Abstract
Whether circulating fatty acids (FAs) play a causal role in the development of cardiovascular disease (CVD) remains unclear. We conducted a Mendelian randomisation study to explore the associations between plasma phospholipid FA levels and 15 CVDs. Summary-level data from the CARDIoGRAMplusC4D, MEGASTROKE, and Atrial Fibrillation consortia and UK Biobank were used. Sixteen single-nucleotide polymorphisms (SNPs) associated with ten plasma FAs were used as instrumental variables. SNPs in or close to the FADS1 gene were associated with most FAs. We performed a secondary analysis of the association between a functional variant (rs174547) in FADS1, which encodes ?5-desaturase (a key enzyme in the endogenous FA synthesis), and CVD. Genetic predisposition to higher plasma α-linolenic, linoleic, and oleic acid levels was associated with lower odds of large-artery stroke and venous thromboembolism, whereas higher arachidonic and stearic acid levels were associated with higher odds of these two CVDs. The associations were driven by SNPs in or close to FADS1. In the secondary analysis, the minor allele of rs174547 in FADS1 was associated with significantly lower odds of any ischemic stroke, large-artery stroke, and venous thromboembolism and showed suggestive evidence of inverse association with coronary artery disease, abdominal aortic aneurysm and aortic valve stenosis. Genetically higher plasma α-linolenic, linoleic, and oleic acid levels are inversely associated with large-artery stroke and venous thromboembolism, whereas arachidonic and stearic acid levels are positively associated with these CVDs. The associations were driven by FADS1, which was also associated with other CVDs.
Link to Full Text
Key Points
Fatty acid desaturase 1 (FADS1) gene encodes ?5-desaturase, which is the key rate-limiting enzyme in the endogenous synthesis of long-chain PUFAs. ?5-desaturase produces arachidonic acid (AA) and eicosapentaenoic acid (EPA) from dihomo-γ-linolenic acid and eicosatetraenoic acid, respectively. Data on the associations of ?5-desaturase activity or FADS1 variants with CVDs are limited. Mendelian randomisation (MR) is a method that strengthens causal inference between risk factors and outcomes by exploiting genetic variants as instrumental variables of an exposure. This technique minimizes confounding as genetic variants are randomly assorted at meiosis, thereby having no association with self-selected lifestyle factors and behaviors. In addition, it overcomes reverse causality since allelic randomisation antedates the disease’s onset.
Here a two-sample MR study was conducted to explore the associations of plasma phospholipid FA levels with risk of 15 CVDs. Since ?5-desaturase plays a major role in FA synthesis, we performed a secondary analysis to investigate the associations of a functional variant (rs174547) within the FADS1 gene, as a proxy for ?5-desaturase activity, with CVD. The main finding of this MR study is that genetically higher plasma levels of ALA, LA, and OA were associated with lower odds of large-artery stroke and venous thromboembolism, whereas higher plasma AA and SA levels were associated with higher odds of these two CVDs. However, these and other suggestive associations of plasma levels of several FAs with CVD were driven by FADS1, encoding the ?5-desaturase enzyme, which showed strong or suggestive associations with six out of 15 CVDs. This study showed that genetic predisposition to higher plasma ALA, LA, and OA levels was associated with lower odds of large-vessel stroke and venous thromboembolism, whereas plasma AA and SA levels were positively associated with these CVDs. However, the associations were driven by FADS1, which was also significantly or suggestively associated with coronary artery disease, abdominal aortic aneurysm and aortic valve stenosis. Thus, it is recommended to reconsider the role of individual FAs in the prevention of different CVDs.