Abstract
This study found that the Camry, a low-cost handgrip dynamometer, is highly reliable, valid, and durable and remains accurate even after many uses.
Handgrip strength tests measure isometric grip strength, which has been demonstrated to be a valid indicator of whole-muscle strength (Artero et al., 2012; Roman-Liu et al., 2021), muscle mass (Bohannon, 2015), and hand function (Pérez-Parra et al., 2024). Optimal hand function is essential to successfully complete daily life activities. This fact makes handgrip strength testing tools widely used across various areas of occupational therapy to monitor rehabilitation progress following hand pathologies or surgeries (Duruöz, 2014; Myles et al., 2023, 2024). Moreover, maximal handgrip strength (hereinafter just handgrip strength) has demonstrated to be an excellent biomarker of health in a wide range of populations (Henriksson et al., 2019; Jeong et al., 2023; Leong et al., 2015; Ortega et al., 2008). Evidence has shown that handgrip strength is negatively associated with many pathological conditions, such as cancer (Zhuang et al., 2020), dementia (Esteban-Cornejo et al., 2022), or cardiovascular disease (Welsh et al., 2020). In addition, handgrip strength is also inversely linked to physical impairments such as sarcopenia (Mayhew et al., 2023) or bone loss (McLean et al., 2021), making it a valuable measure in various health care settings and across various fields of knowledge, to address these concerns. Consequently, the assessment of handgrip strength may be of importance from a clinical, therapeutic, occupational, public health, and sport performance perspective.
Because of this evidence, handgrip has been widely used in research (Bohannon, 2015; Esteban-Cornejo et al., 2022). The assessment of handgrip strength is part of evidence-based fitness test batteries for different age groups, such as the PREFIT battery for children 3 to 5 yr old (Cadenas-Sanchez, Martinez-Tellez, et al., 2016), the ALPHA battery for children 6 to 18 yr old (España-Romero, Artero, et al., 2010), and the ADULT-FIT battery for adults 18 to 64 yr old (Cuenca-Garcia et al., 2022). The test is performed with hand dynamometers, which are easy and portable tools. Some handgrip dynamometers, such as the Jamar (Sammons Preston, Inc., Bolingbrook, IL) and the TKK (Takei, Tokyo, Japan), have been widely used and demonstrated to be reliable and valid for the assessment of handgrip strength (Cuenca-Garcia et al., 2022). Their reliability and validity have been tested with calibrated known weights, showing the TKK dynamometers have the highest test–retest reliability and concurrent validity (Cadenas-Sanchez, Sanchez-Delgado, et al., 2016; España-Romero, Ortega, et al., 2010).
However, the cost of these (Jamar and TKK) dynamometers has been relatively expensive (>$400), which might be a limitation for their use in certain settings, as well as in large-scale fitness monitoring (e.g., in school settings) and surveillance systems. Recently, a more cost-effective version, the Camry EH101 dynamometer (costing roughly $40), has been used in fitness surveillance systems in countries such as Hungary and Slovenia for large-scale nationwide fitness assessment (Csányi et al., 2015; Jurak et al., 2022). Furthermore, handgrip strength testing has been integrated into the FitBack platform (Ortega et al., 2023), offering a global opportunity to develop fitness monitoring systems. Consequently, the cost of dynamometers becomes a crucial factor in determining the feasibility of such systems, and the Camry EH101 is one of the low-cost alternatives. However, the reliability and validity of the Camry EH101 dynamometer have not been tested against known weights. In addition, testing the durability of the dynamometers is important because measurement error could potentially increase after years of heavy use (Cronin et al., 2017).
The main aim of this study was to investigate the test–retest reliability and concurrent validity (a subtype of criterion-related validity when measurements are conducted at the same time) of the Camry EH101 dynamometer by using calibrated known weights. In addition, we compared an old (over 3,000 uses for 8 yr) versus a new (purchased and used for this study) Camry EH101 dynamometer to investigate whether the accuracy of the measurement would change with use and time. Finally, we also included the digital TKK dynamometer (Model 5401, by Takei), for comparison purposes because it was previously shown to be highly reliable and valid (Cadenas-Sanchez, Sanchez-Delgado, et al., 2016; España-Romero, Ortega, et al., 2010).
