Genetics
Low Lp(a) tracking with worse liver disease in cross-sectional studies is reverse causality, UK Biobank and genetic analysis of 270,004 people reveals (Clin Gastroenterol Hepatol 2025)
Original title: Lipoprotein (a) in the Full Spectrum of Metabolic Dysfunction-associated Steatotic Liver Disease: Evidence From Histologically and Genetically Characterized Cohorts
This study combined a multicentre liver-biopsy cohort (n = 332) with a UK Biobank prospective analysis (n = 270,004, median follow-up 12.47 years) and an LPA genetic risk score analysis to clarify Lp(a)'s true relationship with metabolic dysfunction-associated steatotic liver disease (MASLD) across its full severity spectrum. Cross-sectionally, Lp(a) was inversely associated with at-risk steatohepatitis (OR per 1-SD increase 0.64, 95% CI 0.42-0.97), suggesting lower Lp(a) tracks with worse liver disease. But prospectively, the relationship reversed: participants with Lp(a) below 10.75 nmol/L had higher risk of cirrhosis (HR 1.49, 95% CI 1.22-1.81) and hepatocellular carcinoma (HR 1.69, 95% CI 1.12-2.56) than those with Lp(a) in the 10.75-21.5 nmol/L range, with risk rising further above that range, and a similar L-shaped pattern was seen for genetically predicted Lp(a). The authors conclude the cross-sectional inverse association reflects reverse causality (liver disease lowering Lp(a)), while genetically low Lp(a) itself predicts worse liver outcomes, and recommend that Lp(a)-lowering drug trials monitor closely for adverse liver effects in patients who reach extremely low Lp(a) concentrations.
Original abstract
Background & Aims: Lipoprotein(a) (Lp(a)) is an emerging biomarker for cardiometabolic factors. We studied the role of Lp(a) in the full spectrum of metabolic dysfunction-associated steatotic liver disease (MASLD).
Methods: Three independent analyses were implemented using a multi-center, cross-sectional, liver biopsy-based study (n = 332) (Study 1) and the UK Biobank prospective study (n = 270,004) (Study 2; median follow-up, 12.47 years). In Study 1, we studied the cross-sectional association between Lp(a) mass and MASLD stages (Analysis A). In Study 2, we studied the prospective association between Lp(a) concentration and MASLD, liver cirrhosis, and hepatocellular carcinoma (Analysis B). Finally, these analyses were accompanied by a prospective analysis using LPA Genetic Risk Score (Analysis C).
Results: In Study 1, an inverse association between Lp(a) and at-risk metabolic dysfunction-associated steatohepatitis (odds ratioper 1-SD increase, 0.64; 95% confidence interval [CI], 0.42-0.97) was observed. In contrast, when similar associations were examined prospectively (Study 2), subjects with Lp(a) <10.75 nmol/L had a higher risk for cirrhosis (hazard ratio, 1.49; 95% CI, 1.22-1.81) and HCC (hazard ratio, 1.69; 95% CI, 1.12-2.56) compared with subjects with Lp(a) within the 10.75 to 21.5 nmol/L range. Above these levels, the risk increased significantly and positively. Similarly, genetic analysis showed an L-shaped association with LPA Genetic Risk Score.
Conclusions: The inverse association observed in cross-sectional studies should be attributed to reverse causality (ie, the presence of liver disease may decrease Lp(a) levels). Genetically predicted very low Lp(a) levels are also associated with impaired liver health prospectively. Clinical trials evaluating Lp(a)-lowering agents should thus be monitored carefully for adverse liver effects in subjects attaining extremely low concentrations.
Summary written by lp-a.org from the published abstract; figures as published. Page updated 17 August 2026. Methods.