Search tips
Search criteria 


Logo of jathtrainLink to Publisher's site
J Athl Train. 2006; 41(3): 280–285.
PMCID: PMC1569560

Flexible Foil Exercise and Shoulder Internal and External Rotation Strength


Context: The efficacy of exercise using a flexible foil to increase strength in the shoulder rotator muscles is unknown.

Objective: To compare the effects of exercises using a flexible foil (Bodyblade) with exercises using elastic bands on shoulder internal and external rotator muscle strength.

Design: We used a randomized, controlled pretest-posttest design.

Setting: Laboratory.

Patients or Other Participants: Forty young adults with no previous shoulder injury.

Intervention(s): Experimental subjects participated in an 8-week program of internal and external rotation exercises consisting of 3 sessions per week.

Main Outcome Measure(s): Strength was tested by internal and external rotation isometric and isokinetic strength tests at 60°·s −1 and 120°·s −1.

Results: The group exercising with elastic bands had a greater pretest-posttest increase in maximal internal rotation isometric torque at 10° of internal rotation and greater maximal external rotation isometric torque at 65° of external rotation than the control group. The group using a flexible foil did not show an increase in strength significantly different from either the control or elastic band groups. We found no other statistically significant differences.

Conclusions: Our results do not support the use of a flexible foil for strength gains in shoulder internal and external rotation in the asymptomatic young adult population.

Keywords: rotator cuff, stability, torque, oscillating forces, elastic band, Bodyblade

Many techniques are used to provide resistance for strengthening the shoulder muscles during the middle and advanced stages of rehabilitation, including weights, elastic tubing, isokinetic machines, weighted balls, and aquatic exercise. 1–4 Another technique proposed to strengthen shoulder muscles is shaking a weighted flexible foil, commonly known as a Bodyblade (Hymanson, Inc, Playa Del Rey, CA). To use the device, the exerciser holds the foil by the centrally located handgrip and produces and controls oscillations of the foil, encountering forces of up to 151 N and producing up to 270 contractions per minute. 5 This device is recommended for postoperative shoulder strengthening, rotator cuff muscle strengthening, and trunk-stabilizing exercises. 6 7 This type of exercise is also recommended for shoulder musculature improvement among uninjured populations. 5 The Bodyblade has been purported to enhance overall musculature function, including stability, power, strength, and endurance; reduce shoulder instability; and improve balance and coordination. 5

Although the Bodyblade has been on the market for more than 10 years, the authors of only 1 published study 8 identified its efficacy, observing enhanced shoulder proprioception after 4 weeks of exercise. We located no published studies that described the strengthening effects of using this device and none that compared the effects of flexible foil exercise with those of a more traditional exercise program. Therefore, our purpose was to compare strength gains after a program of exercises using a Bodyblade with gains achieved following a more traditional shoulder exercise program using an elastic band.


Experimental Design

We used a randomized, controlled, repeated-measures design. The testing sessions occurred before (pretest) and after (posttest) the 8-week training period.


Forty asymptomatic young adult volunteers (13 men, 27 women) completed the program and test sessions (Table 1). Subjects who had previous or existing shoulder injuries that warranted professional intervention were excluded. Before participating, subjects were given a description of this study, and all signed an informed consent. Our study was approved by the institutional review board of our university.

Table 1
Subject Characteristics


Isometric, concentric, and eccentric muscle strength of the internal and external shoulder rotators was measured by a KinCom isokinetic dynamometer (model 500H; Chattanooga Group, Inc, Chattanooga, TN). Reliability and validity of the KinCom have previously been established. 9 Two lengths of Bodyblades were used: the Bodyblade Pro model and the Bodyblade Classic Black model. Four types (thin/yellow, medium/red, heavy/green, and extra heavy/blue resistance) of TheraBand (The Hygenic Corp, Akron, OH) provided the elastic band resistance. To measure and estimate the prestretch force in each type of TheraBand and at each training position, we used a Nicholas Manual Muscle Tester (model 01160; Lafayette Instrument, Lafayette, IN).



Subjects stretched their left shoulder internal and external rotator muscles for approximately 5 minutes before testing on the KinCom. They were positioned sitting comfortably on the KinCom chair with stabilization straps applied to the thigh, waist, and chest. In accordance with the manufacturer's recommendations, the KinCom head was tilted 55° and the subjects positioned in 30° of left shoulder abduction, neutral flexion, and neutral rotation. 10 The elbow was positioned in 90° of flexion and supported in a V-shaped pad attached to the axis of rotation.

