Low-Load Prolonged Stretch vs.

High-Load Brief
Stretch in Treating Knee Contractures

KATHYE E. LIGHT,
SHARON NUZIK,
WALTER PERSONIUS,
and AUBYN BARSTROM

This study was designed to compare the results of a traditional method of

stretching knee flexion contractures by high-load brief stretch (HLBS) with the
results of an experimental method of prolonged knee extension by skin traction,
low-load prolonged stretch (LLCS). End range of passive knee extension was
measured by standard goniometry. Subjects were 11 nonambulatory residents of
a nursing home who had demonstrated gradually progressive bilateral knee
contractures. Each subject served as his or her own control with one lower limb
receiving LLPS and the other limb receiving HLBS and passive range of motion
(PROM). Sequential medical trials were used as the clinical research design.
Whether comparing the LLPS limb PROM measurements pretreatment and post-
treatment (p ≤ .05) or the HLBS to the LLPS limb PROM recordings posttreatment
(p ≤ .05), the results demonstrated a preference for LLPS in the treatment of
knee contractures in the immobile nursing home resident.
Key Words: Contracture, Knee, Physical therapy.

Many elderly individuals demonstrate limited movement logical restriction of muscle length may be a consequence of
abilities; a frequent consequence is the development of knee spinal segment and supraspinal inputs on the gamma loop
contractures. Despite the efforts of an active physical therapy gain set mechanism.10-12 The result is an alteration of extra-
maintenance program at a county nursing home, including fusal muscle fiber resting length. Therapeutic exercise tech­
daily range of motion (ROM) and passive stretching tech­ niques have been hypothesized to affect this mechanism
niques, these chronic knee contractures continued to be a directly. The therapeutic techniques of contract-relax2 and
problem. This study was designed to compare a more tradi­ proprioceptive neuromuscular facilitation (PNF),3 as well as
tional method of stretching knee contractures, high-load brief the application offluorimethanespray and stretch,1 have all
stretch (HLBS), with an experimental method of prolonged been shown to assist a rapid improvement of restricted joint-
knee extension, low-load prolonged stretch (LLPS). range excursion.
Gross anatomical, histological, and mechanical data impli­
LITERATURE REVIEW cate abnormal intra-articular, periarticular, and extra-articu­
lar connective tissue structures as the cause of limited passive
Experimental and clinical data suggest that the tissue ROM (PROM). These structures include intra-articular adhe­
changes, which may cause restricted joint motion in the sions,5 periarticular joint capsule stiffness,4,6,9 and shortened
bedridden elderly, are physiological1-3 or morphological4-9 extra-articular skeletal muscle.7, 8 After four weeks of immo­
and involve intra-articular,5 periarticular4, 6, 9 and extra-artic­ bilization in a shortened position, cat soleus muscle demon­
ular structures.7, 8 strated a 40 percent decrease in the number of sarcomeres in
Neuromuscular dysfunction appears to be a common cause series7 and, therefore, a decrease in length of the parallel
of extra-articular physiological joint restriction. This physio­ elastic component.8 The decrease in cat soleus muscle length
resulted in a shift of its passive length-tension curve to the
Ms. Light is Assistant Professor, Physical Therapy Program, University of left and a concomitant decrease in ankle ROM. After four
Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San
Antonio, TX 78284 (USA). weeks release from immobilization (and normal activity by
Ms. Nuzik is Research Physical Therapist, Medical College of Virginia the cats in their cages during the interim), muscle extensibility
Hospitals, Virginia Commonwealth University, Richmond, VA 23298.
Mr. Personius is Assistant Professor, Department of Physical Therapy,
and sarcomere number were restored to normal.7 In another
School of Allied Health Professions, Medical College of Virginia, Virginia study, joints of monkeys immobilized in shortened positions
Commonwealth University. exhibited decreased ROM, decreased joint capsule length, and
Ms. Barstrom is a graduate student in the Department of Physical Therapy,
School of Allied Health Professions, Medical College of Virginia, Virginia loss of extensibility.6 Reduction of joint ROM and strength
Commonwealth University. of the medial posterior capsule was reported by Lavigne and
The results of this study were presented in poster format at the Annual Watkins to be appreciable at 16 days and marked after 32
Conference of the American Physical Therapy Association, Anaheim, CA, June
19-23, 1982, and at the 1982 State Meeting of the Virginia Chapter of the days of immobilization.6
American Physical Therapy Association, Williamsburg, VA, where Ms. Nuzik The mechanism for immobilization-induced capsular re­
was awarded Best Clinical Research Paper.
This article was submitted February 2, 1983; was with the authors for striction of ROM is not well understood and may be multi­
revision 25 weeks; and was accepted October 24, 1983. factorial. Histological studies of immobilized tissue have dem-

