Circulating polyunsaturated fatty acids and COVID-19: a prospective cohort study and Mendelian randomization analysis

Abstract

Background: Higher circulating polyunsaturated fatty acids (PUFAs), especially omega-3 ones, have been linked to a better prognosis in patients of coronavirus disease 2019 (COVID-19). However, the effects and causality of pre-infection PUFA levels remain unclear.

Objective: To investigate the observational and causal associations of circulating PUFAs with COVID-19 susceptibility and severity.

Design: We first performed a prospective cohort study in UK Biobank, with 20,626 controls who were tested negative and 4,101 COVID-19 patients, including 970 hospitalized ones. Plasma PUFAs at baseline were measured by nuclear magnetic resonance, including total PUFAs, omega-3 PUFAs, omega-6 PUFAs, docosahexaenoic acid (DHA), linoleic acid (LA), and the omega-6/omega-3 ratio. Moreover, bidirectional two-sample Mendelian randomization (MR) analyses were performed to examine the causal associations of eight individual PUFAs, measured in either plasma or red blood cells, with COVID-19 susceptibility and severity using summary statistics from existing genome-wide association studies.

Results: In the observational association analysis, total PUFAs, omega-3 PUFAs, omega-6 PUFAs, DHA, and LA were associated with a lower risk of severe COVID-19. Omega-3 PUFAs and DHA were also associated with a lower risk of testing positive for COVID-19. The omega-6/omega-3 ratio was positively associated with risks of both susceptibility and severity. The forward MR analysis indicated that arachidonic acid (AA) and docosapentaenoic acid (DPA-n3) might be causally associated with a lower risk of severe COVID-19, with OR (95% CI) per one SD increase in the plasma level as 0.96 (0.94, 0.99) and 0.89 (0.81, 0.99), respectively. The reverse MR analysis did not support any causal effect of COVID-19 on PUFAs.

Conclusions: Our observational analysis supported that higher circulating PUFAs, either omega-3 or omega-6, are protective against severe COVID-19, while omega-3 PUFAs, especially DHA, were also associated with reducing COVID-19 susceptibility. Our MR analysis further supported causal associations of AA and DPA-n3 with a lower risk of severe COVID-19.

Figure 1.

Flowchart of the UK Biobank participants from recruitment to inclusion in the observational analysis.

Figure 2.. Mendelian randomization estimates of the effects of polyunsaturated fatty acids on COVID-19 severity risk. (A) Mendelian randomization analysis based on the release 5 HGI A2. (B) Mendelian randomization analysis based on the release 5 HGI B2. (C) Mendelian randomization analysis based on the release 5 HGI B1.

Odds ratios are scaled to a genetically predicted SD increase in polyunsaturated fatty acids. Associations with p-value < 0.05 were indicated with diamonds, while others with squares. Detailed summary statistics are available in Supplemental Tables 6–8. PUFA, polyunsaturated fatty acid; ALA, α-Linolenic acid; LA: linoleic acid; GLA, γ-Linoleic acid; DGLA, dihomo-γ-linoleic acid; AA, arachidonic acid; DPA-n3, docosapentaenoic acid; DTA, docosatetraenoic acid; DHA, docosahexaenoic acid; OR, odds ratio.

Figure 3.. Mendelian randomization estimates of the effects of polyunsaturated fatty acids on COVID-19 susceptibility risk based on the release 5 HGI C2.

Odds ratios are scaled to a genetically predicted SD increase in polyunsaturated fatty acids. Associations with p-value < 0.05 were indicated with diamonds, while others with squares. Detailed summary statistics are available in Supplemental Table 9. PUFA, polyunsaturated fatty acid; ALA, α-Linolenic acid; LA: linoleic acid; GLA, γ-Linoleic acid; DGLA, dihomo-γ-linoleic acid; AA, arachidonic acid; DPA-n3, docosapentaenoic acid; DTA, docosatetraenoic acid; DHA, docosahexaenoic acid; OR, odds ratio.