SUMMARY OF KEY RESEARCH FINDINGS
This study, presented at the European Society of Cardiology (ESC) Congress 2026, asked whether the long-term pattern of insulin resistance identifies cardiovascular risk that persists once low-density lipoprotein cholesterol (LDL-C) is controlled.1 In 4,051 adults from the population-based Tehran Lipid and Glucose Study, with at least three serial values of the homeostatic model assessment of insulin resistance (HOMA-IR) and free of cardiovascular disease (CVD), group-based trajectory modelling identified three patterns: low stable, moderate stable, and rising. Over a mean of 10.1 years after the index visit, 280 participants developed CVD. The rising pattern was associated with incident CVD both when LDL-C at index was controlled (odds ratio [OR]: 1.35; 95% CI: 1.07–1.77) and when it was not (OR: 1.52; 95% CI: 1.07–2.15), with no evidence of interaction by LDL-C status and little attenuation once LDL-C category and statin use were treated as time-varying covariates. The authors concluded that a deteriorating trajectory of insulin resistance marks residual cardiovascular risk that LDL-focused prevention does not capture.
WHAT CHALLENGE DOES THIS ADDRESS?
Residual cardiovascular risk despite guideline-level LDL-C is well recognised, and insulin resistance is a strong candidate contributor: it is associated with incident CVD independently of conventional risk factors,2 and it travels with atherogenic dyslipidaemia, hypertension, and dysglycaemia. Most supporting evidence, however, rests on insulin resistance measured once. A single value cannot separate a person whose HOMA-IR has been stable for a decade from one in whom it is climbing, and these two people plausibly do not face the same future. Trajectory analyses have begun to close that gap, linking a rising HOMA-IR pattern to incident CVD in a Korean cohort,3 but whether such a signal survives within people whose LDL-C already meets targets, with lipid control treated as time-updated rather than frozen at baseline, remained open. By fixing the trajectory period before the index visit and counting events afterwards, this study put that question on a defensible temporal footing.
RELEVANCE TO EUROPEAN PRACTICE
European prevention is organised around estimation and lipid control: the Systematic Coronary Risk Evaluation 2 (SCORE2) calculator anchors the 2021 ESC prevention guidelines, and neither the algorithm nor the downstream treatment pathway looks at insulin resistance.4,5 A patient whose LDL-C meets the 2019 European dyslipidaemia goals can therefore appear fully managed while their metabolic state drifts year on year.6 In this cohort, that reassurance looked incomplete: risk graded upwards across joint phenotypes, peaking at a crude OR of 5.05 (95% CI: 3.10–8.20) for a rising trajectory with uncontrolled LDL-C against low stable with controlled LDL-C.1
Transportability deserves a direct answer rather than a hopeful one. The Tehran Lipid and Glucose Study is a population-based West Asian cohort studied through a period of nutrition transition,7 with a substantial background burden of central obesity and dysglycaemia.8 Lipid-lowering practices across 1999–2011 predate current European treatment intensity, so controlled LDL-C here often reflects a naturally favourable profile rather than treatment to target, a phenotype distinct from the statin-treated European patient. Absolute risks will not transfer; SCORE2’s regional recalibration is a reminder that they rarely do.5 More likely to travel is the relative pattern: an association consistent across strata, adjustment models, and time-varying lipid control. European cohorts holding serial insulin measurements are the natural place to test it.
WHAT ARE THE NEXT STEPS FOR THE RESEARCH?
Replication comes first, ideally in European cohorts with repeated fasting insulin and time-to-event modelling in place of the ORs reported here. The observational design cannot exclude residual confounding, and no interventional evidence yet shows that flattening a rising trajectory prevents events; insulin resistance is modifiable, but ‘modifiable’ and ‘worth modifying for prevention’ are different claims. HOMA-IR itself is an imperfect instrument, weighted towards fasting hepatic physiology rather than whole-body insulin sensitivity.9
Group-based trajectory modelling imposes discrete classes on continuous behaviour, and class membership carries assignment uncertainty. The definition of control, LDL-C below 100 mg/dL (2.6 mmol/L) at index, is more permissive than current ESC goals for patients at higher risk, so the residual-risk claim should be re-examined against stricter thresholds. Finally, serial insulin measurement is unrealistic in routine practice; simpler longitudinal surrogates deserve evaluation before trajectory-aware prevention can become more than an idea.
