Re\'iew Article

Sports Medicine 8 (2): 10 1-116, 1989

0112-1642/89/0008-0101/$08.00
© ADIS Press Limited
All rights reserved.
SPORT2187a

Isokinetic Dynamometry
Applications and Limitations

V. Baltzopoulos and D.A. Brodie

School of Movement Science, Physical Education and Recreation,
University of Liverpool, Liverpool, England.

Contents Summary .................................................................................................................................... 101

I. Definition of Isokinetics ....................................................................................................... 102
2. Gravitational Effect on Isokinetic Movements .................................................................. 102
3. Inertial Effect on Isokinetic Movements ............................................................................. 104
4. Isokinetic Parameters ............................................................................................................ 105
4.1 Maximum Torque ........................................................................................................... 105
4.2 Angular Position ............................................................................................................. 106
4.3 Torque-Velocity Relationship ........................................................................................ 106
4.4 Reciprocal Muscle Group Ratio .................................................................................... 107
4.5 Muscular Endurance ....................................................................................................... 109
5. Applications of Isokinetics ................................................................................................... 109
5.1 Rehabilitation and Assessment ...................................................................................... 109
5.2 Isokinetic Training .......................................................................................................... 112
5.3 Injury Prevention ............................................................................................................ 112
6. Computer Systems in Isokinetic Analysis ........................................................................... 112

Summary Isokinetic contraction is the muscular contraction that accompanies constant velocity
limb movements around a joint. The velocity of movement is maintained constant by a
special dynamometer. The resistance of the dynamometer is equal to the muscular forces
applied throughout the range of movement. This method allows the measurement of the
muscular forces in dynamic conditions and provides optimal loading of the muscles.
However, during movements in the vertical plane, the torque registered by the dyna-
mometer is the resultant torque produced by the muscular and gravitational forces. The
error depends on the angular position and the torque potential of the tested muscle group.
Several methods have been developed for the correction ofgravitational errors in isokinetic
data.
The torque output also contains artefacts that are associated with the inertial forces
during acceleration and deceleration periods before the development of the constant preset
angular velocity. For an accurate assessment oj muscle junction, only constant velocity
data should be analysed.
The most frequently used isokinetic parameters are the maximum torque and the an-
gular position where it was recorded, the torque output at different angular velocities of
movement, the torque ratio of reciprocal muscle groups and the torque output during re-
peated contractions.
102 Sports Medicine 8 (2) 1989

The unique features of isokinetic dynamometry are optimal loading of the muscles in
dynamic conditions and constant preselected velocity of movement. Thesefeatures provide
safety in the rehabilitation of patients with muscular and ligamentous injuries. 1sokinetic
dynamometry has also been used for the training of various muscle groups in order to
improve the muscular performance in dynamic conditions. The movement velocity of dif-
ferent activities can be simulated during training in order to improve the training effect.
Data acquisition and analysis have been improved by using computer systems inter-
faced to isokinetic dynamometers. Recently developed computer systems provide correction
for gravitational and inertial errors, accurate computation of isokinetic parameters and
real-time display of the torque output.

1. Definition of Isokinetics potential energy and therefore the return move-

ment does not require eccentric contraction to con-
The term 'isokinetics' is defined as the dynamic trol the return of the limb-lever arm system to the
muscular contraction when the velocity of move- initial position (Thistle et al. 1967).
ment is controlled and maintained constant by a Hislop and Perrine (1967) compared muscle
special device (Thistle et al. 1967). The resistance loading during isokinetic and isotonic (uncon-
of the device is equal to the applied muscular torque trolled velocity) testing. The load applied to the
over the range of movement. It is evident from the contracting muscles during isotonic movements is
definition that isokinetic movements require the maximal at points where the mechanical advan-
use of an electromechanical device capable of tage of the muscles is minimal (e.g. at the limits of
maintaining constant the velocity of movement. the range of movement in knee extension-flexion
Thistle et al. (1967) presented the isokinetic con- movements). On the other hand, during isokinetic
traction as a refinement of the controlled motion movements the resistance is equal to the muscular
concept, where the velocity of movement is no capacity and therefore muscle loading is maximal
longer an uncontrolled variable but may be preset at points where the mechanical advantage is max-
according to the specific functional activity of the imal. With the isokinetic method the maximum
contracting muscle groups. The velocity control muscular force that can be applied over a range of
mechanism of the dynamometer is usually an elec- movement can be measured in dynamic condi-
tronic servomotor or a hydraulic valve. The ve- tions, provided that the preset velocity has been
locity of movement is preset and the control mech- attained by the moving limb.
anism is activated only when the preset velocity is
attained by the moving limb. Any increase in mus- 2. Gravitational Effect on
cular torque above this level results in the devel- Isokinetic Movements
opment of an equal-magnitude resistive force by
the control mechanism of the dynamometer (Mof- During isokinetic tests involving movements in
froid et al. 1969). The muscular force varies at dif- the vertical plane (e.g. knee extension-flexion), the
ferent joint angles because of different biomechan- forces acting on the limb-lever system are the mus-
ical properties of the musculoskeletal system. With cular force (Fm) and the gravitational force (Fg)
the isokinetic method, if maximum force is applied generated by the mass of the limb and the lever
to the dynamometer over a range of movement, arm (fig. I). The torque registered by the dyna-
the resistance of the dynamometer is proportional mometer is not the actual muscular torque but the
to the muscular capacity at different joint angles, torque generated by the resultant of the muscular
offering optimal loading of the muscles in dynamic and gravitational forces (Herzog 1988; Winter et
conditions. Furthermore, isokinetic dynamome- al. 1981). Because the gravitational force remains
ters, unlike gravity-loaded systems, do not store constant for the same testing conditions, the per-
Isokinetic Dynamometry 103

centage error in the recorded torque depends on maximum torque over 50 knee extensions at 180
the magnitude of the muscular force applied. In degrees per second and was expressed as a per-
knee flexion movements the error is greater than centage of the initial maximum torque (Thorstens-
the error in extension because the hamstrings are son & Karlsson 1976). The error between corrected
usually less powerful than the quadriceps, while the and uncorrected fatigue indices ranged from -6.5
gravitational torque remains the same for both to 26% and the correlation coefficient was r=0.80,
movements. indicating that the error is not a constant factor
Winter et al. (1981) investigated the effect of since the maximum torque is produced at different
gravitational forces on the recorded torque by the joint angles as muscular fatigue increases. It was
dynamometer during movements in the vertical
suggested that the relationship between fatigue in-
plane. A correction factor was introduced to elim-
dex and relative distribution of fast twitch fibres
inate the gravitational error in the calculation of
as reported by Thorstensson et al. (1976) could
mechanical work generated by the muscular forces
substantially change if the data were corrected for
during knee extension-flexion movements. The
the effect of gravitational forces. The results of this
correction factor was the work generated by the
gravitational forces and it was determined using a study indicated the importance of gravity correc-
piezoresistive accelerometer placed on the lever arm tion in the assessment of muscle function with iso-
of the dynamometer. The magnitude of the grav- kinetic dynamometers. Nelson and Duncan (1983)
itational error was demonstrated by comparing the presented a simplified method for the computation
mechanical work computed from the torque re- of the gravitational torque during knee extension-
corded by the dynamometer with the mechanical flexion movements. This method required only the
work corrected for the effect of gravitational forces. recording of the gravitational torque generated by
In the above study, 4 subjects performed 2 minutes the weight of the limb-lever arm system at a spe-
of alternating knee extension and flexion on an iso- cific angular position within the range of move-
kinetic dynamometer at 20, 40 and 60 degrees per ment, while the limb-lever arm system is allowed
second. The error when the gravitational forces were to fall passively against the resistance of the dy-
not considered varied from 26 to 43% in extension namometer. The gravitational torque at every an-
and from 55 to 510% in flexion. gular position is then calculated and this correction
The effect of gravitational forces in the deter- factor is added to the maximum torque produced
mination of the fatigue index was also investigated. by muscle groups opposed by gravity (quadriceps
Fatigue index was defined as the mean decline in in the knee extension-flexion example) or sub-
tracted from the recorded torque produced by
muscle groups facilitated by gravity (e.g. ham-
strings).
This method is accurate and simpler than the
method proposed by Winter et aI. (1981), requiring
only the measurement of the gravitational torque
in a specific position within the range of move-
ment. However, in order to obtain valid results with
this method the muscles must remain fully relaxed
during the passive fall against the resistance offered
by the dynamometer. In practice, several trials
should be performed in order to obtain the actual
Fig. 1. Action of muscular (Fm) and gravitational (Fg) forces gravitational torque, typically the minimum torque
during isokinetic knee extension testing. value recorded from the repeated trials.
104 Sports Medicine 8 (2) 1989

