Reliability of hand-held dynamometric strength testing in people with diabetes/chronic conditions.
Abstract: We evaluated the intra- and intertester reliability of a newly-designed handheld dynamometer (HHD) testing muscle strength on subjects with diabetes/ chronic diseases. Thirteen adults (aged 67.15 [+ or -] 6.11 years) with diabetes/chronic conditions participating in a community exercise programme took part in the study. One registered physiotherapist and two physiotherapy students measured the maximal isometric strength of the knee extensors and elbow flexors. The registered physiotherapist retested the measurements during a second session. Intraclass correlation coefficient (ICC) and standard error of measurement (SEM) were calculated to investigate the agreement between testers and between test-retest sessions. The 95% smallest real difference ([SRD.sub.95]) was quantified for the test-retest data. [ICC.sub.31] (95% CI) for test-retest agreement was 0.98 (0.95-0.99) for the knee extensors and 0.94 (0.77-0.98) for the elbow flexors. [ICC.sub.2.1] for inter-tester agreement was 0.95 for knee extensors, and 0.83 (0.56-0.94) for elbow flexors. The SEM for re-tests for knee extensors was 1.45kg and for elbow flexors 2.17 kg. [SRD.sub.95] was 4.03kg for knee extensors and 2.85kg and for elbow flexors. The HHD demonstrates "high" to "very high" intra- and intertester reliability in people with diabetes/chronic conditions. This study provides guidelines for minimum reliable differences for muscle strength of clients with chronic conditions.

Keywords: dynamometry, reliability, diabetes mellitus
Article Type: Report
Subject: Dynamometer (Usage)
Chronic diseases (Diagnosis)
Diabetes (Diagnosis)
Medical screening (Usage)
Authors: Sole, Gisela
Wright, Lexie
Wassinger, Craig
Higgs, Chris
Hansson, Maria
Johansson, Sandra
Todd, Nancy
Pub Date: 07/01/2010
Publication: Name: New Zealand Journal of Physiotherapy Publisher: New Zealand Society of Physiotherapists Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2010 New Zealand Society of Physiotherapists ISSN: 0303-7193
Issue: Date: July, 2010 Source Volume: 38 Source Issue: 2
Geographic: Geographic Scope: New Zealand Geographic Code: 8NEWZ New Zealand
Accession Number: 263992601

Regular exercise is part of the management of many patients with type 2 diabetes mellitus (Derouich and Boutayeb 2002, Gill and Cooper 2008, Gordon et al Fraser 2009, Roumen at al 2009) often presented in the community setting as exercise classes. The pre-participation examination may include assessment of muscle strength with hand-held dynamometry (HHD) (Kolber et al 2007). These devices are portable, relatively inexpensive, and easy to use (Bohannon 1997, Knols et al 2009, Lu et al 2007, O'Shea et al 2007). The reliability of HHD has been evaluated in various populations, including healthy, young and elderly adults (Bohannon 1997) adults with neuropathic weakness (Kilmer et al 1997), and adults with chronic obstructive pulmonary disease (O'Shea et al 2007). However, to our knowledge, reliability of HHD has not yet been studied in a community setting in people with diabetes. Different clinicians, including health professional students, may be involved in the assessment of clients with chronic conditions wishing to attend community-based exercise programmes, thus, the inter-tester reliability for HHDs needs to be established. As clients are re-assessed, the intra-tester reliability needs to be determined to establish the minimal value needed to be beyond measurement error. Therefore, the aim of this study was to determine intra- and inter-tester reliability of HHD when measuring isometric muscle strength of people with diabetes/chronic conditions.


Participants were recruited from the Diabetes Community Exercise Programme, Dunedin, New Zealand, and inclusion criteria were passing the programme's pre-participation examination, conducted by a registered physiotherapist (CH) and a registered nurse (NT), and completion of a 12-week basic exercise programme. Participants gave written informed consent after information about the study was provided, in accordance with ethics approval granted by the Lower South Regional Ethics Committee, New Zealand. Participants were familiarised with the procedures at least one week prior to data collection. Exclusion criterion was reported pain during active movement of knee and/ or elbow or during the familiarisation session.

