Effects of intrinsic foot muscle training combined with the lower extremity resistance training on postural stability in older adults: a randomised controlled trial

Abstract

Background

As the population ageing, the problem of increased incidence of falls and higher healthcare expenditure in the elderly will be further accentuating. Intrinsic foot muscles play an important role in postural control and its function is significantly associated with the risk of falls. Therefore, this study aimed to determine the effects of intrinsic foot muscle training on postural control in older adults.

Methods

A single-blind randomized controlled trial was conducted on 123 older participants. They were randomly divided into four groups, including short-foot combined the lower extremity resistance training group (SF-RT group), towel-curl training combined the lower extremity resistance training group (TC-RT group), lower extremity resistance training group (RT group) and control group. Three intervention groups performed resistance training and/or additional foot muscle training three times a week for 8weeks. The Sensory Organization, Limit of Stability, Motor Control tests were used to determine postural stability. The ergoFet dynamometer and colour Doppler ultrasound system were used to determine the foot muscle strength and morphology. Mobility was assessed using the Berg Balance Scale as well as the Timed Up and Go test. The multivariate of analysis for repeated measurements (MANOVA) test was used to determine the differences of these measurements.

Results

For postural stability, in Sensory Organization Test, compared to RT or the control groups, the SF-RT group had significant improvement at equilibrium, strategy and overall scores (p < 0.05). In the Limit of Stability test, there were significant differences in movement velocities between SF-RT and control groups (p < 0.05). Moreover, there were significant improvement at foot muscle strength and morphology found in SF-RT and TC-RT group.

Conclusion

Compared with regular intervention grogramme, additional short-foot training militated extra effect on improving static postural stability, mobility, foot muscle strength and morphology in the elderly.

Trial registration

This study has been registered at Chinese Clinical Trial Registry a priori as a clinical trial (ID ChiCTR2000033623) on June 7th, 2020.

Background

Balance disturbance, as the main symptom in older adults [1], including the age-related postural instability, has been an important risk factor for their quality of life and longevity. Decreased postural control has been reported to be present in approximately 35% of older adults (over 70 years of age) as well as 61% of older adults (over 80 years of age) [2].

The human foot is the starting segment of lower-extremity kinetic chain, and its highly complex structure contributes to stability by modulating its structure, posture and muscle activity [3, 4]. There is growing evidence that foot muscles exert positive biomechanical effects on the medial longitudinal arch, and improve dynamic postural control [5–7]. Intrinsic foot muscles (IFMs), as the local stabilizer of foot core system, were speculated to support the foot functional half dome, contribute the foot core stability, thus providing mechanical and neuromuscular regulation overall posture control [8, 9]. An overview of foot’s muscular structure, previous studies has predominantly focused on the extrinsic foot muscles (EFMs) and their role in toe flexion. There has shown that toe flexor strength was correlated with the maximum stability range of the human body, and was related to its fall risk, which considered as one of the important contributing factors to balance function and mobility in the elderly [10, 11]. However, our recent study revealed that the decline in foot muscle strength in older adults was not limited to flexion strength, but also included the IFMs strength, moreover its strength was correlated with the maximum stability range [12].

Recently, the relationship between IFMs and human postural stability has been widely concerned. Wallace et al. [13] reported that IFMs could play a certain role in compensating postural disturbance caused by vestibular electrical stimulation through specific muscle activities. Similarly, our team’s recent work has found that during a quiet upright stance, the muscle activity of IFMs increased with the sensory input disturbance in the elderly, and there was a positive correlation between their muscle activity and limit of posture stability [14].

Several cross-sectional studies have been shown that there was a correlation between foot muscles and human stability, including IFMs and EFMs [10, 11]. However, a significant challenge lies in providing objective evidence of IFMs in postural stability in older adults, especially through precise measurement methods rather than using several functional tests or scales, including the Timed Up to Go Test, and Berg Balance Scales [15, 16]. Moreover, for IFMs, several strengthening exercises has been widely used in clinical setting [17, 18], such as the patients with chronic ankle instability and flatfoot, rather than employed in the elderly. Therefore, there was a lack of evidence directly indicating the effect of IFMs strengthening exercises on the postural control of the elderly.

As the population ages, the problems associated with reduced postural control in the elderly, increased incidence of falls, and higher medical expenditures will be further highlighted. It is of practical and theoretical significance to investigate the foot muscle function training to further improve the effect of existing intervention programs on increasing the postural control ability and reducing the fall risk of the elderly. Therefore, the primary aim of this study is to determine the IFMs exercise for the older adults and to compare the effectiveness of different IFMs exercise on postural stability.

Methods

Study design and setting

This study was a single-blind randomized controlled trial and its intervention was conducted at several community centres in Shanghai, China and its evaluations were conducted at the Sport Medicine and Rehabilitation Centre, Shanghai University of Sport. Moreover, it was approved by the Ethics Committee of the Shanghai University of Sport (Ref. No.: 102772020RT001) and registered at Chinese Clinical Trial Registry a priori as a clinical trial (ID: ChiCTR2000033623). This study met the ethical aspects based on Resolution 466/12 of the National Health Council and the Declaration of Helsinki. In this study, the researcher and evaluators were blinded to the group allocation. All these measurements were conducted by trained research assistants who blinded to group assignment and conducted in randomized order. Furthermore, the researchers responsible for data collection and data analysis were blinded for the random allocation. The flow diagram of this study was shown in Fig. 1. Three different intervention programs were employed for postural stability promotion in the elderly, while the participants were randomly divided into four groups, including a control group without intervention. The study was grouped separately according to community to avoid the influence of confounding factors on the findings, such as the distribution of age and gender, and education level among these communities. And the evaluations of postural stability, IFMs function, and its morphology were managed at pre-intervention and post-intervention.

Fig. 1.

