Bilateral lower-limb asymmetry and phase-specific trunk acceleration during the Timed Up and Go in older women: an analysis based on muscle morphology and antagonist coactivation

Abstract

The Timed Up and Go (TUG) test is widely used to assess mobility and fall risk in older adults, but total completion time does not capture phase-specific motor control. This study quantified bilateral lower-limb asymmetry in muscle morphology and neuromuscular control and examined its association with phase-specific trunk acceleration during the TUG. Twenty-seven healthy older women (78.96 ± 2.44 years) participated. Bilateral muscle thickness was measured by B-mode ultrasound, and bilateral surface electromyography and L4 trunk acceleration were recorded during a five-phase instrumented TUG. Mean and peak coactivation index (CCI) and asymmetry index (ASI) of the knee (QUAD-BF) and ankle (TA-MG) muscle pairs, together with anteroposterior and mediolateral trunk acceleration RMS (RMSAP, RMSML), were calculated. No significant group-level left–right differences in muscle thickness were detected, although ASI exceeded 15% for all four muscles. After FDR correction, significant bilateral differences were retained for mean CCIQUAD-BF during sit-to-stand (p< 0.001, q = 0.005), mean CCITA-MG during sit-to-stand (p = 0.001, q = 0.013) and peak CCIQUAD-BF during sit-to-stand (p = 0.002, q = 0.013); the difference for mean CCIQUAD-BF during return walk (p = 0.016) did not survive correction. Coactivation ASI was generally higher at the knee than the ankle. After Benjamini–Hochberg false discovery rate correction, three peak coactivation asymmetry correlations with RMSAP survived: peak ASITA-MG during outbound walk (r = 0.72, q< 0.001), peak ASIQUAD-BF during turn (r = 0.53, q = 0.043) and peak ASITA-MG during sit-to-sit (r = 0.57, q = 0.044). The negative correlations of ASIMG with RMSAP during sit-to-stand and outbound walk (r = −0.52, r = −0.51) did not survive FDR correction. No ASI correlated significantly with RMSML. Bilateral asymmetry was heterogeneous in direction, and phase-specific peak coactivation asymmetry was more consistently associated with anteroposterior trunk acceleration than muscle thickness. These measures may provide complementary information beyond muscle morphology or total TUG time, but their reliability and clinical relevance require further investigation.