Maximum isometric knee extensors and elbow flexors strength was assessed during the one-hour exercise class as a "station" in the circuit training offered as part of the programme. One male registered physiotherapist (CH) and two female final year physiotherapy students (MH, SJ) performed the measurements (inter-tester reliability), according to participants' availability. The order of testers for the inter-tester measurements was randomised. To determine intra-tester reliability, the registered physiotherapist performed a retest on each participant for the knee extensors and elbow flexors a few days after the first assessment.

A novel HHD (Industrial Research Limited, Christchurch, New Zealand) was used that measures muscle force and joint range of motion simultaneously (Figure 1). Excellent instrument reliability has been documented for this dynamometer when compared to an isokinetic dynamometer (Janssen and Le-Ngoc 2009). For the knee extensor assessment, participants were seated on a high chair with back support, arms crossed over chest and the pelvis stabilised with a seat belt. The dynamometer was placed on the anterior tibia, 10 cm proximal to the medial malleolus, with a soft towel between the force pad and the leg. Using the unique angle warning function on the device, the knee was placed in 60[degrees] flexion prior to each contraction. For the measurement of elbow flexors, participants were seated on a regular chair, with the upper arm along the body, the elbow joint in 90[degrees] and forearm supinated.


Only the non-dominant side was tested. All testers performed two measurements each on the participant. The participants were encouraged to "press" or "push" as hard as they could for 3-5 seconds. A 10 second rest period was given between contractions and a 1-minute rest between each new tester. The two trials performed by each tester were averaged and used for statistical analysis.

Data analysis

Sample size was calculated using the method described by Walter et al (1998). For an expected Intraclass Correlation Coefficient (ICC) between 0.6-0.9, a minimum of 12 participants were required for intra-tester reliability (two observations), and a minimum of 8 participants for the inter-tester reliability (three observations).

To determine whether there were learning or fatigue effects of the participant for the inter-tester reliability, Analysis of Variance (ANOVA) tests were conducted for the three trials, irrespective of the tester. ANOVAs were also performed to determine whether significant differences existed between the findings of the three testers. Paired t-tests were used to determine whether significant differences were evident for the test-retest sessions. The level of significance was set at P < 0.05.

Intraclass Correlation Coefficients ([ICC.sub.2.1]) for inter-tester reliability and [ICC.sub.3,1] for intra-tester reliability were calculated to determine relative reliability. The ICCs were classified as follows: 0.50 to 0.69 as "moderate"; 0.70 to 0.89 as "high" and values above 0.90 as "very high" (Munro and Visintainer 2005). Absolute reliability was determined with the standard error of measurement (SEM) and smallest real difference (SRD). SEM was calculated with the formula SEM = SD [square root of 1 - ICC] where the SD represents the mean SD of all measurements. The SEM was also expressed as a percentage of the mean (SEM%). In order to determine the minimal value needed to be 95% confident a real change had occurred on an individual patient level, [SRD.sub.95] computed with the formula SRD = 1.96 x SEMx [square root of 2]) (Flansbjer et al 2005), was calculated for the test-retest sessions. Statistical analysis was performed using SPSS 14.0 (Norusis/SPSS Inc Chicago, Illinois, USA).


Thirteen participants (six men, seven women, mean age 67.15 [+ or -] 6.11 years) met the inclusion criteria and volunteered for the study (Table 1). Of the thirteen available for testing of elbow flexors, eleven were available for inter-tester assessment. Eleven participants had a medical history of diabetes and two had other chronic diseases. Six of thirteen participants reported a history of knee, back, or arm operation. One participant experienced knee pain and a further participant shoulder pain following the first session, and were excluded for the assessment of these segments.

No statistically significant differences were found when comparing the means of the three consecutive trials, irrespective of the tester, for elbow flexor (P = 0.441) and knee extensor strength (P = 0.887) during the first sessions. Further, no statistically significant differences were found when comparing the findings of the three testers (Table 2). Similarly, no significant differences were found for the means of the test-retest session (Table 3).

