Reliability and validity of the hand-held dynamometer “NOD”: a new instrument for assessment of isometric grip strength

Abstract

Introduction

Hand grip strength (HGS) is an important measure in a physiotherapy assessment and for this purpose it is necessary to have valid and reliable instruments to measure it. In this study we aimed at investigating the reliability, validity, and agreement of the new hand-held dynamometer NOD (OT-Bioelettronica, To-Italy) compared to Jamar® hydraulic dynamometer (JD), the gold standard.

Methods

Fifty healthy subjects were selected; 9 trials for the dominant hand and 9 trials for the non-dominant hand were administrated to each of them: 3 trials of HGS with the JD in rung #3, 3 trials with the JD-adapted-grip (like the NOD one), and 3 trials with NOD. To verify the reliability of NOD, the Intraclass Correlation Coefficient (ICC 3,1) was calculated with a mixed effects model with the addition of adjustment variables (age, gender, dominant / non-dominant limb, trials). The model used single HGS measurements to estimate variance components, so reflecting both degree of correspondence and agreement among devices.

To assess concurrent validity NOD was compared to the “gold standard” JD, in terms of ICCs and through Pearson correlation. The agreement between the methods of measurement was calculated with the Limits of Agreement (LoA) and the plots of Bland–Altman.

Results

All ICCs show high inter-reliability; the results are very similar for both dynamometers. The value of the adjusted ICC of NOD was 0.90. For validity, Pearson correlations of NOD towards JD and JD-adapted-grip were high (r = 0.87 and 0.88). However, the LoA and the plots of Bland–Altman demonstrated that there is no agreement between NOD and JD and between NOD and JD-adapted-grip, with NOD showing lower mean scores than JD.

Conclusions

NOD is a reliable and valid instrument for HGS. However, even if it cannot be considered interchangeable with JD because there is no agreement between them in free-living adults, NOD is easier to carry than other dynamometers, it has a Bluetooth® connection with a free App and it is a multi-purpose tool that should be considered both in daily practice and in clinical settings.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12891-024-08222-2.

Introduction

In hand rehabilitation, strength assessment is necessary to make a diagnosis, establish realistic goals, make correct decisions, using specific measure outcomes [1–6]. Interestingly, handgrip strength is strongly related to several health outcomes, including mortality, disability, and the ability to perform activities of daily living [7, 8]. As a result, an accurate and reliable assessment of grip strength is essential for diagnosis, establishing realistic goals, and making informed clinical decisions.

Manual muscle testing (MMT), and hand-held dynamometry (HHD) are two common methods to assess muscle strength in occupational and physical therapy. In particular, hand-held dynamometry provides a direct and objective measurement of muscle strength by quantifying the peak isometric force in units of Newtons [1, 9–14]. Unlike MMT, which lacks a standardized unit of measurement, HHD provides quantifiable and more reliable results also in patients with lower levels of strength [15].

To date, there are several models of hand-held dynamometers (HHD) available for measuring isometric hand grip strength (HGS): hydraulic, pneumatic, electronic, with adjustable or modified anatomical shape, with adjustable rectified and complacent handle shape, parallelepiped, or cylindrical handle shape [10, 12, 16].

The American Society for Surgery of the Hand (ASSH) and the American Society of Hand Therapists (ASHT) identified the Jamar® dynamometer (JD) as the most used in clinical setting, and they considered it as the gold standard for isometric HGS measurement particularly used in geriatric subjects [1, 9, 17–21]. This small and portable instrument has excellent test–retest reliability, interrater reliability, and availability of normative data for children and adults [22]. The JD has been used as a standard criterion to validate other dynamometers [12, 13, 23] and to determine agreement, assessing the similarity or the consistency between two or more methods evaluating the same phenomenon, or parallel reliability, which refers to the consistency of results between two or more equivalent versions of a test, designed to measure the same construct with similar difficulty and structure, ensuring the tests produce comparable outcomes. [24, 25]. So, parallel reliability assesses consistency between equivalent test forms, while agreement measures how closely different raters or methods provide similar results.

The use of hand-held dynamometers and/or algometry in everyday clinical and home settings is well supported in recent studies [26]. Recently, Jayaseelan et al. [26] reported that when force is normalized by circumference of the applicator, a HHD was found to be a valid and reliable tool for measuring Pressure pain threshold (PPT). Clinicians may use HHD to detect relevant pain mechanisms at fault in their evaluation and treatment of pain. However, while the Jamar dynamometer is well-established, there remains a need for devices that offer additional functionalities that could optimize clinical processes and reduce potential experimental errors.

