European Journal of

Applied Physiology

Eur J Appl Physiol (1988) 57:526--530

A new dynamometer measuring concentric and eccentric muscle strength in accelerated, decelerated, or isokinetic movements

and Occupational Physiology 9 Springer-Verlag 1988

Validity and reproducibility Jan Y. Seger 1, Stephen H. Westing 2, Mats Hanson 3, Eddy Karlson 2, and Bj6rn Ekblom 2 1 University College of Physical Education, Linding6v~gen 1, S-11433 Stockholm 2 Department of Physiology III, Karolinska Institute, S-11433 Stockholm 3 Department of Machine Elements, Royal Institute of Technology, S-11433 Stockholm, Sweden

Summary. A new computerized dynamometer (the SPARK System) is described. The system can

Introduction

measure concentric and eccentric muscle strength (torque) during linear or nonlinear acceleration or deceleration, isokinetic movements up to 4 0 0 ~ -1, and isometric torque. Studies were performed to assess: I. validity and reproducibility of torque measurements; II. control of lever arm position; III. control of different velocity patterns; IV. control of velocity during subject testing; and, V. intra-individual reproducibility. No significant difference was found between torque values computed by the system and known torque values (p > 0.05). No difference was present between programmed and external measurement of the lever arm position. Accelerating, decelerating and isokinetic velocity patterns were highly reproducible, with differences in elapsed time among 10 trials being never greater than 0.001 s. Velocity during concentric and eccentric isokinetic quadriceps contractions at 30 ~ 9 s -1, 120 ~ -~ and 270 ~ s -1 never varied by more than 3 ~ 9 s-1 among subjects ( N = 21). Over three days of testing, the overall error for concentric and eccentric quadriceps contraction peak torque values for 5 angular velocities between 30 ~ s -~ and 270 ~ 9 s -1 ranged from 5.8% to 9.0% and 5.8% to 9.6% respectively ( N = 25). The results indicate that the SPARK System provides valid and reproducible torque measurements and strict control of velocity. In addition, the intra-individual error is in accordance with those reported for other similar devices.

Natural movement patterns are a result of a series of eccentric, concentric and isometric muscle contractions. Therefore a comprehensive picture of the force potential of human skeletal muscle can best be obtained by studying all three types of muscle contraction. Isometric strength and concentric strength during isokinetic (constant velocity) movements have been studied for decades (for recent reviews, see Petrofsky et al. 1986 and Osternig 1986), but little work has been published concerning eccentric isokinetic muscle strength. This is probably due to the difficulty in constructing an accurate and safe isokinetic dynamometer for the measurement of eccentric strength, since a large external load must be applied to the body segment. Recently, the K I N C O M (Farrel and Richards 1986), the LIDO and the BIODEX, dynamometers that are capable of measuring eccentric isokinetic strength, have become commercially available. Furthermore, body segments do not normally move isokinetically, but rather in a series of accelerations and decelerations which it has not been possible to study up till now. The purpose of this paper is to describe a new dynamometer (the SPARK System) capable of measuring concentric and eccentric torque during linear or nonlinear acceleration or deceleration, and isokinetic and isometric torque. Data will be presented concerning the validity and reproducibility of torque measurements, the control of lever arm position, the control of different velocity patterns and velocity control during subject testing, and the intraindividual reproducibility of the torque output.

Key words: Acceleration -- Eccentric -- Concentric -- Isokinetic -- Method error Offprint requests to: J. Y. Seger

J. Y. Seger et al.: A new dynamometer for measuring muscle strength

1. The experimental setup for the SPARK System. (A) load cell signal amplifier; (B) real-time embedded microcomputer for control, measurement and communication; (C) remote computer system for programming and displaying data; (D) power and control system; (E) load cell (Bofors KIS 3, Bofors Elektronik, Bofors, Sweden; manufacturer reported combined error of measurement, +0.02%); (F) adjustable lever arm; (G) gear box; (H) reversible electric motor; (I) incremental encoder for monitoring angular displacement Fig.

