The normal shoulder
during freestyle
swimming An electromyographic and twelve muscles
cinematographic analysis
of
MARILYN PINK,* MS, PT, JACQUELIN PERRY, MD, ANTHONY BROWNE, MD, MARY LYNN SCOVAZZO, MD, AND JOHN KERRIGAN, MSE From the Biomechanics
Laboratory, Centinela Hospital Medical Center, Inglewood, California
ABSTRACT
demanding
such stress, the swimmer’s shoulder will show
signs of wear and tear. This is confirmed in the literature with shoulder problems being reported in 42% to 67% of competitive swimmers. 1,7,15 To understand the mechanism of injury, and thereby prevent injury, it is important to understand how the shoulder functions during swimming. The biomechanical and electromyographical analysis of swimming has been hampered by unique instrumentation problems such as the water-electrode interface, harnessing, and integration/quantification of the signal. 2-4,6,8,9, 12, 13 There are no published studies that use the advanced technology of fine wire electrodes and telemetered EMG that has been integrated and quantified to analyze the normal shoulder while swimming. After developing the appropriate instrumentation, this study was designed. The purpose of the study is to describe the muscle activity patterns in 12 muscles of the normal shoulder during freestyle swimming.
The shoulder in swimming is subjected to multiple factors that can lead to a high injury rate. To prevent injury, one must understand the biomechanics of swimming. This paper describes the electromyographic and cinematographic findings of 12 shoulder muscles in competitive swimmers without shoulder pain. The results show the three heads of the deltoid and the supraspinatus functioning in synchrony to place the arm at hand entry and exit, the rhomboids and upper trapezius to position the scapula for the arm, the pectoralis major and latissimus dorsi to propel the body, the subscapularis and serratus anterior as muscles with constant muscle activity, the teres minor functioning with the pectoralis major, and the infraspinatus active only to externally rotate the arm at midrecovery. This information is important to design optimal preventative and rehabilitative exercise programs.
MATERIALS AND METHODS One hundred million Americans are swimmers.’ This makes swimming the most popular participation sport in the United States. All ages are represented and people swim for recreation, for competition, and for fitness. Competitive swimmers often practice most of the year, 6 to 7 days a week, and cover 8,000 to 20,000 meters a day.&dquo; This equates to approximately 18,000 shoulder revolutions per week. Thus, the shoulder is subject to strain caused by the high repetition rate, the extremes of range of motion, and the force required for propulsion. As with all sports
Twenty collegiate and masters level competitive swimmers volunteered for this study. All swimmers completed a questionnaire about any episodes of shoulder pain, and underwent a physical examination to check for signs of shoulder instability, apprehension, or impingement. Any subject currently reporting pain was excluded from this population. Many swimmers had past episodes of pain, but were now asymptomatic and considered to be normal. Five of the swimmers volunteered to be tested for both the Group A muscles and the Group B muscles. Thus, the total data sample included 25 shoulders studied in 20 swimmers. They had an average of 9 years of competitive swimming experience and were currently training 2500 to 4000 yards a day,
*
Address correspondence and repnnt requests to Manlyn Pmk, MS, PT, Biomechanics Laboratory, Centinela Hospital Medical Center, 555 East Hardy Street, Inglewood, CA 90301.
569
570
3 to 5 days a week. Fifty-two percent of the reported the freestyle stroke as their best stroke.
swimmers The mean age of the swimmers was 39 years. There were 15 male swimmers and 5 female swimmers.
Procedure
Twelve shoulder muscles were selected for this study, and divided into two groups of six each. Group A muscles included the anterior, middle, and posterior deltoids, serratus anterior, upper trapezius, and rhomboid major. Group B muscles included the subscapularis, supraspinatus, infraspinatus, teres minor, latissimus dorsi, and pectoralis major. Group A muscles were monitored in 16 shoulders, and Group B muscles in 9 shoulders. Testing was done in one of two pools that were equipped with underwater windows. Motion was recorded from the lateral projection using two 16 mm high-speed motion picture cameras operating at 100 frames per second. The underwater view filmed the pull-through phase of the stroke cycle. The surface view filmed the recovery phase of the stroke cycle. Marks were electronically placed on the film and EMG data to allow for synchronization. Recording of the EMG signal was done using the Bas-
without placing undue tension on the leads or interfering with the stroke style. The EMG data were displayed on the oscilloscope, bandpass filtered at 100 to 1000 Hz, and recorded on a multichannel instrument recorder (Model 3968A, Hewlett Packard, Palo Alto, CA) for later retrieval and review. Prior to recording the swim data, the resting EMG was recorded while the swimmers did a prone float. Next, a maximal isometric manual muscle test was done in the water for each muscle. The resting and maximal muscle test recordings were repeated after the swims to assure that damp-
majian single-needle technique.’ Following proper skin preparation and isolation of the muscle, dual 50 micron insulated wires with 2 to 3 mm bared tips were inserted into the muscle using a 25 gauge needle as a cannula. Accuracy of placement was determined on dry land by manual muscle examination and signal display on an oscilloscope. The wires from each muscle were attached to leads that insulated and connected to a ground plate taped to the subject’s back. Finally, waterproof transparent dressings (Bioclusive, Johnson & Johnson, New Brunswick, NJ) were placed over all the insertion points and wires to secure the system. Taping was done in a manner that allowed unencumbered movement of the shoulder. Signals from the leads were transmitted using an FM-FM were
telemetry system (Model 42000-A, Bio-Sentry Telemetry, Torrance, CA). The battery-operated FM transmitters were secured within an empty 2.5 gallon plastic water jug and the wires were carefully sealed around the site of wire entry. The container was secured to the subject’s waist. This enabled the transmitter to float and trail behind the swimmer
.............-
Figure 1.
