Core Muscle Activity During Exercise on a Mini Stability Ball

The Journal of Applied Research • Vol. 7, No. 3, 2007 255

with the floor crunches. The lightestexercise (sitting crunches with the ministability ball behind the back) was aboutequal to half of the work per second asfloor crunches. However, the mostintense exercises with the mini ball wereas much as 4 times the work as abdomi-nal crunches per second of exercise. Thegreatest difference in the mini stabilityball exercise was seen when the degreeof flexion/extension was increased from50 to 90 degrees. This degree of flexioncannot be accomplished with standardfloor crunches or with the Swiss ball(due to its larger diameter and size),thereby giving the mini stability ball asignificant advantage in working themuscles harder and at a better range ofmotion.

Core Muscle Activity DuringExercise on a Mini Stability BallCompared With AbdominalCrunches on the Floor and on aSwiss BallJerrold S. Petrofsky, PhD*Jennifer Batt, BS†Nicceta Davis, PhD*Everett Lohman, DPT Sc, DSc*Michael Laymon, DPT Sc,DSc†Gerson E. De Leon, BS*Heidi Roark, BS*Tony M. Tran*Enzo Gabrial Ayson, BS*Krista M Vigeland, BS*Claire E. Payken*

*Department of Physical Therapy, Loma Linda University, Loma Linda, California†Department of Physical Therapy, Azusa Pacific University, Azusa, California

KEY WORDS: exercise, exertion, coremuscle strength, fitness

ABSTRACTTen subjects were examined to deter-mine muscle use that occurred duringcore body exercise using a 7-inch diame-ter mini stability ball produced bySavvier LP (Santa Fe Springs, Ca) com-pared with abdominal crunches on thefloor and on a Swiss ball. Muscle usewas evaluated through the surface elec-tromyogram recorded above the abdom-inal and lower back muscles. Threelevels of core exercise were tested withthe mini stability ball. The resultsshowed that crunches on the Swiss ballused approximately 50% more musclework per second of exercise as did work

Vol. 7, No. 3, 2007 • The Journal of Applied Research256

By varying the angle of the back dur-ing the exercise, the mini stability ballhad a wide range of exercise intensitiesthat could be accomplished by thebeginner, people with poor conditioning,or provide an intense workout for physi-cally fit people.

INTRODUCTIONCore muscle exercise is common intraining programs.1,2 Typically, abdomi-nal crunches are used to train the rectusabdominis and oblique muscles.1,2

However, recent studies show that con-siderable muscle activity can beachieved with abdominal exercisedevices including Swiss balls,3,4 followingexercise videos,5 or both.

This type of exercise activity canhave numerous benefits. These includestabilization of the core of the body,6reduction in lower back pain,7,8 reduc-tion in plasma lipids,9 increased recoveryand oxygen kinetics following exercisetraining,10 improvement in function incardiac patients,11 better blood pressurecontrol in stabilizing orthostatic intoler-ance,12 and increased skeletal muscleblood flow, which has been linked tonitric oxide production.13

Although most of the studies con-cerning aerobic exercise involve run-ning, cycling, skiing, and other types ofsports,14 considerable core and lowerbody activity can be achieved in manyother types of exercise.15 For example,abdominal core strengthening activitywith a portable abdominal machine,3,4

abdominal shaping machines,16 or otherabdominal training devices,17,18 all causecore strengthening and muscle training.

The abdominal or core muscles, how-ever, are unique. Strengthening thesemuscles not only has strong centraleffects in the body, such as cardiovascu-lar training, but also causes an increasein stability of the trunk to extend duringreach and functional activities.19 These,in turn, reduce the risk of back injury.Lower and upper back injuries havealways been a major problem in theUnited States and throughout theworld.20-23 Generally speaking, the firstback injury usually occurs when peopleare in their twenties, but they do notexperience a reoccurrence of the injuryuntil their forties and fifties.24,25 Theseback injuries cost the American publicbillions of dollars each year in medicalcare and lost wages.26 They are especiallytaxing on the Worker’s CompensationSystem in that they commonly occur inthe work environment.27,28 Numerousstudies have shown that the core mus-cles in the body, in particular theabdominal muscles, are correlated to theincidence of back injury29-31 becausethese muscles are used to stabilize thetrunk; strength in these muscle groupsstabilizes the spine and balance. For thisreason, the United States Army usescore muscle strength as a predictor ofback injury in recruits.32

Traditionally, abdominal cruncheshave been used to train the core musclesof the body.17,18 But, abdominal crunchesprovide only low levels of muscle activi-ty since the only resistance to core mus-cle activity is the body weight.3,4,19

Therefore, large numbers of repetitionsare needed to train. To increase musclework, devices such as Swiss balls havebeen used as an adjunct to abdominalcrunches. By performing exercise on aSwiss ball, there is decreased balancestability during exercise, thereby increas-ing muscle work in the core muscles.33-35

The Swiss ball offers the advantage overfloor crunches in that exercise on theSwiss ball allows a greater extension

Table 1. General Characteristics of theSubjects

Age (yrs) Height (cm) Weight (kg)24.5 ± 2.3 172.7 ± 8.6 75.3 ± 14.0

All results given as mean ± standard deviation


muscles during exercise. In the presentinvestigation, the electromyogram(EMG) was used to assess muscle activi-ty during exercise on a mini stability ballcompared with the muscle use duringabdominal crunches on the floor and ona Swiss ball.

during abdominal exercise comparedwith lying on the floor.33-35 This offersconsiderable improvement over floorcrunches but muscle activity is still notoptimal for training. Therefore, a newmini stability ball was tested here in avariety of exercises to examine its abilityto cause recruitment of the key core

Figure 1. Typical subject on Swiss ball. Figure 2. Level 1 abdominal exercise is illus-trated here with 30 degrees of flexion at thewaist.

