Fried表型和衰弱指数与死亡风险关联的比较分析  

Comparison of Fried phenotype and frailty index and their associations with risk of mortality

作　　者：樊俊宁 孙至佳 余灿清 郭彧[1,3] 孙点剑一 裴培[1,3] 杜怀东 陈君石 陈铮鸣[1,4] 吕筠 李立明[1,2] Fan Junning;Sun Zhijia;Yu Canqing;Guo Yu;Sun Dianjianyi;Pei Pei;Du Huaidong;Chen Junshi;Chen Zhengming;Lyu Jun;Li Liming(the China Kadoorie Biobank Collaborative Group;Department of Epidemiology&Biostatistics,School of Public Health,Peking University/Peking University Center for Public Health and Epidemic Preparedness&Response/Key Laboratory of Molecular Cardiovascular Sciences(Peking University),Ministry of Education,Beijing 100191,China;Chinese Academy of Medical Sciences,Beijing 100730,China;Medical Research Council Population Health Research Unit at the University of Oxford/Clinical Trial Service Unit and Epidemiological Studies Unit,Nuffield Department of Population Health,University of Oxford,Oxford OX37LF,UK;China National Center for Food Safety Risk Assessment,Beijing 100022,China)

机构地区：[1]中国慢性病前瞻性研究项目协作组 [2]北京大学公共卫生学院/北京大学公众健康与重大疫情防控战略研究中心/北京大学分子心血管学教育部重点实验室,100191 [3]中国医学科学院,北京100730 [4]英国牛津大学医学研究委员会人口健康研究组/英国牛津大学临床与流行病学研究中心纳菲尔德人群健康系,OX37LF [5]国家食品安全风险评估中心,北京100022

出　　处：《中华流行病学杂志》2021年第7期1179-1187,共9页Chinese Journal of Epidemiology

基　　金：国家自然科学基金(81941018);中国香港Kadoorie Charitable基金。

摘　　要：目的比较Fried表型以及由不同数量的疾病缺陷构成的衰弱指数(FI)对衰弱状态评价的一致性以及与死亡风险的前瞻性关联。方法利用中国慢性病前瞻性研究(CKB)第二次重复调查的23615名研究对象的数据,采用5种表型指标构建Fried表型,并分别纳入28个和40个疾病缺陷构建FI-28和FI-40。计算加权Kappa系数比较3种指标对衰弱状态分类的一致性。采用Cox比例风险模型分析衰弱指标与死亡风险的关联。结果采用Fried表型、FI-28和FI-40计算的衰弱率分别为5.4%、7.9%和4.0%。Fried表型与FI-28和FI-40的Kappa系数分别为0.357和0.408,FI-28与FI-40的Kappa系数为0.712。经过(3.9±0.5)年的随访,死亡755人。当采用Fried表型时,与无衰弱组相比,衰弱前期和衰弱组的死亡风险均增加,多因素调整后的风险比(HR)(95%CI)分别为1.60(1.32~1.94)和2.90(2.25~3.73);采用FI-28时,衰弱前期和衰弱组的死亡HR值分别为1.71(1.39~2.11)和2.52(1.95~3.27);采用FI-40时,衰弱前期和衰弱组的死亡HR值分别为1.98(1.60~2.44)和3.71(2.80~4.91)。衰弱状态与死亡风险的关联在不同年龄组间存在差异,在低年龄组中的关联强度高于高年龄组。结论Fried表型和基于不同数量的变量构建的FI表现出较好的一致性,都能较好地预测死亡风险。Objective To compare the consistency of frailty status measured by Fried phenotype and frailty index composed of different numbers of deficits,and their prospective associations with risk of mortality.Methods Data of 23615 participants from the second resurvey of the China Kadoore Biobank(CKB)was used.Fried phenotype was constructed using five phenotypes,and frailty indexes(FI)were constructed using 28 and 40 deficits,respectively.We calculated the Weighted Kappa coefficient to compare the consistency of three measures in the classification of frailty status.Cox regression was performed to analyze the association of frailty status with risk of mortality.Results The frailty prevalence calculated by Fried phenotype,FI-28,and FI-40 were 5.4%,7.9%,and 4.0%,respectively.The Kappa coefficients of Fried phenotype with FI-28 and FI-40 were 0.357 and 0.408,respectively.The Kappa coefficients of FI-28 and FI-40 was 0.712.During an average of(3.9±0.5)years of follow-up,755 participants died.When Fried phenotype was used,compared with the robust participants,the prefrail and frail participants had increased risk of mortality,the multivariable-adjusted HRs were 1.60(95%CI:1.32-1.94)and 2.90(95%CI:2.25-3.73),respectively.When FI-28 was used,the corresponding HRs were 1.71(95%CI:1.39-2.11)and 2.52(95%CI:1.95-3.27)for prefrail and frail participants,and when FI-40 was used,the corresponding HRs were 1.98(95%CI:1.60-2.44)and 3.71(95%CI:2.80-4.91).The association of frailty status with mortality differed in different age groups,with the association stronger in younger adults than in older adults.Conclusion Fried phenotype and frailty index constituted with different numbers of deficits showed good consistency;which can be used to well predict the risk of mortality.

关 键 词：Fried表型 衰弱指数 死亡 前瞻性队列研究 中国人 成年人 

分 类 号：R195[医药卫生—卫生统计学]