BrJ Sports Med 1997;31:109-1 13

An investigation into the relation between stepheight and ground reaction forces in step exercise:a pilot study

Mark C Maybury, Jackie Waterfield

AbstractThe aim of this study was to investigatethe effect that changing step height had onground reaction force. Using a ran-domised crossover design, 12 volunteerswith no previous experience of step aero-bics were recruited to perform at threedifferent step heights: 6, 8, and 10 inches.Subjects performed a basic step at acadence of 120 beats/min and performedthree one minute trials during whichground reaction force was measured.Measurement of peak impact force, timeto achieve peak impact, and total time offoot contact was made, and impulse of theforce was calculated. Statistically signifi-cant differences were found to exist forpeak impact force between the 6 and 8inch and 6 and 10 inch, but not betweenthe 8 and 10 inch conditions. No signifi-cant differences were found in any otherparameters. The study supports thepresent advice that participants shoulduse low step heights, and possible mecha-nisms of injury are discussed.(BrJt Sports Med 1997;31:109-1 13)

Keywords: step exercise; ground reaction forces;moments; eccentric contraction

Department ofPhysiotherapy, GoodHope Hospital, SuttonColdfield, WestMidlands B75 7RR,United KingdomM C Maybury

School ofHealth andSocial Sciences,Coventry University,Coventry, UnitedKingdomJ Waterfield

Correspondence to:MrM C Maybury.

Accepted for publication26 February 1997

This study investigates the relation betweenstep heights and ground reaction forces. Thehypothesis to be explored was that differencesexist between three step height conditions-6,8, and 10 inches-in terms of peak verticalimpact, impulse of the force, time to peakimpact, and time of total foot contact. Litera-ture relating to step exercise is reviewed, andthe mechanisms of injury that are possiblyassociated with this type of activity arediscussed.

Step aerobics is claimed to be a highintensity low impact aerobic workout, carryinga low injury risk, which conditions the lowerbody and can also condition the upper body.The intensity of the workout can be manipu-lated by adjusting step height and/or by the useof hand weights. Proponents of step aerobicsclaim that ground reaction forces (GRFs) aresimilar to those of walking.' In a UK televisionconsumers programme, step aerobics cameunder scrutiny because of these claims. Safetyfears were expressed for participants because ofclaims that the research cited to authenticatestep aerobics was inherently flawed. Injury

potential arising from incorrect stepping tech-niques was also highlighted.There are few reported studies on the effects

of step aerobics on GRF. Johnson et arcompared 40 minutes of bench stepping on an8 inch step with "normal" walking, slowjogging, low impact marching, and high impactdouble hop knee lifts, and found that benchstepping had lower recorded GRFs (1.46corrected for body weight (BW), 1.1 3BW,2.26BW, 1.74BW and 3.14BW) than all otherexercises, excluding walking. Farrington andDyson4 measured forces in three planes overfour step heights: 4, 6, 8, and 10 inches. Theyfound that the vertical component of the GRFranged from 1.453 to 1.863BW for the basicstep; similar values were reported for other steppatterns. They concluded that step aerobicsmay not be as low impact an exercise as origi-nally thought, and they suggested that to mini-mise injury risk an optimal step height relatedto aerobic benefit could be used in combina-tion with manipulation of the routine by theaddition of hand weights and by using morecomplex step patterns.

Nisell et ar quantified ankle and kneemoments, femoral shear, and compressiveforces that occurred during four different stepdown activities from a 20 cm height. Theirresults indicated that step downs with the ballof the foot gave softer vertical GRFs, with lowerforces recorded for backward ball steps than forforward ball steps. They went on to suggestthat in forward jumps and steps the momentarms are almost equal for the vertical force (Fz)and the anteroposterior force (Fx) at the anklejoint. However, at the knee the Fx moment armis relatively larger than the Fz moment arm,thus exerting a greater influence over knee loadmoment. Unfortunately they make no mentionof forces related to backward steps.

