Background and Purpose. Mental practice has been shown to be effective in increasing the force production of the abductor digiti minimi muscle in the hand. The aim of this study was to determine whether mental practice could produce strength gains in the larger ankle dorsiflexor muscles, which are important during walking. Subjects. Twenty-four subjects were randomly assigned to a physical practice group, a mental practice group, or a control group (8 subjects per group). Methods. In the practice groups, subjects either physically or mentally practiced producing maximal isometric contractions for 3 sets of 10 repetitions, 3 times per week for 4 weeks. Changes in mean peak isometric torque normalized to body weight and the resulting percentage of improvement were analyzed across the 3 groups. Results. Differences in raw torque production after training in the 2 practice groups resulted in significant percentages of improvement for the physical practice group (25.28%) and the mental practice group (17.13%), but not for the control group (−1.77%). The 2 practice groups were not statistically different in their maximal torque-generating capacity after training. Discussion and Conclusion. These findings show that mental practice in people without impairments can lead to an increase in torque production similar to that produced by physical practice. Such a technique may prove to be a useful adjunct to traditional treatment options aimed at increasing muscle strength.

Treatment plans developed for a wide variety of orthopedic and neurologic diagnoses frequently include the goal of increasing the strength (force production) of specific muscles or muscle groups. Techniques used by physical therapists to improve muscle strength frequently include resistance exercises with weights, elastic bands, and isotonic and isokinetic machines and the use of neuromuscular electrical stimulation. In all of these techniques, the patient is required to contract the muscles being trained. With some orthopedic and neurologic lesions, however, muscle contraction may cause pain or may not even be possible. Neurophysiological research1,2 has suggested that, through the use of mental practice, it may be possible to improve muscle strength without actually requiring significant muscle contraction.

Mental practice is the cognitive rehearsal of a task in the absence of movement.3 Numerous studies39 have shown mental practice to be an effective training technique for enhancing the performance of motor skills when used in combination with physical practice and even when used in isolation. Although research examining the efficacy of mental practice for motor skill acquisition is common, the effect of mental practice on muscle strength has received much less attention. Notably, 2 studies1,2 have examined the ability of mental practice to improve the strength of an isolated muscle in the hand, the abductor digiti minimi. Our study expanded on this work by examining whether mental practice can increase the strength of the ankle dorsiflexor muscle group, a clinically important muscle group necessary for safe ambulation.10

Yue and Cole1 were the first to compare strength increases in the abductor digiti minimi muscle in mental practice and physical practice groups after 4 weeks of training. Increases in muscle strength were found in both groups when compared with a control group, with muscle force improvements of 30% and 22% for the physical and mental practice groups, respectively. In an attempt to determine whether improvements as a result of mental practice are predicated on the presence of low-level activity in the relevant musculature, the researchers recorded electromyographic (EMG) activity from the abductor digiti minimi muscle. Analysis of the EMG results, however, did not provide a definitive answer to this question.

In a more recent investigation, Smith et al2 also examined the effect of mental practice on strength gains in the abductor digiti minimi muscle, and they also recorded EMG activity during practice. In this study, subjects in the physical practice group trained twice a week for 4 weeks and produced 20 maximal voluntary contractions each practice session. The mental practice group followed the same schedule, but, instead of physical practice, they were instructed to imagine themselves performing 20 maximal contractions. Observation of a dynamometer ensured that no motion in the finger occurred during the mental practice sessions. Like Yue and Cole,1 Smith et al2 found an increase in muscle force production (strength) in both the physical practice group (53%) and the mental practice group (23%). Furthermore, the researchers found that, although the mental practice group did elicit significant EMG readings during practice, there was no statistically significant relationship between the magnitude of the EMG recordings and the amount of improvement in muscle force.

It is clear, however, that despite the lack of a definitive understanding of the processes underpinning the effect of mental practice, strength gains have been achieved with this practice technique. Unfortunately, in the studies by Yue and Cole1 and Smith et al,2 strength improvements were examined only in the abductor digiti minimi, a small muscle in the hand. Yue and Cole1 point out that hand muscles have a disproportionately large representation in the primary motor cortex and that the abductor digiti minimi muscle is relatively unused in activities of daily living. Thus, the researchers were reluctant to speculate as to the possible therapeutic benefit of mental practice for clinically significant strength improvement but suggested that such potential benefits be examined in the future.

