Test-retest reliability and concurrent validity of three different handgrip dynamometers (Jamar, Jamar Plus+ and Biodex) in healthy young adults

Abstract

Background

Handgrip strength (HGS) is a key indicator of upper limb function and overall health. While the Jamar dynamometer is considered the gold standard for HGS assessment, alternatives like the Jamar Plus+ and Biodex have emerged, though their reliability and agreement remain underexplored.

Aim

To evaluate the test–retest reliability and agreement of the Jamar, Jamar Plus+, and Biodex dynamometers in healthy young adults.

Methods

This exploratory cross-sectional study included 35 participants (22 females, 13 males; mean age 23.4 years). HGS was assessed using the three devices, following standardised protocols, with retest after 7 days. Reliability was analysed using the Intraclass Correlation Coefficient (ICC), Standard Error of Measurement (SEM), and Minimal Detectable Change (MDC). Agreement was examined through Bland–Altman plots and Limits of Agreement (LoA %).

Results

Jamar and Jamar Plus+ showed excellent reliability (ICC = 0.96–0.98) and strong agreement, with narrow LoA percentages and low measurement error, especially on the non-dominant hand. In contrast, comparisons involving the Biodex revealed wider LoA (up to ±73%) and higher SEM and MDC values, indicating lower agreement with handheld devices. These discrepancies are likely due to differences in measurement principles, hand positioning, and unit conversion. Male participants showed significantly higher grip strength across all instruments.

Conclusion

The Jamar and Jamar Plus+ can be used interchangeably. The Biodex may be considered when appropriate adjustments are made, though its agreement with handheld dynamometers is limited. These findings highlight the need for standardised protocols and further research to ensure consistent and reliable HGS assessment across devices.

Introduction

Hand function is understood as specific and refined skills that emphasize grip, strength, action and coordination to understand participation in a variety of human occupations while maintaining good performance. ¹

Handgrip strength (HGS) assessment is essential to outline behaviors and monitor patients’ parameters before, during, and after conservative (rehabilitation) or surgical treatments. ² HGS can be correlated with overall strength,^3,4 risk of falls in older people,^5,6 quality of life in adults, ⁷ fatigability and upper limb dysfunction,^8,9 other technological equipment ¹⁰ and bone mineral density. ¹¹ In addition, reference values for specific populations and regions consider rheumatological diseases, ¹² healthy adolescents, ¹³ healthy adults,^14–17 healthy older adults ¹⁵ and bone mineral density. ¹⁸

A dynamometer is an essential tool for evaluating the upper limbs, which allows strength determination through static muscular effort. ¹⁹ There are different types of handgrip strength dynamometers, with different prices and analysis systems (hydraulic, digital, pneumatic and spring), these being: Jamar, ²⁰ Jamar Plus+,^21,22 Saehan, ²³ Saehan DHD-1, ²⁴ Saehan Squeeze, ²⁵ GripAble, ²⁶ Camry, ²⁷ K-Force, ²⁸ Smedley Spring, ²⁹ and Smedley Digital, ³⁰ Kern Map, ³¹ Baseline BIMS (BIMS), ³² E-LINK© ³³ and Biodex ³⁴ (Table 1).

Table 1.

Different types of dynamometers.

Dynamometers	Systems
	Hydraulic	Digital	Pneumatic	Spring
Jamar 20	+	−	−	−
Jamar Plus+21,22	−	+	−	−
Saehan 23	+	−	−	−
Saehan DHD−1 24	−	+	−	−
Saehan squeeze 25	−	−	+	−
GripAble 26	−	+	−	−
Camry 27	−	+	−	−
K-force 28	−	+	−	−
Smedley29,30	−	+	−	+
Kern map 31	−	+	−	+
BIMS 32	−	+	−	−
E−LINK© 33	−	+	−	−
Biodex 34	−	+	−	−

+ present | − absent.

Among these, Jamar has been recommended as the gold standard for HGS testing by the American Society of Hand Therapists (ASHT). ³⁵ A prerequisite for comparing Jamar HGS results in research and clinical practice is whether the device used has defined measurement properties ³⁶ which allow its results to be compared with those of other dynamometers.

A reliable instrument allows predictable accuracy measurements under stable conditions. One relevant factor that can influence this measure, is the measurement assessment instrument itself, to ensure the reproducibility of the measure and maintain objective communication between professionals. ³⁷

Studies have shown excellent reliability and validity in assessing handgrip strength in healthy adults using the Jamar and Rolyan. ³⁸ In another study, the sphygmomanometer dynamometer was shown to be valid when compared to the Jamar, also in healthy adults. ³⁹ Other studies have also shown good reliability and validity results between the Jamar and other dynamometers, such as the Takei dynamometer and EMG System Manual Transducer in healthy adults. ¹⁹

Finally, Jamar showed good validity and moderate - excellent reliability results when associated with others dynamometers, such as Roylan, ³⁸ Sammons Preston Rolyan Bulb, Eisenhut, ⁴⁰ CAMRY, ⁴¹ Smedley,^40,41 K-force ⁴² and Jamar Plus+.^43,44 On the other hand, Jamar Plus+ has also been used in some studies to assess the reliability of HGS with other devices, such as BIMS^45,46 and GripAble, ⁴⁷ in which both can be used interchangeably.