The flow diagram of this study. Abbreviation: SF-RT group, the short-foot training combined the lower extremity resistance training group; TC-RT group, the towel-curl training combined the lower extremity resistance training group; RT group, the lower extremity resistance training group

Participants

The G*POWER software 3.1 software was used for the suitable sample size for the multivariate of analysis for repeated measurements (MANOVA) test. With the settings of α = 0.05, power (1 − β) = 0.80 and effect size = 0.25, power analysis showed that 23 participants for each group were the required sample size. Considering the drop-out rate (30%), 30 participants in each group at least were recruited in this study.

The voluntary participation was requested via posters, and expert lectures on community centres in Shanghai, China. The study team (clinicians, physical therapists and exercise specialists) screened all participants based on the following inclusion and exclusion criteria. Prior to the formal testing, the interviewer gave a detailed introduction of the trial process and intervention settings to all the eligible participants, explaining the purpose of the study and its significance in real-life situations, and informing them of the contribution of participation in the study to future health promotion among older adults, the possible benefits to them as individuals, as well as the possible dangers and adverse effects. The interviewer distributed an informed consent form to each participant to obtain his/her agreement and signature.

Recruited participants were grouped by computer-generated random numbers after completing the initial screening, and the grouping information was transferred into opaque envelopes for storage and blinded to all examiners. A total of 125 eligible elderly participants were recruited into this study and randomly divided into 4 groups, respectively, the lower extremity resistance training group (RT group), short-foot training combined the lower extremity resistance training group (SF-RT group), towel-curl training combined the lower extremity resistance training group (TC-RT group), and a control group through the automatic random number generation method.

Inclusion criteria

1. Able to maintain the standing position;

2. Able to walk alone, without others, prosthesis or mobility aids;

3. Normal cognitive function, able to understand the intervention and evaluation process;

4. Adults older than 60 years old;

5. Availability: three times a week over 8 weeks.

Exclusion criteria

1. Abnormal foot posture which has been shown to be associated posture instability [19] (It was determined by FPI-6 scale, and the cutoff scores were defined at + 1 to + 8 [20]);

2. Abnormal foot structure (toe deformity, hallux valgus, etc.);

3. Participants diagnosed with diseases related to postural control, such as vestibular dysfunction, Alzheimer’s disease, Parkinson’s disease and motor neuron disorders;

4. History of lower limb trauma in the past year;

5. Participants with medical contraindications for exercise, including severe cardiovascular diseases, respiratory diseases, musculoskeletal disorders, and any situation where exercise causes pain or discomfort;

6. Currently undertaking a structured exercise programme for postural control, such as the balance function rehabilitation training used in clinical settings.

Interventions

In the present study, participants in each intervention groups performed three different intervention programs, whereas participants in the control group did not undergo any intervention, and were asked to maintain their previous lifestyles. The short-foot and towel-curl training, as foot muscle training, are common interventions to promote foot function. The difference is that the short-foot training focuses more on the intrinsic foot muscles [21] and improves their ability to support the arch by enhancing the neuromuscular control, whereas the towel-curl training stimulates the intrinsic foot muscles to a certain extent but focuses more on the flexion of metatarsophalangeal joints and relies on the extrinsic foot muscles to complete the training maneuvers.

The intervention groups were required to complete eight consecutive weeks of training, three times per week, with a minimum of one day between each session. A metronome was used to standardize the duration and number of intervention sessions, and several qualified physical therapists provided supervision and timely guidance to actively encourage the participants to complete the intervention sessions. The foot muscles trainings, the short-foot training and towel-curl training, were performed after the resistance training. 5-min warm-up and 5-min cool-down exercise were conducted before and after each intervention session, which consisted of joint mobility training and muscle stretching. The resting time about one minute was provided between two set of exercises.

In addition, any reason for absence from training or withdrawal from the study, including personal reasons or adverse reactions, was recorded with the date and reason. There were 24 sessions of intervention and if a participant was absent more than 20% [5] of the total number of times, that participant’s data would not be used in subsequent data analyses. Similarly, if control group additionally participated in other intervention components, their data would be excluded.

The following is a detailed description of the arrangements during the intervention period for the three intervention groups as well as the control group:

1. The lower extremity resistance training group (RT group)

   The RT program included resistance training for hip joints (flexion, extension, adduction, abduction), knee joints (flexion, extension), and ankles (plantar flexion, dorsiflexion) bilaterally (Table 1). Two different strengths of elastic bands (20 lbs, and 30 lbs) were used for resistance training, with three sets of 8–12 repetitions of blocks, with the elastic band intervention protocol referenced to Kwak et al. [22].

2. The short-foot training combined the lower extremity resistance training group (SF-RT group)

   Same as the RT group, participants in SF-RT group were also required to complete elastic band lower extremity strength training that lasted for 8 weeks, three times per week under the supervision of trainer. Differently, SF-RT group were additionally required to perform short foot training. Under the direction of trainers, participants raised their foot arch for 5 s and pulled the metatarsal heads toward the heel bone without flexing the metatarsophalangeal joints. The trainers placed hand over participants’ arch to determine if the arching foot maneuver is completed correctly. As with the lower extremity strength training, participants performed short-foot training with 3 sets of 12 repetitions. The short-foot training was performed progressively in the seated, bipedal standing, and unilateral standing position (at the unilateral standing position section, the SF training was conducted on the dominant side and non-dominant side) [23, 24], with the participants moving on to the next position when he or she could perform the training maneuver skillfully and easily in the previous position.

3. The towel-curl training combined the lower extremity resistance training group (TC-RT group)

   The participants in TC-RT group performed additional towel roll training, in addition to RT. The trainer placed a towel under the feet of participants, who trained by flexing their toes to the best of their ability and curling the towel up, and continued for 5 s. Similarly with SF training, participants performed towel-curl training with 3 sets of 12 repetitions. And the intensity was progressively increased by foot loading (seated, bipedal standing, and unilateral standing position). Similarly, at the unilateral standing position section, the TC training was conducted on the dominant side and non-dominant side.

4. The control group

   The control group did not undergo any intervention training involved in this study. Additionally, this group was required to maintain their usual daily activities and not to participant any additional training or treatments related to postural control promotion during study period. And the participants of control group received treatment, including the resistance training and foot muscle training after the end of study.