Mital et ar studied several riser and treadcombinations to determine the values that gaveminimal hip, knee, and ankle joint moments.They found that riser height had a profoundeffect on all moments, except hip moment. Asthe riser height increased, so did both knee andankle joint force. Andriacchi et afi studied bothascending and descending stairs, and reportedthat descending brought about the largestmoments, larger than walking. However, themechanics of walking are different from thoseof ascending or descending stairs. In addition,phasic muscle activity is different, with in-creased activity in muscles responsible for ver-tical activity, and knee extensor force generatedin stair climbing is higher than in level walking.

109

on June 30, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.31.2.109 on 1 June 1997. Dow

nloaded from

http://bjsm.bmj.com/

Maybury, Waterfield

Andriacchi et af suggest that if joint force isassumed to be proportional to external mo-ment at the joint, then the magnitude of theknee joint force generated while descendingstairs could be more than six times body weightat 500 knee flexion. In walking, this amount offorce tends to occur near full extension.

Nisell and Mizrahi8 studied the effects of dif-ferent step down activities from step heights of0.2 and 0.43 m (8 and 17 inches respectively).They reported that during step down activitiesthe lowering leg is subject to more force thanthe impact leg, thus acting to decelerate andsoften the impact of the lowered leg. When cal-culated, the compressive joint load whendescending stairs was found to be four timesbody weight.

THE MUSCULOSKELETAL SYSTEMA major role played by the musculoskeletalsystem is energy dissipation. During day to dayactivities joints and bones are subjected to highimpulsive loads, leading to high peak dynamicloads. The idea that impulsive loads areinvolved in joint degeneration was proposed byRadin et al.9 Joints are subjected to two types offorce: shear force caused by articulation andcompressive force produced by longitudinalloading. Further work by Radin and others"0-2suggests that degeneration of joints could be aconsequence of repetitive loading, the natureand degree being more important than thetotal force. Gymnasts successfully attenuatelanding forces by using large degrees of jointflexion," although the heights used in gymnas-tics are greater than those used in step aerobics.The body also relies on active and passivemechanisms.'4 '5 Active attenuation, as in toestrike running, is achieved by proprioceptiveactivation of muscle tone to soften impactforces; this is also thought to lead to muscleand bone strengthening.'6 Footwear also im-proves the shock attenuating ability of thebody.'7 Pratt" suggests that in modern runningshoes forefoot cushioning can decrease by 20%after 350 miles of wear, even though rearfootcushioning may be unaffected. He suggests thatpoor or worn shoes may give rise to patellofemoral joint pain, peritendinitis of the tendoAchillis, and tibialis posterior tendinitis. In stepaerobics, the ball of the foot is placed downfirst, followed by the heel. A loss of forefootcushioning'4 would lead to reduced shockattenuation by the shoe material and increasedshock attenuation by the body's active mecha-nisms, such as the intrinsic muscles of the footand plantar flexors.

Stepping activities rely heavily on eccentricmuscle work. Pain and stiffness are signs ofmuscle damage according to Strauber,'9 andoccur at different times after eccentric muscleaction as a result of shortening of thenon-contractile elements of the muscle.20 Mus-cle soreness can be classified into immediateand delayed, with maximum muscle traumabeing caused by long duration high torqueforce producing exercises.2' This may be exac-erbated by leg length inequality (anisomelia),which is often cited by some authors as a pos-sible factor that may predispose individuals to

an increased risk of injury. With injuries such aslow back pain, stress fractures, and osteoarthri-tis, compensatory mechanisms and increasinginternal joint forces can be introduced andthese can be aggravated by repetitive impulseloading.22 Obviously, leg length inequalitycould have a bearing on injury incidenceduring step exercise, individuals with large dis-crepancies being more at risk.

MethodsTo investigate the hypothesis, a randomisedcrossover design was implemented, using aKistler force plate to measure GRFs. After sat-isfactory completion of a health screeningquestionnaire, a convenient self-selected sam-ple of 12 female subjects were randomlyassigned to three exercise sessions to performat all the three heights under investigation. Thethree sessions were required to avoid warm updecrement and because of equipment and sub-ject availability. The order in which the stepheights were used by each subject was also ran-domised. Random assignment to the exercisesessions and to the order of step heights withineach session was achieved by means of the ran-dom drawing of names.23 Novice steppers wereused, experienced steppers being excludedfrom the study.

