Integrating genetic and imaging information to enhance cardiovascular risk stratification: rationale and design of the CVRISK-IT randomized controlled trial

Abstract

Aims

Cardiovascular disease (CVD) remains the leading cause of morbidity and mortality worldwide. Current guidelines recommend algorithms to estimate 10-year CVD risk in apparently healthy individuals, but their precision remains limited. Whether the integration of imaging and genetic information can improve individual risk stratification remains uncertain.

Methods and results

The CVRISK-IT trial is a multicentre, open-label, randomized controlled study designed to enhance CVD prevention through precision-based risk assessment and targeted communication strategies. A total of 30 000 adults aged 40–80 years without previous CVD or diabetes are recruited across Italy. In Phase I, participants undergo a baseline assessment including clinical examination, blood sampling, and calculation of CVD risk. Approximately 12 000 eligible enter Phase II and are randomized equally into four intervention arms: (i) standard care (control), (ii) imaging-based risk refinement (coronary artery calcium score or carotid ultrasound), (iii) polygenic risk score (PRS), or (iv) combination of imaging and PRS. The co-primary endpoints are differences in estimated 10-year CVD risk at 12 months and the incidence of major CVD events, CVD mortality, and all-cause mortality at 5 years. Secondary endpoints include differences in CVD risk factors, prescription and adherence to preventive drug adherence, and behavioural or psychological measures.

Conclusion

By prospectively evaluating the feasibility and impact of incorporating genetic and imaging information in CVD risk stratification, CVRISK-IT will generate robust evidence to inform its integration into routine screening practices and public health policy.

Trial registration

ClinicalTrials.gov Identifier NCT06832644.

Graphical Abstract

Graphical Abstract.

Key Learning Points.

What is already known

• Conventional CVD risk scores have limited precision in individuals with similar risk factor profiles.

• Imaging and genetic information can improve risk stratification as integrated risk modifiers.

• Evidence from large-scale randomized trials on behavioural or clinical impact of using these methods is lacking.

What this study adds

• CVRISK-IT is a large-scale randomized trial that evaluates the impact of incorporating imaging and genetic information into CVD risk stratification, to improve targeting of preventive interventions in primary care.

• It evaluates short- and long-term effects on estimated CVD risk, behaviour, preventive therapy uptake, and outcomes.

• Findings will gain insights on the integration of precision tools into routine CVD prevention and public health policy.

Background

Cardiovascular disease (CVD) remains the leading cause of death and disability worldwide, accounting for over one-third of all deaths each year.¹ In Europe, despite major advances in prevention and treatment, CVD continues to impose a substantial public health and economic burden.^2,3 Effective primary prevention strategies often involve identifying individuals at increased risk in order to enable timely implementation of targeted lifestyle and pharmacological interventions. The European Society of Cardiology (ESC) currently recommends use of the SCORE2 and SCORE2-Older Persons (SCORE2-OP) algorithms for estimating 10-year CVD risk in apparently healthy adults.⁴ These models combine traditional risk factors such as age, sex, blood pressure, smoking, and lipid levels to guide preventive strategies across Europe.^5,6

However, CVD risk can be substantially different even among people with similar conventional risk factor profiles.⁷ To improve risk stratification, the ‘2021 ESC Guidelines on cardiovascular disease prevention in clinical practice’ encourage the use of additional ‘risk modifiers’ including biomarkers such as high-sensitivity C-reactive protein.⁴ Recent guidelines and consensus statement have also extended this approach by incorporating both imaging and genetic information as valuable risk modifiers.^8-10 Coronary artery calcium (CAC) scoring is recognized as a practical imaging-based tool to detect subclinical atherosclerosis and to reclassify individuals at borderline or intermediate risk, particularly when treatment decisions are uncertain.⁸ When CAC scoring is unavailable or not feasible, carotid ultrasound may be considered as an alternative to detect atherosclerotic disease and refine risk classification around treatment thresholds.¹⁰ Similarly, polygenic risk scores (PRS), which aggregate thousands of genetic variants, can quantify inherited cardiovascular susceptibility, improve risk prediction, and enable earlier preventive action.⁹ However, no large-scale randomized studies have examined the impact of integrating both imaging and genetic information for CVD risk assessment on health-related behaviour and clinical outcomes. Most studies to date have been observational or limited by small sample sizes, heterogeneous populations, or short follow-up periods.^11-18

The CVRISK-IT trial was established to address these knowledge gaps by evaluating the short- and long-term effects of integrating imaging and genetic information into CVD risk communication. The study aims to determine whether providing individuals with personalized risk information derived from these modalities enhances risk perception, supports more tailored lifestyle and treatment recommendations, improves adherence to preventive measures, and ultimately reduces overall CVD risk.

