Epidemiology
Pooling MESA, CARDIA, JHS, FHS and ARIC over 21 years, top-decile Lp(a) raises ASCVD risk 46% overall and 92% in people with diabetes (J Am Coll Cardiol 2024)
Original title: Lipoprotein(a) and Long-Term Cardiovascular Risk in a Multi-Ethnic Pooled Prospective Cohort
This study pooled data from five major US prospective cohorts, MESA, CARDIA, JHS, FHS-Offspring, and ARIC, comprising 27,756 people aged 20-79 without prior ASCVD (55.0% women, 35.6% Black, 7.6% with diabetes), followed for a mean 21.1 years. Compared with Lp(a) below the 50th percentile, adjusted hazard ratios for ASCVD events rose progressively across percentile groups: 1.06 (50th-75th), 1.18 (75th-90th), and 1.46 (90th percentile or above, 95% CI 1.33-1.59). Elevated Lp(a) predicted ASCVD similarly across risk groups, sex, and race or ethnicity, but significantly more strongly in patients with diabetes than without (P for interaction = 0.0056), with hazard ratios of 1.92 (95% CI 1.50-2.45) versus 1.41 (95% CI 1.28-1.55) at the 90th percentile or above. Lp(a) also independently predicted myocardial infarction, revascularisation, stroke, and coronary heart disease death, though not total mortality. In this large, long-term US pooled cohort, higher Lp(a) raises ASCVD risk across populations, with a notably amplified effect in patients with diabetes.
Original abstract
Background: Lipoprotein(a) (Lp[a]) is a causal genetic risk factor for atherosclerotic cardiovascular disease (ASCVD). There are limited long-term follow-up data from large U.S. population cohorts.
Objectives: This study examined the relationship of Lp(a) with ASCVD outcomes in a large, pooled, multi-ethnic U.S.
Methods: The study included data on Lp(a) and ASCVD outcomes from 5 U.S.
Prospective Studies: MESA (Multi-Ethnic Study of Atherosclerosis), CARDIA (Coronary Artery Risk Development in Young Adults), JHS (Jackson Heart Study), FHS-OS (Framingham Heart Study-Offspring), and ARIC (Atherosclerosis Risk In Communities). Lp(a) levels were classified on the basis of cohort-specific percentiles. Multivariable Cox regression related Lp(a) with composite incident ASCVD events by risk group and diabetes status.
Results: The study included 27,756 persons without previous ASCVD who were aged 20 to 79 years, including 55.0% women, 35.6% Black participants, and 7.6% patients with diabetes, with mean follow-up of 21.1 years. Compared with Lp(a) levels <50th percentile, Lp(a) levels in the 50th to <75th, 75th to <90th, and ≥90th percentiles had adjusted HRs of 1.06 (95% CI: 0.99-1.14), 1.18 (95% CI: 1.09-1.28), and 1.46 (95% CI: 1.33-1.59), respectively for ASCVD events. Elevated Lp(a) predicted incident ASCVD events similarly by risk group, sex, and race or ethnic groups, but more strongly in patients with vs without diabetes (interaction P = 0.0056), with HRs for Lp(a) levels ≥90th percentile of 1.92 (95% CI: 1.50-2.45) and 1.41 (95% CI: 1.28-1.55), respectively. Lp(a) also individually predicted myocardial infarction, revascularization, stroke, and coronary heart disease death, but not total mortality.
Conclusions: The study shows, in a large U.S. pooled cohort, that higher Lp(a) levels are associated with an increased ASCVD risk, including in patients with diabetes.
Summary written by lp-a.org from the published abstract; figures as published. Page updated 18 August 2026. Methods.