Figure 3. Subject accomplishing abdominalexercise with the mini ball and the handsabove the knees.

Figure 4. Subject accomplishing abdominalexercise with the mini stability ball under theupper gluteus muscle.

Figure 5. A subject performing exercise 1,level 3, with the mini stability ball.

Figure 6. Subject performing the level 3 wiperexercise with the mini stability ball.


SUBJECTSThe subjects in this study were 3 maleand 7 female subjects in the age range of18 to 35 years. Subjects were fit and freeof any cardiovascular, neuromuscular, ororthopedic injuries that would preventtheir inclusion in these studies. All meth-ods and procedures were explained toeach subject who then signed a state-ment of informed consent. The studiesand consent form were approved by theHuman Review Committee at AzusaPacific University. The general charac-teristics of the subjects are listed inTable 1. The number of subjects waschosen such that, based on the varianceof the data in previous studies, statistical

significance could be achieved as perpower analysis.

METHODSDetermination of Muscle ActivityTo determine muscle activity, the EMGwas used. EMG was recorded by 2 elec-trodes and a ground electrode placedabove the active muscle.36-41 The rela-tionship between tension in muscle andsurface EMG amplitude is linear.37,42

Thus, the amplitude of the surface EMGcan be used effectively as a measure ofactivity of the underlying muscle by sim-ply normalizing the EMG in terms of amaximal effort. Muscle activity wastherefore assessed by first measuring the

Figure 7. The muscle activity of the oblique,rectus abdominis, and back extensor musclesduring abdominal crunches in the forward,left, and right directions during floor crunches.Muscle activity is shown for all 4 musclegroups examined as the mean ± the standarddeviation for the group.

Figure 8. The muscle activity of the oblique,rectus abdominis, and back extensor musclesduring abdominal crunches in the forward,left, and right directions during crunches onthe Swiss ball. Muscle activity is shown for all 4muscle groups examined as the mean ± thestandard deviation for the group.


EMG of the muscle during a maximaleffort and then, during each exercise,assessing the percent of maximum EMGto calculate the percent of muscle activi-ty.37,38 Two EMG electrodes wereapplied, 1 over the muscle and one 2 cmdistal to the belly of the muscle. A thirdelectrode, the guard, was attached within4 cm of the 2 active electrodes. The elec-trode placement for the rectus abdomin-is was just above the umbilicus and

parallel to the muscle fibers. The place-ment for the oblique muscles was 2 cmabove the anterior superior iliac spineand at the pelvis of the posterior superi-or iliac spine for the back extensors andon and parallel to the direction of themuscle fibers.

The electrical output from the mus-cle was amplified with a biopotentialamplifier with a gain of 5000 and fre-quency response that was flat from DCto 1000 Hz (Biopac Inc, Goletta, CA).The amplified EMG was digitized with a16-bit analog-to-digital converter andsampled at a frequency of 500 samples/s(Biopac Inc.). The software used to ana-lyze the EMG was Acknowledge 3.8.3software on an MP100 system (BiopacInc.). The amplitude of the EMG wasanalyzed by integrating the digitizeddata.

ExerciseStandard Floor CrunchesThese were accomplished on the floorwith the knees bent at 90 degrees, thehips at 45 degrees, the chest raised by 35degrees of flexion, and the handscrossed on the chest. Standard floorcrunches were accomplished in the for-ward flexion direction and with left andright flexion to a rotation of 30 degreesto exercise the oblique muscles.

Swiss Ball CrunchesA second set of exercises was accom-plished on a Swiss ball. Here the subjectsat on a ball (Figure 1). The subject thensat back with the hip extending to 0degrees and then flexed to 90 degrees.The knees were at 90 degrees and the

Table 2. Comparing Data on the 3 Principal Muscle Groups for Floor and Swiss Ball Crunches.Each Point is the Corresponding P Value

Rectus Right Oblique Left ObliqueForward 0.03 0.53 0.65Right 0.03 0.05 0.02Left 0.02 0.05 0.05

Figure 9. The results of the level 1 exercise.Muscle activity is shown for all 4 musclegroups examined as the mean ± the standarddeviation for the group.


hands were folded on the chest. Thisexercise was repeated with the subjectsfacing forward and the trunk rotated tothe right and left by 35 degrees.

Mini Stability Ball Exercises Level 1: There were 3 exercises in level1. Subjects sat on the floor with theknees at 90 degrees and the hips initiallyat 110 degrees of flexion. The trunk wasextended for the exercise and the ministability ball (Savvier LP, Santa FeSprings, CA) was placed against the

sacrum (about 16 cm [7 inches]) indiameter placed at the mid sacrum)until, in different exercises, the back wasextended to either 60, 40, or 20 degrees.The back was held in place for 1 secondand then flexed to the initial position(Figure 2). By extending the back, forexample, to 60 degrees from neutral, theangle at the back and hips was increasedfrom 110 degrees of flexion to 60degrees of extension or a total move-ment of 50 degrees. Thus the range ofmotion of the exercise was 50, 70, and 90degrees for the 3 exercises. These exer-cises were repeated with the trunk rotat-ed 35 degrees to the right and left to

Table 3. P Values Comparing the Level 1 Exercise With the Floor Crunches


Figure 10. The muscle activity of the 4 musclegroups examined during level 1 exercise withthe subject flexing forward where the move-ment flexed the back by 70 degrees (upperpanel) and 90 degrees (lower panel) from theinitial starting position. Muscle activity is shownfor all 4 muscle groups examined as themean ± the standard deviation for the group.

