孟德尔随机化
肌萎缩
医学
骨质疏松症
全基因组关联研究
内科学
单核苷酸多态性
股骨颈
多效性
内分泌学
肿瘤科
生物信息学
遗传学
遗传变异
基因型
基因
表型
生物
作者
J. ZHU,Shijie Zhou,Ming Jiang,Wenying Wang,Jun Yan
出处
期刊:Medicine
[Wolters Kluwer]
日期:2025-08-29
卷期号:104 (35): e44153-e44153
标识
DOI:10.1097/md.0000000000044153
摘要
Studies show excessive cortisol is linked to osteoporosis (OP). However, the impact of mild cortisol excess (MCE) on bone mineral density (BMD) remains unclear. Sarcopenia may play a key role in this, particularly in aging or stress contexts. To investigate the association between MCE and OP outcomes and the proportion of this association that is mediated through SP using Mendelian randomization (MR). MR study using summary statistics with traits were obtained from publicly available genome-wide association studies (GWAS) of OP and sarcopenia-related traits. Three single-nucleotide polymorphisms (SNPs) associated with plasma cortisol concentrations in the CORtisol NETwork consortium were used as instrumental variables. All participants were of European ancestry. Random-effect inverse-variance weighted (IVW) method was used as the main analysis of MR, and a series of sensitivity analyses were performed to detect heterogeneity and horizontal pleiotropy. A 2-step MR approach was used to investigate whether the mediating pathway from MCE to OP was mediated by sarcopenia-related traits. Cortisol increases OP risk in the femoral neck (FN, OR = 0.874, 95% CI: 0.786–0.973, P = .0138), while reducing it in the lumbar spine (LS, OR = 1.140, 95% CI: 1.008–1.291, P = .037) and heel (OR = 1.027, 95% CI: 1.000–1.054, P = .047). The mediation analysis via 2-step MR showed that sarcopenia mediates up to 8.4% of the cortisol-induced osteoporosis risk in the LS. This MR analysis suggests that MCE primarily increases OP risk in the FN, rather than affecting the body systemically. It also shows that SP mediates up to 8.4% of the cortisol-induced OP risk in the lumbar spine, highlighting the importance of muscle-strengthening exercises in preventing OP.
科研通智能强力驱动
Strongly Powered by AbleSci AI