In this context, a new multi-purpose instrument was created—NOD device by OT Bioelettronica Company (Turin, Italy)—which has the value of combining three functions: hand-held dynamometer, algometer, and EMG-biofeedback. In particular, the NOD device might measure the peak isometric force generated by specific muscle groups with a force range between 0 and 50 kg. In addition, the device includes an algometer function, which might quantify PPT, helping in objectively assessing pain levels in clinical setting. Lastly, the EMG-biofeedback component of the NOD device monitors muscle activity in real time, providing insightful data for both diagnostic and therapeutic purposes.

Despite these considerations, to date, there is no evidence assessing reliability and validity of the hand-held dynamometer NOD in measuring HGS. According to the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) definition [27, 28], interrater reliability assesses the consistency of measurements made by different operators when using the same measurement instrument or method. On the other hand, concurrent validity is defined by COSMIN [27, 28] as the degree to which the scores of a particular measurement instrument relate to the scores of a gold standard measurement instrument. In this context, parallel reliability evaluates the consistency between equivalent test methods, and differs from the COSMIN framework assesses the quality of health measurement instruments across reliability, validity, and responsiveness.

Therefore, the primary objectives of this study were to evaluate the interrater reliability of the handheld dynamometer NOD compared to the gold standard (JD), i.e. the variability between the two devices measuring the same subjects, and to compare the devices in terms of response stability over measurements.

The secondary objectives were estimating and comparing HGS means between the two dynamometers, together with assessing concurrent validity and agreement. Moreover, the automation of data acquisition in the NOD device is intended to minimize potential experimental errors, providing a more optimized and error-reduced measurement process compared to traditional methods.

Methods

Study design and ethics

This prospective validation study conforms to the COSMIN Guidelines [28]. Adult participants (over 18 years of age) were recruited through a convenience sampling between March and December 2023 at the University Hospital SS. Antonio e Biagio e Cesare Arrigo, Alessandria, Italy. The study adhered to the ethical principles set forth in the Declaration of Helsinki [29]. Throughout the study, researchers ensured the protection of participants' privacy. Personal data were anonymized, and access to sensitive information was restricted to authorized personnel only. All participants provided informed consent, explicitly allowing the use of their records for research purpose. The study was approved by the Ethics Committee of the University Hospital SS. Antonio e Biagio e Cesare Arrigo, Alessandria, Italy (ASO.RRF.23.03; Prot. n. 14,310).

Instruments

The JD is a standardized hydraulic tool used to measure isometric HGS; it is considered to have the greatest precision and accuracy for measuring HGS [10, 12, 30]. It allows to adapt the width of the grip, so that it is possible to test HGS in 5 different positions. This property gives the examiner the opportunity to modify the instrument according to the anthropometric sizes of subjects and to assess the force–length relationship of the muscle involved. The higher value of isometric muscle force is considered (peak of isometric force). Guidelines recommend a frequency of three measurements of HGS during a single testing session [31, 32].

NOD device is a multi-purpose digital instrument that combines the functionality of three different clinical applications: algometer, EMG-biofeedback, and hand-held dynamometer. This last modality would seem to be able to measure both the isometric strength of the muscles of the body districts and the HGS. NOD device is a rigid platform with a parallelepiped shape (79 × 194 × 17 mm). Two semi-cylindrical magnetized pads can be applied over the platform thus obtaining an increase in the size of the instrument (52 mm) during the grip, comparable to the same handle spacing of the JD in rung #3 position (51 mm). The detected force is transmitted via Bluetooth® connection to a smartphone/tablet where a dedicated application will allow the display and recording of the force. The NOD device has been tested with reference to the EN 60601–1 and EN 60601–1-2 Standards.

Sample size and study participants

A minimum sample size of 35 participants was determined based on the following parameters: hypothesized Intraclass Correlation Coefficients (ICC) = 0.80, type I error = 0.05, type II error = 0.20. However, sample size was enlarged to 50 subjects with an equal-allocation design, to enhance study power (1-β = 0.90).

