RESUMO
OBJECTIVES: To determine whether physical performance measures commonly used in clinical settings can discriminate fallers from nonfallers and predict falls in older adults with dementia. DESIGN: Systematic review and meta-analysis. SETTING AND PARTICIPANTS: Older adults with dementia residing in the community, hospitals, and residential care facilities. METHODS: MEDLINE, Embase, PsycINFO, CINAHL, SPORTDiscus, the Cochrane Library, and the PEDro databases were searched from inception until December 27, 2023 (PROSPERO registration number: CRD42022303670). Retrospective or prospective studies that evaluated the associations between physical performance measures and falls in older adults with dementia were included. A random effects model was used to calculate the standardized mean difference (SMD) and 95% CI for each physical performance measure between fallers and nonfallers. Sensitivity analyses were conducted on the longitudinal studies to determine the ability of physical performance measures to predict future falls. RESULTS: Twenty-eight studies were included in this review (n = 3542). The 5-time chair stand test [SMD = 0.23 (0.01, 0.45)], the Berg Balance Scale [SMD = -0.52 (-0.87, -0.17)], postural sway when standing on the floor [SMD = 0.25 (0.07, 0.43)] and on a foam surface [SMD = 0.45 (0.25, 0.66)], and the Short Physical Performance Battery total score [SMD = -0.46 (-0.66, -0.27)] could discriminate fallers from nonfallers. Sensitivity analyses showed that gait speed could predict future falls in longitudinal cohort studies [SMD = -0.29 (-0.49, -0.08)]. Subgroup analyses showed that gait speed [SMD = -0.21 (-0.38, -0.05)] and the Timed Up and Go test [SMD = 0.54 (0.16, 0.92)] could identify fallers staying in residential care facilities or hospitals. CONCLUSIONS AND IMPLICATIONS: The 5-time chair stand test, the Berg Balance Scale, postural sway when standing on the floor and a foam surface, and the Short Physical Performance Battery can be used to predict falls in older adults with dementia. Gait speed and the Timed Up and Go test can be used to predict falls in institutionalized older adults with dementia. Clinicians are recommended to use these physical performance measures to assess fall risk in older adults with dementia.
RESUMO
BACKGROUND: Promoting an erect posture in standing has been advocated to offer superior protection to the spine when compared to hyperlordotic and swayback postures. RESEARCH QUESTION: Do postural adjustments towards external perturbation differ between erect, hyperlordotic and swayback postures? If so, which posture offers better protection to lumbar spine? METHODS: Forty-four healthy adults received top-down perturbations under unpredictable (without visual-and-auditory input) and predictable (with visual-and-auditory input) conditions in three simulated postures: erect, hyperlordotic, and swayback. Postural adjustments namely the centre of pressure parameters, joint angle onsets, and neuromuscular responses measured by muscle onsets and co-contraction between muscle pairs upon the perturbation were compared using the two-way repeated measures ANOVA. Post-hoc analysis with Bonferroni correction was conducted to identify the between-posture differences for the respective postural adjustment parameters. RESULTS: Path length, ellipse area and average velocity of centre of pressure were significantly greater under unpredictable condition as compared to predictable condition (p < 0.001). Significant between-posture difference was detected in centre of pressure path length (p < 0.035), pelvic tilt onset (p < 0.038) and all muscle co-contraction indexes (p < 0.001). Post-hoc analysis revealed significantly smaller centre of pressure path length in erect posture as compared to hyperlordotic and swayback postures (p < 0.01) under unpredictable conditions. Significantly greater co-contraction indexes of lumbar multifidus and erector spinae, and internal oblique and lumbar multifidus were found in hyperlordotic as compared to erect and swayback postures (p < 0.05). SIGNIFICANCE: Compared to erect posture, adoption of hyperlordotic and swayback postures altered the contributions of the active and passive subsystems of the spine that regulates postural control upon external perturbations. Such differences in neuromuscular control may lessen the capacity of the human spine to withstand loading and shear forces. Prospective studies are required to validate if habitually adopted hyperlordotic and swayback postures contribute to an earlier/ higher prevalence of spinal dysfunctions.