Can clinical and physical-functional factors predict falls in cognitively impaired older adults?

Abstract Objective: to investigate the frequency of falls and clinical and physical-functional factors associated with falls, and the accuracy of such factors to identify the risk of falling in cognitively impaired older adults. Method: a cross-sectional study with cognitively impaired older adults was carried out using the Mini-Mental State Examination. The dependent variable was a history of falls in the previous six months. The independent variables were self-reported mental confusion, hearing and visual impairment, physical fatigue, muscle weakness, dizziness, body imbalance, insecurity when walking, diagnosed depression, hospitalizations, continuous use medications (form from study used), muscle mass (calf circumference) and handgrip strength (dynamometry), functional capacity (Pfeffer) and mobility (SAM-Br). Descriptive statistics were applied. The groups were compared using the Mann Whitney U test, the risk factors were identified by univariate and multivariate logistic regression, and the area under the ROC curve (AUC) was calculated for the associated factors. Results: 216 cognitively impaired older adults were included in the analysis, 41.7% of whom were fallers. Multivariate regression analyzes indicated that complaints of visual impairment (OR=2.8; p=0.015) and body imbalance (OR=2.7; p=0.004), and greater medication use (OR=1.1; p=0.038) were associated with a history of falls. The AUC found poor accuracy for quantity of medications as a screening tool for fallers (AUC=0.6 [0.5; 0.7]; p=0.028). Conclusion: cognitively impaired older adults had a high frequency of falls. Complaints of visual impairment, body imbalance and polypharmacy were predictors of falls. The early assessment of these factors can contribute to the identification of cognitively-impaired older adults at risk of falling in clinical practice and research.

摘要 目的：探讨认知障碍老年人群的跌倒发生率、与跌倒相关的临床及躯体功能因素，以及此类因素识别跌倒风险的效能。方法：采用简易精神状态检查（Mini-Mental State Examination）筛选受试者，开展认知障碍老年人的横断面研究。本研究的因变量为受试者近6个月内的跌倒史；自变量包括自述精神错乱、视听障碍、躯体疲劳、肌无力、头晕、躯体失衡、行走时不稳感、确诊抑郁症、住院史、长期用药情况（本研究采用的用药类型）、肌肉量（小腿围）、握力（握力计测量）、功能能力（Pfeffer量表）以及活动能力（SAM-Br量表）。统计分析采用描述性统计；组间比较采用曼-惠特尼U检验（Mann Whitney U test）；通过单变量及多变量logistic回归分析识别跌倒危险因素，并计算相关因素的ROC曲线下面积（AUC）。结果：本研究共纳入216名认知障碍老年受试者，其中41.7%存在跌倒史。多因素回归分析显示，视觉障碍主诉（比值比OR=2.8；P=0.015）、躯体失衡主诉（OR=2.7；P=0.004）以及用药量增加（OR=1.1；P=0.038）与跌倒史显著相关。以用药量作为跌倒筛查工具时，其诊断效能较差（AUC=0.6，95%置信区间0.5~0.7；P=0.028）。结论：认知障碍老年人群的跌倒发生率较高。视觉障碍主诉、躯体失衡及多重用药是跌倒的预测因素。对上述因素进行早期评估，有助于在临床实践与研究中识别存在跌倒风险的认知障碍老年人群。