- Email: [email protected]
The Short-Term Reliability of Grip Strength Measurement and the Effects of Posture and Grip Span
The Short-Term Reliability of Grip Strength Measurement and the Effects of Posture and Grip Span Tadayoshi Watanabe, MD, Yamagata, Japan, Kazuya Owashi, MD, Sakata, Japan, Yumiko Kanauchi, MD, Nariyuki Mura, MD, Masatoshi Takahara, MD, Toshihiko Ogino, MD, Yamagata, Japan
Purpose: Because of the difficulty in verifying the reliability and validity of grip strength there is still no consensus regarding its measurement, particularly short-term reliability. The present study was conducted to investigate the short-term reliability of grip strength measurement and the effects of posture and grip span. Methods: One hundred healthy subjects (50 men, 50 women; mean age, 38.2 y; range, 22–58 y) were evaluated. Grip strength was measured twice as a single set by using a dynamometer and the mean value for each hand was recorded. First 3 sets of measurements were performed using 2 different approaches: (1) continuous measurement without rest and (2) interval measurement with a 1-minute rest after each set. Next 1 set of measurements was performed with 3 types of grip span: standard grip span (which was measured as one half the distance between the index finger tip and the metacarpophalangeal joint flexion crease at the base of the thumb), ⫹10% of the standard grip span, and –10% of the standard grip span. Finally 1 set of measurements was performed in 3 postures: standing, sitting, and supine. Results: During continuous measurement the grip strength decreased gradually as the number of sets increased. During interval measurement, however, there was no change among sets for both genders and each hand. On the basis of this result subsequent studies were performed using interval measurement. There was no significant difference in maximum grip strength between the standard and ⫹10% of standard grip span measurements; however, the –10% of standard model gave the minimum grip strength in both genders. With regard to posture the minimum grip strength in both genders was obtained when the subject was supine, with no difference between standing and sitting. Conclusions: Our study showed that interval measurement with a 1-minute rest after each set yielded a constant value; therefore, we advocate this approach for rapid evaluation of grip strength under different conditions. In addition the influence of grip span and posture should be considered to maximize data accuracy. (J Hand Surg 2005;30A:603– 609. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Body position, fatigue, grip span, grip strength, measurement.
From the Department of Orthopaedic Surgery, Yamagata Comprehensive Rehabilitation and Education Center, Kaminoyama City, Yamagata, Japan; the Department of Orthopaedic Surgery, Municipal Sakata Hospital, Sengoku Town, Sakata City, Yamagata, Japan; and the Department of Orthopaedic Surgery, Yamagata University School of Medicine, Iida-Nishi, Yamagata, Japan. Received for publication February 17, 2004; accepted in revised form December 9, 2004. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Corresponding author: Tadayoshi Watanabe, MD, Department of Orthopaedic Surgery, Yamagata Comprehensive Rehabilitation and Education Center, 3-7-1 Kawasaki, Kaminoyama City, Yamagata 999-3145, Japan; e-mail: [email protected]. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A03-0028$30.00/0 doi:10.1016/j.jhsa.2004.12.007
The Journal of Hand Surgery
603
604
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
The measurement of grip strength is a standard feature of patient examination in orthopedic clinics and is a simple test that gives practical information on muscle, nerve, and joint disorders. It is also a good index of the rehabilitation process in lesions affecting the upper extremities. Grip strength is measured in several sport disciplines and on admission tests for various types of jobs. Practical research has been performed to develop highly reproducible and valid procedures for measuring grip strength. There have been many studies of grip strength in different situations (different angle of shoulder, elbow, forearm, and wrist;1–9 posture; 3,10 –12 trial frequency3,13–15; and grip span3,16 –19) but many properties of grip strength remain undefined. In particular the short-term reproducibility of tests in overcoming the fatigue effect seen in continuous measurement has not been defined. In the literature on fatigue effects during short-term measurements3,20 –23 there has been debate over the interval required between each test to maintain the maximum grip strength score. Rest intervals of 1 minute after each test have been used in some studies to investigate the effects of different situations on grip strength.5,6,9,19 In this study we compared the effects of fatigue in continuous measurement with that in interval measurement with a 1-minute rest. Our first aim was to clarify the effects of fatigue in continuous measurement and to determine whether interval measurement with a 1-minute rest could achieve reproducibility over multiple short-term tests. In addition, based on results of the first part of our study, our second aim was to investigate the effects of grip span and posture on grip strength.
Figure 1. Measure of grip span (arrow). Standard grip span is half the distance between the index finger tip and the metacarpophalangeal joint flexion crease at the base of the thumb.
measured as the standard grip span (S-GS) in this study (Fig. 1).