Means (and 95% CI) of the strength measurements and their analyses of inter- and intra-tester reliability are depicted in Tables 2 and 3. Corresponding ICC values for inter-tester agreement were 0.97 and 0.87 for knee extensors and elbow flexors, respectively (Table 2). ICCs for intra-tester agreement were 0.97 and 0.95 for knee extensors and elbow flexors respectively (Table 3).


Overall our study showed "very high" reliability for strength measures using a HHD in a cohort of patients with chronic diseases. The lowest ICC value (0.87), found for inter-tester assessment of elbow flexors, is categorized as "high". The [SRD.sub.95] calculated for the intra-tester reliability indicate that differences of more than 5kg are needed when retesting a patient to be 95% certain that a change beyond measurement error has occurred.

Our study found similar relative reliability for intra- and inter-assessments, in contrast to earlier reports in which intra-tester assessment has generally been considered to be more reliable than inter-tester assessment (Byl et al 1988, Verschuren et al 2008). Further, our results demonstrated higher relative reliability in the stronger muscle group (knee extensors) compared to the weaker muscle group (elbow flexors).

Maximal muscle strength tests are also influenced by motivation, concentration, and co-operation of the patient. In this study the setting of a community exercise class within a gymnasium was chosen to improve construct validity for these rehabilitation programmes. The order the participants were tested was random for all measurements. The individual tests could thus be performed at any time during the class, which could result in different fatigue levels for the retest sessions. However, the analyses indicated that there was no order effect for the trials, thus fatigue or learning effects were unlikely to have affected the group findings.

Although we found no statistically significant differences for the strength measurements performed by the three testers, observations of individual participant's measurements indicate slightly higher strength values when assessed by the registered physiotherapist (Tester 2) compared to the physiotherapy students (Table 2). It is possible that tester 3 (male) was able to generate greater force than the student physiotherapists (female). In order to keep the HHD in a fixed position, the tester must be able to apply a force meeting the participant's strength. Earlier studies have shown that stronger testers perform more reliable measurements (Katoh and Yamasaki 2009, Wikholm and Bohannon 1991). However, our study included a weaker patient population (Kilmer et al 1997) and all testers were, in a pre-test, able to get higher values than the highest value documented among the participants. Further, it was observed that the testers may not have used a consistent technique during the assessments. Two methods can be used when measuring isometric muscle strength, the "break-test" and the "make-test" (Burns and Spanier 2005). Both techniques require the tester to apply a resistance force that equals the participants' strength in order to keep the HHD in a fixed position (Katoh and Yamasaki 2009, Lu et al 2007). The "make" technique requires the participant to produce a maximal voluntary contraction (Burns and Spanier 2005, Morris et al 2008). With the "break" technique the tester applies an adequate force which initially "breaks" the participant's resistance and thereby produces an eccentric contraction. Both these methods show excellent reliability (ICC >0.90) (Burns and Spanier 2005).

Dealing with clients with diabetes and muscle strength testing requires special attention due to possible sensitivity of the extremities to touch therefore a soft towel was used between force pad and extremity. From the fact that two participants experienced pain after the first session, we conclude that caution must be taken with some patients with previous injuries using this device. In these cases we would recommend using the 'make'-technique to minimize the risk of pain or discomfort following the test.

This study was performed during a community exercise class, and participants were not always available due to other factors such as transport opportunities, ill health, etc. Participants formed a heterogeneous group, and thus the current study does not account for possible differences in reliability for male and female participants, or for previous experience with exercise equipment and muscle strength assessments. Despite the "high" ICC for elbow flexors, the mean differences for the three testers' findings, can be described as marginally significant (P = 0.07). The SEM% for this measurement was also slightly higher than the anecdotal clinically acceptable level of 10%, which implies that caution must be taken for individual assessments. Moreover, lack of standardised techniques and external stabilisation may have affected the result.

Results from our study indicate that a novel HHD is a reliable instrument for strength testing groups of people with diabetes/chronic disease. A minimum difference of 5kg is needed when retest individuals by one clinician to be 95% confident that a change beyond measurement error has occurred. To maintain high reliability in the clinical setting it is important to consider the factors that may contribute to source of error. Our results indicate that students and registered clinicians can perform reliable strength assessments using this device after familiarisation sessions when group assessments are sought for patients with diabetes or other chronic conditions.