Variables thought to influence GRF wereaddressed in the following way.* The stepping frequency was set at 120

beats/minute; lower stepping frequenciesmay have decreased, while higher frequen-cies may have increased, GRFs.

* No arm movements were taught, althoughthey are often used in step classes to increasethe work load, as these may have had aneffect on GRF.

* Fatigue was prevented by the short durationof the exercise session; fatigue would havealtered GRF and this factor was not underinvestigation because of limitations of timeand equipment.

* Footwear type was limited to training shoes.* Practice effects were controlled by randomi-

sation of the order of step heights for eachsubject.

A Reebok step was positioned adjacent to aKistler force plate, type 9281B, sunk into asprung walkway. This was to allow the subjectto step one foot centrally on the force platformand the other on to the walkway (fig 1). Theforce plate was connected to a Kistler forceplate control unit, and type 9851 data werecollected by a BBC Master microcomputerwith customised software. The ranges were seton the control unit at 2 kN for vertical force(Fz) and at 500 N for the horizontal force (Fx)and anterior posterior force (Fy).

Before testing, heights and weights weremeasured (table 1), and subjects participatedin a standardised warm up and stretching pro-gramme lasting 15 minutes." Only one subjecthad a leg length discrepancy of 1 cm; this wasmeasured from the inferior medial malleolus tothe anterior superior iliac spine as advocated byMcCaw and Bates.22A two minute habituation period preceded

the trials which consisted of three one minute

110

on June 30, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.31.2.109 on 1 June 1997. Dow

nloaded from

http://bjsm.bmj.com/

111Relation between step height and ground reaction forces

Figure 1 Diagrammatic representation of the position ofthe subject, step, and Kistlerforce plate.

Table 1 Subject characteristics

Age (years) Height (cm) Weight (kg)

Mean 19 167.4 61.6Minimum 17 158.5 50.6Maximum 25 178.0 77.6

stepping periods interspersed with 20 secondintervals of marching on the top of the step.During the test, subjects stepped to a metro-nome rate of 120 beats/minute, correspondingto a stepping rate of 30 steps/minute. On eachinitial step down, the Kistler force plate wastriggered, the GRF was measured, and theforces were recorded. Data collection for theplate was set at one second time periods. It wasdecided beforehand that the second attemptwould be taken for analysis; subjects were naiveto this so as not to influence their stepping pat-tern in any way. The above procedure wasadopted because ofoperating constraints oftheforce plate, thus enabling the force platform tobe reset. The data collection sessions werevideo taped to assist in data interpretation; thisalso allowed stepping frequency to be moni-tored.

STATISTICAL METHODSData were analysed using a one way repeatedmeasures analysis ofvariance, with F correctedfor heterogeneity of variance if necessary. Thecritical value for statistical significance wastaken as P0.05).

Analysis of the data for time to achieve peakforce at the various step heights showed no sig-nificant differences between the heights (F =1.98; df = 22,2; P>0.05).Analysis of the data for impulse at the

various step heights showed no statistically sig-nificant differences between the heights (F =0.84; df = 22,2; P>0.05).

Analysis of the data for peak force at footcontact at the various step heights showed asignificant within-subject effect (F = 8.67; df =22,2; P

Maybury, Waterfield

1300 r

1200

1100 H-

z

a.)L-

oLL

1000

900

800

700 _

600

500

400Peak force Peak force at Peak forceat 6 inches 8 inches at 10 inches

Figure 2 Box and whisker plot ofpeak force scores at 6, 8, and 10 inches. The horizontalband within the box denotes the median value, and the boundaries of the box represent theinterquartile range. The whiskers indicate the total range of scores.

forward descension of stairs. However, inbackward steps, moment forces are reducedcompared with those calculated for forwardstepping, because of the direction of theanteroposterior force; no magnitudes weregiven. Muscle activity also contributes to anincrease in knee moment force.8 This, in addi-tion to impulse loads, as suggested by Radin etal,9 gives rise to a possible mechanism by whichinjury may occur during step. Mital et aPsuggested that during stair climbing, raising thestep height increases the moment force aboutthe knee joint. It is a reasonable assumptionthat, as the step height is increased, momentforces acting about the knee joint on both thelowering leg during step down and the leadingleg during step up activities increase, leading todamage of the articular cartilage of thepatellofemoral joint and tibiofemoral joint.Therefore impact forces may not necessarily beas important as moment forces in any possiblemechanism of injury to both the lowering legand the leading leg.

