Journal Logo

Basic Sciences: Original Investigations

Cardiovascular and metabolic costs of forward, backward, and lateral motion


Author Information
Medicine& Science in Sports & Exercise: September 1998 - Volume 30 - Issue 9 - p 1419-1423
  • Free


Considerable research has evaluated the energy cost and cardiovascular demands of forward motion, particularly walking and running (1,9,13). There has also been significant scientific inquiry related to the biomechanics and energy cost of backward running and walking (2,3,7,11,12). Recently, there has been interest in lateral movement. The research concerning lateral movement has primarily focused on the lateral motion associated with slideboard exercise training (14-16). Slideboard training requires the exerciser to maneuver laterally to the right and left with a skating-like movement over a short (1.5- to 1.8-m) distance (14-16). No research is available, however, concerning the production of continuous lateral movement in one direction and/or how this form of exercise compares with repetitive movement in other directions.

Many athletic events require the athlete to maneuver by employing a variety of movement patterns in forward, backward, and lateral directions. Football, soccer, basketball, tennis, and other racquet sports are examples of sports that commonly demand drills or movement in many different directions. In addition to training athletes, different movement patterns have been prescribed as a means for rehabilitating injuries. Specifically, backward running and walking have been recommended as forms of rehabilitation for individuals who have been diagnosed with overuse injuries resulting from forward running (7). One study reported that backward running reduced the overuse injuries associated with the patellofemoral joint and forward running (7). Lateral motion training has also been recommended by physical therapists and orthopedic surgeons to strengthen the lateral musculature of the knee (5). It has also been suggested that cardiovascular fitness can be maintained with backward running when forward running must be reduced (7).

Previous research concerning lateral motion or slideboard exercise shows that, at the same V˙O2, lateral motion exercise produced a higher heart rate than treadmill walking or running (14-16). Therefore, it has been hypothesized that other forms of lateral movement would also require a greater metabolic and cardiorespiratory demand compared with movement in the forward direction. Because of the increased interest in different directional movement patterns and the need for more scientific investigation in this area, the purpose of this study was to evaluate and compare the metabolic and cardiovascular demands imposed by forward, backward, and lateral motion at two speeds.


Subjects. Thirteen international athletes, seven male and six female collegiate tennis players, volunteered to participate. None of the subjects had prior knee or ankle injuries that might have influenced the cardiorespiratory or movement responses to various forms of locomotion. Following a detailed explanation of the investigation, informed consent was obtained from each subject as described by the American College of Sports Medicine (1).

Data collection procedures. All subjects were tested in November, which is considered the precompetitive tennis season. Their preseason training consisted of flexibility exercises, weight training, aerobic training (1-2 miles 3 d·wk−1), agility drills, and tennis matches. The agility drills required training in forward, backward, and lateral directions. The subjects were evaluated in the Human Performance Laboratory and were measured for height to the nearest 0.5 cm and weight to the nearest 0.01 kg. Percent body fat was calculated from the sum of seven skinfolds (8). The seven-site formula included the following skinfolds: chest, midaxilla, triceps, subscapula, abdomen, suprailium, and thigh.

Before testing, a pilot project was initiated to determine the subjects' ability to perform the different movement patterns on the treadmill. Each subject attempted to exercise on the treadmill performing FM, BM, and LM at three different speeds (80.45, 134.08, and 160.8 m·min−1 or 3, 5, and 6 mph). The tennis players could all perform FM, BM, and LM on the treadmill at 80.45 and 134.08 m·min−1. However, despite being highly fit and accustomed to executing BM and LM drills, approximately 50% of the subjects could not maintain the BM or LM at the 160.8 m·min−1 pace. Therefore, we concluded that for our subjects it was unsafe to attempt to evaluate them at the fastest speed. In a similar investigation Flynn et al. (7) evaluated forward and backward running and walking at 107.2 and 160.8 m·min−1(4 and 6 mph, respectively). Four of their 10 male subjects could not complete the backward running because of difficulties with coordination at speeds greater than 134.08 m·min−1. Therefore, 134.08 m·min−1 was also employed as the running speed for evaluation purposes in their investigation.

