Chen Huiling, Huang Shixin, Zhou Kaiji, Zhang Tao, Tse Mimi Mun Yee, Tse Gary
Department of Applied Social Sciences, The Hong Kong Polytechnic University, Hong Kong.
Department of Applied Social Sciences, The Hong Kong Polytechnic University, Hong Kong.
Arch Gerontol Geriatr. 2025 Aug 16;139:105991. doi: 10.1016/j.archger.2025.105991.
Heart rate variability (HRV) is a potential biomarker for frailty, which is burdensome in older adults. However, existing studies report inconsistent findings regarding their relationship, and no meta-analysis has been conducted. This systematic review and meta-analysis aim to examine the association between HRV and frailty, as well as the boundary conditions that influence this relationship.
We systematically searched eight databases for studies examining three domains of HRV parameters - time-, frequency-, and nonlinear-domain - and their associations with frailty. Using a three-level random-effects meta-analytic model, we investigated the pooled effect sizes (Hedges' g) for short-term (≤ 5 minutes) and long-term (24-hour) HRV, and potential variables moderating the relationship between HRV and frailty.
We included 16 studies (n = 2365) in the review from 1132 records, and 13 of them were meta-analyzed (n = 1677) contributing 90 effect sizes. Both short-term (7 studies, 47 effect sizes) and long-term (6 studies, 43 effect sizes) models indicated lower HRV in (pre)frail individuals compared to their non-frail counterparts, while the effects were not statistically significant. For long-term HRV, the relationship was moderated by HRV domain: time-domain parameters exhibited a larger effect size (g = 2.097) than frequency-domain parameters (g = 1.698).
Our meta-analyses revealed a trend towards lower HRV in (pre)frail individuals; however, the effects in both models were not statistically significant. Although HRV may be a potential biomarker of frailty, the current evidence remains inconclusive. Further prospective studies are necessary, considering the heterogeneity and potential publication bias observed.
心率变异性（HRV）是衰弱的一种潜在生物标志物，衰弱在老年人中是一个负担。然而，现有研究关于它们之间的关系报告了不一致的结果，并且尚未进行荟萃分析。本系统评价和荟萃分析旨在研究HRV与衰弱之间的关联，以及影响这种关系的边界条件。
我们系统地检索了八个数据库，以查找研究HRV参数三个领域（时域、频域和非线性域）及其与衰弱关联的研究。使用三级随机效应荟萃分析模型，我们研究了短期（≤5分钟）和长期（24小时）HRV的合并效应量（Hedges' g），以及调节HRV与衰弱之间关系的潜在变量。
我们从1132条记录中纳入了16项研究（n = 2365）进行综述，其中13项进行了荟萃分析（n = 1677），贡献了90个效应量。短期（7项研究，47个效应量）和长期（6项研究，43个效应量）模型均表明，（预）衰弱个体的HRV低于非衰弱个体，但其效应无统计学意义。对于长期HRV，这种关系受HRV领域的调节：时域参数的效应量（g = 2.097）大于频域参数（g = 1.698）。
我们的荟萃分析揭示了（预）衰弱个体中HRV较低的趋势；然而，两个模型中的效应均无统计学意义。尽管HRV可能是衰弱的潜在生物标志物，但目前的证据仍不确凿。考虑到观察到的异质性和潜在的发表偏倚，有必要进行进一步的前瞻性研究。