Figure 11. The muscle activity of the 4 musclegroups examined during level 1 exercise withthe subject flexing to the right, where themovement flexed the back by 70 degrees(upper panel) and 90 degrees (lower panel)from the initial starting position. Muscle activi-ty is shown for all 4 muscle groups examinedas the mean ± the standard deviation for thegroup.


recruit the transverse abdominis and theoblique muscles. During these exercises,the hands were placed under the kneesfor support.

Level 2: There were 5 exercises in level2. These consisted of first sitting on thefloor with the knees at 90 degrees andthe hips initially at 110 degrees of flex-ion. The hands were held in the air par-allel to the floor. The back was restingagainst the mini stability ball and, theback was extended to 60, 40, or 20degrees in different exercises, held for 1second and then returned back to theinitial starting position (Figure 3). Thisexercise was repeated with the trunkrotated 35 degrees to the right and leftto exercise the transverse abdominis andthe oblique muscles.

The final 2 exercises consisted ofhaving the subject sit on the mini stabili-ty ball with the knees at 90 degrees and

the hips at an angle of 90 degrees withthe legs parallel to the floor. The legswere then alternately extended to touchthe floor with the toes pointed as shownin Figure 4. The hands rested along thefloor with the shoulder at 45 degrees forstability.

Level 3: There were 3 exercises in level3 accomplished to the right and left sideof the body. The first exercise consistedof having the subject sit on the floorwith the ball behind his/her back. Herethe hips were at an angle of 110 degreesand the knees at 75 degrees. The handswere placed behind the head and as 1leg was flexed, the opposite elbowtouched the knee as shown in Figure 5.This was performed on 1 side of thebody and then the other. The secondexercise was similar to the first but themovement was performed in rapid repe-titions.

The third exercise, the wiper exer-cise (Figure 6), set consisted of placingthe hands on the floor with the shoul-ders abducted 45 degrees to the side ofthe body for support and placing theball between the knees with the hips and

Figure 12. The muscle activity of the 4 musclegroups examined during level 1 exercise withthe subject flexing to the left, where themovement flexed the back by 70 degrees(upper panel) and 90 degrees (lower panel)from the initial starting position. Muscle activi-ty is shown for all 4 muscle groups examinedas the mean ± the standard deviation for thegroup.

Figure 13. The work of the 4 muscles in thelevel 1 exercise with the flexion in the forwarddirection with flexion to 50, 70, and 90degrees from the starting position. Muscleactivity is shown for all 4 muscle groups exam-ined as the mean ± the standard deviationfor the group.


knees at 90 degrees. The hips were slow-ly rotated to the right and then the leftthrough full range of motion.

Statistical and Data AnalysisStatistical analysis involved the calcula-tions of means, standard deviations, andpaired and nonpaired t-tests. The level ofsignificance was P<0.05. To analyzework, a work index was calculated. Thisindex used the average EMG and timeto calculate a work index accomplishedin 1 exercise cycle. The work of the 4muscles was added together to give thetotal work for the muscles examined.Work was expressed as average workper second of exercise to make the exer-cises comparable because different exer-

cises required different time periods toaccomplish.

PROCEDURESAll subjects in the study were engagedin the same procedures. First, the EMGwas assessed during a maximum effortfor the 4 muscle groups examined. Thesewere the right oblique, left oblique, rec-tus abdominis, and back extensor mus-cles. Next, abdominal floor crunches inthe forward, left, and right directionswere performed. This was followed bySwiss ball crunches. Finally, mini stabilityball exercises were accomplished using avideo for timing. For each exercise,EMG was sampled to assess muscle use.

RESULTSFloor CrunchesThe muscle use for the crunches per-formed on the floor is shown in Figure7. The total work per each second ofexercise (for all 4 muscle groups addedtogether as work per second of exercise)was 54.1± 5.3 units per second for exer-cise facing forward (upper panel), 73.7 ±4.8 units per second facing to the right(middle panel), and an average of 76.2 ±9.3 units per second when facing to theleft (lower panel). The greatest averagemuscle activity facing forward was forthe rectus abdominis muscles averaging23.3 ± 6.8% of maximum muscle activity.The greatest activity of the oblique mus-cles was seen when facing in the direc-tion of the exercise as shown in theother 2 panels. The peak muscle activityfor a 0.5-second period for the rectusabdominus, when exercising in the for-ward direction, was 76.1 ± 26.2% of themuscle, but the peak was short and wasonly at the greatest flexion point of theexercise. The oblique muscles, with thesubject exercising facing the sides forthe same 0.5-second period, showed 45.0± 22.7% of maximum muscle activity forthe right-facing crunches and the rightoblique muscles and 62.3 ± 24.5% of

Figure 14. The results of the first exercise inlevel 2. Here the subjects held their arms par-allel to the floor and did crunches forward(upper panel), to the right (middle panel),and to the left (lower panel). Muscle activity isshown for all 4 muscle groups examined asthe mean ± the standard deviation for thegroup.


maximum muscle activity for crunchesfacing to the left. The average durationof the floor crunches was 1.84 ± 0.52 sec-onds.