Method
Instruments
A digital TKK 5401 dynamometer (Takei, Tokyo, Japan) with a range of measurement from 5.0 kg to 100.0 kg, two Camry digital hand dynamometers EH101 (Sensun Weighing Apparatus Group Ltd, Guangdong, China; one new and one old with more than 3,000 uses and 8 yr in school settings) with a range of measurement from 0.0 kg to 90.0 kg (Figure 1), and known weights were used for this study. For the verification of the weights, we used a high precision SECA scale (Model 769; SECA, Hamburg, Germany). Dynamometers, weights, and the scale were calibrated by the manufacturer. Following previous studies in this field (Cadenas-Sanchez, Sanchez-Delgado, et al., 2016; España-Romero, Ortega, et al., 2010), we assumed the validity of SECA scale as the criterion method because we could not test the dynamometers against any other gold standard method for weight. However, we tested its test–retest reliability by assessing the intertrial difference when measuring twice the weights in SECA scale (M = 0.06 kg, SD = 0.12), which indicated high reliability.
Camry EH101 (left) and TKK5401digital (right) dynamometers used in the study, and graphical illustration of the setup for the measurement, with the TKK5401 as example.
Procedure
We tested the dynamometers in a randomized order by using the known weights. Measurements ranged from 1 kg to 70 kg, starting from 1 kg and with 1 kg increments up to 20 kg, followed by 5 kg increments up to 70 kg. In total, 30 weights measurements were taken twice (test–retest) with each dynamometer. The weights, as well as the repeated measures, were also applied and conducted in a random order. The dynamometer was placed on two tables, and the weights were suspended from the dynamometer with a loading belt, hanging between the tables (Figure 1). The center of the dynamometer’s handle was previously marked with tape for the consistent placement of the loading. The dynamometers have adjustable handle, allowing them to adapt to a person’s hand sizes. For all measurements, we used a 5 cm grip span, previously measured from the base of the dynamometers to the grip handle with a measuring tape (see Figure 1), because this grip span falls within the range of optimal span observed for men and women (Ruiz et al., 2002).
Statistical Analysis
We applied the Bland–Altman method (Bland & Altman, 1986) to investigate intrainstrument (test–retest) and interinstrument (comparing the three dynamometers: new and old Camry EH101 and TKK 5401) reliability and concurrent validity (comparing dynamometers with known weights). Mean difference (systematic bias) between measurements and 95% limits of agreement (mean difference ±1.96 standard deviation of the difference) were calculated. Bland–Altman plots were created to visually represent individual variation of the measurements in the relationship between the measurement’s differences and means (Bland & Altman, 1986).
We calculated heteroscedasticity, considered as nonconsistency of error among weights increments, as the comparison of the measurement’s differences between weights ≤15 kg and >15 kg. We used this grouping to simulate populations with low handgrip performance, such as older adults, people with illnesses, patients with physical conditions, or very young children, and to test whether the reliability and validity would differ between low and medium-high performers. For heteroscedasticity analyses, we transformed the differences to absolute values (i.e., multiplying the negative values by –1) and compared them by using one-way analysis of variance, with the absolute differences as dependent variable and the weight groups (≤15 kg and >15 kg) as the fixed variable. Significant differences, expressed as p < .05, would confirm heteroscedasticity. In addition, visual inspection of the Bland–Altman plots can inform on whether there is heteroscedasticity. All statistical analyses were conducted using R software (Version 4.3.1).
