Predicting subclinical atherosclerosis in adults with type 1 diabetes using cardiovascular risk models: a cross‑sectional study from Italy.

Abstract

Background: Cardiovascular disease (CVD) is the leading cause of morbidity and mortality among individuals with type 1 diabetes (T1D); nevertheless, cardiovascular risk stratification remains an unmet need in T1D. Cardiovascular risk tools were developed to predict 10‑year clinical events. Whether they identify subclinical atherosclerosis - a potential gatekeeper for intensified prevention - remains uncertain.

Methods: We conducted a cross-sectional analysis of consecutive T1D adults without a history of CVD enrolled at a secondary centre in Italy. All participants underwent coronary computed tomography angiography (CCTA) and carotid ultrasound. The primary outcome was a composite imaging endpoint, defined as any coronary plaque on CCTA and/or maximum carotid intima-media thickness (cIMT_max) ≥ 1.5 mm. We compared the performance of T1D-specific CV risk models (2019 European Society of Cardiology [ESC] Guidelines, Steno Type 1 Risk Engine [ST1RE], and Scottish-Swedish CVD risk prediction tool) in predicting the presence of subclinical atherosclerosis. Discrimination was assessed using the Area Under the receiver operating characteristic Curve (AUC), and differences between AUCs were evaluated using the DeLong test; calibration was visualised with calibration plots and summarised by Brier score. Sensitivity analyses evaluated CCTA‑only and carotid‑only endpoints.

Results: One hundred and one patients (52.5% males) were included in the study, with a mean age of 45 ± 13 years and a median diabetes duration of 21 years (Q1-Q3 13; 32). The composite imaging endpoint occurred in 66/101 (65.4%); 55 patients (54.5%) had coronary plaques, and 41 patients (40.6%) presented cIMT_max above threshold. Both ST1RE and the Scottish-Swedish risk scores showed strong discrimination for the combined imaging endpoint (AUC 0.899 and 0.907, respectively). Apparent calibration to the imaging endpoint was acceptable, with Brier scores of 0.126 for ST1RE and 0.128 for the Scottish-Swedish model.

Conclusions: In an Italian cohort of asymptomatic T1D adults, T1D-specific risk scores demonstrated a strong discrimination ability for imaging-defined subclinical atherosclerosis. They might help triage asymptomatic individuals who could benefit from targeted cardiovascular imaging in specific settings. Since these tools predict events rather than imaging findings, recalibration and prospective validation against imaging endpoints are needed before using predicted probabilities as absolute risks informing clinical decision‑making.

Trial registration: ClinicalTrials.gov Identifier: NCT06290544. Registered on 4 March 2024.