Following the familiarization trials, maximal treadmill testing was conducted on separate days(Quinton 1860, Quinton Instrument Company, Bothell, WA). The maximal treadmill test consisted of a modified Bruce Protocol. Before evaluation, each subject was fitted with a digital heart rate monitor (Polar CIC, Inc., Port Washington, NY) and allowed to rest in a seated position for 10 min to obtain the resting heart rate (RHR). During treadmill testing, inspired gas volumes were measured by a dry gas meter (Rayfield, RAM-2000, Waitsfield, VT) and continuous samples of expired air were withdrawn from a mixing chamber. An analysis of the expired oxygen and carbon dioxide was determined via a computer-based system (Apple II E, Apple Computer, Cupertino, CA; Rayfield Enhanced Software) interfaced with electronic gas analyzers (Applied Electrochemistry CD-3A and S-3A, Ametek, Inc., Pittsburgh, PA). The gas analyzers were calibrated before each test with gases previously verified by gas chromatography. Ratings of perceived exertion (RPE) were taken every 30 s and at maximal exercise effort with the Borg 20-point scale (6). The maximal treadmill test was terminated when the subjects achieved maximal exertion and attained at least two of the three following criteria: a peaking and leveling of V˙O2 (<2.0 mL·kg−1·min−1) at greater work rates, a plateau of HR within 10 beats of the age-predicted maximum, or a respiratory exchange ration(RER) exceeding 1.0.

After the maximal treadmill test and the pilot testing, the subjects reported to the laboratory on three separate days to perform FM, BM, or LM at two different speeds (80.45 m·min−1 and 134.08 m·min−1). A cross-over design was used to determine test order (FM, BM, or LM). All conditions were evaluated by open-circuit spirometry and consisted of a 5-min bout of exercise with 10 min of rest between conditions (80.45 and 134.08 m·min−1).

Videographic analysis of stride frequency and stride length. The subjects were filmed during each condition and the testing protocols were similar to those described by Flynn et al. (7). Specifically, a video camera(Panasonic model WV-F300, Secaucus, NJ) was mounted perpendicularly to the plane of motion in order to film each subject exercising on the treadmill. Stride frequencies were determined between minutes 4 and 5 of exercise using a video cassette recorder with a speed adjustable search and a color video monitor. Stride length (SL) was calculated by dividing the distance traveled by stride frequency (SF).

Statistical analysis. Because of the relatively small sample size, all statistical analyses were based on the responses of the total number of subjects (N = 13). For each of the evaluated variables, a repeated measures ANOVA was employed to detect differences between FM, BM, and LM at 80.45 and 134.08 m·min−1. When a significant F-ratio was found, a post-hoc procedure(Scheffe F-test) was employed to compare the means. A probability level of 0.05 was established as the criterion of significance. In addition, the data were applied to the ACSM equations for predicting the energy cost of forward walking and running. Specifically, the predicted and observed means associated with FM, BM, and LM at both speeds were analyzed with a repeated measures ANOVA and Scheffe comparisons(P < 0.05).


Table 1 shows the descriptive characteristics of the seven female and six male athletes. As illustrated by Table 1, the males were slightly older, taller, had a greater mass, and had higher values for V˙O2.

Descriptive characteristics of subjects (mean ± SD).

Table 2 shows the mean V˙O2, RER, HR, and RPE values for each condition at each speed. At the walking speed, V˙O2 and HR were 28 and 17% greater for BM compared with FM, respectively. When LM was compared with FM, V˙O2 and HR were 78 and 24% greater, respectively.

Comparison of metabolic and cardiovascular variables of forward, backward, and lateral motion at two different speeds (mean ± SD).

At the running speed, comparisons between BM and FM revealed that the V˙O2 and HR were both 15% greater for BM compared with FM. Similar differences were also shown for LM when comparing LM to FM (see Fig. 1). Specifically, V˙O2 and HR were 20% and 16% greater for LM compared with FM, respectively. In addition, Figure 1 shows that the metabolic cost of BM and LM (vs FM) increased more during the slower walking speed compared with the faster running speed.

Figure 1
Figure 1:
Mean V˙O2 values of forward, backward, and lateral motion.