3. Inertial Effect on Isokinetic Movements a resistive force is exerted by the dynamometer, in

order to decelerate the limb to the level of the pre-
The torque output during isokinetic movements set velocity (fig. 3). The overshoot in the torque
frequently contains a prominent initial spike, which output represents this 'reaction' of the dynamom-
may be followed by torque oscillations of decreas- eter to the overspeeding limb-lever arm. Sapega et
ing amplitude (Sapega et al. 1982). This phenom- al. (1982) reported that in the hip abduction test
enon is usually referred to as the 'torque overshoot' the torque overshoot occurred during this deceler-
and always appears in the initial part of the move- ation period. It was calculated that the torque ov-
ment (fig. 2). ershoot was the torque required by the dynamom-
The feedback mechanism of the dynamometer
eter to produce the deceleration of the limb-lever
is not activated if the velocity of movement is lower
system.
than the preset angular velocity. During this period
The torque of a rotating system is proportional
the limb is free to accelerate as there is no resistive
force exerted by the dynamometer (fig. 3). Subse- to the angular acceleration and the moment of in-
quently the velocity of the limb is increased above ertia of the system. During proximal joint testing,
the preset angular velocity. Sapega et al. (1982) where a greater limb mass and a longer distance
filmed 2 tests on an isokinetic dynamometer using between the axis of rotation and the centre of mass
inert weights and a hip abduction movement. An- are involved, the magnitude of the torque over-
alysis of the high-speed film revealed that during shoot increases. Another factor affecting the mag-
this free acceleration period, the angular velocity nitude of the torque overshoot is the mass of the
exceeded the preset velocity by II % and 200% in dynamometer lever arm used for the test. The dur-
the inert weights and hip abduction tests, respec- ation of the acceleration period is affected by the
tively. When the feedback mechanism is activated, level of the present angular velocity and the power

300

150
E
~
Q)
::>
c-
o
I- 0

<:
,g

ii]~~~~ Time

Fig. 2. Torque overshoot during knee extension-flexion movements. The angular velocity and acceleration of the limb-lever
arm system in the initial part of a knee extension movement (boxed area in figure) are illustrated in figure 3.
lsokinetic Dynamometry 105

N of movement and during the acceleration period.

!'" \
The resisted acceleration method allowed a smooth
~ \ transition from the acceleration to the constant ve-
g> 500 \ locity phase, with minimal torque oscillations. The
:s \,
c
o '-, acceleration period, however, was increased and the
.~ .... _-------- preset velocity was attained later in the range of
~ 100
~ o+-+-+--+~~~~~~~~ movement.
~
Another method to overcome the inertial arte-
~ Cll
"5
0>
=> fact is to use torque data only from constant ve-
c ~
« ~~~~--------------~O~ locity periods of the movement (Osternig et al.
0.1 1982; Perrine & Edgerton 1978). Since oscillations
Time (sec)
in the torque output represent alternating periods
Fig. 3. Torque (---), angular velocity ( - - ) and an- of acceleration and deceleration, artefact-free data
gular acceleration ( .... ) of the limb-lever arm system dur- can be obtained from the portion of the movement
ing knee extension. The angular velocity of the dynamometer
was preset at 30 o/sec. Notice that the preset velocity is ex-
where the angular velocity remains constant and
ceeded by the velocity of movement during the free accel- equal in magnitude with the preset velocity setting
eration period and the torque overshoot is the torque re- of the dynamometer.
quired to decelerate the limb. The velocity of movement
becomes constant and equal to the preset velocity after a 4. Isokinetic Parameters
series of acceleration and deceleration periods.
4.1 Maximum Torque

of the muscle group involved relative to the mass The maximum torque during isokinetic move-
of the limb and the dynamometer lever arm. ments is a measure of the muscular force applied
During isokinetic testing, the overshoot is fre- in dynamic conditions. Various testing protocols
quently the peak-point in the torque output. If this have been used for the assessment of maximum
peak is interpreted as the subject's maximum torque. The main difference between these proto-
torque, the muscular capability will be overesti- cols is the number of repetitions required in order
mated, influencing bilateral comparisons and re- to develop the maximum torque. Sawhill et al.
ciprocal muscle group ratios. (1982) investigated the number of repetitions re-
The damp of the torque signal is a method that quired to achieve stable measurements during iso-
has been used to control the torque overshoot. Sin- kinetic testing at angular velocities ranging from
acore et al. (1983) investigated the effect of damp 200 to 400 degrees per second. They suggested that
on isokinetic measurements and they reported that 4 maximal repetitions are required in order to ob-
the damp resulted in a reduction of the torque sig- tain stable isokinetic data. Johnson and Siegel
nal amplitude throughout the range of movement (1978) reported that 3 submaximal followed by 3
and a displacement of the torque curve in the time maximal repetitions are essential fOf stable isoki-
axis. The effects of the damp method introduce er- netic data in knee extension movements. Appen
rors in the maximum torque measurement and the and Duncan (1986) investigated the knee extensor
torque-position relationship. Signal damp is there- and flexor muscles using 5 submaximal followed
fore not an effective method for the elimination of by 3 maximal repetitions. The testing protocol for
the inertial artefact (Bemben et al. 1988; Murray the measurement of maximum torque of the knee
1986). extensors and flexors used by Jenkins et al. (1984)
Gransberg and Knutsson (1983) connected a consisted of 5 maximal reciprocal (i.e. extension
computer to the velocity control mechanism of a followed by flexion) repetitions, whereas Dibrezzo
dynamometer in order to increase the acceleration et al. (1985) used only 2 maximal repetitions. Bal-
period. The limb was resisted before the initiation tzopoulos et al. (1988) used 6 reciprocal repetitions
106 Sports Medicine 8 (2) 1989

for the measurement of maximum torque in knee

extension-flexion movements. 250-
It is evident from the above studies that max-
imum torque is always evaluated from the first 2 200-
to 6 maximal repetitions and is defined as the max-
imum single torque value measured during these 150-

repetitions. E
However, Patton and Duggan (1987) defined 6100-
~
maximum torque as the mean torque from 5 max- 2"
imal repetitions and Morris et al. (1983) used the := 50-

