Effects of upper limb immobilization on isometric muscle strength, movement time, and triphasic electromyographic characteristics.

Effects of Upper Limb Immobilization on Isometric Muscle Strength, Movement Time, and Triphasic Electromyographic Characteristics Limb immobilization resulting from traumatic or pathological conditions produces characteristic negative effects as a result of disuse muscle atrophy. In addition to economic and personal restrictions during the period of immobilization, a long period following immobilization must be used to restore muscle strength and endurance deficits.

Although numerous physiological responses to limb immobilization are well known, adequate information is not available concerning electrical activity and effects on functional activities such as lifting, pushing, or pulling objects. If these effects can be better defined and identified, the negative consequences of limb immobilization may be lessened.

One of the potential origins of muscle atrophy is neurological. Several authors have suggested a link between muscle atrophy and the response of the nervous system to limb immobilization. [1-3] If the effects of limb immobilization on neurological activity can be established more clearly, the connection between muscle atrophy and limb immobilization may become more apparent and allow the development of methods for reducing rehabilitation time after immobilization.

Numerous studies have noted large and significant muscle strength reductions associated with immobilization. [4-12] Exactly how these strength decreases occur is difficult to ascertain, although several factors have been found to play a role.

Several investigatos have shown that disruption of joint structues plays an important role in muscle strength changes even without immobilization. [13] When joint capsules were distended, significant decreases in associated muscle strength occurred. [13] These results suggest that when a joint sustains an injury, reflex inhibition occurs, which is manifested in decreases in muscle tension output. [14]

In several studies assessing the effect of immobilization on an uninjured limb, muscle strength deficits of 23% to 41% were noted. [15-17] These studies suggest that at least some portion of the total strength deficits noted in the casted patient occur secondary to immobilization. The question remains as to what is the mechanism or trigger for the decreases in muscle strength measurements.

A primary concern in the assessment of disuse atrophy is the degree of functional loss that occurs. Because the goal of any rehabilitation program is to restore as much function as possible, it seems natural to assess disuse atrophy in those terms. Functional changes, howwever, are difficult to quantify. The physical therapist must use simple models of function and assess the changes that occur in these models. One of these models it the ballistic forearm motion. Rapid ballistic movement along with slow ramp movements constitute most normal limb movements. Because it is the only component that can be measured accurately, the ballistic forearm motion can be used to assess simulated functional changes.

During a ballistic limb movement, a triphasic electromyographic pattern has been noted. [18-20] The triphasic EMG pattern consists of first a burst of activity in the agonist, then a burst in the antagonist during which the agonist is quiet, followed by a second agonist burst that occurs after the cessation of motion (Fig. 1). The first agonist burst is the initiation of the ballistic. movement. The antagonist burst is the stopping of the motion. The second agonist burst probably provides some stabilization of the joint when halting motion, although this action has been disputed. [21] It has been shown that the second agonist burst remains unchanged even when the antagonist muscle is paralyzed. [21] This finding leads to the explanation that the second agonist burst helps maintain the readiness of the agonist in case subsequent contractions are necessary. [21]

In numerous studies examining EMG activity in injured limbs, marked decreases were found during and following immobilization. [1, 3, 11, 22] These results imply some shutting down or blocking of neurological impulse transmissions. Exactly where this action occurs is unclear. It is also unclear whether these decreases in motoneuron activity have some effect on limb function. As a result of the difficulty in quantifying EMG data, no studies to date have shown how much of a decrease in any EMG characteristic does occur.

Kinematic characteristics such as movement time can also be studied during the ballistic movement. Movement time is defined for the purposes of this study as the time it takes to move a limb volitionally through a set arc of motion as rapidly as possible.

It seems reasonable, therefore, to expect limb immobilization to have a diminishing effect on muscle strength, integrated electromyographic activity, and limb movement time. The purpose of this study was to investigate the effects of upper limb immobilization on isometric elbow flexion and extension strength, movement time, and peak IEMG activity of the triphasic pattern during ballistic forearm flexion and extension.

