The mechanism of static postural control in the impact of lower limb muscle strength asymmetry on gait performance in the elderly

Participants

Using G*Power (version 3.1) for a priori analysis, it was determined that a minimum total sample size of 34 individuals is necessary to achieve a statistical power of 0.80 for a medium effect size (f = 0.25) in a 2 × 2 repeated measures ANOVA (Cohen, 1992), with a significance level of α = 0.05. Participant recruitment was conducted within communities in Yangpu District, Shanghai, and a total of 55 participants (23 males and 32 females) were enrolled, which met the study’s sample size requirements. The inclusion criteria included: 1) aged 60 to 75 years; 2) capable of independent living and unassisted walking; 3) no history of lower limb surgery; 4) no lower limb pain in the week preceding the test. Participants were excluded if they had: 1) severe cardiovascular or cerebrovascular diseases; 2) neurological disorders such as Parkinson’s disease, spinal conditions, or stroke; 3) visual or auditory impairments; 4) a history of severe lower limb injury within the past 6 months; 5) pain during the maximum strength test that prevented test completion. Informed consent was obtained from all participants after a thorough explanation of the potential risks and discomforts. The participants were divided into two groups: symmetrical strength (SG, n = 29) and asymmetrical strength (AG, n = 26), based on a 15% asymmetry threshold. The study protocol underwent scrutiny and received endorsement from the Ethics Committee of Shanghai University of Sport (No. 102772022RT123).

Lower limb muscle strength asymmetry

The “gold standard” for assessing human muscle strength (Kambič, Lainščak & Hadžić, 2020), Isokinetic muscular strength testing equipment (CON-TREX MJ, Physiomed elektromedizin AG, Schnaittach, Germany) was used to evaluate bilateral knee extensor strength. Before commencing the test, participants were instructed to engage in a 5-min warm-up session on a power cycle, followed by dynamic stretching of their lower limb muscles. Subsequently, the participants assumed a seated posture, and the torso and non-examined limb were fastened to the testing chair to ensure safety. Prior to commencing the formal test, participants were required to acquaint themselves with the range of motion and successfully complete five tests at maximal effort. As follows, the angular velocity of movement was set at 60°/s, and the range of motion was measured at 90°. The isokinetic concentric mode was used for testing. The measurable parameter was peak torque (Nm). The lower limb muscle strength asymmetry calculation was based on Laroche, Cook & Mackala’s (2012) work and set the threshold at 15% based on most scholarly definitions of asymmetry (Barber et al., 1990; Perry et al., 2007; Parkinson et al., 2021). More precisely, lower limb muscle strength asymmetry was defined as a difference of peak torque of knee extension of more than 15%.

$$\mathrm{A}\mathrm{s}\mathrm{y}\mathrm{m}\mathrm{m}\mathrm{e}\mathrm{t}\mathrm{r}\mathrm{y}\left(\mathrm{\%}\right)={\displaystyle \frac{\left|\left(\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{f}\mathrm{L}\mathrm{e}\mathrm{f}\mathrm{t}\mathrm{L}\mathrm{i}\mathrm{m}\mathrm{b}-\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{f}\mathrm{R}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{L}\mathrm{i}\mathrm{m}\mathrm{b}\right)\right|}{\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{f}\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{n}\mathrm{g}\mathrm{L}\mathrm{i}\mathrm{m}\mathrm{b}}\ast 100\mathrm{\%}}$$

Static postural control

The ability to maintain static postural control was assessed by using a force platform (Physio sensing, Sensing Future Technologies, Coimbra, Portugal). Participants were instructed to maintain a straight-ahead binocular gaze and to place both palms on their hips throughout the test. Each assessment protocol incorporated eyes open (EO) and eyes closed (EC) visual conditions, and comfortable standing (CS) and narrow standing (NS) widths (Riemann & Piersol, 2017; Scoppa et al., 2017). The sway velocity index (SVI) was determined once all protocol conditions had been met. The natural logarithm function was applied to normalise the result of dividing the participant’s height by the mediolateral velocity to calculate SVI (Riemann et al., 2018). Mediolateral velocity was calculated by dividing the displacement of the centre of pressure during the trial by the duration of the trial in milliseconds. The sway displacement in the mediolateral direction was calculated for every trial during every acquisition (100 acquisitions per second) and the software calculated the mean result at the conclusion. Moreover, the area of the ellipse (AOE) was calculated by utilising the trajectory of the pressure centre for each test.

