Predicting normal grip strength for rheumatoid arthritis patients

Transcription

1 Rheumatology 1999;38: Predicting normal grip strength for rheumatoid arthritis patients A. Fraser1,2, J. Vallow1, A. Preston1 and R. G. Cooper1 1Department of Rheumatology, Pinderfields Hospital, Aberford Road, Wakefield WF1 4DG and 2Rheumatology and Rehabilitation Research Unit, University of Leeds, 36 Clarendon Road, Leeds LS2 9NZ, UK Abstract Objective. An ability to predict accurately normal grip strength in rheumatoid arthritis (RA) patients would facilitate a more accurate assessment of the degree of their functional loss. This, in turn, would allow the setting of more meaningful treatment goals aimed at restoring hand function towards normal. This study carefully measures three modalities of hand grip strength and their correlation with multiple simple anthropometric parameters in normal subjects. We aim to determine which of these parameters are best correlated to grip strength, and whether this correlation is strong enough to allow the accurate prediction of what normal grip strength should be in RA patients. Methods. In 81 normal subjects (67 female), power, pinch and tripod grip strength measurements were made using an MIE digital pinch grip analyser. These strength data were correlated with specific local forearm anthropometric measurements: forearm circumference, forearm length, forearm volume, hand circumference, hand length, hand volume, hand and forearm volume, and various general anthropometric parameters (weight, height and age). These normal subjects had been chosen so as to be age and sex matched with 83 RA patients (67 female) in whom the same strength and anthropometric parameters were assessed and correlated. In patients, the grip strength results were additionally correlated with two markers of disease activity: a modified Ritchie Articular Index local to the hand and forearm (mrai) and a visual analogue scale ( VAS) assessing subjective pain severity. Results. In normal subjects, clear correlations were demonstrated between hand grip strengths and all specific anthropometric variables, the strongest correlation being with forearm and hand volume (r = and for dominant and non-dominant hands, respectively; P < 0.01 for both). The patients were considerably weaker than normal subjects. Markers of disease activity showed a negative correlation with grip strength. In normal subjects, the dominant hand was significantly stronger than the non-dominant hand, and on average by 8%, while the opposite was true in patients, who were 20% weaker on the dominant side. Conclusion. Simple anthropometric measurements, and forearm and hand volume in particular, would be useful at baseline for predicting normal hand grip strength in RA patients, both in the clinical setting and in research trials aimed at improving grip strength and hand function. KEY WORDS: Grip strength, Anthropometry, Rheumatoid arthritis. more meaningfully the clinical relevance of any improvements following such interventions, it would be neces- sary to know the degree of strength lost from normal prior to the intervention, and the degree of recovery towards normal after it. Unfortunately, there remains a paucity of data regarding methods for reliably estimating normal grip strength in RA patients. There have been attempts to develop normative data for grip strength in healthy subjects [11 15], but the reliability of some of the tools employed in these studies has been questioned [1, 2, 16]. Furthermore, detailed anthropometric data, which would have facilitated more accurate between-subject comparisons, were not Loss of hand grip strength and function is a major cause of disability in patients with rheumatoid arthritis ( RA) [1 3]. Such dysfunction results from pain or fear of pain, reflex inhibition [4], disuse atrophy [ 5, 6] and eventually mechanical disruption [7]. Strength testing has been used to monitor the therapeutic response of patients to medical therapies [8], hand surgery [9] and to orthotic interventions [10]. In order to determine Submitted 6 April 1998; revised version accepted 29 January Correspondence to: A. Fraser, Rheumatology and Rehabilitation Research Unit, University of Leeds, 36 Clarendon Road, Leeds LS2 9NZ, UK British Society for Rheumatology