Measurement of hand grip strength. A digital
Our subjects were 100 healthy hospital workers (50 men, 50 women; doctors, nurses, therapists and administrators). Their mean age was 38.2 years (range for both men and women, 22–58 y; men: mean, 37.6 y; range, 22–58 y; women: mean, 38.8 y; range, 22–58 y). All subjects were in good health and free of any diseases or impairments of the upper limbs. Subjects were encouraged to do their best when performing the tests.
hand dynamometer (TKK 5401 Grip-D; Smedley, Takei, Tokyo, Japan) was used to measure grip strength (Fig. 2). Grip span could be adjusted for every 1 mm between 40 and 70 mm to obtain the standard grip span for each patient. Although the reported accuracy of the device was ⫾2 kg we ignored the accuracy in this study, according to the work of Reddon et al.23 During the test the subjects stood in the standard fashion as described in the Japanese physical fitness diagnosis test (1964)25 with the shoulder adducted and neutrally rotated, elbow extended fully, forearm and wrist in the neutral position, and feet shoulder-width apart. Subjects had to keep the dynamometer away from any part of the body. Grip strength was measured twice for each hand (first for the dominant hand and then for the nondominant hand) and the mean of the 2 trials for each hand was recorded as the score for a single set.
Methods Measurement of standard grip span. In accor-
Study of short-term grip-strength measurement. First, 3 sets of measurements (total, 6 trials)
dance with the method of Wasai24 half the distance between the index fingertip and the metacarpophalangeal joint flexion crease at the base of thumb was
were performed successively without rest as the continuous measurement sets (Fig. 3). On another day 3 sets of measurements with a 1-minute rest after each
Materials and Methods Subjects
Watanabe et al / Reliability of Grip Strength
605
Factors apart from grip span were the same as previously in this study. On another day, another set of measurements was performed in 3 postures: standing, sitting, and supine. In the supine position subjects lifted the elbow and hand about 3 to 5 cm from the floor. Factors apart from posture were the same as in the previous study. Whether continuous measurement or interval measurement was used depended on the results of the previous study.
Statistical Analysis
Figure 2. Digital hand dynamometer (Smedley; Takei, Tokyo, Japan).
set were performed as the interval measurement sets. The assessments of the 2 procedures were performed between 10 AM and 5 PM to avoid the effect of time.26
Study of grip span and posture. One set of measurements was performed with 3 sizes of grip span: S-GS, ⫹10% of the S-GS (⫹10% GS), and ⫺10% of the S-GS (⫺10% GS). For example, in the case of 50 mm as S-GS, ⫹10% GS is equivalent to 55 mm and ⫺10% GS is equivalent to 45 mm.
Statistical software (SPSS; SPSS Inc., Chicago, IL) was used for the statistical analysis of grip strength data. One-way repeated-measures analysis of variance was used to analyze the effect of continuous measurement, interval measurement, grip span, and posture on grip strength. Multiple comparison tests were adopted to analyze the data when the result of ANOVA was significant. An unpaired t test was used to compare the effects of age and gender on grip span. A p value less than .05 was considered evidence of a statistically significant finding.
Results There was no significant difference between genders in terms of mean age (p ⫽ .543). A significant difference in grip span was found between genders (men ⫽ 58.2 ⫾ 3.5 mm, women ⫽ 54.1 ⫾ 3.3 mm; p ⬍ .001). The grip strength in the first set of the dominant hand in men (DM) was 45.2 ⫾ 7.0 kg, in the nondominant hand in men (NM) it was 42.4 ⫾
Figure 3. Difference between continuous measurement and interval measurement. T, trial; R, 1-minute rest; D, measurement of dominant hand; N, measurement of nondominant hand.
606
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
Figure 4. Graph showing continuous measurement of 3 sets. † p ⬎ .05 from multiple comparisons.
5.9 kg, in the dominant hand in women (DW) it was 28.4 ⫾ 4.2 kg, and in the nondominant hand in women (NW) it was 26.5 ⫾ 3.5 kg. There was a significant difference in the first set of grip strengths between the dominant and nondominant hands in both genders. Therefore, statistical analyses were performed for each of the 4 groups: DM, NM, DW, NW. In continuous measurement of the 3 sets grip strength diminished significantly without regard to gender or dominant hand (p ⬍ .001) and there was a significant difference between each set in all 4 groups except between the second and third sets of DM and DW (p ⬍ .05) (Fig. 4). The degree of decrease between the first and second sets (DM ⫽ 1.9 kg, NM ⫽ 1.6 kg, DW ⫽ 0.7 kg, NW ⫽ 1.0 kg) was more remarkable than that between the second and third sets (DM ⫽ 0.7 kg, NM ⫽ 1.2 kg, DW ⫽ 0.4 kg, NW ⫽ 0.6 kg) in all 4 groups. When the sets were broken into trials, the grip strength decreased significantly in all 4 groups (p ⬍ .001) (Fig. 5). There was a significant difference between the first trial and all trials after the second trial in multiple comparison testing (p ⬍ .05). Between the first and second trials there was a significant difference only in NW (p ⬍ .001). In the interval measurement with a 1-minute rest no significant difference was confirmed between each set in any of the 4 groups (Fig. 6). Based on these results we used interval measurement in the second part of our study. Maximum grip strength was obtained by S-GS in men, but in DW
Figure 5. Graph showing continuous measurement of 6 trials. †p ⬎ .05 from multiple comparisons.
the maximum grip strength was obtained by ⫹10% GS and in NW it was obtained by both S-GS and ⫹10% GS. There was no significant difference in maximum grip strength between S-GS and ⫹10% GS in all 4 groups (Fig. 7). There was, however, a significant decrease between S-GS and ⫺10% GS in all groups (p ⬍ .05). There was also a significant difference between ⫹10% GS and ⫺10% GS in 3 groups (NM, DW, NW) (p ⬍ .05). Only in DM was there no significant difference between ⫹10% GS and ⫺10% GS.