mean of 3 repetitions. The maximum torque de-

O~~ __ ~~~-L~~ __~LJ~~~_
pends on the angular position (i.e. the joint posi- 30 60 90 120 150 180 210 240
tion) where it was recorded (Caiozzo et al. 1981; Angular velOCIty (degrees/sec)
Osternig 1975; Osternig et al. 1983; Thorstensson
et al. 1976). The mean torque calculated from Fig. 4. Maximum torque at different angular velocities of
knee extension (1lII) and flexion (_) [data from Baltzopoulos
torque values recorded at different angular posi-
& Brodie 1987].
tions is not a meaningful measure of muscle func-
tion, because there is no information about the an-
gular position. This method is useful only when the extension tended to converge near the 60 degree
torque value is recorded at a specific predeter- position. Moffroid et al. (1969) also reported that
mined angular position in every repetition. In this the optimal position in extension was at 63 de-
case, however, the recorded torque at the prede- grees.
termined specific angular position may not be the However, with increasing velocity, the acceler-
maximum torque in that repetition. ation period before the activation of the resistive
mechanism of the dynamometer is longer and the
4.2 Angular Position limb may pass past the optimal position during this
period. As a result the maximum torque tends to
The angular position IS Important in the as- occur later in the range of movement with increas-
sessment of muscle function because it provides ing velocity and not in the optimal joint position.
information about the mechanical properties of the Consequently, analysis of maximum torque data
contracting muscles. It can be used to evaluate the irrespective of angular position may lead to erron-
optimum joint angle for maximum muscular force. eous conclusions about muscle function.
The maximum torque position is affected by the
angular velocity of movement. Thorstensson et al. 4.3 Torque-Velocity Relationship
(1976) reported that during knee extensions the
maximum torque occurred later in the range of The muscular torque exerted during isokinetic
movement as the preset angular velocity increased. testing decreases with increasing angular velocity
Osternig et al. (1983) reported a transfer in the flex- of movement (Barnes 1980; Campbell 1979; Gre-
ion maximum torque position from 32 to 61 de- gor et al. 1979; Moffroid et al. 1969; Osternig et al.
grees of knee flexion during an increase from 50 to 1983; Thorstensson et al. 1976; Yates & Kamon
400 degrees per second, respectively. The transfer 1983) [fig. 4]. This decline in torque output has
observed in the extension maximum torque posi- been attributed to different neurological activation
tion was from 87 to 63 degrees of knee flexion dur- patterns of motor units at different velocities
ing an increase from 50 to 400 degrees per second. (Barnes 1980; Milner-Brown et al. 1975). Moffroid
They also reported that with increasing velocity the et al. (1969) recorded the torque in knee extension
maximum torque optimal position in flexion and movements at a specific position (65 degrees of knee
Isokinetic Dynamometry 107

flexion). With the velocity of movement increasing portant implications in comparisons between the
from 0 to 108 degrees per second, they reported a in vitro and in vivo force-velocity relationship. The
decrease in the torque output. However they ob- velocity in the in vitro curve represents the actual
served an initial plateau in the torque output be- velocity of the contraction, whereas the velocity in
tween 0 and 36 degrees per second. This plateau the in vivo curve represents the velocity of the
was attributed to possible human subject reluct- moving limb under the influence of the contracting
ance to exert more force at the slower velocities. muscle. Hinson et al. (1979) reported that during
Perrine and Edgerton (1978) tested the torque elbow flexion and with the lower arm moving with
of the knee extensors at angular velocities of move- constant angular velocity, the contraction velocity
ment ranging from 0 to 288 degrees per second. of the elbow flexors is not constant but contains
The torque was recorded at an angle of 70 degrees only periods of acceleration and deceleration. They
of knee flexion, in order for the muscle to develop concluded that the term 'isokinetics' denotes the
maximum tension and attain the preset velocity. type of muscular contraction which accompanies
An intial plateau in the torque output was ob- constant angular velocity movements and not con-
served between 0 and 144 degrees per second and stant velocity of muscular contraction.
then the torque decreased with increasing velocity. Another problem in the in vivo and in vitro force-
Lesmes et al. (1978) tested the maximum torque velocity comparison is the angular position of the
of the knee extensors and flexors irrespective of maximum torque during isokinetic testing. Theo-
angular position at angular velocities ranging from retically the maximum torque in the in vivo testing
oto 300 degrees per second. The maximum torque is generated at a joint angle where the contracting
decreased with increasing velocity, but they also muscle has an optimal mechanical advantage, pro-
reported an initial plateau in the torque output be- vided that the muscle has developed maximum
tween 0 and 60 degrees per second for both exten- tension. Since it takes a finite amount of time for
sion and flexion movements. In the above studies individual muscle fibres to develop maximum ten-
the obtained torque-velocity curve was compared sion, the decrease in torque with increasing angular
to the classical in vitro force-velocity curve (Fenn velocity could be a reflection of the muscle's in-
& Marsh 1935; Hill 1938). The in vivo isokinetic ability to develop maximum tension at the optimal
torque-velocity curve was similar to the in vitro joint angle (Coyle et al. 1979). Increasing angular
hyperbola at higher velocities of movement. In velocity would position the limb away from the
lower velocities, however, a plateau was observed optimal joint angle, when the muscle develops
in the torque output, whereas in the in vitro curve maximum tension.
an increase in force occurs with decreasing veloc- Despite these problems, the torque-velocity re-
ity. This difference was attributed to a neural lationship during isokinetic testing provides im-
mechanism which limits the muscle tension de- portant information about muscle function at dif-
velopment in lower velocities of movement during ferent movement velocities, especially when the
isokinetic evaluation of the torque-velocity rela- muscle function is assessed in relation to the ve-
tionship (Perrine & Edgerton 1978). However, Par- locity of a particular activity.
ker et al. (1983) tested knee extension at 54, 108,
162, 216, 270 and 300 degrees per second and con- 4.4 Reciprocal Muscle Group Ratio
cluded that the quadriceps torque-velocity rela-
tionship observed was in accordance with the Hill The reciprocal muscle group ratio is an indi-
equation. cator of muscular balance or imbalance around a
The Hill equation was derived from experi- joint. The hamstring to quadriceps ratio of the knee
ments with animal muscles free of the joint and joint is one of the more important parameters in
therefore the force was acting in the same line as isokinetic assessment because the knee is one of
the actual tension development. This has very im- the largest and most complex joints in the human
108 Sports Medicine 8 (2) 1989

body and its normal function is important for in-

jury prevention. It has been suggested that the
hamstring to quadriceps ratio is more important ~~~~----------------~~90·

