Santos-Rocha, Rita^1,2; Veloso, António²; Machado, Maria Lourdes²

Introduction

Step exercise has been described in a previous study (²²). Besides its cardiovascular benefits (^11,12,25,28), the structure of step exercise sessions, concerning the rate and magnitude of musculoskeletal loading, may improve the osteogenic potential of physical activity (^3,31). This activity involves a large number of loading cycles during each session (²²), but when the Step Reebok program was presented, its proponents claimed that ground reaction forces (GRFs) were similar to those of walking (²⁰). Originally introduced as a low-impact activity, step classes now include propulsive movements that have changed the nature of impact of the activity (^13,22,29). Step exercise is claimed to be a high-intensity, low- to medium-impact aerobic workout that carries a low injury risk and conditions the lower body (²⁹). Two forms of controlling the intensity of the workout are by adjusting stepping rate (125-150 bpm) and by selecting the types of movements included in choreography (e.g., propulsive movements).

A major concern is how to control the intensity of the workout, maintaining safe and effective levels of mechanical load. The GRFs of a step session depend on the type and number of movements performed (²²). Regular exposure to moderately high magnitudes of force is desirable within certain levels, because mechanical stress will induce adaptation on biological structures (²⁷). However, the same forces might produce undesirable effects such as discomfort, pain, and injury, especially when forces are too repetitive in a period of time (⁶). Clinical evidence suggests that workout intensity plays a major role in the development of overuse injuries of the musculoskeletal system (¹⁴). Several authors have indicated that step exercise seems to induce greater loads compared with walking, and at increased stepping rates its impact loading could be compared with those obtained during comfortable running and high-impact aerobics, but with a lower risk of injury (^{2,4,7,13,18,24,34}).

The assessment of biomechanical loading is quite important for exercise prescription and injury prevention in the scope of exercise biomechanics. It has been suggested that there is an optimal amount of loading that healthy individuals should maintain and that loading above a certain limit might be related to the risk of injury (¹⁸). High skeletal loading intensity has been defined as a peak GRF > 4 times body weight (BW), moderate intensity as 2-4 BW, and low intensity as GRF < 2 BW; a minimum osteogenic effect has been related to GRFs of 1-2 BW (^27,31,37). Nevertheless, the result of the loading on the body depends on the magnitude of the force, on the rate at which the force is applied, and on the repetition of load application (⁶). However, most studies have characterized movements concerning only the peak vertical GRF. Measurements of GRF reflect a general indicator of mechanical loading that allows establishment of "biomechanical intensity," the comparison between different activities, establishment of comparisons between healthy people and patients, calculation of internal forces, and calculation of the osteogenic index. Additionally, the GRF characteristics have been positively correlated with metabolic energy requirements. Nevertheless, the magnitude of GRF has been associated, although never verified, with the high incidence of lower-extremity injuries in fitness instructors (²¹). Most studies that have involved step exercise have reported the effects of peak vertical GRFs during the descending phase of the basic step (^4,14,24,26). The biological tissues are also sensitive to the loading rate or the rate of force development during the support, particularly the neuromuscular system, but, to our knowledge, no studies have been performed concerning the loading rate induced by step exercise. During gait, the vertical and horizontal forces change with cadence, and the maximum and minimum peaks increase with cadence (natural, fast, and slow walking). Generating horizontal propulsive GRFs comprises a substantial (~30%) fraction of the total energetic cost of running (^15,36). Few studies have reported the peak anterior-posterior and mediolateral GRFs of step exercise (^4,33). Also, to our knowledge, no studies have been performed concerning the average GRF and impulse during step exercise.

