台灣阿茲海默症與失智症患者照護者負擔： 一項社區式橫斷研究

Abstract

隨著台灣人口快速高齡化，阿茲海默症與其他類型失智症（Alzheimer's disease and other dementias, ADOD）盛行率持續上升，其造成的失能與照護需求已成為重要的公共衛生議題。ADOD不僅影響患者本身之認知功能與生活品質，亦對家庭照護者之身心健康與社會功能造成顯著衝擊。然而，現行以全球疾病負擔（Global Burden of Disease, GBD）架構為主之疾病負擔評估，多聚焦於病人端之失能調整生命年（DALY），尚未建立可與病人端失能權重（Disability Weight, DW）並列比較、並銜接於DALY架構之「照護者端」健康負擔指標。有鑑於此，本研究以2020年台灣NHWS（National Health and Wellness Survey）之社區樣本為基礎，運用GBD架構之獨立乘法共病模型，建立ADOD照護者之綜合失能權重（Combined Disability Weight, Combined DW），並與其他慢性病照護者及非照護者進行比較，同時輔以多維度量表描繪ADOD照護者身心負荷之具體樣貌。 本研究採橫斷性研究設計，依受訪者自我回報之照顧責任與被照護者疾病類型，分為ADOD照護者組（N=716）、其他慢性病照護者組與非照護者組。以受訪者自陳之多重臨床共病，依GBD失能權重（Disability Weight, DW）參數，優先參照臺灣疾病負擔中心（TBDC）本土實證資料，其次以健康指標評估研究所（IHME）數據逆向推估，賦予各共病項目相對應之DW，再透過獨立乘法模型（Combined DW = 1 − Π(1 − DWi)）計算個人層級之綜合失能權重。為控制組間年齡與性別分布之基準線不平衡，本研究以傾向分數匹配法（Propensity Score Matching, PSM）分別建構ADOD照護者組與另外兩組之可比對照樣本。組間Combined DW之比較採Kruskal-Wallis檢定，並以Bonferroni校正之成對Wilcoxon檢定進行事後比較。此外，本研究另計算個別共病之群體層級預期健康損失貢獻度，以釐清驅動各組健康損失之主要臨床因素，並運用SF-12v2、EQ-5D-5L、PHQ-9、GAD-7、WPAI及CRA等量表，描繪ADOD照護者之健康相關生活品質、憂鬱焦慮症狀、工作生產力損失及主觀照顧反應。 結果顯示， ADOD照護者組之Combined DW平均為0.259，其他慢性病照護者組為0.242，非照護者組為0.136；三組間存在顯著之整體分布差異（p<0.001）。事後比較顯示，ADOD照護者組與其他慢性病照護者組之Combined DW，皆顯著高於非照護者組（皆p<0.001），但ADOD照護者組與其他慢性病照護者組間未達顯著差異（p=0.98）。另以中風照護者作為疾病別對照，ADOD照護者組之Combined DW顯著較高（0.254 vs 0.211，p=0.013）。疾病貢獻度分析顯示，憂鬱症與失眠為驅動ADOD照護者組群體健康損失之兩大主要共病項目，其貢獻度皆高於其他慢性病照護者組與非照護者組。多維度量表結果顯示，ADOD照護者之EQ-5D-5L效用值平均為0.83，SF-12v2心理健康總分（43.37）低於生理健康總分（49.82）及一般常模；PHQ-9與GAD-7顯示分別有18.9%與12.1%照護者達中度以上憂鬱或焦慮症狀；就業中之照護者整體工作生產力損失平均為28.1%；CRA五大構面中，照護自豪感平均3.18分，高於量表中點，顯示正向照護評價之存在；四項負向構面則以健康影響（3.07）得分最高，呈現照護經驗中正負向並存之特質。 上述發現與既有以NHWS為基礎之國際文獻方向一致。本研究採橫斷性設計，僅能呈現組間差異，無法確立照顧角色與健康負擔間之因果關係。共病資料悉數仰賴受訪者自陳，其準確性可能受回憶偏誤與社會期許偏誤影響；PSM僅納入年齡與性別兩項共變量，仍可能存在未被匹配之殘餘干擾因子，上述限制於解讀本研究結果時應予考量。 本研究以台灣NHWS資料，首次建立可與病人端失能權重並列比較、並銜接於DALY架構之ADOD照護者端健康負擔指標，結果顯示ADOD照護者所承受之健康負擔顯著高於非照護者；與其他慢性病照護者組雖未達顯著差異，惟改以疾病組成單一之中風照護者作為對照後仍顯著較高，進一步支持ADOD照顧所伴隨之健康負擔為值得重視的公共衛生議題。本研究之Combined DW方法學，可作為未來建立照護者健康負擔常態性監測工具之參考，亦呼應既有照護者文獻中「第二病人」之概念，對台灣及東亞高齡化社會之ADOD照顧政策研擬具參考價值。

