Measurement of Grip Strength: Validity and Reliability of the Sphygmomanometer and Jamar Grip Dynamometer George F. Hamilton, MS, PT, OCS' Carolyn McDonald, BS, PT2 Thomas C. Chenier, PhD3 Copyright 1992. All rights reserved. H and grip strength is a fundamental procedure used by therapists and physicians to assess patient status fbllowing injuries, surgical techniques, and treatment procedures to the hand and upper extremity. An accurate, quantifiable assessment of hand grip strength helps the clinician establish realistic treatment goals, provides treatment outcome data, and is frequently utilized during determination of hand disability ratings. Over the years, several instruments and methods for measuring hand grip strength have been developed. In the 1950s. instruments such as the Sklar, Narrangansett, Geckler, and Collins dynamometers were used but were proven unsatisfactory for standardized grip strength measurement (5, 1 I). In 1954, Bechtol introduced the Jamar dynamometer. It consisted of a sealed hydraulic system with adjustable hand spacings that registered hand grip force in pounds per square inch (PSI) (1). It became the most accepted instrument in California for quantitative measurement of hand grip (5). Brewer et al discussed two methods for hand grip strength measurements. The Lansbury method, used Quantitative measurement of grip strength is an important variable when plotting the progress of a hand-injured patient. When utilizing traditional commercially available apparatuses, obtaining meaningful grip strength measurement in these subjects is frequently difficult due to severe deformity, high tissue sensitivity, and low levels of force generated. The purpose of this study was to measure hand grip strength using two instruments having different physical characteristics and units of measurement to determine the reliability of repeated measures with each instrument. Additionally, validity of the sphygmomanometer for strength measurement was established through comparison with the values obtained from measurements using the research-validated lamar dynamometer. Twenty-nine right hand dominant female college-age subjects volunteered to perform hand grip strength testing. Measurements were taken with a sphygmomanometer and a lamar dynamometer while utilizing standardized measurement procedures. A Spearman Rho correlation coefficient test utilized in measuring within-instrument reliability showed a high correlation for each instrument at.85 for the sphygmomanometer and.82 for the lamar dynamometer. Construct validity testing performed to determine validity of the measurements by the sphygmomanometer compared with the lamar dynamometer produced a.75 correlation. A formula for conversion of the sphygmomanometer scores into lamar units was developed to enhance reporting of sphygmomanometer scores utilizing the lamar standard. The study showed that the sphygmomanometer and lamar dynamometer exhibit good within-instrument reliability. Validity of the sphygmomanometer as a grip measurement device is acceptable and reportable using the conversion formula developed. Therefore, it can be utilized with confidence as essentially equal to the lamar unit for grip strength measurement. Key Words: grip strength, hand injury, reliability Professor, Department of Physical Therapy, School of Allied Health Sciences, East Carolina Uniwrsity, Creenville, NC 27858 Staff physical therapist, Department of Physical Therapy, Nash General Hospital, Rocky Mount, NC 'Assistant professor, Epidemiology and Statistics, School of Allied Health Sciences, East Carolina University, Creenville. NC prior to 1958, consisted of a sphygmomanometer cuff folded twice and inflated to 20 mmhg. The Winthrop torqueometer was a device used to measure hand grip and rotational ability (2). Kirkpatrick reported the findings of a California Commission that examined grip strength measurement instruments based upon pneumatic, spring, and hydraulic systems. They concluded that grip is a force rather than a pressure and that it should be measured in force units such as Ibs or gms; they recommended use of the Jamar dynamom- JOSPT Volume 16 Number 5 November 1992
