台灣社區長者身、心、社會衰弱量表：切點初探

Abstract

研究背景與目的 世界人口快速老化，台灣邁入高齡社會，為提早篩檢衰弱長者進行介入措施，以延緩不良健康結果的發生。我國政策使用Study of Osteoporotic Fractures (SOF)量表篩檢身體衰弱。此外，已經有台灣研究證實身體衰弱(SOF量表、SOF+緩慢量表、SOF+握力量表)與Fried’s Frailty Phenotype量表的信效度驗證研究。不過近年心理、社會衰弱也為當今國際研究趨勢。然而，身、心、社會衰弱量表卻尚未被驗證於評估台灣長者的量表的衰弱切點。本研究目的為探討身(SOF+緩慢、SOF+握力)、心、社會衰弱量表、以及整合性衰弱量表合適台灣族群最佳切點。 研究方法 本研究使用三份地區，包括台北市南港區、中正區、彰化縣次級資料分析，最終分析樣本數在南港區471筆，中正區703筆，彰化縣806筆。資料分析分為四大部分：(1)以斯皮爾曼等級相關係數( Spearman’s Rank Correlation Coefficient ) 探討身、心、社會衰弱相關程度。(2)將身、心、社會衰弱量表與整合性衰弱量表以(Fried’s Frailty Phenotype、失能、跌倒、生活品質)為指標，進行ROC 曲線 ( Receiver Operating Characteristic Curve )探討測量工具準確性，從敏感度(Sensitiviy )、特異度(Specificity) 、約登指數(Youden’s index) 、陽性概似比(Likelihood Rate ,LR+ )、陰性概似比(Likelihood Rate ,LR-)指標尋找最適切點。(3)比較中正區、彰化縣、彰化縣非偏遠鄉鎮、彰化縣偏遠鄉鎮長者在心理、社會衰弱切點的差異(4) 以多元邏輯斯迴歸(Multiple logistic regression )分析在控制其他變項後，將身、心、社會衰弱量表與整合性衰弱量表比較Fried’s Frailty Phenotype、失能、跌倒、生活品質是否有差異。 研究結果 本研究發現身體衰弱和心理或社會衰弱達顯著中度正相關。衰弱也會提高不良健康事件的風險。 身體衰弱(SOF+緩慢、SOF+握力)指標切點「衰弱」≥2分，「衰弱前期」=1分；心理衰弱切點「衰弱」≥3分，「衰弱前期」=2分；社會衰弱切點「衰弱」≥2分，「衰弱前期」=1分。整合性指標「衰弱」≥4分，「衰弱前期」=3分，且其鑑別力比任何單一面向衰弱量表更好。根據這些切點南港區、中正區身體衰弱(SOF+緩慢)盛行率分別為22.3%、11.9%；身體衰弱(SOF+握力)盛行率分別為18.7%、14.9%；中正區、彰化縣、彰化縣非偏遠鄉鎮、彰化縣偏遠鄉鎮心理衰弱(≥3分)盛行率為9.0%、18.2%、19.5%、12.8% ；社會衰弱(≥2分)盛行率為20.0%、22.5%、22.7%、22.1%。中正區整合身(SOF+緩慢、SOF+握力)、心、社會衰弱衰弱前期分別為13.4%、16.4%，衰弱盛行率分別為24.6%、25.5%。在不良健康結果上，發現身體或心理或社會衰弱確實會使跌倒風險、不好的生活品質風險提高。 結論與建議 本研究結果證實身、心、社會衰弱彼此有顯著正相關。在社會衰弱的盛行率比起身體和心理衰弱盛行率更高。證實心理、社會衰弱重要性，及早期篩檢心理、社會衰弱有助於避免惡化成身體衰弱。 在量表的切點上，身體「衰弱」切點≥2分，「衰弱前期」=1分。心理「衰弱」切點≥3分，「衰弱前期」=2分。社會「衰弱」≥2分，「衰弱前期」=1分。整合性指標「衰弱」≥4分，「衰弱前期」=3分。整合性衰弱量表的鑑別力與信度會高於單一衰弱面向。 在城鄉心理、社會衰弱切點相同。然而，彰化縣心理、社會衰弱對以跌倒、生活品質為指標的鑑別力未達顯著，因此若採用中正區切點，則發現心理、社會衰弱比例彰化縣非偏鄉 >彰化縣全>彰化偏鄉>中正區。未來仍建議城鄉切點差異需要不同結果指標探討。此外建議各地區依需求選擇適合量表。

