Abstract
Objective
The purpose of this study was to quantify the variability in the force applied during 20 cycles of Maitland's grade IV anteroposterior ankle mobilization measured on 2 occasions.
Methods
Thirteen healthy adults (mean age, 25 ± 5 years; height, 170 ± 7 cm; weight, 71 ± 16 kg) received 20 cycles of Maitland's grade IV ankle mobilization on 2 sessions separated by 1 week. A force transducer was used to measure the peak force, loading rate, and impulse applied during each load cycle. Mean within-session coefficient of variation, standard error of measurement, and 95% level of agreement were estimated during each mobilization session.
Results
The mean peak force during the anteroposterior mobilization technique was 70 ± 12 N and 58 ± 10 N during sessions 1 and 2, respectively. The mean within-session coefficients of variation in peak force, loading rate, and impulse applied during 20 loading cycles were 10% to 13%, 15%, and 21% to 43%, respectively. There was a significant difference between sessions in mean peak force (−17%, t12 = 2.445, P = .031) and impulse (−51%, t12 = 2.306, P = .040), with large 95% levels of agreement in applied peak force (±33 N) and impulse (±128 N s) compared to their mean values (approximately ±50% and 110%, respectively).
Conclusion
The peak force and loading rate applied by an experienced practitioner during a Maitland's grade IV anteroposterior talar mobilization session varied over 20 loading cycles. Variability between repeated mobilization sessions by the same practitioner was even greater, with respect to peak applied force and loading rate. The large variability in force applied during a Maitland's grade IV talar mobilization may underpin differential clinical effects reported in the joint-mobilization literature. The findings of this study highlight the need for strategies that standardize the application of force during talar mobilization.
Key Indexing Terms: Musculoskeletal Manipulations, Reproducibility of Results, Therapeutics
Introduction
Ankle-joint mobilization techniques are used by many health practitioners to treat musculoskeletal conditions of the lower limb, such as chronic ankle instability, ankle equinus, and plantar heel pain.1, 2, 3 These techniques are thought to reduce pain and improve the extensibility of soft tissues surrounding the ankle joint.4 A systematic review concerning the clinical benefits of joint mobilization on ankle sprains reported a high percentage of studies without significant findings for the Maitland's mobilization technique.5 Twelve of 21 studies (57%) using the Maitland's technique found no significant findings. That review also highlighted the fact that irrespective of the mobilization technique used in the studies included, a proportion of them found no significant findings. This could be due to the outcomes not being related to the mechanism of action of ankle mobilization, or due to differences in force application during the mobilization technique. None of the studies included in the review reported measuring peak force or a dose applied during the technique. Surprisingly, little is known about the force applied during an ankle mobilization technique. It is important to understand the variability in forces applied during mobilization sessions, as this may affect treatment outcomes.
To date, only 2 published studies have reported the variability in force applied during ankle mobilization.3,6 Both studies evaluated the consistency of the applied peak force during 30 seconds of grade IV Maitland's anteroposterior ankle mobilization. Both reported poor to good levels of intrarater agreement in applied peak force, with intraclass correlation coefficients ranging between 0.39 and 0.99. While insightful, intraclass correlation coefficients only provide estimates of relative agreement, and therefore they are specific to the characteristics of the test group. There is a need to determine absolute estimates of agreement, as these allow direct comparison of the force applied during repeated mobilization cycles and sessions. In a systematic review of 7 studies investigating the consistency of peak force applied during various mobilizations, only 3 studies reported absolute measures of agreement, and none during an ankle-joint mobilization technique. The aim of this study was to quantify the variability in the force applied by an experienced practitioner during 20 cycles of Maitland's grade IV anteroposterior ankle mobilization measured on 2 occasions separated by 1 week.
Methods
Participants
Thirteen healthy adults (5 women, 8 men) with a mean age of 25 ± 5 years, height of 170 ± 7 cm, and weight of 71 ± 16 kg participated in the study. No participant reported a history of foot or ankle surgery or disorders of connective tissue, neurology, or movement that might affect their ability to receive the mobilization technique. All participants provided written informed consent. Ethical approval was obtained from the Queensland University of Technology human research ethics committee (approval number 1300000053), and the study was undertaken in accordance with the Declaration of Helsinki.
