The impact of variation in the device used to measure grip strength on the identification of low muscle strength: Findings from a randomised cross-over study.

Grip strength is commonly used to identify people with low muscle strength. It is unclear what impact the type of dynamometer used to measure grip strength has on the identification of low muscle strength so we aimed to assess this. Study participants were 118 men and women aged 45-74y from a randomised, repeated measurements cross-over study. Maximum grip strength was assessed using four hand-held dynamometers (Jamar Hydraulic; Jamar Plus+ Digital; Nottingham Electronic; Smedley) in a randomly allocated order. EWGSOP2 cut-points were applied to estimate prevalence of low muscle strength for each device. Agreement between devices was compared. Prevalence of low muscle strength varied by dynamometer ranging between 3% and 22% for men and, 3% and 15% for women. Of the 13 men identified as having low muscle strength by at least one of the four dynamometers, only 8% were identified by all four and 54% by just one. Of the 15 women classified as having low muscle strength by at least one of the four dynamometers, only 7% were identified by all four and 67% by only one. Variation in the measures of grip strength acquired by different hand-held dynamometers has potentially important implications when identifying low muscle strength. Copyright:

There is increasing recognition of the important role of skeletal muscle for health and disease. This is exemplified by a growing awareness of the clinical importance of sarcopenia[1] - ‘a progressive and generalised skeletal muscle disorder that involves the accelerated loss of muscle mass and function’[2] – which in the last 5 years has been assigned an ICD-10 code[3].

Despite progress, there remain well-documented challenges for clinical practice and research on sarcopenia[1,2]. One of the most important is the ongoing debate relating to how sarcopenia should be operationally defined[1]. Of a number of consensus definitions proposed over the last decade, the European Working Group on Sarcopenia in Older People’s (EWGSOP) definition has gained considerable traction[4]. It was therefore noteworthy when an extended working group, EWGSOP-2, published a revised definition in 2019 to reflect updates to relevant evidence[5].

In working towards the aim of a true consensus definition and improved understanding of sarcopenia, each time a new definition is proposed it is important to compare this with existing definitions[6-9]. Work has thus been undertaken to compare the EWGSOP-2 definition with the EWGSOP and other definitions[10-14]. These comparisons have highlighted that prevalence estimates vary markedly and often have limited overlap i.e. identify different groups of people within a population as sarcopenic. This is a well-recognised challenge; a systematic review in 2019 reported that estimates of sarcopenia prevalence varied between 9.9 and 40.4% depending on the definition used and emphasised the lack of agreement between definitions[9].

The differences that these comparison studies highlight have important implications that need to be resolved. However, alongside these differences, it is also important to consider the impact of variation in how the core measures on which any one definition is based are assessed; this has been given considerably less attention. Identification of low muscle strength is an important step in the case-finding of sarcopenia as defined by EWGSOP2 and other groups. The EWGSOP2 propose doing this via assessment of handgrip strength or chair rises[5]. A number of sources of variation in the protocols commonly used to assess grip strength have been documented[15-18] one of which is the type of hand-held dynamometer used. Some studies have shown variation in the grip strength achieved by individuals when tested using different makes and model of hand-held dynamometer[19-22]. However, only one study, of community dwelling Japanese adults aged 69-89 years, has considered the implications of this for the case-finding of low muscle strength. In this comparison of two types of hand-held dynamometer (Jamar hydraulic and Smedley), there was marked variation in the prevalence of weakness identified[22].

In another recent cross-over study, measurements of maximum grip strength achieved by the same community-dwelling individuals assessed in a random order using four different makes and model of commonly used hand-held dynamometer, including electronic devices not previously examined were compared. Mean differences in maximum grip strength of between 4 and 5 kg were found when comparing hydraulic or spring-gauge devices with electronic devices[23]. In this report, we aimed to assess the impact of these measurement differences on the case-finding of low muscle strength to help inform future work on sarcopenia definitions.

We utilise data from a randomised, repeated measurements cross-over study, full details described elsewhere[23]. In summary, participants were 118 men and women, aged 45 to 74 years, who had previously taken part in a market research study, resided in London or the South East of England and who agreed, on invitation after checks to ensure their eligibility, to participate in a trial to compare different devices commonly used to measure grip strength, blood pressure and lung function. This study received ethical approval from the local UCL ethics committee (ref: 6338/001) and all participants gave written informed consent.

During attendance at a research office in London between October 2015 and January 2016, study participants completed a 2-page questionnaire on sociodemographics and health status. This included questions on whether they had arthritis or other musculoskeletal conditions that affect their hands and have difficulty because of long-term health problems holding something heavy like a full kettle or removing a stiff lid from a jar. They then had their grip strength, blood pressure and lung function assessed by a trained researcher.

Grip strength was assessed using four types of hand-held dynamometer in a randomly allocated order, with 5-10 minutes between each set of grip strength measurements to prevent participant fatigue. These devices were those most commonly used in UK longitudinal studies: Jamar Hydraulic; Jamar Plus+ Digital; Nottingham Electronic; Smedley. The same standardised protocol[15,23] was used for each set of measurements (see Appendix 1). After each test, the strength (kg) achieved was recorded. For each device, two measurements were assessed in each hand. The use of a standardised protocol and assessments in both hands ensured that any variations in grip strength measurements observed could be attributed to dynamometer type and not to other factors (such as handedness or position during testing).

After identifying the maximum grip strength achieved (from all four measures, two in each hand) by each individual using each device we applied the EWGSOP2[5] recommended cut-points for low muscle strength of <16 kg for women and <27 kg for men. We then summarised the prevalence of low muscle strength for each device and assessed the agreement between pairs of devices using kappa statistics.

