key: cord-0947842-ug6s7g0s
authors: Li, Jiuling; Tian, Aowen; Zhu, Haoxue; Chen, Lanlan; Wen, Jianping; Liu, Wanqing; Chen, Peng
title: Mendelian Randomization Analysis Reveals No Causal Relationship Between Nonalcoholic Fatty Liver Disease and Severe COVID-19
date: 2022-02-03
journal: Clin Gastroenterol Hepatol
DOI: 10.1016/j.cgh.2022.01.045
sha: f60c9b5bd18ddaf4b8a7d2e2406677945a60dc3f
doc_id: 947842
cord_uid: ug6s7g0s

Background & Aims The pandemic of coronavirus disease 2019 (COVID-19) has witnessed more than 4.5 million deaths as of the time of writing. Whether nonalcoholic fatty liver disease (NAFLD) increase the risk for severe COVID-19 remains unclear. We sought to address this question using two-sample Mendelian randomization (TSMR) analysis approaches in large cohorts. Methods We performed large-scale TSMR analyses to examine whether there is a causal relationship between NAFLD, serum alanine aminotransferase (ALT), grade of steatosis, NAFLD Activity Score (NAS), or fibrosis stage and severe COVID-19. To maximize the power of this analysis, we performed a genome-wide meta-analysis (GWMA) to identify single-nucleotide polymorphisms (SNPs) associated with NAFLD. We also examined the impact of 20 major co-morbid factors of NAFLD on severe COVID-19. Results Univariate analysis of the UK Biobank (UKB) data demonstrated a significant association between NAFLD and severe COVID-19 (OR=3.06, p=1.07×10−6). However, this association disappeared after demographic and co-morbid factors adjusted (OR=1.61, p=0.08). TSMR study indicated that NAFLD (OR=0.97, p=0.61), ALT level (OR=1.03, p=0.47), grade of steatosis (OR=1.08, p=0.41), NAS (OR=1.02, p=0.39), and fibrosis stage (OR=1.01, p=0.87) were not associated with severe COVID-19. Among all NAFLD-related co-morbid factors, BMI (OR=1.73, p=7.65×10−9), waist circumference (WC, OR=1.76, p=2.58×10−5), and hip circumference (HC, OR=1.33, p=7.26×10−3) were the only ones demonstrated a causal impact on severe COVID-19. Conclusions There is no evidence supporting that NAFLD is a causal risk factor for severe COVID-19. Previous observational associations between NAFLD and COVID-19 are likely attributed to the correlation between NAFLD and obesity.

instrumental variables are valid 4 . In addition, we conducted MR-PRESSO tests to correct for horizontal pleiotropic effect by removing IV outliers 5 . Furthermore, we estimated the heterogeneity of instrumental variables using Cochran's Q test.

The F-statistic was calculated by the following formula to estimate the statistical power of instrumental variables:

where N is the sample size, K is the number of IVs, and R 2 is the proportion of the variability of the exposure explained by instrumental variables. R 2 was calculated with reference to the work of Nikos et al 6 . MR Power calculation was performed using mRnd 7 .

J o u r n a l P r e -p r o o f [Abbreviations: Nsnp-number of IVs used in the TSMR analysis, Beta-causal effect of the exposure on severe COVID-19, SE-standard error, P-p-value of the TSMR analysis, Het-p-p-value of J o u r n a l P r e -p r o o f

Ple-p-p-value of MR-Egger intercept, OR-An odds ratio is a measure of association between the risk factor and severe COVID-19

the effect of the exposure on severe COVID-19 is expressed the change in severe COVID-19 for each 2.72-fold increase in the exposure

UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age

Interpreting findings from Mendelian randomization using the MR-Egger method

Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: the role of the I2 statistic

Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator

Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases

Physical activity and risks of breast and colorectal cancer: a Mendelian randomisation analysis

Calculating statistical power in Mendelian randomization studies

There is no evidence supporting that NAFLD, ALT level, grade of steatosis, NAFLD Activity Score, and fibrosis stage were causally associated with severe COVID-19

Among common NAFLD-related co-morbid factors, BMI, waist circumference, and hip