Journal of Strength and Conditioning Research, 2006. 20(21, 422-428© 2006 National Strength & Conditioning Association

EFFECTS OF BALANCE TRAINING ON SELECTEDSKILLS

JAMES A. YAGGIE' AND BUIAN M. CAMPBELL^

'Applied Biomechatiks Laboratory, Department of Exercise and Nutritional Sciences, San Diego State University,San Diego, CA 92182; -Kinesiology Division, Bowling Green State University, Bowling Green, OH 43403.

ABSTRACT. Ya^gie, J.A., and B.M. Campbell. Effects of balancetraining on selected skills. J. Strength Cond. Res. 20(2):422^28.2006.—The purpose of this study was to determine the effect ofa 4-week haiance training program on specified functional tasks.Thirty-six suhjects lage - 22.7 ± 2.10 years; height = 168.30 ±9.55 cm: weight = 71.15 ± 16.40 kg) were randomly placed intocontrol IC; n ^ 19) and experimental groups (Tx; ii = 17). TheTx group trained using a commercially available balance train-ing device (BOSU). Postural limits (displacement and sway) andfunctional task (time on ball, shuttle run, and vertical jump)were assessed during a pretest (Tl), a posttest (T2), and 2 weeksposttraining (T3). Multivariatf repeated measures analysis (« =0.05) revealed significant difTerences in time on hall, ^buttle run,total sway, and fore/aft displacement after the exerci.se inter-vention (T2). T3 assessment revealed that total sway and timeon ball remained controlled; however, no other measures wereretained. Balance training improved performance of selectedsport-related activities and postural control mea.sures, althoughit is unclear whether the effect of training would transfer togeneral functional enhancement.

KKY WOKU.S. core stability, exercise, proprioception, fitness

INTRODUCTION

ompetitive and recreational sports are depen-dent on multiple components of training and thedevelopment of stren^h, power, and endurance(13, 15, 19). Balance training is a relatively re-

cent phenomenon in the fitness industry that has devel-oped into a primary point of interest for consumers andfitness professionals (5, 14, 16, 23). Balance is comprisedof the dynamic reactions of involuntary sensations andimpulses that maintain an upright stance and is neces-sary for most functional movements (7, 10, 17, 19). Suc-cess in athletic and recreational activities depends onboth balance and functional movements. The proper func-tion of al! active muscles and the velocities at which thesemuscular forces are applied are crucial (23). Many rec-reational activities require lateral, forward, and back-ward movements during which the center of gravity(COG) is often at the edge of the base of support (BOS)(23). To maintain balance, it is necessary to have a func-tional awareness of the BOS to better accommodate thechanging COG (6). The goal of balance training is to im-prove balance tbrough perturbation of the musculoskel-etal system that will facilitate neuromuscular capability,readiness, and reaction (5, 21, 33).

In recent years, several commercial products havebeen developed to enhance and improve proprioceptivetraining. The development of the Biomechanical AnklePlatform System (BAPS) board mirrored the uniaxial andmultiaxial boards that were designed for rehabilitative

purposes to increase proprioeeptive activity in injured an-kles (9, 29). Reebok's Core board was designed to increaseproprioception and core stability and was targeted forthose outside the commercial rehabilitative setting. TheKinesthetic Ahility Trainer (KAT) (18) and the BalanceMaster (12, 14, 27) are computerized mechanical plat-forms designed to enforce calculated perturbations andvisual stimulations to challenge the muscular and visualsystems.

The Both Sides Up balance trainer (BOSU; FitnessQuest, Canton, OH) is an apparatus that was designedfor balance training within the athletic and recreationallyactive population. The design of the BOSU provides a sol-id plastic base integrated with an inflatable rubber blad-der that resembles a halved Swiss ball. The BOSU has asolid surface facing down that provides an unstable sur-face on stable ground (Figure 1). Furthermore, it is de-signed to improve stability not only while tbe user main-tains an upright position, but also when the user is in ahorizontal position (i.e., during abdominal exercises).

Functional ability can be exemplified by tbe perfor-mance of a sport-related task (35). Tbese tasks requireappropriate control of the neuromuscular and musculo-skeletal systems, including the proprioceptive system. Itis presumed that balance training has the most profoundeffect on the somatosensory and proprioceptive controlsystems (2, 23, 25, 35); however, conventional means ofassessment must quantify the training effects gained byproprioceptive control, and not proprioception itselfThrough skill assessment, inferences regarding interven-tions may offer insight into the effects on the propriocep-tive system (23, 24, 32).