Method

Subjects

Six subjects (two female, four male) with a mean age of 31.2 years (range = 25-37 years) volunteered to participate in the study. All subjects had no history of upper extremity injury. The research protocol was approved by the Human Investigation Committee of the Department of Exercise Science, University of Massachusetts. In addition, each subject was informed of all procedures to be used and signed an informed consent document.

Muscle Strength Assessment

Isometric muscle strength for flexion was assessed using a strain gauge with the limb position at 90 degrees of elbow flexion and the forearm in a neutral position between pronation and supination. The subjects were seated with the elbow supported and the shoulder in approximately 70 degrees of flexion. The strain gauge was attached to an immovable support on one side and at the other end to a leather cuff that was secured to the subject's wrist. The gauge was placed so as to measure force generated by the elbow flexors.

To measure forearm extensor strength, the upper extremity was maintained in the same position with the elbow flexed at 90 degrees. The strain gauge was placed on the side of the limb such that the direction of pull was into extension. The strain gauge was calibrated before beginning data collection on the first subject. Tension output was recorded and quantified on a Beckman dynagraph. (*1)

On command, a maximal isometric contraction of the agonist musculature was performed until a peak followed by a level plateau was observed on the dynagraph. The subjects were given verbal encouragement throughout each contraction. Each contraction was held for approximately five seconds. The pretests consisted of two trials on each of two days for the flexors and extensors of both upper limbs.

During the posttest, two trials for both flexion and extension of both upper extremeties were completed with changes in peak isometric tension output being analyzed. The dominant limb ws not immobilized and was used as the control.

Movement Time Measurement

To measure movement time, the subject was seated at a table with the limb to be tested strapped to a moveable wooden arm by a wrist cuff. The axis of the wooden arm was aligned with the axis of the elbows. The testing apparatus was designed such that movement of the arm triggered microswitches at the onset of movement and at the end of a 70-degree arc of motion. The switches triggered marks on the dynagraph paper from which movement time was calcualted (Fig. 1). To analyze the flexion motion, the limb started in 20 degrees of flexion and volitionally ended at 90 degrees of flexion.

Measurement of the extension motion began at 90 degrees of flexion and was volitionally ended at 20 degrees of flexion. On command, subjects moved the forearm as rapidly as possible in the direction being tested. A digital clock, measuring in milliseconds and triggered by the microswitches, was positioned to provide feedback to the subject. This clock was not used for analysis, but the sujects were encouraged to attempt to "beat" the previous movement time indicated on the clock. Two pretests were conducted on separate days with 10 trials for both limbs in flexion and extension performed each day. Changes in flexion and extension movement time were assessed following limb immobilization.

Integrated Electromyographic

Measurement

During the ballistic movement of the forearm, IEMG peak amplitude data were collected. Silver-silver chloride surface electrodes wee used to record EMG data for the agonist and antagonist musculature. Standard electrode application was used involving alcohol cleansing and light skin abrasion. The electrodes were fixed to the belly of the biceps brachii muscle and to the belly of the lateral head of the triceps brachii muscle because they were most prominent and allowed for accurate reapplication of the electrodes for the posttest. The skin preparation caused the skin to be marked such that all electrodes were easily applied to the same position for testing following immobilization. Using an ohmmeter, the skin resistance was evaluated to ensure it was below 10 k[omega] before all test sessions. The EMG signals were processed using a Beckman Type 9852 EMG integrator coupler. (*1) The coupler has a frequency response of 5,000 Hz and a common mode rejection ratio of 300,000:1 at 60 Hz; the mode rejection ratio is virtually infinite using direct current (DC). The EMG signals were amplified with a high-input inpedance differential amplifier. They were full-wave rectified and integrated by a low-pass filter, converted into DC, and filtered for frequencies above 40 Hz, thus producing a linear envelope. The integration time constant was 0.2 seconds.

