Lp(a) mortality risk varies by race and ethnicity in NHANES III, 22.6-year mean follow-up (Am J Cardiol 2026)

Lipoprotein(a) [Lp(a)] is a genetically determined and likely causal independent risk factor for cardiovascular (CV) outcomes and mortality, with levels >50 mg/dL considered risk-enhancing. Over 90% of variation in levels is genetically determined, with levels varying by race/ethnicity. Evidence on whether Lp(a) risk thresholds vary by race/ethnicity, and remains inconsistent. This study examines whether the association between Lp(a) and mortality differs by race/ethnicity. We analyzed survey-weighted data from a nationally representative muti-ethnic cohort of US adults from NHANES III with mortality follow-up through 2019. Participants were stratified into non-Hispanic White, non-Hispanic Black, or Mexican-American. Associations between Lp(a) and mortality outcomes were estimated using multivariable Cox and Fine-Gray competing risk models. Lp(a) were analyzed as continuous variables, logarithmically transformed, and divided into three groups (<50, 50 to 75, and >75 mg/dL). A total of 50,519,751 survey-weighted records were included. Mean follow-up was 22.6 years. Median Lp(a) concentrations were higher among non-Hispanic Black participants (36 mg/dL, IQR 22 to 66) than non-Hispanic White (12 mg/dL, IQR 3 to 30) and Mexican-American (8 mg/dL, IQR 2 to 22) participants. Mexican American participants with Lp(a) >75 mg/dL had a higher risk of CV mortality that persisted after multivariable adjustment (sHR 2.93, 95% confidence intervals 1.01 to 8.56, p value 0.049). Among non-Hispanic Black participants, higher Lp(a) was linked to all-cause and CV mortality in unadjusted models but not after adjustment. No significant association was detected in non-Hispanic White participants. In conclusion, Lp(a) distributions and their relationship with clinical outcomes vary by race/ethnicity. Our findings suggest that prognostic thresholds for Lp(a) may differ, supporting the need to define and validate race/ethnicity-specific cut-offs that best predict CV outcomes and improve risk stratification.