Parallels can be drawn between step aerobicsand toe strike running, which places morestress on the plantar flexors of the ankle andflexors of the toes. However, toe strike runningis only used for short periods of time to helpcope with the increase in force that accompa-nies increased speeds; it could be theorised thatthe length of time spent stepping could causeinflammation of the flexor aponeurosis leadingto plantar fasciitis."6 One might expect thatthroughout the whole of the step downsequence, shock absorption is under consciouscontrol, the plantar flexors being preactivatedto aid shock attenuation in the early stages, butas fatigue sets in the active attenuation may beovershadowed by passive mechanisms. Largejoint flexions, particularly at the knee, may beused to assist in shock attenuation." It is thislast point that is important when consideringthe effect of power steps, where the use oflarger hip flexions would help to soften theimpact of landing. Reebok's instructions tosteppers' advise participants not to step downwith straight knees but to allow them to flex onlanding.

FURTHER FACTORS THAT MAY BE RELATED TOINJURY IN STEP EXERCISEEccentric muscle activity is the predominantform of muscle contraction used in stepping.Essentially, a step class can be viewed as a pro-longed activity that produces torque forces atthe hip, knee, and ankle,6 8 which, according toCleak and Eston," result in fatigue, maximalmuscle soreness, and damage. Eccentric con-tractions produce greater muscular work thanconcentric contractions, but rely on the recruit-ment of fewer fibres. Evans27 supports this viewand reports that fewer fibres are activated,resulting in muscle trauma caused by the gen-eration of high intramuscular tensions. Eccen-tric muscle activity could potentially lead todamage of nervous tissue, especially at nerve-muscle mechanical interfaces. Evidence thatthe main healing process within the muscle isthrough fibrosis is cited by Evans,27 and thusthe nerve itself may become adhered todamaged muscle fibres giving rise to adversemechanical tension type injuries.

Fatigue is a realistic problem that novicesteppers have to cope with, which was control-led in this study by subjects stepping for fiveminutes only. However, it is common fornovices to participate in a one hour class, ofwhich 40 minutes would be stepping, consist-ing of both basic steps and more forceful steps.It is conceivable that, in the poorly condi-tioned, both cardiorespiratory fitness and localmuscle endurance in the leg and thigh wouldbe compromised leading to an increase in GRFand ultimately increasing the risk of injury. Theconcept of active and passive shock attenuationintroduced by Nigg et al'4 gives an insight intothe mechanism by which this could occur.Under the influence of fatigue the activemechanisms of shock attenuation could bereduced and the body would be more reliant onits passive mechanisms. Joints are prime areasto be damaged. Radin et aP suggest that carti-lage readily resists rubbing but is damaged bylongitudinal compression. If the musculaturebecame compromised by fatigue, the increasedimpact force may result in longitudinal com-pression leading to micro fracture of subchon-dral bone. This could be followed by destruc-tion of the articular cartilage at the knee andpossibly the hip, and eventually stress fracturescould develop as the result of overuse. Injuriesneed not be confined to the lower limbs.According to Smeathers,"8 force would betransmitted to the spine, and consequentlydegeneration of intervertebral joints and head-aches may occur. In an individual with a latentmechanical back condition, this may causeuntold damage from the increased shockattenuation that would be placed on the spine.Work by Ricard and Veatch"6 indicates thatpassive impact forces can cause shearing andstretching of soft tissues, including nervous tis-sue.

