周德定, 徐乃婷, 高宁, 施燕, 彭娟娟, 喻彦. 应用结构方程模型分析上海市中心城区老年人跌倒的影响因素[J]. 环境与职业医学, 2019, 36(8): 703-709. DOI: 10.13213/j.cnki.jeom.2019.19223
引用本文: 周德定, 徐乃婷, 高宁, 施燕, 彭娟娟, 喻彦. 应用结构方程模型分析上海市中心城区老年人跌倒的影响因素[J]. 环境与职业医学, 2019, 36(8): 703-709. DOI: 10.13213/j.cnki.jeom.2019.19223
ZHOU De-ding, XU Nai-ting, GAO Ning, SHI Yan, PENG Juan-juan, YU Yan. A structural equation model of fall-related risk factors among the elderly in urban areas of Shanghai[J]. Journal of Environmental and Occupational Medicine, 2019, 36(8): 703-709. DOI: 10.13213/j.cnki.jeom.2019.19223
Citation: ZHOU De-ding, XU Nai-ting, GAO Ning, SHI Yan, PENG Juan-juan, YU Yan. A structural equation model of fall-related risk factors among the elderly in urban areas of Shanghai[J]. Journal of Environmental and Occupational Medicine, 2019, 36(8): 703-709. DOI: 10.13213/j.cnki.jeom.2019.19223

应用结构方程模型分析上海市中心城区老年人跌倒的影响因素

A structural equation model of fall-related risk factors among the elderly in urban areas of Shanghai

  • 摘要: 背景 老年人跌倒的发生往往是多种影响因素共同作用的结果,世界卫生组织将其影响因素归为生物学因素、行为因素、环境因素和社会经济因素四个大类。

    目的 分析影响城市老年人跌倒发生的影响因素,为开展干预工作提供依据。

    方法 对来自上海市7个中心城区3 480例老年人进行问卷调查、步态和平衡功能测试及居家环境评估。以跌倒发生因子为应变量,生物学因子、社会经济因子、行为因素和居家环境因素为自变量建立结构方程模型,分析跌倒发生的影响因素。潜变量跌倒发生因子以跌倒次数、跌伤次数、住院天数和休息天数为观测变量,潜变量生物学因子以骨质疏松风险得分、计时起立行走测试(TUG)时间、体质风险评估得分、年龄和性别为观测变量,潜变量社会经济因子以有无同住人和教育程度为观测变量。

    结果 实际完成调查3 385人。13.06%的调查对象在过去一年内发生过跌倒。女性(χ2=25.83,P < 0.05)、70~和80~岁组(χ2=52.27,P < 0.05)、小学及以下文化程度(χ2=10.05,P < 0.05)和无同住人(χ2=3.98,P < 0.05)的老年人更容易发生跌倒。跌倒组体质风险得分(Z=12.51,P < 0.05)、TUG时间(Z=9.29,P < 0.05)、骨质疏松风险得分(Z=8.46,P < 0.05)、行为因素得分(Z=4.91,P < 0.05)和居家环境因素得分(Z=4.66,P < 0.05)均高于未跌倒组。结构方程模型显示,生物学因子、行为因素和居家环境因素对跌倒发生因子的标准化回归系数(β)分别为0.564(P < 0.01)、0.070(P < 0.01)和0.083(P < 0.01),社会经济因子的β无统计学意义,在生物学因子中,体质风险、TUG时间的β分别为0.658(P < 0.01)和0.477(P < 0.01),高于其他观测变量。女性比男性更容易发生跌倒,女性老年人中居家环境因素对跌倒发生因子的β为0.111(P < 0.01),男性老年人中居家环境因素的β无统计学意义。

    结论 生物学因素、行为因素和居家环境因素都是老年人跌倒发生的影响因素。其中生物学因素的影响最大,生物学因素与体质风险评估得分、TUG时间的关系最为密切。女性跌倒风险要高于男性。居家环境因素是女性特有的影响因素。

     

    Abstract: Background Falls occur as a result of a complex interaction of risk factors. The World Health Organization categorizes the risk factors into four dimensions:biological, behavioral, environmental, and socioeconomic factors.

    Objectve This study is designed to analyze the factors affectng the falls of the elderly in urban areas and provide evidence for targeted interventon programs.

    Methods A total of 3 480 elderly partcipants from 7 urban areas of Shanghai were surveyed by questonnaires, tested for gait and balance functon, and evaluated for home environment. A structural equaton model was developed to examine the relatonships between independent variables (biological, socioeconomic, behavioral, and home environmental factors) and dependent variable (falls). The latent variables of falls were indicated by manifest variables including number of falls, number of injuries, days of hospitalizaton, and days of rest; the latent variables of biological factors were indicated by manifest variables including osteoporosis risk score, timed up and go test (TUG) time, physical risk assessment score, age, and gender; the latent variables of socioeconomic factors were indicated by manifest variables including whether living alone and educaton.

    Results Of the 3 385 subjects who completed the questonnaire survey, 13.06% fell in the past year. The rates of falls were signifcantly higher in women (χ2=25.83, P < 0.05), the 70- years and 80- years age groups (χ2=52.27, P < 0.05), those with elementary school or less educaton (χ2=10.05, P < 0.05), and those with no cohabitant (χ2=3.98, P < 0.05). The elderly partcipants who fell had higher physical risk score (Z=12.51, P < 0.05), TUG time (Z=9.29, P < 0.05), osteoporosis risk score (Z=8.46, P < 0.05), behavioral factor score (Z=4.91, P < 0.05), and home environmental factors score (Z=4.66, P < 0.05) than those who did not fall. The structural equaton model showed that the standardized regression coefcients (β) of the biological factors, behavioral factors, and home environmental factors were 0.564 (P < 0.01), 0.070 (P < 0.01), and 0.083 (P < 0.01), respectvely, and the regression coefcient (β) of socioeconomic factors was not statstcally signifcant. Among the biological factors, the regression coefcients (β) for physical risk and TUG time were 0.658 (P < 0.01) and 0.477 (P < 0.01) respectvely, higher than the coefcients of other manifest variables. Women were more likely to fall than men. The regression coefcient (β) of home environmental factors of falls in female elderly people was 0.111 (P < 0.01), and the coefcient (β) of male elderly people was not signifcant.

    Conclusion Biological, behavioral, and home environmental factors may affect the falls of the elderly. Among them, biological factors are dominant and are most closely associated with physical risk score and TUG time. The risk of falls is higher for women than for men. Home environment as a risk factor is unique to women.

     

/

返回文章
返回