Skip to main content
Elsevier - PMC COVID-19 Collection logoLink to Elsevier - PMC COVID-19 Collection
letter
. 2021 Sep 14;84(2):248–288. doi: 10.1016/j.jinf.2021.09.002

Mendelian randomization reveals potential causal candidates for COVID-19 in 123 blood metabolites

Shizheng Qiu a,1,, Donghua Wang b,1, Yu Zhang c,1, Yang Hu a,
PMCID: PMC8437744  PMID: 34530029

Dear editor

We read with interest the recently published letter in Journal of Infection by Yang et al., who conducted a quantitative meta-analysis of twenty-seven studies containing 146,364 cases and found a lack of association between dyslipidemia and COVID-19 mortality.1 However, Hilser et al. investigated the relationship between metabolic syndrome-related serum biomarkers and the severity of COVID-19 in UK Biobank and found that an increase of 10 mg/dl HDL-Cholesterol reduced the risk of suffering from COVID-19 by 13% (OR = 0.87, 95% CI: 0.79–0.94, P = 1.2E-03).2 In fact, most studies collected by Yang et al. neglected to distinguish the subtypes of dyslipidemia in detail, such as the levels of low-density lipoprotein (LDL), high-density lipoprotein (HDL), triglyceride and cholesterol, etc.1 Inconsistent findings from previous studies sparked our interest to further investigate the causal relationship between blood metabolite levels and COVID-19.

In this study, we aimed to clarify causal candidates for influencing COVID-19 infection rates among 123 blood metabolites. In stage 1, we obtained the genome-wide association study (GWAS) summary statistics of 123 blood metabolites on nearly 25,000 individuals by Kettunen et al.3 We obtained COVID-19 GWAS summary statistics from COVID-19 Host Genetics Initiative Round 4, including 14,134 COVID-19 cases and 1,284,876 controls. All the participants were of European descent. In stage 2, according to the assumptions of Mendelian randomization (MR) model, we chose SNPs that were strongly correlated with exposure (blood metabolites) as instruments (P < 5E-08).4 We only kept the blood metabolites with more than 2 instruments, allowed SNPs with LD Rsq value > 0.8 as proxy SNPs in the outcome (COVID-19), and aligned strands for palindromic SNPs.4 In stage 3, we performed MR analysis using the inverse variance weighted (IVW) method, and assessed the horizontal pleiotropy of the instruments using a sensitivity test4. MR analysis results showed that three blood metabolite concentrations were causally related to the higher risk of COVID-19, increasing the COVID-19 infection rate by 24%, 13%, and 10%, respectively, but none of them passed the adjusted significance threshold (P < 0.05/123 = 0.00041) (Table 1 ).

Table 1.

MR analysis of blood metabolites and COVID-19.

Metabolites Methods NSNP Beta SE P OR
Citrate IVW 3 0.213 0.106 0.0452 1.24 (1.01–1.52)
MR-Egger 3 0.914 1.67 0.681 2.49 (0.09–65.84)
M.VLDL.C IVW 11 0.121 0.0573 0.0349 1.13 (1.01–1.26)
MR-Egger 11 −0.0653 0.134 0.6374 0.94 (0.72–1.22)
S.VLDL.P IVW 12 0.0944 0.0462 0.0411 1.10 (1.00–1.20)
MR-Egger 12 −0.0664 0.119 0.590 0.94 (0.74–1.18)

VLDL: Very low density lipoprotein; M.VLDL.C: Total cholesterol in medium VLDL; S.VLDL.P: Concentration of small VLDL particles. SE: standard error. The statistically significant association of IVW is defined to be P < 0.05/123 = 0.00041. The statistically significance of no horizontal pleiotropy is defined to be P > 0.05.

In conclusion, we supported that the blood metabolite concentration might not be related to the risk of COVID-19, and therefore patients with dyslipidemia might not need specific treatment and medication, so as to avoid potential waste of medical resources.1

Acknowledgment

This work was supported by the National Key R&D Program of China (2017YFC1201201, 2018YFC0910504 and 2017YFC0907503), the Natural Science Foundation of China (61801147 and 82003553) and Heilongjiang Postdoctoral Science Foundation (LBH-Z6064).

We thank MR-base R package and COVID-19 Host Genetics Initiative for providing GWAS summary statistics.

Reference

  • 1.Yang H., Hou H., Liang X., Xu J., Wang Y. Lack of significant association between dyslipidemia and COVID-19 mortality. J Infect. 2021;82(6):276–316. doi: 10.1016/j.jinf.2021.03.001. JunPubMed PMID: 33684401. Pubmed Central PMCID: PMC7934672. Epub 2021/03/09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Hilser J.R., Han Y., Biswas S., Gukasyan J., Cai Z., Zhu R., et al. Association of serum HDL-cholesterol and apolipoprotein A1 levels with risk of severe SARS-CoV-2 infection. J Lipid Res. 2021;62 doi: 10.1016/j.jlr.2021.100061. Mar 2PubMed PMID: 33667465. Pubmed Central PMCID: PMC7923911. Epub 2021/03/06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Kettunen J., Demirkan A., Wurtz P., Draisma H.H., Haller T., Rawal R., et al. Genome-wide study for circulating metabolites identifies 62 loci and reveals novel systemic effects of LPA. Nat Commun. 2016;7:11122. doi: 10.1038/ncomms11122. Mar 23PubMed PMID: 27005778. Pubmed Central PMCID: PMC4814583. Epub 2016/03/24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hemani G., Zheng J., Elsworth B., Wade K.H., Haberland V., Baird D., et al. The MR-base platform supports systematic causal inference across the human phenome. Elife. 2018;7 doi: 10.7554/eLife.34408. May 30PubMed PMID: 29846171. Pubmed Central PMCID: PMC5976434. Epub 2018/05/31. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Infection are provided here courtesy of Elsevier

RESOURCES