Equipment
The posterior surface of each participant's heel was positioned on a 6-degree-of-freedom force transducer (Mini 45; ATI Industrial Automation, Apex, North Carolina). The force transducer had a diameter of 4.5 cm, with a measurement resolution of 0.25 N to a full scale of 580 N in the vertical direction (Fz) and 290 N in the transverse directions (Fx, Fy) and a resonant frequency of 5600 Hz. A square metal plate (5 × 5 cm, and 3 mm thick) was rigidly fixed to the transducer, allowing a larger surface area for the heel to be positioned (Fig 1a). The transducer was connected via cable to a desktop computer using the ATI stand-alone ActiveX software interface (version 2). Force data were sampled at 17 Hz.
Fig 1.
Participant setup for the Maitland's grade IV anteroposterior talar mobilization of the ankle. The posterior surface of the heel was positioned over a 6-degree-of-freedom force transducer (a). The practitioner applied a posteriorly directed force to the medial and lateral aspects of the talar head via the thumbs, while the index fingers applied an anteriorly directed force (b).
Procedure
A Maitland's grade IV anteroposterior (AP) talar mobilization,4 which incorporated a small-amplitude movement, was applied to the anterior talus. Each participant was positioned supine, with the knee in 10° flexion and the ankle in 20° of plantarflexion to achieve a so-called loose-packed talocrural position.7 The practitioner's thumbs were placed on the medial and lateral aspects of the talar head, with the index fingers positioned posteriorly at the medial and lateral posterior processes of the talus (Fig 1b). To ensure that the foot did not move from its position on the force transducer, the proximal leg was secured to the treatment plinth by a stabilizing strap.
The talus was glided posteriorly with downward force applied by the thumbs and anteriorly with the index finger.7,8 Each cyclic load application (posterior-anterior glide) lasted 6 seconds, with 20 cycles applied over a 2-minute period. A 2-minute mobilization session was used, as this duration has been shown to increase ankle range of motion more than shorter testing sessions.7 A metronome (Metronome Beats, version 3.6.1; London, UK) was used to provide a 0.3 Hz auditory cue to maintain a constant cyclic rhythm. The mobilization technique was applied by a single practitioner with more than 10 years’ experience in joint mobilization. Twenty cycles were captured during each mobilization session. The process was repeated during a second mobilization session that occurred 1 week after the initial session.
Data Management
Custom computer code (MATLAB, version R2016a; MathWorks, Natick, Massachusetts) was used to analyze force data. As shear forces constituted less than 0.5% of the applied force, the resultant force was calculated and the weight of the lower limb offset (zeroed). For each AP loading cycle, the peak force, impulse (force-time integral), loading rate (force-time differential), and cycle time were calculated.9,10 The mean peak force, impulse, loading rate, and cycle duration over the 20 mobilization cycles were subsequently calculated for each session.
Statistical Analysis
Statistical analysis was conducted using IBM SPSS (version 23; IBM Corporation, Armonk, New York). All data were evaluated for normality using the Shapiro-Wilk test. Within each session, absolute agreement of the peak force, impulse, loading rate, and load duration were estimated by calculating coefficients of variation (CoVs) for each participant. The mean within-session CoV for each session was subsequently calculated. The standard error of measurement was also calculated and expressed as a percentage of the mean value to aid comparison across variables.11 Agreement of the mean peak force, impulse, loading rate, and cycle duration between mobilization sessions was assessed using the approach with bias and 95% limits of agreement outlined by Bland and Altman.12 Potential differences between mobilization sessions in mean peak force, impulse, loading rate, and cycle duration were investigated using paired t tests. We used α = 0.05 for all tests of significance.
Results
Figure 2 highlights a typical cyclic force profile applied during each mobilization session from a representative participant. The force applied during the grade IV AP talar mobilization generally followed a sawtooth pattern, in which posteriorly applied forces showed distinct force peaks compared to anteriorly directed forces, which tended to show a more graduated trough.
Fig 2.