Swiss Ball CrunchesThe muscle use for the 3 Swiss ballcrunch exercises is shown in Figure 8.The average work index for the muscles(total work of all 4 muscles per secondof work) was 76.3 ± 11.6 units per sec-ond for exercise facing forward, 117.6 ±13.2 units per second for exercise facingright, and an average of 114.7 ± 15.2units per second when exercising to theleft. The greatest activity while facingforward was seen in the rectus abdomin-is muscles, which averaged 32.0 ± 14.4%of total muscle activity. The greatestactivity of the oblique muscles was seenwhen facing in the direction of the exer-cise. The right and left oblique musclesshowed peak muscle activity of 34.1 ±12.2% and 25.5 ± 9.9% of maximummuscle activity in their respective direc-tions. The peak muscle activity over a0.5-second period at the peak of exerciseduring each crunch for the rectus abdo-minis muscle was 81.0 ± 14.2% in theforward direction, 58.8 ± 29.5% for theright oblique muscle in the right-facingdirection, and 67.6 ± 37.3% for the leftoblique muscle when exercising in theleft-facing direction. The average dura-tion of the crunches was 2.25 ± 0.34 sec-onds.

Swiss Ball Compared with FloorCrunchesThe muscle activity of the 4 muscle

groups examined was significantly high-er during most corresponding exercisefor the Swiss ball compared with thefloor crunches as was the duration of theexercise as shown in Table 2. For theprincipal muscle movers for the 3 exer-cises and the total work done, the differ-ence was significant.

Mini Stability Ball ExercisesLevel 1Fifty degrees of flexion/extension: Theresults of the level 1 exercise for 50degrees of flexion-extension are shownin the 3 panels of Figure 9. The totalwork (work for all 4 muscle groups persecond of exercise) was 25.7 ± 4.5 unitsper second for exercise facing forward,27.4 ± 3.2 units per second exercising tothe right, and an average of 27.8 ± 4.9work units per second of exercise whenfacing to the left. The greatest averagemuscle activity facing forward was seenwith the extensor muscles, which aver-aged 7.9 ± 3.1% of total muscle activity,and the greatest activity of the right andleft oblique muscles was shown whenfacing their respective sides as shown inthe other 2 panels. Muscle use was sig-nificantly less in the level 1 exercise thanthat in the floor crunches as shown forthe 3 key muscles in Table 3. The aver-age duration of the exercise was 5.41 ±1.7 seconds.

The peak muscle activity of the rec-tus abdominis muscle measured over a0.5-second period for the forward-facingexercise was 43.4 ± 9.5% of the maxi-mum muscle activity. The peak activityof the muscle was 49.6 ± 24.1% and 55.3

Table 4. P Values Comparing the Level 2 Exercise at 50 Degrees of Flexion/Extension With FloorCrunches



± 25.9%, respectively, for the right andleft oblique muscles for the correspon-ding side-bending exercises.

Exercise at 70 and 90 degrees of flexionextension: For a basis of comparison,Figures 10 through 12 show the muscleactivity for 70 degrees of movement(upper panels) and 90 degrees of move-ment (lower panels) for exercise accom-plished while facing forward (Figure 10),to the right (Figure 11), and left (Figure12) sides. These figures show the averagemuscle activity of the 4 muscles exam-ined here. Increasing the extent of flex-ion/extension from 50 to 70 and 90degrees for any body position caused anexponential increase in muscle use. Infact, as an illustration, the calculatedmuscle work for the 4 muscle groups isshown in Figure 13. Here the work forthe 4 muscle groups in each second ofexercise increased from 25.7 units persecond for 50 degrees of exercise to 38.8

units per second for 70 degrees of trunkmovement, and 126.2 work units per sec-ond for the exercise facing forward atflexion/extension movement of 90degrees. This increase was significant(ANOVA, P<0.01). Similar increases inwork were seen for the other 2 exercisepositions.

Comparing the muscle use and workfor these exercises at flexion/extensionof 70 and 90 degrees of movement, themuscle use and work were significantlyhigher than that of the abdominalcrunches or the Swiss ball exercises(P<0.01).

Level 2 Exercise 1:Fifty degrees of flexion/extension—Theresults of the first level 2 exercises at 50degrees of flexion/extension are shownin the 3 panels of Figure 14. The total

Figure 15. The muscle activity of the 4 musclegroups examined during level 2 exercise withthe subject flexing forward where the move-ment flexed the back by 70 degrees (upperpanel) and 90 degrees (lower panel) from theinitial starting position. Muscle activity is shownfor all 4 muscle groups examined as themean ± the standard deviation for the group.

Figure 16. The muscle activity of the 4 musclegroups examined during level 2 exercise withthe subject flexing to the right where themovement flexed the back by 70 degrees(upper panel) and 90 degrees (lower panel)from the initial starting position. Muscle activi-ty is shown for all 4 muscle groups examinedas the mean ± the standard deviation for thegroup.


work (for all 4 muscle groups in eachsecond of exercise) was 41.7 ± 4.9 unitsper second for exercise facing forward,47.5 ± 6.2 units per second twisting tothe right, and an average of 42.0 ± 7.1units per second when rotating to theleft. The activity facing forward for therectus abdominis muscles averaged 12.6± 5.1% of total muscle activity, while thegreatest activity of the oblique muscleswas when facing their respective sides asshown in the other 2 panels. The averageduration of the exercise was 6.2 ± 0.68seconds.

The peak muscle activity of the rec-tus abdominis muscle measured over a0.5-second period was 68.4 ± 29.3%when facing forward. The peak activitywas 59.9 ± 32.2% and 68.4 ± 30.7% forthe right and left oblique muscles for thecorresponding side bending exercises,respectively. Compared with the floorcrunches, the muscle activity was signifi-

cantly different for the key muscles ineach exercise as shown in Table 4.