Objectives

Primary objective

The primary objective of the CVRISK-IT trial is to evaluate the health benefits of incorporating additional CVD risk information, derived from imaging and genetic data, into conventional risk assessment. Specifically, the study aims in the short term to determine whether providing personalized information based on imaging and/or PRS improves changes in estimated 10-year CVD risk, as calculated by the SCORE2/SCORE2-OP algorithms, compared with standard care using conventional risk factors alone. In the longer term, CVRISK-IT will investigate the impact of integrating imaging and genetic information on incidence of major CVD events, CVD mortality, and all-cause mortality.

Secondary objectives

The secondary objectives are to assess how enhanced risk communication influences modifiable CVD risk factors and to evaluate changes in preventive drug prescriptions and adherence to recommended therapies. The study will also examine behavioural and psychological outcomes, including changes in risk perception, anxiety, and motivation for lifestyle modification.

Methods

Study design

CVRISK-IT is a multicentre, open-label, randomized controlled trial conducted within the framework of the Italian Cardiology Network (ICN), involving 17 Scientific Institutes for Research, Hospitalisation and Health Care (known by the Italian acronym IRCCS) and their affiliated centres distributed across Italy (Figure 1 and Supplementary material online, eTable S1). The study is centrally coordinated by the IRCCS Policlinico San Donato Hospital, with operational support from a dedicated trial management group responsible for protocol implementation, data management, and quality assurance. Oversight is provided by national and institutional ethics committees to ensure compliance with all regulatory and ethical requirements. All study procedures are conducted in accordance with Good Clinical Practice and the ethical principles of the Declaration of Helsinki. The trial is registered at ClinicalTrials.gov NCT06832644.

Figure 1.

Geographic distribution of the enrolling centres involved in the CVRISK-IT trial. Red dots: main recruiting centres (IRCCS); Pins: current affiliated centres.

The trial includes two phases: Phase I, the establishment of a large national bioresource of adults undergoing CVD risk profiling, and Phase II, a randomized evaluation of different strategies for risk stratification (Figure 2).

Figure 2.

Study design and flowchart of the CVRISK-IT trial.

Participants and recruitment

In Phase I, 30 000 apparently healthy adults aged 40 to 80 years without a history of CVD or diabetes are recruited across Italy through workplace screening programmes, care homes, volunteer organizations, and participating hospitals (see Supplementary material online, eTable S1). A participant flow diagram is presented in Supplementary material online, eTable S2 of the Supplementary Material.¹⁹ Individuals must have sufficient proficiency in Italian to provide informed consent and complete study questionnaires. Exclusion criteria include a previous medical history related to CVD (e.g. heart failure, congenital heart disorder, coronary heart disease) based on medical diagnosis; previous history of type 1 or type 2 diabetes (as considered already at very high risk); medical diagnosis of mental disorders; pregnancy based on self-reported information; oncology patients (based on physician indication); candidates who have previously participated in other interventional trial on CVD prevention;²⁰ and inability or unwillingness to provide biological samples (see Supplementary material online, eTable S3).

Baseline visit procedures

Participants provide informed consent, complete a web-based baseline questionnaire, undergo clinical examination, and provide fasting blood samples for biobanking (see Supplementary material online, Appendix A). CVD risk is estimated using SCORE2 or SCORE2-OP algorithms as recommended by the 2021 ESC Guidelines on Cardiovascular Disease Prevention.⁴ Individuals identified as having very high CVD risk are referred directly for clinical management and excluded from randomization, but monitored for long-term incident CVD event. All participants contribute a national biobank and data repository to support future research into chronic disease determinants in Italy.

Randomization and allocation concealment mechanism

Participants identified as having low-to-moderate or high predicted CVD risk, as defined by the 2021 ESC Guidelines on Cardiovascular Disease Prevention (Table 1), are invited to participate in Phase II. Invitations to Phase II initially follow a first-come, first-served process to ensure balanced representation across participating centres. Subsequently, recruitment will transition to an adaptive allocation framework designed to maintain proportional enrolment across sites and study groups. This pragmatic strategy allows continuous recruitment while maintaining operational feasibility within the multicentre network. Approximately 12 000 individuals will be randomly assigned in equal proportions to one of four study interventions: a ‘control group’ receiving standard SCORE2/SCORE2-OP risk assessment; a ‘genetic risk group’ receiving additional risk information based on PRS; an ‘imaging risk group’ receiving additional risk information derived from imaging-based assessment, including CAC scoring or carotid ultrasound, depending on centre capacity; and a ‘combined risk group’ receiving additional risk information based on PRS and imaging (Figure 3). Randomization is performed centrally using a secure, computerized system employing block randomization stratified by the centre. Block sizes vary randomly, and their sequence and size are concealed from study investigators. Because of the nature of the intervention, participants and clinicians cannot be blinded to group assignment, but data analysts and outcome assessors remain masked to allocation.