Phases of the
freestyle swimming stroke cycle.
Figure 2. Muscle activity firing patterns in the deltoids.
571
ening of the signal due
to any water
leakage had
not
oc-
when the humerus
was
perpendicular
to the water surface
curred.
(four intervals).
The subject started swimming at a moderate pace at approximately 20 yards away from the end of the pool, parallel to the underwater window. Data for two to four stroke cycles were obtained. The subject was asked to swim three trials of the freestyle stroke. The EMG data were converted from analog to digital signals by computer at a sampling rate of 2500 Hz, and were quantitated by computer integration. After excluding the noise identified by the resting recording, the peak 1 second EMG signal during a manual muscle test (MMT) was selected as a normalizing value (100%). Activity patterns were assessed every 20 msec and expressed as a percentage of the normalization base. These activity patterns were synchronized with the film to obtain percent muscle activity values at separate phases of the subjects’ motion.
4. Late recovery: beginning at the completion of early recovery and ending at hand entry (two intervals). The data were analyzed to determine whether there was any significant difference due to hand dominance (right or
Data
were
calculated.
analysis
was divided into 25 equal time intervals. The film was synchronized with the EMG to discern four phases. The four phases and the number of intervals per phase were as follows (Fig. 1). 1. Early pull-through: beginning with the hand entry into the water and ending when the humerus was perpendicular to the axis of the torso (12 intervals). 2. Late pull-through: beginning at the completion of early pull-through and ending as the hand left the water
The freestyle stroke cycle
(seven intervals). 3.
left) or breathing side (right, left, or bilateral). An independent t-test was used for the hand dominance question, and an analysis of variance for the breathing side question (P < 0.05). There were no significant differences in either of these factors, thus the data for all normals were able to be grouped. The data were assessed for normality. Outlying data were checked to determine whether they were artifact or noise, and the decision to keep or delete the outlying data was made accordingly. The percent MMT for each subdivision of each phase was then averaged and standard deviations
Early
recovery:
beginning
at hand exit and
ending
RESULTS Deltoids The three heads of the deltoid fired sequentially at the end of pull-through. Peak activity was first displayed by the posterior deltoid, followed by the middle deltoid, then the anterior deltoid. They all reached peaks of activity between 70% and 76% MMT. The activity dropped in recovery. The anterior and middle deltoids demonstrated a second peak of activity at hand entry (45% to 51% MMT). This was fol-
TABLE 1 Muscle
activity at
25
points during the freestyle stroke in normal shoulders
572
by low level activity until they repeated the peak activity at the end of pull-through (Fig. 2, Table 1). lowed
Rotator cuff The four rotator cuff muscles each demonstrated a different activity pattern. This confirmed that each muscle has a discrete function. The supraspinatus activity was similar to that of the anterior and middle deltoids. There were two distinct periods of activity. The muscle was active at hand entry (54% MMT ± 29), and then dropped to lower levels (6% to 19% MMT) for middle pull-through. The second distinct period of activity was seen at the end of pull-through (66% MMT ± 12) and early recovery (74% MMT ± 37). During middle and late recovery, the muscle activity decreased slightly until hand entry occurred once again (Fig. 3, Table 1). The infraspinatus had its highest level of muscle activity at midrecovery (34% MMT ± 34). Otherwise, there were relatively low levels of participation during pull-through (below 16% MMT) (Fig. 3, Table 1). The teres minor demonstrated a gradual rise in muscle activity in pull-through, up to 57% MMT at middle pullthrough. Activity then diminished and stayed between 16% MMT and 19% MMT in recovery (Fig. 3, Table 1). The subscapularis was active throughout the stroke, with the intensity range between 26% and 71% MMT. Late pullthrough and early recovery each had peaks of activity (64% and 71% MMT, respectively). The lowest level of activity in the subscapularis was during early pull-through (26% to 49% MMT) (Fig. 3, Table 1).
Scapular muscles The rhomboids and the upper trapezius demonstrated a similar pattern of muscle activity in the normal shoulders. They both had two peaks of muscle activity: one at hand entry and the other at hand exit. The serratus anterior had a pattern distinct from these other two scapular muscles. The rhomboids revealed sharp activation at hand entry (49% MMT ± 23). It then dropped to a range of 9% to 13% MMT. Once the humerus passed the point where it was perpendicular to the trunk, the rhomboids gradually became more active. By early recovery it reached the peak activation of 76% MMT ± 47. Activity then decreased until hand entry
began once again (Fig. 4, Table 1). At hand entry, the upper trapezius had a magnitude of activity of 64% MMT ± 42. The muscle activity quickly dropped to a resting level of 4% to 16% MMT until the humerus passed the point where it was perpendicular to the trunk. At this time, there was a sharp rise in activity to 80% MMT ± 32. This was of a similar magnitude as the rhomboids, and just slightly before the rhomboids. When the hand came out of the water, the muscle activity dropped 3. Muscle muscles.
