AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Risk-Weighted Apolipoprotein B Highlights Residual Cardiovascular Risk from Lipoprotein(a) and Triglyceride-Rich Remnants Beyond SCORE2
By Yin Y, Zheng K, Yuan L et al.
Background: SCORE2 estimates lipid-related risk using total and high-density lipoprotein (HDL) cholesterol, primarily reflecting the apolipoprotein B (apoB)/low-density lipoprotein pathway, but does not fully capture residual atherogenic particle burden from lipoprotein(a) [Lp(a)] and triglyceride-rich remnants. Risk-weighted apoB (RW-apoB) is a recently proposed composite metric integrating apoB, Lp(a), and triglycerides using pre-specified weights based on their relative atherogenic potency in coronary heart disease. Whether this weighted residual lipid signal translates to the broader cardiovascular disease endpoint used in SCORE2 remains unclear. Methods: Among 281,974 UK Biobank participants aged 40-69 without baseline atherosclerotic cardiovascular disease or diabetes, we fitted four nested Fine-Gray competing-risk models for 10-year incident cardiovascular disease (fatal/non-fatal myocardial infarction, stroke, cardiovascular death): M0=SCORE2; M1=+apoB; M2=+RW-apoB; M3=+apoB+Lp(a)+triglycerides. Models were compared by change in area under the curve (ΔAUC) and categorical net reclassification improvement (catNRI), with bootstrap 95% confidence intervals (CIs). Results: Over a median follow-up of 10 years, 8,613 cardiovascular events occurred. Adding apoB alone did not improve SCORE2 (ΔAUC +0.0014; catNRI +0.0013, p=0.48). RW-apoB modestly improved discrimination and reclassification (ΔAUC +0.0079; catNRI +0.0110, 95%CI –0.001 to +0.022). Similar performance was observed when apoB, Lp(a), and triglycerides were entered separately (M3: ΔAUC +0.0078, 95%CI +0.002 to +0.017; catNRI +0.0141, 95%CI +0.002 to +0.027, p=0.02), with no difference between M2 and M3 (ΔAUC difference –0.0007, p=0.72; catNRI difference +0.003, p=0.72). In decomposition analyses, Lp(a) and triglycerides remained independently associated with cardiovascular risk, whereas apoB did not after SCORE2 adjustment. Reclassification improvement was greatest among participants at high SCORE2 risk and those with elevated Lp(a) (≥125 nmol/L). Calibration was preserved. Conclusions: RW-apoB is associated with residual cardiovascular risk beyond SCORE2 not captured by apoB alone, largely driven by lipoprotein(a) and triglyceride-rich remnants. It performs similarly to models using the three components separately, providing a simplified composite approach for risk assessment in primary prevention. The association appears stronger in individuals with elevated Lp(a) or higher baseline SCORE2 risk.