Resultant force-time curve obtained during a 2-minute Maitland's grade IV anteroposterior talar mobilization session in a typical participant (blue line) and during the repeated mobilization session undertaken 1 week later (red line). The dashed lines represent the mean peak force for the corresponding mobilization session. Posteriorly directed force is represented by an upward slope, anteriorly directed force by a downward slope.
Figure 3 displays the within-session variability for each loading parameter during each mobilization session. Cycle time displayed the lowest within-session variability, with a mean within-session CoV of less than 10%. Although peak force and loading rate displayed larger within-session variations (CoVs of 10%-15%) than cycle time, it was the cyclic impulse applied during each mobilization session that varied most considerably, with a mean within-session CoV of 20% to 40%.
Fig 3.
Within-session coefficient of variation for cycle time (■), peak force (●), impulse (◆), and loading rate (▴) applied during each anteroposterior talar mobilization session. The mean coefficient of variation for each session is represented by a red line.
Table 1 shows the mean peak force, impulse, loading rate, and cycle duration applied during each mobilization session. There was a statistically significant reduction between sessions in mean peak force (−17%, t12 = 2.445, P = .031) and impulse (−51%, t12 = 2.306, P = .040). The 95% limits of agreement in applied peak force (±33 N) and impulse (±128 N s) were large compared to their mean values.
Table 1.
Mean (Standard Deviation) Force Parameters and Variability in the Application of 20 Cycles of Maitland's Grade IV Ankle Mobilization Undertaken on 2 Measurement Occasions
| Variable | Session 1 | Session 2 | Bias | 95% LOA | ||||
|---|---|---|---|---|---|---|---|---|
| Value | CoV | SEM | Value | CoV | SEM | |||
| Cycle time (s) | 5.9 (0.1) | 8 (2) | 0.5 (8) | 5.9 (0.1) | 10 (4) | 0.6 (10) | 0.0 | ±0.2 |
| Peak force (N) | 70 (12) | 10 (5) | 7 (10) | 58 (10) | 13 (6) | 8 (14) | 12a | ±33 |
| Loading rate (N/s) | 83.4 (14.5) | 15 (3) | 12.8 (15) | 75.8 (15.7) | 15 (3) | 11.9 (16) | 3.8 | ±17.0 |
| Cycle impulse (N s) | 135.4 (38.9) | 21 (11) | 26.8 (20) | 93.6 (47.1) | 43 (33) | 30.6 (33) | 47.8a | ±128.0 |
CoV, coefficient of variation; LOA, limit of agreement; SEM, standard error of measurement (as % of mean).
Statistically significant difference between mobilization sessions (P < .05).
Discussion
This study investigated the within-session and between-sessions variability in cyclic force applied during a grade IV AP ankle mobilization, in which 20 loading cycles were performed over a 2-minute period. The standard error of measurement and within-session CoV for applied peak force and loading rate ranged between approximately 10% and 15%, which is comparable to results reported by de Souza et al,3 in which force application during a grade IV Maitland's ankle mobilization technique was augmented with visual feedback (15%). In contrast to the within-session variability, however, the between-sessions 95% limits of agreement for peak force (±33 N, or approximately ±50%), loading rate (±17 N/s, or approximately ±20%), and impulse (±128 N s, or approximately 112%) demonstrated substantially greater variability. The large variations in peak force and loading rate highlight the difficultly in applying consistent force during ankle mobilizations and may underpin the differential clinical effects reported with ankle mobilization in the literature.13 This large variation in force application does not allow for a comparative dose to be applied to the tissues, with some patients therefore receiving smaller doses of force. This has the potential for leading a researcher to find smaller clinical effects than if a larger dose were applied.
