High Triglycerides

2026-01-30 · A genome-wide association meta-analysis of cholesterol synthesis intermediates identifies three associations for lanosterol. EBioMedicine. 2026. DOI:10.1016/j.ebiom.2026.106144

Studying cholesterol's own precursor molecules finds a new gene, and shows its heart-disease effect runs entirely through cholesterol itself

Blood cholesterol's genetics are well studied, but the molecules the body builds cholesterol from -- synthesis intermediates like lanosterol and desmosterol -- had barely been examined genetically. This meta-analysis combined six European cohorts (up to 10,519 people) to study 14 traits related to lanosterol, desmosterol and cholesterol levels, finding 4 genome-wide significant associations not previously in the GWAS Catalog. Two were genuinely new loci with no prior cholesterol connection: one near CYP51A1 (7q21.2, associated with lanosterol) and one at 11q14.1 (associated with free cholesterol). Two more were new associations with lanosterol specifically at already-known cholesterol loci: HMGCR (5q13.3, the gene statins target directly) and the APO gene cluster (11q23.3). Using Mendelian randomization mediation analysis, the study found lanosterol has a significant total effect on coronary artery disease (CAD) risk, but once total cholesterol is accounted for, lanosterol's own direct effect disappears -- meaning lanosterol's apparent CAD risk is essentially fully explained by, and mediated through, cholesterol itself, not an independent pathway. This site's high triglycerides page carries 75 variants; CYP51A1 and this lanosterol-specific HMGCR finding are not currently among them.

Genes and variants:

• APOA5 rs662799 (also referenced by its historical promoter position as −1131T>C). A meta-analysis pooling data across studies (Xu et al., 2013, PLoS One; PMID 23468858) found the minor (C) allele frequency to be approximately 7.7% in European-ancestry populations and 28.6% in East Asian-ancestry populations (24,563 East Asian and 11,583 European individuals analyzed). C-allele carriers had mean triglyceride levels 0.30 mmol/L higher (95% CI, 0.26–0.33) than TT homozygotes, along with smaller shifts in total cholesterol (+0.08 mmol/L), LDL-cholesterol (+0.04 mmol/L), and HDL-cholesterol (−0.05 mmol/L). The pooled odds ratio for metabolic syndrome among C-allele carriers was 1.33 (95% CI, 1.16–1.53), driven mainly by the East Asian subgroup (OR 1.43) with a non-significant estimate in the European subgroup (OR 1.30, 95% CI 0.94–1.78).
• rs964184, an intergenic variant near the BUD13/ZNF259/APOA5-APOA4-APOC3-APOA1 gene cluster on chromosome 11q23. Fine-mapping and admixture-mapping analyses (Parra et al., 2017, PLoS One; PMID 28245265) identified rs964184 as the principal driver of the triglyceride association signal in this region. A separate cohort study of patients with clinically manifest vascular disease (van de Woestijne et al., 2014, PLoS One; PMID 24979386) confirmed the association between rs964184 and elevated plasma triglycerides but found no independent association with incident vascular events in that cohort. rs964184 has also been reported as showing a smaller, secondary association with HDL-cholesterol in some genome-wide association studies, consistent with the well-documented inverse epidemiological relationship between triglycerides and HDL-cholesterol; triglycerides is the trait most consistently and strongly linked to this variant across replication studies.

Clinical thresholds: Fasting triglyceride categories in common clinical use: normal <150 mg/dL, borderline high 150–199 mg/dL, high 200–499 mg/dL, very high ≥500 mg/dL. The Endocrine Society's 2012 clinical practice guideline (Berglund et al., 2012, J Clin Endocrinol Metab; PMID 22962670) further subdivides severe hypertriglyceridemia into a "severe" category (1,000–1,999 mg/dL) and a "very severe" category (≥2,000 mg/dL), and identifies levels above approximately 1,000 mg/dL as the range in which pancreatitis risk becomes clinically significant, while noting the specific cutoff is not a sharp biological boundary.

Interpreting these variants: Common variants such as rs662799 and rs964184 are appropriately understood as contributing to the polygenic architecture of triglyceride levels — each has a small, population-level average effect and neither is diagnostic for an individual. They are distinct from rare, high-penetrance APOA5 loss-of-function variants or other monogenic causes of severe hypertriglyceridemia (e.g., familial chylomicronemia syndrome, which involves genes such as LPL, APOC2, APOA5, GPIHBP1, or LMF1 and produces much larger elevations). This entry concerns the common, population-level variants only, not rare monogenic disease.

No treatment or dietary recommendations are implied by carrying either variant; triglyceride management is guided by an individual's actual measured lipid panel, overall cardiovascular risk profile, and clinical evaluation, not by genotype at these loci.