Vertical force is believed to have the mostinfluence on longitudinal loading.8 This maycause the horizontal and anterioposteriorforces to be overlooked. These forces maybecome important in ankle injury as stepheight is increased. As the ankle moves from a

112

on June 30, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.31.2.109 on 1 June 1997. Dow

nloaded from

http://bjsm.bmj.com/

Relation between step height and ground reaction forces 113

loose packed position to a close packedposition from toe down to heel down, it relieson surrounding muscular control for support.It has been observed that subjects hesitatebefore stepping down from a higher platformand this hesitation may result in a poorlybraced ankle, leading to an ankle sprain.Fatigue could also exacerbate this.Aging footwear may play a role in the mech-

anism of injury. The implication for stepenthusiasts is that a reduction in forefoot cush-ioning may lead to an increased risk of injury,because it is the forefoot that contacts theground first. In step exercise, the ball of thefoot is placed down first, followed by the heel.A loss of forefoot cushioning'4 would lead toreduced shock attenuation by the shoe materialand increased shock attenuation by the body'sactive mechanisms, such as the intrinsicmuscles of the foot and plantar flexors. Thismay lead to plantar fasciitis or peritendinitis ofthe tendo Achillis. Loss of rearfoot cushioningwould not have such consequences, as ex-plained by Lafortune and Henning.'7A factor that may influence the risk of injury

is leg length. McCaw and Bates22 report that leglength inequality predisposes individuals to anincreased risk of injury. It could be postulatedthat on stepping down, the shorter leg wouldtravel slightly further before it contacted theground, which would mean the longer loweringleg would be subjected to slightly highermoment forces which would cause increaseddamage to the patellofemoral joint. In addition,the resulting muscle imbalance in the pelvis,brought about by intrinsic compensations forthe leg length inequality, may predispose theindividual to lumbar spine problems.

It is recognised that this pilot study has limi-tations; the sample is not truly representative ofthe general population of step participants interms of age, gender, and fitness levels. Furtherwork may need to be carried out to investigateif stratifying for age and fitness level influencesGRF in response to altering step height. Diffi-culties were observed in this study in step tim-ing and the maintenance of correct techniquewhich may need to be considered further,although this study did reflect what frequentlyhappens with novice steppers. Errors may havealso been introduced because ofthe orientationof the step platform.

ConclusionThe results of this study support the use of lowstep heights. Raising the step height appears toincrease the forces exerted on the stepper,although no such relation was found to exist interms of the impulse, time of foot contact, ortime to achieve peak force. From the findings ofthis study, mechanisms of injury can beinferred. Eccentric muscle work predominatesin step aerobics and this type of musclecontraction may lead to widespread muscledamage, producing adverse mechanical tensiontype injuries to nervous tissue. Fatigue, oftenexperienced by novices, could lead to stressfractures in the tibia or calcaneus and anincreased risk of osteoarthritis in later years.This is likely to arise through damage to

articular surfaces caused by impulsive loads,related to the reduced effectiveness of thebody's active shock attenuating mechanisms.However, patellofemoral and hip joint damagein stepping up and stepping down activitiescould be due to the moment forces about thesejoints in addition to GRFs. The relationbetween step height and moment forceswarrants further investigation.

We would like to thank Dr Julius Sim, Principal Lecturer inResearch Methods, Coventry University, for his statistical andgeneral advice on this paper.

1 Reebok UK Ltd. Step Reebok Instructors Manual. London:Reebok, 1990.

2 Robinson A. Watch Dog, BBC TV programme. BroadcastOctober 1993.

3 Johnson BF, Rupp JC, Berry SA, Rupp DA. Peak verticalground reaction forces (PVGRFs) and time to peak force(T=PF) in bench stepping and other activities. Med SciSports Exerc 1992;24(suppl 131):783.

4 Farrington T, Dyson RJ. Ground reaction forces during stepaerobics. Journal ofHuman Movement Studies 1995;29:89-98.

5 Nisell R, Mizrahi J, Ekholm J. Knee and ankle loads in stepdown activities. In: Jonsson B, editor. International series onbiomechanics. Biomechanics X Champaign, IL: HumanKinetics Publishers, 1987:1137-42.

6 Mital A, Fard HF, Khaledi H. A biomechanical evaluationof staircase riser heights and tread depths during stairclimbing. Clinical Biomechanics 1987;2: 162-4.