We hypothesized that the step patterns with propulsion would present higher loads than nonpropulsive movements, and that loading would increase with faster stepping rates. Our purposes were 1) to analyze the GRFs produced by step exercise-namely, the force-time curves, average GRF, peak GRF, impulse, and loading rate-for both the ascending and descending phases, in 18 skilled women, and 2) to investigate the differences that exist between 4 stepping-rate conditions (125, 130, 135, and 140 bpm) and between 4 step patterns (basic step, knee lift, run step, and knee hop) performed with the right and left leading legs. The step patterns studied were the most commonly used, and one should also acknowledge that even if the number of different movements in a step session were countless, they would present a limited mechanical variability that could be differentiated in walking, running, squatting, or jumping variations. One of each of these types of movements was selected. The following variables were studied: vertical average GRF (AVERAGE-FZ), anterior-posterior average GRF (AVERAGE-FX), mediolateral average GRF (AVERAGE-FY), vertical impulse (IMPULSE-FZ), anterior-posterior impulse (IMPULSE-FX), mediolateral impulse (IMPULSE-FY), total time of contact (TOTAL-TIME-OF-CONTACT), vertical first peak GRF (PEAK-FZ), vertical first peak loading rate (LOADING-RATE-FZ), time to vertical first peak (TIME-TO-PEAK-FZ), anterior-posterior first peak GRF (PEAK-FX), anterior-posterior first peak loading rate (LOADING-RATE-FX), time to anterior-posterior first peak (TIME-TO-PEAK-FX), mediolateral first peak GRF (PEAK-FY), mediolateral first peak loading rate (LOADING-RATE-FY), and time to mediolateral first peak (TIME-TO-PEAK-FY).

Methods

Experimental Approach to the Problem

One particular source of loading on the body is the GRF. Characteristics such as force profile, impact peak, loading rate, impulse, average force, and others can be readily quantified and have a functional relationship to the performance (⁸). The impact peak has been related to injury (^17,18). The AVERAGE-FZ was referred as an extremely stable indicator (¹⁵). Impulse (represented by the area enclosed between the baseline and the force curve) was referred as a variable for assessment of performance that reflects the change in velocity of the center of mass. Vertical impulse was the only GRF characteristic significantly correlated to aerobic demand (⁸). The initial part of the vertical GRF-time curve may be effectively characterized by the loading rate or rate of force development. The loading rate of a force-time signal is the time derivative of the force-time curve. This parameter indicates how fast the force changes in time and can be depicted as the slope of the force-time curve. It is often assumed that the loading rate is associated with the development of movement-related injuries, but it also is associated with the osteogenic potential of the load (¹⁷). Four step patterns were performed at varying cadences on 2 force platforms. The selected biomechanical parameters were used for input in statistical analysis.

Subjects

Eighteen step-experienced women (mean ± SD age 29.1 ± 6.8 years; body mass 58.9 ± 6.4 kg; height 1.66 ± 0.06 m; Caucasian), with no history of foot, ankle, or knee musculoskeletal/neuromuscular trauma or disease, were led through a sequence of stepping tasks, using approved choreography. These women were experienced fitness instructors with at least 3 years of teaching experience in step exercise. After being informed about the aims and procedures of the investigation, all subjects were screened for health status (¹) and gave their consent to participate in this study before performing any exercise trials. The subjects were allowed to familiarize to each speed by performing a few steps before data collection. They were allowed as many practice trials as they wished before testing. Each participant was given approximately 60-90 seconds of rest between trials to reduce the potential effects of fatigue.

The study was approved by the review committee of the faculty. The subjects performed the sequence of 8 step movements: right basic step, right knee lift, left basic step, left knee lift, right run step, right knee hop, left run step, left knee hop. This procedure was adopted to ensure mechanical balance between both lower limbs and to better represent the real conditions of practice. The motor tasks chosen are the basis of many other movements (²²). No arm movements were added. None of the subjects felt discomfort during stepping over the 2 force platforms, and none suggested that the laboratory conditions had influenced their stepping style. Our previous study showed that metal force-platforms surfaces are suitable to assess mechanical loads of stepping among experienced subjects (²³). For each condition of stepping rate, one successful sequence was collected. Fitness music was used to maintain cadence. All experimental trials were conducted in a “crescent cadence” order. These procedures were adopted so that the result would reflect typical class conditions. Verbal instruction was provided during the tests. To reduce error, participants wore similar Reebok shoes, because the type of footwear can alter foot mechanics, influencing peak pressure, braking, and propulsive forces (^9,16).