During the maximum ballistic arm movement trials, IEMG activity was recorded on the dynagraph. Each trial produced the distinct triphasic pattern, which allowed analysis of the agonist first burst peak amplitude, the antagonist peak amplitude, and agonist second burst peak amplitude (Fig. 1e.

Treatment Procedure

On completion of the pretesting, each subject underwent plaster immobilization of the nondominant upper limb. The cast was applied by an orthopedist using standard application procedures and remained intact for 14 days. The cast extended from the axilla to the base of the fingers with the elbow fixed at 90 degrees and the wrist midway between pronation and supination. During this period, the subjects wore an arm sling to provide support for the casted limb and to further discourage use of the immobilized limb. The subjects were requested to avoid use of the immobilized limb and to generally refrain from strenuous activity. At the end of the casting period, the cast was removed by the orthopedist. Posttesting was completed within 24 hours after cast removal.

Data Analysis

The dynagraph tracings of agonist and antagonist IEMG activity and the movement time markers were digitized and computer analyzed. The data provided movement time and the three peak IEMG characteristics. Means and standard deviations were also calculated (Tab. 1).

An intraclass correlation coefficient (ICC:3,1) was calcualted for each of the triphasic IEMG characteristic measurements obtained on all test days to determine the consistency of the data. A two-way analysis of variance (ANOVA) for repeated measures was used to determine the reliability of the pretest data. A two-way ANOVA for repeated measures was also used to assess significant changes between pretest and posttest data in both casted and noncasted limbs. If significant interactions or significant differences across all test days were found, then a further two-way ANOVA was performed to determine whether either limb had significantly changed from pretest 2 to the posttest.

Results

Muscle strength

Intraclass correlation coefficients calculated to determine the subjects' consistency in reproducing test scores were quite high for all muscle strength measures (.96 and .99 for flexion strength in the noncasted limb and the casted limb, respectively). Similar results of .88 and .94 occurred for extension strength of the nonimmobilized limb and the immobilized limb, respectively.

No significant changes occurred in the extensor musculature between the two pretest days, indicating good repeatability of muscle strength values for extension. A significant change did occur in flexion strength of the dominant limb (p [is less than] .05). This change most likely occurred as a result of some practice effect during the testing. No significant change in strength occurred in flexion of the nondominant limb.

As a result of immobilization, flexor strength in the immobilized limb showed a significant decrease (p [is less than] .01) resulting from the casting. Extension strength in the casted limb showed no significant chane (Tab. 2).

Movement Time

Intraclass correlation coefficients for flexion movement time were very low in the casted limb (ICC = .32) but considerably higher in the noncasted limb (ICC = .71). No ICC could be calculated for extension of the casted limb because negative true values were obtained. These values resulted from a greater variation occurring within subjects' scores than between subjects' scores, probably because of the nondominant limb being used as the test limb. Because this movement required a certain amount of precision in volitionally stopping the limb, it appears the dominant (noncasted) limb was more accurate, hence consistent. For the purpose of this study, these results were considered inconsistent. Extension of the noncasted limb resulted in an ICC of .36. These results showed an inability of the subjects to produce consistent values during an extension movement.

Analyses of variance revealed no sigificant changes between pretest values in either flexion or extension movement time, indicating reliable test values between pretest days. No significant change in movement time occurred between the pretests and the posttest during the extension movement (Tab. 3). A significant interaction (p [is less than] .05) occurred between the two limbs in flexion movement (Tab. 3). Further analysis of this interaction showed no significant change in either limb between test days (Tab. 4).

Integrated Electromyographic

Peak Amplitude

Intraclass correlation coefficients for peak IEMG amplitude of all musculature in flexion were high. Agonist peak amplitude had ICCs of .68 in the casted limb and .86 in the noncasted limb. Antagonist peak amplitude was very high at .95 in the immobilized limb and .91 in the nonimmobilized limb. Similar results occurred in the agonist second burst amplitude with ICCs of .78 in the casted limb and .88 in the noncasted limb.