Figure
activity firing patterns
in the rotator cuff
·
573
until the humerus was perpendicular to the water’s surface and preparing for hand entry once again (Fig. 4, Table 1). The serratus anterior demonstrated a constant level of activity throughout the swim cycle, with most action being between 20% and 40% MMT. The two notable exceptions were an increase at the middle of pull-through (48% MMT), and at hand exit (45% MMT) (Fig. 4, Table 1). Shoulder extensors
pectoralis major and latissimus dorsi demonstrated sharp peaks of muscle activity during middle pull-through.
The
Figure 5. Muscle activity firing patterns in the shoulder extensors.
The pectoralis major reached its peak activity before the latissimus dorsi. The intensity of activity in the two muscles was similar: 71% MMT ± 32 for the pectoralis major and 75% MMT ± 49 for the latissimus dorsi. Immediately after these peaks, the muscle activity dropped to resting levels for each of these muscles (2% to 38% MMT) (Fig. 5, Table 1).
DISCUSSION
Pull-through The
Figure cles.
4. Muscle
activity firing patterns
in the
scapular mus-
points selected on the film to mark pull-through were picked because they were repeatable observations. Muscle activity, however, showed three different phases within pullthrough. The first phase was reaching forward and gliding. From the point that the hand entered the water to the point of maximal elbow extension, there was no actual pulling. The underwater reach was really a transition phase between recovery and the actual pulling. Pulling began after the reach. Reciprocally, the pulling actually stopped when the palm approached the thigh, as opposed to when the hand exited. As the palm approached the thigh, it was rotated inward
so
that it could exit the water with minimal
drag.lo
574
These ideal points (i.e., maximal reach and palm rotation) could not be consistently and accurately discerned on the film. Thus, the points mentioned in the methods were chosen and the phases were subdivided to delineate the different occurrences.
Reach
Reach began as the hand entered the water. The hand entered forward of and lateral to the head, and medial to the shoulder. The elbow was flexed, with the elbow above the hand, so that the fingers were the first to enter the water
Figure 8. Teres minor and pectoralis major activity during pulling.
rn..~~
Figure
6.
Primary
movers
at hand
entry and exit.
Figure 9. Muscle and teres minor.
activity firing patterns
in the
infraspinatus
with the palm facing out.10 At the time of hand entry, there was a predominance of phasic activity in the upper trapezius, rhomboids, supraspinatus, and the anterior and middle deltoids. There was also an increase in the serratus anterior. The serratus anterior was upwardly rotating and protracting the scapula while the upper trapezius was elevating it and the rhomboids were retracting it. This muscle action positioned the glenoid fossa for the humeral head as the arm was abducted and flexed by the supraspinatus and anterior and middle deltoids (Fig. 6).
Pulling
Figure 7. The S-shaped
curve
in
pull-through.
The hand followed an S-shaped curve during the pullthrough phase (Fig. 7).14 During the most propulsive part, the humerus adducted powerfully, causing the hand to cross under the chest. Then the hand travelled laterally and passed the pelvis.&dquo; The pectoralis major was responsible for the initial powerful adduction and extension. At the same time, the teres minor probably contributes to extension. The pectoralis major also caused internal rotation, thus, the teres minor fired to provide an antagonistic external rotation force (Fig. 8). After the burst of activity from the pectoralis major and teres minor, the humerus crossed the point where it was
575
perpendicular to the body, and the latissimus dorsi had the mechanical advantage. The latissimus dorsi then continued the pulling by forcefully extending the shoulder. The humerus was internally rotating at that time with the subscapularis assisting the latissimus dorsi. Throughout the propulsive movements of the pectoralis major and latissimus dorsi, the serratus anterior was active. The serratus anterior (by reversing its origin and insertion), along with the latissimus dorsi and pectoralis major, was acting to move the body over the arm and through the water. Also, by virtue of upwardly rotating the scapula, the serratus anterior was able to keep the scapula with the humerus and thus maintain joint congruency.
When the latissimus dorsi finished its activity, the posterior deltoid fired. The function of the posterior deltoid was to transition between pulling (i.e., shoulder extension) and lifting the shoulder out of the water. The other muscles that became active were the same as those at hand entry (Fig. 6). The middle deltoid became active as the arm began to abduct and continued to lift the arm. The supraspinatus was actively assisting the middle deltoid to abduct the arm. As the hand began to exit, the anterior deltoid then fired to initiate flexion of the shoulder. The upper trapezius was active to upwardly rotate the scapula, while the rhomboids were active to retract the scapula. The serratus anterior assisted the scapular rotation and protraction in order to avoid impingement of the coracoid process or acromion on the humerus.