However, recently there have been more studies evaluating the Biodex for the lower limbs and measuring its correlation and high reliability with other equipment (Cybex and Con-Trex).^48–53 In another study, Biodex showed good or excellent reliability of isokinetic strength of the ankle, knee and hip in physically active adults. ⁵³ Furthermore, Biodex demonstrated moderate reliability for shoulder assessment, specifically shoulder abduction and rotation in healthy individuals. ⁵⁴ It showed excellent reliability for elbow flexion in atypical children and for wrist flexion and extension in healthy individuals.^55,56 Despite this, only one study was found that analysed the reliability of the HGS considering Jamar and Biodex, in which Jamar showed more consistency than Biodex in assessing people with advanced cancer. ⁵⁷

While there is evidence that the Jamar and Jamar Plus+ are interchangeable, there is a gap in the literature regarding the use of the Biodex in the assessment of HGS. Therefore, this study was designed to verify the test-retest reliability of the isometric grip strength assessment between the Jamar, Jamar Plus+ and Biodex dynamometers in healthy young adult volunteers, with the aim of obtaining reproducibility estimates.

Methods

Hypothesis

It is expected that there will be moderate to excellent reliability and agreement between the dynamometers, as evidenced by:

• -Reliability between the Biodex with (a) Jamar and (b) Jamar Plus+ dynamometers.

• -Reliability between the Jamar with the Jamar Plus+ in young adults, as the study by Savas et al. ⁴³ was conducted with an elderly sample.

Reliability values of ICC = 0.7, α = 5%, power = 0.8 are expected, as well as adequate limits of agreement suggesting that the devices are interchangeable.

Study design and outcome measures

This is an exploratory cross-sectional study. It was carried out in three stages: (1) sociodemographic assessment and global lateral preference inventory (2) assessment of handgrip strength in healthy young adults, and (3) assessment of reliability using the test-retest method with the Intraclass Correlation Coefficient (ICC), SEM, MDC and Limits of Agreement (LoA) with Bland-Altman plots.

Participants and local

Walter et. al. ⁵⁸ and Mokkink et. al. ⁵⁹ recommend a sample of at least 30 volunteers for device reliability studies. The number of observations was considered to be equal to two (test and retest/k = 2).^58,59

The study consisted of a convenience sample, with the following inclusion criteria: a) volunteers of both sexes, b) regardless of physical activity level, c) aged between 18 and 40 years; d) individuals who did not have pain or discomfort in the cervical spine and/or upper limbs. Participants with: a) a history of trauma, b) recent upper limb pain or surgery; c) inability to perform the proposed tests for physical or cognitive reasons were excluded. All participants (22 women and 13 men) were recruited face-to-face by an invitation. They agreed to take part in the study by reading and signing the informed consent form.

The test and retest data was collected at the Rehabilitation Centre of the Ribeirão Preto Medical School, São Paulo - Brazil.

Instruments

The Global Lateral Preference Inventory (GLPI) is divided into eight parts and can be used as a whole or just for specific topics, according to the laterality dimension desired to be analysed. In its instructions, the person being assessed must observe the description of the questions presented with different everyday activities and indicate their lateral preference for the tasks described. ⁶⁰

The GLPI B dimension comprises the characterisation of upper limb laterality, in which hand preference is analysed through everyday tasks. In this laterality dimension, there are three subdivisions of levels of analysis (GLPI B1, B2 and B3), each with five items, totalling 15 questions. ⁶⁰

The GLPI B is scored from 1 to 5, with 1 being ‘strongly left-handed’, 2 ‘moderately left-handed’, 3 ‘indifferent’, 4 ‘moderately right-handed’ and 5 ‘strongly right-handed’. When the patient doesn’t know how to answer, a * is used, which is equivalent to ‘don’t know’. The average of the scores for each group assessed is then calculated. ⁶⁰ Then the averages are added together, divided by three and rounded off to obtain the value for classifying the individual’s laterality, again using the interpretation of 1 as ‘strongly left-handed’, 2 ‘moderately left-handed’, 3 ‘indifferent’, 4 ‘moderately right-handed’ and 5 ‘strongly right-handed’. ⁶⁰

The Jamar isometric dynamometer consists of a hydraulic system (Table 2) developed by Bechtol in 1954. ⁶¹ It is an instrument capable of directly reading handgrip strength in kilograms (Kg/f) or pounds. Its use is recommended by the ASHT ³⁵ as the gold standard for measuring isometric hand strength.^62–64

Table 2.

Properties of Jamar, Jamar PLUS+ and Biodex.

The Jamar Plus+ Digital is an electronic version of the hydraulic Jamar, but in principle, it has electronic measurement through the reading of the force sensor that captures the information through a load cell and it is lighter than Jamar with minor differences.^47,65

The Biodex System 4 Pro™ (Biodex) isokinetic dynamometer is a mechanical-electrical system capable of measuring the muscular action of the largest muscles in the body through multiple modes, being able to estimate the values by isokinetic, isometric, isotonic and eccentric methods, obtaining data including: force (Newtons), torque (Newton meter), degree of movement (angles), angular velocity (number of angles per second) and duration in seconds.^66,67

Furthermore, all the equipment used in the study was factory-calibrated and stored according to its instructions, with the exception of the Biodex: it is calibrated automatically with its computerised system.