Table 1.

The lower extremity resistance exercise program

Items	Description	Duration
Warm-up exercise	Joint mobility training		5 min
The lower extremity resistance exercise	1、Hip flexion (8–10 times *3 sets)		40 min, 3 times/week (8 weeks) week1-4, green band for 20lbs; week5-8, blue band for 30lbs
	2、Hip extension (8–10 times *3 sets)
	3、Hip adduction (8–10 times *3 sets)
	4、Hip abduction (8–10 times *3 sets)
	5、Knee flexion (8–10 times *3 sets)
	6、Knee extension (8–10 times *3 sets)
	7、Ankle flexion (8–10 times *3 sets)
	8、Ankle extension (8–10 times *3 sets)
Cool-down exercise	Muscle stretching	5 min

Measurement

This randomized controlled study included measurement in the pre-test (pre-intervention, week 0) and post-test (post-intervention, week 9). To avoid bias, all these measurements were conducted by trained research assistants who blinded to group assignment and conducted in randomized order. Moreover, to minimize the potential confounding effects of fatigue associated with resistance training, the post-intervention assessment was scheduled two days after the completion of the intervention period and conducted within a seven-day timeframe.

Primary outcome measurements

Postural stability

The NeuroCom Balance Manager System (Version 9.3, Copyright ©1989–2016 Natus Medical Incorporated) were used to determine static and dynamic postural stability of older participants. The balance platform is equipped with 23 cm×46 cm dual-force plate to trace the trajectory of center of pressure (COP) of each participant, and its sampling frequency is at 100 Hz. We employed the commonly used sensory organization test (SOT), limits of stability test (LOS), and motor control test (MCT), the test details of which have been reported in previous studies [12, 25]. In SOT, LOS and MCT test, the tasks were performed in a randomized order. All participants were informed and demonstrated of all procedures in advance and performed barefoot at the NeuroCom Balance Manager System.

1. Sensory organization test (SOT).

   In this test, participants were required to maintain postural stability in six different conditions, each of which needed to be performed three times each and each Task needed to be maintained for 20 s. The change of viewing, visualization, and support surface are combined with each other to make up the six conditions of SOT test, which in turn interferes with the participant’s postural control. In the SOT test, the system evaluated the participants’ postural perturbations, thus produced equalization scores, strategy scores for each of the six test conditions, and a composite postural stability score which using a weighted average of these scores and be used for analysis. A high equilibrium score means a low postural sway, indicating greater postural stability [26].

2. Limits of stability test (LOS)

   This test is an active motor function test in the NeuroCom Balance Manager System. In LOS, the participants were barefoot and stood in a bipedal standing position on dual-force plate. The research assistants instructed the participants with the test procedure and assisted in adjusting their initial stance and foot position. The test required the participants to actively move the center of gravity (COG) toward the limit of stability in eight directions (positive forward, right forward, positive right, right backward, positive backward, left backward, positive left, left forward), without lifting the feet off balance and without significant movement of the hip joints. To avoid ambiguity, we uniformly defined the dominant side as the right side and the non-dominant side as the left side. Based on the participant’s shift in 8 directions, the system calculated following parameters based on the changes in COP of each trial: Reaction Time (RT), Movement Velocity (MVL), Endpoint Excursions (EPE), Max Excursions (MXE), and Directional Control (DCL). The system calculated the values of the above parameters in eight directions and made a weighted average of the scores according to each direction, and finally took the scores of the above five parameters in the four directions of forward, backward, dominant direction, and nondominant direction as well as the composite score.

3. Motor Control Test (MCT)

   This test aims to assess the dynamic postural stability of participants in the anterior-posterior direction. During the procedure, participants stood barefoot in a bipedal stance on a support surface that underwent sudden horizontal displacements in the anterior-posterior direction. Participants were required to maintain postural stability, without informing of either the direction or timing of the displacement. The displacement speed is categorized into three distinct levels: low (base threshold: 2.8°/s), medium (moderate level: 6.0°/s), and high (saturation speed: 8.0°/s). For each combination of direction and speed, participants were required to complete three trials in random order. Comprehensive safety measures have been implemented to ensure the protection of participants from potential adverse events. The primary parameters were calculated in this study: (1) the Latency (ms): the response time from the movement of support surface to movement of COP, (2) the Amplitude Scaling: the capacity to apply appropriate force and effectively counteract the perturbation of the support surface to restore postural stability.

Secondary outcome measurements

Balance and walking ability

We performed the Berg Balance Scale (BBS) and the Timed Up and Go Test (TUGT) to assess balance and walking function in the elderly, both of which are commonly used in clinical evaluations as well as research in this population [27, 28]. In this study, the same skilled research assistant conducted the pre- and post-intervention assessments of BBS and TUGT.

Foot muscle strength

An ergoFet ergometer (Hoggan Health, USA, sampling frequency of 100 Hz) was used to determine the foot muscle strength of dominant side. Foot Muscle Strength consisted of EFMs and IFMs strength. In particular, the IFMs strength test (arching foot), accomplished by activating IFMs, was performed on customized frame with the dynamometer. The EFMs strength consisted of the 1st toe flexion strength (T₁), the 2nd–3rd toe flexion strength (T₂₃) and the 2nd–5th toe flexion strength (T₂₃₄₅). The participants were encouraged to pull “T”-shaped metal bar or screw buckle as best he or she could by corresponding toes. The “T”-shaped metal bar or screw buckle were secured to the dynamometer in a custom wood frame. The described strength test required the participants to complete three valid tests, and the data were initially processed and then averaged for statistical analysis and normalized to the participants’ body weights. These strength tests and its reliability have been reported in our previous studies [12, 29].