Figure 6. Graph showing interval measurement with a 1-minute rest. †p ⬎ .05 from multiple comparisons.
Watanabe et al / Reliability of Grip Strength
Figure 7. Graph showing the effects of grip span. *p ⬍ .05 from multiple comparisons.
Figure 8 shows the effects of posture on grip strength. In all 4 groups the maximum grip strength was obtained when the subject was standing and the minimum grip strength was obtained when the subject was lying supine. Grip strength when lying supine was significantly lower than when in the other 2 postures (p ⬍ .001). Between the standing and sitting postures there was no significant difference within any group (Fig. 8).
Discussion There are some reports in the literature of the effects of fatigue on grip strength in short-term trials. Reddon et al23 reported a significant decrease in grip strength over 10 trials with a 30-second rest period between them but they used only 12 subjects. Mathiowetz15 mentioned that there were no significant differences in 3 trials with a 15-second rest period between them. Patterson and Baxter21 emphasized that using a protocol with a 1-minute rest between trials kept the maximum force the same during 3 trials, although using another protocol with a 5-second rest between trials led to a gradual reduction in the maximum force. The study of Patterson and Baxter21 suggested that interval measurement with a 1-minute rest during 3 trials allowed recovery from the fatigue effect and maintenance of maximum grip strength. We were unable to find any accurate studies of the effect of interval measurement using a 1-minute rest period, however, although this procedure has been used in other studies to investigate the effects of different situations on grip strength.5,6,9,19
607
Our study showed that interval measurement with a 1-minute rest after each set maintained constant grip strength over 3 sets, unlike continuous measurement, which led to gradually decreased grip strength. Presumably a 1-minute rest counteracted the effects of fatigue. This procedure is therefore acceptable for the investigation of grip strength in different situations. In addition we think this procedure gave us more accurate values to adapt to the patient’s condition during a limited clinical time. In continuous measurement the degree of decrease between the first and second sets was more remarkable than that between the second and third sets in all 4 groups. The dominant hand of both genders, however, did not show a significant decrease between the second and third sets compared with the nondominant hand in both genders. This result shows that late continuous measurements are not affected by fatigue, especially in the dominant hand. Further studies of the progression of this fatigue effect are warranted. In continuous measurement all groups except NW maintained their grip in the second trial; presumably the NW group had the weakest grip and was therefore the first to suffer from fatigue. Certain measurements of grip strength currently are used widely. One of these was suggested by the American Society of Hand Therapists in 1981.27 According to their protocol the subject is seated with the elbow flexed at 90° and a dynamometer (Jamar; Fit System Inc., Calgary, Canada) is used to measure the results of 3 trials. Another protocol has been performed in Japan since 1964 as a physical fitness
Figure 8. Graph showing the effects of posture (standing, sitting, supine). *p ⬍ .05 from multiple comparisons.
608
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
diagnosis test25: it uses a dynamometer (Smedley; Takei, Tokyo, Japan) with the subject standing with the elbow fully extended and measurements are taken from the mean of 2 trials. We adopted the earliermentioned Japanese procedure with a dynamometer (Smedley) but we varied the grip span measurement protocol. In the Japanese method the grip span is measured from the first web to the proximal interphalangeal joint of the index finger, which is flexed at 90°. This method seemed to influence easily the dynamic hand factor, so instead we used the method of Wasai and Shimada,24 which is based on static hand position. There have been discussions about which is the best grip span to give maximum grip strength or reliably reproducible grip strength. There also has been debate in many studies as to the merits of a fixed grip span that does not take into account the size of each subject versus a changeable grip span that is adaptable to each subject’s size.16 –19 RuizRuiz et al19 reported a gender-based distinction in that the male grip span that gave maximum grip strength was a fixed value (55 mm), but in women it was influenced by individual hand size. Fransson and Winkel17 reported that the grip span that gave maximum values was 50 to 60 mm for women and 55 to 65 mm for men. We used a grip span measured as half the distance between the index fingertip and the metacarpophalangeal flexion crease at the base of the thumb, according to the method of Wasai and Shimada24 and our mean sizes were 58.3 mm in men and 54.1 mm in women. Our results showed that a 10% reduction in grip size (52.5 mm in men, 48.7 mm in women) decreased grip strength significantly but that a 10% increase in grip size (64.1 mm in men, 59.5 mm in women) did not decrease grip strength significantly, especially in the women’s case. We therefore suspect that the grip span measurement that achieves maximum grip strength is between the distance determined by Wasai and Shimada24 and a 10% increase of that distance. These values are not far from those of the study by Fransson and Winkel.17 Because of the limited nature of our study and the small number of subjects we were unable to determine distinctions in results between the factors of gender and dominant hand. Any further studies should include both fixed and changeable measurements. With regard to the effects of posture on grip strength measurement Otsuka et al10 indicated that grip strength declined in order of standing, sitting,
and supine lying and found a significant difference only between supine lying and the others. Teraoka12 reported that there were highly significant differences in the order of standing, sitting, and supine. Richards,11 however, using a hand dynamometer (Jamar), found no significant differences between sitting and supine with the elbow in 90° flexion. Although these results are variable our results were consistent with those of the study by Otsuka et al.10 Presumably supine lying had the lowest value because of the greater effects of gravity and standing had the highest value because the mental and physical strain on the body was greater than in sitting or lying. Therefore, to keep conditions constant during measurement, the sitting position might be better than the standing position to exclude various factors, although standing can yield maximum grip strength. Further investigations of the reproducibility of results in each posture are needed. This study had some limitations. First, we could not determine whether the optimum resting time was universal because of the different degrees of the fatigue effect in individuals. To clarify the optimum resting time to maintain grip strength the optimum frequency and duration of rest periods need to be examined in a number of subjects. Second, different measuring devices, postures, and grip spans might change the results. Finally, we do no know whether these findings can transfer directly to different dynamometers. Some brands are more popular in other countries because of the different grip span scale. Nevertheless our study was performed under constant conditions and therefore we believe that our results are reliable.
