Lower-Limb Phase Angle and Muscle Mass Ratio Are Associated with Slow Timed Up and Go Performance in Community-Dwelling Japanese Adults

Background/Objectives: Lower-limb muscle quantity and quality may be associated with mobility performance, but the relationship between bioelectrical impedance analysis (BIA)-derived lower-limb indices and slow Timed Up and Go (TUG) performance remains insufficiently established. This cross-sectional study examined the associations of lower-limb muscle mass and phase angle (PhA) with TUG performance and explored ROC-derived values for slow TUG performance in 280 community-dwelling Japanese adults. Methods: Lower-limb muscle mass and segmental lower-limb PhA were measured using multifrequency segmental BIA and averaged across both legs. Multiple linear regression analyses were used to examine associations between TUG time and BIA-derived indices after adjusting for age and body size. Receiver operating characteristic analyses were performed for slow TUG performance, defined as TUG ≥ 9 s and ≥10.2 s, and Youden index-derived values were calculated as exploratory estimates. Results: Lower-limb PhA was independently associated with TUG performance across both body-size adjustment models, whereas lower-limb muscle mass was retained only in the model adjusted for BMI. For slow TUG performance, the lower-limb muscle mass-to-body weight ratio showed moderate discriminatory ability, with AUCs of 0.75 for TUG ≥ 9 s and 0.78 for TUG ≥ 10.2 s. Standardized lower-limb PhA showed AUCs of 0.65 and 0.64, respectively. The Youden index-derived standardized PhA value was Z = −1.04 for both TUG thresholds; however, these ROC-derived values were not internally or externally validated. Conclusions: Lower-limb BIA-derived indices, particularly segmental lower-limb PhA, were associated with TUG performance in community-dwelling Japanese adults. The ROC-derived values should be interpreted strictly as exploratory, hypothesis-generating estimates and should not be used for clinical screening or individual-level clinical decision-making without validation in larger independent cohorts.