Department of Global Public Health and Primary Care, University of Bergen, Kalfarveien 31, 5018 Bergen, Norway.
Department of Clinical Science, University of Bergen, Postbox 7804, 5020 Bergen, Norway.
Gait Posture. 2019 Jul;72:142-147. doi: 10.1016/j.gaitpost.2019.06.005. Epub 2019 Jun 7.
Increases in stride-to-stride fluctuations (gait variability) are common among older adults, but little is known about the natural progression of gait variability with increasing age.
Does gait variability change with increasing age in a group of community-living older adults?
The participants were community-living volunteers between 70-81 years, who were tested with a two-year interval between tests. They walked 6.5 m under four different conditions: At preferred speed, at fast speed, during a dual task condition and on an uneven surface. Trunk accelerations in the anteroposterior (AP), mediolateral (ML) and vertical (V) direction were captured using a body-worn sensor worn at the lower back. Gait variability was estimated using an autocorrelation procedure, where coefficients tending towards 1.0 indicated low variability and 0.0 as high variability. To estimate change, we used an ANOVA-procedure with baseline gait speed as a covariate.
At baseline, 85 older adults were tested, and data for 56 of these were available for analysis over a two-year period of time. The average age at inclusion was 75.8 years (SD 3.43) and 60% were women. During preferred speed walking, variability increased in the AP direction (mean difference 0.05, p = .038), during fast speed walking it increased in the V direction (mean difference 0.04, p = .037) and during dual task-walking, it increased in the ML and V directions (mean differences 0.03, p = .032 and 0.09, p = .020 respectively).
The findings from this study could be helpful for discriminating between normal and pathological progression of gait variability in older adults.
步幅波动（步态变异性）增加在老年人中很常见，但对于随着年龄增长步态变异性的自然进展知之甚少。
在一组社区居住的老年人中，步态变异性是否会随着年龄的增长而变化？
参与者是年龄在 70-81 岁之间的社区居住志愿者，他们在两次测试之间间隔两年进行测试。他们在四种不同条件下行走 6.5 米：以惯用速度、快速、双重任务条件和不平坦表面。使用佩戴在腰部的身体佩戴传感器捕获前后（AP）、左右（ML）和垂直（V）方向的躯干加速度。使用自相关程序估计步态变异性，其中接近 1.0 的系数表示低变异性，0.0 表示高变异性。为了估计变化，我们使用了包含基线步态速度作为协变量的 ANOVA 程序。
在基线时，对 85 名老年人进行了测试，其中 56 名老年人在两年的时间内提供了数据进行分析。纳入时的平均年龄为 75.8 岁（标准差 3.43），60%为女性。在惯用速度行走时，AP 方向的变异性增加（平均差异 0.05，p=0.038），在快速行走时，V 方向的变异性增加（平均差异 0.04，p=0.037），在双重任务行走时，ML 和 V 方向的变异性增加（平均差异 0.03，p=0.032 和 0.09，p=0.020）。
本研究的结果有助于区分老年人步态变异性的正常和病理性进展。