The mean data associated with the RER and RPE showed that LM and BM tended to yield higher responses for each variable at both speeds. Specifically, LM and BM produced mean RER values at or above 0.90, and the RPE was significantly higher for LM compared with FM and BM at both speeds.

The kinematic analysis of SL and SF are presented in Table 3. Both the slow and fast speeds produced a significantly different SF and SL. These results were most dramatic between FM and LM. Moreover, the ratio between SF and SL was not only higher at 80.45 compared to 134.08 m·min−1, but the highest of any condition studied.

Comparison of stride length and frequency at constant speed during forward, backward, and lateral motion (mean± SD).

The results associated with applying the data to the ACSM prediction equations showed that there was no significant difference between the observed and predicted V˙O2 for FM at the slowest speed. However, the ACSM prediction equation significantly underestimated the observed values for BM and LM (see Table 4). At the running speed, the ACSM equation yielded values that were within approximately 1 MET for each condition.

Predicted versus observed V˙O2 (mL·kg−1·min−1) across all conditions.


This investigation was one of the first to evaluate and compare the metabolic and cardiovascular responses of FM, BM, and LM. Results of the study show that, compared to movement in the forward direction, motion in the lateral or backward direction increases both the metabolic and cardiorespiratory demand of the activity. As noted previously, many physical activities and athletic events impose movement in forward (FM), lateral (LM), and backward (BM) directions. Tennis, in particular, requires ample maneuvering in several directions and the present group of tennis players reported being highly familiar with forward, backward, and lateral movement drills. The descriptive data also shows that these athletes had a high level of cardiovascular fitness as indicated by their V˙O2max values (59.06 and 49.90 mL·kg−1·min−1 for males and females, respectively). Moreover, the male subjects won the 1995 NAIA National Tennis Championships and the women placed third.

The subjects in the present study were also less efficient when moving laterally at the slower speed. Specifically, greater variability was found in the metabolic measures of LM (compared with FM and BM) at the 80.45 m·min−1 speed. This may indicate that a greater range of economies were present for LM. Table 2 and Figure 1 show that at 134.08 m·min−1, V˙O2 and HR were similar between BM and LM. These findings may be related to the subjects' training program. During training, the present subjects customarily performed drills using BM and LM at "jogging" and "running" speeds. They did not, however, normally train at the slower walking speed. Lack of familiarity with slower movement speeds may result in a greater HR and metabolic demand (10). These training factors may also be related to the difference in the observed versus estimated values for V˙O2 (see Table 4). The present findings show that the ACSM equations for forward motion markedly underpredicted the oxygen consumption rate by 10.55 mL·kg−1·min−1 for LM at the 80.45 m·min−1 speed. In contrast, at the 134.08 m·min−1 speed, the difference between the observed and predicted V˙O2 was only −2.26 mL·kg−1·min−1. Flynn et al. (7) also found a greater difference in energy cost when comparing forward and backward movement at the slowest speed. Specifically, there was a dramatically greater percentage increase in V˙O2 and HR between BM and FM at 4 versus 6 mph.

Table 3 shows that both SL and SF changed progressively between the three directions at both speeds. Therefore, it appears that a change in the direction of walking or running will result in a change in SL and SF. This may also account for some of the large differences in the energy cost, particularly between FM and LM at the slower (80.45 m·min−1) speed. Thus, the smaller SL employed during LM resulted in a greatly increased SF(to maintain the speed of the treadmill).

It has been reported that changing a runner's chosen SL will yield an increase in oxygen uptake (4). Specifically, it has been suggested that a 20% reduction in running stride length(mean = 0.185 m) will yield a 2.6 mL·kg−1·min−1 increase V˙O2. It is difficult to directly compare these findings (which are based solely on forward running) with the present results. However, the results of the present investigation appear to indicate a similar trend: the increased SF when moving backward and laterally (compared with moving forward) appears to result in a greater oxygen utilization. A number of other studies have also found that subjects employ a greater SF and shorter SL during BM, but caution that other variables (i.e., biomechanical and neuromuscular) may also account for the difference in oxygen consumption(2,3,7). Research has shown that the neuromuscular firing patterns are distinct among locomotion in different directions and vary between the phases of these movements (11).