Several studies (2, 9, 23, 24, 32) have found that bal-ance training enhances proprioceptive control. Most ofthese studies have investigated subjects with chronicallyunstable or injured joints of the lower extremities com-pared witb untrained healthy subjects. Few investiga-tions bave examined the efTects of balance training innoninjured individuals (5, 14, 16, 33). Rozzi et al. (28)found that a 4-week balance training program was an ef-fective means of improving joint proprioception and sin-gle-leg standing ability in subjects with unimpaired an-kles. The limited usage of the noninjured population inthis context characterizes an underrepresentation ofhealthy, physically active individuals interested in sta-bility training.

The purpose of this study was to determine the influ-ence of a balance-training protocol, using the BOSU, ondynamic stance and functional performance in healthy,recreationally active individuals.

422

BOSU BALANCE TRAINING 423

a). b).

FIGURE 1. Photographs of the Both Sides Up balanee trainer,la) Diameter of the base is 68 cm. (b) Approximate heightwben inflated is 25 cm.

METHODS

Experimental Approach to the ProblemAll subjects completed a pretest (Tl), a posttest (T2), anda retention test (T3) designed to determine whether in-dividuals experienced a positive response from a balance-training program and whether they retained those posi-tive effects 2 weeks after training was terminated. Priorto the inception of training, postural sway, postural lim-its, vertical jump height, shuttle run time, and time onthe BOSU were assessed (Tl). Repeated assessmentswere performed following a 4-week training protocol (T2)and again following a 2-week suspension of training (T3).Previous research indicates that a 2-week cessation oftraining can result in a significant reduction of trainingeffects, related to physiological and neuromuscular im-plications of deconditioning (4).

SubjectsThirty-six healthy reereationally active volunteers (meanage = 22.47 ± 2.10 years; height = 168.30 ± 9.55 cm;weight = 71.15 ± 16.40 kg) participated in this investi-gation. In compliance with Institutional Review Boardprocedures, all subjects were required to read and signinformed consent documents. Subjects were screened viainterview and self-report regarding activity behavior andinjury status. Those participating in vigorous activity andcardiovascular training (3 to 5 times per week) and mus-culoskeletal resistance activity (2 to 3 times per week)were considered recreationally active. Subjects reportedhaving no lower extremity trauma within the 2 years pri-or to the investigation. All were free of known balancedisorders and were considered to he in good health, ac-cording to the Physical Activity Readiness Questionnaire(PAR-Q). Those with corrective lenses were encouragedto wear them for each testing and training session. Sub-jects were paired for gender and self-reported activity lev-el and were randomly assigned to BOSU-trained (Tx; n= 17) and untrained (C; n ^ 19) groups.

ProceduresLimb Dominance and Postural Sway Atises.'iment. Prior topretesting, lower limb dominance was determined by hav-ing participants complete a ball-kick test and the step-uptest.

To assess postural sway parameters, the subjects wereinstructed to stand on an Advanced Medical Technolo-gies, Inc. (AMTI) force platform (Model OR-6; AdvancedMedical Technologies, Inc., Watertown, MA} using theirdominant leg with the nondominant leg flexed at a 45"angle at the knee joint and the arms placed across their

FIGURE 2. Schematic of sbuttle run.

chest. A visual target was placed approximately 3 m infront of the subjects as a focal point. Subjects were theninstructed to lean forward as far as possible without lift-ing any part of their foot off the platform, attempting tomaintain hip and knee extension at all times. Followinga familiarization trial, 3 15-second trials were performed,with the first 10 seconds standing upright and the last 5seconds leaning forward. Subjects were given 2 minutesof rest between trials. This method is similar to thosefound in the literature (3, 34).

Functional Tasks. Three functional tasks were used inthis investigation, balance on the BOSU (TOB), verticaljump (VJ), and shuttle run (SR). These assessments wereselected because of tbe specificity of the balance device(TOB) and the ability to assess the influence the trainingon power (VJ) and agility (SR).

Time on Ball. Subjects were instructed to stand on theBOSU using their dominant leg. Once comfortable, theywere asked to close their eyes and maintain that positionwithout falling or touching the ground or the BOSU withtheir nondominant leg. If the subject lost control of thatposture, the investigator stopped the watch and recordedthe time spent on the BOSU.

