Authors: Yuko Shimizu, Shigeo Tsutsui, Munenori Katoh
Categories: Original Article, Ankle plantar flexion, Hand-held dynamometer, Intraclass correlation coefficients
Source: Journal of Physical Therapy Science
Doi: 10.1589/jpts.35.619
[Purpose] A hand-held dynamometer (HHD) is less expensive than the isokinetic muscle strength measurement device, and research using HHD is gradually increasing. However, measurement is performed only at a low muscle strength level at which the heel does not take off or heel detachment occurs; therefore, fixation of the foot becomes a problem. This study aimed to determine the validation of measuring ankle plantar flexion strength (with the knee extended) using HHD. [Participants and Methods] Twenty healthy adults (14 males and 6 females) participated in this study. The chair used in this study was for swallowing videofluorography, which was fixed to a wall bar by the belt. The sensor was located at the third metatarsal head. After warming up, the participants sat in a long sitting position on the chair. We conducted the test two times. We used intraclass correlation coefficient (ICC) and Bland–Altman analysis to assess reliability. [Results] The ICC(1, 1) and ICC(2, 1) results were all greater than 0.9. No fixed and proportional errors were present. [Conclusion] The measurement method of this study was both intra- and inter-examiner reliabilities, which were high, and we suggest that sufficient clinical application is possible.
Keywords: Hand-held dynamometer, Ankle plantar flexion, Intraclass correlation coefficients
The triceps surae muscle of the lower leg, which is an ankle plantar flexor, is active when kicking out during walking^1^^)^. Therefore, it is an important muscle in walking in terms of energy efficiency and prevention of falls^1^^)^. In addition, since it is necessary for various movements in sports, it can be used as a standard index for the effect of rehabilitation and performance improvement by measuring the maximum muscle strength of ankle joint plantar flexion^2^^)^. Manual Muscle Testing (MMT) is generally used to measure the triceps surae muscle strength. MMT can be performed without the need for tools, but a score of 3 or above is evaluated based on the number of times a person can bend down against their weights^2^^)^. Therefore, it is believed that the endurance of the muscle, rather than the maximum muscle strength, is reflected in the evaluation. In addition, a one-legged position is required for a measurement score of 3 or more, and the evaluation will be 2+ or less regardless of the actual muscle strength when maintaining such a position is difficult^3^^)^.
Other measurement methods require instruments such as an isokinetic muscle strength measurement device or a Hand-Held Dynamometer (HHD). Using these instruments, maximum muscle strength can be measured numerically, and such changes can be evaluated objectively. It is also possible to exclude the influence of body weight by calculating the Body Weight Ratio (BWR). However, an isokinetic muscle strength measurement device is expensive and requires a large space. Consequently, facilities and hospitals that own such equipment are limited. HHD is less expensive than the isokinetic muscle strength measurement device, and research using HHD is gradually increasing^4^^,^^5^^,^^6^^,^^7^^)^. However, using HHD to measure muscle strength of ankle plan fixation of the foot becomes a problem^6^^, ^^7^^)^. The heel separation caused by the inability to immobilize the foot prevents accurate measurement. We hypothesized that is related to the reliability of the measurement. Therefore, in this study, focusing on the fixation of the foot, which was a problem in the previous studies, we proposed a method for measuring the ankle plantar flexion force using the HHD, and the reliability of the measurement method was examined.
The participants were 20 healthy adults (14 males and 6 females) with a mean age of 31.9 ± 7.6 years, a mean height of 166.2 ± 7.2 cm, and a mean weight of 62.2 ± 11.2 kg. The participants did not have cardiopulmonary dysfunction or functional impairment of the lower limbs. The examiners explained the study to the participants prior to measurement and obtained their consent. This study was approved by the Ethical Review Committee of the hospital (approval 022-002).
In this study, we proposed an evaluation method for measuring the ankle plantar flexion using the HHD, and decided to examine the reliability of this evaluation method. Reliability was examined using relative reliability (intra- and inter-rater reliability) and absolute reliability.
Isometric plantar flexor strength was measured in the knee extension position using the HHD. The measurement was performed by fixing the posture using a belt and chair. The measured limb was the dominant limb (18 right and 2 left), and the dominant foot was judged to be the foot used to kick the ball. μTas F-200 (Anima Corp., Tokyo, Japan) was used as the HHD. The upper limit of μTas F-200 used for the measurements was 200 kg. The chair used was Conver VFX (Takano Co., Ltd, Nagano, Japan), which is used for swallowing videofluorography. The height of the chair was adjusted to a horizontal position by raising the foot support to a horizontal position, and then placing a pad and wrapping a towel to eliminate the height difference between the seat and foot support. Four ratchet lashing belts (Thousandshores Technology Co., Ltd, Tokyo, Japan) were used to fix the chair with a width of 38 mm, maximum load of 500 kg, and breaking load of 1,490 kg. Furthermore, in order to fix the foot and to avoid pain, a 25 mm wide cloth strap was used.
