Department of Basic Psychology, Psychobiology and Methodology of Behavioral Science, Faculty of Psychology, University of Salamanca, Avda. de la Merced 109-131, Salamanca, ES-37005, Spain.
Research Institute of Hospital "12 de Octubre" (i+12), Madrid, Spain; The Biomedical Research Centre Network for Neurodegenerative Diseases (CIBERNED), Carlos III Research Institute, Madrid, Spain; Faculty of Medicine, Complutense University, Madrid, Spain.
Arch Gerontol Geriatr. 2019 Jan-Feb;80:88-94. doi: 10.1016/j.archger.2018.10.008. Epub 2018 Oct 25.
We investigated the mortality rates of three subtypes of disability and their specific explanatory factors in older adults.
Our data come from NEDICES, a population-based longitudinal cohort study of Spanish older adults. We examined 3816 participants without dementia who completed the Pfeffer's Functional Activities Questionnaire (FAQ) and an assessment of self-perceived functional limitations (SFL) associated with health conditions. Subjects were classified into mutually exclusive subtypes of disability: subtype 1 (SFL), subtype 2 (impaired FAQ), and subtype 3 (impaired FAQ plus SFL). Factors related to all disability subtypes were analyzed using a multinomial logistic regression (MLR), whereas Cox regression (CR) models adjusted by covariates were applied to compare survival rates between groups at the 5-year follow up.
The CR models indicated that SFL and FAQ scores were associated with higher risk of mortality at 5-years. After stratifying by subtypes of disability, mortality was significantly higher in subtype 3 than in subtypes 1 and 2. All models were consistent after adjusting by different covariates. The MLR showed that subtype 1 was specifically associated with the number of comorbidities, whereas subtype 2 was associated with lower MMSE scores depression and living in nursing homes.
Our results show that the combination of impaired FAQ plus SFL have an increased differential predictive utility for mortality than approaches based on unique measures. They also indicate that both measures of disability are associated with different explanatory factors.
我们调查了老年人三种残疾亚型的死亡率及其特定的解释因素。
我们的数据来自 NEDICES,这是一项基于人群的西班牙老年人纵向队列研究。我们检查了 3816 名没有痴呆症的参与者,他们完成了 Pfeffer 的功能性活动问卷(FAQ)和与健康状况相关的自我感知功能限制(SFL)评估。受试者被分为相互排斥的残疾亚型:亚型 1(SFL)、亚型 2(FAQ 受损)和亚型 3(FAQ 受损加 SFL)。使用多项逻辑回归(MLR)分析与所有残疾亚型相关的因素,而 Cox 回归(CR)模型则通过协变量进行调整,以比较 5 年随访时各组的生存率。
CR 模型表明,SFL 和 FAQ 评分与 5 年时的死亡率升高相关。在按残疾亚型分层后,亚型 3的死亡率明显高于亚型 1 和 2。在调整不同协变量后,所有模型均保持一致。MLR 表明,亚型 1 与合并症数量有关,而亚型 2 与较低的 MMSE 评分、抑郁和居住在养老院有关。
我们的结果表明,与基于单一措施的方法相比,FAQ 受损加 SFL 的组合对死亡率具有更高的差异预测效用。它们还表明,两种残疾测量方法都与不同的解释因素有关。