Effects of low-frequency vibration training on walking ability and body composition among older adults: a randomized controlled trial

Study design

Participants were randomly assigned to either the experimental or control groups using a lottery method, without stratifying by sex. Group allocation cards labeled “experimental” or “control” were placed in a concealed container, and participants drew one card to determine their group assignment. The process was conducted under the supervision of a single supervisor, who ensured the proper implementation of the randomization procedure. Moreover, to reduce bias, outcome assessors were blinded to the participants’ group assignments. All participants voluntarily provided written informed consent, and personal information was kept confidential to ensure privacy and security.

Participants

Conducted using G*Power (version 3.1), the a priori power analysis determined that a minimum sample size of 34 participants is required to achieve 80% power for a medium effect size (f = 0.25) in a 2 × 2 repeated measures ANOVA (Cohen, 1988). This calculation aligns with a significance level of α = 0.05, following Cohen’s (1988) guidelines. All participants were residents of Shanghai, China, at least 75 years of age, and could execute the tasks and walk independently. Exclusion criteria were (1) stroke, severe cardiac disease, or stent placement that was not appropriate for the level of required exercise; (2) spinal cord injury, endogenous osteosynthesis, knee or hip replacement, pacemaker, cardiovascular illness, and epilepsy; and (3) osteoarthritic conditions that limited ability to exercise, osteoporosis, and physical or cognitive disorders that would hamper training and testing procedures. Participants were randomly assigned to vibration group (n = 25) and control group (n = 25). A drop out in vibration group occurred due to health problems (n = 3) and loss of interest (n = 2) and in the control group due to personal reasons (n = 2), health problem (n = 5) (Fig. 1). The study was approved by the Ethics Committee of Shanghai University of Sport (102772022RT123).

Intervention

Participants in the vibration group stood on a vibration platform (Body Green, Taiwan) three times a week for 8 weeks while maintaining a knee flexion angle of 40–60 degrees (Fig. 2). The platform had handrails for safety. Based on previous studies (Wadsworth & Lark, 2020), we designed the vibration intervention protocol for this study. In our protocol, the vibration frequency for each session was initially set at 5 Hz and then gradually increased over the 8-week period (Table 1). This design aims to provide participants with progressively enhanced stimulation, thereby ensuring safety while improving the effectiveness of the training (Wadsworth & Lark, 2020). The vibration amplitude was maintained at two mm. The vibration frequency used for each session began at 5 Hz but then increased across the 8 weeks as shown in Table 1. The vibration lasted for 1 min each bout, starting with five bouts per session for weeks 1–4 and then increasing to 10 bouts for weeks 5–8. Participants rested for 1 min on the platform between vibration bouts. After vibration training, participants performed 5 min of stretching and relaxation exercises, including two sets of static stretching for the lower limbs (hamstrings, quadriceps, and calves), each held for 15 s per side, followed by two minutes of diaphragmatic breathing exercises to promote relaxation. The control group maintained their normal daily routine, with activities of daily living monitored by an investigator on a weekly basis.

Walking ability

Walking ability is defined as the capacity to generate displacement by utilizing lower limb muscle strength to support body weight against gravity and provide the necessary forward and upward propulsion during ambulation. This ability facilitates critical transitional movements while maintaining dynamic balance by keeping the center of gravity within the base of support (Suwannarat et al., 2021). Walking ability was evaluated using the 30-second Chair Stand Test (30-s CST), Timed Up and Go (TUG) test and the 6 m walk speed.

The 30-s CST is the most widely used test method for assessing lower limb muscular strength and its role in transitional movements during ambulation, such as sit-to-stand and stand-to-walk transitions among older individuals in a community (intraclass correlation coefficient (ICC), 0.84 for men, and 0.92 for women) (Barrios-Fernández et al., 2020). Participants sat in a chair with the back straight, feet shoulder-width apart, knees bent at 90 degrees, and arms crossed in front of the chest. The number of times that a participant stood from a seated position and sat down again within 30 s was recorded. The average of two tests was used to calculate the final exam score (Jones, Rikli & Beam, 1999).

The TUG test is a dynamic balance and mobility test with a high rate of retest reliability (ICC = 0.99) (Suwannarat et al., 2021). TUG measures a person’s ability to maintain dynamic balance while moving about and changing positions. Participants sat in a chair until a tester said “go”, at which time the participant stood and walked three m at their usual walking speed before returning to sit in the chair (Gadelha et al., 2018). The tester noted any assistance needed and recorded the time in seconds it took to complete the test. The average of two measurements used to determine the outcome.

Walking speed, which has a high retest reliability (ICC of 0.92) (Lyons et al., 2015), provides the strongest diagnostic metric for evaluating walking ability in older people (Tai et al., 2021). The first and last two m of a 10-m open space are not included in the test to avoid acceleration and deceleration effects. Participants walked 10 m from a marked starting line to a marked finish line. When the participant’s front toe crossed the two-meter line, the tester started a stopwatch and stopped it when the participant’s toe reached the eight-m line (Zhao, Wu & Wei, 2020). Participants were asked to walk at their maximum pace. The final score was determined by averaging the results of two tests.