much shorter phase than pull-through. Its simply to bring the arm into position to pull once again. In early recovery, the muscles of the normal shoulder were lifting and abducting the arm and rotating the scapula. The arm was internally rotated as it left the water.&dquo; Thus, the activity noted at the end of pull-through in the middle deltoid and supraspinatus carried into early recovery. The rhomboids fired to retract the scapula. The subscapularis was active as it internally rotated the humerus, while the forearm was swinging around. The infraspinatus was also active to depress the humerus, and to provide a stabilizing force for the strong pull of the antagonistic subscapularis. Thus, the sequence of muscle activity in the normal shoulder during recovery was lift and abduction of the humerus (middle deltoid and supraspinatus), retraction of the scapula (rhomboids) followed by internal rotation of the humerus (subscapularis), and depression and stabilization of the humerus (infraspinatus). The hand was then placed in position to begin entry into the water again. It is of interest to note that both the subscapularis and the serratus anterior continually fired above 20% MMT. Monadll found that 15% to 20% of maximal voluntary contraction was the highest level at which sustained activity can be performed without fatigue. Thus, these two muscles would appear to be susceptible to fatigue. Another interesting point lies with the teres minor and infraspinatus. Because of their anatomical proximity, one might think they would function similarly. However, the
Recovery
individually. SUMMARY There are five summary points to be made for the normal shoulder during freestyle swimming: 1. The patterns of muscular activity at hand entry and exit were similar. The upper trapezius and rhomboids complemented each other: the upper trapezius upwardly rotated the scapula, and the rhomboids retracted the scapula. The three heads of the deltoid were active to lift and place the arm in position for hand exit, and the anterior and middle deltoids at hand entry. 2. The supraspinatus fired with the anterior and middle deltoid to abduct the humerus at hand entry and exit. 3. The rotator cuff muscles each had a unique role. 4. The pectoralis major and latissimus dorsi were active during the propulsive phase of the stroke. The serratus anterior and teres minor were also active at this time. 5. The subscapularis and serratus anterior were active throughout the stroke cycle, and thus were susceptible to
fatigue.
Recovery purpose
muscle-firing patterns tell a very different story (Fig. 9). The infraspinatus depressed the humerus in midrecovery to control the strong internal rotation force of the subscapularis, while the teres minor was quite active in pull-through when acting with the pectoralis major. This reinforces the fact that the role of each muscle needs to be considered
was a
was
This information provides a basis for comparison with painful shoulders during swimming, which is the topic of the following paper (pp 577-582). It also provides a biomechanical framework for development of muscle conditioning programs to optimize performance and prevent injury, as well as programs for specific rehabilitative strengthening programs.
ACKNOWLEDGMENT The authors thank Dr. Frank Jobe for all of his contributions to this study.
REFERENCES 1 2
Basmajian JV: Muscles Alive, Their Functions Revealed by Electromyography Baltimore, Williams & Wilkins, 1978 Clarys JP: A review of EMG in swimming: Explanation of facts and/or feedback information, in Hollander AP, Huijing P, De Goats G (eds) Biomechanics and Medicine in Swimming Champaign, IL, Human Kinetics,
3
Clarys JP,
4
Clarys JP, Massez C, Van Den Broeck, et al. Total telemetric surface EMG
1983, pp 123-135 Jiskoot J, Lewillie L A kinematographical electromyographical, and resistance study of water polo and competition front crawl, in Cerguilini . Basel, S Karger, S, Venerando A, Wartenweiler J (eds) Biomechanics III
1973, pp 446-452 of the front crawl, in Matui H, Kobayashi K (eds): Biomechanics IXB , Champaign, IL, Human Kinetics Publishers Inc , 1983, pp 951-958 5 Dominguez RN: Shoulder pain in age group swimmers, in Enckson B, Furberg B (eds). Swimmmg Medicine IV Vol 6 Baltimore, University Park Press, 1978, pp 105-109
576 6 Ikai M, Ishi K, Miyashita M: An electromyographic J Phys Educ 7. 55-87, 1964 7 Johnson D In swimming, shoulder the burden
study of swimming
Res
Sportcare and Fitness (May-June) 24-30, 1988 8 Lewillie L Telemetry of electromyographic and electrogoniometric signals in swimming, in Nelson RC, Moorehouse CA (eds) Biomechanics IV , Basel, S Karger Verlag, 1974, pp 203-207 9 Lewillie L Muscular activity in swimming, in Cerquiglini S, Venerando A, Wartenweiler J (eds) Biomechanics III . Basel, S Karger Verlag, 1973, pp 440-445
10
Maglischo EW Swimming Faster Co, 1982, pp 53-99
Mountain View,
CA, Mayfield Publishing
11 12.
Contractility of muscle during prolonged static and repetitive dynamic activity Ergonomics 28(1) 81-89, 1985 Nuber GW, Jobe FW, Perry J, et al Fine wire electromyography analysis of muscles of the shoulder during swimming Am J Sports Med 14 7-11, Monad H.