Copyright 1992. All rights reserved. eter (5). Most of the recent studies of grip strength measurement have reported the Jamar dynamometer to be the most reliable and accurate device for measurement of hand grip strength (6-9). However, Solgaard et al tested three units utilized in grip strength measurement; two were steel spring force dynamometers ("Collins" and "My-Gripper") and one was pneumatic ("Martin Vigorimeter"). Accuracy of the units was tested by placing each instrument between two compressible plates of a machine that yielded force and deflection information. Machine test results demonstrated near linearity of all units, with the pneumatic unit performing the best. The instruments were also subjected to a clinical test of measured grip strength in 100 normal men and women subjects. The pneumatic unit demonstrated the highest precision and was deemed to be the superior instrument for grip strength measurement by those investigators (1 2). Due to problems of discomfort conformity and convenience in hand-injured patients, many clinicians continue to utilize the soft cuff of the Sphygmomanometer to measure grip strength. The purpose of this study was to determine the repeated measures reliability of the sphygmomanometer under well-controlled test conditions, to compare those reliability outcomes to values obtained for the research-validated Jamar dynamometer, and to test for construct validity of the sphygmomanometer's measurement scale (mmhg) by comparison with the Jamar (PSI) unit. If the sphygmomanometer proved to be valid and reliable, then clinicians might use this instrument with greater confidence when obtaining hand grip measurements. known history of orthopaedic disorders of their dominant hand volunteered to participate in this study. Descriptive characteristics of the subjects are listed in Table 1. Apparatus Apparatuses used in this experiment were an Aneroid type adult sphygmomanometer U.A. Preston, Inc., 60 Page Road, Clifton, New Jersey, 270 12) (Figure 1). which measures force in units of mmhg, and a Jamar dynamometer (Item 3363, G.E. Miller, Inc., 484 Broadway, Yonkers, New York, 10705) (Figure 2). which measures in units of PSI. Many dinicians continue to utilize the soff cuff of the sphygmomanometer to measure grip strength. The sphygmomanometer cuff was evenly rolled, forming a circumference of approximately 7 in to conform with a normal functional hand position for grip.a rubber band was placed around each end of the cuff to hold it in position. The cuff was inflated to 20 mmhg, which was the starting position for measurement of each subject. The Jamar dynamometer was set at the third handle space, which is reported to be the most commonly used position for measurement (1). Additionally, it provided the closest hand grip dimension to the rolled up sphygmomanometer cuff. The Jamar dynamometer was calibrated by suspending a range of weights, comparable to anticipated study values, from the handle of the FIGURE 1. The sphygmomanometer. FIGURE 2. The lamar dynamometer. METHODS Subjects Twenty-nine normal healthy female college-age subjects with no Factor Units Total Mean SD Age (yean) 29 23.8 4.9 Height (in) 29 66.1 2.2 Weight (Ibd 29 131.9 26.3 TABLE 1. Subject characteristics. Volume 16 Number 5 November 1992 JOSPT
Copyright 1992. All rights reserved. fixated instrument. Dial readings obtained were identical to those of the suspended weights, thus, confirming the validity of the Jamar dynamometer with which the sphygmomanometer would be compared. Design Under the direction of the same operator, 29 subjects performed hand grip strength testing for each instrument. A mean score was calculated from three measurements per instrument, obtained at each of three independent measurement sessions occurring approximately 1 week apart. Spearman Rho correlation coefficient tests were administered to determine relationships within and between the instruments. Repeated measures reliability was obtained by calculating correlation coefficients between sessions 1 and 2, sessions 1 and 3, and sessions 2 and 3 for each instrument. The relationship between the two instruments was obtained by applying a Spearman Rho correlation coefficient test utilizing the mean value of all observations from each instrument and by plotting a simple linear regression to assess construct validity of the sphygmomanometer for measurement of hand grip strength. Procedure All subjects signed a personal consent form, after which we recorded