Armstrong Joshua J, Mitnitski Arnold, Launer Lenore J, White Lon R, Rockwood Kenneth
Department of Medicine, Dalhousie University, Nova Scotia, Canada.
Laboratory of Epidemiology, Demography, and Biometry, National Institute on Aging, Bethesda, Maryland.
J Gerontol A Biol Sci Med Sci. 2015 Jan;70(1):125-31. doi: 10.1093/gerona/glu089. Epub 2014 Jun 27.
A frailty index (FI) based on the accumulation of deficits typically has a submaximal limit at about 0.70. The objectives of this study were to examine how population characteristics of the FI change in the Honolulu-Asia Aging Study cohort, which has been followed to near-complete mortality. In particular, we were interested to see if the limit was exceeded.
Secondary analysis of six waves of the Honolulu-Asia Aging Study. Men (n = 3,801) aged 71-93 years at baseline (1991) were followed until death (N = 3,455; 90.9%) or July 2012. FIs were calculated across six waves and the distribution at each wave was evaluated. Kaplan-Meier analyses and Cox proportional hazard models were performed to examine the relationship of frailty with mortality.
At each wave, frailty was nonlinearly associated with age, with acceleration in later years. The distributions of the FIs were skewed with long right tails. Despite the increasing mortality in each successive wave, the 99% submaximal limit never exceeded 0.65. The risk of death increased with increasing values of the FI (eg, the hazard rate increased by 1.44 [95% CI = 1.39-1.49] with each increment in the baseline FI grouping). Depending on the wave, the median survival of people with FI more than 0.5 ranged 0.84-2.04 years.
Even in a study population followed to almost complete mortality, the limit to deficit accumulation did not exceed 0.65, confirming a quantifiable, maximum number of health deficits that older men can tolerate.
基于缺陷累积的衰弱指数（FI）通常有一个约为0.70的次最大极限值。本研究的目的是考察在随访至接近完全死亡的檀香山亚洲老年研究队列中，FI的人群特征如何变化。我们尤其感兴趣的是，该极限值是否会被超过。
对檀香山亚洲老年研究的六波数据进行二次分析。以1991年为基线，纳入年龄在71 - 93岁的男性（n = 3801），随访至死亡（N = 3455；90.9%）或2012年7月。计算六波数据的FI，并评估每一波的分布情况。采用Kaplan - Meier分析和Cox比例风险模型来考察衰弱与死亡率之间的关系。
在每一波中，衰弱与年龄呈非线性相关，且在晚年加速。FI的分布呈右偏态且有长右尾。尽管在后续每一波中死亡率都在增加，但99%的次最大极限值从未超过0.65。死亡风险随着FI值的增加而增加（例如，基线FI分组每增加一级，风险率增加1.44 [95% CI = 1.39 - 1.49]）。根据不同波次，FI大于0.5的人群的中位生存期为0.84 - 2.04年。
即使在一个随访至几乎完全死亡的研究人群中，缺陷累积的极限值也未超过0.65，这证实了老年男性能够耐受的可量化的最大健康缺陷数量。
