Mechanisms
A national database of 87,379 patients confirms no link between Lp(a) and retinal vein occlusion, validating a small-cohort null finding (Can J Ophthalmol 2024)
Original title: Lack of association between Lp(a) and retinal vein occlusion in a single institution and US national database
This two-phase retrospective study examined whether Lp(a) predicts retinal vein occlusion (RVO). In the first phase, 45 patients with RVO and a laboratory Lp(a) value at a single tertiary centre showed no significant associations between Lp(a) status and age at RVO diagnosis, time to RVO development, visual acuity, or central subfield thickness (p > 0.05 for all). In the second phase, the TriNetX US national database identified 35,687 patients with high Lp(a) (above 30 mg/dL or 61 nmol/L) and 51,692 with low Lp(a); elevated Lp(a) was not associated with higher odds of central RVO (odds ratio 1.15, 95% CI 0.88-1.50) or branch RVO (odds ratio 1.01, 95% CI 0.76-1.36). The large-database analysis confirms the small-cohort finding, demonstrating no meaningful association between Lp(a) and RVO risk, and illustrating how national databases can validate or refute laboratory-value associations first identified in small observational cohorts.
Original abstract
Objective: This study examines associations between lipoprotein(a) (Lp[a]), a low-density-like lipoprotein, and renal vein occlusion (RVO) in US cohorts to characterize its prognostic role in the setting of RVO.
Design: A two-phase retrospective cohort study.
Methods: In the first phase, patients with RVO and a Lp(a) quantitative laboratory value at a single tertiary centre were reviewed. Lp(a) status was assessed in association with age of RVO diagnosis, visual acuity, time to development of RVO, and central subfield thickness. In the second phase, the TriNetX US Collaborative Network, a large national database, also was queried for the presence of high or low Lp(a) values and diagnoses of RVO.
Results: The single tertiary care centre identified 45 patients with RVO and a laboratory value of Lp(a), finding no significant associations with respect to Lp(a) status and age of RVO onset, time from the laboratory draw to the development of RVO, visual acuity, and central subfield thickness (p > 0.05 for all). The TriNetX national database identified 35,687 patients with a high Lp(a) value (>30 mg/dL or 61 nmol/L) and 51,692 with a low Lp(a) value. An elevated Lp(a) value was not associated with higher odds of central (odds ratio [OR] = 1.15; 95% CI, 0.88-1.50) or branch RVO (OR = 1.01; 95% CI, 0.76-1.36).
Conclusion: Taken together, this analysis suggests a lack of association between Lp(a) value and risk of RVO. This study highlights the benefit of large national databases in the validation of laboratory value predictors identified through small-cohort observational studies.
Summary written by lp-a.org from the published abstract; figures as published. Page updated 18 August 2026. Methods.