Hui Gao1, Lei Xu2. 1. School of Medicine, Ningbo University, Department of Gastroenterology, Ningbo First Hospital, Ningbo, China. 2. Department of Gastroenterology, Ningbo First Hospital, Ningbo, China.
Dear Editor:I read with interest the article by Li et al, who performed large-scale 2-sample Mendelian randomization analyses to explore the relationship between nonalcoholic fatty liver disease (NAFLD) and severe COVID-19. They concluded that there is no evidence supporting that NAFLD is a causal risk factor for severe COVID-19. I have the following queries about their study.Supplementary Table 5 showed 4.31% of patients with NAFLD were diagnosed as severe COVID-19, higher than those diagnosed as nonsevere COVID-19 (1.45%). Besides, the authors performed univariate regression analysis to assess the impact of NAFLD on severe COVID-19 in Figure 2. The odds ratio was 3.06 (P = 1.07 × 10–6). The association was significant. After adjusting for the remaining risk factors, the odds ratio was 1.61 (P = .08). P value was close to 0.05 but not significant. Thus, the author concluded that NAFLD is not an independent risk factor for severe COVID-19. However, these results seem to suggest a link between NAFLD and COVID-19. I suggest that the author should perform multiple models to prove the robustness of this result in the multivariate regression analysis.Some of the upper 95% confidence intervals of odds ratio were wrong in Supplementary Table 9. The data lack decimal points, which may lead to misinterpretation. Please check again.To further test the hypothesis that other NAFLD-associated comorbid factors may causally confound the associations, I suggest that the authors perform multivariable Mendelian randomization instead of two-sample Mendelian randomization analyses in Figure 3. It could reduce bias as far as possible after adjusting for comorbid factors in multivariable Mendelian randomization.A lot of previous studies have shown that admission C-reactive protein is associated with disease severity in patients with COVID-19.2, 3, 4, 5 However, it seems that no association was found between C-reactive protein and severe COVID-19 in Figure 3. It would be better if the authors explain.Given the large impact of NAFLD and COVID-19 on human health, further studies are required to assess whether NAFLD is a risk factor for severe COVID-19 disease and requires active intervention.