The inclusion criteria for participation were being aged 16 to 78 years, having signed the written informed consent, and having declared to be healthy. Hand dominance was established through verbal confirmation of activities such as writing, ensuring accuracy in participant classification. Exclusion criteria encompassed individuals with a history of trauma or neuro-musculoskeletal disorders within the past three months.

Participants were recruited consecutively and keeping equal cluster size as far as gender is concerned, so enrolment stopped when 25 females and 25 males were recruited.

Testing procedures

According to the ASHT, the protocol established the subjects sat on a stool with the shoulder in neutral rotation, arm adducted, elbow flexed at 90°, and wrist extension at 30° [33].

For NOD device, the hand was positioned with the thumb in extension leaning on the semi-cylindrical magnetized pad while the other fingers were wrapped around the other semi-cylindrical pad (Fig. 1a).

Fig. 1.

NOD grip modality: a NOD, b JD-adapted-grip non-dominant hand, c JD. Abbreviations: JD: Jamar® dynamometer

For JD-adapted-grip to thumb position was like the NOD (Fig. 1b), and for JD the handle was set to position on rung #3 (Fig. 1c).

While the ASHT protocol may not explicitly mandate a warm-up [33], all the participants were instructed about the instruments and invited to familiarize with the tests before starting the strength assessment. All 50 subjects received standardized verbal instructions from the same tester, and prior to the actual testing, three sub-maximal recordings were performed as warmup. After the warmup session, participants received verbal guidance and encouragement “one, two, three, Go! … Strength! … Strength! … Strength! … Relax … “, issuing in a vigorously manner, to guarantee the maximal HGS during the six seconds of each repetition, with ten seconds pause within each measurement [33]. Nine tests were administered for both the dominant arm and nine for the non-dominant arm: three trials for maximal HGS with JD on rung #3, three trials on a maximal HGS with JD-adapted-grip (the most like the NOD’s), and 3 trials on a maximal HGS with NOD. In order to standardize the assessment session, maximal HGS with JD on rung #3 was assessed firstly for three trials with the dominant hand, followed by three trials with the non-dominant hand. Subsequently, HGS with JD-adapted-grip was assessed for three trials with the dominant hand, followed by three trials with the non-dominant. Lastly, a maximal HGS with NOD was assessed firstly for three trials with the dominant hand, followed by three trials with the non-dominant hand.

Participants performed all tests without sight of the scale (JD) or monitor (NOD) of the device and had no visual feedback on their performance.

Statistical analysis

A repeated-measures design was used to investigate different reliability features of the NOD dynamometer and to evaluate concurrent validity and its agreement to the established “gold standard” JD (set in rung #3 and with adapted-grip, like the NOD one).

Reliability for each dynamometer was assessed through ICC with 95% confidence intervals (95%CI), calculated using single HGS measurements to estimate variance components, so reflecting both degree of correspondence and agreement among devices. In general terms, the ICC ranges between 0.00 and 1.00, with values closer to 1.00 representing stronger reliability [34, 35]. The 95%CI was estimated based on the equations proposed by Lu and Shara [35].

ICC were estimated by a mixed model, with single HGS measures as the outcome—model 3.1 [36]—dynamometers as a fixed effect, because they were the only devices of interest and to allow estimation of residual variance components for each instrument. Participants were considered as a random sample from a larger population. As the variability of interest is the one due to subjects, who were considered clusters in which multiple observations are grouped, these ICC represent the relationship between subjects variability on total variance [36]. To take into account possible differences in HGS among participants some demographical, physical, and organizational baseline characteristics that could influence the average outcome were added to the model, and related parameters were estimated. In detail:

• age in completed years, used as a covariate as HGS can decrease with increasing age;

• gender, used as a fixed effect as the study design is gender-balanced, as females are supposed to apply a weaker HGS than males due to biological characteristics;

• dominant / non dominant limb, used as a fixed effect as they are not interchangeable within each subject;

• subsequent trials, used as a fixed effect to account for the fact that HGS measurement followed a fixed sequence, which could bring fatigue and/or memory effects.

Response stability of maximal HGS was estimated for all dynamometers through model-based coefficients of variation (CVs) [37, 38]. CV is defined as the ratio of the standard deviation (SD) to the mean and it is expressed as a percentage; the higher the CV, the greater the dispersion level around the mean [34].

