Original Article Comparison of Reliability of Isometric Leg Muscle Strength Measurements Made Using a Hand-Held Dynamometer with and without a Restraining Belt J. Phys. Ther. Sci. 21: 37 42, 2009 MUNENORI KATOH, PT, PhD 1), HIROSHI YAMASAKI, PT, PhD 2) 1) Department of Physical Therapy, Faculty of Health Science, Ryotokuji University: 5 8 1 Akemi, Urayasu-City, Chiba 279-8567, Japan 2) Department of Physical Therapy, Kochi Rehabilitation Institute Abstract. [Purpose] The aim of this study was to evaluate whether using a belt to restrain a hand-held dynamometer improves reliability of isometric leg muscle strength measurements in healthy subjects. [Subjects] Twenty to 44 healthy subjects participated in the test. [Methods] Two raters, one man and one woman, used a hand-held dynamometer with or without a restraining belt to measure the isometric strengths of the following muscle groups: flexors, extensors, abductors, adductors, internal rotators and external rotators of the hip; flexors and extensors of the knee; and dorsiflexors and plantar flexors of the ankle. [Results] The intraclass correlation coefficient, used to describe interrater agreement, ranged from 0.97 to 0.99 with the belt and from 0.21 to 0.88 without the belt. Pearson s correlation coefficient for measurements with versus without the belt ranged from 0.61 to 0.95 for the man and from 0.31 to 0.87 for the woman. [Conclusion] The interrater reliability of isometric leg muscle strength measurements was improved by use of a belt to restrain the hand-held dynamometer. Key words: Muscle strength measurement, Hand-held dynamometer, Belt fixation (This article was submitted Jul. 17, 2008, and was accepted Sep. 22, 2008) INTRODUCTION Hand-held dynamometers (HHD) are small and relatively simple to use for quantitatively measuring muscle strength, but there is no consensus on the reliability of measurements 1 12). Bohannon et al. 1) evaluated knee extensor strength in 24 healthy women aged around 30 years old and reported, as an estimate of interrater agreement, an intraclass correlation coefficient (ICC) of 0.948, and good intrarater reliability irrespective of experience. In a study by Wikholm et al. 7) in which muscle strengths of healthy subjects were evaluated by 3 raters, ICC was 0.226 for strength measurements of knee extensor muscles, which are strong, and 0.768 and 0.932, respectively, for measurements of elbow flexor and shoulder external rotator muscles, which are weaker. In a study by Agre et al. 8) using 4 healthy subjects and 3 raters, Pearson correlation coefficients for interrater measurements ranged from 0.19 to 0.96 for the lower extremity, but from 0.88 to 0.94 for the upper extremity. Regarding the limitations of measurements using a HDD, Hyde et al. 10) claim that, regardless of the tester s experience or power of resistance, 30 kg is the limit, while Wiles et al. 11) and Roebeck et al. 12), claim respectively that it is difficult to maintain the position of the HDD at forces of 300 N and 85 Nm or higher. Problems of reliability of measurements made by