than the maximum torque in the assessment of

muscle function (Campbell & Glenn 1982). Goslin
and Charteris (1979) tested the knee extension- B
flexion movement of 60 untrained subjects at 30
degrees per second and reported a hamstring to
quadriceps ratio of 0.44. Gilliam et al. (1979) tested
high school football players at 30 and 180 degrees
per second and found hamstring to quadriceps ra-
tios of 0.60 and 0.77, respectively. Scudder (1980)
tested the knee extensors and flexors of 10 normal
untrained subjects and reported an increase in the
hamstring to quadriceps ratio from 0.56 to 0.62 with O·
an increase in the angular velocity from 0 to 72
degrees per second. A similar increase was reported Fig. 5. The gravitational torque in position B, Tb = Fg • db
and is greater than the torque in position A, Ta = Fg • da
by Davies et al. (1981) using professional football
because db> d a while the gravitational force Fg remains the
players. The ratio was increased from 0.61 at 45 same throughout the range of movement.
degrees per second to 0.80 at 300 degrees per sec-
ond. Wyatt and Edwards (1981) reported a similar
increase with female subjects, from 0.71 at 60 de- at 60, 180, 240 and 300 degrees per second. The
grees per second to 0.85 at 300 degrees per second. results of this study demonstrated that although the
Housh et al. (1984) reported that the hamstring to uncorrected ratios were similar to previous studies,
quadriceps ratios in female throwers, jumpers, indicating an increased hamstring to quadriceps ra-
middle distance runners and sprinters were 0.70, tio with increasing angular velocity, the gravity-
0.75,0.81 and 0.71, respectively, at 180 degrees per corrected ratios remain constant with increasing
second. Dibrezzo et al. (1985) reported that the angular velocity. The error in the computation of
mean ratio of 241 females between the age of 18 the ratio, with data not corrected for the effect of
and 28 years was 0.54 at 60 degrees per second. gravity, increased from 18.5% at 60 degrees per sec-
It is evident from the above studies that ham- ond to 37.7% at 300 degrees per second.
string to quadriceps ratio is affected by age, sex and The error increase can be explained by the dif-
activity. It is also evident that the ratio is increased ferent angular position of the maximum torque with
with an increase in the angular velocity of move- increasing angular velocity (Osternig et al. 1983;
ment, indicating a possible decline in the relative Thorstensson et al. 1976). The maximum torque
quadriceps activity. However, it is important to is generated at increased knee joint angle with in-
note that the isokinetic data in the above studies creasing velocity. The gravitational torque also in-
were not corrected for the effect of gravity. creases with increasing knee joint angle because the
Schlinkman (1984) reported that the hamstring horizontal distance between the centre of mass of
to quadriceps ratio of high school football players the limb-lever arm system and the vertical axis of
increased from 0.54 at 60 degrees per second to the dynamometer is increasing (fig. 5). In order to
0.67 at 300 degrees per second, but when the ex- compute the gravity-corrected hamstring to quad-
tension and flexion torque was corrected for the riceps ratio the gravitational torque is added to the
effect of gravity, the ratio was decreased by 8 to denominator (quadriceps) and subtracted from the
12%. Appen and Duncan (1986) computed the cor- numerator (hamstrings) resulting in a decrease of
rected and uncorrected ratio in male track athletes the ratio magnitude. At decreased knee joint angles
Isokinetic Dynamometry 109

the gravitational torque is minimal and the error tween fatigue indices measured by the 2 tests.
is smaller. With increased knee joint angle, the However, the isokinetic data were not corrected for
gravitational torque increases, resulting in a further the effect of gravity. Baltzopoulos et al. (1988) de-
decrease of the hamstring to quadriceps ratio and fined fatigue index as the decline in maximum
a greater error. torque over time, using 30 seconds of repeated re-
It is evident from these studies that the inter- ciprocal contractions with gravity-corrected data.
pretation of the reciprocal muscle group ratio with- The results of this test were compared with the fa-
out considering the gravity effect results in erron- tigue index from the 30-second Wingate test. A sig-
eous conclusions about muscle function (Fill yaw et nificant correlation (r=0.86, p<O.OOI) was found
al. 1986). Consequently, conclusions of previous between the fatigue indices from the two tests.
studies with data uncorrected for the effect of grav- The difference in angular position of the max-
ity must be treated with caution, because the effect imum torque and the reduction of the angular ve-
of the gravitational error in the validity of the re- locity with muscle fatigue may have an effect in
sults is unknown. the computation of fatigue index. The work per-
formed is a more representative measure of muscle
4.5 Muscular Endurance function because it takes into account the force
output throughout the range of movement. How-
Muscular endurance is the ability of the con- ever, Burdett and Swearingen (1987) computed the
tracting muscles to perform repeated contractions ratio of the work produced during the last 5 of 25
against a load. The muscular endurance in dy- maximal contractions to the work during the first
namic conditions using isokinetic dynamometers 5 and reported that the reliability of the work ratio
is assessed by computing a fatigue index. However, was low and that the number of contractions to
different testing protocols and definitions have been 50% of the initial torque level was a more reliable
used for the determination of the fatigue index. The measurement of muscular endurance.
testing protocol used by Thorstensson and Karls-
son (1976) consisted of 50 maximal contractions 5. Applications of Isokinetics
of the knee extensors. Muscular endurance was as- 5.1 Rehabilitation and Assessment
sessed by expressing the mean torque from the last
3 contractions as a percentage of the mean torque The advantages of isokinetic systems include
from the initial 3 contractions. Patton et al. (1978) variable resistance equal to the applied muscular
investigated the shape of fatigue curves using re- force, and constant preselected velocity of move-
peated contractions to exhaustion. Fatigue index ment. These unique features provide safety when
was expressed as the time required for muscular used for rehabilitation of patients with muscular
exhaustion. Barnes (1981), in a similar study, used and ligamentous injuries and accuracy in the as-
a testing protocol consisting of 10 maximal con- sessment of muscular performance at different
tractions and the fatigue index was computed by functional velocities of movement.
expressing the maximum torque in the last con- The purpose of rehabilitation programmes fol-
traction as a percentage of the maximum torque lowing injury or surgery is to restore normal muscle
during the 10 contractions. It is evident from the function of the affected limb. However, the force-
above studies that there is no standardised testing velocity relationship during isokinetic movements
protocol and definition for the fatigue index and and the velocity specific training effects on mus-
the assessment of muscular endurance. Patton and cular strength reported for normal subjects (Caiozzo
Duggan (1987) examined the relationship between et al. 1981; Coyle et al. 1981; Jenkins et al. 1984;
the muscular endurance test introduced by Thor- Parker et al. 1983) had a considerable effect on the
stensson and Karlsson (1976) and the 30-second selection of training velocity in rehabilitation pro-
Wingate test. No relationship was reported be- grammes. Parker (1982) proposed the use of an ap-
110 Sports Medicine 8 (2) 1989