The aim of our study, therefore, was to determine whether mental practice can be used to increase the force production of a muscle group with a much smaller representation in the motor cortex and yet one that is more important in activities of daily living, namely the ankle dorsiflexor muscles. The ability to effectively dorsiflex the ankle is a critical requirement for safe ambulation. Patients with ineffective dorsiflexor muscle performance are more likely to be unable to clear the foot during the swing phase of gait and thus are prone to hazardous falls. If mental practice proves effective in increasing the torque generation of the dorsiflexor muscles, this technique may be particularly valuable in the treatment of patients in the initial stages of neuromuscular or musculoskeletal damage when they may be confined to a bed or physical contractions may not be appropriate or even possible. It also may help strengthen muscles later in rehabilitation when traditional resistance exercises are being prescribed.

To examine the efficacy of mental practice in strengthening the ankle dorsiflexor muscles, we examined the ability of 3 groups to generate torque. One group physically practiced maximal isometric contractions, the second group mentally practiced the same contractions, and the third group (control group) did not practice at all. We hypothesized that mental practice would result in greater increases in dorsiflexor torque production than those found in the control group and that these changes would be equivalent to those achieved by the physical practice group.



Twenty-four physical therapist students volunteered to participate in the experiment. Subjects were 6 men and 18 women who were 19 to 26 years of age (X̄=22.7 years). No subjects were actively engaged in an exercise program or had an existing lower-extremity injury. After the subjects read and signed informed consent forms, they were randomly assigned to a mental practice intervention group, a physical practice intervention group, or a control group, with the restriction that each group have 8 subjects and the same ratio of men to women.


On the initial day of the study, the ability of all subjects to produce maximal torque using the dorsiflexor muscles was measured with a Biodex System 3.0* isokinetic dynamometer. The Biodex dynamometer has been found to yield valid and reliable measurements of torque produced mechanically11,12 and torque produced by various muscle groups, including the knee extensors and flexors13,14 and the ankle dorsiflexors.15 The manufacturer's protocol was followed for subject positioning with each subject seated with his or her hips and knees flexed so that the tibia was resting on a pad parallel to the floor. The foot rested on the vertical footplate with the ankle in a neutral position. The subjects then were given verbal instructions to dorsiflex their foot as hard as they could against the strap that held their foot to the footplate for a 5-second isometric contraction. This process was repeated 10 times with a 5-second rest period between contractions. As in previous research,2 the mean peak torque of the 10 trials was calculated to increase reliability for subsequent data analysis.

Following this pretest, subjects assigned to the mental practice group were given instructions from a prepared script on how to use kinesthetic mental imagery (Appendix). The subjects then performed 3 mental imagery practice trials of isometric ankle dorsiflexion. To ensure that the 2 practice intervention groups received the same number of trials, the physical practice group also was required to perform 3 maximal dorsiflexion contractions following baseline testing. All subjects in the physical practice and mental practice groups then performed 3 sets of 10 repetitions within their assigned practice regimen individually 3 times per week for 4 weeks. Each training session lasted approximately 15 minutes. All procedures used were identical to those of the baseline testing phase and were always performed on the Biodex dynamometer. Subjects were given 20 seconds of rest between sets of 10 repetitions. If assigned to the control group, subjects did not participate in any form of practice during the 4 weeks of the experiment.

At the start of each training session, subjects receiving the mental practice intervention were read the same imagery script given on day 1. The mental practice group then imagined performing the maximal isometric ankle dorsiflexion contraction for 3 sets of 10 repetitions under the same schedule as the physical practice group. During this practice, subjects also were positioned and attached to the dynamometer as in the pretest, just as the physical practice subjects were. Subjects were instructed not to produce a muscle contraction of the ankle dorsiflexors during mental practice. The dynamometer was monitored during imagery trials to ensure that no torque was being generated during the trial and the leg was visually monitored for any signs of muscle contraction. Muscle force might be generated by an isometric co-contraction of the dorsiflexors and the plantar flexors without any net torque being developed against the dynamometer. The instructions for the physical practice group were the same as those given for the pretest. No feedback was given to the physical practice group during training, and all groups were instructed not to perform any training outside their scheduled sessions.

