Background: Hand grip force is a good indicator of general muscle strength and can also be used to predict multiple outcomes such as changes in activities of daily living (ADL), disability, mortality and general upper extremity strength. Hand grip force is often measured as the amount of static grip force a subject can produce when measured with a hydraulic dynamometer such as the Jamar or with an electronic device such as the Grippit. The Grippit device measures an average grip force, a peak grip force and force over a set time period. Grippit has shown good reliability for healthy subjects. Grippit, which was developed over 20 years ago in Gothenburg, Sweden is no longer manufactured. Therefore, the need for anewly developed and modernized measurement instrument for use in evaluating hand rehabilitation has arisen.
Objectives: The aim of this study was to evaluate the test-retest reliability of the newly developed instrument GRIP-it and to describe and validate the relationship between grip force measurements from GRIP-it and the original Grippit device.
Methods: Healthy controls (n=43) were included in the study. Two devices were used to evaluate grip force (Newton, N), (i) GRIP-it a newly developed device and (ii) Grippit. Both instruments were used to measure mean and maximal force over 10 seconds.
Results: GRIP-it displayed a mean measurement error of -1.7 ± 0.5% and the corresponding error for Grippit was -1.6 ± 1.9%. All subjects completed the grip force tests and the results for three attempts for each hand. The test-retest reliability was excellent for both pieces of equipment, with ICCs ranging from 0.963 to 0.947 (CI 95% between 0.103 and 0.041) for GRIP-it and from 0.979 to 0.968 (CI 95% between 0.087 and 0.042) for Grippit.
Relationships between Grippit and GRIP-it
There was a significant difference between the measured values derived from Grippit and GRIP-it for both the dominant hand (P < 0.001) and the non-dominant hand (P < 0.01). Grippit gives in general a higher grip force measurement than GRIP-it which is also indicated by the slope (β1) of the regression lines that deviates from 1. However, there were no substantial differences in the grip force when comparing the measurements for the dominant hand with the non-dominant hand for either Grippit (P = 0.071) or GRIP-it (P = 0.404). Based on these non-significant differences between hands and the fact that the model estimates for the intercept (β0) and the slope (β1) are contained within the confidence intervals of the model estimates for the opposite hand, a combined model was derived. The linear regression analysis, with grip force measurements for both hands included, gives: GRIP-it = 49.0 + 0.779 · Grippit. This explains 89.6% of the variance in grip force analyzed by GRIP-it (P < 0.001) see Figure 2. To enable an estimation of grip force measured by Grippit based on GRIP-it values a regression analysis with Grippit as dependent variable gives: Grippit = -18.1 + 1.15 · GRIP-it, which explains 89.6% of the variance in grip force analyzed by Grippit (P < 0.001).
Conclusions: This study showed that GRIP-it has excellent test-retest reliability. Measurements of grip force with GRIP-it are strongly related to those from the original Grippit. The newly developed GRIP-it shows great potential for use in the assessment of hand function and the evaluation of hand rehabilitation.