propriate velocity according to the condition of the different from the respective ratios of the healthy
injured muscle. The velocity was calculated by subjects. The test-retest reliability for patients with
substituting in Hill's equation for the force-velocity multiple sclerosis was 0.99 (p<O.OOI) with both tests
relationship the maximum isometric torque that a performed in the same week. However, the maxi-
patient is able to exert. Sherman et al. (1982) rec- mum torque was variable after 6 and II weeks and
ommended rehabilitation velocities ranging from it was suggested that when such patients are not
60 to 300 degrees per second in order to ensure familiar with isokinetic equipment, an increase in
that both muscle fibre types were recruited and the maximum torque may not reflect an improve-
trained. Grimby (1985) suggested that the training ment in the functional condition, but a learning
velocity should depend on the phase of rehabili- effect or familiarisation with the isokinetic appa-
tation, type and degree of muscular hypotrophy and ratus. Watkins et al. (1984) examined 15 hemipar-
individual reaction at different velocities. etic patients and 15 healthy subjects. They per-
Campbell and Glenn (1982) assessed the effect formed 5 bilateral consecutive repetitions of the
of rehabilitation programmes for patients with knee extensors and flexors muscles at 30 degrees
chondromalacia, ligamentous repairs and menis- per second in order to evaluate the maximum
cectomies with isokinetic testing. An isokinetic dy- torque and hamstring to quadriceps ratio. The
namometer was used to evaluate the maximum maximum torque of the unaffected side of the
torque and hamstring to quadriceps ratio at 30 and patients was significantly lower than in healthy
180 degrees per second, before and after the re- subjects and furthermore the maximum torque of
habilitation programme of the affected and unaf- the affected side was significantly lower than the
fected limb. Although the rehabilitation pro- unaffected side. The accuracy of isokinetic testing
gramme consisted of isometric contractions and in detecting muscle function deficiencies was docu-
functional activities of the affected limb, a signifi- mented by evaluating the muscle function of the
cant increase in the isokinetic maximum torque was affected side of the patients with manual muscle
reported. The isokinetic test revealed that the ex- testing. Although the maximum torque and
tension maximum torque and the hamstring to hamstring to quadriceps ratio of the affected side
quadriceps ratio were not rehabilitated to the levels using isokinetic dynamometry were significantly
of the unaffected limb but the opposite was found lower than healthy subjects, the recorded grades of
for the flexion maximum torque. manual testing were 'good' to 'normal', indicating
Armstrong et al. (1983) investigated the relia- the superiority of isokinetic dynamometry in de-
bility and safety features of isokinetic dynamo- tecting muscle function deficiencies.
metry in patients with multiple sclerosis. The max- Burnie and Brodie (1986) assessed the effec-
imum torque and hamstring to quadriceps ratio of tiveness of a rehabilitation programme for knee in-
the right knee were evaluated for 10 patients and jury using isokinetic dynamometry. Muscle func-
20 healthy subjects at angular velocities ranging tion of the knee extensors and flexors of a
from 0 to 270 degrees per second. In order to assess professional football player was assessed with an
the reliability of isokinetic dynamometry, the max- isokinetic dynamometer 12 weeks after an injury
imum torque of 3 patients was evaluated after 0, which involved the medial collateral ligament and
6 and II weeks. The results demonstrated that the both the anterior and posterior cruciate ligaments.
maximum torque of patients with multiple scle- Bilateral testing of the knee extensors and flexors
rosis was significantly lower than the maximum was performed at 60 degrees per second 12, 20 and
torque of healthy subjects, although the torque 27 weeks after surgery. The use of an isokinetic
curves were similar in shape. The maximum torque dynamometer was also included in the rehabilita-
output of 50% of the patients at 270 degrees per tion programme during this period. The results in-
second was 0 N· m. Hamstring to quadriceps ra- dicated a significant increase in extension and flex-
tios at all angular velocities were not significantly ion maximum torque of the operated knee (304%
Isokinetic Dynamometry 111

in flexion and 344% in extension), reducing the bi- with the gravitational torque and not 0 N • m as
lateral deficit from 52 to 16% in flexion and from reported, or it was less than the torque signal res-
70 to 26% in extension. The range of movement olution of the isokinetic system.
was increased from 40 to 106 degrees and the Isokinetic dynamometers have also been used
hamstring to quadriceps ratio was improved from to assess the effects of injuries on muscle function
I to 0.87 after the rehabilitation programme. and the effect of various treatment and rehabili-
Similar improvements were reported by Tho- tation techniques. Among other applications, iso-
mee et al. (1987) after rehabilitation of patients with kinetic dynamometry has been used to examine the
anterior cruciate ligament injury. The maximum synergetic action of the anterior cruciate ligament
torque of the knee extensors and flexors at 30, 60, and the thigh muscles in maintaining joint stability
120, 180 and 300 degrees per second of 16 patients
(Solomonow et al. 1987), to assess muscle function
was evaluated before and after a rehabilitation pro-
and evaluate rehabilitation programmes for knee
gramme of 8 weeks. The rehabilitation programme
ligament injuries (Grimby et al. 1980; LoPresti et
consisted of knee extension and flexion at 60 and
al. 1988; Murray et al. 1984; Noyes et al. 1987), to
180 degrees per second using an isokinetic dyna-
examine muscle function after bilateral femoral os-
mometer. After the rehabilitation programme the
operated knee extension maximum torque in- teotomy (Olerud et al. 1984) and for arthroscopic
creased from 56 to 74% and the flexion maximum meniscectomy with and without tourniquet control
torque from 78 to 102% compared with the non- (Thorbland et al. 1985). It has also been used to
operated knee. evaluate the efficiency of a rehabilitation pro-
The results of the above studies indicate that gramme after arthroscopic meniscectomy (Shields
isokinetics is an effective rehabilitation method and et al. 1987), to assess the function of the knee ex-
is also of value for rehabilitation assessment. It is tensors and flexors after diagnostic and operative
also evident that the most frequently used isoki- arthroscopy and open meniscectomy (Hamberg et
netic parameters in the assessment of muscle func- al. 1983), to examine the effect of patella brace on
tion are maximum torque and reciprocal muscle quadriceps torque (Lysholm et al. 1984), to ex-
group ratio. However, the magnitude of errors in amine the results of transcutaneous neural stimu-
the evaluation of these parameters if the isokinetic lation after arthroscopic knee surgery (Jensen et al.
data are not corrected for gravitational and inertial 1985) and to assess muscle function after lateral
effects (Sapega et al. 1982; Winter et al. 1981) dem- reconstruction for anteriolateral rotary instability
onstrate the importance of appropriate filters in or- of the knee (Fleming et al. 1983). Mira et al. (1980)
der to eliminate potential errors. Furthermore, the examined the shape of the isokinetic quadriceps
maximum torque of an injured or operated joint torque in order to determine the type of femoral
is very low, increasing further the magnitude of the
shaft fracture and the level of injury. Knutsson and
percentage gravitational error. A typical example
Martensson (1985) used isokinetic measurements
is the previously reported result by Armstrong et
to examine the origin of hysterical paresis. Treat-
al. (1983) that many patients with multiple scle-
ment methods for achilles tendon injuries have also
rosis were unable to produce extension and flexion
maximum torque greater than 0 N· mat 275 de- been evaluated using isokinetic dynamometry (Be-
grees per second. Assuming that the limb was mov- skin et al. 1987; Inglis et al. 1976; Nistor 1981;
ing in extension for example with a constant ve- Pierre et al. 1984) and it has also been used for
locity of 275 degrees per second, it is evident that postoperative evaluation of shoulder dislocation
the knee extensors were generating force and thus (Miller et al. 1984) and assessment of trunk exten-
a finite amount of torque was applied to the dy- sors and flexors in normal and low back dysfunc-
namometer, but 0 N • m was recorded. In this case tion patients (Kishino et al. 1985; Mayer et al. 1985;
the muscular torque was either equal in magnitude Smidt et al. 1983).
112 Sports Medicine 8 (2) 1989

