Using Mendelian randomization method to elucidate the causal relationship between platelet and hypertension

• Main Authors: Meng-Chun Wu, 吳孟駿
• Other Authors: Tzu-Pin Lu
• Format: Others
• Language: en_US
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/qrme8w

碩士 === 國立臺灣大學 === 流行病學與預防醫學研究所 === 107 === ABSTARCT BACKGROUND Platelet has a crucial role in the increasing thrombotic tendency from both clinical and biological perspectives, and has already been reported correlated to cardiovascular diseases. Some of the previous studies showed that platelet has correlation to hypertension as well, however, with some confounding bias or design restrictions so that the causal inference didn’t succeed. Also, there would be absence in the form of platelet count, which is one of the important platelet indices, in the previous studies. We have utilized several single nucleotide polymorphisms (SNPs) related to platelet count as instrumental variables to perform genetic risk score Mendelian randomization to make the causal inference of platelet count on hypertension. MATERIAL AND METHODS There were 15,996 healthy Taiwanese Han individuals aging from years 30 to 70 randomly picked up from the Taiwan Biobank project in our study. The baseline of platelet count and hypertension were used as an exposure and an outcome, and the clinical data and genomic data with 646,735 SNPs were all coming from the same dataset. We set 4 different scenarios to choose the appropriate SNPs and eliminated the interventions from confounding factors. The genetic risk score development and the further genetic risk score Mendelian randomization in weighted-median model was then executed to elucidate the causal relationship of platelet count on hypertension. RESULTS There was significantly positive relationship of platelet count on hypertension with only age and sex (p-value = 3.80 e-10) and all the confounding included in logistic regression analysis (p-value = 2.85 e-3). The positive causal relationships between platelet count and hypertension were also consistently statistically significant (p-value < 0.05) in 3 different scenarios with more than 2 SNPs as instrumental variables in genetic risk score Mendelian randomization. CONCLUSION Our study had enough evidence to state that platelet count had causal relationship on hypertension. Platelet count could be taken as the risk factor of the hypertension, and then related clinical applications and preventive therapeutic treatments could be considered in the near future.