330 PHYSICAL THERAPY

 RESEARCH

onstrated capsular fibrofatty infiltrate,6 decreased matrix

water and glycosaminoglycans, increased major collagen
cross-links,4 and increased total amount of collagen.9
Warren et al used rat-tail tendons as an experimental model
for clinical stretching techniques.13 The rat-tail tendons were
used to compare two elongation methods similar to those
used in this study. Application of low-load, long-duration
tension at elevated temperatures, followed by cooling with
the load maintained, produced significantly greater residual
elongation with less tearing damage to the tissue than high-
load, short-duration tension at lower temperatures. The
present study included two of the four elements tested by
Warren et al: low-load and long-duration tension.
Existing literature supports LLPS as the preferred method
of lengthening immobilized, shortened tissues in animal
models. Because the very common clinical practice of stretch-
ing contractures with manual high loads for periods of a
minute or less was contradictory to findings in the literature, Fig. 1. Modified Buck's skin traction apparatus applied to the lower
this study sought to compare the PROM effects between the limb to provide LLPS.
clinical treatments of briefly applied manual high loads versus
prolonged low loads by skin traction. We hypothesized that if treatment time required for the HLBS and PROM was ap-
chronic knee flexion contractures of at least 30 degrees in proximately 15 minutes of patient-practitioner time.
nonambulatory geriatric nursing home subjects are treated Transmission of LLPS to the limb was accomplished by
with LLPS instead of HLBS, then the passive knee extension applying a modified Buck's skin traction technique (Fig. 1)
ROM will increase significantly. with the patient in bed. A foam strip measuring 60 in by 3 in
by 1 in* was applied to the subject's lower extremity in a
METHOD stirrup-like fashion. The central portion of the foam strip was
placed at the subject's heel; its length extended alongside the
Subjects
medial and lateral aspects of the limb. Taped to the center of
Criteria for admission to this study were 1) the presence of the foam and positioned at the subject's heel was a 3-in by 3-
bilateral knee flexion contractures of at least three-months in by 0.75-in padded wooden block. A screw eye extended
duration and at least 30 degrees short of full passive extension outward from the center of the block and served as an
and 2) the inability to walk or pivot transfer without maximal attachment for the pulley rope. The foam strip and block
assistance. Eleven geriatric subjects who were nonambulatory were secured to the limb by two or three 6-in ace bandages.
and who had demonstrated gradually progressive bilateral The rope was threaded through a single pulley and a weight
chronic knee-flexion contractures ranging from 30 degrees to was attached to the end. Plastic milk cartons filled with sand
132 degrees short of full passive end-range extension partici- were used as weights. The length of each LLPS treatment was
pated in this study. Subjects served as his or her own control one hour; the patient-practitioner setup time required only
with one lower extremity receiving the LLPS and the other two to five minutes. Each week the traction weight was
receiving a traditional combination of both HLBS and increased as follows: week one, 2.27 kg (5 lb); week two, 3.18
PROM. The choice of treatment for each limb was randomly kg (7 lb); week three, 4.08 kg (9 lb); and week four, 5.44 kg
determined. Each treatment was performed twice daily, five (121b).
days a week for four weeks. In addition to the two methods of stretching described, each
subject received a standard upper extremity and trunk pro-
Procedures gram of guided passive movements once daily. This program
included 10 repetitions of pelvis-on-trunk rotation and upper
The lower extremity not chosen to receive LLPS was treated extremity PNF diagonals bilaterally.
with a traditional regimen of 10 repetitions of passive lower Standard manual goniometric measurements of the limbs
extremityflexion,adduction, and external rotation using PNF were taken before treatment began and after four weeks of
diagonals that were followed by the HLBS. The HLBS pro- treatment. With the subject in the supine position, a physical
cedure was considered to be a routine, forced, passive, man- therapist moved one limb to its end range of knee extension
ual-stretching technique. A maximum manual force was used while also progressing toward maximal hip extension. This
that did not cause injury. The muscles on stretch were palp- end position was maintained for one minute. The knee-
ably very tight. With the LLPS, this was not the case. The extension range was then measured by a second physical
patient's limb was moved manually at a force and velocity therapist. The goniometer arms were aligned parallel to the
that allowed the soft tissue structures to accommodate to the long axis of the femur and the tibia with the axis at midknee
change in ROM without resulting in excessive resistance, pain joint. These two tasks were performed by the same individuals
on the part of the patient, or the possibility of injury. Once throughout the experiment.
the end range was achieved, this position was manually main-
tained for one minute. The force was then reduced for 15
seconds. The sequence of HLBS followed by a 15-second rest
was repeated three times each treatment session. The total * One inch = 2.54 cm.