During extension, the consistency was somewhat less. Agonist peak amplitude ICCs were .77 and .60 for the casted and noncasted limbs, respectively. Antagonist peak amplitudes had values of .65 for the immobilized limb and .78 for the nonimmobilized limb. The agonist second burst peak amplitude ICCs were .22 for the casted limb and .72 for the noncasted limb.

In assessing reliability, no significant changes occurred between pretest days in any of the IEMG variables during flexion or extension of either limb. Reliable pretest values were obtained for both motions.

During the ballistic flexion motion of the immobilized limb, a significant interaction occurred in the agonist peak amplitude and agonist second burst peak amplitude (Tab. 5). A significant decrease was also found across all test days in antagonist peak amplitude (p [is less than] .05) (Tab. 5). Further analysis revealed a significant decrease (p [is less than] .01) only in antagonist peak amplitude of the casted limb between pretest 2 and the posttest (Tab. 4).

In the ballistic extension motion of the casted limb, a significant interaction (p [is less than] .05) was found in antagonist peak amplitude. Significant decreases were also noted across all test days in agonist peak amplitude and antagonist peak amplitude (p [is less than] .05) (Tab. 6). Further analysis of agonist peak amplitude and antagonist peak amplitude revealed a significant decrease in both variables between pretest 2 and the posttest (p [is less than] .05). A significant difference also existed between groups in antagonist peak amplitude (p [is less than] .05). No significant change occurred in agonist second burst peak amplitude (Tab. 4).

Discussion

Limb immobilization produced a significant decrease in elbow flexor strength of 12.7% (Fig. 2). Because the noncasted limb had essentially no reduction in flexion, limb immobilization appears to have had a negative effect on elbow flexor strength. This decrease is somewhat less than those found in other studies. [17, 20] According to Muller, muscle strength loss resulting from limb immobilization decreased by 23% after 7 days of casting reaching a plateau without further reduction from 7 to 14 days after immobilization. [22] The forearm flexors were not measured in any other study, so direct comparison is difficult. The reason is unclear why the decrease in elbow flexor strength was less in this study than in other studies, because the length of immobilization was the same and all subjects were uninjured.

Convrsely, no significant decreases were found in forearm extensor strength (Fig. 2). This result is very different from that of MacDougall et al who noted a 42% decrease in extensor strength in their study. [24] A major difference existed between this study and that of MacDougall et al, however, in that their subjects were casted for five to six weeks. Length of time of immobilization may be an important factor in the amount of muscle strength loss.

Several possible reasons exist for the difference between the effects on flexor and extensor musculature. Stillwell and associates reported that muscle strength deficits normally resulting from immobilization can be eliminated in healthy, uninjured subjects if isometric (static) exercises are performed during casting. [10] If the subjects had performed more contractions with the triceps brachii muscle, perhaps unknowingly, then the effects of the immobilization could easily have been diminished.

A second possible reason is the position of immobilization, because several studies on animals have shown atrophy is either nonexistent or diminished when muscle is immobilized at a length greater than resting position. [25, 26] Resting length of the triceps brachii muscle is with the forearm in full extension, and the position of the immobilization was with the elbow flexed to 90 degrees. In theory, if the muscle is constantly stretched, then the muscle spindles may fire, which may produce enough of a contraction to prevent muscle atrophy from occurring. [25, 26]

A third possible reason is that extensor musculature is affected differently by immobilization than flexor musculature. Edstrom [27] and Eriksson [28] have shown type I (slow-twitch) muscle fiber to be the fiber type primarily affected by immobilization. Johnson et al found forearm flexor musculature to have a higher percentage of type I muscle fibers than forearm extensor musculature. [29] The changes in muscle strength of the upper extremity may be reflective of the effects of immobilization on specific fiber types.