Procedures

The movements were performed on the AMTI force platform (Advanced Mechanical Technology, Inc, Watertown, Mass) of 0.90 × 0.60 × 0.17 m (length/wide/height) for stepping up (substituting the step bench) and on the KISTLER force platform (Kistler AG, Winterthur, Switzerland) of 0.60 × 0.40 m (length, wide) at ground level for stepping down. After calibrating the force platforms, GRFs were measured at 1000 Hz, and the software Acqknowledge 3.7.3 (BIOPAC Systems, Inc., Goleta, Calif) was used to collect and process data. Data were smoothed with a Hamming low-pass digital filter. The optimal cutoff frequencies of 8 Hz were determined by the residual error method proposed by Winter (³⁵). The 3 components of GRFs were obtained. The force profiles for each recording were analyzed and processed using Acqknowledge software. In each condition of stepping rate, the trajectories of the components of the GRFs of all subjects were normalized in time, and the average curve was calculated and presented graphically. The average GRFs (N), impulses (N·s), times to peak, and TOTAL-TIME-OF-CONTACT (seconds) were calculated. The impulse during the ascending phase was determined by integrating the GRF-time curve from the instant the foot touched the AMTI platform to zero (impulse = ∫ force [N]·TOTAL-TIME-OF-CONTACT) and, during the descending phase, from the instant the foot touched the Kistler platform to the instant of peak double support. The average GRFs and impulses were normalized to body weight (BW). Loading rate (N·s⁻¹) was calculated (loading rate = peak force [N]/time to peak [seconds]) and normalized to BW per second (loading rate = peak force [BW]/time to peak [seconds]). Force was normalized to BW. The GRFs were measured in 3 directions-vertical (FZ), anterior-posterior (FX), and mediolateral (FY)-during a sequence of step movements performed at varying cadences. Figure 1 represents the identification of the movements studied during the ascending and descending phases.

Figure 1

Statistical Analyses

The 4 stepping-rate conditions and the 4 step patterns were used as dependent variables. The following independent variables were analyzed statistically:

• AVERAGE-FZ, AVERAGE-FX, AVERAGE-FY, PEAK-FZ, PEAK-FX, and PEAK-FY (force parameters);
• IMPULSE-FZ, IMPULSE-FX, IMPULSE-FY, LOADING-RATE-FZ, LOADING-RATE-FX, and LOADING-RATE-FY (combined parameters); and
• TOTAL-TIME-OF-CONTACT, TIME-TO-PEAK-FZ, TIME-TO-PEAK-FX, and TIME-TO-PEAK-FY (temporal parameters).

All statistical procedures were conducted using SPSS 14.0 for Windows (SPSS Inc., Chicago, Ill). All results are reported as mean, SD, maximum, minimum, range, and coefficient of variation (CV). Kolmogorov-Smirnov normality test and Mauchly's test of sphericity were conducted. A 1-way analysis of variance for repeated measures with 2 within-subjects factors (ANOVA-RM) was used to determine whether there were significant differences in parameters between conditions of stepping rate and between step patterns. In the cases in which sphericity was not assumed, the Huynh-Feldt correction was used. The pairwise comparisons with the Bonferroni confidence interval adjustments were used to identify where differences could be found. In all cases, the level of statistical significance was set at p ≤ 0.050 (³²).

Results

Figures 2, 3, 4, and 5 show the average GRF (N) curves during the ascending and descending phases of basic step, knee lift, run step, and knee hop, respectively, performed at 125, 130, 135, and 140 bpm.

Figure 2

Figure 3

Figure 4

Figure 5

Considering the profile of GRF curves, the vertical component of GRF dominates the impact force-time history in comparison with the other 2 components. The magnitude of vertical GRF increased with increasing speed. Similar results were obtained in running, slow jogging, and walking (¹⁰). During the ascending phase, the vertical GRF curves for the basic step and knee lift exhibited a triple peak, whereas the run step and knee hop exhibited a double peak. In the basic step, the typical time-history GRF curve demonstrated 3 distinctive maximums (triple hump) during the ascending phase, with first peak related to foot contact on the bench, the second peak related to weight transfer between both feet, and the third peak related to the impulsion for the descending phase. In the knee lift, the typical time-history curve also demonstrated 3 distinctive maximums (triple hump) during the ascending phase, with the first peak (the greater) related to foot contact on the bench, the second peak related to knee lift, and the third peak related to the impulsion for the descending phase. In the run step and knee hop, the curve demonstrated 2 distinctive maximums (double hump) during the ascending phase, with the first peak related to foot contact on the bench with propulsion, and the second peak related to the second propulsion and to the impulsion for the descending phase. During the descending phase, all movements exhibited one peak (single hump) related to the first foot contact on the floor; a second peak existed if weight transfer occurred between both feet, and a third peak corresponded to the beginning of another movement, if the movement was followed by another. These patterns did not change at varying cadences. The results show that during stepping at different cadences, the vertical GRF curves were very regular and repetitive among subjects, despite different interval times among conditions. We observed the absence of impact peaks in the movements analyzed.