Background and purpose: Screening for frailty among older people to guard them against adverse health outcomes is a crucial task as the population ages. Taiwanese governments use the Study of Osteoporotic Fractures (SOF) index to screen for physical frailty . In addition, there is evidence that confirms the reliability and validity between the SOF index and Fried’s Frailty Phenotype index in Taiwan. However , psychological and social frailty is a rather unexplored concept. This study sought to find the most optimal cutoff values for physical (SOF-walking speed、SOF-handgrip), psychological (TFI, Tilburg Frailty Indicator), and social frailty indexes (QSFS, Questionnaire to define Social Frailty Status), and integral frailty indexes in Taiwanese community-dwelling older people. Methods: The study used cross-sectional survey data in three secondary sets collected from two municipalities(Nangang and Zhongzheng District ) in Taipei and Changhua , with 471 ,703,and 806 older adult participants, respectively. Spearman’s rank correlation analysis was used to assess the association of three dimensions of frailty. Receiver operator characteristic curves were generated to determine the discriminant functions for these frailty indexes. Chi-square and Multiple logistic regression analysis were applied to test whether the risks of frailty ,disability, falls, and poor quality of life are different in physical, psychological, and social frailty indexes , and integral frailty indexes. Results: Modest associations were found between physical, psychological, and social frailty. The full integrated frailty indexes, which combined the three separate indexes, showed an improved discriminant function (cutoff values: frail ≥4, pre-frail = 3) relative to the single index. The cutoff values suggested for Taiwanese older adults was as follows: physical frailty (frail ≥2, pre-frail = 1), psychological frailty (frail ≥3, pre-frail = 2), and social frailty (frail ≥2, pre-frail = 1). About 22.3% / 18.7% and 11.9% / 14.9% in Nangang and Zhongzheng District of Taipei City were physical (SOF-walking speed/ SOF-handgrip ) frail . The prevalence rates of psychological frailty (≥3 ) in Zhongzheng District, Changhua ,Changhua urban , and Changhua rural areas were 9.0%, 18.2%, 19.5%, and 12.8%, respectively; the prevalence rates of social frailty (≥2) were 20.0%, 22.5%, 22.7%, and 22.1%, respectively. In Zhongzheng District, the prevalence rates of integrated pre- frailty(=3 )(SOF-walking speed、SOF-handgrip) were 13.4% and 16.4%, respectively, and the prevalence rates of frailty (≥4)were 24.6% and 25.5%, respectively. Conclusions and recommendations : Our findings supported the use of the full integrated frailty index and provided cutoff values to screen older community-dwelling people for frailty. We found that physical, psychological or social frailty will increase the risk of falls and poor quality of life. Therefore, screening for physical, psychological, and social frailty should be promoted in the community to improve the health for the elderly in Taiwan. The cutoff values suggested for Taiwanese older adults was as follows: physical frailty (frail ≥2, pre-frail = 1), psychological frailty (frail ≥3, pre-frail = 2), social frailty (frail ≥2, pre-frail = 1) , and the integral frailty (frail ≥4, pre-frail = 3) .The discriminate ability of the integral frailty was better to include SOF-handgrip comparing to include SOF walking speed as physical frailty. Researchers can select either one of the indexes as needed. The psychological and social frailty cutoff values between urban and rural areas are the same. However, the discriminatory ability of psychological and social frailty were not statistically significantly in Changhua.Therefore, we use the cutoff values in Zhongzheng district. Putting the prevalence rates of psychological, social frailty in descending order : Changhua urban > Changhua rural areas > Changhua > Zhongzheng District. In the future, it is still recommended that the difference between urban and rural cutoff values requires different outcome indicators to comfirm.