Gait performance

Gait analysis utilised the plantar pressure gait assessment system (MedTrack, Xinkang Biomedical Technology Co., Hangzhou, CHN). The system comprises a computer, a pressure-sensitive walkway, and specialised analysis software, enabling the uninterrupted collecting of gait metrics at a sample frequency of 400 Hz. The supplementary computer accurately identifies and assesses the participant’s locomotion data to generate spatiotemporal parameters with the assistance of its integrated software. The dimensions of the walkway are 7.2 m in length and 1.0 m in width, with buffer zones of 1 m at both ends. The equipment used in this study is distinct from that in previous research, as it independently calculates the gait parameters for both the left and right limbs of subjects (with data analysis distinguishing between the strong and weak limbs), thereby facilitating a more detailed examination of the influence of muscle strength asymmetry on the spatiotemporal gait parameters in older adults.

Participants were instructed to remove their shoes and walk barefooted at a natural speed for the entire test duration, starting 1 m from the walkway. The test concluded when the participant reached the endpoint. All 14 gait parameters—step length; stride length; step time; stride time; stance time; swing time; single support time (SST); double support time (DST); percentage of stance time in gait cycle (STT%); percentage of swing time in gait cycle (SWT%); percentage of SST in gait cycle (SST%); percentage of DST in gait cycle (DST%)—were used for analysis. All the above parameters except for gait velocity and cadence included both the strong and weak limbs.

Statistical analyses

Experimental data were analysed with SPSS software (IBM, Armonk, NY, USA). A repeated measures analysis of variance (ANOVA) with a 2 × 2 design was used to investigate intergroup variations in data pertaining to comfortable/narrow stance conditions in the static postural control test. In addition, a comparative analysis was performed to examine the disparities in data pertaining to the strong and weak limbs during the gait performance test. Ultimately, Pearson correlation analysis was utilised to evaluate the association between the notable disparities revealed in the static postural control test and gait parameters. A significance level of P < 0.05 was established.

Participant characteristics

The demographic characteristics of the two participant categories are detailed in Table 1. Age, gender, height, weight, body mass index (BMI), preferred limb side, and the strong and weak knee extensor strength were statistically similar between the two cohorts, suggesting that all participants were comparable.

Gait performance

The independent samples t-test results comparing gait velocity and cadence are presented in Table 2. Gait velocity (t₍₅₃₎ = 3.557, P < 0.01, Cohen’s d = 0.97) and cadence (t₍₅₃₎ = 2.371, P < 0.05, Cohen’s d = 0.65) were significantly slower in AG than in SG.

Using SG and AG as between-group factors, and left and right limb lateralization as within-group factors, a repeated measures analysis of variance was employed to examine differences in spatiotemporal gait parameters between the strong and weak limbs within the two participant groups. As shown in Table 3, there was a significant intergroup main effect of step length (F_1,53 = 10.555, P < 0.01, η_p² = 0.166), with post hoc analysis revealing that the step lengths of both strong and weak legs were significantly shorter in AG than in SG (P < 0.01 and P < 0.01, respectively). A significant intergroup main effect was observed in stride length (F_1,53 = 6.849, P < 0.05, η_p² = 0.114); post hoc analysis indicated that both the strong and weak legs stride lengths were significantly shorter in AG than in SG (P < 0.05 and P < 0.05, respectively). In terms of step time, there was a significant intergroup main effect (F_1,53 = 19.278, P < 0.001, η_p² = 0.267), with post hoc analysis revealing that both strong and weak legs step times were significantly longer in AG than in SG (P < 0.05 and P < 0.001, respectively). We found a significant intergroup main effect of stride time (F_1,53 = 10.303, P < 0.01, η_p² = 0.163); post hoc analysis indicated that both strong and weak legs stride times were significantly longer in AG than in SG (P < 0.05 and P < 0.01, respectively). There was a significant intergroup main effect of stance time (F_1,53 = 22.420, P < 0.001, η_p² = 0.297); post hoc analysis showed that both strong and weak legs stance times were significantly longer in AG than in SG (P < 0.05 and P < 0.05, respectively); however, strong leg stance time was significantly shorter than weak leg stance time among AG subjects (P < 0.05). We observed significant intergroup main effects of swing time (F_1,53 = 5.130, P < 0.05, η_p² = 0.088); post hoc analysis revealed that the weak leg swing time in AG was significantly longer than that of SG (P < 0.05). SST demonstrated a significant intergroup main effect (F_1,53 = 5.012, P < 0.05, η_p² = 0.086); post hoc analysis indicated that weak leg SST was significantly greater in AG than in SG (P < 0.05). A significant intergroup main effect was noted for DST (F_1,53 = 31.794, P < 0.001, η_p² = 0.375), and post hoc analysis revealed that both strong and weak legs DST was significantly greater in AG than in SG (P < 0.05 and P < 0.05, respectively). STT%, exerted significant intragroup (F_1,53 = 8.799, P < 0.001, η_p² = 0.142). Post hoc analysis showed that weak side STT% were significantly greater than in AG than in SG (P < 0.05). We observed significant intergroup main effects of SWT% (F_1,53 = 15.584, P < 0.001, η_p² = 0.227). Post hoc analysis indicated that SWT% for both strong and weak sides were significantly smaller in AG than in SG (P < 0.01 and P < 0.01, respectively). A significant intergroup main effect was observed for SST% (F_1,53 = 8.530, P < 0.01, η_p² = 0.139); post hoc analysis showed a significantly smaller strong leg SST% in AG compared to SG (P < 0.05). DST% demonstrated was a significant intergroup main effect (F_1,53 = 20.412, P < 0.001, η_p² = 0.278); post hoc analysis indicated that DST% for both strong and weak sides was significantly greater in AG than in SG (P < 0.001 and P < 0.01, respectively).