Foot muscle morphology

In this study, foot muscles morphology was determined by a portable color Doppler ultrasound system (Diagnostic Ultrasound System, M7 Super, Mindary, China) with a 10 MHz linear broadband array transducer (model: L14-6s), including abductor hallucis (AbH), flexor digitorum brevis (FDB), quadratus plantae (QP), and flexor hallucis brevis (FHB) of the dominant foot [29, 30]. Based on previous studies [4, 31] and findings of our own work [12], it has been established that IFMs, including the AbH, FDB and QP, exhibit a significant correlation with various parameters indicative of human postural stability. Given the sensitivity of ultrasound imaging in evaluating superficial musculoskeletal structures, the morphology of AbH, FDB and QP were assessment for IFMs. All thickness and cross-sectional area (CSA) of these IFMs were standardized with participants’ height. Due to the specificity of the ultrasound test, human soft tissue morphology would change depending on the amount of pressure applied, and tester should apply as little pressure as possible to the skin tissue during ultrasound test. In addition, all pre- and post-intervention muscle morphology tests were performed by the same skilled research assistant to ensure the reliability of the test.

Statistical methods

The SPSS statistical software (version 20.0 for Windows; SPSS, Inc., Chicago, IL, USA) was used for statistical analyses. The Kolmogorov–Smirnov test was used to determine the normality of distribution. The normally distributed continuous variables were calculated as the mean and SDs, and the non-normally distributed continuous variables were described as the median and interquartile range. The one-way analysis of variance (ANOVA) and the chi-square test were used to determine whether there were differences between general characteristics (age, height, weight, etc.) and demographic parameters (male-to-female ratio and dominant side) among these four groups.

The multivariate of analysis for repeated measurements (MANOVA) test (time*group) was used to determine whether there were differences in primary outcomes and secondary outcomes among groups. Moreover, the Tukey correction was used for multiple comparisons, and the partial η² was used to calculate for effect sizes. Statistical significance was set at 0.05 for analyses. Additionally, for post hoc test, the paired samples t-tests for within-group analysis.

Results

A total of 125 eligible older participants were recruited for this study and randomized into four groups. During the interventions, one participant in TC-RT group withdrew due to other medical reasons (not related to the intervention and measurement of this study), and another participant in the control group withdrew from the study due to a short-term absence from Shanghai and inability to participate in the post-intervention measurement. Ultimately, a total of 123 participants completed all the interventions as well as the assessments. Most of the participants reported significant lower extremity muscle soreness at the beginning of the intervention, and the fatigue was basically eliminated after one day of rest, except for this, there were no special adverse effects reported in each group. The baseline demographic characteristics were shown in the Table 2, and there were no significant differences in age, height, weight, BMI, male-female ratio, and dominant side ratio among these groups (p > 0.05). The data of 2 participants who dropped out were not included in the statistical analysis.

Table 2.

Baseline demographic characteristics of participants in groups

	SF-RT (n = 31)	TC-RT (n = 29)	RT (n = 32)	control (n = 31)
Age (years)	66.2 ± 4.5	67.7 ± 4.3	67.5 ± 4.6	66.9 ± 4.9
Height (cm)	161.3 ± 6.1	159.9 ± 7.6	159.3 ± 8.4	161.3 ± 8.0
Weight (kg)	62.2 ± 9.0	63.3 ± 8.7	63.1 ± 9.8	64.1 ± 9.7
BMI(kg/m2)	23.9 ± 3.2	24.7 ± 2.8	24.9 ± 3.6	24.7 ± 4.2
Sex (female/male)	5/26	5/24	7/25	8/23
Dominant side (left/right)	3/28	2/27	1/31	2/29

All the data were expressed as means ± SD or ratio

Abbreviation: SF-RT The short-foot training combined the lower extremity resistance training group; TC-RT The towel-curl training combined the lower extremity resistance training group; RT The lower extremity resistance training group; BMI Body mass index

Postural stability of SOT, LOS, and MCT

Table 3 showed there were significant interaction effects found in equalization scores of conditions 1 and 6 (condition 1, F = 3.561, p = 0.017, the partial η² = 0.092; condition 6, F = 2.734, p = 0.047, the partial η² = 0.068), strategy score of condition 6 (F = 2.811, p = 0.034, the partial η² = 0.069), as well as composite score of SOT (F = 2.901, p = 0.038, the partial η² = 0.073). And significant time effects were found in equalization scores of conditions 3, 4, and 5 (condition 3, F = 13.495, p < 0.001, the partial η² = 0.110; condition 4, F = 6.205, p = 0.014, the partial η² = 0.050; condition 5, F = 19.356, p < 0.001, the partial η² = 0.151), strategy score of condition 3 (F = 4.494, p = 0.036, the partial η² = 0.038) and composite score (F = 31.123, p < 0.001, the partial η² = 0.219). In addition, the following post hoc tests showed significant difference between groups (Fig. 2), respectively, the improvement in equalization score of condition 1 between SF-RT and control groups (mean difference [MD] = 1.43, 95% confidence interval [95%CI] = 0.13 to 2.74, p = 0.025), the improvement in equalization score of condition 6 between SF-RT and RT groups (MD = 8.99, 95%CI = 0.47 to 17.51, p = 0.035), the improvement in strategy score of condition 6 between SF-RT and RT groups (MD = 5.21, 95%CI = 0.16 to 10.26, p = 0.041), and the improvement in composite score between SF-RT and control groups (MD = 3.21, 95%CI = 0.17 to 6.48, p = 0.035).

Table 3.