2 522 A. Fraser et al. recorded. The normative tables available to date assume sepsis or fracture below the humerus in the previous that individuals of the same age, gender and general 2 yr. Also excluded were patients who had undergone body habitus will have similar forearm anthropometry surgery below the humerus in the last 2 yr, or any [11, 12, 14, 17 20]. Clearly, this is not the case, and surgery deemed to have had a detrimental effect on grip, when estimating normal grip strength in patients with e.g. wrist arthrodesis. Those with fixed digital deformimpaired grip, one would expect forearm measurements ities, including swan neck and boutonnière, or wrist to be of relevance. Gilbertson and Barber-Lomax [21] subluxation to the extent that passive or active wrist studied 260 normal volunteers, attempting to define extension was not possible, were also excluded. Patients normative data for British adult grip strength. While with ulnar deviation were included. Eighty-one agethe results included tables of normative data for power, and gender-matched control subjects were recruited pinch and tripod grip strengths, and did show an age- from the medical, paramedical and non-medical staff related decline in grip strength, the Jamar dynamometer of Pinderfields Hospital, and from friends and relatives used had previously been shown to produce variable of RA patients. In these normal subjects, any cause of results [22] and anthropometric data other than age and hand pain caused exclusion, as did clinical evidence of gender were not recorded. A highly significant relation- painless osteoarthritis. No normal subject with DIP/PIP ship between mid-forearm cross-sectional area (CSA) or thumb carpometacarpal joint osteoarthritis was and grip in normal subjects and RA patients has been included. Ingestion of any drug known to cause muscle shown by Helliwell and Jackson [3], but this study dysfunction also caused exclusion in the control subjects, employed a technique not readily applicable to the who were screened by an experienced occupational clinical setting. therapist specializing in hand therapy (JV ), and overseen The widespread use of hand and wrist orthoses in RA by RGC. patients is justified primarily by limited data demonstrat- Three types of grip strength were analysed, using an ing short-term improvements in pain and inflammation, MIE digital power and pinch grip analyser featuring a and consequently grip strength [23 26]. The theoretical visual results display ( MIE, Medical Research, Leeds, and empirical evidence for the usefulness of splinting is UK ). The latter allows visual feedback and the instruthus stronger than either clinical or scientific evidence ment has been shown to produce reliable and repeatable has yet confirmed. Indeed, some studies have failed to results, and is acceptable to patients with rheumatoid show any benefit in short-term disease activity [27] and hands [2]. Pinch grip strength was measured between none has demonstrated an ability to alter long-term thumb pulp and the radial side of the second digit. outcome, i.e. the rate of development of deformity [26]. Tripod grip strength was measured between thumb pulp This is partly explained by the previous lack of precise measuring tools, but also by a lack of good normative and the palmar surfaces of the second and third digits. data for hand grip strength. The cost-conscious health Power grip strength was measured between the partially service of today may thus require more definitive evicounter pressure. Dominant and non-dominant hands flexed fingers and the palm while the thumb applied dence to justify the ongoing use of resources currently allocated to orthotic use. were both assessed. In all instances, patients and control The aim of the present study was to measure carefully subjects were asked to make three maximal voluntary various hand grip strengths and general and specific contractions (MVC), with 1 min between each. Verbal anthropometric parameters in normal subjects, and to encouragement was given and the digital read-out of determine whether any of these parameters are suffi- each contraction, in newtons, allowed visual feedback. ciently correlated to allow accurate prediction of normal If these readings were not within <10% of each other, grip strength. These correlations were then employed in further contractions were undertaken, with 1 min a comparison between the normal subjects and a similar between each, until they were repeatable. Once repeat- sized cohort of RA patients. able, only the maximum contraction was then taken to represent the MVC strength. General data recorded were age, height, weight, Methods gender, previous medical history, drug history and hand This study was undertaken with local ethical committee dominance. Specific anthropometric parameters were approval. Although the primary aim of the study was measured using a non-stretchable plastic tape measure, to produce durable normative data, RA patients were the degree of ulnar deviation was measured using a also studied and this dictated that, to allow meaningful hand-held goniometer and, for the measurement of hand patient/control comparisons, the normal subjects and and forearm volume, a specially designed and locally patients had to be age and gender matched. Thus, the manufactured water displacement column was used. RA patients were recruited initially and normal controls The specific anthropometric parameters assessed for procured to age and gender match with the patients. the forearm and hand were as follows. Eighty-three RA patients were recruited (by RGC) from the out-patient population, and all fulfilled the Maximum forearm circumference, taken one-sixth of revised New York criteria for the diagnosis of RA [28]. the distance between the olecranon process and distal Excluded were any patients with proven or clinically wrist crease in supination. Preliminary assessment in apparent peripheral nerve or tendon dysfunction, open 10 subjects had shown that one-sixth distance consist-