5.2 Isokinetic Training knee maximum torque or an imbalance in the

hamstring to quadriceps ratio was not associated
The constant preselected velocity during isoki- with increased incidence of knee joint injury. Pre-
netic movements allows the training and improve- season maximum torque and hamstring to quad-
ment of muscular performance in dynamic con- riceps ratio of 206 male high school football play-
ditions. Isokinetic training at a specific angular ers were evaluated with an isokinetic dynamometer
velocity increases the maximum torque of the in- at 60 and 240 degrees per second.
volved muscle groups at the training velocity Maximum torque imbalance was defined as a
(Lesmes et al. 1978). A transfer effect at other ve- difference between right and left knee of 10% or
locities (i.e. increased maximum torque at lower more. Hamstring to quadriceps ratio imbalance was
and higher velocities than the training velocity) has defined as the difference between the mean and the
also been reported (Coyle et al. 1981; Lesmes et al. actual ratio of 10% or more. Although an imbal-
1978). In these studies it was reported that maxi- ance was detected for 33% of the tested subjects,
mum torque increased significantly at the training no relationship was found between imbalance and
velocity and velocities below the training velocity. joint injury susceptibility. However, further re-
It was also reported that high velocity training has search is needed to examine the relationship be-
a better transfer effect to lower velocities than low tween muscle imbalance assessed with the isoki-
velocity training to higher velocities of movement. netic method and injury (Grace 1985).
Caiozzo et al. (1981) reported that high velocity
training (240 degrees per second) produced in- 6. Computer Systems in
creased maximum torque at lower velocities with Isokinetic Analysis
an exception of 30 degrees per second. Jenkins et
al. (1984) reported that training at 240 degrees per Manual analysis of isokinetic data involves the
second produced improvements at 240 and 300 de- computation of the isokinetic parameters from the
grees per second, while training at 60 degrees per torque graph printed on a chart recorder. This
second produced improvements at 60 and 180 de- method involves basic measurement techniques
grees per second. Garnica (1986) reported that im- and can be time consuming and inaccurate. Fur-
provements after low velocity training (60 degrees thermore the implementation of appropriate filters
per second) occurred at a higher velocity (180 de- for the gravitational and inertial artefacts is re-
grees per second) and that high velocity training stricted because of the amount and complexity of
increased the maximum torque at the training ve- the mathematical computations involved (Watkins
locity only. et al. 1984). The development of computer systems
The improvement in muscular performance after interfaced to isokinetic dynamometers provides a
isokinetic training has been explained by velocity- solution to the above problems and enhances the
specific adaptation of motor units within the muscle efficiency and accuracy of isokinetic dynamometry
(Milner-Brown et al. 1975; Sale et al. 1983) and for training and rehabilitation.
velocity-specific adaptation within the nervous Richards and Cooper (1982) described the in-
system (Barnes 1980; Sale et al. 1982). However, terface of an Apple III microcomputer to a Cybex
differences in the direction of the transfer effect can II isokinetic dynamometer. The isokinetic para-
be explained by differences in sample size, muscle meters computed from the isokinetic data include
fibre distribution and training period and intensity. maximum torque, work, power, reciprocal muscle
5.3 Injury Prevention group ratio and range of movement. In order to
avoid interpretation of torque overshoot as mus-
In contrast to previous findings (Heiser et al. cular torque, data sampled at the first 0.01 seconds
1984; Mulder 1973; Slagle 1979), Grace et al. (1984) of the movement were not included in the analysis.
reported that an imbalance between right and left Data analysis time is approximately 10 seconds.
Isokinetic Dynamometry 113

300

150

E
~
CD
:>
e-o
I- 0

Time

Fig. 6. Real-time display of the gravity-corrected torque and angular position during a knee extension-flexion test. Notice
that at the end of extension movements a torque amount of about 30 N· m is registered by the system, representing the
muscular torque required to maintain the limb-lever arm system in this upright position. The negative values at the start of
flexion movements indicate that the muscular torque is applied in the opposite direction. Compare also with figure 2 where
the torque output is not gravity-corrected.

The reliability of the system was determined by the ond from 6 repetitions or at 180 degrees per second
intraclass correlation coefficient for the computa- from 20 seconds of continuous repetitions were
tion of torque, work and power. The reliability implemented in the program. After data input
coefficients were greater than 0.99 (p<O.OOI), in- completion the program evaluates maximum
dicating reliable measurement of the isokinetic torque, power, reciprocal muscle group ratio and
parameters. several timing parameters.
Ostemig et al. (1982) developed a computer sys- More recently Baltzopoulos (1988) has devel-
tem for data acquisition and analysis from a mod- oped a computer system for the Akron isokinetic
ified Orthotron isokinetic dynamometer. The an- dynamometer which displays the gravity-corrected
gular velocity of movement is computed from the torque and the angular position in real time (fig.
angular position data. With this method accelera- 6) and corrects the data for inertial errors before
tion and deceleration phases can be identified, al- the computation of the isokinetic parameters de-
lowing the evaluation of maximum torque from scribed previously.
constant velocity data. The replacement of manual data acquisition and
Another computer system for the Cybex dyna- analysis using computer systems, has reduced an-
mometer was developed by Potash et al. (1983). An alysis time and computational error allowing the
Apple II microcomputer was interfaced to the dy- implementation of correction methods for any
namometer. Two testing protocols for the evalua- gravitational or inertial errors, and therefore en-
tion of isokinetic parameters at 30 degrees per sec- hancing the accuracy of isokinetic measurements.
114 Sports Medicine 8 (2) 1989