Seventy and 90 degrees of flexion/exten-sion—As described above, when theexercise was extended to 70 and 90degrees of flexion/extension, the exer-cise required considerably more muscleuse and associated work. Figures 15through 17 show muscle activity for 70degrees of flexion extension exercise(upper panels), and 90 degrees of flexionextension exercise (lower panels) forexercise accomplished while facing for-ward (Figure 15), to the right (Figure16), and left (Figure 17) sides. As can beseen here, increasing the extent of flex-ion/extension from 50 to 70 and 90degrees for any body position caused anexponential increase in muscle use. Thisincrease was significant (P<0.01).Similar increases in work were seen forthe other 2 exercise positions.Comparing the muscle use and workaccomplished with either the Swiss ballor floor crunches, the use and work were

Figure 17. The muscle activity of the 4 musclegroups examined during level 2 exercise withthe subject flexing to the left where themovement flexed the back by 70 degrees(upper panel) and 90 degrees (lower panel)from the initial starting position. Muscle activi-ty is shown for all 4 muscle groups examinedas the mean ± the standard deviation for thegroup.

Figure 18. The results of the second exerciseset in level 2. Here the subjects held their legsoff of the floor and touched their toes to thefloor on the right (upper panel) or left (lowerpanel) sides. Each point is the mean of thegroup data ± the standard deviation for thegroup.


significantly higher (P<0.01) for exerciseat 50 and 70 degrees of back flexion/extension.

Exercise 2: The results of the secondexercise in this sequence are shown inthe 2 panels in Figure 18. The total workfor the 4 muscle groups per second ofexercise was 40.8 ± 5.4 units per secondfor exercise touching the right toes and46.2 ± 6.2 units per second for touchingthe left toes. The average muscle activityduring the exercise for the rectus abdo-minis muscle was 16.8 ± 14.3%, and 7.9± 4.5% and 17.7 ± 6.2%, respectively, forthe left and right oblique muscles duringexercise to the left and right sides. Theaverage duration of the exercise was 6.1± 0.7 seconds.

The peak muscle activity of the rec-tus abdominis muscle measured over a0.5-second period for the toe taps was34.2 ± 16.3% when facing forward. Thepeak activity was 31.6 ± 8.9% and 41.2 ±6.8%, respectively, for the right and leftoblique muscles for the correspondingside tapping exercises. Compared withthe floor crunches, the muscle activitywas significantly different for the keymuscles in each exercise as shown inTable 5. For the key muscles, muscleactivity was significantly less for thecrunches on the floor based on per sec-ond of exercise.

Level 3Exercise 1:Right and left cross crunch—The resultsof the first series of exercise are shownin the 2 panels of Figure 19. The totalwork averaged for each second of workfor the 4 muscle groups was 109.1 ± 9.8units per second for exercise touchingthe right toes and 121.9 ± 13.2 units persecond for touching the left toes. Theaverage muscle activity for the rectusabdominis muscle was 37.5 ± 11.2% forthe right and left cross crunch exercises,and 37.4 ± 11.2% and 33.1 ± 12.1%,respectively, for the left and rightoblique muscles during exercise to theleft and right sides. The average durationof the exercise was 5.4 ± 1.6 seconds.

The peak muscle activity of the rec-tus abdominis muscle measured over a0.5-second period was 69.3 ± 24.7% ofthe muscle when facing forward. Thepeak activity was 89.4 ± 15.1% and 57.1± 18.8%, respectively, for the right andleft oblique muscles for the correspon-ding side tapping exercises. Comparedwith the floor crunches, the muscle activ-


Rectus Right Oblique Left ObliqueRight 0.032 0.001 0.008Left 0.041 0.009 0.046

Figure 19. The results of the first part of thelevel 3 exercises. Here the subjects were in theabdominal crunch position with the ballbehind their backs and alternatively touchedtheir elbows to their knees on the right (upperpanel) or left (lower panel) sides. Each point isthe mean of the group data ± the standarddeviation for the group.


ity was significantly different for themuscles in each exercise as shown inTable 6. For the key muscles, muscleactivity was significantly greater thanthe crunches on the floor based on persecond of exercise.

Exercise 2:Left and right crunch pulse exercise—The results of the second exercise areshown in the 2 panels of Figure 20. Thetotal work for the 4 muscle groups persecond of exercise was 220.8 ± 28.4 forpulse exercises to the right and 187.0 ±18.5 for pulse exercises to the left. Theaverage muscle activity for the rectusabdominis muscle was 76.9 ± 27.4% forboth the right and left exercises, and41.7 ± 20.8% and 58.6 ± 14.2%, respec-

tively, for the left and right oblique mus-cles during exercise for the left and rightsides. The average duration of the exer-cise was 5.45 ± 0.98 seconds. The pulseexercise was one fifth of this duration,and for basis of comparison, the datafrom 5 exercises were added togetherand averaged for comparison with theother exercises in this series.

The peak muscle activity of the rec-tus abdominis muscle measured over a0.5-second period was 76.9 ± 14.2%when facing forward. The peak activitywas 53.2 ± 9.2% and 55.2 ± 7.3% for theright and left oblique muscles for thecorresponding side tapping exercises.Compared with the floor crunches, themuscle activity was significantly differ-ent for the key muscles in each exerciseas shown in Table 7. For the key muscles,muscle activity was significantly greaterthan for the crunches on the floor basedon per second of exercise.

Exercise 3:Left and right wiper exercise—Theresults of the first exercises are shown inthe 2 panels of Figure 21. The total workfor the 4 muscle groups was 33.8 ± 3.7units per second for exercise to the rightand 39.4 ± 2.9 units per second for exer-cise to the left. The average muscleactivity during this exercise for the rec-tus abdominis muscle was 8.35 ± 4.5%of the maximum muscle activity, and19.1 ± 11.2% and 15.1 ± 9.1%, respec-tively, of the maximum muscle activityfor the right and left oblique musclesduring exercise to the left and rightsides. The average duration of the exer-cise was 5.6 ± 1.28 seconds.