3 Predicting normal grip strength for RA patients 523 ently approximated to the point of maximum forearm Pearson s product moment correlation coefficients were circumference. used to determine the significance of correlations 2. Forearm length, measured between the olecranon between variables. process and the distal wrist crease with the wrist extended and in pronation. 3. Hand length, measured between the distal wrist crease Results and the tip of the third digit with the wrist supinated. The normal controls and RA patients were matched for 4. Hand circumference, measured around the hand age, weight and sex (Table 1). between the thumb web and a point on the ulnar Coefficients of variance show that in normal subjects border of the hand, 1 cm distal to the distal wrist ( Table 2a) and RA patients ( Table 2b), test/retest reli- crease, with the wrist in the neutral position and with the hand pronated. TABLE 1. Mean (± 1 S.D.) age, height and weight of study participants 5. Hand volume, measured with the elbow extended and the hand relaxed. The hand was immersed in Normals (n = 81) RA (n = 83) water to the distal wrist crease and held motionless Male Female Male Female until water displacement ceased. (n = 14) (n = 67) (n = 16) (n = 67) 6. Hand and forearm volume, measured with the elbow rested and in extension and the hand relaxed. The Age (yr) 59.9 (12.0) 54.5 (12.69) 54.0 (12.56) 57.0 (11.7) Height (m) 1.72 (0.07) 1.64 (0.07) 1.70 (0.08) 1.59 (0.07) hand and forearm were immersed to the olecranon Weight (kg) (11.59) (10.88) (13.16) (14.16) process until water displacement ceased. 7. Forearm volume was calculated by subtracting (5) from (6). TABLE 2a. Correlation coefficient of variance results in normal sub- 8. Ulnar deviation was measured using a hand-held jects (%) goniometer. The fulcrum was placed on the MCP joint of the third digit with the two arms aligned Non-dominant Dominant with the third metacarpal and the third proximal Mean Highest Lowest Mean Highest Lowest phalanx. F.CIR These specific parameters were all measured by one F.LEN H.LEN individual (JV ). However, to ensure reliability of meas- H.CIR urement in control subjects and RA patients, 10 of each H.VOL group had all parameters blindly remeasured by another H + F.VOL observer 2 weeks later to allow for a test/retest analysis. F.VOL The patients additionally had hand disease activity MAX POW.G MAX TRI.G assessments made using a modified local Ritchie MAX PIN.G Articular Index (mrai), assessing elbow, wrist, MCP and PIP joints with a maximum score of 36, and a Forearm circumference (F.CIR), forearm length (F.LEN ), hand visual analogue scale ( VAS) [29] scoring joint localized length (H.LEN ), hand circumference (H.CIR), hand volume pain in the MCP, PIP and wrist joints. Standardization (H.VOL), hand and forearm volume (H + F.VOL), forearm volume (F.VOL), maximum power grip (MAX POW.G.), maximum tripod of joint position was achieved by employing the grip (MAX TRI.G.), maximum pinch grip (MAX PIN.G.). American Society of Hand Therapists (ASHT) recommendations for joint position during the assessment of TABLE 2b. Correlation coefficient of variance results in RA patients grip strength [30]. During all measurements, room tem- (%) perature was kept constant at 20 C and water displacement bath temperature at 25 C [31]. All assessments in Non-dominant Dominant the RA group were carried out between 1 and 4 p.m. to minimize the known effects of diurnal variation on grip Mean Highest Lowest Mean Highest Lowest [31 33]. RA patients with an mrai of 18 were F.CIR excluded from the study, as pain of this degree was F.LEN H.LEN thought likely to have too great an effect on maximum H.CIR grip readings. H.VOL H + F.VOL Statistical analysis F.VOL MAX POW.G A test/retest coefficient of variance analysis was carried MAX TRI.G out on the results of all the specific parameters from the MAX PIN.G normal subjects and 10 RA patients, who underwent blinded reassessments. This was to clarify the accuracy Forearm circumference (F.CIR), forearm length (F.LEN ), hand length (H.LEN ), hand circumference (H.CIR), hand volume of each assessor, ensuring repeatability and validity. (H.VOL), hand and forearm volume (H + F.VOL), forearm volume The Mann Whitney U-test was used to compare (F.VOL), maximum power grip (MAX POW.G.), maximum tripod differences between normal subjects and patients, and grip (MAX TRI.G.), maximum pinch grip (MAX PIN.G.).