As a warm-up, the participants performed the standing heel-raise ten times. After warming up, the participants sat in a long sitting position on a chair grounded to the vertical pillar of a wall bar. The examiner adjusted the distance between the pillar of the wall bar and the chair such that the sole of the tested limb touched the vertical pillar. At this time, the participant’s sacrum was pressed as much as possible against the backrest of the chair. The examiner fixed the HHD sensor to the pillar of the wall bar using hook-and-loop fasteners. The center, of the HHD sensor covered with a silicone pad, was positioned at the same height as the third metatarsal head. After fixing the sensor to the supporting pillar, the examiner inserted a piece of wood of the same height as the sensor under the heel and made the measurement surface parallel to the sensor. The examiner then fixed the chair using three belts to prevent the chair from moving during measurement.
In order to fix the chair, the belt was positioned so that the backrest of the chair and the wall bar were connected in one place, the chair frame and the wall bar were connected in two places, and the belt was adjusted so that the belt did not loosen as much as possible.
Similarly, the examiner placed the participant’s lower limb on the foot support of a horizontally set chair with the participant’s knee joint extended and the thigh fixed using a belt. It was fixed with a string such that the heel did not leave the wall bar. The foot was fixed from the upper part of the heel to the lower part of the inner and outer malleoli, crossed at the dorsum of the foot, tied once on the back of the supporting pillar, passed through the inside of the string around the ankle, folded back, and tied again on the back of the pillar (Fig. 1). The participant’s unmeasured lower limb was lowered from the chair to prevent force, and the arms were crossed in front of the chest (Fig. 2). After the preparation for the measurement was completed, the examiner explained the measurement method to the participants, including how to apply the force. The explanations consisted of ankle plantar flexion without misaligned foot movements or toe flexion and isometric contraction to reach maximum effort in approximately 3 s and maintain it for up to 5 s. After that, the participant initially exercised at approximately 70% and then at 100% muscle strength capacity to familiarize themselves with the test. If the tested heel left the supporting pillar, the position of the chair and the way the belt was tightened were adjusted again.
Fig. 1. The method of fixing a foot with string.Depending on the way the strap was tied, the examinee sometimes complained of pain during plantar flexion of the ankle joint. In this case, the twisted part of the strap was fixed and the contact area was widened to eliminate the pain.
Fig. 2. Test position.The belt that connecting the backrest of the chair and the wall bar, the examiner adjusted the tightness of the belt in order to avoid pain in the participant’s hamstrings, and ensure that the chair kept its position.
After the examiner performed zero calibration of the HHD, the participant performed isometric plantar flexion for 5 s. The maximum muscle strength was measured during that time. The measured values were checked and recorded by assistants other than the examiner, who blinded the results. The same examiners performed two measurements at an interval of approximately 30 s. The examiners checked for the compensatory movements during the measurement, and if they were detected, the measurement was repeated. Examples of compensatory movements included rapid plantar flexion of the ankle joint, flexion of the toes, and pushing the trunk against a backrest.
These tests were performed by two physical Examiner A (19 years of experience) and Examiner B (5 years of experience). Measurements by examiners A and B were performed at intervals of one to three days, and the order of examiners A and B was random. Inter-rater reliability was examined using the largest of the two measurements. The BWR was calculated based on the maximum value of two measurements by Examiner A.
A Modified R Commander (4.0.2) was used as the analytical software, and the significance level was set at less than 5%. The distribution of normality was analyzed by the Shapiro–Wilk test. Intraclass Correlation Coefficients (ICC) and the Bland–Altman analysis (BAA) were used to assess reliability. Intra-examiner reliability was analyzed using the ICC(1,1), and inter-examiner reliability was analyzed using ICC(2,1). Prior to the study, a sample size design was conducted. When k=2, α=0.05, ρ=0.8, Δ=0.5 in ICC(2,1), resulted in n=20.
The average isometric plantar flexor strength was 658.7 N, 653.1 N, and 677.7 N for Examiner A, while 649.4 N, 670.1 N, and 674.7 N for Examiner B, in order of the first, second, and maximum values, respectively. Regarding intra-examiner reliability, the ICC(1,
Fig. 3. Bland–Altman plot (the first and second tests of Examiner A).(Left) The x-axis represents the average between the first and second tests of Examiner A, and the y-axis represents the bias (difference) between the first and second tests of Examiner A. (Right) The x-axis represents the average between the first and second tests of Examiner A, and the y-axis represents relative value. The solid blue line represents the mean of the difference between the two measurements, and the light black lines represent the upper and lower limits of agreement.