It is noteworthy that the ASHT recommends that for strength measurement, the instrument must be positioned in the second position (4.8 cm away between the anterior and posterior handles) of the Jamar. ⁶⁵ Such positioning promotes a balanced activation of the fingers' intrinsic and extrinsic flexor muscles.^68,69 However, in order to analyse the characteristics of the devices, it was necessary to adapt these guidelines, since the minimum distance for capturing and reading by the Biodex was 6 cm. Accordingly, the handles of the Jamar and Jamar PLUS+ were set to the 3rd position (6 cm). Thus, the volunteers’ positioning pattern during the collection was established (Figure 1).

Figure 1.

6 cm handle adjustment.

Procedures

The procedures were divided into two stages (Stage 1 and 2), both conducted by a single trained hand therapist to reduce inter-rater variability. In the test phase, steps 1 and 2 were carried out. An interval of 7 days was determined to carry out the retest, after which a new date was scheduled to repeat stage 2.

• Stage 1: Initially, the volunteer completed a sociodemographic assessment (gender, age, occupation), and the Global Lateral Preference Inventory (GLPI) was also completed. Finally, a draw was made to randomise the order of device testing in stage 2.

• Stage 2: A warm-up was carried out on a mechanical cycle ergometer for the upper limbs lasting 5 min without adding load, ⁴⁶ followed by strength collection with the three pieces of equipment (Biodex, Jamar and Jamar PLUS+), in which participants were properly positioned in the chair (Table 2).

After positioning, three attempts of HGS were collected from the volunteers for each dynamometer, alternating hands, always starting with the dominant one (Right or Left), which could be R-L-R-L-R-L or L-R-L-R-L-R. The time of contraction was 5 s and the rest was 1 min between trials. ⁷⁰

The command was given in a standardised verbal form so as not to interfere with the reproducibility of the measurement. When the participant was ready, the assessor gave verbal instructions according to Amaral, Mancini and Novo Júnior ¹⁹ : ‘one, two, three, NOW! PUSH, PUSH, PUSH… that’s it… relax!’.

Each participant had 5 min of rest between each of the strength assessment instruments. At the end of the collection, the retest was scheduled, respecting the predetermined time interval of 7 days. The test phase was performed in approximately 2 h and the retest in approximately 1h30 (Figure 2).

Figure 2.

Methodological flowchart.

Ethics

The study was submitted to the Research Ethics Committee of the Hospital das Clínicas of the Faculty of Medicine of Ribeirão Preto, University of São Paulo (CAAE 86927718.3.0000.5440).

Data analysis

After collecting the data from all the volunteers on the 3 dynamometers, the strength test means were calculated for each hand, according to the recommendations of the American Society of Hand Therapists (ASHT).^35,62,71

Initially, the values for the Biodex were converted from Newton metres (N.m) to Kilogram force (Kgf) following the formula. The distance perpendicular to the axis of movement (DPAM) was known and had a value of 0.235 m. ⁷²

$$Torque\hspace{0.17em}\left(Nm\right)=Force\hspace{0.17em}\left(N\right)\hspace{0.17em}*\hspace{0.17em}DPAM\hspace{0.17em}\left(m\right)$$

$$Torque\hspace{0.17em}\left(Nm\right)=Force\hspace{0.17em}\left(N\right)\hspace{0.17em}*\hspace{0.17em}0.235m$$

In our analysis, we first assessed the normality of the data using Q-Q plots (Appendix 1) and the Shapiro-Wilk test (Appendix 1), a widely used method to determine if a sample follows a normal distribution. A significant p-value (p < .05) indicates non-normality, while a non-significant p-value (p ≥ .05) suggests normality. ⁷³

Since the data did not meet the assumption of normality, we applied the Wilcoxon Signed-Rank Test (Appendix 1), a non-parametric alternative to the paired t test, to compare the paired samples. ⁷⁴ This test evaluates whether the median difference between pairs is significantly different from zero.

The Standard Error of Measurement (SEM) was calculated to estimate the precision of individual scores and assess the variability in measurements due to inherent error, using the ICC data and the following formula. ⁷⁵

$$SEM=SD*\sqrt{1-ICC}$$

The Minimum Detectable Change (MDC) was used to assess the agreement between the Jamar, Jamar Plus and Biodex dynamometers; it can be calculated using the formula below, in which 1.96 is corresponding to a 95% confidence interval. ⁷⁵

$$MDC=1.96*\sqrt{2}*SEM$$

The test-retest reliability analysis was performed using the Intraclass Correlation Index (ICC) determination at a 95% confident interval. Reliability between the dynamometers was checked using the mean value in kilograms of force (Kgf) for each individual on the dominant and non-dominant hands.

Values were interpreted based on the Koo et al. ⁷⁶ classification, in which an ICC based on the 95% confident interval of the ICC estimate, values less than 0.50 - poor, between 0.50 and 0.75 - moderate, between 0.75 and 0.90 - good, and greater than 0.90 - excellent reliability.