The peak muscle activity of the rec-tus abdominis muscle measured over a



Figure 20. The results of the second part ofthe level 3 exercises. Here the subjects were inthe abdominal crunch position with the ballbehind their backs and alternatively touchedtheir elbows to their knees on the right (upperpanel) or left (lower panel) sides. Exercisehere was a series of rapid pulses. Each point isthe mean of the group data ± the standarddeviation for the group.


0.5-second period was 34.8 ± 15.4%when facing forward. The peak muscleactivity was 78.9 ± 27.1% and 83.8 ±39.2%, respectively, for the right and leftoblique muscles for the correspondingexercises. Compared with the floorcrunches, the muscle activity was signifi-cantly different for the key muscles ineach exercise as shown in Table 8. Forthe key muscles, muscle activity was sig-nificantly less than that of the cruncheson the floor based on per second ofexercise.

Comparison of the Work for AllExercise with Abdominal Crunches andSwiss Ball CrunchesTable 9 shows a summary of the workfor the various exercises. The extracolumns for some of the exercises inlevel 1 and level 2 are for the extra exer-

cises at 70 and 90 degrees of flexionextension of the trunk. Each numberrepresents the average for all subjects ofthe total work of all 4 muscles examinedper second of exercise. As shown inTable 9, the work per second of time forexercise with the mini stability ball var-ied from less than to many times greaterthan that used during floor crunches.Floor crunches required about twothirds of the work used for the sameexercises on a Swiss ball. Mini ballcrunches showed the greatest work at 90degrees of flexion extension where workwas as high as 2 to 4 times as much workas floor crunches. To show this relation-ship, Table 10 shows the percentage ofwork each exercise required comparedwith a forward facing floor crunch. Thegreatest work was shown with the level3 pulse in the right-facing directionwhere work was 408% of that seen withthe floor crunch.

DISCUSSIONNumerous benefits have been touted foraerobic exercise.1,2 These include reduc-tion in back pain,43 better stabilization ofthe torso muscles for balance andreach,44 reduction in blood lipids,9increased oxygen consumption in recov-ery following exercise,10 and better over-all fitness.45 Although all types ofexercise have good therapeutic bene-fits.46 submaximal intermittent or contin-uous exercise seems to show the greatestbenefits.11,47

One of the most effective types ofexercise is core muscle strengthening.This form of exercise concentrates onthe abdominal and lower back mus-cles.29-31 The advantage of this form ofexercise is that it can reduce lower back



Figure 21. The results of the third part of thelevel 3 exercises. Here the subjects were lyingon their backs on the floor with the ballbetween their knees and rotated their pelvisto the right (upper panel) or left (lower panel)sides. Each point is the mean of the groupdata ± the standard deviation for the group.


pain and reduce back injury by allowingproper alignment of the spinal column.29-32

Furthermore, functional reach can beincreased with good core strengthening,reducing the risk of falling, especially inthe elderly.31

In the present investigation, exercis-ing with a 7-inch mini stability ball wasused as a core muscle exercise program.Three levels of exercise were accom-plished following an exercise video fortiming. These data were compared withexercise with a Swiss ball and floorcrunches.

The work and muscle use duringfloor crunches was significantly less forthe key muscle groups for each exercise

compared with the Swiss ball. Thus thetotal work was greater for any conditionfor the Swiss ball. This is probablybecause of the increase in muscle use ofthe core muscles to stabilize the body tobalance on the Swiss ball compared withfloor exercise32-34 as well as the increasedextension flexion by having the personelevated from the floor when exercising.With floor exercise, the floor providesstability; the Swiss ball removes this sta-bility, which then must be provided bycore muscle use. The Swiss ball providesgreater extension and flexion but theextent of that movement is limited bythe diameter of the ball; the larger thediameter, the less the movement. To gain



Table 9. Total Work of Each of the Exercises Per Second of Exercise Time

Total Work 70 Degrees 90 DegreesCrunch Forward 54.1Crunch Right 73.7Crunch Left 76.2Swiss Forward 76.3Swiss Right 117.6Swiss Left 114.7Level 1 Forward 25.7 38.8 126.2Level 1 Right 27.4 82.1 125.3Level 1 Left 27.8 76.3 144.7Level 2 Forward 41.7 122.7 156.9Level 2 Right 47.5 119.8 188.7Level 2 Left 42.0 132.9 202.6Level 2 Toe Right 40.8Level 2 Toe Left 46.2Level 3 Cross Right 109.1Level 3 Cross Left 121.9Level 3 Pulse Right 220.8Level 3 Pulse Left 187.0Level 3 Wiper Right 33.8Level 3 Wiper Left 39.4


greater extension, the person wouldhave to move off of the ball; this unsup-ported exercise is not recommended bytrainers due to potential injuries becausethe back is not supported in a cant-ilevered position.

With the mini stability ball, thesedifferences were enhanced further whilestill providing support for the lowerback. While the muscle use and work onthe Swiss ball and floor crunches weresimilar to that reported in other stud-ies,32-34 here the increased instabilitycaused more muscle use in the obliquemuscles compared with rectus abdomin-is muscles. This is demonstrated in exer-cises such as the forward facing exercise,where, during a floor crunch, only therectus abdominis is the prime mover.Therefore, because the oblique muscleswere used to stabilize the core while therectus abdominis muscle was contract-

ing, muscle use and total work weregreater. Additionally, the smaller diame-ter ball allowed for greater range ofmotion during the exercise, increasingmuscle use and work. This was especiallytrue of exercise in level 2 at 90 degreesof flexion extension. Here the work wasas high as 2 to 4 times greater than thatof a crunch on the floor.

