Magni E, Bianchetti A, Trabucchi M, Rozzini R
"S. Cuore Fatebenefratelli" Hospital, Via Pilastroni, 4, 25123, Brescia, Italy.
J Cross Cult Gerontol. 1994 Jan;9(1):77-85. doi: 10.1007/BF00972066.
The aim of this 6 year follow-up study is to analyze the role of physical health indicators in predicting mortality. The data used were obtained from a multidimensional survey on a wide population of 70 to 75 year old subjects living at home. Baseline self-reported physical health information regarding the use of medical health services, diagnoses, and medications was collected. Multiple logistic regression analysis was used to define the multivariate association with mortality of the baseline presence of the diseases or of medications used while adjusting for sex and IADL score. The number of diseases and medications related to mortality, and the medical services use index were entered together in the same statistical model to assess which of these factors had a preeminent role in predicting mortality. Indicators such as medications and medical services use were more important mortality predictors than the diagnosis itself. A significant interaction of the variables with gender was also found.
这项为期6年的随访研究旨在分析身体健康指标在预测死亡率方面的作用。所使用的数据来自对广泛的70至75岁居家人群进行的多维调查。收集了关于医疗健康服务使用、诊断和药物治疗的基线自我报告身体健康信息。在调整性别和工具性日常生活活动能力（IADL）得分的同时，使用多元逻辑回归分析来确定疾病基线存在情况或所用药物与死亡率的多变量关联。将与死亡率相关的疾病和药物数量以及医疗服务使用指数纳入同一统计模型，以评估这些因素中哪些在预测死亡率方面具有突出作用。药物和医疗服务使用等指标比诊断本身更是重要的死亡率预测因素。还发现变量与性别之间存在显著交互作用。