References 1. Su C-Y, Lin J-H, Chien TH, Chien T-H, Cheng K-F, Sung Y-T. Grip strength in different positions of elbow and shoulder. Arch Phys Med Rehabil 1994;75:812– 815. 2. Kim L. Elbow positioning for maximum grip performance. J Hand Ther 2000;13:33–36. 3. Mathiowetz V, Rennells C, Donahoe L. Effect of elbow position on grip and key pinch strength. J Hand Surg 1985; 10A:694 – 697. 4. Mathiowetz V. Grip and pinch strength measurements. In: Amundsen L, ed. Muscles Strength Testing. New York: Churchill Livingstone, 1990:163–177. 5. Takagi D, Arai M, Inaniwa C, Oshima N, Saitou H, Satou M. Relationship between grip strength and elbow position in healthy subjects. Jpn Clin Rehabil 1997;25:651– 654. 6. Richards LG, Olson B, Palmiter-Thomas P. How forearm position affects grip strength. Am J Occup Ther 1996;50: 133–138. 7. De Smet L, Tirez B, Stappaerts K. Effect of forearm rotation on grip strength. Acta Orthop Belg 1998;64:360 –362.
Watanabe et al / Reliability of Grip Strength 8. Lamoreaux L, Hoffer MM. The effect of wrist deviation on grip and pinch strength. Clin Orthop 1995;314:152–155. 9. Fong PWK, Ng GYF. Effect of wrist positioning on the repeatability and strength of power grip. Am J Occup Ther 2001;55:212–216. 10. Otsuka T, Domen K, Liu M, Sonoda S, Saitoh E, Tsubahara A, et al. Grip strength of healthy elderly individuals— method of measurement and mean strength. Jpn J Rehabil Med 1994;31:731–735. 11. Richards LG. Posture effects on grip strength. Arch Phys Med Rehabil 1997;78:1154 –1156. 12. Teraoka T. Studies on the peculiarity of grip strength in relation to body positions and aging. Kobe J Med Sci 1979; 25:1–17. 13. Hamilton A, Balnave R, Adams R. Grip strength testing reliability. J Hand Ther 1994;7:163–170. 14. Mathiowetz V, Weber K, Volland G, Kashman N. Reliability and validity of grip and pinch strength evaluations. J Hand Surg 1984;9A:222–226. 15. Mathiowetz V. Effect of three trials on grip and pinch strength measurement. J Hand Ther 1990;3:195–198. 16. Firrell JC, Crain GM. Which setting of the dynamometer provides maximal grip strength. J Hand Surg 1996;21A: 397– 401. 17. Fransson C, Winkel J. Hand strength: the influence of grip span and grip type. Ergonomics 1991;34:881– 892. 18. Härkönen R, Piirtomaa M, Alaranta H. Grip strength and
19.
20.
21. 22. 23.
24.
25.
26.
27.
609
hand position of the dynamometer in 204 Finnish adults. J Hand Surg 1993;18B:129 –132. Ruiz-Ruiz J, Mesa JLM, Gutiérrez A, Castillo MJ. Hand size influences optimal grip span in women but not in men. J Hand Surg 2002;27A:897–901. Newman DG, Pearn J, Barnes A, Young CM, Kehoe M, Newman J. Norms for hand grip strength. Arch Dis Child 1984;59:453– 459. Patterson RP, Baxter T. A multiple muscle strength testing protocol. Arch Phys Med Rehabil 1988;69:366 –368. Pearn J, Bullock K. A portable hand-grip dynamometer. Aust Paediatr J 1979;15:107–109. Reddon JR, Stefanyk WO, Gill DM, Renney C. Hand dynamometer: effects of trials and sessions. Percept Mot Skills 1985;61:1195–1198. Wasai Y, Shimada T. Medical Textbook Series of Rehabilitation No.5: Measurement and Evaluation. 2nd ed. Tokyo: Ishiyaku Publishers Inc., 1987:237–238. Ministry of Education, Culture, Sports, Science and Technology. Japan Fitness Test. 2nd ed. Tokyo: Gyosei Corporation, 2002:98. McGarvey SR, Morrey BF, Askew LJ, An K-N. Reliability of isometric strength testing: temporal factors and strength variation. Clin Orthop 1984;185:301–305. American Society for Surgery of the Hand. The Hand: Examination and Diagnosis. 3rd ed. New York: Churchill Livingstone, 1990:121–122.