Changes in the corresponding parameters of SOT before and after interventions in four groups

		SF-RT(n = 31)	TC-RT(n = 29)	RT(n = 32)	control(n = 31)	p value(time)	p value(interaction)
Equalization Score 1	Pre-intervention	91.90 ± 1.76	92.69 ± 1.85	92.00 ± 1.93	92.44 ± 1.85	0.13	0.02
	Post-intervention	93.17 ± 1.86*	92.72 ± 2.21	91.97 ± 3.11	92.27 ± 1.94
Equalization Score 2	Pre-intervention	90.82 ± 2.60	90.36 ± 3.04	89.86 ± 3.57	91.04 ± 2.61	0.52	0.28
	Post-intervention	91.26 ± 2.08	91.16 ± 2.50	89.87 ± 3.79	90.46 ± 3.36
Equalization Score 3	Pre-intervention	84.68 ± 4.90	84.77 ± 5.79	84.23 ± 6.71	84.63 ± 5.19	<0.001	0.13
	Post-intervention	88.33 ± 3.35*	87.05 ± 4.72	85.32 ± 6.51	85.08 ± 4.93
Equalization Score 4	Pre-intervention	80.25 ± 10.63	79.21 ± 8.63	74.04 ± 13.24	79.95 ± 11.46	0.01	0.43
	Post-intervention	82.23 ± 8.02	83.15 ± 7.03*	77.77 ± 12.73	79.89 ± 10.31
Equalization Score 5	Pre-intervention	64.73 ± 6.50	62.80 ± 9.54	62.17 ± 10.35	64.48 ± 7.83	<0.001	0.37
	Post-intervention	70.81 ± 8.22*	67.55 ± 10.29*	66.09 ± 9.91*	65.90 ± 9.23
Equalization Score 6	Pre-intervention	56.59 ± 12.75	59.33 ± 12.96	61.64 ± 11.92	56.28 ± 13.46	0.06	0.05
	Post-intervention	63.31 ± 10.13*	60.53 ± 10.42	60.38 ± 14.50	57.70 ± 13.41
Strategy Score 1	Pre-intervention	94.94 ± 2.64	95.75 ± 1.48	94.86 ± 2.49	95.06 ± 2.49	0.64	0.10
	Post-intervention	95.66 ± 1.28	95.24 ± 2.03	94.57 ± 2.10	95.52 ± 1.37
Strategy Score 2	Pre-intervention	94.39 ± 2.05	94.03 ± 2.44	93.59 ± 3.61	94.13 ± 2.23	0.14	0.78
	Post-intervention	93.95 ± 2.14	93.77 ± 2.37	93.58 ± 3.10	93.45 ± 3.05
Strategy Score 3	Pre-intervention	91.31 ± 5.48	91.05 ± 7.29	91.29 ± 4.18	91.37 ± 4.54	0.04	0.56
	Post-intervention	93.11 ± 3.12*	92.30 ± 3.41	91.49 ± 4.66	91.88 ± 3.64
Strategy Score 4	Pre-intervention	86.44 ± 4.75	85.80 ± 3.69	84.30 ± 5.82	86.00 ± 4.72	0.34	0.77
	Post-intervention	87.26 ± 2.84	86.63 ± 3.23	84.70 ± 6.50	85.67 ± 5.39
Strategy Score 5	Pre-intervention	76.63 ± 6.54	76.64 ± 8.12	77.14 ± 8.03	75.96 ± 6.91	0.23	0.49
	Post-intervention	79.14 ± 5.61	77.24 ± 8.12	76.68 ± 9.08	76.70 ± 6.08
Strategy Score 6	Pre-intervention	73.57 ± 9.96	74.51 ± 8.73	76.55 ± 8.74	75.36 ± 7.84	0.21	0.04
	Post-intervention	77.92 ± 5.46*	74.61 ± 8.75	75.69 ± 7.92	75.35 ± 7.43
Composite Score	Pre-intervention	74.52 ± 5.12	74.28 ± 5.73	73.43 ± 7.03	74.04 ± 5.94	<0.001	0.04
	Post-intervention	78.68 ± 5.13*	77.00 ± 4.68	75.10 ± 7.03	74.88 ± 6.21

All the data were expressed as means ± SD. *, compared with pre-intervention measurement (significant difference within group), p < 0.05

Abbreviation: SOT Sensory organization test; SF-RT The short-foot training combined the lower extremity resistance training group; TC-RT The towel-curl training combined the lower extremity resistance training group; RT The lower extremity resistance training group

Fig. 2.

The differences of outcomes on SOT among groups. Abbreviation: SOT, sensory organization test; SF-RT group, the short-foot training combined the lower extremity resistance training group; TC-RT group, the towel-curl training combined the lower extremity resistance training group; RT group, the lower extremity resistance training group; #, SF-RT group vs control group, p < 0.05; ##, SF-RT group vs RT group, p < 0.05

In LOS, the weighted average of RT, MVL, EPE, MXE, and DCL in four main directions were analyzed. Among these outcomes, significant time*group interaction effects were found on the MVL of right direction, left direction, and a composite score (right direction, F = 3.976, p = 0.010, the partial η² = 0.094; left direction, F = 3.192, p = 0.026, the partial η² = 0.076; composite score, F = 4.051, p = 0.009, the partial η² = 0.091). As shown in Fig. 3, the post hoc tests revealed that there were significant between-group differences in the MVL of right direction, left direction, and a composite score in the SF-RT group compared with the control group (right direction, MD = 1.01°/s, 95%CI = 0.19 to 1.84, p = 0.009; left direction, MD = 0.83°/s, 95%CI = 0.02 to 1.63, p = 0.041; composite score, MD = 0.64°/s, 95%CI = 0.07 to 1.21, p = 0.022). Moreover, significant time effects were also found for several outcomes, respectively, RT of four directions and composite score, EPE of three directions (forward, right direction, and left direction) and composite score, as well as MXE of three directions (forward, right direction, and left direction) and composite score (all p < 0.05).

Fig. 3.

The differences of outcomes on limits of stability test among groups. Abbreviation: SF-RT group, the short-foot training combined the lower extremity resistance training group; TC-RT group, the towel-curl training combined the lower extremity resistance training group; RT group, the lower extremity resistance training group; #, SF-RT group vs control group, p < 0.05

In MCT, participants were exposed to sudden postural perturbations in the anterior-posterior direction. There were only significant time effects found in Latency and Amplitude Scaling during posterior postural perturbation trial at high speed (Latency: p = 0.040; Amplitude Scaling: p = 0.020).