The action of stepping up from the ground to the bench was expected to produce greater force in the anterior direction. The typical anterior curves of the 4 patterns were similar to the corresponding vertical curves, but with lower magnitude. Thus, the first peak is related to foot contact on the bench, the second peak is related to weight transfer between both feet, and the third peak is related to the impulsion for the descending phase, in the basic step and knee lift. In the run step and knee hop, the typical time-history GRF curve also demonstrated a double hump during the ascending phase, with the first peak related to foot contact on the bench with propulsion, and the second peak related to the second propulsion and to the impulsion for the descending phase. The action of stepping back from the bench to the ground was expected to produce greater force in the posterior direction. This might cause eccentric loading in Achilles tendon. During the descending phase, all movements exhibited one posterior peak (single hump) corresponding to the contact with the ground.

The mediolateral curves for the basic step and the run step during the ascending phase were of lower magnitude than the vertical GRF, and these curves represent right foot contact (positive hump) followed by the left foot (negative hump). The mediolateral curves for the knee lift and knee hop during the ascending phase were also of lower magnitude than the vertical. These curves exhibited 2 positive humps that represent right foot contact during the knee lift or hop, followed by the propulsion of the right foot for stepping down. During the descending phase, the 4 movements exhibited one positive or negative peak, depending on the first foot to contact the ground: the right foot in the basic step and run step, and the left foot in the knee lift and knee hop. The patterns of the horizontal curves also did not change at varying cadences, showing that, during stepping at different cadences, these parameters were very regular among subjects. However, although smaller in magnitude, the GRF horizontal components applied to the lower extremity during the loading phases may also influence biomechanical loading. These results indicate that lower-extremity external loading can be effectively controlled by varying the stepping rate during step classes.

Table 1 shows the results of descriptive statistics of the AVERAGE-FZ (BW). Table 2 shows the results of descriptive statistics of the IMPULSE-FZ (BW·s⁻¹). Table 3 shows the results of descriptive statistics of the vertical peak GRF during the ascending and descending phases of 4 selected step patterns performed at varying cadences. Table 4 shows the results of descriptive statistics of the anterior-posterior peak GRF. Tables 5 and 6 show the results of descriptive statistics of the LOADING-RATE-FZ and the LOADING-RATE-FX, respectively, during the ascending and descending phases of 4 selected step patterns performed at varying cadences.

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Acceptable variability was considered a CV < 10% (⁵). The CVs of the force, temporal, and combined parameters were acceptable and showed that these kinetic variables were more variable at faster conditions, indicating that changes occur as a function of increased stepping velocity. Impulse presented greater variability. The CVs showed that the mediolateral parameters were more variable than the other components. The mediolateral and anterior-posterior parameters presented unacceptable variation. The CVs of the force and combined parameters were more variable in the descending phases of the movements.

Tables 7-10 show the results of ANOVA-RM and Bonferroni pairwise comparisons of the parameters analyzed.

Table 7

Table 8

Table 9

Table 10

The test of within-subject effects has shown no interaction between step pattern and stepping rate in relation to the average GRF, impulse, and temporal variables. The test of within-subjects effects has shown no interaction between step pattern and stepping rate in temporal variables, PEAK-FY, LOADING-RATE-FY, PEAK-FX (descending phase), and LOADING-RATE-FZ (descending phase). There was interaction between conditions in relation to PEAK-FZ (ascending phase, p = 0.001; descending phase, p = 0.011), LOADING-RATE-FZ (ascending phase, p = 0.002), PEAK-FX (ascending phase, p = 0.000), and LOADING-RATE-FX (ascending phase, p = 0.002; descending phase, p = 0.048).

Discussion

Step exercise is performed using music that sets a movement cadence, which involves the repetition of exercises that induce peak GRFs of low magnitude (1-2.5 BW) but of high frequency (3750-4050 loading cycles during a 30-minute session, using a music speed of 125-135 bpm) (22). The FZ component dominates the force-time history in comparison with the horizontal components. However, these components also contribute to the magnitude of loading. Our results show that during stepping at different cadences, the GRF curves were very regular and repetitive among subjects, despite different interval times among conditions. Further research is required to see how these 3 components can be modified with specific training. The profiles of the GRFs are similar among propulsion and among nonpropulsion movements, especially during the ascending phase; however, it seems to be relatively stable and immune to stepping rate.