Shuttle Run. The shuttle run course used in this in-vestigation (Figure 2) was similar to that found in theliterature (34). The course included several directionalchanges and involved sprinting, backpedaling, sidestep-ping, and starting-and-stopping patterns. The time takento complete this running course was measured by the in-vestigator using digital timing pads placed at the startand finish (DT2819, Melbourne, FL). Subjects completedseveral practice trials to familiarize themselves with thecourse. Following the practice trials, 3 running trialswere performed by each subject with 2 minutes of restbetween each trial.

Vertical Jump. The vertical jump protocol used in thecurrent investigation has been described in the literature(34). Subjects performed a standing vertical jump withtheir preferred arm adjacent to the wall. The subjectsreached up with the selected hand, keeping their heels onthe floor and flrmly marked the wall by touching theinked flnger to the wall. The height was recorded.

Subjects were instructed on the proper form of a ver-tical jump. Each participant was allowed a countermove-ment prior to take off as long as a step approach was notused. The subjects were instructed to touch the wall withtheir inked hand at the peak of the jump. The distancebetween the standing reach and vertical jump marks wascalculated. The average from 3 vertical jump trials wascalculated and used for the analyses.

Training Procedure. The balance training on theBOSU began within 2 days following the pretest. The

424 YAGGIE AND CAMPBELL

a). b).

FIGURE 3. Example of exercise progresHion of ball stancewith la) unilateral stance on ball, and (b) unilateral stancewith trunk excursion on ball.

training protocol consisted of exercises progressing fromthe simplest to most complex sessions. The protocol thatwas used in the current investigation was a commerciallydeveloped training program that is provided with theBOSU at the point of sale (8). The prescribed protocolincluded activities that are consistent with exercises de-scrihed in the literature (2, 21, 23, 31).

The Tx group trained on the BOSU 3 times per weekfor approximately 20 minutes. Each week the suhjectswere presented with more difficult variations of exercisesto replace those already mastered. Mastery was definedas remaining on the BOSU for a period that was morethe 2 times longer than the previous session without fall-ing off or adding support. The selected exercises were de-signed to challenge one or more of the sensory systemsintegral in maintaining halance. Additions and modifi-cation to the battery of exercises included rotating thehead laterally, tilting the head upward, keeping the eyesopen or closed, and using the trunk excursion or lean(Figure 3).

All suhjects were asked to maintain a log to track theamount and intensity of elective activity during the test-ing period. Inclusion of each subject was hased on thefrequency of activity (3 to 5 sessions per week) and thenumber of average hours logged (1 to 3 hours per occur-rence). Those that cataloged the minimum hours withoutexceeding a maximum value (15 total hours per week)were included in the study. Activity included total timein participation of recreational activities and sports,strength and resistance training, exercises classes, flexi-bility training, and cardiovascular exercise. Initially,evenly numbered groups were recruited for the study;however, these criteria lead to the exclusion of data from4 subjects (C - 1; Tx - 3).

Instrumentation. Each subject's balance was assessedusing the AMTI force platform in conjunction withBalanceTrak software (Motion Analysis Corporation,Santa Rosa, CA) to detennine specific sway patterns. The

BalanceTrak software produced a trial duration of 15 sec-onds and sampled data at a rate of 200 Hz. The force platewas calibrated prior to data collection hefore each session.

Quiet Stance. Totai sway (TS), medial-lateral sway(MLS), fore-aft sway (FAS), fore-aft displacement (FAD),and medial-lateral displacement (MLD) were collected foreach suhject during each testing session (Tl, T2, and T3)and were compared across time and treatment.

Lean Test. During the unilateral forward lean test de-scribed by Blaszczyk et al. (3), voluntary MLD and FADof the center of foot pressure was estatalished while thesubject maintained a rigid body posture and leaned for-ward at the ankle joint. Correct body posture was verballyexplained, demonstrated, and observed by the investiga-tor prior to the testing sessions. The software calculatedthe excursion of the x and y values and determined themaximum FAD and MLD. Trials exhibiting incorrectleaning posture were discarded and not included in theanalysis.

Statistical Analyses

SPSS for Windows (Version 13.0; SPSS, Inc., Chicago, IL)was used for statistical analyses. Data were recorded andcoded for gender and treatment group. A repeated mea-sures multivariate analysis of variance (MANOVA; a <0.05) was computed to determine significant differencesbetween the Tx and the C groups, as well as between Tl,T2, and T3 testing time points. The dependant variablesthat were examined were TOB (seconds), VJ height (cm),SR time (seconds), TS (cm), MLS (cm), FAS (cm), FAD(cm), and MLD (cm). Bonferroni pairwise comparisonswere used to test main effects.