The mean peak force applied during ankle mobilization (65 ± 5 N) is consistent with that reported in spinal mobilization (55-158 N),9,14 but is approximately one-quarter to one-third of those reported by Resende et al6 and Venturini et al,15 in which peak forces on the order of 160 to 200 N were observed during a grade IV Maitland's AP ankle mobilization technique. The discrepancy in the magnitude of the peak force observed between studies may reflect several methodological differences. Chiefly, the technique outlined by Resende et al and Venturini et al involved the practitioner placing the talus between the thumb and index finger and applying body weight to produce the posteriorly directed force on the talus. In the current study, by contrast, the force was applied to the talus via the tips of thumbs and did not involve the use of body weight. Although participants in all three studies were healthy and free of ankle pathology, none of the studies characterized ankle-joint stiffness, hampering further comparison. It should also be noted that ankle mobilization in the present study was undertaken by a practitioner with more than 10 years of clinical experience in manual mobilization and manipulation. Hence, the force variability we observed may not be applicable to practitioners with differing levels of experience. Nonetheless, the findings of the present study provide clinicians and researchers with an indication of the consistency of the manual loads that are applied during a Maitland's grade IV AP talar mobilization in healthy adults. They also highlight the need for strategies that allow a more consistent application of force, thereby ensuring that a more consistent therapeutic dose is applied to the tissues.
Limitations
This study quantified the variability of applied force from only 1 practitioner. Consequently, variability may differ in novices or practitioners with differing levels of experience. Second, this study did not evaluate treatment effects—for example, measuring improvements for a specific outcome measure—and therefore it cannot be assumed that 20 cycles of consistent force is required to produce a treatment effect. It is possible that an effect could be produced after 1 force cycle, given that no study to date has quantified a minimum dose required to produce a clinical effect.
Given that ankle mobilization treatment durations typically range from 30 seconds to 3 minutes,7,16 there is a need to demonstrate absolute variations in force over clinically recommended durations for mobilization. To date, consistency between ankle mobilization sessions has only been reported for a return period of 48 hours,3 whereas practitioners may have their patients return after longer periods (eg, weekly treatments1), and therefore variation in force application between sessions with longer return periods may be more clinically important. This study provides variability over the longer return period as well as increasing the mobilization duration to 2 minutes, to make the results more clinical applicable.
Conclusion
The peak force and loading rate applied by an experienced practitioner during a Maitland's grade IV AP talar mobilization session varied by about 10% to 15% over 20 loading cycles. Variability between repeated mobilization sessions by the same practitioner was even greater, with peak applied force and loading rate agreeing only within ±33 N (approximately ±50%) and ±17 N/s (approximately ±21%), respectively. These levels of variability in applied force may underpin the differential clinical effects reported with ankle mobilization in the literature. The findings also highlight the need for future research to focus on strategies that ensure a more consistent application of force during talar mobilization.
Acknowledgments
Acknowledgments
The authors thank Scott Wearing for his technical assistance during the experiment.
Funding Sources and Conflicts of Interest
No funding sources or conflicts of interest were reported for this study.
Contributorship Information
Concept development (provided idea for the research): A.J.W.
Design (planned the methods to generate the results): A.J.W.
Supervision (provided oversight, responsible for organization and implementation, writing of the manuscript): A.J.W.
Data collection/processing (responsible for experiments, patient management, organization, or reporting data): A.J.W.
Analysis/interpretation (responsible for statistical analysis, evaluation, and presentation of the results): A.J.W., T.J.
Literature search (performed the literature search): A.J.W.
Writing (responsible for writing a substantive part of the manuscript): A.J.W., T.J.
Critical review (revised manuscript for intellectual content, this does not relate to spelling and grammar checking): A.J.W., T.J.
Practical Applications.
-
•
The peak force applied during 20 cycles of grade IV Maitland's AP ankle mobilization varied markedly both within and between treatment sessions when performed by an experienced therapist.
-
•
The large variation in applied force likely resulted in an inconsistent cyclic stretch of the connective tissues of the ankle, and may underpin differential treatment effects reported with ankle mobilization.