38 J. Phys. Ther. Sci. Vol. 21, No. 1, 2009 Table 1. Subjects characteristics per Exertion Task Exertion Task Subjects Male Female Age (years) Height (cm) Weight (kg) Leg Side/Number Hip flexion 20 9 11 20.0 ± 0.6 163.5 ± 9.2 56.3 ± 7.0 Both/40 Hip extension 20 9 11 19.7 ± 1.7 167.4 ± 9.5 59.8 ± 9.3 Both/40 Hip abduction 44 18 26 19.9 ± 1.4 165.7 ± 7.9 57.8 ± 6.9 Right/44 Hip adduction 44 18 26 19.9 ± 1.4 165.7 ± 7.9 57.8 ± 6.9 Right/44 Hip external rotation 44 18 26 19.9 ± 1.4 165.7 ± 7.9 57.8 ± 6.9 Right/44 Hip internal rotation 43 18 25 20.0 ± 1.4 165.9 ± 7.9 58.1 ± 6.7 Right/43 Knee flexion 43 18 25 20.0 ± 1.4 165.9 ± 7.9 58.1 ± 6.7 Right/43 Knee extension 21 11 10 19.4 ± 0.5 165.8 ± 8.3 59.0 ± 8.6 Both/42 Ankle plantar flexion 44 18 26 19.9 ± 1.4 165.7 ± 7.9 57.8 ± 6.9 Right/44 Ankle dorsiflexion 42 17 25 19.9 ± 1.5 165.5 ± 7.5 58.1 ± 7.6 Right/42 ±: mean ± SD HHDs make comparisons of studies difficult. Therefore, the aim of this study was to assess whether using a restraining belt to steady a HDD with a thin sensor improves reliability of measurements of leg muscle strength. SUBJECTS AND METHODS The number of subjects varied between 20 and 44 depending on which muscle group was evaluated, but since the exertion tasks were performed either bilaterally or unilaterally, the total number of legs per task varied between only 40 and 44 (Table 1). The subjects mean age was 20. None of the subjects had any joint pain or other orthopedic problems of the joints used in the exertion tasks. All the subjects gave their informed consent to participate. The HDD was a µtas MF-01 (Anima Corp., Tokyo). Its metal sensor was covered by a rubber pad. When the belt was used to secure the device, the sensor s Velcro fastener was used to secure the sensor to the part of the body under investigation, but when the belt was not used to secure the device, the sensor s Velcro fastener was used as a grip for the tester to hold when applying resistance. The belt length could be adjusted using a sliding bar buckle and a plate was used to fix the position of the sensor (Fig. 1). The HDD had a measuring range of 0.0 to 80.0 kgf with a precision of 0.1 kgf. Measurements were made with the subjects making the following 10 exertions: flexion, extension, abduction, adduction, internal rotation and external rotation of the hip; flexion and extension of the knee; and dorsiflexion and plantar flexion of the ankle. In the test without the belt, the Fig. 1. Equipment and example of installation in a case of knee extension. The sensor of a hand-held dynamometer with its Velcro surface fastener straps is shown attached to the belt. The Velcro straps are used to fasten the sensor to the surface of a patient s limb when the belt is employed (top right) or as a grip for the tester when the belt is not employed (Fig. 2). A sliding buckle is used to adjust the belt length when anchored to a fixed structure (bottom right). HDD was used in the conventional way, i.e., held in the palm of the hand and pressed directly against the part of the body under test. Measurement was ceased when the tester was unable to hold the HDD steady. In the test with the belt, the sensor s Velcro fastener was used to fix the HDD against the part of the body under test, and the belt was fixed to something relatively immobile, such as the leg or rail of the bed or around the tester s leg, to directly oppose the movement of the part of the body under test (Fig. 2, Table 2). The subjects ramped up to maximum their exertions over 3 s, and then maintained a static state for about 5 s during which the maximum force was noted. Each exertion trial was repeated once after 30 seconds of rest. The highest measurement for each pair of trials was noted.