527

5. adjustable acceleration and deceleration ramps at the beginning and end of the range of motion 6. unidirectional or bidirectional movement 7. repetitive movement up to 75 trials 8. adjustable range of motion up to 90 ~ 8. graphic presentation of force, torque, work, or velocity as a function of angular displacement o r time 9. tabular presentation of peak torque, angle of peak torque, torque at any requested angle, work, and elapsed time every 10 ~ in the range of motion Currently there is no automatic correction for the influence of gravity. Gravity corrections can be made by measuring the torque created by the relaxed leg and adjusting torque scores accordingly.

Methods

Subjects. Two groups of subjects were tested: Group A con-

Description of the dynamometer The primary components of the dynamometer are presented in Fig. 1.

Capabilities 1. measurement of both concentric and eccentric torque 2. measurements during constant acceleration or deceleration (so called isoacceleration or isodeceleration) 3. self-designed run pattern Using the self-designed run pattern feature, it is possible to create virtually any unidirectional, bidirectional or repetitive movement pattern desired. 4. isometric measurements at different specific angles

tained 21 habitually active male subjects (mean age, stature and body mass: 25.3+3.0 (SD) years, 179.8+5.6cm and 74.3+5.0 kg). Data from Group A were used in Study IV. Group B consisted of 25 moderately trained subjects (10 females and 15 males; mean age, stature and body mass of the females and males combined: 29.0+7.7 years, 176.3 +8.5 cm and 68.4+9.8 kg). Group B participated in Study V.

Study" I. Validity and reproducibility of torque measurements. Torque-measurement calibration of the dynamometer was performed by hanging a weight of known mass (40.0 kg) on the load cell at the end of the lever arm. The resulting calibration constant was stored by the controlling software of the system and was used for all tests in Study I. The validity of the torque measurements was established by hanging a series of weights of known mass (20.0, 30.0, 40.0, 50.0, and 60.0 kg) from the load cell while the lever arm was horizontal (0~ verified with a bubble clinometer with a tolerance of __+0.5 ~ The calculated torque values were compared to the known torque values. The reproducibility of these measurements was established by repeating this procedure 3 times in direct succession, 3 separate times per day, for 3 consecutive days. Since the load cell is attached to the end of the lever arm, the measurement of torque is unaffected by movement of the lever arm, so that this procedure was not repeated during dynamic conditions.

Study H. Control of lever arm position. This was tested in the

Features 1. calibration of torque measurements and angular velocity 2. maximum angular velocity of 400 ~ 9 s - 1 3. torque recordings up to 1200 Nm (with 1 Nm resolution) 4. adjustable leve~ of pretension torque for lever arm movement initiation

isometric mode by comparing 50 randomly selected lever arm positions from 0 ~ to 90 ~ with known lever arm positions measured by a bubble clinometer.

Study IH. Control of different velocity patterns. Six different velocity patterns were tested: a. isoacceleration (constant rate of acceleration) up to 400 ~ - s - a ; b. non-linear rate of acceleration up to 400 ~ 9 s - l ; c. isodeceleration (constant rate of deceleration) down from 300 ~ 9s - l ; d. nonlinear rate of deceleration down from 300 ~ - s - i ; e. isokinetic (constant velocity) at 350 ~ s 1 with acceleration and deceleration ramps; and, f. seemingly random accelerations and decelerations. Without a

528

J.Y. Seger et al.: A new dynamometer for measuring muscle strength

subject seated on the apparatus, each pattern was performed 10 times in succession. The reproducibility of these velocity patterns was checked by the elapsed time recorded every 10 ~ in the range of motion by an electronic timer built into the computer system (with a resolution of 0.001 s).