Purpose: Because of the difficulty in verifying the reliability and validity of grip strength there is still no consensus regarding its measurement, particularly short-term reliability. The present study was conducted to investigate the short-term reliability of grip strength measurement and the effects of posture and grip span. Methods: One hundred healthy subjects (50 men, 50 women; mean age, 38.2 y; range, 22–58 y) were evaluated. Grip strength was measured twice as a single set by using a dynamometer and the mean value for each hand was recorded. First 3 sets of measurements were performed using 2 different approaches: (1) continuous measurement without rest and (2) interval measurement with a 1-minute rest after each set. Next 1 set of measurements was performed with 3 types of grip span: standard grip span (which was measured as one half the distance between the index finger tip and the metacarpophalangeal joint flexion crease at the base of the thumb), ⫹10% of the standard grip span, and –10% of the standard grip span. Finally 1 set of measurements was performed in 3 postures: standing, sitting, and supine. Results: During continuous measurement the grip strength decreased gradually as the number of sets increased. During interval measurement, however, there was no change among sets for both genders and each hand. On the basis of this result subsequent studies were performed using interval measurement. There was no significant difference in maximum grip strength between the standard and ⫹10% of standard grip span measurements; however, the –10% of standard model gave the minimum grip strength in both genders. With regard to posture the minimum grip strength in both genders was obtained when the subject was supine, with no difference between standing and sitting. Conclusions: Our study showed that interval measurement with a 1-minute rest after each set yielded a constant value; therefore, we advocate this approach for rapid evaluation of grip strength under different conditions. In addition the influence of grip span and posture should be considered to maximize data accuracy. (J Hand Surg 2005;30A:603– 609. Copyright © 2005 by the American Society for Surgery of the Hand.) Key words: Body position, fatigue, grip span, grip strength, measurement.
From the Department of Orthopaedic Surgery, Yamagata Comprehensive Rehabilitation and Education Center, Kaminoyama City, Yamagata, Japan; the Department of Orthopaedic Surgery, Municipal Sakata Hospital, Sengoku Town, Sakata City, Yamagata, Japan; and the Department of Orthopaedic Surgery, Yamagata University School of Medicine, Iida-Nishi, Yamagata, Japan. Received for publication February 17, 2004; accepted in revised form December 9, 2004. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Corresponding author: Tadayoshi Watanabe, MD, Department of Orthopaedic Surgery, Yamagata Comprehensive Rehabilitation and Education Center, 3-7-1 Kawasaki, Kaminoyama City, Yamagata 999-3145, Japan; e-mail: [email protected]. Copyright © 2005 by the American Society for Surgery of the Hand 0363-5023/05/30A03-0028$30.00/0 doi:10.1016/j.jhsa.2004.12.007
The Journal of Hand Surgery
603
604
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
The measurement of grip strength is a standard feature of patient examination in orthopedic clinics and is a simple test that gives practical information on muscle, nerve, and joint disorders. It is also a good index of the rehabilitation process in lesions affecting the upper extremities. Grip strength is measured in several sport disciplines and on admission tests for various types of jobs. Practical research has been performed to develop highly reproducible and valid procedures for measuring grip strength. There have been many studies of grip strength in different situations (different angle of shoulder, elbow, forearm, and wrist;1–9 posture; 3,10 –12 trial frequency3,13–15; and grip span3,16 –19) but many properties of grip strength remain undefined. In particular the short-term reproducibility of tests in overcoming the fatigue effect seen in continuous measurement has not been defined. In the literature on fatigue effects during short-term measurements3,20 –23 there has been debate over the interval required between each test to maintain the maximum grip strength score. Rest intervals of 1 minute after each test have been used in some studies to investigate the effects of different situations on grip strength.5,6,9,19 In this study we compared the effects of fatigue in continuous measurement with that in interval measurement with a 1-minute rest. Our first aim was to clarify the effects of fatigue in continuous measurement and to determine whether interval measurement with a 1-minute rest could achieve reproducibility over multiple short-term tests. In addition, based on results of the first part of our study, our second aim was to investigate the effects of grip span and posture on grip strength.
Figure 1. Measure of grip span (arrow). Standard grip span is half the distance between the index finger tip and the metacarpophalangeal joint flexion crease at the base of the thumb.
measured as the standard grip span (S-GS) in this study (Fig. 1).
Measurement of hand grip strength. A digital
Our subjects were 100 healthy hospital workers (50 men, 50 women; doctors, nurses, therapists and administrators). Their mean age was 38.2 years (range for both men and women, 22–58 y; men: mean, 37.6 y; range, 22–58 y; women: mean, 38.8 y; range, 22–58 y). All subjects were in good health and free of any diseases or impairments of the upper limbs. Subjects were encouraged to do their best when performing the tests.
hand dynamometer (TKK 5401 Grip-D; Smedley, Takei, Tokyo, Japan) was used to measure grip strength (Fig. 2). Grip span could be adjusted for every 1 mm between 40 and 70 mm to obtain the standard grip span for each patient. Although the reported accuracy of the device was ⫾2 kg we ignored the accuracy in this study, according to the work of Reddon et al.23 During the test the subjects stood in the standard fashion as described in the Japanese physical fitness diagnosis test (1964)25 with the shoulder adducted and neutrally rotated, elbow extended fully, forearm and wrist in the neutral position, and feet shoulder-width apart. Subjects had to keep the dynamometer away from any part of the body. Grip strength was measured twice for each hand (first for the dominant hand and then for the nondominant hand) and the mean of the 2 trials for each hand was recorded as the score for a single set.