Although not measured in the study,the mini stability ball’s diameter allowedfor further extension than 90 degrees.Based on the exponential relationship inthis study between extension/flexion andmuscle use, this would further increasethe exercise.

Similarly, the exercises in level 3,especially pulse right and pulse left,were already so difficult that higher lev-els of flexion extension were not stud-ied, but would have lead to higher work.

One interesting observation was

Table 10. The Level of Work of the Various Exercises Compared with Traditional Forward FacingFloor Crunches. Each Number is a Percentage of the Work of a Crunch

Total Work 70 Degrees 90 DegreesCrunch Forward 99.9Crunch Right 136.1Crunch Left 140.8Swiss Forward 141.0Swiss Right 217.4Swiss Left 211.9Level 1 Forward 47.4 71.7 233.3Level 1 Right 50.7 151.8 231.6Level 1 Left 51.3 141.0 267.5Level 2 Forward 77.2 226.8 290.0Level 2 Right 87.9 221.4 348.8Level 2 Left 77.7 245.7 374.5Level 2 Toe Right 75.3Level 2 Toe Left 85.4Level 3 Cross Right 201.7Level 3 Cross Left 225.3Level 3 Pulse Right 408.2Level 3 Pulse Left 345.6Level 3 Wiper Right 62.4Level 3 Wiper Left 72.9


that the mini stability ball created goodback support in spite of its size. InFigures 9 and 10, when exercise was var-ied from 50 to 90 degrees, although therectus abdominis muscle activityincreased dramatically, there was almostno increase in back extensor activity. Infact, there was almost no back extensoractivity seen, therefore allowing the iso-lation of the abdominal muscles whilenot increasing stress on the back mus-cles.

The variety of exercise here allowsan unfit user to select low levels of workor a fit user to select very hard levels ofwork. Additionally, users can select topreferentially work different musclegroups to emphasize exercise on the rec-tus or the oblique muscles. A user withpoor fitness can start easy and thenprogress over a range of 4 fold from thelightest mini stability ball work to thehardest, providing a good progressiveworkout for long-term training.

REFERENCES1. Astrand PO, Rodahl K. Physiology of work

capacity and fatigue. New York, NY: McGrawHill; 1970.

2. McArdle WD, Katch F, Katch V. ExercisePhysiology. Energy, Nutrition and HumanPerformance. 5th ed. Baltimore, MD:Lipincott, Williams and Wilkins; 2001.

3. Petrofsky JS, Morris A, Bonacci J, et al.Aerobic training on a portable abdominalmachine. J Appl Res Clin Exp Ther.2003;3:402-415.

4. Petrofsky JS, Morris A, Bonacci J, et al.Comparison between an abdominal curl withtimes curls on a portable abdominal machine.J Appl Res Clin Exp Ther. 2003;3:394-401.

5. Petrofsky JS, Hill J, Jorritsma R, et al. Muscleuse during low impact aerobic exercise (glid-ing) compared to conventional weight liftingequipment. J Appl Res Clin Exp Ther.2005;5:136-148.

6. Sternlicht E, Rugg S, Fujii LL, Tomomitsu KF,Seki MM. Electromyographic comparison ofa stability ball crunch with a traditionalcrunch. J Strength Cond Res. 2007;21:506-509.

7. Deyo RA. Conservative therapy for low backpain. Distinguishing useful from useless ther-

apy. JAMA. 1983;250:1057-1062.

8. Donchin M, Woolf O, Kaplan L, Flooman Y.Secondary prevention of low-back pain. Aclinical trial. Spine. 1990;15:1317-1320.

9. Herd SL, Kiens B, Boobis LH, Hardman AE.Moderate exercise, postprandial lipemia, andskeletal muscle lipoprotein lipase activity.Metabolism. 2001;50:756-762.

10. Burnley M, Doust JH, Carter H, Jones AM.Effects of prior exercise and recovery dura-tion on oxygen uptake kinetics during heavyexercise in humans. Exp Physiol. 2001;86:417-425.

11. McConnel TR, Laubach CA III. Elderly car-diac rehabilitation patients show greaterimprovements in ventilation at submaximallevels of exercise. Am J Geriatr Cardiol.1996;5:15-23.

12. Gallagher KM, Fadel PJ, Stromstad M, et al.Effects of exercise pressor reflex activationon carotid baroreflex function during exercisein humans. J Physiol. 2001;15;533(Pt 3):871-880.

13. Agostoni P, Bussotti M. Exhaled nitric oxideand exercise performance in heart failure.Arch Physiol Biochem. 2003;111:293-296.

14. Millet GY, Lepers R. Alterations of neuro-muscular function after prolonged running,cycling and skiing exercises. Sports Med.2004;34:105-116.

15. Stensdotter AK, Hodges PW, Mellor R, et al.Quadriceps activation in closed and in openkinetic chain exercise. Med Sci Sports Exer.2003;35:2043-2047.

16. Warden SJ, Wajswelner H, Bennell KL.Comparison of abshaper and conventionallyperformed abdominal exercises using surfaceelectromyography. Med Sci Sports Exer.1999;31:1656-1664.

17. Sands WA, McNeal JR. A kinematic compari-son of four abdominal training devices and atraditional abdominal crunch. J StrengthCond Res. 2002;16:135-141.

18. Thomas TR, Ridder MB. Resistance exerciseprogram effects on abdominal function andphysique. J Sports Med Phys Fitness.1989;29:45-48.