39 Fig. 2a. Using a hand-held dynamometer with and without the belt for the hip. Photographs show six measurements of isometric muscle strength for the hip, using the belt to anchor the device (left) and without using the belt (right) where the tester applies resistance by hand and the Velcro straps of the surface fastener are used as a grip. Fig. 2b. Using a hand held dynamometer with and without the belt for the knee and ankle. Photographs show two measurements of isometric muscle strength each for the knee and ankle, using the belt to anchor the device (left) and without using the belt (right) where the tester applies resistance by hand and the Velcro straps of the surface fastener are used as a grip. In the assessment of interrater reliability, 1 man and 1 woman with very different physiques were used as testers (Table 3). The subjects rested for at least 1 h between the first and second testers taking measurements. Measurements made with and without the belt were taken at an interval of 1 week. The order in which measurements were made with the belt, without the belt, by the man, or by the
40 J. Phys. Ther. Sci. Vol. 21, No. 1, 2009 Table 2. Subjects posture, dynanometer position and belt fixation method during isometric muscle strength measurements made using a hand-held dynamometer with belt Exertion Task Posture Joint Position Dynamometer Position* Belt Fixation Method Hip flexion Sitting upright (1) Hip flexed 90, popliteal fossa floating Femur head Under table leg Hip extension Prone (2) Hip at 0, distal femur floating Femur head Under table leg Hip abduction Supine (3) Hip at 0, opposite pelvic side on pillow, touching bed rail Femur head Around bed rail Hip adduction Supine (4) Hip at 0, pelvis held by tester s hand Femur head Around tester s lower leg Hip external rotation Sitting upright (1) Hip at 0, thigh horizontal Medial maleolus Around table leg Hip internal rotation Sitting upright (1) Hip at 0, thigh horizontal Lateral maleolus Around table leg Knee flexion Sitting upright (1) Knee flexed 90, thigh horizontal Medial maleolus Around tester s lower leg Knee extension Sitting upright (1) Knee flexed 90, thigh horizontal Medial maleolus Around table leg Ankle plantar flexion Prone (5) Knee at 115, ankle dorsiflexed 15 1st metatarsal head Around subject s thigh Ankle dorsiflexion Sitting on chair (6) Heel on 10 cm block touching bed rail, ankle dorsiflexed 20 1st metatarsal head Around bed rail (1) On edge of a table, (2) On table, more proximal than sensor, (3) On bed, (4) On mat, (5) On table, (6) Lower leg on bed. *lower part of sensor placed on top of part of body under test in opposite direction to movement. Table 3. Testers characteristics per Exertion Task Male Tester Female Tester Exertion Task Age (years) Height (cm) Weight (kg) Age (years) Height (cm) Weight (kg) Hip flexion 26 183.0 76.0 21.0 152.0 43.0 Hip extension 26 183.0 76.0 21.0 152.0 43.0 Hip abduction 26 183.0 76.0 21.0 154.0 65.0 Hip adduction 21 179.0 79.0 21.0 156.0 45.5 Hip external rotation 21 176.0 71.5 21.0 151.0 44.0 Hip internal rotation 21 176.0 71.5 21.0 151.0 44.0 Knee flexion 21 178.5 76.0 21.0 154.0 56.0 Knee extension 26 183.0 76.0 21.0 152.5 43.0 Ankle plantar flexion 21 178.5 76.0 21.0 154.0 56.0 Ankle dorsiflexion 21 178.5 76.0 21.0 152.5 43.0 woman, was random. Before starting the study, the testers practiced making measurements until they were familiar with the technique. Interrater reliability was evaluated using the paired Student s t test, ICC (2,1), and Pearson s correlation coefficient. Measurements made with versus without the belt were analyzed using the paired Student s t test and Pearson s correlation coefficient (SPSS ver.15j for Windows, SPSS Japan Inc., Tokyo) P values of <.05 were considered significant. RESULTS Isometric muscle strength measurements for each exertion task and their statistics are shown in Tables 4 6. On comparing the examiners measurements, for all tasks, there were no significant differences when the belt was used, but when the belt was not used, measurements