Associations of Weight and Glucose Levels With Myasthenia Gravis: A Two-Step, Two-Sample Mendelian Randomization and Retrospective Case-Control Study.
Previous studies suggested links between obesity, glucose regulation, and autoimmune diseases, but the associations of weight and glucose levels with myasthenia gravis (MG) remain uncertain. We examined these associations and a potential indirect pathway through glucose levels.
We investigated the associations of weight and glucose levels with MG risk using Mendelian randomization (MR) and a retrospective case-control study.
We conducted two-sample MR analyses using genome-wide association study summary statistics from European populations. The clinical component retrospectively included patients with MG and healthy controls and used propensity-score matching and logistic regression.
The MR analyses suggested associations among genetically predicted weight, glucose levels, and MG risk. They also yielded a small indirect estimate through glucose levels that was opposite in direction to the direct and total effects. This opposing-direction indirect estimate was sensitive to the inferential method and was not observed in the retrospective case-control analysis. In the clinical analysis, MG cases had higher weight and lower glucose levels than controls, and both variables were associated with MG status.
Two-sample MR suggested that genetically predicted higher weight was associated with higher odds of MG, whereas genetically predicted higher glucose levels were associated with lower odds of MG. The retrospective case-control analysis showed directionally similar associations for measured weight and fasting glucose, but the indirect pathway suggested by MR was not observed in the clinical analysis.
We investigated the associations of weight and glucose levels with MG risk using Mendelian randomization (MR) and a retrospective case-control study.
We conducted two-sample MR analyses using genome-wide association study summary statistics from European populations. The clinical component retrospectively included patients with MG and healthy controls and used propensity-score matching and logistic regression.
The MR analyses suggested associations among genetically predicted weight, glucose levels, and MG risk. They also yielded a small indirect estimate through glucose levels that was opposite in direction to the direct and total effects. This opposing-direction indirect estimate was sensitive to the inferential method and was not observed in the retrospective case-control analysis. In the clinical analysis, MG cases had higher weight and lower glucose levels than controls, and both variables were associated with MG status.
Two-sample MR suggested that genetically predicted higher weight was associated with higher odds of MG, whereas genetically predicted higher glucose levels were associated with lower odds of MG. The retrospective case-control analysis showed directionally similar associations for measured weight and fasting glucose, but the indirect pathway suggested by MR was not observed in the clinical analysis.
Authors
Lin Lin, Zhu Zhu, Qu Qu, Lin Lin, Li Li, Wang Wang, Chen Chen, Wei Wei, Pan Pan, Li Li, Xu Xu, Huang Huang, Weng Weng
View on Pubmed