Alt-text: Unlabelled box
References
- 1.Cleland JA, Abbott JH, Kidd MO, et al. Manual physical therapy and exercise versus electrophysical agents and exercise in the management of plantar heel pain: a multicenter randomized clinical trial. J Orthop Sports Phys Ther. 2009;39(8):573–585. doi: 10.2519/jospt.2009.3036. [DOI] [PubMed] [Google Scholar]
- 2.Collins N, Teys P, Vicenzino B. The initial effects of a Mulligan's mobilization with movement technique on dorsiflexion and pain in subacute ankle sprains. Man Ther. 2004;9(2):77–82. doi: 10.1016/S1356-689X(03)00101-2. [DOI] [PubMed] [Google Scholar]
- 3.de Souza MV, Venturini C, Teixeira LM, Chagas MH, de Resende MA. Force-displacement relationship during anteroposterior mobilization of the ankle joint. J Manipulative Physiol Ther. 2008;31(4):285–292. doi: 10.1016/j.jmpt.2008.03.005. [DOI] [PubMed] [Google Scholar]
- 4.Maitland G. Churchill Livingstone; London: 1991. Peripheral Manipulation. [Google Scholar]
- 5.Weerasekara I, Osmotherly P, Snodgrass S, Marquez J, de Zoete R, Rivett DA. Clinical benefits of joint mobilization on ankle sprains: a systematic review and meta-analysis. Arch Phys Med Rehabil. 2018;99(7):1395-412.e5. [DOI] [PubMed]
- 6.Resende M, Venturini C, Penido M, Bicalho L, Peixoto G, Chagas M. Estudo da confiabilidade da força aplicada durante a mobilização articular ântero-posterior do tornozelo. Braz J Phys Ther. 2006;10(2):199–204. [Google Scholar]
- 7.Holland CJ, Campbell K, Hutt K. Increased treatment durations lead to greater improvements in non-weight bearing dorsiflexion range of motion for asymptomatic individuals immediately following an anteroposterior grade IV mobilisation of the talus. Man Ther. 2015;20(4):598–602. doi: 10.1016/j.math.2015.02.003. [DOI] [PubMed] [Google Scholar]
- 8.Houglum PA. 4th ed. Human Kinetics; Champaign, IL: 2016. Therapeutic Exercise for Musculoskeletal Injuries. [Google Scholar]
- 9.Harms M, Bader D. Variability of forces applied by experienced therapists during spinal mobilization. Clin Biomech (Bristol, Avon) 1997;12(6):393–399. doi: 10.1016/s0268-0033(97)00023-5. [DOI] [PubMed] [Google Scholar]
- 10.Snodgrass SJ, Rivett DA, Robertson VJ, Stojanovski E. Cervical spine mobilisation forces applied by physiotherapy students. Physiotherapy. 2010;96(2):120–129. doi: 10.1016/j.physio.2009.08.008. [DOI] [PubMed] [Google Scholar]
- 11.Wier J. Quantifying test-restest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res. 2005;19(1):231–240. doi: 10.1519/15184.1. [DOI] [PubMed] [Google Scholar]
- 12.Bland M, Altman G. Measuring agreement in method comparison. Stat Methods Med Res. 1999;8(2):135–160. doi: 10.1177/096228029900800204. [DOI] [PubMed] [Google Scholar]
- 13.Gorgos KS, Wasylyk NT, Van Lunen BL, Hoch MC. Inter-clinician and intra-clinician reliability of force application during joint mobilization: a systematic review. Man Ther. 2014;19(2):90–96. doi: 10.1016/j.math.2013.12.003. [DOI] [PubMed] [Google Scholar]
- 14.Snodgrass SJ, Rivett DA, Robertson VJ. Manual forces applied during cervical mobilization. J Manipulative Physiol Ther. 2007;30(1):17–25. doi: 10.1016/j.jmpt.2006.11.008. [DOI] [PubMed] [Google Scholar]
- 15.Venturini C, Penedo MM, Peixoto GH, Chagas MH, Ferreira ML, de Resende MA. Study of the force applied during anteroposterior articular mobilization of the talus and its effect on the dorsiflexion range of motion. J Manipulative Physiol Ther. 2007;30(8):593–597. doi: 10.1016/j.jmpt.2007.08.002. [DOI] [PubMed] [Google Scholar]
- 16.Green T, Refshauge K, Crosbie J, Adams R. A randomised controlled trial of a passive accessory joint mobilization on acute ankle inversion sprains. Phys Ther. 2001;81(4):984–994. [PubMed] [Google Scholar]