The movement time data, although not statistically significant, show an 11.8% slowing of movement in the casted limb during flexion. In contrast, the movement of the noncasted limb actually became faster (Fig. 3). Because flexor strength significantly decreased, it seems reasonable that some parallel effect may have occurred in movement time. These results indicate that immobilization may perhaps have some negative effect on functional activities. According to Marchetti and associates, these decreases may be due to some deterioration of the ability of motor centers to recruit motor units or to fire them at higher frequencis. [1] If motor unit activity is diminished following immobilization, then muscle function must also decrease. Because the ballistic motion of forearm flexion is isolated in terms of muscle activity, the power generation originally present before casting may be decreased following immobilization. No concrete conclusion can be drawn regarding the effects of immobilization on movement time, and more research in this area is needed.

Extension movement time also tended to be slower, although not significantly. This decrease, however, occurred in both limbs, regardless of immobilization (fig. 3). The reasons for the noncasted limb's decreased movement time are unclear. Perhaps an overall decrease in the subject's activity level during the immobilization period had some negative effect. All subjects were very physically active before immobilization; hence, any general decrease in activity may have adversely affected all measurements. One of the problems with extension movement time was the lack of consistency in the data as illustrated by low or nonexistent ICCs. Any real changes occurring may have been masked by these inconsistencies.

Large decreases in the casted limb's IEMG amplitude measurements occurred during flexion. A 36% reduction in agonist IEMG peak amplitude was noted in the casted limb, where a 7% increase occurred in the noncasted arm (Fig. 4). A significant decrease of 48% was obtained in antagonist IEMG peak amplitude in the immobilized limb (Fig. 5). This decrease can be compared with a 28% decrease in the nonimmobilized limb. Similarly, in the agonist second burst peak amplitude, a 44% decrease was found in the casted limb (Fig. 6). These results continually point to some type of inhibition of motoneuron activity resulting from immobilization.

Examination of the quantitative IEMG characteristics during extension shows that several had statistically significant changes. Agonist peak amplitude diminished by 62% in the casted limb and 52% in the noncasted limb (Fig. 4). Antagonist peak amplitude decreased by 54% in the immobilized limb and 38% in the nonimmobilized limb (Fig. 5). Agonist second burst peak amplitude during extension had a 46% decrease in the casted limb and a 40% decrease in the noncasted limb (Fig. 6).

These results show further evidence that a decrease in physical activity seems to have affected the extensor musculature of both limbs. It is puzzling why large decreases in EMG activity in the extensor musculature occurred yet muscle strength did not diminish equally. Perhaps short-term immobilization is very selective in the typeof neuromuscular activity inhibited. The neurons controlling rapid ballistic type activity in the muscles may primarily be affected. If the muscle is given enough time, the neurological system can generate enough activity to produce substantial isometric output. In this study, the subjects generated tension for five to eight seconds during the isometric muscle strength tests.

The difference in the flexor and extensor results may possibly be accounted for by fiber type composition. If the extensor musculature is composed of more predominantly type II fibers, then perhaps the neuromuscular ability to produce a maximal isometric muscle contraction over a period of five seconds is unaffected.

Results from both flexion and extension motions indicate that some type of neural inhibition is occurring with limb immobilization. Sale and MacDougall [2] and Sale et al [30] believed that two possible explanations existed. First, they theorized that disuse may have impaired the ability of the descending motor pathways to excite the spinal motoneurons. Their second explanation is that a period of immobilization inhibits neuromuscular transmission. Furthermore, Marchetti and associates suggested that immobilization is accompanied by some impairment of the peripheral components of the neuromuscular system. [1] The evidence in this study concurs with that found by other investigators. Based on the large decreases in peak IEMG amplitude in the agonist and antagonist muscles, it appears some decrease in muscular electrical activity results from immobilization. Whether this decrease is caused by a reduction in central nervous system generation of neurological transmission, a hindrance of axonal transmission, a decrease in synpatic activity, or a failure of the muscle fibers to fully transmit depolarization along the muscle membrane is unclear.