their ID number, age, height, and weight. The American Society of Hand Therapists' standardized arm position for hand strength testing was utilized (4). Each subject was positioned in a straight back chair with both feet flat on the floor. Arm positioning was demonstrated by the operator, then each subject was instructed to place her left hand on her right thigh and assume a position of adducted and neutrally rotated shoulders. For the arm to be tested, the elbow was flexed to 90". the forearm and wrist were in neutral positions, and the fingers were flexed as needed for a maximal contraction (Figure 3). Fach subject was instructed to breathe in through her nose and blow out through pursed lips as a maxinlum grip effort was made. At this time, a verbal command of "Squeeze! Harder! Harder! Relax!" was given by the examiner. Demonstration of maximum hand grip performance was given prior to the first session, and re-instruction was given prior to the other two sessions as needed. The instrument to be tested first was assigned by random order to each subject. Four minutes or Results demonstrated high within-instrument repeated measures reliability and an acceptable correlation between sphygmomanometer and lamar dynamometer measurements. greater were allowed between each grip measurement. The mean score among three trials of each instrument was recorded for data calculations. The second and third sessions for hand grip measurements were repeated approximately 1 week apart using each instrument. Extraneous variables were controlled by using the same room with an average temperature of 74" F, subjects of the same gender, right hand dominance, and approximately the same age group, and approximately the same time of day. The FIGURE 3. Positioning of a subject while performing hand grip strength measurements using the sphygmomanometer and the lamar dynamometer. same instructions were given to each subject regarding position of the up per extremity, hand grip strength performance, and breathing technique. Each subject was encouraged not to do any strenuous activity with the dominant upper extremity during the study. RESULTS Mean and standard deviation grip strength scores for each instrument at each of the three sessions are presented in Table 2. Spearman Rho correlation coefficient test outcomes for each possible pair of Jamar and sphygmonianometer readings collected over all sessions are presented in Table 3. The average r value for the three observational periods was.85 for the sphygmomanometer and.8 1 for the Jamar (Table 3A), each with a statistically significant relationship (/I <.OO 1) for scores obtained between the three test intervals. Sphygnionianometer vs Jamar mean scores for all observational periods produced a correlation coefficient of.75 (Table 3C). Using the inverse regression technique (3). we calculated an in- JOSPT * Volume 16 * Number 5 * November 1992
Copyright 1992. All rights reserved. Sessions 1 2 3 Mean Sphygmomanometer (mmhg) Mean 205.96 200.17 199.83 203.12 SD 22.53 21.53 14.85 21.10 jamar dynamometer (PSI) Mean 63.09 69.21 66.23 64.83 SD 9.43 8.08 10.50 9.07 TABLE 2. Mean and standard deviation for three grip strength sessions using the sphygmomanometer and the lamar dynamometer, plus the mean and standard deviation values for all grip sessions. A Sphyg 1 vs Sphyg 2 Rho =.8426 p <.001 Sphyg 1 vs Sphyg 3 Rho =.8616 p <.001 Sphyg 2 vs Sphyg 3 Rho =,8401 p <.001 Jamar 1 vs Jamar 2 Rho =,8764 p <.001 Jamar 1 vs Jamar 3 Rho =.8049 p <.001 Jamar 2 vs Jamar 3 Rho =.7844 p <.001 B Sphyg 1 vs Jamar 1 Rho =.6079 p <.001 Sphyg 2 vs Jamar 2 Rho =.7455 p <.001 C Sphyg 3 vs Jamar 3 Rho =.5723 p =.001 Mean scores of Sphyg vs mean scores of jamar Rho =.7497 p <.001 -- -- TABLE 3. Spearman Rho correlation coefficient values showing within (A) and between (6) instrument rekability in hand grip strength measurement for the sphygmomanometer and the lamar dynamometer. FIGURE 4. Inverse regression prediction line showmg the relationship between sphygmomanometer (mmhg) and lamar (PSI) readings and the formulas for conversion of sphygmomanometer to lamar scores. verse regression line showing the re- lationship between Jamar and reliability and an acceptable correlation between sphygmomanometer and Jamar dynamometer measurements. Although statements are available in the literature expressing the inadequacy of pressure-related instrumentation for grip strength measurement (5. 