Volume 64/ Number 3, March 1984 331

 TEST # I: Increase in PROM of LLPS limb > 15° ment (Fig. 2) or the increase in limb measurement of the
LLPS versus HLBS limbs posttreatment (Fig. 3), the results
demonstrated a preference (greater increase in PROM) for
LLPS in treatment of knee contractures of the immobile
nursing home resident. The Table offers a synopsis of the raw
data from this study.

DISCUSSION
A simple, noninvasive, nonstressful treatment is always
desirable, especially in an elderly group. In this sample, pa­
tients did not appear to experience pain while receiving LLPS
treatment. Trained physical therapy aides were instructed to
implement the exercise program and apply the traction ap­
paratus. Inconvenience to the nursing staff was minimal, and
the LLPS device did not impede primary care needs of the
patients, such as bathing and elimination.
Most of the subjects were characterized by a history of
progressive mental confusion, reduction of mobility, and
subsequent physical dependence. The use of traction as a
treatment for knee contractures in this patient group was
initiated long after the early physiological and possible mor­
phological effects of immobility had developed. The goals of

TEST# 2: PROM increase of LLPS limb at least 10° > PROM

Fig. 2. Sequential medical trials grid with a significance level of p ≤ increase of HLBS limb
.05 comparing pretreatment and posttreatment LLPS for Test 1
criterion. Statistical significance is established when the demarcation
lines are passed to the lower right or upper left of the grid.

Data Analysis
Sequential medical trials (SMT) were chosen to be the
method of experimental design and statistical analysis.14, 15
This sequential design was considered ideal because 1) sub­
jects could be admitted to the study as they became available,
2) comparisons could be made within the same subject, 3) no
statistical computations were necessary, and 4) the experiment
could be terminated as soon as statistical significance was
achieved (ie, a predetermined number of subjects was not
required).
The SMT plan was designed by Bross.14 Significance levels
were preestablished at the p ≤ .05 level. The outcome measure,
passive knee-extension end range, was assessed in two separate
test situations. Each test had its own criterion for improve­
ment: Test 1, difference between pretreatment and posttreat­
ment PROM recordings of the LLPS leg must be ≥15°; Test
2, difference in PROM posttreatment change between LLPS
and HLBS limbs must be ≥10°.