As Taiwan's population ages rapidly, the prevalence of Alzheimer's disease and other dementias (ADOD) continues to rise, adding to the burden borne by family caregivers. Current disease burden estimates under the Global Burden of Disease (GBD) framework focus on patient-level Disability-Adjusted Life Years (DALY), but no caregiver-side burden indicator has been established that can be placed alongside patient-level disability weights within the DALY framework. This study used the 2020 Taiwan National Health and Wellness Survey (NHWS) to construct a Combined Disability Weight (Combined DW) for ADOD caregivers under the GBD independent multiplicative comorbidity model, compared it with other chronic-disease caregivers and non-caregivers, and characterized their burden using multiple validated instruments. This cross-sectional study classified respondents by self-reported caregiving status and care-recipient diagnosis into ADOD caregivers (N=716), other chronic-disease caregivers, and non-caregivers. Disability weights for comorbidities were assigned from Taiwan Burden of Disease Center (TBDC) estimates or back-calculated from Institute for Health Metrics and Evaluation (IHME) data. Individual-level Combined DW was calculated using the independent multiplicative model (Combined DW = 1 − Π(1 − DWi)). Propensity score matching (PSM) on age and sex built comparable samples for each pairwise comparison; group differences were tested with the Kruskal-Wallis test and Bonferroni-corrected pairwise Wilcoxon tests. Expected health loss contribution was calculated for each comorbidity, and health-related quality of life, depression/anxiety, work productivity, and caregiving reactions were assessed with SF-12v2, EQ-5D-5L, PHQ-9, GAD-7, WPAI, and CRA. After PSM, the three groups were balanced on age and sex (all ASMD<0.1). Mean Combined DW was 0.259 for ADOD caregivers, 0.242 for other chronic-disease caregivers, and 0.136 for non-caregivers (p<0.001). Both caregiver groups scored significantly higher than non-caregivers (both p<0.001), but ADOD and other chronic-disease caregivers did not differ significantly (p=0.98). In a disease-specific comparison with stroke caregivers, ADOD caregivers had a significantly higher Combined DW (0.254 vs 0.211, p=0.013). Depression and insomnia were the leading contributors to health loss among ADOD caregivers. Mean EQ-5D-5L utility was 0.83; the SF-12v2 mental component summary (43.37) was lower than the physical component summary (49.82) and population norms. Moderate or greater depressive and anxiety symptoms were reported by 18.9% (PHQ-9) and 12.1% (GAD-7) of caregivers; mean work productivity impairment (WPAI) among employed caregivers was 28.1%. On the CRA, caregiving esteem averaged 3.18, above the scale midpoint, indicating the presence of positive caregiving appraisal; among the four negative domains, impact on health scored highest (3.07). These findings are consistent with prior NHWS-based international studies. As a cross-sectional design, this study cannot establish causality between caregiving and health burden. Comorbidity data were entirely self-reported, so their accuracy may be affected by recall bias and social desirability bias; PSM also adjusted only for age and sex, leaving the possibility of residual confounding. Using Taiwan NHWS data, this study is the first to construct a caregiver-side health burden indicator directly comparable with patient-level disability weights within the DALY framework. ADOD caregivers showed significantly higher health burden than non-caregivers; although they did not differ significantly from other chronic-disease caregivers, they scored significantly higher than stroke caregivers in a disease-specific comparison, supporting ADOD caregiving-related health burden as a public health issue that warrants attention. The Combined DW methodology may support future routine monitoring of caregiver health burden, echoing the "second patient" concept in the caregiving literature, and offers evidence for ADOD caregiving policy in Taiwan and other aging East Asian societies.