Fillyaw M, Bevins T, Fernandez L. Importance of correcting iso-

Acknowledgement kinetic peak torque for the effect of gravity when calculating
knee flexor to extensor muscle ratios. Physical Therapy 66: 23-
V. BaItzopoulos is supported by a scholarship from 31, 1986
Fleming RT, Blatz D, McCarroll J. Lateral reconstruction for an-
the Greek Scholarships Foundation.
terolateral rotary instability of the knee. American Journal of
Sports Medicine II: 303-307, 1983
Garnica RA. Muscular power in young women after fast and slow
References isokinetic training. Journal of Orthopaedic and Sports Physical
Therapy 8: 1-9, 1986
Appen L, Duncan WP. Strength relationship of the knee mus- Gilliam T, Sady S, Freedson P, Villanaci J. Isokinetic torque lev-
culature: effect of gravity and sport. Journal of Orthopaedic els for high school football players. Archives of Physical Med-
and Sports Physical Therapy 7: 232-235, 1986 icine and Rehabilitation 60: 110-114, 1979
Armstrong EL, Winant MD, Swasey RP, Seidle EM, Carter LA, Goslin B, Charteris J. Isokinetic dynamometry: normative data
et al. Using isokinetic dynamometry to test ambulatory patients for clinical use in lower extremity (knee cases). Scandinavian
with multiple sclerosis. Physical Therapy 3: 1274-1279, 1983 Journal of Rehabilitative Medicine II: 105-109, 1979
Baltzopoulos V. The development of a computer system for real- Grace T. Muscle imbalance and extremity injury: a perplexing
time display and analysis of isokinetic data. Unpublished M.Phil relationship. Sports Medicine 2: 77-82, 1985
dissertation, University of Liverpool, 1988 Grace T, Sweetser E, Nelson M, Ydens L, Skipper B. Isokinetic
Baltzopoulos V, Brodie DA. The effect of isokinetic training on muscle imbalance and knee joint injuries. Journal of Bone and
the maximum torque output of swimmers using the Akron Joint Surgery 66-A: 734-740, 1984
dynamometer. 5th International Symposium of Biomechanics Gransberg L, Knutsson E. Determination of dynamic muscle
in Sports, Athens, July 13-18, 1987 strength in man with acceleration controlled isokinetic move-
Baltzopoulos V, Eston RG, Mclaren D. A comparison of power ments. Acta Physiologica Scandinavica 119: 317-320, 1983
outputs on the Wingate test and on a test using an isokinetic Gregor R, Edgerton R, Perrine J, Campion D, Debus C. Torque-
device. Ergonomics 31: 1693-1699, 1988 velocity relationship and muscle fiber composition in elite fe-
Barnes W. The relationship of motor unit activation to isokinetic male athletes. Journal of Applied Physiology: Respiratory, En-
muscular contraction at different contractile velocities. Phys- vironmental and Exercise Physiology 47: 388-392, 1979
ical Therapy 60: 1152-1158, 1980 Grimby G. Progressive resistance exercise for injury rehabilita-
Barnes W. Isokinetic fatigue curves at different contractile veloc- tion: special emphasis on isokinetic training. Sports Medicine
ities. Archives of Physical Medicine and Rehabilitation 62: 66- 2: 309-315, 1985
69,1981 Grimby G, Gustafsson E, Peterson K, Renstrom P. Quadriceps
Bemben M, Grump K, Massey B. Assessment of technical ac- function and training after knee ligament surgery. Medicine
curacy of the Cybex II isokinetic dynamometer and analog re- and Science in Sports and Exercise 12: 70-75, 1980
cording system. Journal of Orthopaedic and Sports Physical Hamberg P, Gillquist J, Lysholm J, Oberg B. The effect of diag-
Therapy 10: 12-17, 1988 nostic and operative arthroscopy and open meniscectomy on
Beskin J, Sanders R, Hunter S, Hughston J. Surgical repair of muscle strength in the thigh. American Journal of Sports Med-
achilles tendon ruptures. American Journal of Sports Medicine icine II: 289-292, 1983
15: 1-8, 1987 Heiser TM, Weber J, Sullivan G, Clare P, Jacobs RR. Prophylaxis
Burdett R, Swearingen J. Reliability of isokinetic muscle endur- and management of hamstring muscle injuries in intercolle-
ance tests. Journal of Orthopaedic and Sports Physical Therapy giate football players. American Journal of Sports Medicine
8: 484-488, 1987 12: 368-370, 1984
Burnie J, Brodie DA. Isokinetics in the assessment of rehabili- Herzog W. The relation between the resultant moments at ajoint
tation. Clinical Biomechanics I: 140-146, 1986 and the moments measured by an isokinetic dynamometer.
Caiozzo VJ, Perrine JJ, Edgerton YR. Training induced altera- Journal of Biomechanics 21: 5-12, 1988
tions on the in vivo force-velocity relationship in human muscle. Hill VA. The heat of shortening and the dynamic constants of
Journal of Applied Physiology 51: 750-754, 1981 muscle. Proceedings of the Royal Society of London B 126:
Campbell DE. Generation of horsepower at low and high velocity 136-195, 1938
by sprinters and distance runners. Research Quarterly 50: 1- Hinson M, Smith W, Funk S. Isokinetics: a clarification. Re-
8, 1979 search Quarterly 50: 30-35, 1979
Campbell DE, Glenn W. Rehabilitation of knee extensor and flexor Hislop HJ, Perrine JJ. Isokinetic concept of exercise. Physical
muscle strength in patients with meniscectomies, ligamentous Therapy 47: 114-117 1967
repairs and chondromalacia. Physical Therapy 62: 10-15, 1982 Housh n, Thorland WG, Tharp GD, Johnson GO, Cisar CJ.
Coyle E, Costill D, Lesmes G. Leg extension power and muscle Isokinetic leg flexion and extension strength of elite adolescent
fiber composition. Medicine and Science in Sports II: 12-15, female track and field athletes. Research Quarterly for Exercise
1979 and Sport 55: 347-350, 1984
Coyle, E, Feiring D, Rotkins T, Cote W, Roby F, et al. Specificity Inglis A, Scott N, Sculco T, Patterson A. Ruptures of the tendo
of power improvements through slow and fast isokinetic train- achilles: an objective assessment of surgical and non-surgical
ing. Journal of Applied Physiology 51: 1437-1442, 1981 treatment. Journal of Bone and Joint Surgery 58-A: 990-993,
Davies JG, Kirkendall TD, Leigh HD, Lai LH, Reinhold RT, et 1976
al. Isokinetic characteristics of professional football players: Jenkins W, Thackaberry M, Killiam C. Speed-specific isokinetic
normative data between quadriceps and hamstrings muscle training. Journal of Orthopaedic and Sports Physical Therapy
groups and relative to body weight. Medicine and Science in 6: 181-183, 1984
Sports and Exercise 13: 76-77, 1981 Jensen J, Conn, R, Hazelrigg G, Hewett J. The use of transcu-
Dibrezzo R, Gensch BE, Hinson MM, King J. Peak torque values taneous neural stimulation and isokinetic testing in arthro-
of the knee extensor and flexor muscles of females. Journal of scopic knee surgery. American Journal of Sports Medicine 13:
Orthopaedic and Sports Physical Therapy 7: 65-68, 1985 27-33, 1985
Fenn, WO, Marsh BS. Muscular force at different speeds of short- Johnson J, Siegel D. Reliability of an isokinetic movement ofthe
ening. Journal of Physiology 85: 277-297, 1935 knee extensors. Research Quarterly 49: 88-90, 1978
Isokinetic Dynamometry 115