At last, the Bland-Altman plot was proposed to assess the agreement between the Jamar, Jamar Plus and Biodex scores. A dispersion graph makes it possible to evaluate the difference and the average of two variables. This chart allows the visualisation of the error (the dispersion of the points of the differences around the mean) and the outliers and trends. ⁷⁷

Results

Characterisation of volunteers

The data collected from 35 volunteers were analysed (Table 3), of which 22 (62.9%) are female, and 13 (37.1%) are male, aged between 19 and 38 years, with a measured age of 23.4 (±4.2). Most were students (26%–74.3%) and physiotherapists (4%–11.4%).

Table 3.

Sample characterisation.

Variable	n = 35	Frequency (n)	Percentage (%)
Sex	Female	22	62,9
	Male	13	37,1
Laterality (GLPI)	Moderate left-handed	1	2,9
	Moderate right-handed	22	62,8
	Strongly right-handed	12	34,3
Occupation	Student	26	74,3
	Physiotherapist	4	11,4
	Bartender	1	2,8
	Bioterium intern	1	2,8
	Administrator	1	2,8
	Nurse	1	2,8
	Social worker	1	2,8

During the verification of laterality, the GLPI-B classified 34 of the volunteers as right-handed, of which 22 were moderate 62.8%, 12 were strongly 34.3%, and only 1 (2.9%) were left-handed.

Mean of dynamometry measurements for test-retest (see supplemental material - Table 8 and 9.)

The mean found on the dynamometer for the three instruments, obtained from the test and retest, show that the mean grip strength with the Jamar dynamometer in Kgf was higher than the others obtained with the Jamar Plus and Biodex (Table 4).

Table 4.

Mean of the grip forces with the Jamar, Jamar Plus dynamometers, and for the Biodex in the test and retest.

Variable		SEM	MDC	All volunteers (n = 35)*
				Dominant hand	Non-dominant hand	Mean difference
				Mean Kgf ± SD	Mean Kgf ± SD	Kgf
Test	Jamar (Kgf)	0.335	0.929	36.03 ± 11.23	33.44 ± 11.11	2.59
	Jamar plus (Kgf)	1.364	3.781	34.43 ± 11.85	31.01 ± 10.45	3.42
	Biodex (Kgf)	0.582	1.614	28.05 ± 10.13	27.25 ± 10.35	0.80
Retest	Jamar (Kgf)	0.185	0.513	35.48 ± 10.42	33.50 ± 11.10	1.98
	Jamar plus (Kgf)	1.497	4.149	33.71 ± 11.48	30.90 ± 10.88	2.81
	Biodex (Kgf)	0.899	2.493	27.90 ± 11.40	27.70 ± 09.97	0.20

*Of all the volunteers, 34 had a right-sided preference and only one had a left-sided preference. Kgf - kilogram force; SD - Standard Deviation; SEM - Standard Error of Measurement; MDC - Minimal Detectable Change.

The differences in mean values between the dominant and non-dominant sides for Jamar were 2.59 Kgf, and in relation to the test and retest they were 0.55 Kgf more for the dominant side and −0.06 Kgf for the non-dominant side.

For Jamar Plus, the mean values between the dominant and non-dominant sides were 3.42 Kgf, while the average values for the test and retest were 2.42 Kgf more for the dominant side and 2.81 Kgf for the non-dominant side.

Finally, the mean values between the dominant and non-dominant sides of the Biodex were 0.80 in the test and 0.20 in the retest, the mean values between test and retest were 0.20 Kgf more for the dominant side and −0.40 Kgf for the non-dominant side (Table 4).

Analysing all the volunteers, Jamar’s SEM values (0.335 in the test and 0.185 in the retest) indicate a better performance, with predictions closer to the real values, presenting smaller errors. Jamar Plus, on the other hand, obtained a higher SEM value (in the test 1.364 in the test and 1.497 in the retest), with significantly worse performance, suggesting that its predictions are further away from the real values.

When analysing the MDC values of all the volunteers, Jamar showed the smallest difference between the tests (in the test 0.929 and in the retest 0.513), suggesting that the changes between measurements are sufficiently small and therefore well detected, with a lower chance of random errors impacting the measurement.

The Jamar Plus, on the other hand, showed the highest MDC values (in the test 3.781 and in the retest 4.149), indicating that smaller changes were not well detected, which may suggest lower instrument sensitivity.

This data can be understood due to the standard of collection, that of the Jamar, since it is done through the human eye, which can lead to parallax errors when analysing the movements of the needle, and is therefore not as precise as the Jamar Plus, which is digital. ⁷⁸

Finally, the SEM value for Biodex (0.582 in the test and 0.899 in the retest) indicates moderate performance, and the MDC values for Biodex (at test 1.614 and retest 2.493) were between those of Jamar and Jamar Plus, showing intermediate performance.

The mean by sex were also considered; Jamar values were the highest, the results showing higher mean strength values for males (Table 5) compared to females (Table 6).

Table 5.

Mean of the grip forces with the Jamar, Jamar Plus dynamometers, and for the Biodex in the test and retest considering male volunteers.