4 524 A. Fraser et al. ability was acceptable for all of the specific forearm strength and summed forearm circumference (r = 0.60, anthropometric parameters. The test/retest reliability of P < ). Our study confirms this finding [domin- the three grip strengths was also very reliable in normal ant (r = 0.551, P < 0.01), non-dominant (r = 0.687, subjects, but considerably less so in RA patients, P < 0.01)], and further we have demonstrated that these although for power grip the mean coefficient of variance correlations persist for pinch and tripod grip. Several was still <10% for patients. other simple anthropometric measurements are also The results of mean values for anthropometric param- shown to be strongly correlated with hand strength, but eters and grip strengths in normal subjects and RA the strongest correlation we found for all three modal- patients, and their standard deviations, are shown in ities of hand grip strength was for hand and forearm Table 3. volume. This measurement offers a simple and practical All modalities of hand grip strength were significantly tool for the accurate prediction of what normal grip reduced in RA patients, but hand and forearm anthropometrics strength should be. were not significantly reduced compared to A severe weakness in RA patients in our study is normals. confirmed for all three modalities of hand strength The results for normal subjects demonstrated clear, measured. Hand and forearm volume measurements and sometimes highly significant, correlations between were not significantly different between normal controls the general and specific anthropometric variables tested and RA patients. This appears at variance with the and pinch, tripod and power grip strength ( Tables 4a c). finding of a reduction in forearm anatomical muscle For instance, in Fig. 1a, the results of power grip vs CSA in RA patients ( 25.9 cm2) vs normal controls forearm volume for the dominant side in normal subjects (29.7 cm2) previously reported [3]. However, the afore- are demonstrated. Figure 1b illustrates the 5 and 95% mentioned study calculated forearm anatomical CSA, confidence intervals for this correlation. detecting a true reduction in forearm lean muscle, not The results for RA patients also demonstrated correla- seen when measuring volume alone. However, in view tions between the variables, but these correlations were of the degree of significance of our results using compar- in general not as clear as those found for normal subjects atively crude parameters, the use of more complicated ( Tables 4a c). As expected, patients were considerably methods appears perhaps unnecessary. weaker than their age- and sex-matched normal controls. Predictably, indices of disease activity (mrai and This is perhaps best seen in Fig. 1c, where the majority VAS) were negatively correlated with strength in of RA patient results for power grip are below the 5% patients, as previously noted [3, 7, 34]. Surprisingly, in confidence interval for control subjects. our study, ulnar deviation showed only a poor correl- In normal subjects, the dominant hand was on average ation with hand grip strength. Helliwell and Jackson [3] 8% stronger for power grip strength than the nondominant found a significant correlation between maximum isothe hand. Conversely, in the RA patient group, metric grip strength and a deformity index which dominant hand was on average 20% weaker for included ulnar drift (coefficient = 4.66, S.D. = 1.52, power grip strength than the non-dominant hand, P = 0.003), while Eberhardt et al. [7] also found a and the breakdown by age for both groups is shown in significant weakness in their hand deformity group, Fig. 2. which included ulnar drift, over RA controls (P < 0.05). In RA patients, there were clear and significant negative In this latter study, however, the deformity group had correlations between the mrai (r = 0.882, more active disease and more severe radiographic P < 