Study IV. Control of angular velocity durin 9 subject testin 9. Two velocity-control calibrations were performed for concentric and eccentric lever arm movements respectively. The resulting calibration constants were stored by the controlling software of the system and were used for all tests in Study IV. Group A performed maximal voluntary concentric and eccentric quadriceps contractions with the left leg between 10 ~ and 90 ~ at 3 preset lever arm velocities (30 ~ 9s -1, 120 ~ 9s -a and 270 ~ s-~). The angular velocity of each trial was calculated from the elapsed time every 10 ~ in the range of motion between 30 ~ and 70 ~.

Study V. Intra-individual reproducibility. After a standardized warm-up procedure on a bicycle ergometer and a familiarization session on the SPARK at each testing velocity, Group B performed maximal voluntary concentric and eccentric quadriceps contractions with the left leg between 20 ~ and 90 ~ at 5 lever arm velocities (30 ~ 9s 1, 60 ~ . s - l , 120 ~ . s - l , 180 ~ . s 1 and 270 ~ 9 s - l ) . This test was repeated on 3 separate days. The error of the method was calculated for peak torque from the best trial at each velocity. The "best trial" was defined as the trial with the greatest work performance.

Statistical procedures. Standard statistical techniques were used to calculate means and standard deviations. The method error (Study V) was defined as the percentage variations in relation to the grand mean between duplicate determinations. Differences between mean values (Studies I & V) were tested for significance using a repeated-measure analysis of variance. The level of significance was established at p < 0.05.

Results

Study I. The torque values calculated by the system were not significantly different from the known torque values (p> 0.05), see Table 1. Further, no significant difference was seen among torque values either across trials within a single test day or across the 3 days of testing (p > 0.05).

Study II. No difference was present between the 50 randomly selected programmed lever arm positions and the lever arm positions measured by a bubble clinometer with a tolerance of _+0.5 ~

Table 1. Changes in torque measured by the SPARK system when weights were applied 3 times per day on 3 consecutive days (Study I) Known Torque (Nm)

79 118 157 196 236

Torque measured by the SPARK system (Nm) a Day 1

Day 2

Day 3

a

b

c

a

b

c

a

b

c

82 119 157 193 230

82 118 156 193 231

82 119 158 193 231

82 119 157 194 230

81 120 157 193 230

82 119 157 194 230

81 119 157 194 231

81 118 158 194 231

81 119 157 193 231

Each value represents the mean of 3 trials performed in direct succession, a, b and c denote the trials on Day 1, Day 2 and Day 3

Study IV. The mean directly-timed angular velocity, determined from measurements of elapsed time every 10 ~ from 30 ~ to 70 ~ did not differ by more than 1% across the entire measurement interval for any concentric and eccentric test (with the exception of the 270 ~ s -1 concentric test: 2.7% difference). The elapsed time every 10 ~ for the 270 ~ 9s-1 concentric test ranged from 0.036 s to 0.037 s across all subjects (system resolution = 0.001 s), or angular velocities of 277.8 ~ 9 s and 270.3 ~ s -1 respectively. Due to these findings, we have only presented a single estimation of the mean angular velocity for each test, calculated from the total elapsed time from 30 ~ to 70 ~ (see Table 2). In no case did the standard deviations of the mean directly-timed angular velocities exceed _+0.9 ~ 9s-1 for any test, and the velocities never varied by more than 3 ~ s-1 among subjects.

Study V. No significant difference was seen over the three days of testing for measurements of peak torque or angle-specific torque (every 10 ~ from 30 ~ to 70 ~) at any concentric or eccentric test velocity (p > 0.05). The overall method errors for concentric and eccentric peak torque values ranged from 5.8% to 9.0% and 5.8% to 9.6% respectively.

Study IlL Figure 2 displays 6 velocity patterns programmable with the SPARK System. Each

Discussion

curve consists of the super-imposed tracings of 10 trials performed in succession. For all of these patterns, differences in the elapsed time every 10 ~ in the range of motion were never greater than 0.001 s.

