Familial hypercholesterolaemia does not cause elevated Lp(a); instead, high Lp(a) makes FH more likely to be diagnosed, an analysis of 37,877 people (J Am Coll Cardiol 2020)

Original title: Ascertainment Bias in the Association Between Elevated Lipoprotein(a) and Familial Hypercholesterolemia

J Am Coll Cardiol · 2 June 2020 · 8

Trinder M, DeCastro ML, Azizi H, Cermakova L, Jackson LM, Frohlich J, Mancini GBJ, Francis GA, Brunham LR

Comparing 391 clinically diagnosed familial hypercholesterolaemia (FH) patients with 37,486 UK Biobank participants genetically assessed for FH, the authors found FH cohort patients had significantly higher Lp(a) than the general population, explained by a higher frequency of the rs10455872-G LPA risk allele (15.1% vs 8.8%, P<0.05). However, within the FH cohort, Lp(a) did not differ between those with and without an FH-causing genetic variant (1.43 vs 1.42 log mg/dL, P=0.97), and the same held in the UK Biobank cohort (221 with FH variant vs 37,486 without). A phenome-wide association study across 19,202 phenotypes showed elevated Lp(a) is linked to higher LDL cholesterol, family history of cardiovascular disease, premature coronary artery disease, and clinical FH diagnosis. The findings suggest FH itself does not raise Lp(a), but elevated Lp(a) makes individuals with genetic FH more likely to be clinically recognised, a form of ascertainment bias.

Original abstract

Background: Lipoprotein(a) is an atherogenic low-density lipoprotein-like particle and circulating levels are largely determined by genetics. Patients with familial hypercholesterolemia (FH) have elevated lipoprotein(a); however, it remains unclear why.

Objectives: This study compared the levels of lipoprotein(a) and associated genetic factors between individuals that were ascertained for FH clinically versus genetically.

Methods: We investigated causes of elevated lipoprotein(a) in individuals with clinically diagnosed FH (FH cohort, n = 391) and in individuals with genetically diagnosed FH from the general population (UK Biobank; n = 37,486).

Results: Patients in the FH cohort had significantly greater lipoprotein(a) levels than either the general population or non-FH dyslipidemic patients. This was accounted for by increased frequency of the rs10455872-G LPA risk allele (15.1% vs. 8.8%; p < 0.05). However, within the FH cohort, lipoprotein(a) levels did not differ based on the presence or absence of an FH-causing variant (means = 1.43 log mg/dl vs. 1.42 log mg/dl; p = 0.97). Lipoprotein(a) levels were also not statistically different between individuals with and without an FH-causing variant in the UK Biobank cohort, which represents a population sample not biased to cardiovascular ascertainment (n = 221 vs. 37,486). We performed a phenome-wide association study between LPA genotypes and 19,202 phenotypes to demonstrate that elevated lipoprotein(a) is associated with increased low-density lipoprotein cholesterol, a family history of cardiovascular disease, premature coronary artery disease, and a diagnosis of FH.

Conclusions: These results suggest that FH does not cause elevated lipoprotein(a), but that elevated lipoprotein(a) increases the likelihood that an individual with genetic FH will be clinically recognized.

Summary written by lp-a.org from the published abstract; figures as published. Page updated 18 August 2026. Methods.