Results
Reliability
Table 1 shows the mean differences among repeated measures with the same instrument (intrainstrument reliability) and with different instruments (interinstrument reliability). According to the intrainstrument test–retest reliability, the new Camry dynamometer had the smallest mean error (M = 0.01 kg, SD = 0.49), followed by its old version (M = −0.10 kg, SD = 0.49) and the TKK dynamometer (M = 0.14 kg, SD = 0.76). The intrainstrument reliability is graphically represented in Figure 2 using Bland–Altman plots. When we compared instruments, the mean differences between the two Camry dynamometers (old vs. new) were smaller (M = 0.03 kg, SD = 0.57) than the mean differences between the Camry dynamometers and the TKK (new Camry vs. TKK: M = 0.84 kg, SD = 0.79; old Camry vs. TKK: M = 0.88 kg, SD = 0.85; Table 1). Figure 3 shows Bland–Altman plots for interinstrument reliability. We tested heteroscedasticity by using absolute differences to investigate whether the variability between trials (test–retest) or between instruments changed as the magnitude increased (i.e., comparing weights ≤15 kg and >15 kg). Overall, heteroscedasticity was present in all comparisons and was significant (p < .05) for most of them, indicating that the variability is larger (i.e., the reliability is lower) at higher weights (Table 1). However, the errors were small (i.e., <0.5 kg differences between higher and lower weights).
Intrainstrument and Interinstrument Reliability of the Camry and TKK Dynamometers
Note. All means and standard deviations represent the difference between trials or instruments. Intrainstrument = comparison using the same instrument; Interinstrument = comparison between instruments.
aThe analysis split by weights groups was conducted by transforming the real difference variable into absolute differences by multiplying all negative values by –1, with a higher positive value indicating higher variation between test and retest or between instruments in any direction.
bAnalysis of variance was performed with the absolute differences as dependent variable and weight group as fixed variable.
Bland–Altman plots showing the test–retest reliability (retest minus test).
Bland–Altman plots showing the interinstrument reliability.
Validity
Concurrent validity results (dynamometers measures against known weights) are presented in Table 2. Concurrent validity showed a small-magnitude negative systematic error for the new Camry (M = −0.21 kg, SD = 0.35) and old Camry (M = −0.18 kg, SD = 0.79) dynamometers. This error was found to be slightly larger for the TKK dynamometer (M = −1.07 kg, SD = 0.75). The heteroscedasticity analysis showed a significant increment on the absolute mean difference for weights above 15 kg compared with lighter weights (≤15 kg), with the largest difference observed for the TKK dynamometer (roughly 1 kg larger error in heavier weights) and small differences for the Camry dynamometers (<0.5 kg). We also plotted results from concurrent validity by using the Bland–Altman method for each of the dynamometers studied (Figure 4). Although there was no obvious association between the real differences and the magnitude between the Camry dynamometers and the known weights, there was a clear negative association for the TKK dynamometer, indicating that the larger the magnitude (i.e., higher weights), the larger the underestimation of the TKK dynamometer compared with known weights.
Concurrent Validity of the Digital TKK and Camry Dynamometers Compared Against Known Weights
Note. All means and standard deviations represent the difference of dynamometers’ values minus known weights, with a negative value indicating that the dynamometer is underestimating the known weights and vice versa.
aThe analysis split by weights groups was conducted by transforming the real difference variable into absolute differences by multiplying all negative values by –1, with a higher positive value indicating higher variation between the dynamometer and known weights in any direction.
bAnalysis of variance was performed with the absolute differences as dependent variable and weight group as fixed variable.
Bland–Altman plots showing the concurrent validity of the TKK digital dynamometer and the Camry EH101 dynamometers against known weights.