RESULTS Fig. 3. Sequential medical trials grid with a significance level of p ≤

.05 comparing posttreatment results of LLPS with HLBS for Test 2
A total of 11 subjects participated in this study before a p criterion.
≤ .05 significance level was attained with SMT. The SMT
grids for Test 1 (Fig. 2) and Test 2 (Fig. 3) demonstrate how
quickly and directly the LLPS was found to be the superior this treatment approach were to 1) reduce contractures to aid
treatment. In Test 1, when comparing pretreatment and post- in ease of nursing care and 2) increase active movement
treatment knee extension of the LLPS limb, 10 of the 11 abilities, including standing transfers.
subjects demonstrated greater than a 15-degree increase in Changes in activities of daily living (ADL) were not objec­
PROM. When comparing the HLBS with the LLPS limb in tively measured in this study. None of these subjects, however,
Test 2, a 10-degree PROM difference in favor of LLPS was became ambulatory, and all continued to require maximal
demonstrated in the same 10 subjects. Whether comparing assistance in transferring. Considering the severity and dura­
the LLPS limb measurements pretreatment and posttreat­ tion of the contractures, as well as the general mental and

332 PHYSICAL THERAPY

 RESEARCH

physical debilitation of these subjects, the reduction in con­ TABLE

tractures was probably inadequate to affect functional level. Passive Range of Motion Changes of Each Subject
With another group, ADL changes might have been used as
a criterion for improvement. Future investigations could con­ PROM Post Rx
Change
sider the effects of LLPS for early treatment of contractures (° from 180°) Difference
Sub­ Between
Between
in mentally alert hospitalized patients. ject Rx PreRx
LLPS
Individuals with severe joint restrictions, subjects 2 and 11 No. Pre Rx Post Rx and
and
in the Table, demonstrate greater improvement with the LLPS Post Rx
HLBS
treatment than those subjects whose initial restrictions were
1 LLPS - 50 - 46 4
less dramatic. The number of subjects and study procedures -14
HLBS - 75 - 57 18
cannot confirm this observation, but it does raise the question 2 LLPS -132 -102 30
of whether contractures of different degrees of severity re­ 16
HLBS -113 - 99 14
spond better to different stretching procedures. Other possible 3 LLPS - 95 - 80 15
explanations may lie in the morphological and physiological 10
HLBS -102 - 97 5
changes specific to a given disease process. Patients in this 4 LLPS - 99 - 70 29
16
study had a variety of diagnoses, as do most geriatric nursing HLBS - 80 - 67 13
home residents. 5 LLPS - 45 - 27 18
10
We presumed that the effects obtained in this study were HLBS - 48 - 40 8
the direct result of connective tissue lengthening. Alterations 6 LLPS - 82 - 64 18
12
HLBS - 70 - 67 3
in neurophysiological input, and consequently, the gamma
7 LLPS - 92 - 57 35
loop, may also have been a factor. Neurological changes occur HLBS - 60 - 56 4
31
rapidly and accommodate quickly, and, therefore, were not 8 LLPS - 69 - 50 19
measured or focused on in this study.12 The results of this HLBS - 68 - 74 - 6
25
study are in accordance with those of Warren et al who used 9 LLPS - 75 - 58 17
27
a rat-tail tendon model to study the effects of load on tissue HLBS - 60 - 70 -10
elongation.13 Low-load, long-duration tension produced 10 LLPS - 53 - 20 33
12
greater elongation of tissues than high-load, short-duration HLBS - 54 - 36 18
tension. Temperature, a variable not considered in this exper­ 11 LLPS -112 - 80 32
29
iment with humans, was found by Warren et al to enhance HLBS -105 -108 3
tissue elongation in rats.13
Sapega et al have published a treatment protocol designed
to increase ROM in the postreconstructed and immobilized p ≤ .05 significance level after only 11 subjects were assessed.
knee.16 Their protocol was based on principles evolved from The LLPS was found to be the more effective, acceptable,