0.01) and the VAS (r = 0.567, P < 0.01) and changes than their RA control counterparts, and the dominant power grip strength. A weak negative correl- numbers studied were small. It has long been thought ation was noted between grip strength and the degree that ulnar deviation should have a marked detrimental of ulnar drift, but only for the non-dominant side effect on hand strength due to mechanical disadvantages, (r= 0.225, P < 0.05). and previous data support this. It is not clear why such a poor correlation was found between ulnar drift and grip strength in our study, but it may be that differences Discussion would have been more apparent if a composite deformity Previous studies of grip strength in normal subjects have index had been used, as in other studies [3, 7]. primarily been concerned with age-related grip changes In RA patients, the reversal of the normal strength and the correlation of grip with a limited number of superiority of the dominant over the non-dominant other general variables, i.e. height, weight and gender. hand is in keeping with previous work [35, 36], and In this study, we analysed in detail for the first time appears to confirm the belief that increased physical three functions of hand grip strength and their correl- activity of the dominant side predisposes to joint ation with various general and specific local anthropometric inflammation and damage, resulting in more impaired parameters, thus creating reliable normative data function. for future comparative use in arthritic patients. Kallman et al. [17] concluded that forearm circumference provides the most practical index of grip strength related Conclusion to muscle mass. In 382 normal subjects, his group Many general and specific local anthropometric parameters, found a strong correlation between summed grip and their correlation with hand grip strengths in

5 Predicting normal grip strength for RA patients 525 TABLE 3. Mean (± 1 S.D.) results for anthropometric data and grip parameters in normal and RA subjects Normal RA Male Female Male Female Non-dominant Dominant Non-dominant Dominant Non-dominant Dominant Non-dominant Dominant F.CIR (cm) (2.46) (2.65) (1.87) (2.03) (2.15) (2.12) (2.79) (2.65) F.LEN (cm) (1.31) (1.22) (1.69) (1.78) (2.22) (2.14) (1.89) (1.78) H.LEN (cm) (0.63) (0.04) (1.03) (0.90) (1.42) (1.37) (1.22) (1.37) H.CIR (cm) (1.05) (1.29) (1.12) (1.09) (1.79) 23.18(1.98) (1.37) (1.37) H.VOL(cm3) (52.69) (48.90) (53.59) (52.97) (110.42) (99.80) (68.87) (58.74) H. + F.VOL (cm3) (204.58) (241.19) (216.53) (199.25) (275.60) (306.69) (273.19) (267.49) F.VOL (cm3) (184.14) (218.28) (207.25) (163.92) (212.32) (237.46) (231.46) (233.03) MAX POW.G. (N) (72.08) (77.26) (56.46) (60.20) (85.49) (85.20) (50.23) (54.49) MAX TRI.G. (N) (24.09) (22.53) (17.21) (20.00) (23.35) (25.15) (17.20) (15.45) MAX PIN.G. (N) (12.33) (13.15) (10.43) (9.85) (16.79) (16.03) (12.30) (12.79) Forearm circumference (F.CIR), forearm length (F.LEN ), hand length (H.LEN), hand circumference (H.CIR), hand volume (H.VOL), hand and forearm volume (H + F.VOL), forearm volume (F.VOL), maximum power grip (MAX POW.G.), maximum tripod grip (MAX TRI.G.), maximum pinch grip (MAX PIN.G.).

6 526 A. Fraser et al. TABLE 4a. Correlations (r values) for maximum power grip (POW.G.) Normal (male + female) RA (male + female) Non-dominant Dominant Non-dominant Dominant F.VOL/POW.G ** 0.707** 0.303** 0.337** F.CIR/POW.G ** 0.687** 0.359** 0.327** F.LEN/POW.G ** 0.336** 0.403** 0.452** H + F VOL/POW.G ** 0.729** 0.309** 0.330** H.VOL/POW.G ** 0.667** H.CIR/POW.G ** 0.693** 0.298** H.LEN/POW.G ** 0.559** 0.425** 0.349** AGE/POW.G * 0.347** 0.304** WEIGHT/POW.G ** 0.565** 0.236* 0.236* HEIGHT/POW.G ** 0.478** 0.283* 0.291** Forearm volume (F.VOL), forearm circumference (F.CIR), forearm length (F.LEN), hand and forearm volume (H + F VOL), hand volume (H.VOL), hand circumference (H.CIR), hand length (H.LEN), age, weight and height. *P < 0.05; **P < TABLE 