Methods Measurement of standard grip span. In accor-
Study of short-term grip-strength measurement. First, 3 sets of measurements (total, 6 trials)
dance with the method of Wasai24 half the distance between the index fingertip and the metacarpophalangeal joint flexion crease at the base of thumb was
were performed successively without rest as the continuous measurement sets (Fig. 3). On another day 3 sets of measurements with a 1-minute rest after each
Materials and Methods Subjects
Watanabe et al / Reliability of Grip Strength
605
Factors apart from grip span were the same as previously in this study. On another day, another set of measurements was performed in 3 postures: standing, sitting, and supine. In the supine position subjects lifted the elbow and hand about 3 to 5 cm from the floor. Factors apart from posture were the same as in the previous study. Whether continuous measurement or interval measurement was used depended on the results of the previous study.
Statistical Analysis
Figure 2. Digital hand dynamometer (Smedley; Takei, Tokyo, Japan).
set were performed as the interval measurement sets. The assessments of the 2 procedures were performed between 10 AM and 5 PM to avoid the effect of time.26
Study of grip span and posture. One set of measurements was performed with 3 sizes of grip span: S-GS, ⫹10% of the S-GS (⫹10% GS), and ⫺10% of the S-GS (⫺10% GS). For example, in the case of 50 mm as S-GS, ⫹10% GS is equivalent to 55 mm and ⫺10% GS is equivalent to 45 mm.
Statistical software (SPSS; SPSS Inc., Chicago, IL) was used for the statistical analysis of grip strength data. One-way repeated-measures analysis of variance was used to analyze the effect of continuous measurement, interval measurement, grip span, and posture on grip strength. Multiple comparison tests were adopted to analyze the data when the result of ANOVA was significant. An unpaired t test was used to compare the effects of age and gender on grip span. A p value less than .05 was considered evidence of a statistically significant finding.
Results There was no significant difference between genders in terms of mean age (p ⫽ .543). A significant difference in grip span was found between genders (men ⫽ 58.2 ⫾ 3.5 mm, women ⫽ 54.1 ⫾ 3.3 mm; p ⬍ .001). The grip strength in the first set of the dominant hand in men (DM) was 45.2 ⫾ 7.0 kg, in the nondominant hand in men (NM) it was 42.4 ⫾
Figure 3. Difference between continuous measurement and interval measurement. T, trial; R, 1-minute rest; D, measurement of dominant hand; N, measurement of nondominant hand.
606
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
Figure 4. Graph showing continuous measurement of 3 sets. † p ⬎ .05 from multiple comparisons.
5.9 kg, in the dominant hand in women (DW) it was 28.4 ⫾ 4.2 kg, and in the nondominant hand in women (NW) it was 26.5 ⫾ 3.5 kg. There was a significant difference in the first set of grip strengths between the dominant and nondominant hands in both genders. Therefore, statistical analyses were performed for each of the 4 groups: DM, NM, DW, NW. In continuous measurement of the 3 sets grip strength diminished significantly without regard to gender or dominant hand (p ⬍ .001) and there was a significant difference between each set in all 4 groups except between the second and third sets of DM and DW (p ⬍ .05) (Fig. 4). The degree of decrease between the first and second sets (DM ⫽ 1.9 kg, NM ⫽ 1.6 kg, DW ⫽ 0.7 kg, NW ⫽ 1.0 kg) was more remarkable than that between the second and third sets (DM ⫽ 0.7 kg, NM ⫽ 1.2 kg, DW ⫽ 0.4 kg, NW ⫽ 0.6 kg) in all 4 groups. When the sets were broken into trials, the grip strength decreased significantly in all 4 groups (p ⬍ .001) (Fig. 5). There was a significant difference between the first trial and all trials after the second trial in multiple comparison testing (p ⬍ .05). Between the first and second trials there was a significant difference only in NW (p ⬍ .001). In the interval measurement with a 1-minute rest no significant difference was confirmed between each set in any of the 4 groups (Fig. 6). Based on these results we used interval measurement in the second part of our study. Maximum grip strength was obtained by S-GS in men, but in DW
Figure 5. Graph showing continuous measurement of 6 trials. †p ⬎ .05 from multiple comparisons.
the maximum grip strength was obtained by ⫹10% GS and in NW it was obtained by both S-GS and ⫹10% GS. There was no significant difference in maximum grip strength between S-GS and ⫹10% GS in all 4 groups (Fig. 7). There was, however, a significant decrease between S-GS and ⫺10% GS in all groups (p ⬍ .05). There was also a significant difference between ⫹10% GS and ⫺10% GS in 3 groups (NM, DW, NW) (p ⬍ .05). Only in DM was there no significant difference between ⫹10% GS and ⫺10% GS.