Foot muscle strength and morphology

The changes in foot muscle strength as well as morphology in the four groups before and after the 8-week intervention are shown in Table 4. The results showed that there was a significant increment in the time factor for all foot muscle strength parameters (all p < 0.05). As shown in Fig. 4, the results of between-group analysis showed that the IFMs strength (doming) significantly improved in the SF-RT group compared with control group (MD = 0.13 N/kg, 95%CI = 0.02 to 0.24, p = 0.017). For the T₁, TC-RT had significantly higher T₁ strength as compared to the RT group as well as to the control group (vs. RT, MD = 0.18 N/kg, 95%CI = 0.03 to 0.33, p = 0.011; vs. control, MD = 0.19 N/kg, 95%CI = 0.04 to 0.35, p = 0.007). Similar strength improvement in T₂₃ were found between TC-RT and the RT group or the control group (vs. RT, MD = 0.21 N/kg, 95%CI = 0.04 to 0.37, p = 0.008; vs. control, MD = 0.20 N/kg, 95%CI = 0.04 to 0.37, p = 0.011). And significant differences on T₂₃₄₅ were detected between SF-RT and control group, TC-RT and control group, as well as TC-RT and RT group (SF-RT vs. control, MD = 0.17 N/kg, 95%CI = 0.01 to 0.33, p = 0.042; TC-RT vs. RT, MD = 0.17 N/kg, 95%CI = 0.01 to 0.34, p = 0.040; TC-RT vs. control, MD = 0.19 N/kg, 95%CI = 0.02 to 0.36, p = 0.020).

Table 4.

Changes in foot muscle strength and morphology before and after interventions in four groups

		SF-RT(n = 31)	TC-RT(n = 29)	RT(n = 32)	control(n = 31)	p value(time)	p value(interaction)
Doming (N/kg)	Pre-intervention	0.60 ± 0.17	0.58 ± 0.20	0.61 ± 0.19	0.64 ± 0.17	<0.001	0.02
	Post-intervention	0.75 ± 0.22*	0.62 ± 0.16	0.66 ± 0.13	0.66 ± 0.16
T1 (N/kg)	Pre-intervention	0.62 ± 0.16	0.59 ± 0.16	0.64 ± 0.21	0.67 ± 0.14	<0.001	0.01
	Post-intervention	0.69 ± 0.19	0.77 ± 0.14*	0.65 ± 0.20	0.69 ± 0.21
T23 (N/kg)	Pre-intervention	0.72 ± 0.19	0.68 ± 0.15	0.74 ± 0.24	0.79 ± 0.23	0.04	0.02
	Post-intervention	0.78 ± 0.21	0.82 ± 0.18*	0.72 ± 0.20	0.77 ± 0.26
T2345 (N/kg)	Pre-intervention	0.77 ± 0.21	0.72 ± 0.14	0.82 ± 0.30	0.83 ± 0.27	0.001	0.01
	Post-intervention	0.89 ± 0.26*	0.86 ± 0.19*	0.81 ± 0.28	0.84 ± 0.27
AbH	Pre-intervention	5.07 ± 0.84	5.31 ± 1.29	5.52 ± 1.22	5.63 ± 0.83	0.01	0.04
Thickness(mm/m)	Post-intervention	5.53 ± 1.03*	5.78 ± 1.24*	5.54 ± 0.91	5.59 ± 0.89
AbH	Pre-intervention	0.88 ± 0.21	0.90 ± 0.24	0.93 ± 0.21	0.96 ± 0.18	0.04	0.01
CSA(cm2/m)	Post-intervention	0.93 ± 0.24	1.00 ± 0.24*	0.94 ± 0.24	0.92 ± 0.16
FDB	Pre-intervention	4.80 ± 0.67	5.19 ± 0.86	5.08 ± 1.06	4.89 ± 0.87	0.52	0.19
Thickness(mm/m)	Post-intervention	5.10 ± 0.81	5.23 ± 0.07	5.10 ± 0.70	4.91 ± 0.79
FDB	Pre-intervention	0.93 ± 0.21	0.96 ± 0.23	0.98 ± 0.19	0.98 ± 0.16	0.51	0.41
CSA(cm2/m)	Post-intervention	0.99 ± 0.18	0.98 ± 0.18	0.96 ± 0.19	0.97 ± 0.18
QP	Pre-intervention	4.02 ± 0.84	4.17 ± 1.43	4.34 ± 1.43	4.36 ± 0.91	0.25	0.04
Thickness(mm/m)	Post-intervention	4.48 ± 1.38*	4.29 ± 1.31	4.31 ± 1.37	4.21 ± 0.81
QP	Pre-intervention	0.63 ± 0.18	0.63 ± 0.23	0.73 ± 0.26	0.75 ± 0.14	0.57	0.30
CSA(cm2/m)	Post-intervention	0.96 ± 0.21	0.61 ± 0.20	0.73 ± 0.19	0.73 ± 0.16
FHB	Pre-intervention	5.31 ± 0.94	5.47 ± 1.02	5.50 ± 1.54	4.98 ± 0.78	0.99	0.36
Thickness(mm/m)	Post-intervention	5.70 ± 1.11	5.44 ± 0.74	5.43 ± 1.09	4.68 ± 0.63
FHB	Pre-intervention	0.66 ± 0.21	0.65 ± 0.15	0.64 ± 0.19	0.63 ± 0.16	0.26	0.79
CSA(cm2/m)	Post-intervention	0.72 ± 0.16	0.65 ± 0.12	0.66 ± 0.20	0.64 ± 0.19

All the data were expressed as means ± SD. *, compared with pre-intervention measurement (significant difference within group), p < 0.05

Abbreviation: T₁ The 1 st toe flexion strength; T₂₃ The 2nd–3rd toe flexion strength; T₂₃₄₅ The 2nd–5th toe flexion strength; AbH Abductor hallucis; FDB Flexor digitorum brevis; QP Quadratus plantae; FHB Flexor hallucis brevis; CSA Cross-sectional area; SF-RT The short-foot training combined the lower extremity resistance training group; TC-RT The towel-curl training combined the lower extremity resistance training group; RT The lower extremity resistance training group

Fig. 4.