Discussion
Our study contributes to the existing knowledge by providing novel findings about the objectively assessed (against known weights) reliability and validity of a low-cost handgrip dynamometer, the Camry dynamometer. First, test–retest reliability was excellent (i.e., mean error ≤0.1 kg) for all the dynamometers analyzed, being slightly superior for new and old Camry dynamometers (i.e., lower systematic and random errors) than for the TKK dynamometer. Second, the comparison between the new and old Camry showed strong consistency, indicating good durability. However, the comparability was lower between the Camry dynamometers and the TKK, particularly at high weights. Third, the new and old Camry dynamometers showed high validity against known weights, with a mean difference (bias) of –0.2 kg and a random error (limits of agreement) slightly smaller than 1 kg. On the other hand, the TKK showed on average roughly 1 kg underestimation compared with known weights and progressively increasing underestimation as the weights increased, suggesting it would have a larger error among individuals with high handgrip strength.
Known weights have been previously used for the assessment of reliability and validity of different dynamometers (Cadenas-Sanchez, Sanchez-Delgado, et al., 2016; España-Romero, Ortega, et al., 2010; Mathiowetz, 2002; Shechtman et al., 2005). However, to our knowledge, no study had targeted the reliability and concurrent related validity of the Camry dynamometer using known weights, which has important implications because of its markedly lower cost. When testing reliability and validity of different population groups, it is important to consider the inherent variability of human biology when performing repeated measurements.
Our results from the test–retest reliability analysis demonstrated the excellent reliability of the Camry dynamometer when using known weights. We found a mean systematic error between trials of 0.01 kg (SD = 0.49) for the new Camry dynamometer and –0.10 kg (SD = 0.49) for the old one. These results are in line with Latorre Román et al. (2017), who analyzed test–retest reliability of the Camry EH101 dynamometer on a healthy preschool children population (N = 1,215; mean age = 4.32 yr, SD = 1.05) and reported similar differences between trials (M = 0.11 kg, SD = 0.69), with an intraclass correlation coefficient (ICC; with a value of 1 as perfect reliability) of 0.969. Similarly, Mani et al. (2019) reported an ICC of 0.95 when testing the Camry EH101 reliability among 114 healthy adults, and Cao et al. (2021) found a highly stable ICC (0.737) among 599 female college students (mean age = 18.7 yr, SD = 1.00). In the case of TKK digital dynamometer, we found a mean systematic error of 0.14 kg (SD = 0.76) between test–retest measurements, which can also be considered good reliability and is in line with previous results from España-Romero, Ortega, et al. (2010), who found test–retest systematic error of 0.02 kg for TKK 5101 dynamometer, and Cadenas-Sanchez, Sanchez-Delgado, et al. (2016), who reported intrainstrument systematics errors for several TKK models ranging from a mean of 0.09 kg (SD = 0.65) to –0.33 kg (SD = 0.69). However, the variability of the measure changed slightly as the weights increased, suggesting its reliability might differ slightly between people with lower and higher handgrip performance.
Comparability among different dynamometers is crucial for interpreting and pooling data from different studies. To our knowledge, this is the first study comparing the Camry and TKK digital dynamometers. The TKK digital dynamometer has previously been validated, even showing lower systematic error against the Jamar dynamometer (España-Romero, Ortega, et al., 2010), which is widely and classically used. We found that the Camry dynamometer might show slightly higher handgrip strength values compared with the TKK digital dynamometer. In contrast, a previous study in a geriatric setting (N = 1,064; mean age = 66 yr, SD = 7.7; Huang et al., 2022) showed lower values of the Camry EH101 when compared with the Jamar dynamometer, with a difference of 0.5 kg for men and 0.6 kg for women. Similar results were found by Andrade et al. (2023), who also concluded that the Camry EH101 dynamometer showed lower grip strength, with an average difference of –0.11 kg for the right hand and –0.30 kg for the left hand among 220 older adults (mean age = 73.1 yr, SD = 6.3).
In addition, we assessed the durability of the Camry dynamometer by comparing a new device with an old one. Our findings revealed robust durability, with a mean systematic error of 0.03 kg (SD = 0.57), which suggest there is no need to calibrate the device after many uses, at least up to the roughly 3,000 uses over an 8-yr period of the dynamometer used in this study.