Variable		Male (n = 13)*
		Dominant hand			Non-dominant hand			Mean difference
		Mean Kgf ± SD	SEM	MDC	Mean Kgf ± SD	SEM	MDC	Kgf
Test	Jamar (Kgf)	48.25 ± 07.46	0.775	2.148	45.48 ± 07.95	1.420	3.936	2.77
	Jamar plus (Kgf)	47.55 ± 08.70	3.095	8.579	42.18 ± 07.42	0.732	2.030	5.37
	Biodex (Kgf)	38.80 ± 06.38	0.977	2.708	38.10 ± 06.69	0.977	2.708	0.70
Retest	Jamar (Kgf)	47.12 ± 06.98	0.190	0.527	45.69 ± 07.91	1.165	3.229	1.43
	Jamar plus (Kgf)	46.18 ± 08.53	4.421	12.255	42.75 ± 07.62	0.468	1.298	3.43
	Biodex (Kgf)	38.70 ± 09.00	2.194	6.080	37.10 ± 06.23	2.194	6.080	1.60

*All male volunteers had a right-sided preference. Kgf - kilogram force; SD - Standard Deviation; SEM - Standard Error of Measurement; MDC - Minimal Detectable Change.

Table 6.

Mean of the grip forces with the Jamar, Jamar Plus dynamometers, and for the Biodex in the test and retest considering female volunteers.

Variable		Female (n = 22)*
		Dominant hand			Non-dominant hand			Mean difference
		Mean Kgf ± SD	SEM	MDC	Mean Kgf ± SD	SEM	MDC	Kgf
Test	Jamar (Kgf)	28.81 ± 05.05	0.580	1.607	26.33 ± 04.58	1.150	3.189	2.48
	Jamar plus (Kgf)	26.69 ± 03.89	2.302	6.380	24.42 ± 04.78	0.731	2.026	2.27
	Biodex (Kgf)	21.70 ± 05.37	1.167	3.234	20.90 ± 05.68	1.167	3.234	0.80
Retest	Jamar (Kgf)	28.60 ± 03.77	0.616	1.707	26.30 ± 04.18	1.417	3.927	2.30
	Jamar plus (Kgf)	26.35 ± 04.34	1.388	3.849	23.77 ± 04.27	0.583	1.617	2.58
	Biodex (Kgf)	21.50 ± 07.10	0.954	2.643	22.10 ± 07.16	0.954	2.643	−0.60

*Of all the female volunteers, 21 had a right-sided preference and one had a left-sided preference. Kgf - kilogram force; SD - Standard Deviation, SEM - Standard Error of Measurement; MDC - Minimal Detectable Change.

In the Jamar test, males demonstrated a mean grip strength 19.44 kgf higher than females on the dominant side and 19.15 kgf higher on the non-dominant side, as measured by the Jamar dynamometer. Conversely, the Jamar Plus dynamometer recorded mean kgf values 20.86 and 17.46 higher for the dominant and non-dominant sides, respectively. For the Biodex equipment, the mean differences were 17.10 kgf for the dominant side and 17.50 kgf for the non-dominant side (Table 5).

In the retest, the data remained higher for males than for females. Using the Jamar dynamometer, the average difference in handgrip strength was 18.52 kgf on the dominant side and 19.39 kgf on the non-dominant side. The Jamar Plus dynamometer recorded mean kgf values 19.83 and 18.98 higher for the dominant and non-dominant sides, respectively. Finally, the Biodex equipment showed mean differences of 17.20 kgf and 15.00 kgf for the dominant and non-dominant sides, respectively (Table 6).

When analysing the test and retest SEM (Standard Error of Measurement) and MDC (Minimal Detectable Change) results for the dominant hand of males and females, the Jamar dynamometer exhibited the lowest SEM and MDC values. This indicates greater accuracy in its predictions, with measurements closer to the real values and a relatively low likelihood of random errors. In contrast, the Jamar Plus dynamometer had the highest SEM and MDC values, suggesting significantly poorer performance and a greater deviation from the real values.

The Biodex equipment showed intermediate performance in the test and retest of the dominant hand, with a SEM indicating moderate performance and a MDC between the two other instruments.

For the non-dominant hand of males and females, the Jamar Plus dynamometer showed a lower SEM, indicating relatively accurate measurements and good predictive performance. However, in the test, the Jamar dynamometer recorded a lower error and MDC, suggesting a reduced likelihood of random errors affecting the SEM and MDC values for males. In the retest, the Biodex equipment exhibited higher SEM and MDC values.

For females, the Jamar dynamometer had higher SEM and MDC values in both the test and retest, indicating poorer performance and predictions further from the real values. This result indicates a contrast with the anatomy of women’s hands, known to be generally smaller than men’s, which may have made it difficult to carry out the tests in the 3rd position on the dynamometers.^79,80

Finally, for the non-dominant hand test, the Biodex equipment showed intermediate SEM and MDC values for males, while the Jamar dynamometer did so for females. In the retest, the Jamar dynamometer exhibited intermediate SEM and MDC values for males, and the Biodex equipment did so for females. In both the test and retest, the SEM values indicated moderate performance, and the MDC values fell between those of the other two instruments.

In summary, the Jamar dynamometer was the most precise, with lower error values for males' dominant hands and females' non-dominant hands, although with some variability. The Jamar Plus dynamometer had the highest error values, particularly for the dominant hand. The Biodex equipment performed moderately, serving as an intermediate option for both males and females.