Static postural control

ANOVA was performed to assess the differences in static postural control between two groups, with SG and AG as between-group factors and comfortable standing test and narrow stance test as within-group factors. The analysis aimed to investigate the variations across wide and narrow stance test tasks. Data are reported in Table 4, there was a significant intragroup main effect on EO-SVI (F_1,53 = 14.419, P < 0.001, η_p² = 0.214), and NSEO-SVI was significantly greater than CSEO-SVI (P < 0.01). There was a significant intergroup main effect of EO-SVI (F_1,53 = 14.223, P < 0.001, η_p² = 0.212); simple effect analysis indicated that under both standing distance conditions, EO-SVI significantly higher in AG than in SG (P < 0.05 and P < 0.01, respectively); however, NSEO-SVI was significantly greater than CSEO-SVI (P < 0.001). EC-SVI demonstrated inter-and intragroup interaction effects (F_1,53 = 8.137, P < 0.01, η_p² = 0.133); post hoc analysis revealed that under both standing distance conditions, EC-SVI was significantly greater in AG than in SG (P < 0.01 and P < 0.001, respectively). However, within SG, there was no significant difference (P > 0.05). Within AG, NSEC-SVI was significantly greater than CSEC-SVI (P < 0.001). Neither the main nor interaction effects of EO-AOE were significant (P > 0.05). Post hoc analysis revealed that under the NS condition, EO-AOE was significantly lower in SG compared to AG (P < 0.05). There were significant inter-and intragroup interactions effects on EC-AOE (F_1,53 = 11.342, P < 0.01, η_p²=0.176). Simple effect analysis indicated that under both testing conditions, EC-AOE was significantly larger in AG than in SG (P < 0.01 and P < 0.001, respectively). However, there was no significant difference within SG (P > 0.05). Within AG, NSEC-AOE was significantly larger than CSEC-AOE (P < 0.001).

