BackgroundThe association between brain and sarcopenia has not been clarified. We aim to investigate the causal association between brain structure, function, gene expression and sarcopenia‐related traits.MethodsAll participants were Europeans. GWAS data of Brain Imaging Data Structure (BIDs) was from the UK Biobank. Gene expression in 13 brain regions was acquired from the GTEx Consortium. The sarcopenia‐related traits, including appendicular lean mass (ALM), whole body lean mass (WBLM), grip strength and sarcopenia diagnosed by the European Working Group on Sarcopenia in Older People (EWGSOP) or Foundation for the National Institutes of Health (FNIH), were from the IEU website. The inverse variance weighted (IVW), MR Egger, weighted median, weighted mode and Wald ratio methods were used for a two‐sample Mendelian randomization (MR) analysis between brain structures and sarcopenia‐related traits. The summary‐data‐based MR (SMR) was used to investigate brain genes causally influencing sarcopenia. We calculated the F‐value, odds ratio with 95% CI and p‐value. For the sensitivity analysis, the heterogeneity I2 statistic, Cochrane's Q test, Egger's intercept test, MR‐PRESSO and the leave‐one‐out sensitivity test were used. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway, protein–protein interaction (PPI), transcription factor interaction prediction and miRNA interaction prediction analysis were used to reveal potential signalling pathways and mechanisms.ResultsWe included 3144 imaging phenotypes from 8428 participants in BIDs data. One hundred forty‐one BIDs were identified to causally influence ALM, 160 BIDs showed significant causal effect on the WBLM, and 86 BIDs showed significant causal effect on grip strength. There were 48 or 32 BIDs causally associated with sarcopenia diagnosed by the EWGSOP or FNIH criteria respectively. After FDR correction, there were 35 BIDs showing causal effect on the ALM, 28 BIDs showing causal effect on the WBLM and 7 BIDs showing causal effect on grip strength (p < 0.05). Twelve to forty‐eight genes in different brain regions showed causal effect on all the five sarcopenia‐related traits. MMP24‐AS1, HLA‐DRB6, HLA‐DQA2, DDX42, BAG6, NUSAP1, LINC00940, NME1‐NME2 and AS3MT in the amygdala region showed detrimental effect on all the five sarcopenia‐related traits, whereas HLA‐DRB1, HLA‐DQB1‐AS1 and C6ORF3 showed protective effect (p < 0.05). The gene enrichment analysis indicated these screened genes was mainly enriched in immune‐related signalling.ConclusionWe discovered the causal effect of BIDs and brain gene expression on sarcopenia. The positive genes were mainly enriched in immune‐related signalling, suggesting an immune‐based cross‐organ regulation mechanism of brain–muscle axis.

中文翻译：

---

大脑对肌肉结构的遗传影响：肌肉减少症的孟德尔随机化研究


背景大脑与肌肉减少症之间的关联尚未阐明。我们旨在研究大脑结构、功能、基因表达和肌肉减少症相关特征之间的因果关系。方法所有参与者均为欧洲人。脑成像数据结构 （BID） 的 GWAS 数据来自英国生物样本库。13 个脑区的基因表达是从 GTEx Consortium 获得的。欧洲老年人肌肉减少症工作组 （EWGSOP） 或美国国立卫生研究院基金会 （FNIH） 诊断的肌肉减少症相关特征，包括四肢瘦体重 （ALM）、全身瘦体重 （WBLM）、握力和肌肉减少症，均来自 IEU 网站。逆方差加权 （IVW） 、 MR Egger 、加权中位数、加权模式和 Wald 比率方法用于大脑结构和肌肉减少症相关特征之间的双样本孟德尔随机化 （MR） 分析。基于汇总数据的 MR （SMR） 用于研究因果影响肌肉减少症的脑基因。我们计算了 F 值、比值比和 95% CI 和 p 值。对于敏感性分析，使用异质性 I2 统计量、Cochrane Q 检验、Egger 截距检验、MR-PRESSO 和留一法敏感性检验。采用京都基因与基因组百科全书 （KEGG） 通路、蛋白质-蛋白质相互作用 （PPI） 、转录因子相互作用预测和 miRNA 相互作用预测分析来揭示潜在的信号通路和机制。结果我们在 BID 数据中纳入了来自 8428 名参与者的 3144 种影像学表型。确定 141 个 BID 对 ALM 有因果影响，160 个 BID 对 WBLM 有显著的因果影响，86 个 BID 对握力有显著的因果影响。 分别有 48 或 32 例 BID 与 EWGSOP 或 FNIH 标准诊断的肌肉减少症有因果关系。FDR 校正后，有 35 个 BID 显示对 ALM 的因果影响，28 个 BID 显示对 WBLM 的因果影响，7 个 BID 显示对握力的因果影响 （p < 0.05）。不同大脑区域的 12 到 48 个基因对所有 5 个肌肉减少症相关特征都显示出因果影响。杏仁核区域的 MMP24-AS1、HLA-DRB6、HLA-DQA2、DDX42、BAG6、NUSAP1、LINC00940、NME1-NME2 和 AS3MT 对所有五个肌肉减少症相关性状均显示出不利影响，而 HLA-DRB1、HLA-DQB1-AS1 和 C6ORF3 显示出保护作用 （p < 0.05）。基因富集分析表明，这些筛选的基因主要富集于免疫相关信号。结论我们发现了 BID 和脑基因表达对肌肉减少症的因果影响。阳性基因主要富集于免疫相关信号传导，提示脑肌轴存在基于免疫的跨器官调控机制。