Dr Gisela Sole, Centre for Physiotherapy Research, University of Otago, PO Box 56, Dunedin, New Zealand, Fax: -64-3-4798414, E-mail:

Conflicts of interest: The authors declare that they have no conflict of interest. The company providing the dynamometer, Industrial Research Limited, Christchurch was not involved in the data collection, analysis or preparation of the manuscript.

Funding: No external funding was received for this study.

Ethical Approval: This study was approved by the Lower South Regional Ethics Committee, Dunedin, New Zealand (LRS/09/07/025).


Bohannon RW (1997): Reference values for extremity muscle strength obtained by hand-held dynamometry from adults aged 20 to 79 years. Archives of Physical Medicine and Rehabilitation 78: 26-32.

Burns SP and Spanier DE (2005): Break-technique handheld dynamometry: Relation between angular velocity and strength measurements. Archives of Physical Medicine and Rehabilitation 86: 1420-1426.

Byl NN, Richards S, Asturias J (1988): Intrarater and interrater reliability of strength measurements of the biceps and deltoid using a hand held dynamometer. Journal of Orthopaedic and Sports Physical Therapy 9: 399-405.

Derouich M and Boutayeb A (2002): The effect of physical exercise on the dynamics of glucose and insulin. Journal of Biomechanics 35: 911-917.

Flansbjer, U. B., Holmback, A. M., Downham, D., and Lexell, J. (2005): What change in isokinetic knee muscle strength can be detected in men and women with hemiparesis after stroke? Clinical Rehabilitation, 19(5): 514-522.

Gill JMR and Cooper AR (2008): Physical activity and prevention of type 2 diabetes mellitus. Sports Medicine 38: 807-824. Gordon BA, Benson AC, Bird SR, Fraser SF (2009): Resistance training improves metabolic health in type 2 diabetes: A systematic review. Diabetes Research and Clinical Practice 83: 157-175.

Janssen JC and Le-Ngoc L (2009): Intratester reliability and validity of concentric measurements using a new handheld dynamometer. Archives of Physical Medicine and Rehabilitation 90: 1541-1547.

Katoh M and Yamasaki H (2009): Comparison of reliability of isometric leg muscle strength measurements made using a hand-held dynamometer with and without a restraining belt. Journal of Physical Therapy Science 21: 37-42.

Kilmer DD, McCrory MA, Wright NC, Rosko RA, Kim HR, Aitkens SG (1997): Hand-held dynamometry reliability in persons with neuropathic weakness. Archives of Physical Medicine and Rehabilitation 78: 1364-1368.

Knols RH, Aufdemkampe G, De Bruin ED, Uebelhart D, Aaronson NK (2009): Hand-held dynamometry in patients with haematological malignancies: Measurement error in the clinical assessment of knee extension strength. BMC Musculoskeletal Disorders 10.

Kolber MJ, Beekhuizen K, Cheng MSS, Fiebert IM (2007): The reliability of hand-held dynamometry in measuring isometric strength of the shoulder internal and external rotator musculature using a stabilization device. Physiotherapy Theory and Practice 23: 119-124.

Lu TW, Hsu HC, Chang LY, Chen HL (2007): Enhancing the examiner's resisting force improves the reliability of manual muscle strength measurements: Comparison of a new device with hand-held dynamometry. Scandinavian Journal of Rehabilitation Medicine 39: 679-684.

Morris S, Dodd K, Morris M. (2008): Reliability of dynamometry to quantify isometric strength following traumatic brain injury. Brain Injury 22: 1030-1037.

Munro B and Visintainer M (2005): Statistical methods for health care research (5th ed.): Philadelphia: Lippincott.

O'Shea SD, Taylor NF, Paratz JD (2007): Measuring muscle strength for people with chronic obstructive pulmonary disease: retest reliability of hand-held dynamometry. Archives of Physical Medicine and Rehabilitation 88: 32-36.