Body composition

Body composition was evaluated using the height and weight, body mass index (BMI) and body fat percentage, and circumference. Height and weight were obtained using an SH-200 automatic ultrasonic height and weight measuring device (also known as a medical height and weight scale), an electronic and an electronic scale from AVIC Precision Pressure Sensors to measure weight. BMI and body fat percentage were determined through bioelectrical impedance analysis using a body composition analyzer (Tanita Corporation, MC-980MA, Tokyo, Japen). The instrument has a total of eight contact electrodes to assess a variety of bodily signs quickly enough to perform the test. A soft leather tape measure was used to measure the circumference of the waist and hips to one decimal point (0.1 cm). The circumference measurement method estimates body fat content through the measurement of the circumference of the body parts where subcutaneous fat accumulates more, typically the waist circumference (WC) and hip circumference (HC). WC divided by HC is the waist-to-hip ratio (WHR).

Statistical analyses

Independent samples were assessed, and the findings were analyzed using SPSS version 26.0 software (IBM Corp., Armonk, NY, USA). The pre-intervention baseline data were analyzed using a t-test. The temporal within-group effects and differences between the vibration and control groups were examined using repeated-measures analyses of variance (ANOVAs). If there was a statistically significant interaction between group and time, the data were further analyzed using simple effects test. If the data fail to meet the assumption of normal distribution, non-parametric tests should be utilized for statistical analysis. Effect size classifications: trivial (η_p² < 0.01, minimal impact), small (0.01 ≤η_p² < 0.06, weak), medium (0.06 ≤η_p² < 0.14, moderate), and large (η_p² > 0.14, substantial) (Cohen, 1988). A significance level of p < 0.05 was established.

Participant characteristics

Demographic characteristics for the two participant groups are presented in Table 2 and Table 3. Comparisons of age, walking ability, and body composition between the cohorts revealed no statistically significant differences, indicating that the participant groups were equivalent for the purposes of this study.

Effect of vibration training on walking ability of older adults

With group as the between-subjects factor (VG and CG) and time as the within-subjects factor, a two-way repeated measures ANOVA was conducted to investigate the trend of variation. In the 30-s CST, the results showed that the main effect of the group (F_(1,36) = 0.05, p = 0.81, η_p² = 0.002) was not significant (Fig. 3). In walking speed, the main effect of the group (F_(1,36) = 4.50, p = 0.04, η_p² = 0.11) was significant. In TUG test, using the non-parametric Mann–Whitney test, after the intervention, there was a significant difference between the experimental group and the control group (z =  − 2.72, p = 0.007) (Fig. 4).

Effects of vibration training on body composition

In the WC, the results revealed that the main effect of the group (F_(1,36) = 0.80, p = 0.37, η_p² = 0.02) and the interaction effect (F_(1,36) = 1.96, p = 0.16, η_p² = 0.05) were not significant. Moreover, the effect of time (F_(1,36) = 7.19, p = 0.01, η_p² = 0.16) was significant. The main effect of the group for HC (F_(1,36) = 0.06, p = 0.80, η_p² = 0.002) and WHR (F_(1,36) = 2.00, p = 0.16, η_p² = 0.05) were not significant, but the interaction effects for HC (F_(1,36) = 6.37, p = 0.01, η_p² = 0.15) and WHR (F_(1,36) = 9.08, p = 0.005, η_p² = 0.20) were significant. The main effect of the group for the WHR (F_(1,36) = 2.00, p = 0.16, η_p² = 0.05) (Fig. 5). The main effect of the group for the BMI (F_(1,36) = 0.84, p = 0.36, η_p² = 0.02) and body fat percentage (F_(1,36) = 0.48, p = 0.48, η_p² = 0.01) were not significant (Fig. 6).

Effects of vibration training on walking ability of older adults

The number of counts in the 30-s CST slightly increased in the vibration training group, while a decrease was observed in the control group. However, these changes did not reach statistical significance, indicating that the vibration training did not significantly improve lower limb muscle strength or endurance as measured by the 30-s CST. A study by Zhu et al. (2019) assessed lower limb muscle strength among older men (88.5 ± 3.7 years old) who held a vibrating rope with both hands and flexed their knees while slightly squatting after 5 weeks of vibration training (12–16 Hz, 3–5 mm, 5 times/week) (Zhu et al., 2019). Their results showed that whole-body vibration training group significantly improved lower limb muscle strength, with lower limb strength increased by 15.6%–18.2% following the intervention. The authors indicated that there may be gender differences in the effects of vibration training. A study by Faes et al. (2018) found that muscle activity did not increase with vibration training. However, that study assessed only the acute effects, which may not have provided sufficient time for the stimulation to make a difference. In addition, the participants in the study by Faes et al. (2018) were young adults (20 years or older), and the vibration frequency used was 8 ± 2 Hz (the present study used 5–13 Hz). A previous study showed that high-frequency vibration training had a greater impact than low-frequency vibration training on young people’s muscles (Sañudo et al., 2016). Unlike the study by Faes et al. (2018) that focused on a population with a high degree of physical function, our study examined persons 75 years or older, a population with poorer physical function but with the potential for substantial improvement. According to the findings of the study by Sañudo et al. (2016), vibration training enhances blood flow to the body, increases blood flow, quickens cellular metabolism, and boosts the body’s capacity for sustained movement during exercise (Sitjà-Rabert et al., 2015).