19. Petrofsky JS, Cuneo M, Dial R, et al. Corestrengthening in a geriatric population. JAppl Res Clin Exp Ther. 2005;5:423-433.

20. Schultz IZ, Crook J, Berkowitz J, et al.Predicting return to work after low backinjury using the Psychosocial Risk forOccupational Disability Instrument: a valida-tion study. J Occup Rehabil. 2005;15:365-376.

21. Crill MT, Hostler D. Back strength and flexi-bility of EMS providers in practicing prehos-


pital providers. J Occup Rehabil. 2005;15:105-111.

22. Chibnall JT, Tait RC, Andresen EM, HadlerNM. Race and socioeconomic differences inpost-settlement outcomes for AfricanAmerican and Caucasian Workers’Compensation claimants with low backinjuries. Pain. 2005;114:462-472.

23. Edlich RF, Winters KL, Hudson MA, et al.Prevention of disabling back injuries in nurs-es by the use of mechanical patient lift sys-tems. J Long Term Eff Med Implants.2004;14:521-533.

24. Lind AR, Petrofsky JS. Cardiovascular andrespiratory limitations on muscular fatigueduring lifting tasks. In: Safety in ManualMaterials Handling. Proceedings from theInternational Symposium on Lifting inIndustry, State University of New York atBuffalo. 1978:57-62.

25. Garcy P, Mayer T, Gatchel RJ. Recurrent ornew injury outcomes after return to work inchronic disabling spinal disorders. Tertiaryprevention efficacy of functional restorationtreatment. Spine. 1996;21:952-959.

26. Gluck JV, Oleinick A. Claim rates of compen-sable back injuries by age, gender, occupa-tion, and industry. Do they relate toreturn-to-work experience? Spine.1998;23:1572-1587.

27. Jarvis KB, Phillips RB, Morris EK. Cost percase comparison of back injury claims of chi-ropractic versus medical management forconditions with identical diagnostic codes. JOccup Med. 1991;33:847-852.

28. Wasiak R, McNeely E. Utilization and costsof chiropractic care for work-related lowback injuries: do payment policies make a dif-ference? Spine J. 2006;6:146-153.

29. Petrofsky JS, Bonacci J, Bonilla T, et al. Can aone-week diet and exercise program causesignificant changes in weight, girth and bloodchemistry? J Appl Res Clin Exp Ther.2004;4:369-375.

30. Petrofsky JS, Laymon M, Cuneo M, et al. Abidirectional resistance device for increasingthe strength and tone in upper body coremuscles and chest girth. J Appl Res Clin ExpTher. 2005;5:553-559.

31. Petrofsky JS, Cuneo M, Dial R, et al. Coremuscle strengthening on a portable abdomi-nal machine. J Appl Res Clin Exp Ther.2005;5:460-472.

32. Szasz A, Zimmerman A, Frey E, et al. Anelectromyographical evaluation of the validi-ty of the 2-minute sit-up section of the ArmyPhysical Fitness Test in measuring abdominalstrength and endurance. Mil Med.2002;167:950-953.

33. Behm DG, Anderson K, Curnew RS. Muscleforce and activation under stable and unsta-ble conditions. J Strength Cond Res.2002;16:416-422.

34. Marshall PW, Murphy BA. Core stabilityexercises on and off a Swiss ball. Arch PhysMed Rehabil. 2005;86:242-249.

35. Marshall P, Murphy B. Changes in muscleactivity and perceived exertion during exer-cises performed on a Swiss ball. Appl PhysiolNutr Metab. 2006;31:376-383.

36. Bigland B, Lippold O. The relation betweenforce, velocity and integrated EMG. JPhysiol. 1954;123:214-224.

37. Petrofsky JS. Frequency and amplitude analy-sis of the EMG during exercise on the bicycleergometer. Eur J Appl Physiol. 1979;41:1-15.

38. Petrofsky JS. Computer analysis of the surfaceEMG during isometric exercise. Comp BiolMed. 1980;10:83-95.

39. Petrofsky JS. Quantification through the sur-face EMG of muscle fatigue and recoveryduring successive isometric contractions.Aviat Space Environ Med. 1981;52:545-550.

40. Petrofsky JS, Dahms T, Lind AR. Power spec-trum of the EMG during static exercise.Physiologist. 1975;18:350.

41. Petrofsky JS, Lind AR. Frequency analysis ofthe surface EMG during sustained isometriccontractions exercise. Eur J Appl Physiol.1980;43:173-182.

42. Lind AR, Petrofsky JS. Isometric tensionfrom rotary stimulation of fast and slow catmuscles. Muscle Nerve. 1978;1:213-218.

43. Plamondon A, Trimble K, Lariviere C,Desjardins P. Back muscle fatigue duringintermittent prone back extension exercise.Scand J Med Sci Sports. 2004;14:221-230.

44. Kavcic N, Grenier S, McGill SM. Determiningthe stabilizing role of individual torso musclesduring rehabilitation exercises. Spine.2004;29:1254-1265.

45. Bouchard C, Rankinen T. Individual differ-ences in response to regular physical activity.Med Sci Sports Exer. 2001;33(6 suppl):S446-S451.

46. Howley ET. Type of activity: resistance, aero-bic and leisure versus occupational physicalactivity. Med Sci Sports Exer. 2001;33(6suppl):S364-S369.

47. Fagard R, Thijs L, Amery A. Age and thehemodynamic response to posture and exer-cise. Am J Geriatr Cardiol. 1993;2:23-40

Core Muscle Activity During Exercise on a Mini Stability Ball

Documents