RESULTS

Means and standard deviations for all variables may heviewed in Table 1. Analysis of variance (ANOVA) re-vealed that no main effects of gender were observedacross time and condition ip < 0.001).

Time on Ball

Intraclass correlation coefficients (ICCs) were calculatedfor the TOB values to examine the internal consistencyof the data across time and condition. Consistency as-sessment was run using a 2-way mixed model set for ab-solute agreement of measures. ICCs for the C group forTOB were considered acceptable (R - 0.86).

Repeated measures MANOVA revealed a significantdifference in time and treatment (F = 25.98; observedpower = 1.00). In the C group, pairwise comparison noteda significant difference between Tl and T3 while differ-ences hetween Tl and T2 and between T2 and T3 werenot significant.

In addition, pairwise comparison revealed significantdifferences in the Tx group across Tl and T2 and acrossTl and T3; however T2 and T3 did not differ. It shouldbe noted that both groups displayed a progressively lon-ger TOB across sessions (T1-T2-T3), albeit not statisti-cally significant. Figure 4 illustrates these findings.

Shuttle Run

Intraclass correlation coefficients for SR were calculatedand considered acceptable (R ^ 0.84) for the C group.

A significant main effect in time and treatment {F =4.82; observed power ^ 0.764) was observed. Bonferroni

BOSU BALANCE TRAINING 425

TABLE 1. Mean (± SD) for functional and sway parameters across session and group.^

Functional TasksTOB (s)

VJ (cm)

SR(s)

Sway ParametersTS (cm)

MLS (cm)

FAS (cm 1

FAD (cm)

MLD (cm)

TxCTxCTxC

TxCTxCTxCTxCTxC

Tl

2.35 ± .6982.75 ± 1.06

41.30 ± 10.2147.91 ± 13.3113.16 ± 1.4712.62 ± 2.01

32.25 ± 11.433.18 ± 10.2

.65 ± .26

.62 ± .261.41 ± .251.46 ± .251.59 ± .92

1.5 ± .73.610 ± .217.626 ± .232

T2

3.71 ± 1.863.11 ± 1.27

40.40 ± 9.2448.98 ± 14.2112.45 ± 1.8712.70 ± 2.07

27.9 ± 9.633.56 ± 7.5

.59 ± .28

.67 ± .341.46 ± .33

1.5 ± .342.35 ± .911.58 ± .62.660 ± .323.544 ± .326

T3

3.96 ± 1.553.38 ± 1.57

41.28 ± 9.5549.24 + 13.9312.77 ± 1.3812.38 ± 2.08

29.25 i 7.631.89 ± 9.1

.64 ± .27

.62 ± .251.4 ± .24

1.45 ± .252.02 ± .81.65 ± .75.584 ± .224.640 ± .381

* TOB = balance on the Both Sides Up balance trainer; VJ == vertical jump; SR = shuttle run; TS = total sway; MLS = medial-lateral sway; FAS = fore-aft sway; FAD = fore-aft displacement; MLD = medial-lateral displacement; Tx = experimental group; C= control group; Tl, T2, T3 - testing sessions.

5 -

3 -

cS 2CO

1 -

T2

Testing Sessions

FIGURE 4. Mean (± SD) of time on ball (TOB) task.Significant differences were found in the experimental (Tx)group between sessions (a) T1-T2 and (b) T1-T3. (c) Significantdifferences were noted between T1-T3 in the control (C) group.The.se comparative results may indicate a learning effect ofTOB.

T2

Testing Sessions

DTx

FiGURK 5. Mean ("*: SD) of shuttle run time, (a) A significantdifference was found in the experimental (Tx) group betweensessions Tl and T2. C - control group.

T2Testing Session

FIGURE 6. Mean (-̂ SD) of total sway (TS) parameters acrosssessions Tl, T2, and T3. (a) A significant decrease in TS wasobserved in T2 and (b) T3 when compared to Tl in theexperimental (Tx) group. However, no differences were notedin T2 to T3. No significant differences were noted in thecontrol (C) group.

painvise comparisons indicated that the Tx group expe-rienced a significant decrease in shuttle run time betweenTl and T2; however no discernable differences were ob-served between Tl and T3 or between T2 and T3. No dif-ferences were observed in the C group (Figure 5).