8). test results from Solgaard et al demonstrated that a pneumatic-based (Martin Vigorimeter) ball type unit provided highly linear and reliable data for grip strength measurement (12). Robertson and Deitz have advocated the use of the Martin Vigorimeter for measuring grip strength in preschool children due to its ability to accurately measure spherical grip strength, which they indicated was a more effective indicator of hand monianometer scores and a formula function in the child (1. 0)., to convert sphygmomanometer Our scores for s~hv~momanom- 1, <7 scores to equivalent Jamar values eter and Jamar grip strength testing (Figure 4). The formula is Jamar = were similar, each yielding correla-.!j4 s~h~gnlonlanometer - 45.12. tion values of.80 or greater. These <I findings conform with the previous DISCUSSION work of Mathiowetz et al, who reported a high correlation (.80 or Results demonstrated high greater) in hand grip strength using within-instrument repeated measures the Jamar dynamometer (8). Data obtained from the two instruments were used to construct a formula Uamar =.54 x Sphygmomanometer - 45.12) by which measurements on the sphygmomanometer (mmhg) could be converted into the corresponding Jamar (PSI) measurements. An inverse regression technique was used to make these conversions (3). For these data, it would be inappropriate to apply the usual regresson analysis. One of the assumptions of the usual regression analysis is that the predictor variable is measured without error, or that the error of the predictor varible is small relative to the error of the response variable. When a regression line is fit to data where the predictor variable is measured with error, the resulting estimate of the slope of the regression line is biased. In this instance, measurements associated with the sphygmomanometer that are subject to error are being used to predict the values obtained from the Jamar. When measuring grip strength in the hand-disabled subject, there are multiple advantages of the Sphygmomanometer over the Jamar unit: I) ready availability of a sphygmomanometer in most clinics, 2) a soft compliant surface that may produce less discomfort to the injured hand during testing, and 3) a scale with smaller increments than the Jamar and, therefore, greater sensitivity to small changes in strength. The outcome of this study demonstrates the ability of the sphygmomanometer to provide acceptable levels of repeated measures reliability in the normal population. However, further studies are needed within the hand-injured population to fully validate its application. JOSPT REFERENCES 1. Bechtol CO: The use of a dynarnorneter with adjustable handle spacings. I Bone joint Surg 36A(4):820-824, 1954 2. Brewer K, Cuyatt AR, Scott IT: Corn- Volume 16 Number 5 November 1992 JOSPT
paring grip strength. Physiother 61:118, 1975 3. Draper NR, Smith H: Applied Regression Analysis (2nd Ed), pp 47-5 1. New York: lohn Wiley & Sons, 198 1 4. Fess EE, Moran CA: Clinical assessment recommendations booklet. American Society of Hand Therapists, 1002 Van Dora Springs Road, Suite 10 1, Garner, NC 27529, 1981 5. Kirkpatrick I: Evaluation of grip loss: A factor of permanent disability in California. Ind Med Surg 26:285-289, 1957 6. Mathiowetz V, Kashman N, Volland G, Weber K, Dowe M, Rogers S: Crip and pinch strength: Normative data for adults. Arch Phys Med Rehabil66:69-74, 1985 7. Mathiowetz V, Rennells C, Donahoe L: Effect of elbow position on grip and key pinch strength. I Hand Surg 1 OA:694-697, 1985 8. Mathiowetz V, Weber K, Volland G, Kashman N: Reliability and validity of grip and pinch strength evaluations. 1 Hand Surg 9A:222-226, 1984 9. Mathiowetz V, Wiemer D, Federman S: Crip and pinch strength: Norms for 6 to 19 year olds. Am I Occup Ther 40:705-711, 1986 10. Robertson A, Dietz I: A description of grip strength in pre-school children. Am I Occup Ther 42:647-652, 1988 11. Schmidt RT, Toews /V: Crip strength as measured by the lamar dynamometer. Arch Phys Med Rehabil 5 1:32 1-327, 1970 12. Solgaard S, Kristiansen 6, lensen IS: Evaluation of instruments for measuring grip strength. Acta Orthop Scand 55:569-572, 1984 Copyright 1992. All rights reserved. JOSPT Volume 16 Number 5 November 1992