The subject pushed against a padded force transducer attached to the actuator arm of the KinCom machine. The measurement sequence of internal and external shoulder rotator strength was randomly determined. Subjects warmed up with 3 repetitions of submaximal isometric effort in 10° of shoulder internal rotation and then 65° of shoulder external rotation. After a minute of rest, subjects performed 3 repetitions of 5-second maximal isometric contractions at each position (10° of internal rotation and 65° of external rotation). Isometric strength of the other rotation was subsequently measured in the same manner. After 2 minutes of rest, subjects performed isokinetic warmup and strength testing between 10° of internal rotation and 65° of external rotation: 5 repetitions of submaximal warmup efforts at 60°·s −1 and, after 30 seconds of rest, 5 repetitions of maximal isokinetic concentric and eccentric contractions at 60°·s −1. After 1 minute of rest, subjects performed 5 submaximal warmup repetitions at 120°·s −1. After 30 seconds of rest, subjects performed 5 maximal isokinetic shoulder concentric and eccentric contractions at 120°·s −1. After 1 minute of rest, subjects performed isokinetic shoulder strength testing at 60°·s −1 and 120°·s −1 in the other rotation using the same procedures.


We randomly assigned each subject to the elastic band, Bodyblade, or control group. Subjects in the elastic band group performed resisted shoulder internal and external rotation exercises with repetitions, preload, and elastic band color progressed as described in Table 2. A towel roll was held between the elbow and body in order to keep the shoulder in neutral abduction. With 1 end of the elastic band attached to the wall at elbow level and the elbow flexed to 90°, each subject performed resisted internal and external rotations from about 80° of shoulder external rotation to about 80° of internal rotation. The intensity was increased at each session within a week by positioning the subject farther away from the fixed end, stretching the elastic, and increasing the preload (force in the band at the start of the repetition). If the subject felt more than moderate delayed-onset muscle soreness (verbal report) during the next exercise day, preload was not increased. Figure 1 shows the positioning for exercising the internal rotators. The external rotators were exercised by having the subject face the other direction.

Figure 1
Elastic band exercise of internal rotation
Table 2
Exercise Protocols for Elastic Band and Bodyblade Groups*

Subjects in the Bodyblade group exercised the shoulder internal and external rotators as described in Table 2. The vertically oriented Bodyblade was oscillated horizontally using small-amplitude internal and external shoulder rotation motions with the shoulder in neutral abduction and about 10° of internal rotation. This was repeated with the shoulder in about 65° of external rotation (Figure 2) and again with the shoulder in neutral shoulder rotation. Total exercise duration per exercise started at 1 minute in accordance with the manufacturer's recommendations. 5

Figure 2
Flexible foil exercise at 65° of external rotation

Subjects in the control group did not participate in any internal and external shoulder rotational exercises over the course of this study.


All subjects participated in the posttesting session 8 weeks after the pretesting session. At least 1 day but no more than 7 days of rest separated the last training day from the posttesting session. The posttesting procedures were exactly the same as those used during the pretesting session; the sequence of internal and external rotation strength testing in the posttesting session was the same as that used for the pretesting session for each subject.

Statistical Analysis

Peak isometric torque was averaged across the 3 repetitions for each isometric test condition. Peak isokinetic torque was averaged across the 5 repetitions for each phase (concentric and eccentric) of each test at 60°·s −1 and 120°·s −1. The statistical significance of differences in each dependent measure was tested by a 2 × 3 factorial repeated-measures analysis of variance. This design evaluated the differences within tests (pretest and posttest), groups (elastic band, Bodyblade, and control), and the group × test interaction. Because of the unknown training specificity of exercise with a flexible foil, we tested 6 variables at each rotation strength:

  1. Maximal isometric torque at 10° of internal rotation (short muscle length for internal rotators, long muscle length for external rotators).
  2. Maximal isometric torque at 65° of external rotation (long muscle length for internal rotators, short muscle length for external rotators).
  3. Peak concentric torque at 60°·s −1 (slow-speed concentric torque).
  4. Peak eccentric torque at 60°·s −1 (slow-speed eccentric torque).
  5. Peak concentric torque at 120°·s −1 (faster-speed concentric torque).
  6. Peak eccentric torque at 120°·s −1 (faster-speed eccentric torque).