Vertical Jump

Intraclass correlation coefficients for VJ were calculatedand considered acceptable (R = 0.91) for the C group.

Results revealed no significant differences betweentimes or treatments (F = 2.43; ohserved power = 0.669).

Sway Values

Intraclass correlation coefficients for force plate datawere calculated and considered acceptable {R =0.72) forthe C group.

Significant main effects for time (Tl, T2, and T3) andtreatment (Tx and C) were noted (F = 4.01; observed pow-er = 0.710) for TS (Figure 6). A significant decrease in

426 YACGIE AND CAMPBKIJ.

T2

Testing Session

FIGURE 7. Mean (± SD) of fore-aft displacement (FAD)values across testing session, (a) A significant increase in FADwas noted between Tl and T2 in the experimental (Tx) group;however, the effects of training were not retained. Nodifferences were observed in the control IC) subjects.

TS was observed in the Tx subjects immediate posttrain-ing (T2 = 27.91 ± 9.7 cm) compared to the pretrainingvalues (Tl - 32.25 ± 11.4 cm). Mean retention scores (T3= 29.24 ± 7.6 cm) of TS for trained subjects also showeda significant decrease in sway when compared to those ofTl and did not significantly differ from those of T2, in-dicating that the skill was retained following 2 weeks ofsuspended activity. No differences were noted in the Csubjects across time.

No significant main effects were noted in the MLS (F- 0.022; observed power = 0.053) and FAS iF - 0.415;observed power = 0.111) data for either the trained oruntrained groups.

Lean Test ParametersMANOVA exhibited significant main effects in FAD (F =38.80; observed power = 1.00) for hoth treatment andtime. Pairwise comparisons noted a significant increasein FAD from pretraining (Tl = 1.59 ± 0.6 cm) to post-training (T2 = 2.34 ± 0.9 cm) in the Tx participants. Thechanges in FAD were not statistically significant follow-ing the retention test (T3 - 2.02 ± 0.9 cm), but they dis-played a tendency toward preservation of enhanced skill(p = 0.55) (Figure 7). No significant differences were not-ed in the C group across time. MANOVA did not revealmain effects for MLD {F = 0.792; observed power - 0.591)values across time or treatment.

DISCUSSION

The results of TOB performance indicated the presenceof a training effect relative to the differential data fromTl to T2 in BOSU-trained subjects. Furthermore, the Txgroup had no significant difference between the posttestand retention test when training concluded. TS data re-vealed a similar response to training. In both instances,comparative T2-T3 values did not differ, suggesting thatthe acquired skill level was retained following 2 weeks ofinactivity. The retention of these skills was not expectedbut seems reasonable considering the nature of the activ-ities. The additional quiet-stance parameters did not ex-hibit significant results and were relatively consistentacross session and treatment.

Each assessment was modeled to represent a dynamicposture or stance. The sway vaiues were measured on the

force platform and were a familiar type of activity for anactive, healthy subject. The familiarity of the posture mayhave influenced the quiet stance assessment. The stanceof the TOB assessment typically exhibited a posture thatrequires the rear foot to sink due to the compliance of theBOSU, causing a slight dorsiflexion angle in the ankle.Witb the joint in this closed-packed orientation and load-ed in a single-legged stance, the proprioceptive infiuenceof the joint may have been agitated to meet the demandof stressed joints, eliciting a more controlled stance. Thisnotion could he tested by including individuals with un-stable or chronically injured ankles. It could he predictedthat the information from the sensory systems may havebeen confounded, eliciting alternate results due to pro-prioceptive loss.

Research has shown that individuals with decreasesin postural-sway parameters (better balance) may expe-rience fewer lower extremity injuries. Likewise, thosewith poor balance (higher sway values) experience in-creased injury rates (26). Tropp et al. (32) found thatwhen using uniaxial and multiaxial ankle disks in bal-ance training, there was a significant decrease in posturalsway of soccer players with known ankle injuries. This isfurther supported by Rozzi et al. (28), who found that hal-ance training may he used to restore ankle stability, pre-sumably by training altered afTerent neuromuscularpathways after an injury. Therefore, balance trainingmay improve the ability of proprioceptive pathways thatwere previously injured, resulting in improvement in bal-ance and a decrease in sway parameters.