For concurrent validity, Pearson’s product-moment coefficients (r) between JD/NOD and JD-adapted-grip/NOD were calculated. These correlations were interpreted as absent or low (< 0.25), fair (0.25–0.49), moderate (0.50–0.75), very good (> 0.75) [34]. To better address agreement between JD/NOD and JD-adapted-grip/NOD, and not correlation only, Bland–Altman analysis was performed [39]. So mean difference in measuring HGS with two instruments and the 95% limits of agreement (LoA), i.e. mean ± 2 SD, were calculated, considering all subjects’ measurements with both hands. In general, values close to 0.00 indicate that the two methods are producing similar results. Results were graphically examined by plotting mean differences against the averages for each subject, with the zero line representing no difference [34].

Before model estimation, normal distribution of scores was assessed both statistically (Shapiro–Wilk test) and visually (Q-Q plot). Significance level was set at an alpha value of less than 0.05, for all statistical tests.

Statistical analyses were performed using SAS statistical analysis software – version 9.4 (SAS Institute Inc., Cary, NC, USA). Sample size calculation was done by STATA (statistics/data analysis) – version 18 (StataCorp., Texas, USA). Microsoft Excel was used for data acquisition and graph generation.

Results

A total of 50 participants were enrolled in the study, comprising 25 females (average age 41.52 ± 15.38 years) and 25 males (average age 36.04 ± 18.68 years). The majority of participants were right-handed (n = 43; 86%), only 7 were left-handed (14%).

Table 1 describes reliability measures and average HGS for each device, estimated through a mixed model, so to also control for differences in participants’ characteristics. Inspection of outcome distributions detected no departure from normality.

Table 1.

Model-based reliability comparing NOD, Jamar-adapted-grip, and Jamar

Average Hand Grip Strength in kg (standard error)	Coefficient of variation	Variance components		Intraclass correlation coefficient (95% CI)
		Source of variation	Estimates (standard error)
NOD
18.69 (0.90)	4.81%	participant (σ2s1)	40.02 (8.58)	0.90 (0.84–0.94)
		residual (σ21)	4.28 (0.39)
Jamar-adapted-grip
23.23 (1.16)	4.99%	participant (σ2s2)	66.17 (13.68)	0.90 (0.83–0.94)
		residual (σ22)	7.28 (0.65)
Jamar
30.48 (1.25)	4.10%	participant (σ2s3)	76.80 (16.19)	0.89 (0.82–0.94)
		residual (σ23)	9.31 (0.85)

Estimated HGS resulted in lower mean scores for NOD (average 18.69 ± 0.90 kg) than for JD-adapted-grip (average 23.23 ± 1.16 kg) and for JD (average 30.48 ± 1.25 kg), so showing a weaker effect of NOD compared to JD, when holding all the other variables in the model constant.

As far as variability of average measurements of maximal HGS are concerned, JD showed a smaller CV (4.10%) in respect of NOD (4.81%); though the difference between them was tiny.

Table 1 also shows, for each dynamometer, the variance due to participants and the residual one, adjusted for demographic-physic-organizational variables. The residual variances of the three instruments differ, so suggesting some heterogeneity between them, ICC estimated by variance components (see Appendix for further details). However they indicated good and similar reliability for all devices: ICC = 0.90 (95%CI: 0.84–0.94) for NOD and for JD-adapted-grip (95%CI: 0.83–0.94), and ICC = 0.89 (95%CI: 0.82–0.94) for JD, with non significant difference between them in terms of 95%CI. Moreover, NOD showed the smallest residual variance component (VC = 4.28; 9.66% of total variance) compared to JD-adapted-grip (VC = 7.28, 9.91% of total variance) and to JD (VC = 9.31; 10.81% of total variance).

Table 2 shows estimated parameters for adjustment variables. Age had a negative coefficient, suggesting that participants’ HGS significantly decreased of about 0.11 kg for every increasing year of age. As expected, males performed a greater strength than females (mean difference: 11.46 kg; p-value < 0.0001), and measurement with the dominant limb were stronger than with the non-dominant one (mean difference: 2.55 kg; p-value < 0.0001). HGS measured on subsequent trials statistically decreased going from trial1 to trial3, ranging from 25.05 kg to 24.26 kg, and to 23.09 kg, with significant mean differences between each couple (p-value < 0.001). However, eight subjects (16%) showed an increasing sequence in HGS measures from trial to trial (one subject using JD-adapted-grip and seven subjects using NOD).

Table 2.