The time needed to perform the TUG test after 8 weeks of low-frequency vibration was significantly reduced in older adults compared with both before the training and with the control group. These findings suggested that low-frequency vibration training improved the dynamic balance of older adults. Tsuji et al. (2014) reported that short-term whole-body vibration training improved functional mobility more than training without vibration stimulation in a TUG assessment; however, this benefit persisted only for 30 min (Tsuji et al., 2014).

Vibration training improves ankle and knee stability (Moezy et al., 2008), which can enhance postural control and is connected with greater neuromuscular control and decreased joint stiffness. The vibration produced by the vibration platform is transmitted through proprioceptive receptors, and this transmission of mechanical vibration triggers the tonic vibration reflex, a complex spinal and supraspinal neurophysiological response. This reflex in turn activates muscles and improves physical performance, and vibration-stimulated haptic feedback and intrinsic proprioception assist in controlling postural stability (Fischer et al., 2019). Capicíková et al. (2006) demonstrated that the muscles of the lower limbs of the leg can alter the state of postural balance during standing and are dependent on the activation and duration of proprioceptive stimulation. Vibration training stimulates the body’s proprioceptors, promoting better balance by triggering brain plasticity and maintaining the body’s postural position. This is so because during standing, some muscles are engaged to a certain extent, but while vibrating, all muscles are always engaged.

We found that 8 weeks of low-frequency vibration training markedly increased the walking speed of older adults compared with both before the vibration training and with the control group. The improvement may depend on the individual’s baseline level. For example, a previous review reported that vibration training increased walking speed among older adults by 36%, while walking speed in the untrained group increased by 18.1%, suggesting that the starting level of individuals may be a significant influencing factor (Pollock, Martin & Newham, 2012). Increases in single-leg support time following vibration training and the body’s enhanced postural control are additional reasons why vibration training may improve walking. When the lower limbs are supported, walking becomes more similar to normal gait kinematics, and this stability forms the basis for the swing leg’s forward motion during walking. This lower limb support encourages an increased walking pace (Chan et al., 2012). The ankle joint requires plantarflexion as the foot is forced upward off the vibration platform during walking. An improvement in lower extremity joint range of motion caused by vibration training has been demonstrated (Szopa et al., 2021). Therefore, improved joint range of motion and ankle muscle strength may also contribute to the increase in walking speed. Joint mobility, however, was not assessed in the study and should be investigated in future studies.

Effects of low-frequency vibration training on the body composition

Vibration training did not show a statistically significant effect on BMI or body fat percentage. While a reduction in body fat mass was observed in the vibration group and a slight increase in the control group, these changes were not sufficient to conclude that vibration training effectively lowers BMI in older adults. Our findings are consistent with those of Jo et al. (2021) showing that vibration training had no appreciable impact on body fat percentage and BMI in older adults (Jo et al., 2021; Reis-Silva et al., 2023). However, most previous studies focused on older women; thus, gender should be investigated in future studies. Fjeldstad et al. (2009) (Fjeldstad et al., 2009)found that vibration training with a frequency of 15–40 Hz and an amplitude of three mm has a beneficial effect on body fat mass among postmenopausal women. Additionally, Pérez-Gómez et al. (2020) reported that vibration training (12–24 Hz, 3 times/week, three mm) decreased body fat percentage (4.4%) among middle-aged women. It is possible that vibration training alters the body’s metabolic rate, resulting in an increase in energy expenditure and an inhibition of adipogenesis (Cristi-Montero, Cuevas & Collado, 2013) as well as changes plasma levels of staphylococcal or senescence marker protein 30 (SMP30), which is primarily involved in hepatic lipid regulation and lipid metabolism in women (Kondo & Ishigami, 2016). Additionally, several studies have shown a positive correlation between changes in the resistance index, mean velocity, and peak velocity of blood flow with changes in body composition after vibration training, demonstrating that changes in blood flow are linked to changes in fat mass (Sañudo et al., 2013). The present study showed that low-frequency vibration training had a significant impact on WHR and WC. Dietary factors also impact changes in body composition, but diet was not taken into account in this study.

Study limitations

This study has limitations, including unmonitored physical activity and sedentary behavior, potential influence of control group activities, lack of participant blinding, absence of follow-up for long-term effects, and adherence rates of 80% in the vibration group and 72% in the control group, which may limit the interpretation and generalizability of the results. Finally, due to the high rate of data loss caused by health conditions or unforeseen events (e.g., illness or death), intention-to-treat analysis was not applied.