Thus, our results suggest some decisions in the evaluation of HGS in clinical practice, considering the evaluation variables (device, price, calibration, training, adjustments and positioning), the hierarchy of choice of dynamometers would be to first use the Jamar, then the Jamar Plus and finally the Biodex.

Reliability

Intraclass correlation coefficient

In the test-retest reliability between Jamar and Jamar Plus, excellent reliability was observed in both the test (ICC 0.95 dominant; 0.93 non-dominant) and retest (ICC 0.94 and 0.91, respectively),through the intraclass correlation coefficient (ICC_2.1),. The SEM ranged from 1.75 to 3.12 kgf, with MDC between 4.86 and 8.66 kgf, indicating high accuracy and sensitivity in detecting real changes. These results reinforce the possibility of using the two dynamometers interchangeably in both hands (Table 7).

Table 7.

ICC_2,1 and LoA values between Jamar, Jamar Plus and Biodex was moderate-excellent in the test and retest for both hands.

Dynamometers	Dominant hand				Non-dominant hand
	ICC	SEM	MDC	LoA (%)	ICC	SEM	MDC	LoA (%)
Jamar x Jamar plus test	0.95	1.75	4.86	−12.95/+23.44	0.93	1.98	05.51	−09.27/+24.71
Jamar x Jamar plus retest	0.94	2.74	7.59	−12.60/+25.23	0.91	3.12	08.66	−11.21/+29.22
Jamarx Biodex test	0.69	4.18	11.52	−09.34/+62.52	0.78	3.52	09.76	−05.08/+49.23
Jamar x Biodex retest	0.70	4.23	11.73	−18.13/+73.71	0.77	3.55	09.86	−08.77/+48.98
Jamar plus x Biodex test	0.75	3.89	10.80	−15.83/+58.72	0.89	2.42	06.73	−09.52/+38.40
Jamar Plusx Biodex retest	0.77	3.83	10.62	−20.38/+63.86	0.81	3.13	08.68	−20.83/+43.12

ICC: Intraclass Correlation Interval, SEM: Standard Error of Measurement; MDC: Minimal Detectable Change, LoA (%): Percentage of the Limits of Agreement (- Lower/+ Upper).

On the other hand, reliability between the Jamar and Biodex was moderate to good, with ICCs between 0.69 and 0.70 (test and retest, dominant hand) and 0.77 to 0.78 (non-dominant). SEM (3.52 - 4.23 kgf) and MDC (9.76 - 11.73 kgf) values were higher, suggesting lower precision and greater variability in measurements, especially in the dominant hand. This limits direct substitution between these instruments.

Finally, between the Jamar Plus and Biodex, reliability was good in the test (ICC 0.75 dominant; 0.89 non-dominant) and retest (ICC 0.77 and 0.81, respectively). The SEM (2.42 - 3.89 kgf) and MDC (6.73 - 10.80 kgf) values were lower compared to the previous comparison (Jamar vs Biodex), especially in the non-dominant hand, showing greater alignment between the devices and making it viable to use them as an alternative to Biodex in clinical contexts.

In summary, the data show that the Jamar and Jamar Plus dynamometers have high reliability and low variability, which makes them the preferred option for assessing manual strength in clinical practice and valid for the population of healthy young adults. It is understood that accuracy and the ability to detect real changes are essential, so these devices can be used interchangeably.

On the other hand, the Biodex showed greater variability, limiting its direct replacement. In short, as previously suggested, the initial use of the Jamar or Jamar Plus dynamometers for routine assessments, reserving the Biodex for situations requiring complementary or specialised assessments.

Limits of agreement: Bland-Altman plots

The Bland-Altman analysis is widely used to assess whether two methods of measurement exhibit an acceptable degree of disagreement. ⁷⁷ The plots display the differences between measurements obtained from the three dynamometers during the test (Figure 3) and retest (Figure 4).

Figure 3.

Bland & Altman graphs: agreement between Jamar, Jamar Plus and Biodex considering the dominant and non-dominant side during the test. a) Bland-Altman analysis with Jamar vs Biodex - Dominant Hand; b) Bland-Altman analysis with Jamar vs Biodex - Non-Dominant Hand; c) Bland-Altman analysis with Jamar Plus vs Biodex - Dominant Hand; d) Bland-Altman analysis with Jamar Plus vs Biodex - Non-Dominant Hand; e) Bland-Altman analysis with Jamar vs Jamar Plus - Dominant Hand; f) Bland-Altman analysis with Jamar vs Jamar Plus - Non-Dominant Hand.

Figure 4.

Bland & Altman graphs: agreement between Jamar, Jamar Plus and Biodex considering the dominant and non-dominant side during the retest. a) Bland-Altman analysis with Jamar vs Biodex - Dominant Hand; b) Bland-Altman analysis with Jamar vs Biodex - Non-Dominant Hand; c) Bland-Altman analysis with Jamar Plus vs Biodex - Dominant Hand; d) Bland-Altman analysis with Jamar Plus vs Biodex - Non-Dominant Hand; e) Bland-Altman analysis with Jamar vs Jamar Plus - Dominant Hand; f) Bland-Altman analysis with Jamar vs Jamar Plus - Non-Dominant Hand.