4b. Correlations (r values) for maximum tripod grip (TRI.G.) Normal (male + female) RA (male + female) Non-dominant Dominant Non-dominant Dominant F.VOL/TRI.G ** 0.576** 0.265* F.CIR/TRI.G ** 0.565** 0.320** F.LEN/TRI.G ** 0.224* 0.368** 0.301** H + F VOL/TRI.G ** 0.583** 0.267* H.VOL/TRI.G ** 0.508** H.CIR/TRI.G 0.598** 0.517** 0.239* H.LEN/TRI.G ** 0.420** 0.335** AGE/TRI.G * 0.359** 0.294* WEIGHT/TRI.G ** 0.469** HEIGHT/TRI.G ** 0.273* 0.232* 0.226* Forearm volume (F.VOL), forearm circumference (F.CIR), forearm length (F.LEN), hand and forearm volume (H + F VOL), hand volume (H.VOL), hand circumference (H.CIR), hand length (H.LEN), age, weight and height. *P < 0.05; **P < TABLE 4c. Correlations (r values) for maximum pinch grip (PIN.G.) Normal (male + female) RA (male + female) Non-dominant Dominant Non-dominant Dominant F.VOL/PIN.G ** 0.630** * F.CIR/PIN.G ** 0.562** 0.278* 0.276* F.LEN/PIN.G ** 0.348** 0.345** 0.359** H + F VOL/PIN.G ** 0.633** * H.VOL/PIN.G ** 0.529** H.CIR/PIN.G ** 0.598** 0.244* H.LEN/PIN.G ** 0.486** 0.237* 0.237* AGE/PIN.G * WEIGHT/PIN.G ** 0.474** * HEIGHT/PIN.G ** 0.356** * Forearm volume (F.VOL), forearm circumference (F.CIR), forearm length (F.LEN), hand and forearm volume (H + F VOL), hand volume (H.VOL), hand circumference (H.CIR), hand length (H.LEN), age, weight and height. *P < 0.05; **P < normal and RA subjects, are studied here for the first time. These normative data, in conjunction with the use of water displacement for volume measurement of the hand and forearm, allow accurate estimation of normal hand grip strengths in future studies where baseline assessment, the setting of meaningful treatment goals and the measurement of outcome, is required in RA and other treatment groups. A worthwhile assessment of the usefulness of orthotic and other interventions to improve hand function should now be possible.

7 Predicting normal grip strength for RA patients 527 Acknowledgements We would like to thank all patients and volunteer controls for their kind participation, Sister Christine White and the Mayor s fund and the WRVS of Wakefield who provided financial support. References 1. Lee P, Baxter A, Carson-Dick W, Webb J. An assessment of grip strength measurement in rheumatoid arthritis. Scand J Rheumatol 1974;3: Helliwell PS, Howe A, Wright V. Functional assessment of the hand: utility, reproducibility and acceptability of a new system for measuring strength. Ann Rheum Dis 1987;46: Helliwell PS, Jackson S. Relationship between weakness and muscle wasting in rheumatoid arthritis. Ann Rheum Dis 1994;53: Joint injury and muscle weakness. [Leading Article] Lancet 1984;2: Heymsfield SB, McManus C, Smith J, Stevens V, Nixon DW. Anthropometric measurement of muscle mass: revised equations for calculating bone-free arm muscle area. Am J Clin Nutr 1982;36: Bruce SA, Newton D, Waledge RC. Effect of subnutrition on normalised muscle force and relaxation rate in human subjects using voluntary contractions. Clin Sci 1989; 76: Eberhardt K, Malcus Johnson P, Rydgren L. The occurrence and significance of hand deformities in early rheumatoid arthritis. Br J Rheumatol 1991;30: Hart FD, Huskisson EC. Measurement in rheumatoid arthritis. Lancet 1972;i: Opitz JL, Linscheid RL. Hand function after metacarpophalangeal joint replacement in rheumatoid arthritis. Arch Phys Med Rehabil 1978;59: FIG. 1. (a) The correlation between dominant power grip (N) 10. Falconer J. Hand splinting in rheumatoid arthritis. and dominant forearm volume (cm3), where grip = Arthritis Care Res 1991;4: volume 24.3 for normal males and females. (b) The 11. Montoye HJ, Lamphier DE. Grip and arm strength in 95% confidence intervals for dominant power grip (N) vs males and females, age 10 to 69. Res Q 1979;48: dominant forearm volume (cm3) where, for normal males and 12. Chatterjee S, Chowdhuri BJ. Comparison of grip strength females; 95% CI is grip = 0.3 volume , 5% CI is and isometric endurance between the right and left hands grip = 0.2 volume (c) Dominant power grip vs dom- of men and their relationship with age and other physical inant forearm volume for RA patients, superimposed on 5 parameters. J Hum Ergol 1991;20: and 95% CI for normal subjects. 