Model-based estimates of participants’ characteristics, by adjustment variable

Variable	Estimated parameter from HGS (in kg)		T-test p-value
AGE	β coefficient		0.0024
	-0.11
	estimated mean	mean difference
Gender
Males	29.87	11.46 (males vs females)	< 0.0001
Females	18.40
LIMB
Dominant	25.41	2.55 (dominant vs non-dominant)	< 0.0001
Non-dominant	22.86
TRIALa
Trial 1	25.05	0.78 (trial 1 vs trial 2)	0.0004
Trial 2	24.26	1.96 (trial 1 vs trial 3)	< 0.0001
Trial 3	23.09	1.18 (trial 2 vs trial 3)	< 0.0001
TOOL*
NOD	18.69	-4.54 (NOD vs JD-ag)	0.0068
Jamar-adapted-grip	23.23	-11.79 (NOD vs JD)	< 0.0001
Jamar	30.48	-7.25 (JD-ag vs JD)	0.0001

^aTukey-Kramer adjusted for multiple comparisons

As far as concurrent validity is concerned, Pearson’s correlation coefficient between NOD and both JD and JD-adapted-grip was very high and significant: r = 0.87, p < 0.000, and r = 0.88, p < 0.000, respectively.

The agreement between NOD and JD consistently resulted in lower mean scores for NOD (Fig. 2); while between NOD and JD-adapted-grip the spread of mean scores was slightly more evenly distributed above and below the zero line (Fig. 3). In both comparisons, the difference in scores taken by NOD and the other dynamometer (JD or JD-adapted-grip) increases as the average value does. This pattern, suggesting some form of heteroscedasticity, is particularly evident when considering agreement between NOD and JD, while when comparing NOD to JD-adapted-grip the data still show an increase in scores at increasing mean values, but of a lesser entity. This is confirmed by the fact that the means of the difference scores between NOD and JD was equal to -11.79 kg, i.e. 95% LoA lied between -25.10 and 1.51 kg, a range of about 27 kg. The means of the difference scores between NOD and JD-adapted-grip was -4.54 kg, i.e. 95% LoA was expected to vary between -15.24 and 6.16 kg, a range of about 21 kg.

Fig. 2.

Parallel agreement: Bland–Altman plots of NOD vs Jamar. Difference scores between Nod and Jamar, plotted against mean scores for each subject. Dashed line shows the mean difference score (-11.79 kg). Solid lines show 95% upper and lower limits of agreement, representing 2 standard deviations above and below the mean difference score (-11.79 ± 13.30). Green line represents the no error point (complete agreement)

Fig. 3.

Parallel agreement: Bland–Altman plots of NOD vs Jamar-adapted-grip. Difference scores between Nod and Jamar-adapted-grip, plotted against mean scores for each subject. Dashed line shows the mean difference score (-4.54 kg). Solid lines show 95% upper and lower limits of agreement, representing 2 standard deviations above and below the mean difference score (-4.54 ± 11.06). Green line represents the no error point (complete agreement)

Discussion

Our results based on recruited subjects’ characteristics, suggest the NOD device can be considered a valid and reliable HHD as supported by the ICC = 0.9 of the NOD device compared to the gold standard, adjusted for demographic-physic-organizational factors. While validity refers to the accuracy of the NOD in measuring hand strength, caution should be posed when substituting one device for the other. In particular, interchangeability goes beyond validity alone and potential systematic errors as shown by the parallel agreement assessed by the Bland–Altman analysis.

Usually, the CV has been used as an indicator of measurement error during maximal HGS assessments [37, 40], and the cutoff value in the related literature varies greatly and ranges from 7 to 20%, with 11% and 15% being the most common [38, 40]. However, for some authors the CV measurement variability of maximal HGS among trials is not an appropriate method for determining the consistency of voluntary effort, especially in patients with compromised hand strength [37, 38]. The model-based CV values obtained in this study correspond to about 5%, in the lower boundary of literature data.

Stability of measures over time was assessed through a test–retest approach, with ten seconds pause within each measurement and keeping all testing conditions constant [34]. The mixed model used to ICCs estimation also includes some factors, as age, gender, dominant / non-dominant limb, subsequent trials, to take into account possible differences in HGS related to these facets, as suggested in literature.

In adult populations, HGS grows weaker with aging with a peak at around 20 years of age. Our results, showing a decrease of about 0.11 kg for every additional year, are consistent with published literature [41, 42].