7 Andriacchi TP, Andersson GBJ, Fermier RW, Stern D,Galente JO. A study of lower-limb mechanics during stair-climbing. J Bone Joint Surg [Am] 1980;62:749-57.

8 Nisell R, Mizrahi J. Knee and ankle joint forces during stepsand jumps down from two different heights. Clinical Bio-mechanics 1988;3:92-100.

9 Radin EL, Parker HG, Pugh JW, Steinberg RS, Paul IL,Rose RM. Response of joints to impact loading. III.Relationship between trabecular microfractures and carti-lage degeneration. J Biomech 1973;6:51-7.

10 Anderson DD, Brown TD, Yang KH, Radin EL. A dynamicanalysis of impulsive loading of the extension-splinted rab-bit knee. Y Biomech Eng 1990;112:119-28.

11 Farksa T, Boyd RD, Schaffler MB, Radin EL, Burr DB.Early vascular changes in rabbit subchondral bone afterrepetitive impulsive loading. Clin Orthop 1987;219:259-67.

12 Lukoschek M, Boyd RD, Schaffler MB, Burr DB, Radin EL.Comparison of joint degeneration models. Surgical insta-bility and repetitive impulsive loading. Acta Orthop Scand1986;57:349-53.

13 McNitt-Gray J. Kinematics and impulse characteristics ofdrop landings from three heights. International Journal ofSport Biomechanics 1991;7:201-24.

14 Nigg, BM, Denoth J, Kerr S, Leuthi S, Smith S, Stacott A.Load sports shoes and playing surfaces. In: Frederick EC,editor. Sports shoes and playing surfaces. Champaign, IL:Human Kinetics Publishers, 1984:1-23.

15 Jefferson RJ, Collins JJ, Whittle MW, Radin EL, O'ConnorJJ. The role of the quadriceps in controlling impulsiveforces around heel strike. Proceedings of the Institution ofmechanical Engineers. Part H. Journal ofEngineering in Medi-cine 1990;204:21-8.

16 Ricard MD, Veatch S. Comparison of impact forces in highand low impact aerobic dance moves. International JournalofSport Biomechanics 1990;6:67-77.

17 Lafortune MA, Henning EM. Cushioning properties offootwear during walking: accelerometer and force platformmeasurements. Clinical Biomechanics 1992;7: 181-4.

18 Pratt DJ. Mechanisms of shock attenuation via the lowerextremity during running. Clinical Biomechanics 1989;4:5 1-7.

19 Strauber WT, Clarkson PM, Fritz VK, Evans WJ. Extracel-lular matrix disruption and pain after eccentric muscleaction. JAppl Physiol 1990;69:868-74.

20 Newham DJ, Jones DA, Clarkson PM. Repeated high forceeccentric exercise: effects on muscle pain and damage. JAppl Physiol 1987;63:1381-6.

21 Cleak MJ, Eston RG. Delayed onset muscle soreness:Mechanisms and management. Y Sports Sci 1992;10:325-41.

22 McCaw ST, Bates BT. Biomechanical implications of mildleg length inequality. BrJ Sports Med 199 1;25:10-13.

23 Portney LG, Watkins MP. Foundations of clinical research:applications to practice. Norwalk: Appleton & Lange, 1993.

24 American College of Sports Medicine. Guidelines for exercisetesting and prescription. 4th ed. Philadelphia: Lea & Febiger,1991.

25 Hinton P. Statistics explained. London: Routledge, 1995.26 Collins JJ, Whittle MW. Impulsive forces during walking and

their clinical implications. Clinical Biomechanics 1989;4:179-87.

27 Evans WJ. Exercise induced skeletal muscle damage. Physi-cian and Sports Medicine 1987;15:89-100.

28 Smeathers JE. Measurement of transmissibilty for thehuman spine during walking and running. ClinicalBiomechanics 1989;4:34-40.

on June 30, 2021 by guest. Protected by copyright.

http://bjsm.bm

j.com/

Br J S

ports Med: first published as 10.1136/bjsm

.31.2.109 on 1 June 1997. Dow

nloaded from

http://bjsm.bmj.com/