国际内分泌代谢杂志2024,Vol.44Issue(5) :299-304.DOI:10.3760/cma.j.cn121383-20240722-00093

社会行为特征与1型糖尿病:一项双向孟德尔随机化研究

Causal relationships between socio-behavioural traits and type 1 diabetes: a bidirectional mendelian randomization analysis

陈阳 李爽 刘静飞 杨涛 顾愹
国际内分泌代谢杂志2024,Vol.44Issue(5) :299-304.DOI:10.3760/cma.j.cn121383-20240722-00093

社会行为特征与1型糖尿病:一项双向孟德尔随机化研究

Causal relationships between socio-behavioural traits and type 1 diabetes: a bidirectional mendelian randomization analysis

陈阳 1李爽 1刘静飞 1杨涛 1顾愹1
扫码查看

作者信息

  • 1. 南京医科大学第一附属医院内分泌科,南京 210029
  • 折叠

摘要

目的 通过孟德尔随机化分析(Mendelian randomization, MR)探讨社会行为特征与1型糖尿病(type 1 diabetes mellitus, T1DM)之间的因果关系。 方法 本研究使用社会行为特征和T1DM的全基因组关联研究(Genome-wide association study,GWAS)数据进行双向MR分析。社会行为特征主要包括饮食习惯、吸烟、饮酒、体力活动、身心健康、受教育程度等。正向分析中,选择与社会行为特征显著相关的遗传变异作为工具变量,通过逆方差加权法(inverse variance weighted,IVW)为主的多种方法进行MR分析,并从异质性、水平多效性等方面进行敏感性分析。在反向分析中,选择与T1DM显著相关的遗传变异作为工具变量,进一步探索T1DM与社会行为特征的因果关联。 结果 正向MR分析显示,体重指数(body mass index,BMI)与T1DM呈正相关(OR=1.059,95% CI:1.027~1.092,P FDR=0.004),而教育程度与T1DM呈负相关(OR=0.790,95% CI:0.730~0.855,P FDR<0.001)。反向MR分析表明,T1DM与非肿瘤疾病负担呈正相关(OR=1.030,95% CI:1.015~1.045,P FDR=0.002)。敏感性分析中,BMI、非肿瘤疾病负担结果存在异质性(Cochran’s Q检验P<0.05),受教育时长结果不存在异质性(Cochran’s Q检验P>0.05);双向结果均无水平多效性(MR-Egger截距P>0.05)。 结论 较高的BMI、较低的受教育程度在T1DM发病风险中有重要作用。 Objective To explore the causal relationship between socio-behavioural traits and type 1 diabetes mellitus (T1DM) through Mendelian Randomization (MR) analysis. Methods A bidirectional MR analysis using data from genome-wide association study (GWAS) of socio-behavioural traits and T1DM was performed. Socio-behavioural traits mainly include dietary habits, smoking, alcohol consumption, physical activity, physical and mental health, and educational attainment. In forward analysis, genetic variants significantly associated with socio-behavioural traits were selected as instrumental variables (IV). Multiple MR methods, including inverse variance weighted (IVW), were used for MR analyses, along with sensitivity analyses focusing on heterogeneity and horizontal pleiotropy. In the reverse analysis, IVs associated with T1DM were selected to further explore the causal association of T1DM with social behavior. Results The forward MR analysis revealed a significant positive causal relationship between body mass index (BMI) and T1DM (OR=1.059, 95% CI: 1.027-1.092, P FDR=0.004). Additionally, there was a significant negative association between education attainment and T1DM (OR=0.790, 95% CI: 0.730-0.855, P FDR<0.001). The reverse MR analysis indicated a positive causal link between T1DM and non-cancer illness burden (OR=1.030, 95% CI: 1.015-1.045, P FDR=0.002). Sensitivity analyses were then conducted to ensure the robustness of the results. Conclusions The present study identifies the important role of higher BMI and lower educational attainment in the risk of T1DM onset and emphasized the wide-ranging impact of T1DM on the health status of patients. Regularly checking weight fluctuations in individuals at risk for T1DM may help in early screening and timely intervention.

Abstract

Objective To explore the causal relationship between socio-behavioural traits and type 1 diabetes mellitus (T1DM) through Mendelian Randomization (MR) analysis. Methods A bidirectional MR analysis using data from genome-wide association study (GWAS) of socio-behavioural traits and T1DM was performed. Socio-behavioural traits mainly include dietary habits, smoking, alcohol consumption, physical activity, physical and mental health, and educational attainment. In forward analysis, genetic variants significantly associated with socio-behavioural traits were selected as instrumental variables (IV). Multiple MR methods, including inverse variance weighted (IVW), were used for MR analyses, along with sensitivity analyses focusing on heterogeneity and horizontal pleiotropy. In the reverse analysis, IVs associated with T1DM were selected to further explore the causal association of T1DM with social behavior. Results The forward MR analysis revealed a significant positive causal relationship between body mass index (BMI) and T1DM (OR=1.059, 95% CI: 1.027-1.092, P FDR=0.004). Additionally, there was a significant negative association between education attainment and T1DM (OR=0.790, 95% CI: 0.730-0.855, P FDR<0.001). The reverse MR analysis indicated a positive causal link between T1DM and non-cancer illness burden (OR=1.030, 95% CI: 1.015-1.045, P FDR=0.002). Sensitivity analyses were then conducted to ensure the robustness of the results. Conclusions The present study identifies the important role of higher BMI and lower educational attainment in the risk of T1DM onset and emphasized the wide-ranging impact of T1DM on the health status of patients. Regularly checking weight fluctuations in individuals at risk for T1DM may help in early screening and timely intervention.

关键词

1型糖尿病/社会行为特征/体重指数/教育程度/孟德尔随机化

Key words

Type 1 diabetes mellitus/Socio-behavioural traits/BMI/Educational attainment/Mendelian Randomization

引用本文复制引用

出版年

2024
国际内分泌代谢杂志
中华医学会,天津医科大学

国际内分泌代谢杂志

CSTPCD
影响因子:0.842
ISSN:1673-4157
段落导航相关论文