Identifying potential causal effects of Parkinson's disease: A polygenic risk score-based phenome-wide association and mendelian randomization study in UK Biobank

Changhe Shi,Dongrui Ma,Mengjie Li,Zhiyun Wang,Chenwei Hao,Yuanyuan Liang,Yanmei Feng,Zhengwei Hu,Xiaoyan Hao,Mengnan Guo,Shuangjie Li,Chunyan Zuo,Yuemeng Sun,Mibo Tang,Chengyuan Mao,Chan Zhang,Yuming Xu,Shilei Sun
DOI: https://doi.org/10.1038/s41531-024-00780-5
2024-09-06
Abstract:There is considerable uncertainty regarding the associations between various risk factors and Parkinson's Disease (PD). This study systematically screened and validated a wide range of potential PD risk factors from 502,364 participants in the UK Biobank. Baseline data for 1851 factors across 11 categories were analyzed through a phenome-wide association study (PheWAS). Polygenic risk scores (PRS) for PD were used to diagnose Parkinson's Disease and identify factors associated with PD diagnosis through PheWAS. Two-sample Mendelian randomization (MR) analysis was employed to assess causal relationships. PheWAS results revealed 267 risk factors significantly associated with PD-PRS among the 1851 factors, and of these, 27 factors showed causal evidence from MR analysis. Compelling evidence suggests that fluid intelligence score, age at first sexual intercourse, cereal intake, dried fruit intake, and average total household income before tax have emerged as newly identified risk factors for PD. Conversely, maternal smoking around birth, playing computer games, salt added to food, and time spent watching television have been identified as novel protective factors against PD. The integration of phenotypic and genomic data may help to identify risk factors and prevention targets for PD.
What problem does this paper attempt to address?