Conclusions

Based on the results of this study, upper limb immobilization appears to have affected some, but not all, of the assessed variables:

1. Strength of the forearm flexors was significantly reduced by immobilization, whereas the forearm extensors were unchanged.

2. Although ballistic flexion of the immobilized limb was slower, it was insignificant. Flexion movement time of the nonimmobilized limb lessened, although insignificantly. Movement time in extension was longer in both limbs, although insignificantly.

3. Significant decreases were noted during flexion in the antagonist peak IEMG amplitude. Large decreases were also noted in agonist peak IEMG amplitude and agonist second burst peak IEMG amplitude during flexion, indicating a general reduction in volume of muscular electrical activity. During extension significant decreases occurred in agonist peak IEMG amplitude and antagonist peak IEMG amplitude.

Perhaps the focus of rehabilitation should be on the prevention of muscle atrophy by blocking the decreases in neuromuscular activity associated with limb immobilization. Although this focus may not eliminate all atrophy attributable to injury and subsequent immobilization, it may provide a significant enough reduction to allow a substantial savings in disability and rehabilitation time. Certainly, further studies are needed to explore safe, effective methods of achieving these savings.

Acknowledgments

I acknowledge the assistance of Walter Kroll, PhD, and Dave Gundy, MD, in conducting the study and of Sarah Vaughan, RN, and Donna Cozzolina in the preparation of this manuscript.

(*1) Beckman instruments, Inc, 3900 River Rd, Schiller Park, IL 60176.

References

[1] Marchetti M, Salleo A, Figura F, et al: Electromyographic and phonomyographic patterns in muscle atrophy in man. In: Biomechanics IV. Baltimore, MD, University Park Press, 1974, pp 388-393

[2] Sale DG, MacDougall JD: Effects of immobilization upon moto-excitability in man. Landry F, Orban W (eds): International Congress of Physical Activity Sciences: Book 6. Biomechanics of Sports and Kinanthropometry. Miami, FL, Symposia Specialists, 1978, pp 79-83

[3] Wolf E, Magora A. Gonen B: Disuse atrophy of the quadriceps muscle. electromyography 11: 479-489, 1971

[4] Fried T, Shephard RJ: Assessment of a lower extremity training program. Can Med Assoc J 103:260-266, 1970

[5] Grimby G, Gustafsson E, Peterson L, et al: Quadriceps function and training after knee ligament surgery. Med Sci Sports Exerc 12:70-75, 1980

[6] Imms FJ, Hackett AJ, Prestidge SP, et al: Voluntary isometric muscle strength of patients undergoing rehabilitation following fractures of the lower limb. Rheumatol Rehabil 16:162-171, 1977

[7] Imms FJ, Prestidge SP, Mayes FB: The decline of aerobic capacity and muscle strength following fractures of the lower limb. Injury 11: 219-224, 1980

[8] Jenkins DG, Imms FJ, Prestidge SP, et al: Muscle strength before and after menisectomy: A comparison of methods of post operative management. Rheumatol Rehabil 15:153-155, 1976

[9] Mendler MH: Knee extensor and flexor force following injury. Phys Ther 47:35-45, 1967

[10] Stillwell DM, McLarren GL, Gersten JW: Atrophy of quadriceps muscle due to immobilization of the lower extremity. Arch Phys Med Rehabil 48:289-295, 1967

[11] Stoboy H, Friedebold G, Strand F: Evaluation of the effect of isometric training on function and organic muscle atrophy. Arch Phys Med Rehabil 49:508-514, 1968

[12] Stoboy H, Friedebold G: Changes in muscle function in atrophied muscles due to isometric training. Bull NY Acad Med 44:553-559, 1968

