Fig. 2). This supports the assumption that changes in walking gait with exercise duration are subject dependent, leading to a substantial but unpredictable adaptation of the race walking style.

Figure 2
Figure 2:
Individual changes in energy cost of walking (Cw), stride length (SL), maximal vertical (HeVexc), and horizontal (HeHexc) excursion of the heel and in maximal ankle flexion at toe-off(Pflex) recorded between test 1 and test 2.

In conclusion, in a generalized viewpoint, results from this study demonstrate that competitive race walkers are able to maintain their walking gait with fatigue despite a systematic increase in energy cost. However, it appears that an individual and systematic change in walking kinematics may be observed within the limits of race walking adaptability.


1. Bates, B. T., S. L. James, L. R. Osternig, and J. A. Sawhill. An assessment of subject variability, subject-shoe interaction, and the evaluation of running shoes using ground reaction force data. J. Biomech. 16:181-191, 1983.
2. Bates, B. T. Single-subject methodology: an alternative approach. Med. Sci. Sports Exerc. 28:631-638, 1996.
3. Brisswalter, J., B. Fougeron, and P. Legros. Effect of three hours race walk on energy cost, cardiorespiratory parameters and stride duration in elite race walkers. Int. J. Sports Med. 17:182-186, 1996.
4.Brisswalter, J., and P. Legros. Use of energy cost and variability in stride length to assess an optimal running adaptation. Percept. Mot. Skills 80:99-104, 1995.
5. Bruckner, J. C., G. Atchou, C. Capelli, et al. The energy cost of running increases with distance covered. Eur. J. Appl. Physiol. 62:385-389, 1991.
6. Cairns, M. A., R. G. Burdett, J. C. Pisciotta, and S. R. Sheldon. A biomechanical analysis of race walking gait. Med. Sci. Sports Exerc. 18:446-453, 1986.
7. Capozzo, A. Three-dimensional analysis of human walking: experimental methods and associated artefacts. Hum. Mov. Sci. 10:589-602, 1991.
8. di Prampero, P. E. The energy cost of human locomotion on land and in water. Int. J. Sports Med. 7:55-72, 1986.
9.Douglas, B. L., and G. E. Garrett. Biomechanics of elite junior race walkers. In Sports Biomechanics. J. Terrauds, K. Barthels, E. Kreighbaum, R. Mann, and J. Crakes (Eds.). Del Mar, CA: Academic Publishers, 1984, pp. 91-96.
10. Elliot, B., and T. Ackland. Biomechanical effects of fatigue on 10,000 meter running technique. Res. Q. Exerc. Sport 52:160-166, 1981.
11. Evans, S. A., N. G. Keanmun, and S. Macdowell. The effect of fatigue on lower limb kinematics in female distance runners. Res. Q. Exerc. Sport 65:A-17, 1994.
12. Freund, H., and P. Zouloumian. Lactate after exercise in man, IV: physiological observations and model predictions.Eur. J. Appl. Physiol. 46:161-176, 1981.
13.Guezennec, C. Y., A. X. Bigard, J. M. Vallier, and A. Durey. The energy cost of running increases at the end of a triathlon. Eur. J. Appl. Physiol. 73:440-445, 1996.
14. Hausswirth, C., A. X. Bigard, M. Berthelot, M. Thomaïdis, and C. Y. Guezennec. Variability in energy cost of running at the end of a triathlon and a marathon. Int. J. Sports Med. 17:574-581, 1996.
15. Holt, K. G., J. Hamill, and R. O. Andres. Predicting the minimal energy cost of human walking. Med. Sci. Sports Exerc. 23:491-498, 1991.
16. Holt, K. G., J. Hamill, and M. M. Slavin. Running at resonance: is it a learned phenomenon? In:Proceedings of the Canadian Society for Biomechanics. Quebec: CSB Organizing Committee, 1991.
17. Hopkins, J. Biomechanics of race walking. Athl. Coach 15:22-26, 1981.
18. Howley, E. T., D. R. Basset, and H. G Welch. Criteria for maximal oxygen uptake: review and commentary. Med. Sci. Sports Exerc. 27:1292-1301, 1995.
19. Knicker, A., and M. Loch. Race walking technique and judging: the final report of the International Athletic Foundation research project, News Stud. Athletics 3:79-87, 1990.
20. Kreider, R. B., T. Boone, W. R. Thompson, S. Burke, and C. W. Cortes. Cardiovascular responses of triathlon performance. Med. Sci. Sports Exerc. 20:385-390, 1988.
21. Lake, M. J., and P. R. Cavanagh. Six weeks of training does not change running mechanics or improve running economy. Med. Sci. Sports Exerc. 28:860-869, 1996.
22. Martin, P. E., and Morgan, D. W. Biomechanical considerations for economical walking and running. Med. Sci. Sports Exerc. 24:467-474, 1992.
23. Martin, P. E., G. D. Heise, and D. W. Morgan. Interrelationships between mechanical power, energy transfers, and walking and running economy. Med. Sci. Sports Exerc. 25:508-515, 1993.
24. Morgan, D. W., P. E. Martin, F. D. Baldini, G. S. Krahenbuhl. Effects of prolonged maximal run on running economy and running mechanics. Med. Sci. Sports Exerc. 6:834-840, 1990.
25. Murray, M. P., G. Gutten, L. Mollinger, and G. Gardner. Kinematic and electromyographic patterns of olympic race walkers. Am. J. Sports Med. 11:68-74, 1983.
26. Nelson, W. Physical principles for economies of skilled movements. Biol. Cybern. 46:135-147, 1983.
27. Nigg, B. M., R. W. De Boer, and V. Fisher. A kinematic comparison of overground and treadmill running. Med. Sci. Sports Exerc. 27:98-105, 1995.
28. Payne, A. H. A comparison of ground forces in race walking with those in normal walking and running. In: Biomechanics, Vol. 6A. E. Asmussen and K. Jorgensen (Eds.). Baltimore: University Park Press, 1978, pp. 293-302.
29. Siler, W. L., and P. E. Martin. Changes in running pattern during a treadmill run to volitional exhaustion: fast versus slow runners. Int. J. Sports Biomech. 7:12-28, 1991.
    30.White, S. C., and D. Winter. Mechanical power analysis of the lower limb musculature in race walking. Int. J. Sports Biomech. 1:15-24, 1985.
    31. Williams, K. R., and P. R. Cavanagh. Relationship between distance running, running mechanics, running economy, and performance. J. Appl. Physiol. 63:1236-1245, 1987.
    32. Williams, K. R., J. E. Jones, and R. E. Snow. Mechanical and physiological adaptations to alterations in running stride length. Med. Sci. Sports Exerc. 23:S6, 1991.


    ©1998The American College of Sports Medicine