时间标签：数据日期 2026-10-31 · 生成时间 2026-11-01 05:30（Asia/Shanghai）｜ 检索窗口 PubMed edat 2026/10/28–2026/10/31 ｜ 情报发生日期 2026-10-31
研究问题
However, current assessment methodologies are often constrained by limited precision and suboptimal ecological validity.；While sensor-based gait analysis has improved identification accuracy, its scope remains limited by the specificity of the motor tasks performed [].；To bridge this gap, dual-task protocols—which synchronize cognitive interference with postural challenges—have emerged as a more ecologically valid paradigm, effectively simulating the complex, real-world conditions under which falls typically occur [].
方法
A total of 216 older adults were initially recruited for this study.；Of these, 7 individuals were excluded due to lower limb joint pain.；Consequently, 209 community-dwelling older adults aged 65 years and older (70.2 ± 3.9 years; 162.8 ± 8.0 cm; 63.9 ± 10.3 kg; Table ) were included in the final analysis, including 119 females and 90 males.；Inclusion criteria were: (1) age ≥ 65 years; (2) ability to walk independently without assistive devices; (3) absence of medications that could affect balance (sedatives, hypnotics, antidepressants, and benzodiazepines); and (4) stable health status sufficient to complete balance and cognitive tasks.
核心结论
Regarding functional performance, the Fall group demonstrated significantly longer TUGT durations compared with the Non-fall group (9.6 ± 1.2 s vs. 8.9 ± 1.1 s, p < 0.001).；However, no significant intergroup differences were observed in balance and strength measures, including BBS, SPPB, SEBT-Left, SEBT-Right, FTSTS, EC-SLST, and HGS (all p ≥ 0.069, all FDR-adjusted p ≥ 0.097).；Regarding mean spatiotemporal parameters, the Fall group exhibited significantly prolonged cycle time, stance time, step time, and double-limb support time (all p < 0.001, all FDR-adjusted p < 0.001), as well as significantly shorter step length and stride length (p < 0.001, FDR-adjusted p < 0.001), and lower walking speed, steps per minute, and strides per minute (all p < 0.001, all FDR-adjusted p < 0.001), compared with the Non-fall group. ‖ 结论：These results suggest that the synergy between using optical motion capture to analyze gait under dual-task challenges and probabilistic machine learning can facilitate early, personalized screening and evaluation, helping to identify high-risk individuals and potentially guiding timely interventions to mitigate the socioeconomic impact of falls in the aging population.
局限
Nevertheless, these 10 metrics should be interpreted as promising discriminative candidates within our study population rather than universal, definitive biomarkers, and their long-term stability warrants further validation across external datasets.；First, NB is highly efficient for relatively small sample sizes (n = 209), as it requires fewer parameters to be estimated and is less prone to overfitting than more complex, non-linear models [].；Despite the promising results, several limitations warrant consideration.
对本项目的价值
★☆☆☆☆ 衰弱/肌少症是少有的已有监管认可终点的老年综合征，是临床转化的最短路径。
全文来源
PMC JATS 全文（正文 41,483 字符，已过标题/DOI 逐字归属校验）
原文题录
- 英文标题：Machine learning-based fall risk prediction in community-dwelling older adults using dual-task spatiotemporal gait parameters.
- 期刊与标识：BMC Geriatr · PMID 42947007 · DOI 10.1186/s12877-026-07998-3
A 级证据分级：A 级（同行评议人体研究）
本文属人体研究证据，但需注意：观察性或小样本干预的结论不等于已证实临床获益；效应量与长期安全性仍需更大规模、更长随访的试验验证。
关于本文的文献来源：本站依据 IMA 知识库「五刊监控·深度文献卡片」撰写，依据已获取的期刊全文（方法、样本量与结论已回到原文核对）；未能获取全文的同批条目另行标注「摘要级」，全程未做任何数值推算补全。