Event-anchored piecewise trajectories of the triglyceride-glucose index before and after incident cardiovascular disease: a propensity score-matched longitudinal study from the English Longitudinal Study of Ageing

The triglyceride-glucose (TyG) index is an established surrogate marker of insulin resistance associated with cardiovascular disease (CVD) risk. However, existing studies have primarily examined TyG as a pre-event predictor, and the dynamic metabolic trajectory surrounding incident CVD—including how TyG evolves in the years leading up to, at the time of, and following first-reported CVD—remains poorly characterized. This study aimed to delineate the temporal trajectory of the TyG index anchored to the wave of first-reported CVD and to evaluate whether pre-event metabolic divergence constitutes a retrospectively detectable metabolic signal. This retrospective longitudinal cohort study utilized data from Wave 2 (2004) through Wave 9 (2018) of the English Longitudinal Study of Ageing. Participants aged ≥ 50 years free of CVD at baseline with complete fasting biomarker data were included. Incident CVD cases were matched 1:3 to non-CVD controls via propensity score matching. Piecewise quadratic mixed-effects models with event-anchored time scales were fitted to characterize pre- and post-event TyG trajectories. Pre-event warning signal analyses, subgroup analyses by sex, hypertension status, and age, and nine sensitivity analyses were conducted. Among 2327 eligible participants, 470 incident CVD cases were matched to 1306 controls (analytical cohort n = 1776). The CVD group exhibited a significantly steeper pre-event TyG rise compared with controls (β = 0.0505, P < 0.001), an abrupt decline at the first-reported CVD wave (β = −0.1769, P < 0.001), and an accelerated post-event decline (β = −0.0666, P < 0.001), whereas the control group trajectory remained flat. Between-group TyG divergence became statistically significant from 8 years before the first-reported CVD wave, widening progressively through the event. Subgroup analyses confirmed consistent trajectory patterns across sex, hypertension status, and age strata. Excluding users of lipid- or glucose-lowering medications attenuated the post-event decline, suggesting a potential pharmacological contribution to the observed post-event metabolic shift, although causal attribution was not possible. All nine sensitivity analyses yielded directionally consistent results. The TyG index follows a distinctive asymmetric trajectory around incident CVD, characterized by progressive pre-event elevation detectable up to 8 years before clinical onset, followed by abrupt and sustained post-event decline potentially related to pharmacological intervention. These findings suggest that longitudinal TyG monitoring may complement cardiovascular risk stratification in aging populations, pending prospective validation.