Although no significant differences were ohserved inthe C group, a trend of enhanced performance had de-veloped across sessions, indicating a learning effect as-sociated with assessment. The Tx group also displayedthe trend of increasing the TOB with each session. There-fore, it is possible that both training and learning of thisspecific skill could have enhanced the performance of theTOB task across sessions. The learning effect ohserved inbalance assessment has heen addressed in the literature(1, 11, 25, 35). However, covariation of these effects wasconsidered through the use of multivariate analysis.

Training also influenced performance of the SR intrained subjects. A significant decrease in SR time wasohserved between the pretest (Tl) and posttest (T2) forthe Tx group, hut not for the C group. Furthermore, thetraining effect was diminished as the SR times returnedtoward the pretest values after a 2-week retention inter-val (T3). A similar result was noted in the FAD in Txsubjects, suggesting that BOSU training was influentialin enhancing dynamic sway parameters involving excur-sion of the center of gravity or mass.

It is well documented that consistent activity andtraining of the lower extremities may infiuence the re-action time, proprioception, and muscle activation of cru-ral musculature (1, 2, 22, 24, 25, 35). Peroneal musclereaction time has been examined in multiple researchmodels and disease states. One of the hallmarks of achronically unstable ankle is the loss of reactionary con-trol of the lateral musculature of the crurum. This typeof condition leads to poor muscle activation, joint motion,and alteration of the center of pressure of the foot (20).Furthermore, Lentell et al. (22) reported that instabilityof the ankle is not a result of muscle weakness but isassociated with the presence of proprioceptive deficits.These kinematic outcomes result in modifications

BOSU BALANCE TRAINING 427

throughout the kinetic chain that alter the inverse dy-namics of the knee and hip. This results in a delay in theinherent mechanisms and refiex loops used to control pos-ture and balance. Therefore, the training of these mus-cles, among others, will enhance reaction and propriocep-tive influences of the lower extremity and will result inimproved postural control.

The objectives of balance training are to augment theafferent pathways to enhance the sensation of joint move-ment (23). The influence of sensory enhancement is moreohvious in those that have impairment or are unfamiliarwith specific tasks. In the present study, it was ohservedthat the dynamic tasks of SR and FAD improved withtraining {T1-T2) and returned toward basehne once train-ing had been suspended (T2-T3). Improvements in per-formance of dynamic or functional tasks may not be asapparent when examining the uninjured, competitiveathlete. Therefore, the specificity of athletic performanceand the level of competition should be considered whendesigning programs for highly trained individuals, orthose participating in athletics.

This idea was supported by Soderman et al. (30), whoinvestigated the use of halance board training in the pre-vention of traumatic lower extremity injuries. Although,no traditional balance measures were assessed, the re-sults revealed that there were no significant changes insoccer players receiving balance-training interventionwith respect to the number, incidence, or type of trau-matic injury to the lower extremities. The lack of statis-tical findings may have heen masked by the rigor of thesoccer training regimen. The authors contend that bal-ance hoard training may not have sufficiently stressedthe neuromuscular system to notice any significant im-provements when examined during the soccer season. Al-though changes were expressed in the present study, theinfluence of training on dynamic performance may be dif-ficult to discern in young healthy populations.

Furthermore Hakkinen and Myllyla (13) reported thatforce production and reaction in athletes training for ac-tivities using power and strength are significantly moreforceful and responsive than those athletes that are en-durance trained. This paradigm seems reasonable consid-ering the demand of muscle activation for power andstrength activity. However, the current investigationlacks support of enhancement of strength and power giv-en the nonsignificant findings in vertical jump perfor-mance for both the Tx and C groups. If the increases inagility skill (SR) were coupled with those of VJ, the im-provement in performance could he explained throughmuscular adaptations and strength gains. The presenceof significant improvements in dynamic tasks indicatesthat agility, skill, and excursion may he attrihuted to neu-rological adaptation to activity and proprioceptive actionof the trained joints and soft tissues. This model is con-sistent with the observational gains in ahility followingshort-term training (less than 4 weeks). Perhaps gains instrength and power would hecome apparent with modi-fication to the training protocol.

The recruitment of recreationally active participantsincluded those participating in sport and exercise 3-5times per week hut excluded those in training or compe-tition. The exclusion of tbe highly trained allowed a morehomogeneous model representative of the fit, young, andable population that predominates the membership of fit-ness centers and gyms. It was also assumed that this

group represented a target market of commercially de-veloped exercise products. Future studies may find causeto investigate the differential effects of balance trainingin those in competitive sports, including power, strength,and endurance athletes.