The SPARK System appears capable of valid and reproducible torque measurements, and can maintain a strict control of angular velocity during different velocity patterns and actual subject testing.

J. Y. Seger et al.: A new dynamometer for measuring muscle strength Velocity (~

,00

Velocity (~

/

400

350

350

300

300

250

250

200 -

2:00

150

150

-

f/

100

100.

50

50-

l

0

o

~~

l

~

i

I i rime

0

velocit

529

400

0

i

(ms)

I

i

I ~

i

Time

BOO (ms)

Velocity (~

(~

400

d

350 300 -

300 - '

250 -

250 -

200 9

200

150 -

150 -

100

I00 -

-

50,

-

SO-

!

0

i

i

i

i

i

Time 400 (ms)

i

i

I ~

0

Velocity (O/S)

i

i

i

i

i

t i 700

Time


Velocity (~

400 ~

e

400

t

350-

350 T

300-

300

250,

250

f

T

200. 150-

I00" 50" 0 0

--

"

.

.

.

.

. O'

Angle (o)

~176

. . . . . . .

0

The estimates of the method error for peak torque obtained in Study V were similar to those reported for the CYBEX II (Thorstensson 1976), the ORTHOTRON (Ingemann-Hansen and Halkjaer-Kristensen 1979) and the ACE system (Jacobs and Pope 1986). The directly timed mean angular velocity values in Study IV did differ somewhat from the preset angular velocity values (Table 2). This probTable 2. Changes in directly-timed angular velocity during actual subject isokinetic testing ( N = 21; S t u d y IV) Preset Velocity

(O.s-,)

Directly timed angular velocity a

Concentric

Eccentric

Velocity M e a n _ S D & Range

Velocity Mean + S D & Range

25-• 0.8 (24-27) 117 + 0.9 (116-119) 272 __+0.9 (272-274)

33 + 0 . 8 (31-34) 121 _+ 0.8 (120-122) 267 + 0.9 (265-268)

(~

30 120 270 a

1)

(o. s-, )

Calculated from the elapsed time between 30 ~ and 70 ~

90 I~ ~

Fig. 2. Six velocity patterns programmable with the SPARK System. Each curve consists of the super-imposed tracings of 10 trials performed in succession: (a) isoacceleration (constant rate of acceleration) up to 400 ~ 9 s - I ; (b) nonlinear rate of acceleration up to 400 ~ - s - ' ; (c) isodeceleration (constant rate of deceleration) down from 300 ~ 9 s - l ; (d) nonlinear rate of deceleration down from 300 ~ 9 s - ] ; (e) isokinetic (constant velocity) at 350 ~ 9 s -1 with acceleration and deceleration ramps; (f) seemingly random accelerations and decelerations

lem is due to the fact that the concentric and eccentric velocity calibration constants employed in this study were not sufficient to cover the complete range of testing velocities. To ensure highly valid velocities, it is necessary to either calibrate separately for low and high velocity tests (see for example Fig. 2e, an isokinetic test at 350 ~ s -~ performed after a high-velocity test calibration), or request a slightly lower or higher velocity (for example, requesting 27 ~ s -1 to ensure a valid 30 ~ 9 s - 1 eccentric test). A frequent problem during high-velocity isokinetic testing is the occurrence of overshoot and undershoot torques at the beginning of the torque-displacement curves. The attempts with CYBEX II to electronically "dampen" these oscillations has been criticized (Sapega et al. 1982). However, acceleration-controlled isokinetic movements have been tested on the CYBEX II, resulting in considerably smaller oscillations at the beginning of the high-velocity torque-displacement curves (Gransberg and Knutsson 1983). With the SPARK System (as with the KINCOM system; see Farrel and Richards 1986), the oscillation

