Handgrip strength (HGS), a measure of muscular strength, might be a risk indicator for cognitive functioning, but the evidence is not consistent. Using a new prospective study and meta-analysis of published observational cohort studies, we aimed to evaluate the prospective associations of HGS with poor cognitive outcomes including cognitive impairment, dementia and Alzheimer’s disease (AD). Handgrip strength, measured using a Martin-Balloon-Vigorimeter, was assessed at baseline in a population-based sample of 852 men and women with good cognitive function in the Kuopio Ischemic Heart Disease cohort. Hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated for cognitive outcomes. Relevant published studies were sought in MEDLINE, Embase and Web of Science from inception until October 2021 and pooled using random effects meta-analysis. During a median follow-up of 16.6 years, 229 dementia cases were recorded. Comparing extreme tertiles of HGS, the multivariable adjusted HR (95% CI) for dementia, AD and vascular dementia was 0.77 (0.55–1.07), 0.75 (0.52–1.10) and 0.49 (0.16–1.48), respectively. In a meta-analysis of 16 population-based prospective cohort studies (including the current study) comprising 180,920 participants, the pooled multivariable adjusted relative risks (95% CIs) comparing the top vs bottom thirds of HGS levels were as follows: 0.58 (0.52–0.65) for cognitive impairment; 0.37 (0.07–1.85) for cognitive decline; 0.73 (0.62–0.86) for dementia; 0.68 (0.53–0.87) for AD; and 0.48 (0.32–0.73) for vascular dementia. GRADE quality of evidence ranged from low to very low. Meta-analysis of aggregate prospective data suggests that HGS may be a risk indicator for poor cognitive outcomes such as cognitive impairment, dementia and AD. Systematic review registration: PROSPERO 2021: CRD42021237750.

握力 (HGS) 是衡量肌肉力量的指标，可能是认知功能的风险指标，但证据并不一致。我们使用一项新的前瞻性研究和对已发表的观察性队列研究的荟萃分析，旨在评估 HGS 与不良认知结果（包括认知障碍、痴呆和阿尔茨海默病 (AD)）的前瞻性关联。握力强度是使用 Martin-Balloon-Vigorimeter 测量的，在库奥皮奥缺血性心脏病队列中对 852 名具有良好认知功能的男性和女性的人口样本进行了基线评估。对认知结果的风险比 (HR) 和 95% 置信区间 (CI) 进行了估计。从一开始到 2021 年 10 月，在 MEDLINE、Embase 和 Web of Science 中寻找相关已发表的研究，并使用随机效应荟萃分析进行汇总。在 16.6 年的中位随访期间，记录了 229 例痴呆病例。比较 HGS 的极端三分位数，痴呆、AD 和血管性痴呆的多变量调整 HR (95% CI) 分别为 0.77 (0.55–1.07)、0.75 (0.52–1.10) 和 0.49 (0.16–1.48)。在一项包含 180,920 名参与者的 16 项基于人群的前瞻性队列研究（包括当前研究）的荟萃分析中，比较 HGS 水平最高和最低三分之一的合并多变量调整相对风险（95% CI）如下：0.58 (0.52 –0.65) 认知障碍；认知能力下降 0.37 (0.07–1.85)；痴呆症 0.73 (0.62–0.86)；AD 为 0.68 (0.53–0.87)；血管性痴呆为 0.48 (0.32–0.73)。GRADE 证据质量从低到极低。对汇总前瞻性数据的荟萃分析表明，HGS 可能是认知障碍、痴呆和 AD 等不良认知结果的风险指标。系统评价注册：PROSPERO 2021：CRD42021237750。