PRACTICAL APPLICATIONS

These data denote that balance training improves perfor-mance of selected activities. The absence of such effectsupon retention assessment (T3) suggests that benefits aretransient and subject to issues related to compliance oftraining. The improvements in TS and FAD in the Txgroup indicate that training may influence proprioceptiveinput, reaction time, and specified muscular strength, inexisting postural control mechanisms via neuromuscularadaptation to activity. There was no ohservable change inthe performance of vertical jump posttraining, which maysuggest that BOSU training may not affect those activi-ties that relate to power skills. BOSU balance trainingmay improve performance of dynamic skills and sway pa-rameters; however, it is unclear if those skills are trans-ferable to the performance of recreational or competitivesport and activity. Practically, the benefits of balancetraining may provide an enhanced sense of control to theclient or user, warranting its use as a training techniqueor a prescriptive activity for the exercise professional.

REFERENCES1. A|)LP;RT()N, A.K., AND U. MoHITZ. Does calf muscle fatigue af-

fect standing balance? Scand. J. Med. Sci. Sports 6:211-215.1996.

2. BERNii-:ii, J.N.. AND D.H. PKKKIN. Effect of coordination train-ing on proprioception of the functionally unstable ankle. J. Or-thop. Sports Phys. Tlwr. 27(4):264-275. 1998.

3. BlASZrzYK, J.W.. D.L. LdWE, AND P.D. HA.MSON. Ranges of pos-tural stability and their changes in the elderly. Gait Posture 2:11-17. 1994.

4. Cm M.M.Y., C.S. NINT/., AND E.F. COYLE. Effects if detrainingon enzymes of energy metabolism in individual muscle fibers.Am. J. Physiol. 244:C276-C287. 1983.

5. Cox, E.D.. S.M. LEPMART. AND J.J. IRRANC;. Unilateral balancetraining of noninjured individuals and tbe effects on posturalsway. J. Sport Rehabil. 2(2):87-96. 1993.

6. DiKTZ, V. (Jating of refiexes in ankle muscles during humanstance and gait. Prog. Brain Res . 97:181-188. 1993.

7. DIET/,, V.. U. ScnMinTBLEiCHEK, AND J. NOTH. Neurona! mech-anism of human locomotion. J. Neurophysiol. 42:1212-1222.1979.

8. DW FITNESS, LLC. BOSU Balance Trainer Instructional Man-ual. Madison, NJ: DW Fitness, LLC. 2001.

9. FK)RE, R.D., AN!) J.S. LKARH. A functional approach in the re-habilitation of the ankle and rear foot. J. Athl. Train. 15:231-2.35. 1980.

10. FRANK, J.S., AND M. EARL. Coordination of posture and move-ment. Phys. Ther. 70:885-863. 1990.

11. GOLDIE, P.A., E.M. EVANS, AND T.M. BACH. Steadiness in one-legged stance: Development of a reliable force-platform testingprocedure. Arc/?. Phys. Med. Rehabil. 73:348-354. 1992.

12. HACEMAN, P.A., J.M. LEiiiowiTZ, AND D. BLANKE. Age and gen-der effects on postural control measures. Arch. Phys. Med. Reh-ahil. 76:961-965. 1995,

13. HAKKINEN, K., AND E. MYI.LYLA. Acute effects of muscle fatigueand recovery on force production and relaxation in endurance,power and strengtb atbletes. J. Sports Med. Phys. Fitness 30( 1):5-12. 1990.

14. HAMANN, R.G., 1. MEK-JAViC, A.I. MAI.LIN'SON. AND N.S. LoNG-RIDC.E. The effects during repeated therapy sessions of balancetraining visual feedback. Arch. Phys. Med. Rehabil. 73:738-744.1992.

428 YAGGIE AND CAMPBELL

15. HiCKSON, R. Interference of strength development by simul-taneously training for strength and endurance. Eur. J.AppLPhysiol. 45:255-263. 1980.

16. HOFFMAN, M., AND V.G. PAYNE. The effects of proprioceptiveankle disk training on healthy subjects. J. Orthop. Sports Phys.Ther. 21(2):90-93. 1995.

17. HoRAK, F.B., L.M. NASHNP:K, AND H.C. DIENER. Postural strat-egies associated with somatosensory and vestibular los.s, Exp.Brain Res . 82:167-177. 1990.