The AVERAGE-FZ ranged from 0.7 to 0.8 BW and decreased as stepping rate increased. Greater values were registered in the run and descending phases. These results were lower than the 1.4 BW obtained for running speed at 3 m·s⁻¹ (¹⁵). To our knowledge, no other studies were performed concerning the magnitude of the AVERAGE-FZ in step exercise. The horizontal components exhibited a characteristic shape that was similar among movements with propulsion and among the nonpropulsion movements, differing in magnitude according to the stepping-rate conditions. The AVERAGE-FX values were about 0.2 BW in the ascending phase and −0.2 BW in the descending phase. The AVERAGE-FY values ranged from −0.002 to 0.005 BW in the ascending phase and from −0.006 to 0.003 BW in the descending phase.

The TOTAL-TIME-OF-CONTACT significantly decreased as stepping rate increased. The TOTAL-TIME-OF-CONTACT ranged from 1.14 to 1.24 seconds in the ascending phase and from 0.48 to 0.67 seconds in the descending phase.

Impulse increased as stepping rate decreased. The IMPULSE-FZ ranged from 0.8 to 1 BW·s⁻¹, and it decreased significantly among conditions of stepping rate. In the descending phase, the impulse was lower than in the ascending phase, ranging from 0.4 to 0.6 BW·s⁻¹. Significant differences were found among step patterns, except for the basic/run and knee lift/knee hop. The IMPULSE-FX was about 0.2 and −0.1 BW·s⁻¹ in the ascending and descending phases, respectively. The CV values show that the mediolateral parameters of GRFs were more variable than the other components. The IMPULSE-FY values were about 0.001 to 0.006 and −0.005 to 0.002 BW·s⁻¹ in the ascending and descending phases, respectively. To our knowledge, no studies were performed concerning the magnitude of the impulse in step exercise, except that of Maybury and Waterfield (¹⁴), who have reported (in absolute values) a range of 0.5-0.9 BW·s⁻¹ and mean values of 0.7 ± 0.1 BW·s⁻¹ during the descending phase of the basic step at 120 bpm. Our results reflect that experienced participants become more economical as the stepping rate increases, such as in running (¹⁵).

The results indicate that lower-extremity external loading can be effectively controlled by varying the stepping rate during step classes and by choosing movements mechanically similar to those analyzed in the present study. As an example, the run step clearly induced greater forces, which might be more related to injury.

Table 7 shows the results of the statistical analysis performed with average GRF, impulse, and total-time-of-contact parameters, as well as the summary of the confirmation of the hypothesis that differences exist among stepping-rate conditions and among step patterns.

The GRF may provide a surrogate measure for the strain experienced by bone on a variety of loading activities such as step movements. But most studies have focused only the vertical component because it is of much greater magnitude than the other components, and because the major interest in landings has been the effect of vertical loads or impacts on the human body.

In the basic step, the PEAK-FZ had a mean value of up to 1.4 BW and a maximum of 1.5 BW for the ascending phase, and a mean value of up to 1.7 BW and maximum of 2.2 BW for the descending phase. These results for the descending phase were greater than those reported by other authors who used slower cadences (120 bpm) (^2,4,14). The results for the descending phase are in line with those obtained by Teriet and Finch (³⁰) and with those obtained in our previous studies (²⁴). In the knee lift, PEAK-FZ had a mean of up to 1.3 BW and maximum of 1.6 BW for the ascending phase and a mean of up to 1.8 BW and maximum of 2.3 BW for the descending phase. The results for the descending phase were greater than those reported by Farrington and Dyson (⁴), who used slower cadences (120 bpm). The results in both phases are in line with those obtained by Panda (¹⁹). In the run step, the PEAK-FZ had a mean value of up to 2.3 BW and a maximum of 2.7 BW for the ascending phase, and a mean of up to 1.8 BW and a maximum value of 2.4 BW for the descending phase. Tagen and Zebas (²⁹) have reported 2.5 BW during the ascending phase of running at 126 bpm. In the knee hop, the PEAK-FZ had a mean of up to 1.8 BW and a maximum of 2.2 BW for the ascending phase, and a mean of up to 1.6 BW and maximum of 2.2 BW for the descending phase. The results for the ascending phase are in line with those reported by Machado and Abrantes (¹³), but the authors also used slower cadences (120 bpm). The results for the ascending and descending phases of all movements performed at 130 and 140 bpm were about 0.1 to 0.2 times smaller than those obtained in our previous studies using pressure insoles (²³). In walking, the vertical component had a maximum value of 1-1.2 BW, and, in running, the maximum value can achieve 3-5 BW (¹⁵). Therefore, step exercise seems to produce greater loads compared with impact loading of walking, and at increased stepping rates its impact loading could be compared with the loads obtained during comfortable running.

