A phenome-wide association study of genetically mimicked statins

Abstract

Background Beyond their success in cardiovascular disease prevention, statins are increasingly recognized to have sex-specific pleiotropic effects. To gain additional insight, we characterized associations of genetically mimicked statins across the phenotype sex-specifically. We also assessed whether any apparently non-lipid effects identified extended to genetically mimicking other widely used lipid modifiers (proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors and ezetimibe) or were a consequence of low-density lipoprotein cholesterol (LDL-c). Methods We performed a sex-specific phenome-wide association study assessing the association of genetic variants in HMGCR, mimicking statins, with 1701 phenotypes. We used Mendelian randomization (MR) to assess if any non-lipid effects found were evident for genetically mimicked PCSK9 inhibitors and ezetimibe or for LDL-c. Results As expected, genetically mimicking statins was inversely associated with LDL-c, apolipoprotein B (ApoB), and total cholesterol (TC) and positively associated with glycated hemoglobin (HbA1c) and was related to body composition. Genetically mimicking statins was also inversely associated with serum calcium, sex hormone-binding globulin (SHBG), and platelet count and positively associated with basal metabolic rate (BMR) and mean platelet volume. Stronger associations with genetically mimicked statins were evident for women than men for lipid traits (LDL-c, ApoB, and TC), calcium, and SHBG, but not for platelet attributes, body composition, or BMR. Genetically mimicking PCSK9 inhibitors or ezetimibe was also associated with lower lipids, but was not related to calcium, SHBG, BMR, or body composition. Genetically higher LDL-c increased lipids and decreased BMR, but did not affect calcium, HbA1c, platelet attributes, or SHBG with minor effects on body composition. Conclusions Similar inverse associations were found for genetically mimicking statins on lipid traits in men and women as for other lipid modifiers. Besides the positive associations with HbA1c, BMI (which may explain the higher BMR), and aspects of body composition in men and women, genetically mimicking statins was additionally associated with platelet attributes in both sexes and was inversely associated with serum calcium and SHBG in women. This genetic evidence suggests potential pathways that contribute to the effects of statins particularly in women. Further investigation is needed to confirm these findings and their implications for clinical practice.

Investigating the potential anti-depressive mechanisms of statins: a transcriptomic and Mendelian randomization analysis

“…Interestingly, despite extensive validation of the effects of HMGCR inhibition on cardiometabolic traits by previous studies [44,45], we observed HEIDI significance for the association of HMGCR expression in blood with CAD and LDL-C (Fig. 3), which suggested that the variant associated with HMGCR expression was distinct from those associated with CAD risk and LDL-C levels.…”

Statins and antidepressants induce similar in vitro gene expression responses

“…Interestingly, we observed HEIDI significance for the association of HMGCR expression in blood with CAD and LDL-C (Figure 4), which suggested that the variant associated with HMGCR expression was distinct from those associated with CAD risk and LDL-C level, despite extensive validation of the effects of HMGCR inhibition on these traits by previous studies [43, 45]. The HEIDI significance may be due to the complex LD structures in the GWAS studies, which may not be accurately accounted for by the LD reference (Supplementary Figure 9).…”

“…We observed no association between HMGCR expression and inflammatory cytokine levels, namely IL6 and CRP. Interestingly, we observed HEIDI significance for the association of HMGCR expression in blood with CAD and LDL-C (Figure 2), which suggested that the variant associated with HMGCR expression was distinct from those associated with CAD risk and LDL-C level, despite extensive validation of the effects of HMGCR inhibition on these traits by previous studies (18,19). This is likely due to the complex LD structures in the GWAS studies (Supplementary Figure 11).…”

“…To investigate the on-target effect of statins, we selected rs12916, which is widely used as an instrument for HMGCR inhibition [43-45]. The genetic associations between rs12916-T allele and metabolomic traits are shown to mirror the actual effects of statin exposure on the metabolome [46].…”

Cross-sectional association between blood cholesterol and calcium levels in genetically diverse strains of mice

“…Two interventional studies using statins have showed a reductions in calcium levels [45,46], whereas another single-arm study showed declines that did not reach significance [47]. A Mendelian Randomization approach using LDL-C as the instrument was also associated with elevated calcium levels [48], further supporting a potential causal relationship with cholesterol driving calcium levels. As this is a cross-sectional associative relationship we are not able at present to define directionality, much less the underlying biological mechanism(s).…”