Among the 118 community-dwelling men and women who participated in this study, 20 (17%) reported arthritis or another musculoskeletal condition that affects their hands and 24 (20%) reported some or a lot of difficulty holding something heavy. Prevalence estimates of low muscle strength varied by dynamometer type with a range of 3% to 22% for men and 3% to 15% for women (Table 1).

Table 1

Prevalence of low muscle strength and mean maximum grip strength (kg) by dynamometer type.

	Men (N=59)	Women (N=59)
Age (y) – mean (SD)	59.3 (7.7)	59.9 (8.4)
Low muscle strength* – n (%)
Jamar Hydraulic	5 (8.5)	9 (15.3)
Jamar Plus+ Digital	2 (3.4)	3 (5.1)
Nottingham Electronic	2 (3.4)	9 (15.3)
Smedley	13 (22.0)	2 (3.4)
Maximum grip strength (kg) – mean (SD)
Jamar Hydraulic	35.2 (8.5)	20.5 (5.7)
Jamar Plus+ Digital	39.4 (8.5)	25.2 (5.3)
Nottingham Electronic	40.9 (10.3)	24.4 (6.9)
Smedley	32.4 (7.4)	22.9 (5.0)

Low muscle strength defined as maximum grip strength <27 kg for men and <16 kg for women[

Of the 13 men identified as having low muscle strength by at least one of the four dynamometers, only one participant was identified by all four dynamometers and seven by just one device (and in all cases this was the same device) (Table 2 and Appendix Supplementary Figure 1). When comparing pairs of devices, kappa statistics ranged from 0.22 to 0.55 suggesting poor to moderate agreement. Agreement between devices among women was lower with kappa statistics ranging from 0.13 to 0.46; of the 15 women classified as having low muscle strength by at least one of the four dynamometers, only one was identified by all four and ten by only one.

Table 2

Number of men and women classified as low muscle strength by number of dynamometers.

Classified as low muscle strength by:	Men	Women
All 4 dynamometers	1	1
3	1	1
2	4	3
1 dynamometer	7	10
At least 1 dynamometer	13	15

Appendix Supplementary Figure 1

Overlap between men and women classified as low muscle strength by four different hand-held dynamometers (Jamar Hydraulic (Jamar H); Jamar Plus+ Digital (Jamar D); Nottingham Electronic; Smedley).

In a study comparing four commonly used hand-held dynamometers, we have shown that identification of low muscle strength is influenced by the type of dynamometer used to assess grip strength. The differences in prevalence estimates and the limited overlap in case-finding between devices that we have identified are likely to have important implications for research and clinical practice that need to be carefully considered in future work on sarcopenia.

Our findings concur with those from a previous study of older Japanese adults which compared two of the four devices we assessed[22]. As we were unable to compare the strength of associations between low muscle strength, identified by each of the four devices, and important health outcomes due to limited statistical power and no longitudinal assessment of relevant health outcomes, further research is required to clarify the implications of our findings. As our study population were aged 45 to 74 years, additional research is also required to establish whether these results are generalizable to older populations where sarcopenia is more prevalent.

It has previously been argued that differences in the absolute measures of grip strength recorded when using different makes and models of hand-held dynamometers may not be a major concern when pooling grip strength data for use in epidemiological studies of association on the assumption that different devices rank people equivalently[24,25]. However, these new findings highlight that when applying absolute cut-points to grip strength measures, systematic measurement differences between different types of dynamometer are a cause for concern. This suggests that there may be benefits of avoiding the application of cut-points and instead using continuous measures in research whenever possible.

The application of cut-points is necessary for case-finding in clinical practice and so cannot always be avoided. Research is ongoing to validate cut-points for grip strength and other measures required for sarcopenia case-finding. Our findings suggest that more work is also required to understand how best to standardise the measurement of each of the different components of any one sarcopenia definition to which these cut-points are applied. For low muscle strength, it is important to recognise that there are a number of potential sources of variation in the measurement of grip strength that may have an impact on case-finding[15,17,26-28]. While excellent work has been done to highlight some of these factors and promote the standardisation of assessments[15], it may be unrealistic to expect that all research and clinical facilities will ever have access to exactly the same measurement devices and be able to follow precisely the same protocols. In recognition of this, options to consider include the development of correction factors that can be applied to take account of measurement differences between devices. Alternatively, it may be necessary to consider developing different reference values and cut-points for different types of measurement device. However, whether this is feasible remains to be seen, especially as it can be expected that existing measurement devices will be updated and new devices will continue to be developed and introduced. Creating standard protocols that are followed to validate initial results prior to a diagnosis being confirmed and interventions being implemented, similar to those followed in diagnosing hypertension could also have value[29]. This could involve taking initial grip strength measurements using whatever device is available and, where low muscle strength is indicated undertaking additional assessments of muscle strength using a gold-standard method.

Funding

This project was funded by CLOSER (www.closer.ac.uk), whose mission is to maximise the use, value and impact of longitudinal studies. The CLOSER consortium is currently funded by the Economic and Social Research Council (award reference: ES/K000357/1). CLOSER was funded by the Economic and Social Research Council and the Medical Research Council between 2012 and 2017. CL is supported by an ESRC Doctoral Training Programme grant (ES/J500161/1) with supervision from Patrick Sturgis and Dave Martin. AW is supported by the UK Medical Research Council (MC_UU_12019/06). RH is Director of CLOSER.