Correlation between static postural control and gait performance

Correlations between static postural control and gait performance were examined through Pearson correlation analysis. As shown in Fig.1. In the CSEO condition, the SVI exhibited negative correlations with strong side stride length (r = −0.298, P = 0.027), weak side stride length (r = −0.343, P = 0.010). In the CSEC condition, SVI was negatively correlated with strong side step length (r = −0.301, P = 0.026); and weak side swing time (r = −0.295, P = 0.029); however, SVI showed positive correlations with strong and weak sides stance time (r = 0.302, P = 0.025; r = 0.328, P = 0.014); strong and weak side DST (r = 0.340, P = 0.011; r = 0.345, P = 0.010); weak side STT% (r = 0.295, P = 0.029); and weak side DST% (r = 0.271, P = 0.046). In the NSEO condition, SVI positively correlated with strong side step time (r = 0.440, P = 0.001); strong and weak sides stance time (r = 0.461, P = 0.000; r = 0.449, P = 0.001); strong and weak sides DST (r = 0.423, P = 0.001; r = 0.478, P = 0.000); strong and weak sides STT% (r = 0.459, P = 0.000; r = 0.428, P = 0.001); and strong and weak sides DST% (r = 0.387, P = 0.004; r = 0.331, P = 0.014). However, within NSEO condition, SVI negatively correlated with strong and weak sides SWT% (r = −0.459, P = 0.000; r = −0.428, P = 0.001). In the NSEC condition, SVI correlated negatively with gait velocity (r = −0.350, P = 0.009); cadence (r = −0.281, P = 0.038); strong side step length (r = −0.322, P = 0.016); weak side SWT% (r = −0.469, P = 0.000); and strong side SST% (r = −0.415, P = 0.002); however, SVI exhibited positive correlations with strong and weak sides step time (r = 0.345, P = 0.01; r = 0.519, P = 0.000); strong and weak sides stride time (r = 0.417, P = 0.002; r = 0.373, P = 0.005); strong and weak sides stance time (r = 0.477, P = 0.000; r = 0.546, P = 0.000); strong side swing time (r = 0.298, P = 0.027); weak side SST (r = 0.376, P = 0.005); strong and weak sides DST (r = 0.609, P = 0.000; r = 0.587, P = 0.000); weak side STT% (r = 0.469, P = 0.000); and strong and weak sides DST% (r = 0.430, P = 0.001; r = 0.437, P = 0.001). In the CSEO condition, AOE showed negative correlations with weak side step length (r = −0.377, P = 0.005); strong and weak sides stride length (r = −0.419, P = 0.001; r = −0.388, P = 0.033); strong side swing time (r = −0.313, P = 0.020); weak side SST (r = −0.301, P = 0.025); and strong side SWT% (r = −0.321, P = 0.017). However, in the CSEO condition, AOE showed positive correlations with strong side STT% (r = 0.321, P = 0.017). In the CSEC condition, AOE exhibited negative correlations with gait velocity (r = −0.273, P = 0.044); weak side step length (r = −0.272, P = 0.045); strong and weak sides stride length (r = −0.349, P = 0.009; r = −303, P = 0.025); and strong and weak sides SWT% (r = −0.405, P = 0.002; r = −0.322, P = 0.013); however, in the CSEC condition, AOE exhibited positive correlations with strong and weak sides stance time (r = 0.303, P = 0.025; r = 0.321, P = 0.017); strong and weak sides DST (r = 0.390, P = 0.003; r = 0.415, P = 0.002); strong and weak sides STT% (r = 0.405, P = 0.002; r = 0.332, P = 0.013); and strong and weak sides DST% (r = 0.415, P = 0.002; r = 0.359, P = 0.007). In the NSEO condition, AOE positively correlated with strong side step time (r = 0.301, P = 0.026); strong and weak sides stance time (r = 0.328, P = 0.014; r = 0.288, P = 0.033); weak side DST (r = 0.320, P = 0.017); strong side SST% (r = 0.423, P = 0.001); and strong side DST% (r = 0.266, P = 0.050); however, in the NSEO condition, AOE negatively correlated with strong side SWT% (r = −0.423, P = 0.001). In the NSEC condition, AOE negatively correlated with gait velocity (r = −0.387, P = 0.004); cadence (r = −0.279, P = 0.039); strong side step length (r = −0.315, P = 0.019); weak side stride length (r = −0.285, P = 0.035); strong and weak SWT% (r = −0.459, P = 0.000; r = −0.669, P = 0.000); and strong side SST% (r = −0.453, P = 0.001). However, in the NSEC condition, AOE exhibited positive correlations with strong and weak sides step time (r = 0.508, P = 0.000; r = 0.498, P = 0.000); strong stride time (r = 0.411, P = 0.002); strong and weak sides stance time (r = 0.669, P = 0.000; r = 0.692, P = 0.000); strong swing time (r = 0.283, P = 0.037); strong and weak DST (r = 0.722, P = 0.000; r = 0.709, P = 0.000); strong and weak sides STT% (r = 0.459, P = 0.000; r = 0.669, P = 0.000); and strong and weak sides DST% (r = 0.543, P = 0.000; r = 0.592, P = 0.000).