Fig. 4. Bland–Altman plot (the first and second tests of Examiner B).(Left) The x-axis represents the average between the first and second tests of Examiner B, and the y-axis represents the bias (difference) between the first and second tests of Examiner B. (Right) The x-axis represents the average between the first and second tests of Examiner B, and the y-axis represents relative value. The solid blue line represents the mean of the difference between the two measurements, and the light black lines represent the upper and lower limits of agreement.
Fig. 5. Bland–Altman plot (the maximum values of Examiners A and B).(Left) The x-axis represents the average between the maximum values of Examiners A and B, and the y-axis represents the bias (difference) between the maximum values of Examiners A and B. (Right) The x-axis represents the average between the maximum values of Examiners A and B, and the y-axis represents relative value. The solid blue line represents the mean of the difference between the two measurements, and the light black lines represent the upper and lower limits of agreement.
This study showed the high intra- and inter-reliability of the method for measuring ankle plantar flexion muscle strength during knee extension using an HHD with belt fixation. Based on these results, if the pelvis and ankle joints can be immobilized, the method of measuring ankle plantar flexion muscle strength by the HHD has both high intra- and inter-examiner reliability and is clinically applicable. This statement is supported by the ICC and BAA assessed in this study.
First, according to Landis et al.^8^^)^, 0.81–1.00 is considered almost perfect as the criterion for the kappa coefficient. When applied to Landis et al.’s standard ICC, the results of ICC(1, 1) and ICC(2, 1) were all 0.9 or higher, and the intra- and inter-examiner reliability was considered high, regardless of years of experience.
In addition, in the study of error using the BAA, there was no systematic error (fixed error or proportional error) between the first and second tests of Examiner A, the first and second tests of Examiner B, and the maximum values of Examiners A and B. The intra-examiner random error was approximately 15%, while the inter-examiner error was approximately 25%; therefore, when considering true value, those errors should have taken into account. However, the chance error was relatively large, approximately 15% among each examiner and 25% between the examiners. Therefore, it was considered necessary to judge the increase or decrease in muscle strength based on these errors.
A study showed that the BWR in the knee flexion position using an HHD with a belt and metal plate is 136.9% on average for men and 127.3% for women^9^^)^. The measured value in this study was lower than the ankle plantar flexion in the knee flexion position, as shown in previous studies. This may indicate that the average age of the participants could have been affected the results. In the previous study, the number of exercises was greater than in the present study because the participants performed heel raise on a two-legged position, heel raises on a one-legged position, and exerting muscles at 50%, 75%, 90%, and 100% prior to the study. Muscle strength measurements using the HHD require the participants to demonstrate maximum muscle strength over a period of three to five seconds, unlike normal muscle exertion. Therefore, the more times the participants in the measurement limb position, the more habituation may occur and the easier it may be for the participants to exert muscle strength.
The limitations of this study are that the participants were healthy adults with no cardiopulmonary or lower limb dysfunction, with a mean age of 31.9 ± 7.6 years. Therefore, it is unclear whether similar results can be obtained for older adults and adults with diseases. Second, when used in clinical practice, there is no comparison target, such as the average value for each age group at this stage. It is not possible to compare it with the average value for each age when used clinically. In the future, we would like to calculate the average value for each age group by increasing the target age.
This study examined the reliability of a method for measuring plantar flexor strength in the knee extension position using an HHD with belt fixation. Both intra- and inter-examiner reliabilities were high, and we suggest that sufficient clinical application is possible. However, the error in inter-examiner reliability was relatively large; therefore, it is necessary to consider the error when comparing the measured values by different examiners.
The data that support the findings of this study are openly available in OSF at http://doi.org/10.17605/OSF.IO/PGDNE.
All authors (1) made substantial contribution to the study concept or the date analysis or interpretation; (2) drafted the manuscript or revised it critically for important intellectual content; (3) approved the final version of the manuscript to be published; and (4) agreed to be accountable for all aspects of the work.
There is no conflict of interest to be disclosed concerning this study.
We thank Yusuke Nakagawa, Takahiro Sumino, Satoshi Shinohara for the acquisition of and maintaining the data. We are grateful to Dr. Hiroaki Yoshioka, Manami Ozaki for their continuous support and thoughtful guidance through the program.
The data that support the findings of this study are openly available in OSF at http://doi.org/10.17605/OSF.IO/PGDNE.