Prospective directional associations between lipid measures and digital volume pulse-derived arterial stiffness parameters: evidence from cross-lagged panel analyses in the UK Biobank

Background Cardiovascular disease (CVD) is the leading cause of morbidity and mortality worldwide. Arterial stiffness and dyslipidemia are considered risk factors for CVD. However, no relevant prospective inference studies have been reported in the middle-aged and elderly populations. This study aimed to investigate the prospective directional associations between dyslipidemia and arterial stiffness. Methods The study included 20,344 participants from the UK Biobank with available first follow-up data. After applying the inclusion and exclusion criteria, 2,989 participants remained, with a median follow-up interval of 3.0 years (interquartile range, 2.8–3.2). Prospective directional associations between arterial stiffness and dyslipidemia were confirmed by the classical cross-lagged panel model and three improved nonlinear cross-lagged panel models. Results The baseline age of the participants was 58.7 ± 7.2 years, of which 51.7% were male. In males, cross-lagged analyses revealed a prospective directional association from apolipoprotein B (ApoB) to pulse wave reflection index (PWRI; False discovery rate-corrected, FDR p = 0.038). In females, triglyceride (TG) levels showed prospective directional associations with subsequent digital volume pulse-derived arterial stiffness index, PWRI, and pulse wave peak-to-peak time, as well as pulse rate (FDR p < 0.05). Sex differences were observed in the associations of ApoB and TG with subsequent PWRI ( p < 0.05). However, after incorporating SBP as a third endogenous variable, these associations were no longer statistically significant (FDR p > 0.05). Conclusion The results of this study suggest that dyslipidemia is prospectively associated with arterial stiffness, with predominantly directional associations from lipid measures to arterial stiffness parameters. However, the associations were no longer statistically significant in the sensitivity analysis incorporating SBP, suggesting that the observed associations may be sensitive to SBP.