The results obtained for peak vertical forces suggest that step exercise is a low to moderate activity, depending on the inclusion of nonpropulsion or propulsion, but not on stepping rate (with experienced participants). However, the results of the present study are referred to the superimposed GRF, which profile represents a summation of bilateral force profiles during the double support phase of the movements. Teriet and Finch (³⁰) have suggested that the faster loading and unloading rates of the musculature attributable to the faster stepping rates (122-130 bpm) caused less control of the movement, resulting in a 4% increase in the PEAK-FZ; therefore, the use of faster tempos in a beginning-level class could be a source of elevated risk for potential injury. Our results support the conclusion of Scharff-Olson et al. (²⁶) that experience with step exercise may improve one's ability to make uniform and force-absorbing adjustments in PEAK-FZ at increased speeds.

The anterior-posterior component exhibits a characteristic shape similar to the vertical, but of lower magnitude. The PEAK-FX reaches 0.15 BW in walking, and 0.25-0.30 BW in running at 3 m·s⁻¹, for both braking and propulsion (¹⁵). Concerning step exercise, Farrington and Dyson (⁴) found PEAK-FX values of 0.06-0.30 BW in the basic step (120 bpm/15.2 cm/descending phase). Panda (¹⁹) has reported 0.06 BW for the knee lift, 0.07 BW for the basic step, and 0.10 BW for the run (126 bpm/15 cm) in the ascending phase and 0.29 BW for the knee lift, 0.26 BW for the basic step, and 0.38 BW for the run (126 bpm/15 cm) in the descending phase. Wieczorek et al. (³³) have reported 0.10-0.13 BW for the basic step (120 bpm/30 cm) in the descending phase and 0.12-0.15 BW (120 bpm/20 cm) in the ascending phase. In the present study, the mean PEAK-FX ranged from 0.3 to 0.6 BW (propulsive) in the ascending phase and from −0.4 to −0.3 BW (braking) in the descending phase. These results are in line with those obtained by Panda (¹⁹) and Miller (¹⁵) for running at 3 m·s⁻¹. No further studies were found concerning the magnitude of PEAK-FX with similar conditions tested.

The mediolateral component also exhibited a characteristic shape similar in the 2 movements with propulsion and in the 2 nonpropulsion movements, differing in magnitude according to the stepping-rate conditions. The PEAK-FY values ranged from 0.01 BW in walking to 0.10-0.20 BW in running (^6,15). Concerning step exercise, Wieczorek et al. (³³) have reported PEAK-FY values of 0.2-0.3 BW (132 bpm/20 cm/descending phase). No further studies were found concerning the magnitude of PEAK-FY with similar conditions tested. In the present study, the mean PEAK-FY ranged from −0.02 to 0.02 BW in the ascending phase and from −0.03 to 0.01 BW in the descending phase. In both phases, there were no significant differences among stepping rates and among step patterns. These results are in line with those obtained for walking by Wieczorek et al. (³³) and by Hamill and Caldwell (⁶).

