Department of Epidemiology and Biostatistics, College of Public Health, Zhengzhou University, Zhengzhou, 450001, Henan, People's Republic of China.
Tulane Center for Biomedical Informatics and Genomics, School of Medicine, Tulane University, New Orleans, LA, 70112, USA.
Osteoporos Int. 2021 Jul;32(7):1359-1367. doi: 10.1007/s00198-020-05786-2. Epub 2021 Jan 13.
This study estimates causality of physical activity (PA) on bone mineral density (BMD) by conducting multivariable Mendelian randomization (MR). The findings suggest that habitual vigorous PA increases lumbar spine BMD, and higher overall acceleration average would improve forearm BMD. The results could promote PA intervention targeting individuals with optimized type.
Evidence from epidemiologic studies showed type, frequency, and duration of PA influenced BMD. However, these observational studies may be confounded by many factors, resulting in spurious associations. We aimed to conduct multivariable MR to estimate the causal effect of self-reported and device-measured PA on osteoporosis.
Three self-reported and two device-measured PA-related traits were selected as exposures. Outcomes were BMD at different skeletal sites: femoral neck BMD (FN BMD), lumbar spine BMD (LS BMD), and forearm BMD (FA BMD). Exposure datasets were obtained from UK Biobank with total 377,234 subjects. Outcome datasets were obtained from GEFOS consortium with 53,236 subjects. Standard MR analysis and multivariable MR were conducted to assess the total and direct causal effect of PA on BMD.
For self-reported PA, inverse-normalized moderate-to-vigorous had a direct causal effect on FN BMD independently (β = - 1.116 (95% confidence interval, 95%CI: - 2.210, - 0.023), P = 0.045); vigorous PA showed a direct effect (β = 3.592 (95%CI: 0.310, 6.874), P = 0.032) on LS BMD independently. While overall acceleration average and fraction of accelerations both had a direct causal effect on FA BMD independently.
Habitual vigorous PA could increase LS BMD. Individuals with higher overall acceleration average would have a higher FA BMD.
本研究通过多变量孟德尔随机化（MR）来估计体力活动（PA）对骨密度（BMD）的因果关系。研究结果表明，习惯性剧烈 PA 可增加腰椎骨密度，而较高的总体加速度平均值可改善前臂骨密度。这些结果可以促进针对优化类型个体的 PA 干预。
来自流行病学研究的证据表明，PA 的类型、频率和持续时间会影响 BMD。然而，这些观察性研究可能受到许多因素的混淆，导致虚假关联。我们旨在进行多变量 MR，以估计自我报告和设备测量的 PA 对骨质疏松症的因果影响。
选择三种自我报告和两种设备测量的 PA 相关特征作为暴露因素。结果是不同骨骼部位的 BMD：股骨颈 BMD（FN BMD）、腰椎 BMD（LS BMD）和前臂 BMD（FA BMD）。暴露数据集来自英国生物库，共 377234 名受试者。结果数据集来自 GEFOS 联盟，共 53236 名受试者。标准 MR 分析和多变量 MR 用于评估 PA 对 BMD 的总影响和直接因果关系。
对于自我报告的 PA，中等至剧烈的反向归一化具有独立的 FN BMD 直接因果效应（β=-1.116（95%置信区间，95%CI：-2.210，-0.023），P=0.045）；剧烈 PA 对 LS BMD 具有独立的直接效应（β=3.592（95%CI：0.310，6.874），P=0.032）。而总体加速度平均值和加速度分数都具有独立的 FA BMD 直接因果效应。
习惯性剧烈 PA 可增加 LS BMD。总体加速度平均值较高的个体 FA BMD 较高。