13. Fraser C, Benten J. A study of adult hand strength. Br J Occup Ther 1983;October: Bassey EJ, Harries UJ. Normal values for handgrip strength in 920 men and women over 65 years, and longitudinal changes over 4 years in 620 survivors. Clin Sci 1993;84: Mathiowetz V, Kashman N, Volland G, Weber K, Dowe M, Rogers S. Grip and pinch strength: Normative data for adults. Arch Phys Med Rehabil 1985;66: Solgaard S, Kristiensen Jensen JS. Evaluation of measuring grip strength. Acta Orthop Scand 1984;55: Kallman DA, Plato CC, Tobin JD. The role of muscle loss in the age related decline of grip strength: Crosssectional and longitudinal perspectives. J Gerontol 1990;45: Pearson MB, Bassey EJ, Bendall MJ. The effects of age on muscle strength and anthropometric indices within FIG. 2. The percentage by which the dominant hand is stronger a group of elderly men and women. Age Ageing than the non-dominant hand for normal subjects (&) and RA patients (%), and divided into age cohorts. 1985;14: Burke WE, Tuttle WW, Thompson CW, Janney CD,

8 528 A. Fraser et al. Weber RJ. The relation of grip strength and grip strength Rheumatism Association 1987 revised criteria for the endurance to age. J Appl Physiol 1988;5: classification of rheumatoid arthritis. Arthritis Rheum 20. MacLennan WJ, Hall MRP, Timothy JJ, Robinson M. Is 1988;31: weakness in old age due to wasting? Age Ageing 29. Downie WW, Leatham PA, Rhind VM, Pickup ME, 1980;9: Wright V. The visual analogue scale in the assessment of 21. Gilbertson L, Barber-Lomax S. Power and pinch grip grip strength. Ann Rheum Dis 1978;37: strength recorded using the hand held Jamar dynanometer 30. Fess EE, Moran CA. Clinical assessment recommendations, and B and L hydraulic pinch gauge: British normative 2nd edn. Chicago: American Society of Hand data for adults. Br J Occup Ther 1994;57: Therapists, Flood JM, Mathiowetz V. Grip strength measurement: a 31. Pearson R, MacKinnon MJ, Meek AP, Mijers DB, Palmer comparison of the three Jamar dynamometers. Occup DG. Diurnal and sequential grip function in normal Ther J Res 1987;7: subjects and effects of temperature change and exercise of 23. Partridge REH, Duthie JJR. Immobilisation of the joints the forearm on grip function in patients with rheumatoid in rheumatoid arthritis. Ann Rheum Dis 1963;12:91 9. arthritis and in normal controls. Scand J Rheumatol 24. Gault SJ, Spyker JM. Beneficial effect of immobilisation 1982;11: of joints in rheumatoid arthritis. A splint study using 32. Wright V. Some observations on diurnal variation of grip. sequential analysis. Arthritis Rheum 1969;12: Clin Sci 1959;18: Feinberg J. Effect of the arthritis health professional on 33. Harkness JAL, Richter MB, Panayi, Van de Pette K, compliance with use of resting hand splints by patients Unger A, Pownall R. Circadian variation in disease activity with rheumatoid arthritis. Arthritis Care Res 1992;5: in rheumatoid arthritis. Br Med J 1982;284: Prevoo ML, van Riel PL, van t Hof MA et al. Validity 26. Malcus Johnson P, Sandkrist G, Eberhardt K, Liang B, and reliability of joint indices. A longitudinal study in Herrlin K. The usefulness of nocturnal resting splints in patients with recent onset rheumatoid arthritis. Br the treatment of ulnar deviation of the rheumatoid hand. J Rheumatol 1993;32: Clin Rheumatol 1992;11: Swanson AB, Matev IB, de Groot G. The strength of the 27. Anderson K, Maas F. Immediate effect of working splints hand. Bull Prosth Res 1970;10: on grip strength of arthritic patients. Aust Occup Ther 36. Petersen P, Petrick M, Connor H, Conklin D. Grip J 1987;34: strength and hand dominance: Challenging the 10% rule. 28. Arnett FC, Edworthy SM, Block DA et al. The American Am J Occup Ther 1989;43:444 7.