1986 13 Okamoto T, Wolf SL Underwater recording of electromyographic activity using fine wire electrodes, in Terands J, Bedingfield W (eds): Swimming III Baltimore, University Park Press, 1979, pp 160-166 14. Richardson AB. The biomechanics of swimming: The shoulder and knee. Clin Sports Med 5 103-113, 1986 15. Richardson AB, Jobe FW, Collins HR: The shoulder in competitive swimming. Am J Sports Med 8 159-163, 1980
during freestyle
swimming An electromyographic and twelve muscles
cinematographic analysis
of
MARILYN PINK,* MS, PT, JACQUELIN PERRY, MD, ANTHONY BROWNE, MD, MARY LYNN SCOVAZZO, MD, AND JOHN KERRIGAN, MSE From the Biomechanics
Laboratory, Centinela Hospital Medical Center, Inglewood, California
ABSTRACT
demanding
such stress, the swimmer’s shoulder will show
signs of wear and tear. This is confirmed in the literature with shoulder problems being reported in 42% to 67% of competitive swimmers. 1,7,15 To understand the mechanism of injury, and thereby prevent injury, it is important to understand how the shoulder functions during swimming. The biomechanical and electromyographical analysis of swimming has been hampered by unique instrumentation problems such as the water-electrode interface, harnessing, and integration/quantification of the signal. 2-4,6,8,9, 12, 13 There are no published studies that use the advanced technology of fine wire electrodes and telemetered EMG that has been integrated and quantified to analyze the normal shoulder while swimming. After developing the appropriate instrumentation, this study was designed. The purpose of the study is to describe the muscle activity patterns in 12 muscles of the normal shoulder during freestyle swimming.
The shoulder in swimming is subjected to multiple factors that can lead to a high injury rate. To prevent injury, one must understand the biomechanics of swimming. This paper describes the electromyographic and cinematographic findings of 12 shoulder muscles in competitive swimmers without shoulder pain. The results show the three heads of the deltoid and the supraspinatus functioning in synchrony to place the arm at hand entry and exit, the rhomboids and upper trapezius to position the scapula for the arm, the pectoralis major and latissimus dorsi to propel the body, the subscapularis and serratus anterior as muscles with constant muscle activity, the teres minor functioning with the pectoralis major, and the infraspinatus active only to externally rotate the arm at midrecovery. This information is important to design optimal preventative and rehabilitative exercise programs.
MATERIALS AND METHODS One hundred million Americans are swimmers.’ This makes swimming the most popular participation sport in the United States. All ages are represented and people swim for recreation, for competition, and for fitness. Competitive swimmers often practice most of the year, 6 to 7 days a week, and cover 8,000 to 20,000 meters a day.&dquo; This equates to approximately 18,000 shoulder revolutions per week. Thus, the shoulder is subject to strain caused by the high repetition rate, the extremes of range of motion, and the force required for propulsion. As with all sports
Twenty collegiate and masters level competitive swimmers volunteered for this study. All swimmers completed a questionnaire about any episodes of shoulder pain, and underwent a physical examination to check for signs of shoulder instability, apprehension, or impingement. Any subject currently reporting pain was excluded from this population. Many swimmers had past episodes of pain, but were now asymptomatic and considered to be normal. Five of the swimmers volunteered to be tested for both the Group A muscles and the Group B muscles. Thus, the total data sample included 25 shoulders studied in 20 swimmers. They had an average of 9 years of competitive swimming experience and were currently training 2500 to 4000 yards a day,
*
Address correspondence and repnnt requests to Manlyn Pmk, MS, PT, Biomechanics Laboratory, Centinela Hospital Medical Center, 555 East Hardy Street, Inglewood, CA 90301.
569
570
3 to 5 days a week. Fifty-two percent of the reported the freestyle stroke as their best stroke.
swimmers The mean age of the swimmers was 39 years. There were 15 male swimmers and 5 female swimmers.
Procedure
Twelve shoulder muscles were selected for this study, and divided into two groups of six each. Group A muscles included the anterior, middle, and posterior deltoids, serratus anterior, upper trapezius, and rhomboid major. Group B muscles included the subscapularis, supraspinatus, infraspinatus, teres minor, latissimus dorsi, and pectoralis major. Group A muscles were monitored in 16 shoulders, and Group B muscles in 9 shoulders. Testing was done in one of two pools that were equipped with underwater windows. Motion was recorded from the lateral projection using two 16 mm high-speed motion picture cameras operating at 100 frames per second. The underwater view filmed the pull-through phase of the stroke cycle. The surface view filmed the recovery phase of the stroke cycle. Marks were electronically placed on the film and EMG data to allow for synchronization. Recording of the EMG signal was done using the Bas-
without placing undue tension on the leads or interfering with the stroke style. The EMG data were displayed on the oscilloscope, bandpass filtered at 100 to 1000 Hz, and recorded on a multichannel instrument recorder (Model 3968A, Hewlett Packard, Palo Alto, CA) for later retrieval and review. Prior to recording the swim data, the resting EMG was recorded while the swimmers did a prone float. Next, a maximal isometric manual muscle test was done in the water for each muscle. The resting and maximal muscle test recordings were repeated after the swims to assure that damp-
majian single-needle technique.’ Following proper skin preparation and isolation of the muscle, dual 50 micron insulated wires with 2 to 3 mm bared tips were inserted into the muscle using a 25 gauge needle as a cannula. Accuracy of placement was determined on dry land by manual muscle examination and signal display on an oscilloscope. The wires from each muscle were attached to leads that insulated and connected to a ground plate taped to the subject’s back. Finally, waterproof transparent dressings (Bioclusive, Johnson & Johnson, New Brunswick, NJ) were placed over all the insertion points and wires to secure the system. Taping was done in a manner that allowed unencumbered movement of the shoulder. Signals from the leads were transmitted using an FM-FM were
telemetry system (Model 42000-A, Bio-Sentry Telemetry, Torrance, CA). The battery-operated FM transmitters were secured within an empty 2.5 gallon plastic water jug and the wires were carefully sealed around the site of wire entry. The container was secured to the subject’s waist. This enabled the transmitter to float and trail behind the swimmer
.............-
Figure 1.