Kishino N, Mayer T, Gatchel R, McCrate Parrish M, Anderson Parker M, Ruhling R, Bolen T, Edge R, Edwards S. Aerobic train-
C, et al. Isometric and isokinetic lifting simulation in normal ing and the force-velocity relationship of the human quadri-
subjects and low back dysfunction patients. Spine 10: 921-927, ceps femoris muscle. Journal of Sports Medicine 23: 136-147,
1985 1983
Knutsson E, Martensson A. Isokinetic measurements of muscle Patton J, Duggan A. An evaluation of tests of anaerobic power.
strength in hysterical paresis. Electroencephalography and Aviation, Space and Environmental Medicine 3: 237-242, 1987
Clinical Neurophysiology 61: 370-374, 1985 Patton WR, Hinson M, Arnold BR, Lessard MA. Fatigue curves
Lesmes GR, Costill DL, Coyle FE, Fink WJ. Muscle strength and of isokinetic contractions. Archives of Physical Medicine and
power changes during maximal isokinetic training. Medicine Rehabilitation 59: 507-509, 1978
and Science in Sports and Exercise 10: 262-269, 1978 Perrine J, Edgerton VR. Muscle force-velocity and power-velocity
LoPresti C, Kirkendall D, Street G, Dudley D. Quadriceps in- relationships under isokinetic loading. Medicine and Science
sufficiency following repair of the anterior cruciate ligament. in Sports 10: 159-166, 1978
Journal of Orthopaedic and Sports Physical Therapy 9: 245- Pierre R, Andrews L, Allman F, Fleming L The Cybex II eval-
249, 1988 uation of lateral ankle ligamentous reconstructions. American
Lysholm J, Nordin M, Ekstrand J, Gillquist J. The effect of a Journal of Sports Medicine 12: 52-56, 1984
patella brace on performance in a knee extension strength test Potash R, Burn S, Grace P, Harris B, Zaris B, et al. Design of a
in patients with patella pain. American Journal of Sports Med- computer based system for isokinetic testing and analysis. Ath-
icine 12: 110-112, 1984 letic Training 18: 176-178, 1983
Mayer T, Smith S, Kondraske G, Gatchel R, Carmichael T, et al. Richards J, Cooper J. Implementation of an on-line isokinetic
Preliminary data on isokinetic torso rotation testing with my- analysis system. Journal of Orthopaedic and Sports Physical
oelectric spectral analysis in normal and low-back pain sub- Therapy 4: 36-38, 1982
jects. Spine 10: 912-920, 1985 Sale D, McComes J, McDougall D, Upton A. Neuromuscular ad-
Miller L, Donahue J, Good R, Staerk A. The Magnuson-Stack aptation in human muscles following strength training and im-
procedure for treatment of recurrent glenohumeral disloca- mobilization. Journal of Applied Physiology 53: 419-424, 1982
tions. American Journal of Sports Medicine 12: 133-137, 1984 Sale D, McDougall D, Upton A, McComes J. Effect of strength
Milner-Brown HS, Stein RB, Lee RG. Synchronization of human training upon motoneuron excitability in man. Medicine and
motor units: possible role of exercise and supraspinal reflexes. Science in Sports and Exercise 15: 57-62, 1983
Electroencephalography and Clinical Neurophysiology 38: 245- Sapega A, Nicholas J, Sokolow D, Sarantini A. The nature of
254, 1975 torque "overshoot" in Cybex isokinetic dynamometry. Medi-
Mira A, Kitty Markley C, Greer R. A critical analysis of quad- cine and Science in Sports and Exercise 14: 368-375, 1982
riceps function after femoral shaft fracture in adults. Journal Sawhill J, Bates B, Osternig L, Hamill, J. Variability ofisokinetic
of Bone and Joint Surgery 62-A: 61-67, 1980 measures. Medicine and Science in Sports and Exercise 14:
Moffroid M, Whipple R, Hofkosh J, Lowman E, Thistle H. A 177, 1982
study of isokinetic exercise. Physical Therapy 49: 735-742, 1969
Schlinkman B. Norms for high school football players derived
Morris A, Lussier K, Bell G, Dooley J. Hamstrings/quadriceps
from the Cybex data reduction computer. Journal of Ortho-
strength ratios in collegiate middle-distance and distance run-
paedic and Sports Physical Therapy 5: 243-254, 1984
ners. Physician and Sportsmedicine II: 71-77, 1983
Scudder NG. Torque curves produced of the knee during iso-
Mulder H. Ice hockey injuries. Journal of Sports Medicine I: 41-
metric and isokinetic exercise. Archives of Physical Medicine
42, 1973
and Rehabilitation 61: 68-72, 1980
Murray D. Optimal filtering of constant velocity torque data.
Medicine and Science in Sports and Exercise 18: 603-611, 1986 Sherman, W, Pearson D, Plyley M, Costill A, Habansky A, et al.
Murray MS, Warren FR, Otis CJ, Kroll M, Wickiewicz LT. Isokinetic rehabilitation after surgery: a review of factors which
Torque-velocity relationship of the knee extensors and flexors are important for developing physiotherapeutic techniques after
muscles in individuals sustaining injuries of the anterior cru- knee surgery. American Journal of Sports Medicine 10: 155-
ciate ligament. American Journal of Sports Medicine 12: 436- 161, 1982
440, 1984 Shields C, Silva I, Yee L, Brewster C. Evaluation of residual in-
Nelson S, Duncan P. Correction of isokinetic torque recordings stability after arthroscopic meniscectomy in anterior cruciate
for the effect of gravity. Physical Therapy 63: 674-676, 1983 deficient knees. American Journal of Sports Medicine 15: 129-
Nistor L. Surgical and non-surgical treatment of achilles tendon 131, 1987
rupture. Journal of Bone and Joint Surgery 63-A: 394-399, 1981 Sinacore, D, Rothstein J, Delitto, A, Rose S. Effect of damp on
Noyes F, Mangine R, Barber S. Early knee motion after open and isokinetic measurements. Physical Therapy 63: 1248-1250, 1983
arthroscopic anterior cruciate ligament reconstruction. Amer- Slagle GW. The importance of pre-testing the knee joint. Athletic
ican Journal of Sports Medicine 15: 149-160, 1987 Training 14: 225-226, 1979
Olerud S, Wallenstein R, Olsson E. Muscle strength after bilateral Smidt G, Herring T, Amundsen K, Rogers M, Russel A, et al.
femoral osteotomy. Journal of Bone and Joint Surgery 66-6: Assessment of abdominal and back extensor function: a quan-
792-793, 1984 titative approach and results for chronic low-back patients.
Osternig L. Optimal isokinetic loading and velocities producing Spine 8: 211-219,1983
muscular power in human subjects. Archives of Physical Med- Solomonow M, Baratta R, Zhou B, Shoji H, Bose W, et at The
icine and Rehabilitation 50: 152-155, 1975 synergistic action of the anterior cruciate ligament and thigh
Osternig L, Sawhill J, Bates B, Hamill J. A method for rapid muscles in maintaining joint stability. American Journal of
collection and processing of isokinetic data. Research Quart- Sports Medicine 15: 207-213, 1987
erly 53: 252-256, 1982 Thistle H, Hislop H, Moffroid M, Hofkosh J, Lowman E. Iso-
Osternig K, Sawhill J, Bates B, Hamill J. Function of limb speed kinetic contraction: a new concept of exercise. Archives of
on torque patterns of antagonist muscles. In Matsui & Ko- Physical Medicine and Rehabilitation 48: 279-282, 1967
bayashi (Eds) Biomechanics VIII-A, pp. 251-257, Human Ki- Thomee R, Renstrom P, Grimby G, Peterson L. Slow and fast
netic Publishers, Champaign, 1983 isokinetic training after knee ligament surgery. Journal of Or-
Parker M. Calculation of isokinetic rehabilitation velocities for thopaedic and Sports Physical Therapy 8: 475-479, 1987
the knee extensors. Journal of Orthopaedic and Sports Phys- Thorbland J, Ekstrand J, Hamberg P, Gillquist J. Muscle reha-
ical Therapy 4: 32-35, 1982 bilitation after arthroscopic meniscectomy with or without
116 Sports Medicine 8 (2) 1989

tourniquet control: a preliminary randomized study. American Wyatt MP, Edwards AM. Comparisons of quadriceps and
Journal of Sports Medicine 13: 133-135, 1985 hamstrings torque values during isokinetic exercise. Journal of
Thorstensson A, Grimby G, Karlsson J. Force-velocity relations Orthopaedic and Sports Physical Therapy 3: 48-56, 1981
and fibre composition in human knee extensor muscles. Jour- Yates J, Kamon E. A comparison of peak and constant angle
nal of Applied Physiology 40: 12-16, 1976 torque-velocity curves in fast and slow-twich populations.
Thorstensson A, Karlsson J. Fatigueability and muscle fibre com- European Journal of Applied Physiology, 51: 67-74, 1983
position in human skeletal muscle. Acta Physiologica Scan-
dinavica 198: 318-322, 1976
Watkins PM, Harris AB, Kozlowski BA. Isokinetic training in
patients with hemiparesis. Physical Therapy 64: 184-189, 1984 Authors' address: D.A. Brodie, School of Movement Science,
Winter DA, Wells RP, Orr GW. Errors in the use of isokinetic
dynamometers. European Journal of Applied Physiology 46: Physical Education and Recreation, University of Liverpool, P.O.
397-408, 1981 Box 147, Liverpool L69 3BX, England

Adenosine and A TP
Progress in Research and Therapeutic Potential
25-26 September 1989
Royal College of Physicians, London
For further information, please contact:
IBC Technical ervices Ltd
Bath House (3rd Floor)
56 Holborn Viaduct
London ECIA 2EX
ENGLAND