We set the level of significance for the study a priori at alpha = .05. The analysis of interest in our study was the group × test interaction. Main effects of group (with test scores combined) and of test (with group scores combined) were not of interest and are neither presented nor discussed. When we obtained a significant interaction, we compared groups using a priori planned independent t tests of the pretest-posttest difference scores. Because multiple t tests were used, we adjusted the level of significance of the independent t tests a priori to P < .016 (Bonferroni correction = 0.05/3). Statistical analyses were performed using SPSS (version 11.0; SPSS Inc, Chicago, IL).


Forty subjects completed the study, and 7 subjects did not complete the study. Two subjects lost contact immediately after the pretesting session and did not complete any of the exercise sessions. One subject in the Bodyblade group could not complete the study because of time constraints of changed employment. One subject in the elastic band group experienced prolonged shoulder soreness after the seventh week of the training period and discontinued participation. One additional subject in the same group could not participate in the posttest session within the time allowed because of illness. These subjects were not included in any of the analyses. The attendance rate at the exercise sessions was 86.5% for the elastic band group and 90.5% for the Bodyblade group. If a subject missed 3 exercise sessions in a row, that subject was excluded from the analysis. One subject from each experimental group was so eliminated.

Internal Rotation Strength

A significant group × test interaction occurred only for maximal internal rotation isometric torque at 10° of internal rotation (F 2,37 = 5.377, P < .009, effect size = .54, power = 80%). The elastic band group's strength in this position increased 14% (3.0 ± 1.4 Nm), which was statistically more than the control group's increase of 2% (0.4 ± 2.3 Nm) ( P < .003). The increase in strength in the Bodyblade group of 5% (1.3 ± 2.1 Nm) was not significantly different from that of either of the other 2 groups (Table 3).

Table 3
Internal Rotation Strength (Mean ± SD)

External Rotation Strength

A significant group × test interaction occurred only for maximal external rotation isometric torque at 65° of external rotation (F 2,37 = 4.678, P < .015, effect size = .50, power = 74%) and peak external rotation concentric torque at 120°·s −1 (F 2,37 = 3.272, P < .049, effect size = .42, power = 55%). The elastic band group's isometric strength at 65° of external rotation increased 34% (3.3 ± 3.0 Nm), which was statistically more than the control group's decrease of 3% (−0.4 ± 2.0 Nm, P < .001). The increase in isometric strength in the Bodyblade group of 8% (1.0 ± 4.0 Nm) was not significantly different from either of the other groups (Table 4). The planned t tests indicated no difference between the groups in the increase in peak external rotation concentric torque at 120°·s −1.

Table 4
External Rotation Strength (Mean ± SD)


Eight weeks of flexible foil exercise, progressively increased in intensity, did not improve shoulder internal or external rotator strength. This result was unexpected, because the motions and positions of the exercises were specifically designed to target the internal and external rotator muscles. Additionally, the intensity, duration, and frequency of the exercises were selected to follow the guidelines recommended by the manufacturer. 5 However, we did not have any objective way of measuring or ensuring maintenance of intensity during the exercise. To enhance intensity in our study, we increased the oscillation amplitude for each session of the week by instructing the subjects to increase the shaking intensity from mild to hard and finally to maximum. At the fifth week, the Classic model was discontinued and the more difficult Pro model was used. Most of the subjects in the Bodyblade group reported a “hard” or “tired” feeling in their shoulder muscles in the third session of each week; this was especially pronounced during the third, fourth, seventh, and eighth weeks. We cannot determine if these reports are attributable to intensity (strengthening) or duration (endurance) stimuli. We could find no published studies on the strengthening effects of flexible foil exercise. The author of 1 abstract 11 described no increase in strength after a 10-week training program, supporting our results.

The oscillating nature of flexible foil exercise is unusual. As a result, we were uncertain what mode of muscle strength (concentric, eccentric, or isometric) would be most affected by this type of exercise. Although the test procedures were intended to capture these types of muscle strength, we did not observe any strength increase in the Bodyblade group over the control and elastic band groups. We found only 1 group 12 that published an abstract comparing exercise with the Bodyblade with other types of strengthening devices. They observed greater electromyographic activity during exercise with a Bodyblade than with either cuff weights or elastic bands. 12 It is possible that exercise with a Bodyblade could increase strength in a type of motion we did not test (eg, short-amplitude dynamic motion) or could increase endurance, and these possibilities should be investigated.

