Lipid markers and age differences in atherosclerotic cardiovascular disease risk

Background Atherogenic lipoproteins, including LDL-c and ApoB, are well-established risk factors for atherosclerotic cardiovascular disease (ASCVD). However, whether different lipid markers demonstrate distinct patterns across ASCVD subtypes and age groups remains incompletely defined. Methods We used data from the UK Biobank to establish a cohort of participants free of cardiovascular disease and not using cholesterol-lowering medications at baseline. We examined associations of 10 lipid markers (lipoprotein(a), ApoB, ApoA, ApoB/ApoA ratio, total cholesterol, remnant cholesterol, non-HDL-c, HDL-c, LDL-c, and triglycerides) with incident ASCVD. Metabolomic markers were also evaluated to provide complementary insight into lipid metabolism of ASCVD. To evaluate whether age modifies lipid-associated ASCVD risk, we fitted multivariable Cox proportional hazards models including interaction terms between each lipid marker and baseline age. We quantified age modification by deriving age-specific hazard ratios and calculating ratios of hazard ratios (RHR) and 95% CI comparing younger and older individuals. We then used restricted cubic splines with three knots to characterize the dose-response relationships across age strata. Findings Among 401,228 individuals included in the study (232,152 (57.9%) women, mean (SD) age at baseline 55.4 (8.1) years), 31,539 developed incident ASCVD over a median (IQR) of 13.2 [12.3, 13.9] years. Across the 10 lipid markers investigated, all atherogenic lipids showed consistent and significant age-related heterogeneity. RHR were about two-fold stronger in younger (<50 years) compared with older individuals (≥65 years), particularly ApoB (RHR 2.12 [95% CI 1.81-2.48]), non-HDL-c (2.16 [1.84-2.55]), and LDL-c (2.20 [1.86-2.59]). In contrast, atheroprotective markers (ApoA and HDL-c) showed no evidence of age modification, and lipoprotein(a) stood out with comparatively modest heterogeneity by age (Figure 1). Within age strata, associations demonstrated graded dose-response relationships, without clear threshold effects. For all atherogenic markers except lipoprotein(a), age-specific curves demonstrated steeper gradients in younger individuals, progressive attenuation across older age groups, and often converged at the highest lipid concentrations. For lipoprotein(a), the shape and strength of the associations were similar across age groups (Figure 2). Additional metabolomic measures showed similar age-related attenuation of risk, consistent with the patterns observed for lipid markers. Interpretation Across atherogenic lipid markers, dose-response gradients of association with ASCD were steeper in younger individuals and progressively attenuated with advancing age, except for lipoprotein(a). These findings could support more targeted screening and prevention strategies in early adulthood, currently predominantly concentrated in later life when lipid-associated risk gradients appear attenuated.