The Bland–Altman analysis conducted during the initial test indicated good agreement between the Jamar and Jamar Plus dynamometers, with narrower limits of agreement (e.g., −12.95% to +23.44% for the dominant hand) and high ICC values (0.93–0.95), suggesting that these devices produce consistent measurements.

In contrast, the comparison between Jamar and Biodex showed poor agreement, with wide limits of agreement (e.g., up to +62.52% for the dominant hand) and lower ICCs (0.69–0.78), indicating significant variability between devices (Table 7). The comparison between Jamar Plus and Biodex demonstrated moderate agreement, with intermediate ICC values (0.75–0.89) and still relatively wide limits of agreement (e.g., up to +58.72%). These findings suggest that only the Jamar and Jamar Plus pair can be considered reliable for interchangeable use in the test setting.

The results from the retest followed the same pattern observed in the initial test, reinforcing the consistency of the findings. Agreement between the Jamar and Jamar Plus dynamometers remained high, with ICC values ranging from 0.91 to 0.94 and limits of agreement similar to those in the test (e.g., −12.60% to +25.23%).

Conversely, the retest comparison between Jamar and Biodex continued to show substantial variability, with similarly low ICCs (0.70–0.77) and very wide limits of agreement (reaching up to +73.71%), reaffirming that these devices should not be used interchangeably. The comparison between Jamar Plus and Biodex again showed moderate agreement, with ICCs between 0.77 and 0.81 and limits of agreement extending up to +63.86%. As with the test results, the retest findings confirm that only the Jamar and Jamar Plus dynamometers demonstrated sufficiently narrow agreement to be considered interchangeable (Table 7).

Discussion

It is well established that handgrip strength is influenced by a range of individual factors, including age, sex, hand shape, and asymmetry due to hand dominance. Other contributing elements may include lifespan, life stage, and occupational demands.^81,82 In light of these variables, this study was conducted in alignment with the guidelines proposed by Fess ⁸³ for the evaluation of measurement instruments.

The sample consisted predominantly of female, right-handed volunteers. The results demonstrated excellent agreement between the Jamar and Jamar Plus dynamometers for both the dominant and non-dominant hands across test and retest phases. This is in accordance with the literature, which recognises the Jamar as a gold standard instrument, known for its high inter-device reliability and precision, particularly when compared to other hand-held dynamometers, such as the NOD. ⁸⁴

These findings are also consistent with existing studies reporting moderate-to-excellent agreement between the Jamar Plus and other devices, including the Jamar, ³⁸ GripAble, ⁴⁷ Camry EH101, ⁸⁵ and Bodygrip. ⁸⁶ However, to date, no inter-instrument reliability studies have explored comparisons involving force assessment using the Biodex in upper limb rehabilitation.

One previous study assessing inter-instrument reliability between the Jamar and Biodex in patients with advanced cancer found that the Jamar demonstrated greater consistency over time, reinforcing its suitability as a preferred tool for monitoring strength progression. ⁵⁴ Accordingly, our findings partially support the initial hypothesis, demonstrating moderate to good agreement between the dynamometers, particularly between the Jamar and Jamar Plus.

The Bland–Altman analysis confirms good agreement between the Jamar and Jamar Plus dynamometers, while only moderate-to-poor agreement was observed when either device was compared to the Biodex dynamometer. The stronger agreement observed on the non-dominant side may reflect reduced variability in force application, potentially due to less dominance-related asymmetry during grip performance.

While the Jamar and Jamar Plus dynamometers demonstrated consistent reliability, the Biodex exhibited weaker agreement, likely due to differences in measurement principles, hand positioning, device configuration and testing protocols.^54,86

Our findings suggest that the Jamar and Jamar Plus can be considered interchangeable for clinical use. However, the Biodex dynamometer may require additional considerations, including its larger physical footprint (requiring significant storage space), higher cost (potentially limiting accessibility for many clinicians, who may favour more affordable handheld devices), and the need for specialised training to ensure proper positioning, calibration, and use.

Although one of the initial hypotheses of this study was that all three instruments would demonstrate good reliability and agreement, the results did not confirm this assumption, particularly in comparisons involving the Biodex. This underscores the need for further research to better understand the sources of variability and disagreement between different dynamometers. Future studies should explore the impact of unit conversion, device-specific protocols, and testing standardisation, particularly when including isokinetic devices like the Biodex. ⁵⁴

Finally, further research is warranted to examine the influence of economic conditions and accessibility on the availability and use of such equipment in routine assessments. Despite the variability observed, the present study supports the use of the Jamar and Jamar Plus as reliable tools, and highlights the need for caution when interpreting or comparing grip strength data obtained from different types of dynamometers, especially when including the Biodex.

Limitations

One limitation of this study concerns the standardisation of handle spacing across dynamometers. Given that each device differs in its structure, we opted to standardise the handle distance in centimetres rather than by slot position. This approach was chosen particularly because the Biodex offers greater flexibility in handle adjustment compared to the more fixed positions of the handheld dynamometers.

Additionally, while we standardised the dynamometer position for all participants, measurements may still have been influenced by hand size differences between sexes, potentially introducing systematic error or parallax distortion during data collection.