Phases of the
freestyle swimming stroke cycle.
Figure 2. Muscle activity firing patterns in the deltoids.
571
ening of the signal due
to any water
leakage had
not
oc-
when the humerus
was
perpendicular
to the water surface
curred.
(four intervals).
The subject started swimming at a moderate pace at approximately 20 yards away from the end of the pool, parallel to the underwater window. Data for two to four stroke cycles were obtained. The subject was asked to swim three trials of the freestyle stroke. The EMG data were converted from analog to digital signals by computer at a sampling rate of 2500 Hz, and were quantitated by computer integration. After excluding the noise identified by the resting recording, the peak 1 second EMG signal during a manual muscle test (MMT) was selected as a normalizing value (100%). Activity patterns were assessed every 20 msec and expressed as a percentage of the normalization base. These activity patterns were synchronized with the film to obtain percent muscle activity values at separate phases of the subjects’ motion.
4. Late recovery: beginning at the completion of early recovery and ending at hand entry (two intervals). The data were analyzed to determine whether there was any significant difference due to hand dominance (right or
Data
were
calculated.
analysis
was divided into 25 equal time intervals. The film was synchronized with the EMG to discern four phases. The four phases and the number of intervals per phase were as follows (Fig. 1). 1. Early pull-through: beginning with the hand entry into the water and ending when the humerus was perpendicular to the axis of the torso (12 intervals). 2. Late pull-through: beginning at the completion of early pull-through and ending as the hand left the water
The freestyle stroke cycle
(seven intervals). 3.
left) or breathing side (right, left, or bilateral). An independent t-test was used for the hand dominance question, and an analysis of variance for the breathing side question (P < 0.05). There were no significant differences in either of these factors, thus the data for all normals were able to be grouped. The data were assessed for normality. Outlying data were checked to determine whether they were artifact or noise, and the decision to keep or delete the outlying data was made accordingly. The percent MMT for each subdivision of each phase was then averaged and standard deviations
Early
recovery:
beginning
at hand exit and
ending
RESULTS Deltoids The three heads of the deltoid fired sequentially at the end of pull-through. Peak activity was first displayed by the posterior deltoid, followed by the middle deltoid, then the anterior deltoid. They all reached peaks of activity between 70% and 76% MMT. The activity dropped in recovery. The anterior and middle deltoids demonstrated a second peak of activity at hand entry (45% to 51% MMT). This was fol-
TABLE 1 Muscle
activity at
25
points during the freestyle stroke in normal shoulders
572
by low level activity until they repeated the peak activity at the end of pull-through (Fig. 2, Table 1). lowed
Rotator cuff The four rotator cuff muscles each demonstrated a different activity pattern. This confirmed that each muscle has a discrete function. The supraspinatus activity was similar to that of the anterior and middle deltoids. There were two distinct periods of activity. The muscle was active at hand entry (54% MMT ± 29), and then dropped to lower levels (6% to 19% MMT) for middle pull-through. The second distinct period of activity was seen at the end of pull-through (66% MMT ± 12) and early recovery (74% MMT ± 37). During middle and late recovery, the muscle activity decreased slightly until hand entry occurred once again (Fig. 3, Table 1). The infraspinatus had its highest level of muscle activity at midrecovery (34% MMT ± 34). Otherwise, there were relatively low levels of participation during pull-through (below 16% MMT) (Fig. 3, Table 1). The teres minor demonstrated a gradual rise in muscle activity in pull-through, up to 57% MMT at middle pullthrough. Activity then diminished and stayed between 16% MMT and 19% MMT in recovery (Fig. 3, Table 1). The subscapularis was active throughout the stroke, with the intensity range between 26% and 71% MMT. Late pullthrough and early recovery each had peaks of activity (64% and 71% MMT, respectively). The lowest level of activity in the subscapularis was during early pull-through (26% to 49% MMT) (Fig. 3, Table 1).
Scapular muscles The rhomboids and the upper trapezius demonstrated a similar pattern of muscle activity in the normal shoulders. They both had two peaks of muscle activity: one at hand entry and the other at hand exit. The serratus anterior had a pattern distinct from these other two scapular muscles. The rhomboids revealed sharp activation at hand entry (49% MMT ± 23). It then dropped to a range of 9% to 13% MMT. Once the humerus passed the point where it was perpendicular to the trunk, the rhomboids gradually became more active. By early recovery it reached the peak activation of 76% MMT ± 47. Activity then decreased until hand entry
began once again (Fig. 4, Table 1). At hand entry, the upper trapezius had a magnitude of activity of 64% MMT ± 42. The muscle activity quickly dropped to a resting level of 4% to 16% MMT until the humerus passed the point where it was perpendicular to the trunk. At this time, there was a sharp rise in activity to 80% MMT ± 32. This was of a similar magnitude as the rhomboids, and just slightly before the rhomboids. When the hand came out of the water, the muscle activity dropped 3. Muscle muscles.