[13] deAndrade J, Grant A, Dixon A: Joint distention and reflex muscle inhibition in the knee. J Bone Joint Surg [Am] 47:313-322, 1965

[14] Spencer JD, Hayes KC, Alexander IJ: Knee joint effusion and quadriceps reflex inhibition in man. Arch Phys Med Rehabil 65:171-177, 1984

[15] Hettinger T: Das Verholten der Kraft eines Trainierten Muskels Qahrend und nach Mertagiger Ruhestellung. Internationale Zeitschrift fur Angewandte Physiologie Einschiesslich Arbeitsphysiologie 17:357-360, 1960 (German)

[16] MacDougall JD, Elder GC, Sale DG, et al: Effects of strength training and immobilization on human muscle fibres. Eur J Appl Physiol 43: 23-34, 1980

[17] Muller E: Influence of training and of inactivity in muscle strength. Arch Phys Med Rehabil 51:449-462, 1970

[18] Angel RW: Electromyographic patterns during ballistic movements in normals and in hemiplegic patients. In Desmedt JE (ed): Motor Unit Types: Recruitment and Plasticity in Health and Disease. Basel, Switzerland, S Karger AG, Medical and Scientific Publishers, 1981, pp 347-357

[19] Hallet M, Shahani BT, Young RR: EMG analysis of stereotyped voluntary movements in man. J Neurol Neurosurg Psychiatry 38:1154-1162, 1975

[20] Hallet M, Marsden CD: Physiology and pathophysiology of the ballistic movement pattern. In Desmedt JE (ed): Motor Unit Types: Recruitment and Plasticity in Health and Disease. Basel, Switzerland, S Karger AG, Medical and Scientific Publishers, 1981, pp 331-346

[21] Angel RW: Electromyographic patterns during ballistic movement of normal and spastic limbs. Brain Res 99:387-392, 1975

[22] Fuglsang-Fredriksen A, Scheel U: Transient decrease in number of motor units after immobilization in man. J Neurol Neurosurg Psychiatry 41:924-929, 1978

[23] Muller E, Hettinger T: Uber Untershiede der Trainings-geschwindigkiet atrophierter und normaler Muskeln. Arbeitsphysiologie 15:223-230, 1953 (German)

[24] MacDougall JD, Ward GR, Sale DG, et al: Biochemical adaptation of human skeletal muscle to heavy resistance training and immobilization. J Appl Physiol 43:700-703, 1977

[25] Eisenhauer J, Key JA: Studies in muscle atrophy. Arch Surg 51:154-163, 1945

[26] Summers TB, Hines HGM: Effect of immobilization in various positions upon the weight and strength of skeletal muscle. Arch Phys Med 32:142-145, 1951

[27] Edstrom L: Selective atrophy of red muscle fibers in the quadriceps in long standing knee joint dysfunction: Injuries to the anterior cruciate ligament. J Neurol Sci 11:551-558, 1970

[28] Eriksson E: Sports injuries of the knee ligaments: Their diagnosis, treatment, rehabilitation, and prevention. Med Sci Sports 8: 133-144, 1976

[29] Johnson MA, Polgar J, Weightman D, et al: Data on the distribution of fibre types in thirty-six human muscles: An autopsy study. J Neurol Sci 18:111-129, 1973

[30] Sale DG, McComas AJ, MacDougall JD, et al: Neuromuscular adaptation in human thenar muscles following strength training and immobilization. J Appl Physiol: Respirat Environ Exercise Physio 53:419-424, 1982

V Vaughan, MS, PT, is Supervisor of Physical Therapy, Temple Physical Therapy, 60 Temple St, New Haven, CT 06510 (USA).

This article was adapted from a thesis completed in partial fulfillment of the requirements for Mr Vaughan's Master of Science degree in exercise science at the University of Massachusetts-Amherst.

This article was submitted October 21, 1986; was with the author for revision for 54 weeks, and was accepted August 9, 1988.