The manufacturer claimed that exercise with a Bodyblade also enhances stabilization, endurance, power, and balance/coordination. 5 We could find only 1 research article that addressed these claims. 8 In a study of 40 collegiate athletes (25 males and 15 females, mean age = 19.28 years), Schulte and Warner 8 observed a statistically significant increase in shoulder proprioception of 1.5° after 4 weeks of flexible foil exercise. However, although statistically significant, the clinical effect of this magnitude of change is unclear.

Our observation of the elastic band group's increase in isometric strength could be explained by the stretch of the elastic band, which provided maximal resistance at 10° of internal rotation for the internal rotators and at 65°of external rotation for the external rotators. Velocity specificity of training may have also played a role, as the elastic band was maximally stretched at the midrepetition turning point, when the working muscles were exhibiting a slow dynamic or an isometric contraction. The slow dynamic or isometric contraction at positions with maximal resistance force and muscle force could contribute to this group's increase in isometric force measurement but not during the faster isokinetic testing.

Concentric external rotation strength at 120°·s −1 showed a statistically significant time × group interaction ( P < .049), but the independent t tests failed to identify a statistically significant group difference in the pretest-posttest difference scores in any variable. The largest difference in pretest-posttest increases in the peak external rotation concentric torque at 120°·s −1 was between the Bodyblade group (2.8 ± 2.2 Nm) and the elastic band group (1.2 ± 3.6 Nm), but the P value of this comparison (.023) was above the Bonferroni-corrected critical value. It is possible that the significant time × group interaction term was caused by a comparison that was not of interest in this study—for example, between the pretest values of one group and the posttest values of another.

Our study had several limitations. The subjects in the elastic band group had fewer exercise workouts per week than those in the other groups, and they experienced more shoulder soreness during the training period, especially after the fourth week. This factor might have contributed to the lower attendance rate of this group, and the assumed high relative training load that led to this soreness might partially explain why this group gained more strength. The control group had greater height, weight, and strength values because this group had relatively more male subjects than the other groups. Unfortunately, these factors were not balanced during random assignment of groups but could be a confounding factor in our study. Finally, although subjects did not start or change their regular exercise practice during the period of the study, many subjects continued their own exercise routines, including resistance training. However, no subjects had specific shoulder internal rotation and external rotation muscular strengthening exercises outside of our study.


This study was partially supported by the Graduate Programs Fund, College of Human Science and Services, University of Rhode Island, Kingston (D.S.).


  • Baechle TR. Essentials of Strength Training and Conditioning. Champaign, IL: Human Kinetics; 1994:393.
  • Williams GR, Kelley M. Management of rotator cuff and impingement injuries in the athlete. J Athl Train. 2000;35:300–315. [PMC free article] [PubMed]
  • Newton RU, Jones J, Kraemer WJ, Wardle H. Strength and power training of Australian Olympic swimmers. J Strength Cond Res. 2002;24:7–15.
  • Treiber FA, Lott J, Duncan J, Slavens G, Davis H. Effects of Theraband and lightweight dumbbell training of shoulder rotation torque and serve performance in college tennis players. Am J Sports Med. 1998;26:510–515. [PubMed]
  • Hymanson B. Bodyblade. Available at: Accessed January 10, 2006 .
  • Blackburn TA, Guido JA. Rehabilitation after ligamentous and labral surgery of the shoulder: guiding concepts. J Athl Train. 2000;35:373–381. [PMC free article] [PubMed]
  • Leggin BG, Kelley MJ. Rehabilitation of the shoulder following rotator cuff surgery. Univ Pa Orthop J. 2000;13:10–17.
  • Schulte RA, Warner C. Oscillatory devices accelerate proprioception training. Biomechanics. May 2001:85–91 .
  • Farrell M, Richards JG. Analysis of the reliability and validity of the kinetic communicator exercise device. Med Sci Sports Exerc. 1986;18:44–49. [PubMed]
  • Kin-Com Clinical Manual. Hixson, TN: Chattanooga Group Inc; 1992.
  • Lake DA. The effects of a Bodyblade training protocol on shoulder strength and throwing velocity. J Orthop Sports Phys Ther. 2002;32:A-51–A-52.
  • Signorile JF, Lister JL, Rossi GD. Electromyographic analysis of scapular stabilizers using Bodyblade, cuff weights and Theraband resistance [abstract] Med Sci Sports Exerc. 2005;37:S147. et al.

Articles from Journal of Athletic Training are provided here courtesy of National Athletic Trainers Association