社區高齡者社會心理因素與死亡風險之相關性探討

Abstract

背景：隨全球高齡化趨勢加劇，高齡人口比例持續上升，平均壽命延長使高齡者健康議題逐漸由以急性疾病為主，轉向多重慢性疾病共存、功能退化與整體健康風險交織而成之複雜健康樣態。高齡者之死亡風險已不再僅由單一疾病所決定，而是由慢性病負荷、功能狀態下降與社會心理脆弱性等多重因素相互影響所形成。其中，社區高齡者之社會心理因素在同時考量慢性疾病與功能狀態等相關因素後，其與全死因死亡風險之關聯仍有待進一步釐清。 目的：本研究旨在探討社區高齡者之基本人口學特徵、慢性疾病、失能狀態、認知功能、社會心理因素（包含社會性衰弱、心理性衰弱及憂鬱症狀）與全死因死亡風險之關聯。 方法：本研究採回溯性世代研究設計，納入2019年至2025年8月31日期間設籍彰化縣、並參與高齡者整合式健康檢查之65歲以上社區高齡者。研究以整合式健檢資料串聯戶役政系統死亡登記資料，追蹤至2025年8月31日之全死因死亡事件。以Cox比例風險迴歸模型進行單變項與多變項分析，評估各人口學、健康狀態、失能與社會心理因素與死亡風險之關聯。 結果：本研究共納入8,222位65歲以上社區高齡者，追蹤期間存活7,581人（92.2%）、死亡641人（7.8%）。多變項Cox模型顯示，男性（aHR = 2.14，95% CI: 1.28–3.59）、年齡較高（75–84歲aHR = 2.52，95% CI: 1.98–3.22；≥85歲aHR = 5.33，95% CI: 4.06–6.99）、BMI過輕（aHR = 2.00，95% CI: 1.42–2.81）、喪偶（aHR = 1.25，95% CI: 1.04–1.50）、糖尿病（aHR = 1.62，95% CI: 1.36–1.93）與腎臟病（aHR = 1.47，95% CI: 1.12–1.93）皆與較高全死因死亡風險獨立相關。失能方面，中重度依賴者之死亡風險顯著高於完全獨立者（aHR = 2.03，95% CI: 1.62–2.54），顯示失能為高齡死亡風險之重要指標。社會心理因素中，社會性衰弱與死亡風險呈穩定正向關聯（衰弱前期aHR = 1.25，95% CI: 1.04–1.50；衰弱aHR = 1.26，95% CI: 1.01–1.58），且社會性衰弱指標項數呈劑量反應趨勢（aHR = 1.09，95% CI: 1.00–1.19；χ² = 4.034，df = 1，p = 0.045）。心理性衰弱僅在單變項模型中與死亡風險呈顯著正相關（衰弱前期HR = 1.24，95% CI: 1.01–1.52；衰弱HR = 1.48，95% CI: 1.22–1.80），惟在多變項模型中其與死亡風險之關聯不再顯著。憂鬱症狀於單變項分析中亦呈現死亡風險上升趨勢，其中僅輕度與中度憂鬱達統計顯著（輕度HR = 1.52，95% CI: 1.20–1.92；中度HR = 1.67，95% CI: 1.07–2.61；中重度HR = 1.72，95% CI: 0.77–3.85；重度HR = 2.06，95% CI: 0.51–8.27）。然而，在同時調整慢性疾病、失能狀態與社會心理衰弱等因素後，憂鬱症狀與全死因死亡風險之關聯亦未達統計顯著。 結論：本研究顯示，在彰化縣社區高齡者中，男性、高齡、BMI過輕、喪偶、糖尿病、腎臟病與中重度失能，皆為全死因死亡風險之重要且獨立相關因子；其中，社會性衰弱除與死亡風險呈穩定正向關聯外，亦呈現指標項數增加之劑量反應趨勢，顯示其可提供在慢性病與失能以外之額外風險辨識價值。相對地，心理性衰弱與憂鬱症狀之風險訊號在多變項調整後不再顯著，可能反映其影響部分透過慢性病、失能與社會性衰弱等因素所中介或重疊。建議將社會性衰弱納入整合式健檢風險分層，以利及早連結社會支持與社區資源、強化介入策略，進一步降低高風險高齡者之死亡風險。

Background With accelerating global population aging, the health profile of older adults has shifted from predominantly acute illnesses to complex patterns characterized by multimorbidity, functional decline, and interacting health risks. All-cause mortality in later life is increasingly shaped by the combined effects of chronic disease burden, functional status, and psychosocial vulnerability. However, the extent to which psychosocial factors are associated with mortality risk among community-dwelling older adults, after accounting for chronic conditions and functional status, remains incompletely understood. Objectives To examine the associations of demographic characteristics, chronic diseases, functional dependence, cognitive function, and psychosocial factors (social frailty, psychological frailty, and depressive symptoms) with all-cause mortality risk among community-dwelling older adults. Methods This retrospective cohort study included community-dwelling adults aged ≥65 years who were registered residents of Changhua County, Taiwan, and participated in an integrated health screening program between 2019 and August 31, 2025. Screening data were linked to official death registry records to ascertain all-cause mortality through August 31, 2025. Univariate and multivariate Cox proportional hazards regression models were used to evaluate associations between study variables and mortality risk. Results Among 8,222 participants, 7,581 (92.2%) survived and 641 (7.8%) died during follow-up. In multivariate Cox models, higher all-cause mortality risk was independently associated with male (adjusted hazard ratio [aHR] = 2.14, 95% confidence interval [CI]: 1.28–3.59), older age (75–84 years: aHR = 2.52, 95% CI: 1.98–3.22; ≥85 years: aHR = 5.33, 95% CI: 4.06–6.99), underweight body mass index (aHR = 2.00, 95% CI: 1.42–2.81), widowhood (aHR = 1.25, 95% CI: 1.04–1.50), diabetes mellitus (aHR = 1.62, 95% CI: 1.36–1.93), and chronic kidney disease (aHR = 1.47, 95% CI: 1.12–1.93). Moderate-to-severe functional dependence was a strong predictor compared with full independence (aHR = 2.03, 95% CI: 1.62–2.54). Regarding psychosocial factors, social frailty was consistently associated with increased mortality risk (prefrail: aHR = 1.25, 95% CI: 1.04–1.50; frail: aHR = 1.26, 95% CI: 1.01–1.58). In trend analysis, each additional social frailty criterion was associated with higher mortality risk (aHR = 1.09, 95% CI: 1.00–1.19; χ² = 4.034, df = 1, p = 0.045), indicating a dose–response relationship. Psychological frailty and depressive symptoms were associated with mortality risk only in univariate analyses; after adjustment for chronic diseases, functional dependence, and psychosocial frailty measures, neither psychological frailty nor depressive symptoms remained significantly associated with all-cause mortality risk. Conclusions In community-dwelling older adults in Changhua County, male, older age, underweight BMI, widowhood, diabetes, chronic kidney disease, and moderate-to-severe functional dependence were independent predictors of all-cause mortality risk. Social frailty showed a stable association with mortality risk and a dose–response pattern, suggesting incremental risk information beyond chronic disease and functional dependence. In contrast, the risk signals of psychological frailty and depressive symptoms were attenuated after multivariate adjustment, potentially reflecting overlap or mediation through chronic conditions, functional dependence, and social frailty. Incorporating social frailty into risk stratification within integrated community screening may help identify high-risk older adults and facilitate early linkage to social support and community resources.