Figure
activity firing patterns
in the rotator cuff
·
573
until the humerus was perpendicular to the water’s surface and preparing for hand entry once again (Fig. 4, Table 1). The serratus anterior demonstrated a constant level of activity throughout the swim cycle, with most action being between 20% and 40% MMT. The two notable exceptions were an increase at the middle of pull-through (48% MMT), and at hand exit (45% MMT) (Fig. 4, Table 1). Shoulder extensors
pectoralis major and latissimus dorsi demonstrated sharp peaks of muscle activity during middle pull-through.
The
Figure 5. Muscle activity firing patterns in the shoulder extensors.
The pectoralis major reached its peak activity before the latissimus dorsi. The intensity of activity in the two muscles was similar: 71% MMT ± 32 for the pectoralis major and 75% MMT ± 49 for the latissimus dorsi. Immediately after these peaks, the muscle activity dropped to resting levels for each of these muscles (2% to 38% MMT) (Fig. 5, Table 1).
DISCUSSION
Pull-through The
Figure cles.
4. Muscle
activity firing patterns
in the
scapular mus-
points selected on the film to mark pull-through were picked because they were repeatable observations. Muscle activity, however, showed three different phases within pullthrough. The first phase was reaching forward and gliding. From the point that the hand entered the water to the point of maximal elbow extension, there was no actual pulling. The underwater reach was really a transition phase between recovery and the actual pulling. Pulling began after the reach. Reciprocally, the pulling actually stopped when the palm approached the thigh, as opposed to when the hand exited. As the palm approached the thigh, it was rotated inward
so
that it could exit the water with minimal
drag.lo
574
These ideal points (i.e., maximal reach and palm rotation) could not be consistently and accurately discerned on the film. Thus, the points mentioned in the methods were chosen and the phases were subdivided to delineate the different occurrences.
Reach
Reach began as the hand entered the water. The hand entered forward of and lateral to the head, and medial to the shoulder. The elbow was flexed, with the elbow above the hand, so that the fingers were the first to enter the water
Figure 8. Teres minor and pectoralis major activity during pulling.
rn..~~
Figure
6.
Primary
movers
at hand
entry and exit.
Figure 9. Muscle and teres minor.
activity firing patterns
in the
infraspinatus
with the palm facing out.10 At the time of hand entry, there was a predominance of phasic activity in the upper trapezius, rhomboids, supraspinatus, and the anterior and middle deltoids. There was also an increase in the serratus anterior. The serratus anterior was upwardly rotating and protracting the scapula while the upper trapezius was elevating it and the rhomboids were retracting it. This muscle action positioned the glenoid fossa for the humeral head as the arm was abducted and flexed by the supraspinatus and anterior and middle deltoids (Fig. 6).
Pulling
Figure 7. The S-shaped
curve
in
pull-through.
The hand followed an S-shaped curve during the pullthrough phase (Fig. 7).14 During the most propulsive part, the humerus adducted powerfully, causing the hand to cross under the chest. Then the hand travelled laterally and passed the pelvis.&dquo; The pectoralis major was responsible for the initial powerful adduction and extension. At the same time, the teres minor probably contributes to extension. The pectoralis major also caused internal rotation, thus, the teres minor fired to provide an antagonistic external rotation force (Fig. 8). After the burst of activity from the pectoralis major and teres minor, the humerus crossed the point where it was
575
perpendicular to the body, and the latissimus dorsi had the mechanical advantage. The latissimus dorsi then continued the pulling by forcefully extending the shoulder. The humerus was internally rotating at that time with the subscapularis assisting the latissimus dorsi. Throughout the propulsive movements of the pectoralis major and latissimus dorsi, the serratus anterior was active. The serratus anterior (by reversing its origin and insertion), along with the latissimus dorsi and pectoralis major, was acting to move the body over the arm and through the water. Also, by virtue of upwardly rotating the scapula, the serratus anterior was able to keep the scapula with the humerus and thus maintain joint congruency.
When the latissimus dorsi finished its activity, the posterior deltoid fired. The function of the posterior deltoid was to transition between pulling (i.e., shoulder extension) and lifting the shoulder out of the water. The other muscles that became active were the same as those at hand entry (Fig. 6). The middle deltoid became active as the arm began to abduct and continued to lift the arm. The supraspinatus was actively assisting the middle deltoid to abduct the arm. As the hand began to exit, the anterior deltoid then fired to initiate flexion of the shoulder. The upper trapezius was active to upwardly rotate the scapula, while the rhomboids were active to retract the scapula. The serratus anterior assisted the scapular rotation and protraction in order to avoid impingement of the coracoid process or acromion on the humerus.