The differences of foot muscles strength among groups. Abbreviation: SF-RT group, the short-foot training combined the lower extremity resistance training group; TC-RT group, the towel-curl training combined the lower extremity resistance training group; RT group, the lower extremity resistance training group; #, SF-RT group vs control group, p < 0.05; *, TC-RT group vs RT group, p < 0.05; **, TC-RT group vs control group, p < 0.05

For the foot muscle morphology, there were significant time*group interaction effects found in the thickness and CSA of AbH, and the thickness of QP (thickness of AbH, F = 4.353, p = 0.006, the partial η² = 0.073; CSA of AbH, F = 4.219, p = 0.007, the partial η² = 0.102; thickness of QP, F = 2.688, p = 0.048, the partial η² = 0.073), while other foot muscle morphologic parameters did not change significantly.

Balance and walking ability

There was only a significant time effect before and after the intervention for BBS and TUGT outcomes among these four groups (Berg, F = 24.42, p < 0.001, the partial η² = 0.035; TUGT, F = 4.357, p = 0.039, the partial η² = 0.035) without a significant interaction effect.

Discussion

In this study, we conducted 8-week interventions programs including IFMs training combined with lower extremity resistance training to investigate the effects of different IFMs training on postural stability, mobility, foot muscle strength and morphology in the elderly. The participants performed SF training exhibited improved postural stability, as shown in the improvements of static postural control outcomes (equalization score and strategy score) in SOT test and dynamic postural control outcomes (movement velocity) in LOS test. In addition, participants with different IFMs training developed growth with different kinds of IFMs strength and morphology, as follows significant improvements of IFMs strength (arching) and T₂₃₄₅ strength in SF training whereas improvements of toe flexion strength (T₁, T₂₃, and T₂₃₄₅) in TC training.

Among multitude risk factors for falls, gait alterations as well as balance deficits are stronger determinants which more commonly in geriatric population. As reported in previous studies [12, 32], the decline of IFMs strength with age might lead to a decline in mobility in the elderly from two aspects, including a decline in posture control during standing and a decline in the ability to generate propulsion during activity [33]. The foot, as the only structure in direct contact between human body and the ground, is an important component in the postural control mechanism. Our results showed that SF training, as an IFMs training, combined with lower limb strength training would effectively improve postural stability in healthy elderly population. In addition to the overall postural perturbation parameters, we found that the participants in RT group tended to adopt hip strategy to actively cope with postural perturbations, as measured by the strategy score of SOT. However, the results in the SF-RT group presented a significant increasement in the strategy score after the addition of IFMs training and a significant difference compared with RT group, suggesting that the addition of IFMs training resulted in older participants relying on the distal muscles of the body to maintain postural stability. Interestingly, there was a significant difference in equilibrium scores between these two groups, suggesting that the addition of IFMs training to conventional resistance training reduced the range of COP swing during perturbations in older adults and effectively improved their postural stability. This finding is consistent with a previous study by Kao et al. [15].

Including the study mentioned here, limited studies have targeted older adults for interventional training related to foot muscles and explored the effects on their balance function and mobility. In an earlier study, mobility (measured by functional reaching test) as well as toe dorsiflexion strength was significantly improved in elderly who performed foot and ankle-based intervention training for 8 weeks [34]. Another study devised a 12-week progressive foot and ankle exercise program, and its results showed the balance function of older adults who participated in the exercise improved as evidenced by an increase in their unilateral stance [35]. Moreover, a meta-analysis study has shown that functional foot training might help improve balance and mobility in older adults; however, due to inconsistencies in intervention programs and the fact that most of them were comprehensive foot-ankle health interventions, it is not yet possible to determine the specific effects of functional foot training on older adults’ related functions [16, 36].

In this study, two techniques of foot muscle training commonly utilized in clinical setting were implemented to improve foot function, with the primary distinction being the specific muscle groups targeted during execution [37]. Short-foot training involves a movement in which IFMs contract to approximate the first metatarsophalangeal joint toward the calcaneus, while simultaneously elevating the medial longitudinal arch. This technique emphasizes maintaining the toes in a non-flexed position throughout the movement. By shortening the foot via intrinsic muscle activation, this method effectively engages intrinsic musculature such as AbH, FDB and QP. In contrast, the towel-curl training involves flexing the toes to grasp and pull a towel or similar object, primarily activating the flexor hallucis longus and flexor digitorum longus muscles. Further imaging studies have shown that different foot movements tended to involve different kinds of foot muscles with different levels of activation [38], while the SF training was recommended as an effective method to improve the neuromuscular control of IFMs. Therefore, the present study explored the role of two different foot muscle training separately, rather than a combination of IFMs and EFMs training, which is different from previous studies [15, 34, 35]. And our results suggested the contribution of IFMs training to postural stability in older adults, which was consistent with our previous EMG study [14].

Based on current understanding of foot core system, we speculated that the promotion of postural control in the elderly by IFMs function training might be associated with foot arch support, neuromuscular control of foot muscles, and foot sensory input. Firstly, IFMs as integral functional units, assist in resisting subtalar pronation, which was presented as calcaneus valgus combined with medial displacement of scaphoid bone and increased vertical height. There was a more widely accepted explanation, that is, enhancing the strength of IFMs increases the support for foot arch and provides mechanical support for human posture stability by improving the alignment of lower extremity and the coordination between the muscles [11]. The elderly population exhibits a marked decline in foot muscle function, which is associated with reduced postural stability and mobility [12, 33]. Moreover, the elderly employs higher co-contraction of foot and ankle muscles, adopt their stiffening strategy to maintain postural stability [33]. The intervention program used in this study, which involves training of the muscles around hip, ankle, and foot arch, may have improved stability by increasing the aforementioned mechanical support.