Figure 6. Graph showing interval measurement with a 1-minute rest. †p ⬎ .05 from multiple comparisons.
Watanabe et al / Reliability of Grip Strength
Figure 7. Graph showing the effects of grip span. *p ⬍ .05 from multiple comparisons.
Figure 8 shows the effects of posture on grip strength. In all 4 groups the maximum grip strength was obtained when the subject was standing and the minimum grip strength was obtained when the subject was lying supine. Grip strength when lying supine was significantly lower than when in the other 2 postures (p ⬍ .001). Between the standing and sitting postures there was no significant difference within any group (Fig. 8).
Discussion There are some reports in the literature of the effects of fatigue on grip strength in short-term trials. Reddon et al23 reported a significant decrease in grip strength over 10 trials with a 30-second rest period between them but they used only 12 subjects. Mathiowetz15 mentioned that there were no significant differences in 3 trials with a 15-second rest period between them. Patterson and Baxter21 emphasized that using a protocol with a 1-minute rest between trials kept the maximum force the same during 3 trials, although using another protocol with a 5-second rest between trials led to a gradual reduction in the maximum force. The study of Patterson and Baxter21 suggested that interval measurement with a 1-minute rest during 3 trials allowed recovery from the fatigue effect and maintenance of maximum grip strength. We were unable to find any accurate studies of the effect of interval measurement using a 1-minute rest period, however, although this procedure has been used in other studies to investigate the effects of different situations on grip strength.5,6,9,19
607
Our study showed that interval measurement with a 1-minute rest after each set maintained constant grip strength over 3 sets, unlike continuous measurement, which led to gradually decreased grip strength. Presumably a 1-minute rest counteracted the effects of fatigue. This procedure is therefore acceptable for the investigation of grip strength in different situations. In addition we think this procedure gave us more accurate values to adapt to the patient’s condition during a limited clinical time. In continuous measurement the degree of decrease between the first and second sets was more remarkable than that between the second and third sets in all 4 groups. The dominant hand of both genders, however, did not show a significant decrease between the second and third sets compared with the nondominant hand in both genders. This result shows that late continuous measurements are not affected by fatigue, especially in the dominant hand. Further studies of the progression of this fatigue effect are warranted. In continuous measurement all groups except NW maintained their grip in the second trial; presumably the NW group had the weakest grip and was therefore the first to suffer from fatigue. Certain measurements of grip strength currently are used widely. One of these was suggested by the American Society of Hand Therapists in 1981.27 According to their protocol the subject is seated with the elbow flexed at 90° and a dynamometer (Jamar; Fit System Inc., Calgary, Canada) is used to measure the results of 3 trials. Another protocol has been performed in Japan since 1964 as a physical fitness
Figure 8. Graph showing the effects of posture (standing, sitting, supine). *p ⬍ .05 from multiple comparisons.
608
The Journal of Hand Surgery / Vol. 30A No. 3 May 2005
diagnosis test25: it uses a dynamometer (Smedley; Takei, Tokyo, Japan) with the subject standing with the elbow fully extended and measurements are taken from the mean of 2 trials. We adopted the earliermentioned Japanese procedure with a dynamometer (Smedley) but we varied the grip span measurement protocol. In the Japanese method the grip span is measured from the first web to the proximal interphalangeal joint of the index finger, which is flexed at 90°. This method seemed to influence easily the dynamic hand factor, so instead we used the method of Wasai and Shimada,24 which is based on static hand position. There have been discussions about which is the best grip span to give maximum grip strength or reliably reproducible grip strength. There also has been debate in many studies as to the merits of a fixed grip span that does not take into account the size of each subject versus a changeable grip span that is adaptable to each subject’s size.16 –19 RuizRuiz et al19 reported a gender-based distinction in that the male grip span that gave maximum grip strength was a fixed value (55 mm), but in women it was influenced by individual hand size. Fransson and Winkel17 reported that the grip span that gave maximum values was 50 to 60 mm for women and 55 to 65 mm for men. We used a grip span measured as half the distance between the index fingertip and the metacarpophalangeal flexion crease at the base of the thumb, according to the method of Wasai and Shimada24 and our mean sizes were 58.3 mm in men and 54.1 mm in women. Our results showed that a 10% reduction in grip size (52.5 mm in men, 48.7 mm in women) decreased grip strength significantly but that a 10% increase in grip size (64.1 mm in men, 59.5 mm in women) did not decrease grip strength significantly, especially in the women’s case. We therefore suspect that the grip span measurement that achieves maximum grip strength is between the distance determined by Wasai and Shimada24 and a 10% increase of that distance. These values are not far from those of the study by Fransson and Winkel.17 Because of the limited nature of our study and the small number of subjects we were unable to determine distinctions in results between the factors of gender and dominant hand. Any further studies should include both fixed and changeable measurements. With regard to the effects of posture on grip strength measurement Otsuka et al10 indicated that grip strength declined in order of standing, sitting,
and supine lying and found a significant difference only between supine lying and the others. Teraoka12 reported that there were highly significant differences in the order of standing, sitting, and supine. Richards,11 however, using a hand dynamometer (Jamar), found no significant differences between sitting and supine with the elbow in 90° flexion. Although these results are variable our results were consistent with those of the study by Otsuka et al.10 Presumably supine lying had the lowest value because of the greater effects of gravity and standing had the highest value because the mental and physical strain on the body was greater than in sitting or lying. Therefore, to keep conditions constant during measurement, the sitting position might be better than the standing position to exclude various factors, although standing can yield maximum grip strength. Further investigations of the reproducibility of results in each posture are needed. This study had some limitations. First, we could not determine whether the optimum resting time was universal because of the different degrees of the fatigue effect in individuals. To clarify the optimum resting time to maintain grip strength the optimum frequency and duration of rest periods need to be examined in a number of subjects. Second, different measuring devices, postures, and grip spans might change the results. Finally, we do no know whether these findings can transfer directly to different dynamometers. Some brands are more popular in other countries because of the different grip span scale. Nevertheless our study was performed under constant conditions and therefore we believe that our results are reliable.