obtained by the male examiner were significantly higher than those obtained by the female examiner (p<0.001). ICC ranged from 0.99 to 0.97 when the belt was used, but ranged from 0.88 to 0.21 when it was not used. Pearson s correlation coefficient ranged from 0.99 to 0.97 when the belt was used, but ranged from 0.91 to 0.26 when it was not used. Measurements were significantly higher for both examiners when the belt was used than when it was not used (p<0.01). Pearson s correlation coefficient for measurements made with versus without the belt ranged from 0.95 to 0.61 for the male examiner and from 0.87 to 0.31 for the female examiner. DISCUSSION Gross et al. 13) reported ICCs of between 0.92 and
41 Table 4. Isometric muscle strength measurements of young healthy adults made by one male and one female tester using a hand-held dynamometer with and without a belt Exertion Task n With Belt Without Belt Male Tester Female Tester Male Tester Female Tester Hip flexion 40 25.0 ± 9.7 25.8 ± 9.8 19.8 ± 4.5 17.3 ± 3.6 Hip extension 40 30.8 ± 8.3 31.0 ± 7.6 27.0 ± 5.1 23.6 ± 3.3 Hip abduction 44 29.4 ± 7.7 29.0 ± 7.5 25.7 ± 6.6 19.2 ± 4.7 Hip adduction 44 19.5 ± 7.0 19.5 ± 7.0 17.1 ± 6.2 15.3 ± 4.9 Hip external rotation 44 17.2 ± 5.8 17.2 ± 5.4 15.5 ± 5.0 14.5 ± 4.0 Hip internal rotation 43 17.2 ± 6.4 17.1 ± 6.2 14.5 ± 5.1 12.8 ± 3.4 Knee flexion 43 29.2 ± 8.0 28.6 ± 7.8 22.5 ± 5.3 20.5 ± 4.3 Knee extension 42 46.8 ± 13.7 46.6 ± 13.6 32.3 ± 5.0 20.2 ± 2.6 Ankle plantar flexion 44 48.1 ± 13.8 48.0 ± 13.6 33.0 ± 4.9 26.5 ± 3.9 Ankle dorsiflexion 42 32.4 ± 9.0 32.5 ± 8.7 25.7 ± 6.7 23.2 ± 4.7 mean ± SD (kgf). Table 5. Interrater comparison (Intraclass correlation coefficient) of isometric muscle strength measurements made using a hand-held dynamometer with and without a belt Without Belt With Belt Exertion Task 95%CI 95%CI ICC (2,1) ICC (2,1) Lower Upper Lower Upper Hip flexion 0.66 0.11 0.86 0.98 0.96 0.99 Hip extension 0.61 0.06 0.85 0.98 0.96 0.99 Hip abduction 0.41 0.10 0.41 0.99 0.97 0.99 Hip adduction 0.83 0.58 0.93 0.98 0.97 0.99 Hip external rotation 0.82 0.67 0.90 0.97 0.95 0.98 Hip internal rotation 0.74 0.45 0.87 0.99 0.97 0.99 Knee flexion 0.79 0.40 0.91 0.98 0.96 0.99 Knee extension 0.04 0.04 0.16 0.98 0.97 0.99 Ankle plantar flexion 0.33 0.10 0.68 0.99 0.98 0.99 Ankle dorsiflexion 0.74 0.38 0.88 0.98 0.96 0.99 ICC: intraclass correlation coefficient. 95%CI: 95% confidence interval. Table 6. Comparison of isometric muscle strength measurements: Interrater comparison with and without a belt, and for each tester, a comparison of measurements with versus without the belt (Pearson s correlation coefficient) Exertion Task Interrater comparison Pearson s correlation coefficient for measurements with vs without belt With Belt Without Belt Male Tester Female Tester Hip flexion 0.99 0.80 0.75 0.49 Hip extension 0.98 0.88 0.93 0.80 Hip abduction 0.99 0.72 0.92 0.67 Hip adduction 0.98 0.91 0.95 0.85 Hip external rotation 0.97 0.86 0.92 0.87 Hip internal rotation 0.99 0.86 0.93 0.82 Knee flexion 0.99 0.88 0.82 0.75 Knee extension 0.98 0.26 0.61 0.31 Ankle plantar flexion 0.99 0.69 0.72 0.53 Ankle dorsiflexion 0.98 0.85 0.85 0.70
42 J. Phys. Ther. Sci. Vol. 21, No. 1, 2009 0.97 in a study using two types of isokinetic dynamometers to measure the force of knee extensions performed at 60 degrees/s and 180 degrees/s. Wikholm et al. 7) and Agre et al. 8) reported that the interrater reliability of strength measurements made using a HHD was poorer for the lower than the upper limbs. In the present study, ICC for interrater reliability was 0.97 or higher for all measurements made with the HHD restrained by the belt, while compared with measurements made with a belt, ICC without the belt was low. Therefore, for testing healthy subjects, isometric muscle strength measurements made by the HHD restrained by a belt are at least as reliable as measurements made by an isokinetic HHD, and more reliable than isometric measurements made by a HHD without a restraining belt. The most plausible reason that measurements