A biomechanical study has demonstrated that during static weight-bearing tasks, after neuromuscular blockade of IFMs, the calcaneus-metatarsal angle exhibits relatively modest increasement, indicating the limited effect of IFMs on arch morphology [39]. Therefore, several researchers have started to investigate whether foot muscles affect posture stability through alternative mechanisms. As demonstrated by previous studies, IFMs exert a direct and independent facilitatory effect in the neuromuscular control mechanism responsible for regulating static postural stability, thereby enabling the body to effectively address postural perturbations and enhance overall stability [13, 14, 40]. In the study conducted by Okai et al. [41], they conducted electrical stimulation experiments on foot muscles to verify their role in posture control. The results indicated that electrical stimulation-induced contraction of FDB could result in a distinct anterior displacement of COP without activating the muscles of the lower extremity. This finding highlights the independent neuromuscular control role of IFMs in posture regulation. Consequently, we propose that targeted training of the intrinsic foot muscles to enhance their neuromuscular control capabilities may directly improve posture control ability in older adults.

In addition, these relatively small IFMs could provide immediate sensory feedback regarding changes in foot arch posture via the stretch reflex [11]. In previous studies, several researchers conducted fatigue of IFMs by repeatedly contracting the metatarsophalangeal joints, and found a significant reduction in the navicular drop index and proprioception of other lower limb regions during standing [42, 43], indicating that the functionality of IFMs may influence foot posture and even the proprioception of other lower extremity. Conversely, functional training of IFMs may potentially enhance postural stability by increasing the sensitivity, as well as the muscle and tendon receptors surrounding the ankle. Although all participants in this study were elderly individuals with normal plantar sensation, it cannot be excluded that IFMs training may enhance the sensitivity of intramuscular receptors. However, there is currently a lack of validated methods for assessing the sensitivity of these receptors. Future research could further investigate the facilitative effects of IFMs training on sensory input.

The MCT test was employed to determine the effect of IFMs training on dynamic postural control in older adults. No significant differences in dynamic postural stability were found among these four intervention groups, despite the presence of time effects suggesting a trend toward improvement for several parameters of MCT. Previous EMG studies of static postural tasks have shown a correlation between muscle recruitment in IFMs and postural oscillations of COP in medial-lateral direction and anterior-posterior direction of static postural tasks, including the AbH, FDB, and QP [14, 44]. As for dynamic postural control, there was limited study designed to determine the relationship of recruitment in IFMs and postural oscillations of COP. During the process of regulating human body posture, the muscle activities of the ankle joint and the hip joint generate torques to resist posture disturbances [45]. Therefore, we hypothesized that this might have contributed to the lack of improvement in dynamic postural stability in the older adults due to IFMs’ anatomical characteristics. As a consequence, we hypothesized that the foot muscles, which have a smaller muscle size compared to the ankle muscles, have a limited role to play in more difficult dynamic tasks, but rather assist in postural stabilization in less difficult static tasks. In addition, the latency and the amplitude scaling, the parameters of MCT, respectively represented the reaction time and the capacity to apply appropriate force and effectively counteract the perturbation of the support surface to restore postural stability, which primarily relied on somatosensory input [46]. The intervention training implemented in this study emphasized the functional capacity of lower limb muscles. However, it may not have been sufficient to elicit significant functional improvements in the elderly population, thus limiting its effectiveness in enhancing dynamic balance ability. Furthermore, the limited sample size and relatively short intervention duration in this study may have contributed to the inability to observe significant improvements in dynamic postural stability. Future research should include more multi-center, large-sample, randomized controlled trials, as well as in-depth electromyographic investigations, to further elucidate the role of IFMs in the dynamic postural stability of older adults.

Notably, although our study furtherly confirmed the positive role of IFMs training on postural stability in older adults, we conceded that this study has some limitations. Firstly, only elderly volunteers with normal foot function (posture, and plantar sensation) were included in this randomized controlled trial, rather than classifying elderly participants into subgroups. In addition, due to physical functional requirements of interventions used in this study, the community-dwelling older adults with no or low fall risk were recruited as participants, which might limit the application of our findings to population with high fall risk or hospitalized older adults. Consequently, the impact of intervention protocol utilized in this study on postural stability remains undetermined for patient populations beyond those specifically addressed herein. Moreover, considering the high variability of baseline physiological function and physical function of the elderly, such as frail or post-fall populations, future studies with larger cohorts and extended intervention periods are needed to develop more comprehensive and effective fall prevention programs and interventions to promote postural control.

Conclusion

Our study shows that 8-week fall prevention program, coupled with IFM training (short-foot training) and lower extremity resistance training, might be beneficial for older adults, as postural stability, mobility, foot muscle strength and morphology improved significantly. Importantly, commonly used as extrinsic foot muscle training, the towel-curl training combined with lower extremity resistance training did not promote significant postural stability improvement. These findings emphasize the efficacy of additional IFM training militated extra effect on improving postural stability in older adults.

Abbreviations

IFMs

Intrinsic foot muscles

EFMs

Extrinsic foot muscles

RT group

The lower extremity resistance training group

SF-RT group

The short-foot training combined the lower extremity resistance training group

TC-RT group

The towel-curl training combined the lower extremity resistance training group

COP

Center of pressure

SOT

Sensory organization test

LOS

Limits of stability test

MCT

Motor control test

COG

Center of gravity

RT

Reaction time

MVL

Movement velocity

EPE

Endpoint excursions

MXE

Max excursions

DCL

Directional control

BBS

Berg balance scale

TUGT

The timed up and go test

EFMs

Extrinsic foot muscles

AbH

Abductor hallucis

FDB

Flexor digitorum brevis

QP

Quadratus plantae

FHB

Flexor hallucis brevis

CSA

Ccross-sectional area

BMI

Body mass index

Author contributions

ZL and LW contributed to the conception and design of the trial and drafted the manuscript. MC and PG participated in trial registration, communication, and monitoring. RW, and GZ participated in the statistical analysis design. All authors read and approved the final version of the manuscript.

Funding

This work was supported by the key Project of Zhejiang Administration of Traditional Chinese Medicine Co-construction Science and Technology Program (No. GZY-ZJ-KJ-24075) and the research special project of Zhejiang University of Traditional Chinese Medicine Affiliated Hospital in 2023 (No. 2023FSYYZQ10).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the Shanghai University of Sport (Ref. No.: 102772020RT001) and all procedures were conducted according to the Declaration of Helsinki. All participants were informed about the study procedures and given written informed consent in advance.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.