The relationship between polyunsaturated fatty acids and inflammation: evidence from cohort and Mendelian randomization analyses

Abstract

Background: Omega-3 (n-3) and omega-6 (n-6) polyunsaturated fatty acids (PUFAs) are thought to have anti- and pro-inflammatory roles, respectively, and influence the risk of various chronic diseases. However, it is unclear whether these associations are causal.

Methods: We examined the associations of dietary polyunsaturated FAs with biomarkers of systemic inflammation: C-reactive protein (CRP), glycoprotein acetyls (GlycA), and interleukin 6 (IL-6) in two cohort datasets-Avon Longitudinal Study of Parents and Children (N = 2802) and UK Biobank (N = 12 401)-by using multivariable analyses. We investigated causality by using two-sample Mendelian randomization (MR). In addition to the inverse-variance weighted (IVW) method, we used sensitivity analyses to strengthen the causal inference. We conducted multivariable MR (MVMR) to investigate the causal effects of n-3 and n-6 on inflammation, accounting for the low-density lipoprotein (LDL) cholesterol, triglycerides, monounsaturated FAs, and saturated FAs.

Results: Cohort analyses show a positive association between the n-6:n-3 ratio and each biomarker. Total n-3 and n-6 PUFAs were associated with higher GlycA levels [mean difference = 0.33; 95% confidence interval (CI) = 0.29, 0.36, and 0.52; 95% CI = 0.48, 0.55, respectively]. The MR results suggest that total n-3 FAs cause higher circulating CRP (IVW = 0.09; 95% CI = 0.03, 0.16) and GlycA levels (0.12; 95% CI = 0.04, 0.21). The positive association between n-3 FAs and GlycA remained in the MVMR analysis after accounting for LDL cholesterol, triglycerides, monounsaturated FAs, and saturated FAs.

Conclusion: We find no convincing evidence of a simple pro- and anti-inflammatory dichotomy regarding the function of n-6 and n-3 PUFAs. Further research is needed to better understand the mechanisms underlying the effects of PUFAs on specific immune biomarkers.

Keywords: ALSPAC; CRP; ELOVL2; FADS; GlycA; IL-6; Mendelian randomization; biomarkers; inflammation; polyunsaturated fatty acids.

Figure 1.

Summary figure of results from the cohort, MR, and MVMR analyses. Green represents a positive association whereas red represents a negative association. Transparency indicates the strength of the association, where bold colours are used for smaller P-values and muted colours are used for larger P-values. For cohort analyses (ALSPAC and UKB)—M1: no adjustment; M2: adjustment for sex, age, maternal and paternal smoking patterns during pregnancy, and SEP; M3: adjustment for covariates from M2, triglycerides, and LDL cholesterol; M4: covariates included in M3, SFAs, and MUFAs. For MVMR analyses—M1: inclusion of both total n-3 and total n-6 as exposures; M2: inclusion of exposures from M1, triglycerides, and LDL cholesterol; M3: inclusion of exposures from M2, SFAs, and monounsaturated FAs.

Figure 2.

Association between fatty acids and inflammatory biomarkers using ALSPAC data at age 24 y. M1: no adjustment; M2: adjustment for sex, age, maternal and paternal smoking patterns during pregnancy, and SEP; M3: adjustment for covariates from M2, triglycerides, and LDL cholesterol; M4: adjustment for covariates included in M3 and SFAs and monounsaturated fatty acids.

Figure 3.

Association between fatty acids and biomarkers of inflammation using ALSPAC data at age 24 y stratified by sex at birth. Analyses presented were adjusted for household social class, maternal highest education qualification, maternal and paternal smoking status during pregnancy, and participant age and smoking and drinking status when attending the ALSPAC 24-y clinic.

Figure 4.

Multivariable MR analysis investigating the direct effect of n-3 and n-6 on three biomarkers of inflammation (CRP, IL-6, and GlycA) independently of each other by accounting for their mutual relationships.

Figure 5.

Multivariable MR analysis estimating the effects of n-3, n-6, triglycerides, and LDL cholesterol on three biomarkers of inflammation (CRP, IL-6, and GlycA) independently of each other by accounting for their mutual relationships. 95% CI, 95% confidence interval.

Figure 6.

Multivariable MR analysis estimating the direct effects of n-3, n-6, triglycerides, LDL cholesterol, SFAs, and monounsaturated fatty acids on CRP, IL-6, and GlycA independently of each other by accounting for their mutual relationships. 95% CI, 95% confidence interval.