References 1. Su C-Y, Lin J-H, Chien TH, Chien T-H, Cheng K-F, Sung Y-T. Grip strength in different positions of elbow and shoulder. Arch Phys Med Rehabil 1994;75:812– 815. 2. Kim L. Elbow positioning for maximum grip performance. J Hand Ther 2000;13:33–36. 3. Mathiowetz V, Rennells C, Donahoe L. Effect of elbow position on grip and key pinch strength. J Hand Surg 1985; 10A:694 – 697. 4. Mathiowetz V. Grip and pinch strength measurements. In: Amundsen L, ed. Muscles Strength Testing. New York: Churchill Livingstone, 1990:163–177. 5. Takagi D, Arai M, Inaniwa C, Oshima N, Saitou H, Satou M. Relationship between grip strength and elbow position in healthy subjects. Jpn Clin Rehabil 1997;25:651– 654. 6. Richards LG, Olson B, Palmiter-Thomas P. How forearm position affects grip strength. Am J Occup Ther 1996;50: 133–138. 7. De Smet L, Tirez B, Stappaerts K. Effect of forearm rotation on grip strength. Acta Orthop Belg 1998;64:360 –362.
Watanabe et al / Reliability of Grip Strength 8. Lamoreaux L, Hoffer MM. The effect of wrist deviation on grip and pinch strength. Clin Orthop 1995;314:152–155. 9. Fong PWK, Ng GYF. Effect of wrist positioning on the repeatability and strength of power grip. Am J Occup Ther 2001;55:212–216. 10. Otsuka T, Domen K, Liu M, Sonoda S, Saitoh E, Tsubahara A, et al. Grip strength of healthy elderly individuals— method of measurement and mean strength. Jpn J Rehabil Med 1994;31:731–735. 11. Richards LG. Posture effects on grip strength. Arch Phys Med Rehabil 1997;78:1154 –1156. 12. Teraoka T. Studies on the peculiarity of grip strength in relation to body positions and aging. Kobe J Med Sci 1979; 25:1–17. 13. Hamilton A, Balnave R, Adams R. Grip strength testing reliability. J Hand Ther 1994;7:163–170. 14. Mathiowetz V, Weber K, Volland G, Kashman N. Reliability and validity of grip and pinch strength evaluations. J Hand Surg 1984;9A:222–226. 15. Mathiowetz V. Effect of three trials on grip and pinch strength measurement. J Hand Ther 1990;3:195–198. 16. Firrell JC, Crain GM. Which setting of the dynamometer provides maximal grip strength. J Hand Surg 1996;21A: 397– 401. 17. Fransson C, Winkel J. Hand strength: the influence of grip span and grip type. Ergonomics 1991;34:881– 892. 18. Härkönen R, Piirtomaa M, Alaranta H. Grip strength and
19.
20.
21. 22. 23.
24.
25.
26.
27.
609
hand position of the dynamometer in 204 Finnish adults. J Hand Surg 1993;18B:129 –132. Ruiz-Ruiz J, Mesa JLM, Gutiérrez A, Castillo MJ. Hand size influences optimal grip span in women but not in men. J Hand Surg 2002;27A:897–901. Newman DG, Pearn J, Barnes A, Young CM, Kehoe M, Newman J. Norms for hand grip strength. Arch Dis Child 1984;59:453– 459. Patterson RP, Baxter T. A multiple muscle strength testing protocol. Arch Phys Med Rehabil 1988;69:366 –368. Pearn J, Bullock K. A portable hand-grip dynamometer. Aust Paediatr J 1979;15:107–109. Reddon JR, Stefanyk WO, Gill DM, Renney C. Hand dynamometer: effects of trials and sessions. Percept Mot Skills 1985;61:1195–1198. Wasai Y, Shimada T. Medical Textbook Series of Rehabilitation No.5: Measurement and Evaluation. 2nd ed. Tokyo: Ishiyaku Publishers Inc., 1987:237–238. Ministry of Education, Culture, Sports, Science and Technology. Japan Fitness Test. 2nd ed. Tokyo: Gyosei Corporation, 2002:98. McGarvey SR, Morrey BF, Askew LJ, An K-N. Reliability of isometric strength testing: temporal factors and strength variation. Clin Orthop 1984;185:301–305. American Society for Surgery of the Hand. The Hand: Examination and Diagnosis. 3rd ed. New York: Churchill Livingstone, 1990:121–122.