18. JOHNSTON, R.B., M.E. HOWARD, P.W. CAWLEY, AND G.M. LOS-SE. Effect of lower extremity muscular fatigue on motor controlperformance. Med. Sci. Sports Exerc. 30:1703-1707. 1998.

19. KOMI, P.V. Biomechanics and neuromuscular performance.Med. Sci. Sports Exerc. 16a):26-28. 1984.

20. KoNRAnsKN, L., AND J.P. RAVN. Ankle instability caused byprolonged peroneal reaction time. Acta Orthop. Scand. 61(5):338-390. 1990.

21. LASKOWSKI, E.R., A.K. NEWCOMBR, AND J. SMITH. Refining re-habilitation with proprioception training: Expediting return toplay. Physician and Sports Med. 25(10):89-102. 1997.

22. LENTELI., G.L., L.L. KATZMAN, AND M.R. WALTERS. The rela-tionship between muscle function and ankle stahility. J. Or-thop. Sports Phys. Ther. ll(12):605-611. 1990.

23. LP:PHART. S.M., D.M. PINCIVERO, J.L. GIRAI.DO, AND F.H. Fu.The role of proprioception in tbe management and rehabilita-tion of athletic injuries. Am. J. Sports Med. 25ll):13a-137.1997.

24. LEPHART, S.M., D.M. PINCTVERO, AND S.L. Rozzi. Propriocep-tion of the ankle and knee. Sports Med. 25(3):149-155. 1998.

25. LuNDiN, T.M., J.W. FEUERBACII, AND M.D. GRABINKK. Effect ofplantarfJexor and dorsiflexor fatigue on unilateral posturalcontrol. J. AppL Biomech. 9:191-201. 1993.

26. McGiriNE, T.A., J.J. GREENE. T. BEST, AND G. LEVERSON. Bal-ance as a predictor of ankle injuries in high school hasketballplayers. Ciin. J. Sport Med. 10(4):239-244. 2000.

27. NICHOLS, D.S., T.M. GLENN, AND K.J. HUTCHINSON. Changesin mean center of balance furring balance testing in youngadults. Phys. Ther. 75(8):699-706. 1995.

28. ROZZI, S.L., S.M. LEFHART, R. STKRNER, AND L. KULIGOWSKI.Balance training for persons with functionally unstable ankles.-/ Orthop. Sports Phys. Ther. 29(8):478-486. 1999.

29. SoDERBERc, G.L., T.M. COOK, S.C. RIDER, ANII B.L. STEPHEN-ITCH. Electromyographic activity of selected leg musculature in.subjects with normal and chronically sprained ankles perform-ing on a BAPS Board. Phys. Ther. 71(7):514-522. 1991.

30. SODERMAN, K., S. WERNER, T. PIETILA, B. ENGSTROM, AND H.ALFREDSON. Balance board training: Prevention of traumaticinjuries of tbe lower extremities in female soccer players? Aprospective randomized intervention study. Knee Surg. SportsTraumatol. Arthrosc. 8(6):356-363. 2000.

31. TROPP, H., C. ASKLING, AND C. GILQUIST. Prevention of anklesprains. Am. J. Sports Med. 13(4):259-262. 1985.

32. TROI'P, H., J . EK.STKAND, AND J. GILL(JUI.ST. Factors affectingstabiliometry recordings of single limb stance. Am. J. SportsMed. 12(3):185-188. 1984.

33. WOLF, S.L., H.X. BARNHART, G.L. ELLLSON, C.E. C(KX:LER, ANDATLANTA F I C S I T GROUI'. The efTect of Tai Chi Quan and com-puterized balance training on postural stability in older sub-jects. Phys. Ther. 77(4):371-381. 1997.

34. YAGGIE, J.A., AND S.J., MCGRKCOR. Effects of isokinetic anklefatigue on the maintenance of balance and postural limits.Arch Phys Med Rphahil 83:224-228. 2002.

35. YAGCIE J.A., AND S.J. KINZEY. A comparison of selected ankleortboses on the vertical jump and shuttle run. J. Sports Reh-abil. 10:174-183. 2001.

AcknowledgmentsThe authors would like to acknowledge the efforts of Kristen C.Cummings in her assistance with data collection and testingduring this project.

The results of the study do not constitute endorsement of theproduct by the authors or the National Strength andConditioning Association. All equipment was procuredindependent of external funding or corporate sponsorship.

Address correspondence to James A. Yaggie, jyaggie®mail.sdsu.edu.