In conclusion, at 80.45 m·min−1, FM, BM, and LM resulted in significantly different cardiorespiratory measures. LM produced the greatest energy cost and highest HR, and these variables were also higher between BM and FM. At 134.08 m·min−1, there was no difference between LM and BM, but both conditions produced a significantly higher cardiorespiratory demand compared with FM. The present findings indicate that performing relatively high speed continuous movements in either a backward or lateral manner may significantly increase the energy cost compared with moving forward at the same speed. Training factors associated with unfamiliarity and inefficiency in performing LM at the slowest speed apparently resulted in substantially higher oxygen consumption rates compared with forward walking. At the speed of 134.08 m·min−1, there was little difference between the cardiovascular and metabolic cost of lateral and backward running. These findings support previous research comparing forward and backward walking and running and provide new information which may prove useful when evaluating the cardiorespiratory demands of different directional movements.


1. American College of Sports Medicine. Guidelines for Graded Exercise Testing, and Prescription (5th Ed). Baltimore: Williams & Wilkins, 1995, pp. 1-373.
2. Bates, B. T. and S. T. McCaw. A comparison between forward and backward walking. In: Proceedings of the North American Congress on Biomechanics: Human Locomotion. In: V. P. Allard and M. Gargon (Eds.). Montreal: Microform Publication, 1986, pp. 307-308.
3. Bates, B. T., E. Morrison, and J. Hamill. A comparison between forward and backward running. In: Proceedings of the 1984 Olympic Scientific Congress: Biomechanics, M. Adrian and H. Deutsch (Eds.). Eugene, OR: Microform Publications, 1984, pp. 127-135.
4. Cavanaugh, P. R. and K. R. Williams. The effect of stride length variation on oxygen uptake during distance running. Med. Sci. Sports Exerc. 14:30-35, 1982.
5. Bergfield, J. A. and T. E. Anderson. Achieving mobility, strength, and function of the injured knee. In: L. Y. Hunter and F. J. Funk (Eds.). Rehabilitation of the Injured Athlete. St. Louis: C. V. Mosby, 1985, p. 39.
6.Borg, G. Perceived exertion as indicator of somatic stress. Scand J. Rehab. Med. 2:92-98, 1970.
7. Flynn, T. W., S. M. Connery, M. A. Smutok, R. J. Zeballo, and I. M. Weisman. Comparison of cardiopulmonary responses to forward and backward walking and running. Med. Sci. Sports Exerc. 26:89-94, 1994.
8. Jackson, A. S. and M. L. Pollock. Practical assessment of body composition. Phys. Sportsmed. 13:76-90, 1985.
9. Montoye, H. J., T. Ayen, F. J. Nagle, and E. T. Howley. The oxygen requirement of horizontal and grade walking on a motor driven treadmill. Med. Sci. Sports Exerc. 17:640-646, 1985.
10.Schwane, J. A., B. G. Walttous, S. R. Johnson, and R. B. Armstrong. Delayed-onset muscle soreness and plasma CPK and LDH activities after downhill running. Med. Sci. Sports Exerc. 15:51-56, 1983.
11. Thorstensson, A. How is the normal locomotor program modified to produce backward walking? Exp. Brain Res. 61:664-668, 1986.
12. Vilensky, J. A., E. Gankiewicz, and G. Gehlsen. A kinematic comparison of backward and forward walking in humans. J. Hum. Mov. Studies 13:29-50, 1987.
13.Williams, K. R. Biomechanics of running. Exerc. Sports Sci. Rev. 13:389-441, 1985.
14. Williford, H. N., D. L. Blessing, M. Scharff-Olson, and J. Brown. Injury rates and physiological changes associated with lateral motion training in females. Int. J. Sports Med. 17:452-457, 1996.
15. Williford, H. N., M. R. Scharff-Olson, L. A. Richards, D. L. Blessing, and N. Wang. Determinants of the oxygen cost of slideboard exercise. J. Strength Cond. Res. 9:90-94, 1995.
16.Williford, H. N., M. R. Scharff-Olson, N. Wang, D. L. Blessing, and J. Kirkpatrick. The metabolic responses of slideboard exercise in females. J. Sports Med. Phys. Fitness 35:43-49, 1995.


©1998The American College of Sports Medicine