the results of in vitro, connective-tissue studies in animals, nonstressful treatment to the patient over a four-week period.
including the study of Warren et al, and involved the use of
REFERENCES
heat in addition to low-load, long-duration tension. Their
recommended treatment sequence is 1) heating of the short­ 1. Halkovich LR, Personius WJ, Clamann HP, et al: Effect of fluorimethane
ened tissues, 2) applying the load while maintaining the spray on passive hip flexion. Phys Ther 61:185-189,1981
2. Medeiros JM, Smidt GL, Burmeister LF, et al: Influence of isometric
elevated tissue temperature, 3) maintaining both the load and exercise and passive stretch on hip joint motion. Phys Ther 57:518-523,
heat throughout the treatment period, 4) cooling the tissue 1977
3. Tanigawa MC: Comparison of the hold-relax procedure and passive mo­
below the normal body temperature before the load is re­ bilization on increasing muscle length. Phys Ther 52:725-735,1972
moved, and 5) removing the load. No experimental data were 4. Akeson W, Amiel D, Woo S: Immobility effects on synovial joints: The
presented. Sapega et al did not incorporate a traction com­ pathomechanics of joint contracture. Biorheoiogy 17:95-110,1980
5. Enneking WF, Horowitz M: The intra-articular effects of immobilization on
ponent into their stretching technique.16 We believed joint the human knee. J Bone Joint Surg [Am] 54:973-985,1972
mechanics to be an important consideration. We attempted 6. Lavigne AB, Watkins RP: Preliminary results on immobilization-induced
to apply the load in line with the tibia, so as to distract the stiffness of monkey knee joints and posterior capsules. In Perspectives in
Biomedical Engineerings: Proceedings of a Symposium. Biological Engi­
knee, and reduce compressive forces that may be created by neering Society, University of Strathyclyde, Glasgow, Scotland, Baltimore,
lengthening of imbalanced shortened tissues. The results of MD, University Park Press, 1972, pp 177-179
7. Tabary JC, Tabary C, Tardieu C: Physiological and structural changes in
this study may partially reflect the difference in treatment the cat's soleus muscle due to immobilization at different lengths by plaster
times between the two groups. A future study is planned to casts. J Physiol 224:231-244,1972
compare results of light-load versus heavy-load, skin-traction 8. Tardieu C, Tabary JC, Tabary C, et al: Adaptation of connective tissue
length to immobilization in the lengthened and shortened positions in cat
progression with equal treatment time for groups. soleus muscle. J Physiol 78:214-220,1982
9. Peacock E: Comparison of collagenous tissue surrounding normal and
immobilized joints. Surgical Forum 14:440-441,1963
CONCLUSION 10. Granit R, Burke RE: The control of movement and posture. Brain Res
53:1-28,1973
11. Kots YM: The Organization of Voluntary Movement: Neurophysiological
Chronic knee flexion contractures are a familiar clinical Mechanisms. New York, NY, Plenum Publishing Corp, 1977, pp 5-25
problem to most physical therapists. This study was per­ 12. Matthews PBC: Mammalian Muscle Receptors and Their Central Actions.
formed to compare the results of two methods of treatment Baltimore, M D, Williams & Wilkins, 1972, pp 143-192, 481-606
13. Warren CG, Lehman JF, Koblanski JN: Heat and stretch procedures: An
(LLPS vs HLBS) in increasing the passive knee extension evaluation using rat tail tendon. Arch Phys Med Rehabil 57:122-126,1976
ROM of knee flexion contractures in a group of elderly, 14. Bross I: Sequential medical plans. Biometrics 8:189-205,1952
15. Gonnella C: Designs for clinical research. Phys Ther 53:1276-1283,1973
nonambulatory, nursing-home patients. The SMT design al­ 16. Sapega AA, Quedenfeld TC, Moyer RA, et al: Biophysical factors in range-
lowed the obvious preference for LLPS to be attained at the of-motion exercise. Physician and Sports Medicine 9:57-65,1981

Volume 64/ Number 3, March 1984 333