made by the HHD without the belt are poor is that the strength of the subject may exceed the resistance that the tester is able to apply to prevent the HHD from moving. In the present study, for both examiners, measurements made when using the belt were significantly higher than those made when not using the belt. Pearson s correlation coefficients for measurements made with versus without the belt were higher for the male tester than for the female tester. The average forces exerted during abduction, internal rotation and external rotation of the pelvis were all 20 kgf or less, which is clearly less that the reported limit of 30 kgf for measurements made by holding the HHD conventionally, that is without the aid of a restraining device 10). Even the male tester could not adequately maintain the position of the HHD without the belt, so the actual practical strength limit for using the HHD conventionally is probably less than the reported limit of 30 kgf in some circumstances. Therefore, isometric muscle strength measurements made using the restraining belt, are more reliable that those made conventionally, that is, without the belt, probably because the belt is able to compensate for the limit at which the HDD can be fixed in position by just the hand. Furthermore, this hand-held limit of 30 kgf was in practice demonstrated to be impractical, especially in the case of the female tester. We believe that when the belt is used to fix the device, the dynamometer can be reliably applied to quantitatively measure muscle strength of grade 3 (fair) or lower. However, because the trunk, lower back and other factors were not fixed or determined when the belt was used, future studies should evaluate whether values of measurements differ from those made with an isokinetic dynamometer. REFERENCES 1) Bohannon RW, Wikholm JB: Measurements of knee extension force obtained by two examiners of substantially different experience with a hand-held dynamometer. Isokinetics and Exercise Science, 1992, 2: 5 8. 2) Smidt GL, Albright JP, Deusinger RH, et al.: Pre-and postoperative functional changes in total knee patients. J Orthop Sports Phys Ther, 1984, 6: 25 29. 3) Bohannon RW, Andrews AW: Interrater reliability of hand-held dynamometry. Phys Ther, 1987, 67: 931. 4) Kwoh CK, Patrick MA, Munin MC: Inter-rater reliability for function and strength measurements in the acute care hospital after elective hip and knee arthroplasty. Arthritis Care Res, 1997, 10: 128 134. 5) McMahon LM, Burdett RG, Whitney SL: Effects of muscle group and placement site on reliability of handheld dynamometry strength measurements. J Orthop Sports Phys Ther, 1992, 15: 236 242. 6) Leggin BG, Neuman RM, Iannotti JP, et al.: Intrarater and interrater reliability of three isometric dynamometers in assessing shoulder strength. J Shoulder Elbow Surg, 1996, 5: 18 23. 7) Wikholm JB, Bohannon RW: Hand-held dynamometer measurements: tester strength makes a difference. J Orthop Sports Phys Ther, 1991, 13: 191 198. 8) Agre JC, Magness JL, Hull SZ, et al.: Strength testing with a portable dynamometer: reliability for upper and lower extremities. Arch Phys Med Rehabil, 1987, 68: 454 458. 9) Bandinelli S, Benvenuti E, Del Lungo I, et al.: Measuring muscular strength of the lower limbs by hand-held dynamometer: A standard protocol. Aging Clin Exp Res, 1999, 11: 287 293. 10) Hyde SA, Goddard CM, Scott OM: The Myometer: The development of a clinical tool. Physiotherapy, 1983, 69: 424 427. 11) Wiles CM, Karni Y: The measurement of muscle strength in patients with peripheral neuromuscular disorders. J Neurol Neurosurg Psychiatry, 1983, 46: 1006 1013. 12) Roebroeck ME, Harlaar J, Lankhorst GJ: Reliability assessment of isometric knee extension measurements with a computer-assisted hand-held dynamometer. Arch Phys Med Rehabil, 1998, 79: 442 448. 13) Gross MT, Huffman GM, Phillips CN, et al.: Intramachine and intermachine reliability of the Biodex and Cybex II for knee flexion and extension peak torque and angular work. J Orthop Sports Phys Ther, 1991, 13: 329 335.