Moreover, on the principle of maximal muscular effort is represented by the one-repetition maximum (1RM) test. As the name suggests, 1RM is defined as the maximum weight that can be lifted once and only once for that movement [43]. Consequently, the subsequent repetitions must have lower and decreasing values. In accordance with this principle, the HGS tests showed a progressive decrease in the values recorded by the recruited participants, for all instruments and after controlling for potential differences due to increasing age, genders, and dominant / non-dominant hand. Similarly, in maximum voluntary isometric contraction (MVIC) tests, while the contraction is static rather than dynamic, repeated trials often result in decreasing force outputs due to fatiguability, paralleling the decline observed in 1RM tests. Both methods demonstrate a decreased ability to exert maximal strength over subsequent repetitions, emphasizing the principle of muscle fatigue and the progressive decrease in strength over time [44]. Yet, the eight subjects with an increasing sequence represent an interesting exception. As seven of them showed this pattern when measured with NOD, this is suggestive of two opposite issues. On the one hand, NOD is more exposed to a learning effect than JD, which is probably due to its grip; on the other hand, NOD being an electronic device returns more precise data reading than JD. This observation underscores the potential for performance improvement through familiarization with the device, rather than a mere memory effect.

According to the literature, the dominant hand has a greater capability for HGS than the non-dominant hand in both genders [20, 45], with the general rule assuming that the dominant hand is 10% stronger, compared to the non-dominant one [18, 46–49]. In our study, the results are in line with the literature, obtaining for the dominant hand higher values of 11.15%.

Males were on average stronger than females from adolescence onwards: males’ peak median grip was 51 kg between ages 29 and 39, compared to 31 kg in females between ages 26 and 42 [41]. Also, when comparing HGS between males and females, our results are similar to published literature [1, 9], with males applying an average strength of 29.87 kg versus 18.40 kg among females.

However, the HGS values within subjects obtained from NOD (18.49 kg) are lower than those from JD (30.48 kg), so showing a weaker effect in this last comparison with a mean difference of 11.79 kg. This is suggestive of a “measuring bias” as NOD systematically detects less strength than JD, and the possible differences may be due to the different “grip system” used. As in other studies, the JD recorded HGS values higher than those of other dynamometers [10, 16]. Amaral et al. [10] demonstrated that there is an influence of the dynamometer’s handle shapes on the measurements of HGS. The NOD in fact has a more cylindrical handle than the JD or similar "two-pole". This difference in values may be because the intrinsic thenar musculature and thumb flexors during taking with the NOD would not be able to completely transmit the force, but it would mainly the function of maintaining the thumb resting on the instrument. In fact, if we compare the grip NOD with the JD-adapted-grip this difference in strength tends to decrease (NOD vs. JD-adapted-grip: -4.54 kg).

As already stated, reliability of the NOD was obtained through a mixed model adjusted by age, gender, dominant / non-dominant limb, and subsequent repetitions. Although ICC were similar between NOD and JD, NOD obtained the smallest residual variance component, accounting for 9.66% of the total variance.

As far as concurrent validity is concerned, Pearson correlation between NOD and JD was above 0.87, so demonstrating a good correlation among them, even if this index does not totally reflects the extent of agreement in the data but it well describes the relationship between the two sets of measures [34]. This result was slightly lower than reported correlations between the JD and other dynamometers.

While agreement is visually represented by Bland–Altman plots, which showed lower mean scores for NOD than for JD (with and without adapted grip) and 95% LoA pointed out a range of variation of about 27 kg for the means of the difference scores between NOD and JD and of about 21 kg for NOD against JD-adapted-grip. These values are higher than other studies [12, 23] and therefore it can be said that, unless the systematic bias was estimated and the measurement of one of the two instruments was corrected, NOD cannot be considered interchangeable with JD in clinical practice, but when it is the only dynamometer to be used its scores are certainly repeatable and consistent.

Moreover, Bland–Altman plots display some form of heteroscedasticity, as the difference in scores taken by NOD and the other dynamometer (JD or JD-adapted-grip) increases as the average value does. This is probably due to the grip system explained above and it is consistent with other studies that compared the agreement between JD and other models of dynamometers for HGS measurement have exhibited variability in the provided results [12, 23]. Differences in HGS values from 2.2 to 8.7 kg and with mean LoA, were previously found between the JD and other dynamometers. The large values of LoA obtained with the NOD showed values higher than those previously mentioned, thus making the instrument not interchangeable with the JD.