Another limitation involves the decision to use the third handle position across all devices. Although this ensured consistency in handle distance, it diverges from the commonly used second position, which is often considered more anatomically appropriate and biomechanically optimal. This could have affected grip performance and test outcomes.

The conversion of force units from N metres (Nm) to kilograms-force (kgf) may also be considered a limitation in clinical settings. In this study, we addressed this by applying a validated formula, but this may pose a barrier to broader clinical application without automated tools.

Finally, Bland–Altman analysis, while appropriate for evaluating agreement, is less robust in studies with small sample sizes and limited variability. This may lead to overestimation or underestimation of agreement levels. For this reason, future studies with larger and more diverse samples are recommended.

Implications for practice

• -While the Jamar and Jamar Plus demonstrated good reproducibility, the Biodex did not show the same level of consistency and sensitivity. Clinicians should exercise caution when selecting the appropriate device, particularly when aiming for reproducible measurements over time.

• -These findings underscore the importance of standardised testing procedures and highlight the need for ongoing training, positioning accuracy, calibration, and device adjustment to ensure valid and consistent measurements across different instruments.

• -The study reinforces the clinical value of reproducibility studies, enabling hand therapists to make informed decisions about reliable and interchangeable instruments while maintaining a focus on standardisation and precision.

• -Although the Biodex showed weaker agreement in this study, its use in grip strength analysis should depend on the clinician’s clinical judgment and the specific requirements of the rehabilitation setting. Clinicians should be aware of its limitations and consider the necessary adjustments when incorporating it into practice.

Conclusions

Assessing measurement properties, such as test–retest reliability and inter-instrument agreement, is essential when comparing different grip strength devices. Based on our findings, the Jamar and Jamar Plus dynamometers can be used interchangeably, while the Biodex may be utilised with caution, as its agreement and reproducibility were weaker compared to the handheld devices. Appropriate adjustments, such as handle positioning and unit conversion, are critical when considering the Biodex for clinical use.

Furthermore, the variability observed with the Biodex underscores the need for careful consideration and further standardisation in its use.

Finally, the findings point to a promising future for the integration of technologically advanced instruments in rehabilitation. The study opens avenues for future research aimed at improving the consistency of grip strength measurements and exploring the role of different dynamometers in clinical practice, thus contributing to the ongoing development of more precise and accessible assessment tools.

Appendix 1.

Table 8.

Shapiro-Wilk test.

	Variable	W	p value
Test	MJDH	0.9190626	0.0133979
	MJNDH	0.8972364	0.0033352
	MJPDH	0.8685675	0.0006266
	MJPNDH	0.8958309	0.0030605
	MBDH	0.9447844	0.0783613
	MBNDH	0.879473	0.0011607
Retest	MJDH	0.8902208	0.002181
	MJNDH	0.8840993	0.0015184
	MJPDH	0.8844226	0.0015474
	MJPNDH	0.8944451	0.0028131
	MBDH	0.9434416	0.071278
	MBNDH	0.8926397	0.002522

MJDH - Mean JAMAR Dominant Hand; MJNDH- Mean JAMAR Non-Dominant Hand; MJPDH - Mean JAMAR Plus Dominant Hand; MJPNDH - Mean JAMAR Plus Non-Dominant Hand; MBDH - Mean BIODEX Dominant Hand; MBNDH - Mean BIODEX Non-Dominant Hand.

Table 9.

Wilcoxon Signed-Rank test.

	Comparison		Statistic	p value
Test	MBDH	MJPDH	8	<0.001
	MBDH	MJPDH	35	<0.001
	MBNDH	MJNDH	10	<0.001
	MBNDH	MJPNDH	40	<0.001
	MJPDH	MJPDH	468.5	<0.05
	MJNDH	MJPNDH	566	<0.001
Retest	MBDH	MJPDH	25	<0.001
	MBDH	MJPDH	53	<0.001
	MBNDH	MJNDH	50	<0.001
	MBNDH	MJPNDH	104	<0.001
	MJPDH	MJPDH	509.5	<0.001
	MJNDH	MJPNDH	592	<0.001

MJDH - Mean JAMAR Dominant Hand; MJNDH- Mean JAMAR Non-Dominant Hand; MJPDH - Mean JAMAR Plus Dominant Hand; MJPNDH - Mean JAMAR Plus Non-Dominant Hand; MBDH - Mean BIODEX Dominant Hand; MBNDH - Mean BIODEX Non-Dominant Hand.

ORCID iDs

Gabriel Morais Xavier dos Santos https://orcid.org/0000-0001-5923-5742

Leonardo Dutra de Salvo Mauad https://orcid.org/0000-0003-1914-6352

Heloísa Corrêa Bueno Nardim https://orcid.org/0000-0001-9110-2317

Flávia Pessoni Faleiros Macedo https://orcid.org/0000-0002-4130-6853

Gabriela Rezende https://orcid.org/0000-0002-1355-3945

Raquel Metzker Mendes Sugano https://orcid.org/0000-0003-4585-1709

Elaine Caldeira de Oliveira Guirro https://orcid.org/0000-0002-1045-8053

Marisa de Cássia Registro Fonseca https://orcid.org/0000-0001-8187-5834