Abstract
Background The multitrajectory model can identify joint longitudinal patterns of different lipids simultaneously, which might help better understand the heterogeneous risk of premature cardiovascular disease (CVD) and facilitate targeted prevention programs. This study aimed to investigate the associations between multitrajectories of lipids with premature CVD. Methods and Results The study enrolled 78 526 participants from the Kailuan study, a prospective cohort study in Tangshan, China. Five distinct multitrajectories of triglyceride, low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol over 6-year exposure were identified on the basis of Nagin's criteria, using group-based multitrajectory modeling. During a median follow-up of 6.75 years (507 645.94 person-years), 665 (0.85%) premature CVDs occurred. After adjustment for confounders, the highest risk of premature CVD was observed in group 4 (the highest and increasing triglyceride, optimal and decreasing LDL-C, low and decreasing high-density lipoprotein cholesterol) (hazard ratio [HR], 2.13 [95% CI, 1.36-3.32]), followed by group 5 (high and decreasing triglyceride, optimal and increasing LDL-C, low and decreasing high-density lipoprotein cholesterol) (HR, 2.07 [95% CI, 1.45-2.98]), and group 3 (optimal and increasing triglyceride, borderline high and increasing LDL-C, optimal and decreasing high-density lipoprotein cholesterol) (HR, 1.90 [95% CI, 1.32-2.73]). Conclusions Our results showed that the residual risk of premature CVD caused by increasing triglyceride levels remained high despite the fact that LDL-C levels were optimal or declining over time. These findings emphasized the importance of assessing the joint longitudinal patterns of lipids and undertaking potential interventions on triglyceride lowering to reduce the residual risk of premature CVD, even among individuals with optimal LDL-C levels.