Despite this, the NOD shows some advantages compared to the JD, such as its lightweight (300 g versus the 700 g of the JD) and the Bluetooth® connection with a free App or computer, which allows the examiner to save all the collected data. The multiple use of one instrument with rapid and automatic transferal of measurements to the computer system prevents loss of results and error in reporting, and it supports optimal use of staff time.

Lastly, NOD does not require yearly calibration like JD [13], as the NOD device utilizes high-quality, stable sensors designed to maintain their calibration over time, reducing the need for frequent recalibration.

Compared to JD (but not to other hand-held dynamometers), NOD also has some disadvantages such as the non-adjustment of the handle. This lack of adjustability is particularly concerning when using the NOD for comparisons between individuals, as it may lead to inaccuracies in measuring grip strength across different users.

In future studies it would be interesting to analyze the validity and reliability of the NOD with cylindrical grip dynamometers, on other body areas and on subjects with disabilities.

Despite these findings, this study is not free from limitations. In particular, the main limitation is that JD was not adjusted to fit the hand dimensions of each individual participant, with potential implications for accuracy and consistency of the hand grip strength measurements since each participant's hand size might influence the force exerted. Future research should be conducted with a larger sample size per age, gender, and pathologies.

Regarding technical limitations, this study did not include the used JD handle position #2 which allows the highest grip force values as NOD has a load limit between 0 and 50 kg. The exclusion of adjustable regulation could potentially result in recorded values exceeding 50 kg, which is a significant limitation.

However, it should be noted that no study participants exceeded the 50 kg limit during the study suggesting that this limit did not significantly affect our findings.

However, the loading limit might have negative implications in the applicability of NOD device to specific subgroup populations characterized by higher muscle strength.

Lastly, we validated the NOD device as a handheld dynamometer, without providing evidence for its algometry and EMG biofeedback functions. Further studies are needed to characterize the advantages of these additional functionalities, exploring their potential to enhance clinical assessments and interventions in rehabilitation.

Conclusion

Taken together, findings of this study suggested excellent reliability (ICC = 0.9) of the NOD device compared to the gold standard in both male and female subjects. Although HGS measurements with the NOD are systematically lower, the data showed that it is an instrument very stable with a very small variability.

Therefore, this study confirmed that NOD is a reliable tool that can be used by clinicians to measure HGS. Furthermore, it is lighter and smaller than other dynamometers and it is a multi-purpose tool that should be considered both in daily practice and in clinical settings.

Further research should clarify the role of NOD in assessing muscle strength of body parts other than hands, and in subjects with disability.

Abbreviations

COSMIN

COnsensus-based Standards for the selection of health Measurement Instruments

HGS

Hand grip strength

JD

Jamar® hydraulic dynamometer

SEM

Standard error of measurement

LoA

Limits of agreement

ICC

Intraclass correlation coefficient

CV

Coefficient of variation

SD

Standard deviation

SE

Standard error

r

Pearson’s product-moment coefficients

kg

Kilograms

mm

Millimeters

HHD

Hand-held dynamometer

MMT

Manual muscle testing

1RM

One-repetition maximum

VAS

Visual analogue scale

NRS

Numeric rating scale

PROM

Patient reported outcome measure

RoM

Range of motion

PPT

Pressure pain threshold

95%CI

Confidence interval for 95% confidence level

VC

Variance components

Authors’ contributions

SF, LL and IM participated in the design of this study. BS, and ZD participated in data acquisition and collection. VS, FG, IM, and GF analyzed the data. SF, LL, BS, and ZD drafted the manuscript. dSA and IM revised the manuscript draft. GF, FG, and SV performed the statistical analysis. All authors read and approved the final manuscript.

Funding

This publication is part of the project NODES which has received funding from the MUR – M4C2 1.5 of PNRR with grant agreement no. ECS00000036.

Data availability

Data are available on request from the corresponding author at lorenzo.lippi@uniludes.ch.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of the University Hospital SS. Antonio e Biagio e Cesare Arrigo, Alessandria, Italy (ASO.RRF.23.03; Prot. n. 14310). Informed consent was obtained from all subjects involved in the study.

Consent for publication

Informed consent was obtained from all subjects involved in the study.

Competing interests

The authors declare no competing interests.