The time to PEAK-FZ ranged from 0.20 to 0.28 seconds for the ascending phase and from 0.21 to 0.22 seconds for the descending phase. The time to PEAK-FX ranged from 0.18 to 0.26 seconds for the ascending phase and from 0.20 to 0.22 seconds for the descending phase, and the time to PEAK-FY ranged from 0.18 to 0.24 seconds for the ascending phase and from 0.18 to 0.21 seconds for the descending phase. The interval time decreased with stepping rate, meaning that the same movement has to be performed in the same form but with less time. This is reflected by the increase in loading rate. No previous studies were found concerning the behavior of loading rate in step exercise. Loading rate was positively related to running speed at 3 m·s⁻¹, associated with an average of 77 BW·s⁻¹ (¹⁵). In the present study, the mean LOADING-RATE-FZ in the ascending phase ranged from 4.9 BW·s⁻¹ in the knee lift (125 bpm) to 10.2 BW·s⁻¹ in the run (140 bpm). It increased with stepping rate, and the greatest value was found in the ascending phase of the run. During the descending phase, it ranged from 7.4 BW·s⁻¹ in the knee hop (125 bpm) to 8.5 BW·s⁻¹ in the run (140 bpm). It increased significantly with stepping rate. The mean LOADING-RATE-FX in the ascending phase ranged from 1.7 to 3 BW·s⁻¹, and in the descending phase it ranged from −1.96 to −1.5 BW·s⁻¹. The mean LOADING-RATE-FY in the ascending phase ranged from −0.08 to 0.16 BW·s⁻¹, and in the descending phase it ranged from −0.15 to 0.05 BW·s⁻¹. In both phases there were no significant differences among stepping-rate conditions or among step patterns. The larger peaks and loading rates indicate a loss of shock-absorbing capacity. This may increase their susceptibility to lower-extremity overuse injuries.

Tables 8-10 show the results of the statistical analyses performed with vertical, anterior-posterior, and mediolateral parameters, respectively, as well as the summary of the confirmation of the hypothesis that differences exist among stepping-rate conditions and among step patterns. The results provide an answer to the proposed hypothesis.

The results indicate that lower-extremity external loading can be effectively controlled by varying the stepping rate during step classes, and by choosing movements mechanically similar to those analyzed in the present study. As an example, the run step clearly induced greater forces and loading rate, which might be more related to injury.

These findings indicate the relative contributions of stepping rate and different choreographic movements to the external forces experienced during step exercise. Further research is needed focusing other step patterns to select those that are more appropriate to be included in exercise and rehabilitation programs. The results of our previous studies with older people have suggested that these programs might have a positive impact on the aging delay in spite of the improvements in well-being and quality of life, and they might reduce the risk of making older people prone to falling or to other accident-related injuries (²⁴). The present investigation provides data that may be used as the basis of comparison with patients, elderly people, and beginners who participate in step programs, in future biomechanical research. However, the present results are based on a sample of 18 experienced and physically active instructors; the force characteristics of the tasks might be different if participants with less experience in step were used, and establishing norms for other populations will require understanding other factors that affect GRFs. Further research is required to select other step patterns that are appropriate for inclusion in exercise and/or rehabilitation programs. Also, these results are related to the mechanical characteristics of this physical activity, and they might be analyzed under the ergonomic perspective, because the group of subjects comprised experienced step instructors.

Practical Applications

This study presents the analysis of the GRF parameters during step exercise at varying cadences. This study investigated the external loading experienced by the lower extremity during 4 common step movements performed at various cadences of stepping rate. The results contribute to our understanding of how skilled participants deal with the increase of the external load. Instructors must be aware of participants' fitness and levels of expertise when choosing appropriate stepping rates and movements during step classes. Understanding the biomechanics of the lower limb during step exercise is very important for instructors to prescribe exercise correctly and for therapists to design rehabilitation programs. Our results indicate that lower-extremity external loading can be effectively controlled by varying the stepping rate during step classes and by selecting step patterns. Assuming that walking or running are “safe” activities to be included in exercise and rehabilitation programs, oriented stepping exercise seems relatively safe with respect to the magnitude of loading. Despite being created as an exercise program, the origins of this physical activity are in a bench used in rehabilitation programs. For this reason, it is acceptable that the actual form of performing some of the step patterns might be included in rehabilitation programs. However, our search found few studies that have used step exercise as a form of rehabilitation exercise. Our results are relevant to determine which patterns and cadences can be recommended for inclusion in rehabilitation programs where walking and running are prescribed. The knowledge about the magnitude of loading helps in selecting the proper stepping rates and movements for inclusion in classes.

Acknowledgments

The authors wish to thank all participants of this study; Carlos Ferreira, PhD (Faculty of Human Kinetics); Helô Isa André, MSc (Faculty of Human Kinetics); Maria Fátima Ramalho, MSc (Sport Sciences School of Rio Maior); and Maria João Valamatos, MSc (Faculty of Human Kinetics) for their contributions to this manuscript; and Pedro Aguiar, MSc (National School of Public Health) and Isabel Carita, PhD (Faculty of Human Kinetics) for their advice in statistical procedures.

References