摘 要

失智症是造成病人殘障的高風險疾病，很少研究指出一般族群中失智症的累進風險。本研究希望評估台灣族群中失智症的共病風險以及存活率。我們從全民健康保險研究數據庫(TNHIRD)分析，從1999-2008收集共48710位失智症病人，平均年齡≧18歲，根據ICD-9代碼(290-290.9，294，331-失智症診斷碼)，按照年齡和性別，以1：1方式隨機匹配對照組。利用邏輯回歸比較失智症共病比例，及Cox- Proportional模型，了解相對風險值(RR=Relative Risks)，並分析病人10年累積存活率。結果年齡增加失智症風險亦提高。而具有共病症者失智症更高，包含腦血管疾病、糖尿病、高血壓、冠心病、週邊動脈疾病、充血性心臟疾病、慢性肺阻塞性疾病、胃腸道出血、以及癌症。由回歸模式分析，失智症比起非失智症約有1.1〜1.7倍的相對風險(RR)；其中腦血管疾病的6.42倍最高，具統計意義(個別P值 <,'Times New Roman'; font-family:'新細明體',serif, 'Times New Roman'">。長期追蹤失智症病人，其累積生存率隨之下降(第1年0.93，第5年0.82，第9年0.77，log-rank test, p<,'Times New Roman';font-family:'新細明體',serif,'Times New Roman'">。調整及多變數回歸分析，失智症的死亡平均風險，約多了2.23倍 (RR2.23，95％信賴區間CI=2.13-2.34)。10年分析結果顯示失智症和許多共病有關。而年齡因素更與高死亡率相關。如何於一般族群中尋找具有認知功能障礙者，早期篩選可能共病症加以控制，或許可以改善失智症的不良預後。

關鍵詞：失智症，風險及死亡率，共病症，十年資料分析

Abstract

The aim of this study was to assess the risk and mortality of dementia in Taiwan. We analyzed the data collected from Taiwan National Health Insurance Research Database (TNHIRD) 1999-2008, and got a cohort of 48710 subjects aged≧18 yrs old. Dementia was defined based on ICD-9 code (290-290.9, 294,331) with a randomly matched control 1:1 by age and sex. The logistic regression and Cox proportional model were adopted for the baseline demographics and relative risk (RR). The cumulative survival years were estimated between dementia and non-dementia patients in general population. Results showed dementia was increasing with age, and co-morbidities were higher in dementia together with cerebrovascular disease, diabetes, hypertension, coronary artery disease, peripheral artery disease, congestive heart disease, chronic pulmonary obstructive disease, gastrointestinal bleeding, or cancer. By conditional regression models, we found that RRs for dementia were 1.1~1.7 times higher than those for non-dementia patients, and that the highest RRs appeared in cerebral vascular disease, reaching 6.42 times, with statistically significance (individual p<0.01). The cumulative survival rate decreased with years in dementia than in non-dementia (1-year 0.93, 5-year 0.82, 9-year 0.77, log-rank test, p<0.0001). Adjusting confounders with multivariate regression led us to the finding that RRs for dementia was 2.25 times higher (RR 2.23, 95% confidence interval (CI) 2.13–2.34). In brief, the 10-year analysis indicated that dementia were associated with other diseases, and age played a key role. Implications emerge that screening for cognitive impairment in advance might help control dementia and its co-morbid disorders.

Keywords: Dementia, Risk and Mortality, Co-Morbidities, 10-Year Registration