Roumen C, Blaak EE, Corpeleijn E. (2009): Lifestyle intervention for prevention of diabetes: determinants of success for future implementation. Nutrition Reviews, 67: 132-146.

Verschuren O, Ketelaar M, Takken T, van Brussel M, Helders P, Gorter JW (2008): Reliability of hand-held dynamometry and functional strength tests for the lower extremity in children with Cerebral Palsy. Disability and Rehabilitation 30: 1358-1366.

Walter SD, Eliasziw M, Donner A (1998): Sample size and optimal designs for reliability studies. Statistics in Medicine, 17: 101-110.

Wikholm JB, Bohannon RW (1991): Hand-held dynamometer measurements: Tester strength makes a difference. Journal of Orthopaedic and Sports Physical Therapy 13: 191-198.

Gisela Sole, BSc(Physio), MSc(Med)Exercise Science, PhD

Lexie Wright, BSc(Kinesiology), DPT

Craig Wassinger, PT, PhD

Chris Higgs, BSc(Hons)

Centre for Physiotherapy Research, University of Otago, Dunedin

Maria Hansson

Sandra Johansson

Physiotherapy, University of Umed, Sweden

Nancy Todd, RN

Mornington Primary Health Organisation, Dunedin
Table 1: Characteristics of participants *

                          Men              Women              All
                         (n=8)             (n=7)            (n=13)

Age (years)          68.17 (4.49)      66.29 (7.48)      67.15 (6.11)

Weight (kg)          93.97 (23.91)     84.35 (16.14)    89.16 (20.09) *

Height (m)             1.700.13)        1.62 (0.09)       1.65 (0.11)

BMI(kg/[m.sup.2])    32.75 (7.50)      28.34 (13.4)     30.37 (10.93) *
* Data are mean (SD)

([dagger]) One participant was unwilling to report

Table 2: Muscle strength of knee extensors and elbow flexors
assessed by three testers and statistical analysis of Inter-tester

                                Tester 1            Tester 2
                                Mean (SD)           Mean (SD)

Knee extensors (n = 11)     25.32 (11.13) kg    23.00 (10.41) kg

Elbow flexors (n = 13)       15.28 (6.58) kg     13.24 (4.49) kg

                               Tester 3       P *      [ICC.sub.2,1]
                              Mean (SD)                  (95% CI)

Knee extensors (n = 11)     27.13 (10.11)    0.56    0.95 (0.86-0.99)

Elbow flexors (n = 13)       19.16 (0.64)    0.07    0.83 (0.56-0.94)

                                   SEM                SEM%
                                  (kg)             ([dagger])

Knee extensors (n = 11)          2.32 kg              9.15%

Elbow flexors (n = 13)           3.11 kg             20.34%

* Analysis of Variance (ANOVA); SD = standard deviation; ICC =
intraclass correlation coefficient; SEM = standard error of

([dagger]) SEM% = SEM as percentage of grand mean

Table 3: Muscle strength assessed by one tester and statistical
analysis of intra-tester reliability of knee extensors and elbow

                  Test 1            Test 2            P *
                  Mean (SD)         Mean (SD)

Knee extensors    27.13 (10.11)     26.00 (10.28)     0.161
(n = 12)

Elbow flexors     19.16(0.64)       21.45 (8.86)      0.177
(n = 11)

                  [ICC.sub.3,1]       SEM
                  (95% CI)            (kg)

Knee extensors    0.98 (0.95-0.99)    1.45 kg
(n = 12)

Elbow flexors     0.94 (0.77-0.98)    2.17 kg
(n = 11)

                  SEM%                [SRD.sub.95]
                  ([dagger])          (kg)

Knee extensors    5.59                4.03 kg
(n = 12)

Elbow flexors     10.11               6.01 kg
(n = 11)

* Paired t-tests; SD = standard deviation; ICC = intraclass
correlation coefficient; SEM = standard error of measurement;
([dagger]) SEM% = SEM as percentage of grand mean; [SRD.sub.95] =
smallest reliable difference at 95% confidence level
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.