530

J. Y. Seger et al.: A new dynamometer for measuring muscle strength

problem appears to be solved by the establishment of an acceleration ramp at the beginning of the range of motion (for an example of an acceleration ramp, see Figure 2e). Acceleration ramps on the SPARK System over the first 5 ~ (ca. 10 ~ 9 s -1 to 100 ~ s - l ) , 10 ~ (ca. 100 ~ S - 1 to 200 ~ s - l ) , 15 ~ (ca. 2 0 0 ~ -1 to 3 0 0 ~ -1) or 20 ~ (ca. 300 ~ 9 s -1 to 400 ~ 9s -~) of the range of motion appear to be sufficient to eliminate the oscillation problem (unpublished findings). The magnitude of these S P A R K acceleration ramps are similar to those of Gransberg and Knutsson (1983) for CYBEX II of 2 ~ (30 ~ s - l ) , 8 ~ (120 ~ s -1) and 18 ~ (240~ s - 1). Synchronizing the initiation of the movement of the body segment with the movement of the motor-driven lever arm is important, especially during testing at high velocities. This is obtained in the SPARK System by establishing a pretension torque level that must be overcome by the subject before the movement is initiated. This method has also been employed on CYBEX II (Gransberg and Knutsson 1983) and the K I N C O M system (Farrel and Richards 1986). An additional benefit of this feature (compared with the option of the test leader initiating the lever arm movement) is that subjects can perform the trials exactly when they choose, thus greatly increasing their ability to concentrate. With the SPARK System, the researcher may programme virtually any concentric or eccentric movement pattern. Concentric isoacceleration or eccentric isodeceleration are two possible velocity patterns for future testing (see Thorstensson et al. 1986 for some preliminary results). By investigating muscle strength during movements with a constant rate of acceleration or deceleration, the researcher maintains the advantageous aspects of isokinetic measurements (i. e., a high degree of experimental control), while possibly more closely imitating natural movement patterns. Since eccentric decelerated and concentric accelerated movements can be performed in succession with a

short coupling time, another possible area of investigation is the stretch-shortening cycle (for a recent review, see Komi 1984). In the light of these findings, it appears safe to conclude that the SPARK System can confidently be used for future research work. Acknowledgements. This investigation was supported by research funds from the Folksam Insurance Company, The Karolinska Institute, and the Research Board fo the Swedish Sports Federation, all of Stockholm, Sweden. We wish to thank Dr. A. Thorstensson for his valuable comments while preparing this manuscript.

References Farrel M, Richards JG (1986) Analysis of the reliability and validity of the kinetic communicator exercise deviee. Med Sci Sports Exerc 18:44-49 Gransberg L, Knutsson E (1983) Determination of dynamic muscle strength in man with acceleration controlled isokinetic movements. Acta Physiol Scand 119 :317-320 Ingemann-Hansen T, Halkjaer-Kristensen J (1979) Force-velocity relationships in human quadriceps muscles. Scand J Rehab Med 11:85-89 Jacobs I, Pope J (1986) A computerized system for muscle strength evalution: measurement reproducibility, validity and some normative data. NSCA J 8:28-33 Komi PV (1984) Physiological and biomechanical correlates of muscle function: Effects of muscle structure and stretchshortening cycle on force and speed. Exc Sport Sci Rev 12:81 121 Osternig LR (1986) Isokinetic dynamometry: implications for muscle testing and rehabilitation. Exc Sport Sci Rev 14:45-80 Petrofsky JS, Phillips CA (1986) The physiology of static exercise. Exc Sport Sci Rev 14:1-44 Sapega AA, Nicholas JA, Sokolow D, Saraniti A (1982) The nature of torque "overshoot" in Cybex isokinetic dynamometry. Med Sci Sports Exerc 14:368-375 Thorstensson A (1976) Muscle strength, fibre types and enzyme activities in man. Acta Physiol Scand [Suppl 443] Thorstensson A, Oddsson L, Karlson E, Seger J (1986) Does acceleration influence the force-velocity relationship of concentric and eccentric concentrations? Med Sci Sports Exerc 18:$63

Accepted February 2, 1988