According to concurrent validity (comparison with known weights), our analysis revealed a small-magnitude negative systematic error for the three dynamometers. The new and old Camry dynamometers showed better agreements with the known weights, with a mean systematic error of –0.21 kg (SD = 0.35) and –0.18 kg (SD = 0.79), respectively, whereas the TKK digital dynamometer presented greater underestimation, with a mean systematic error of –1.07 kg (SD = 0.75) and a tendency to increase the error more markedly as the magnitude of the measure increased (i.e., heteroscedasticity). These findings are in concordance with those described by España-Romero, Ortega, et al. (2010) and Cadenas-Sanchez, Sanchez-Delgado, et al. (2016), whose studies showed a mean systematic error of 0.49 kg and between –0.94 and –2.02 kg, respectively. Heteroscedasticity was observed as well among the TKK digital dynamometers in these two studies, reinforcing the suggestion that the TKK’s measuring error might differ depending on the handgrip performance. Likewise, these previous studies also found the Jamar, TKK, and DynEx dynamometers underestimated between 0.5 lg and 2.6 kg compared with known weights (Cadenas-Sanchez, Sanchez-Delgado, et al., 2016). Thus, Camry dynamometers seem to have similar and even better agreement with known weights compared with other dynamometers.
The findings of this study have implications for occupational therapy, such as monitoring initial evaluation, progression, and recovery of hand pathologies and after hand surgery; assessing weakness in neurological conditions, including stroke or Parkinson’s disease; evaluating physical capability for workplace tasks to prevent overuse injuries; and detecting age-related functional decline to address frailty or fall risk. Occupational therapists can use the results to adjust interventions, track progress, and recommend ergonomic or functional modifications, thus promoting improved motor skills, independence, and safety.
The main strength of this study was that we determined the reliability and validity of the low-cost dynamometer Camry EH101 by using known weights, which avoided human variability. Future studies should replicate our findings and include other low-cost dynamometers, as well as other well-known dynamometers (e.g., Jamar), for comparability purposes.
Implications for Occupational Therapy Practice
This study has the following implications for occupational therapy practice: ▪ Occupational therapists and other health or fitness-related practitioners can use the Camry dynamometer to evaluate handgrip strength, because it is markedly cheaper than other dynamometers and has demonstrated good concurrent validity as well as test–retest reliability. This study therefore contributes to make handgrip strength testing more affordable and accessible. ▪ The Camry dynamometer is durable and does not need calibration, as shown in this study’s comparison of the validity and reliability of an old, heavily used model with a new model. ▪ The results give clear estimates of measurement errors, allowing researchers and practitioners to compare two timepoints and distinguish between the natural variability of the instruments and actual changes in handgrip strength.
Conclusion
The findings of this study indicate that the Camry EH101 dynamometer has excellent reliability and validity and can therefore be used for handgrip strength assessment of different populations with higher or lower handgrip strength levels. Our results also show that these properties remain after heavy use for several years, suggesting that these devices are durable and do not need recalibration. Most important, because of its lower cost, the Camry seems an excellent value-for-money alternative for the assessment of handgrip strength in large-scale population studies or for monitoring and surveillance systems, as well as for individual end-users in clinical, educational, and sport settings.
Footnotes
Acknowledgments
This study is mainly supported by Grant PID2020-120249RB-I00, funded by MCIN/AEI/10.13039/501100011033 and by the Andalusian government (Junta de Andalucía, Plan Andaluz de Investigación, ref. P20_00124). Lucía Sánchez-Aranda and Javier Fernández-Ortega are supported by the Spanish Ministry of Science, Innovation and Universities (FPU 21/06192 and FPU 22/03052, respectively). Isabel Martín-Fuentes is supported by the Spanish Ministry of Science, Innovation and Universities (JDC2022-049642-I). Angel Toval has received funding from the Junta de Andalucia, Spain, under Postdoctoral Research Fellows (Ref.POSTDOC_21_00745). This work is part of a PhD thesis conducted in the Doctoral Programme in Biomedicine of the University of Granada, Granada, Spain.