Data Analysis

During the posttest, all subjects were again tested for maximal dorsiflexion isometric torque for 10 trials as in previous studies.1,2 For each set of trials, the Biodex system reported the coefficient of variability (CV) as an index of the variation in the peak torque across the 10 trials. All testing was performed by an researcher experienced in the use of the Biodex system. Changes in mean peak torque normalized with respect to a subject's body weight were analyzed using a 3 × 2 (group × test) analysis of variance (ANOVA). The percentage of change in mean peak torque production between the pretest and posttest for each subject also was calculated and then was submitted to a 1-way between-subjects ANOVA. Post hoc analysis of significant F ratios (P<.05) was conducted using the Tukey honestly significant difference (HSD) procedure.


No subjects in the mental practice group were observed producing torque against the dynamometer during training. Furthermore, no subjects appeared to use a co-contraction strategy, because there was no indication of muscle contraction in the dorsiflexor muscles of any subject in the mental practice group.

The subject pretest and posttest data are shown in the Table and expressed graphically in Figure 1. These data indicate that the majority of subjects were quite consistent in torque generation across the 10 repetitions. The mean peak torque calculated by the Biodex system software was used in subsequent data analyses.

Figure 1.

Pretest and posttest mean peak torque as a function of group assignment.


Pretest, Posttest, and Percentage of Change in Mean Peak Torque Data as a Function of Group Assignment

The group × test ANOVA on the mean peak torque data revealed significant main effects for group (F=6.2; df=2,21; P<.01) and test (F=22.4; df=1,21; P<.0001). A significant group × test interaction (F=8.1; df=2,21; P<.01) superseded these main effects. Follow-up analysis using the Tukey HSD procedure indicated that there were no significant differences among the groups during the pretest (control group: X̄=0.44 N·m/kg, SD=0.027; mental practice group: X̄=0.44 N·m/kg, SD=0.014; physical practice group: X̄=0.45 N·m/kg, SD=0.032). Analysis also revealed that posttest performance in the mental practice (X̄=0.52 N·m/kg, SD=0.047) and physical practice (X̄=0.57 N·m/kg, SD=0.100) groups was significantly greater than pretest performance, but this was not the case for the control group (X̄=0.44 N·m/kg, SD=0.031). Furthermore, both the mental practice and physical practice groups generated greater posttest torque than the control group; however, the difference between the 2 practice groups during the posttest was not statistically significant.

For a more generalizable assessment of strength gains, the percentage of change in mean peak torque production from pretest to posttest was calculated for each subject. Group means revealed an increase in mean peak torque of 25.28% (SD=20) for the physical practice group, an increase of 17.13% (SD=13) for the mental practice group, and a decrease of 1.77% (SD=6) for the control group (Fig. 2).

Figure 2.

Percentage of change in mean peak torque as a function of group assignment.

A 1-way ANOVA performed on these data revealed a significant effect of intervention (F=7.6; df=2,21; P<.01). Follow-up analysis using the Tukey HSD procedure indicated that the interventions for both the mental practice group and the physical practice group resulted in significantly greater percentage of improvement in torque than in the control group (P<.05). The difference in improvement in torque production between the physical practice and mental practice groups was not statistically significant.


Our study was designed to explore an intervention strategy that physical therapists might consider when the inability to produce dorsiflexion torque is a limiting factor in a patient's functional capability. Our results show that a mental practice intervention can increase maximal torque generation of the dorsiflexor muscles significantly without actual physical practice. Although the physical practice group had an 8% greater improvement in mean torque than the mental practice group, this difference was not statistically significant. Our study, however, had a relatively small number of subjects per group. With a greater number of subjects, a difference of 8% might have reached statistical significance. The general finding of increased torque production with mental practice, however, is consistent with previous studies in which isometric strength gains were found with mental practice in muscles in the hand.1,2 These studies were designed to explore the mechanisms by which mental practice can mediate increases in motor performance. To that end, EMG data were analyzed in an attempt to determine the relationship between such activity and performance enhancement.