much shorter phase than pull-through. Its simply to bring the arm into position to pull once again. In early recovery, the muscles of the normal shoulder were lifting and abducting the arm and rotating the scapula. The arm was internally rotated as it left the water.&dquo; Thus, the activity noted at the end of pull-through in the middle deltoid and supraspinatus carried into early recovery. The rhomboids fired to retract the scapula. The subscapularis was active as it internally rotated the humerus, while the forearm was swinging around. The infraspinatus was also active to depress the humerus, and to provide a stabilizing force for the strong pull of the antagonistic subscapularis. Thus, the sequence of muscle activity in the normal shoulder during recovery was lift and abduction of the humerus (middle deltoid and supraspinatus), retraction of the scapula (rhomboids) followed by internal rotation of the humerus (subscapularis), and depression and stabilization of the humerus (infraspinatus). The hand was then placed in position to begin entry into the water again. It is of interest to note that both the subscapularis and the serratus anterior continually fired above 20% MMT. Monadll found that 15% to 20% of maximal voluntary contraction was the highest level at which sustained activity can be performed without fatigue. Thus, these two muscles would appear to be susceptible to fatigue. Another interesting point lies with the teres minor and infraspinatus. Because of their anatomical proximity, one might think they would function similarly. However, the
Recovery
individually. SUMMARY There are five summary points to be made for the normal shoulder during freestyle swimming: 1. The patterns of muscular activity at hand entry and exit were similar. The upper trapezius and rhomboids complemented each other: the upper trapezius upwardly rotated the scapula, and the rhomboids retracted the scapula. The three heads of the deltoid were active to lift and place the arm in position for hand exit, and the anterior and middle deltoids at hand entry. 2. The supraspinatus fired with the anterior and middle deltoid to abduct the humerus at hand entry and exit. 3. The rotator cuff muscles each had a unique role. 4. The pectoralis major and latissimus dorsi were active during the propulsive phase of the stroke. The serratus anterior and teres minor were also active at this time. 5. The subscapularis and serratus anterior were active throughout the stroke cycle, and thus were susceptible to
fatigue.
Recovery purpose
muscle-firing patterns tell a very different story (Fig. 9). The infraspinatus depressed the humerus in midrecovery to control the strong internal rotation force of the subscapularis, while the teres minor was quite active in pull-through when acting with the pectoralis major. This reinforces the fact that the role of each muscle needs to be considered
was a
was
This information provides a basis for comparison with painful shoulders during swimming, which is the topic of the following paper (pp 577-582). It also provides a biomechanical framework for development of muscle conditioning programs to optimize performance and prevent injury, as well as programs for specific rehabilitative strengthening programs.
ACKNOWLEDGMENT The authors thank Dr. Frank Jobe for all of his contributions to this study.
REFERENCES 1 2
Basmajian JV: Muscles Alive, Their Functions Revealed by Electromyography Baltimore, Williams & Wilkins, 1978 Clarys JP: A review of EMG in swimming: Explanation of facts and/or feedback information, in Hollander AP, Huijing P, De Goats G (eds) Biomechanics and Medicine in Swimming Champaign, IL, Human Kinetics,
3
Clarys JP,
4
Clarys JP, Massez C, Van Den Broeck, et al. Total telemetric surface EMG
1983, pp 123-135 Jiskoot J, Lewillie L A kinematographical electromyographical, and resistance study of water polo and competition front crawl, in Cerguilini . Basel, S Karger, S, Venerando A, Wartenweiler J (eds) Biomechanics III
1973, pp 446-452 of the front crawl, in Matui H, Kobayashi K (eds): Biomechanics IXB , Champaign, IL, Human Kinetics Publishers Inc , 1983, pp 951-958 5 Dominguez RN: Shoulder pain in age group swimmers, in Enckson B, Furberg B (eds). Swimmmg Medicine IV Vol 6 Baltimore, University Park Press, 1978, pp 105-109
576 6 Ikai M, Ishi K, Miyashita M: An electromyographic J Phys Educ 7. 55-87, 1964 7 Johnson D In swimming, shoulder the burden
study of swimming
Res
Sportcare and Fitness (May-June) 24-30, 1988 8 Lewillie L Telemetry of electromyographic and electrogoniometric signals in swimming, in Nelson RC, Moorehouse CA (eds) Biomechanics IV , Basel, S Karger Verlag, 1974, pp 203-207 9 Lewillie L Muscular activity in swimming, in Cerquiglini S, Venerando A, Wartenweiler J (eds) Biomechanics III . Basel, S Karger Verlag, 1973, pp 440-445
10
Maglischo EW Swimming Faster Co, 1982, pp 53-99
Mountain View,
CA, Mayfield Publishing
11 12.
Contractility of muscle during prolonged static and repetitive dynamic activity Ergonomics 28(1) 81-89, 1985 Nuber GW, Jobe FW, Perry J, et al Fine wire electromyography analysis of muscles of the shoulder during swimming Am J Sports Med 14 7-11, Monad H.
1986 13 Okamoto T, Wolf SL Underwater recording of electromyographic activity using fine wire electrodes, in Terands J, Bedingfield W (eds): Swimming III Baltimore, University Park Press, 1979, pp 160-166 14. Richardson AB. The biomechanics of swimming: The shoulder and knee. Clin Sports Med 5 103-113, 1986 15. Richardson AB, Jobe FW, Collins HR: The shoulder in competitive swimming. Am J Sports Med 8 159-163, 1980