A causative relationship between performance enhancement and minute EMG recordings during mental imagery is the basis of the psychoneuromuscular explanation of mental practice. Richardson16 developed this theoretical explanation based on previous work showing minute innervations of task-specific musculature during mental imagery.1719 This innervation was thought to provide ongoing kinesthetic feedback, allowing subsequent corrections in the motor program and facilitation of neuromuscular coordination.16 Similarly, it has been suggested that neuromuscular activity enables the motor cortex to develop motor schema20 or primes corresponding muscle movement nodes.21 In a more cognitive interpretation, researchers have proposed that mental practice increases performance by improving the preparation and anticipation of movements rather than movement execution.22 Support for this proposal comes from the finding, obtained from positron emission tomography, of increased activity in the medial aspect of the orbitofrontal cortex and decreased activity in the cerebellum following mental practice.22 This cerebral functional reorganization was similar to that observed after physical practice of the sequential foot movement task by these researchers.

Our study was not designed to produce data that could contribute to this theoretical debate. Despite our inability to determine the underlying mechanism, mental practice used by the subjects in our study resulted in significant gains in the ability to produce maximal torque. This finding perhaps should not be surprising considering that the majority of initial strength gains from resistive exercise are not due to changes in the physiological capacity of the muscle fibers themselves. Muscle hypertrophy resulting from the addition of contractile proteins is a gradual process that takes many weeks to occur.23,24 The ability to increase force production during the initial weeks of training, therefore, is thought to be the result of neural adaptations.23,25 Possible mechanisms of these neural adaptations include the extent of motor unit activation,26 improved coordination,27 and decreased co-contraction of antagonist muscles.28 In our study, subjects were trained for only 4 weeks as was done in previous research,1,2 and, therefore, we would expect changes in force production to be primarily a function of these neural adaptations. Therefore, given the length of training in our study, it appears that whatever neurological changes in muscle recruitment were taking place in the physical practice group also were being stimulated by the mental practice intervention. If the training period continued for another month, strength differences between the 2 practice groups would be expected to emerge as physiological hypertrophy became evident in the physical practice group.23

A potential limitation of our study arises from the fact that our sample consisted of only students who were healthy. Extrapolating the findings to clinical populations, therefore, might be questionable. In many orthopedic injuries (eg, bony, ligamentous, and muscular lesions), however, the patient's nervous system remains relatively unaffected, and mental practice may be an appropriate technique to use with these diagnoses. It also should be noted that mental practice has been found to be an effective technique for motor skill acquisition in an elderly population.29 The question remains, however, whether mental practice would be effective in increasing force production in people with neurological involvement, such as those diagnosed with stroke. The answer to this question must remain speculative, but a number of review articles have supported the role of mental practice in the treatment of patients with stroke.3032 Furthermore, a number of recent empirical studies examining various motor skills—including weight bearing,33 wrist movements,34 line-tracing accuracy,35 and general arm movements,3638—have shown mental practice to be a valuable rehabilitative tool in patients with stroke. This evidence holds the hope that mental practice also may be effective in increasing strength in patients with stroke.

Finally, in our study, subjects were tested on their ability to produce torque under isometric conditions and then were physically or mentally trained under these conditions as was done in previous studies.1,2 Although concentric and eccentric activations were not trained, the mental practice literature does not suggest that the effectiveness of mental practice would interact with the type of contraction being imagined. Therefore, we believe that mental practice would be equally effective for other types of muscle activation that were not tested in our study.


The current study demonstrated that the use of mental practice is effective in increasing the torque production of the ankle dorsiflexor muscles in subjects without impairments. The ankle dorsiflexor muscles were chosen because they are a clinically significant muscle group in activities such as gait and stair climbing. The next step then is to examine the role of mental practice in strength improvements in patients with neurological or musculoskeletal involvement. The strong effect found for mental practice does hold the hope that this intervention may have a role in the rehabilitation of these patients either as a stand-alone intervention for patients with acute injuries or as a useful adjunct to traditional therapeutic exercise later in the rehabilitation process.



Script for the Mental Practice Group


  • Both authors provided concept/idea/research design and data collection. Dr Sidaway provided writing and data analysis. Ms Trzaska provided subjects.

    This research was presented, in part, at the Combined Sections Meeting of the American Physical Therapy Association; February 20-24, 2004; Boston, Mass.

    This study was approved by the institutional review board of Husson College.

  • * Biodex Medical Systems Inc, 20 Ramsay Rd, Shirley, NY 11967-4704.

  • Received September 27, 2004.
  • Accepted March 22, 2005.


View Abstract