Table 1.

10-year CVD risk categories based on SCORE2 and SCORE2-OP in apparently healthy people stratified by age group

Cardiovascular risk	<50 years	50–69 years	≥70 years
Low-to-moderate CVD risk: risk factor treatment generally not recommended	<2.5%	<5%	<7.5%
High CVD risk: risk factor treatment should be considered	2.5 to <7.5%	5 to <10%	7.5 to <15%
Very high CVD risk: risk factor treatment generally recommended	≥7.5%	≥10%	≥15%

Thresholds suggested by 2021 ESC guidelines on cardiovascular disease prevention in clinical practice.⁴

Figure 3.

Factorial 2 × 2 randomization design used in the CVRISK-IT trial. *Individuals are further randomized to CAC scoring or carotid ultrasound (IMT) based on centres feasibility.

Interventions

Risk information is communicated through a standardized in-person or remote medical consultation that combines visual aids and personalized discussion to enhance understanding and motivation. All participants receive structured CVD prevention counselling based on ESC recommendations using a dedicated digital platform, ‘MyCardioSpace’.²¹ Recommendations include dietary habits, promotion of physical activity, and smoking cessation,⁴ and, when indicated, pharmacological prevention, including lipid-lowering and antihypertensive therapy.

Participants allocated to the control arm will receive standard CVD risk estimates derived from SCORE2 or SCORE2-OP, accompanied by the corresponding evidence-based prevention recommendations. Participants in the genetic arm will receive additional risk information derived from a PRS, computed from genome-wide genotyping data and categorized into relative risk strata. Participants in the imaging arm will undergo cardiovascular imaging, either CAC scoring or carotid ultrasound. Imaging findings will be classified according to predefined quantitative thresholds.^21-23 For participants in the genetic and imaging arms, the PRS and/or imaging findings will be used solely for risk reclassification purposes within the overall CVD risk assessment framework. Clinical management recommendations will not be individualized based on specific PRS percentiles or CAC scores but will instead follow standardized guideline-directed preventive strategies corresponding to the participant’s final risk category after reclassification. This approach ensures consistency of preventive advice across study arms while allowing evaluation of the impact of risk modifiers on overall risk stratification (see Supplementary material online, Appendix B). Downward reclassification based on imaging or genetic findings is systematically captured for analytical purposes; however, to preserve adherence to current clinical guidelines, it does not lead to de-escalation of recommended preventive strategies within the trial.

Follow-up and outcome measures

Participants are followed up at 12 months after risk communication visit through repeat questionnaires, blood collection, and clinical assessments. Over long-term follow-up, the trial aims to demonstrate whether short-term improvements translate into lower incidence of major cardiovascular events. Long-term outcomes will be tracked through linkage with national registries of hospital admissions and mortality records, in particular CVD events and cause-specific mortality. The co-primary outcomes are the difference in estimated 10-year CVD risk, calculated using SCORE2 or SCORE2-OP, at 12 months and incidence of CVD events (i.e. non-fatal myocardial infarction, acute coronary syndrome, percutaneous transluminal coronary angioplasty, coronary artery bypass grafting, non-fatal stroke, transient ischaemic attach, and new diagnosis or hospitalization for heart failure and atrial fibrillation), CVD mortality, and all-cause mortality at 5 years.

Secondary outcomes include between-group differences in modifiable risk factors such as systolic blood pressure, lipids, smoking status, and body weight (both in isolation and in combination using CVD risk algorithms other than SCORE2;²² differences in initiation of preventive drug prescriptions and adherence; and behavioural and psychological measures, including perceived risk and motivation for lifestyle change. Longer term follow-up is also planned at 10 years.

Data management and quality assurance

All data are collected electronically through a secure web-based system managed by the central coordinating centre. Data quality is ensured through automated validation checks and regular monitoring visits. Biological samples are processed and stored at the national biobank, ensuring traceability and long-term preservation (see Supplementary material online, Appendix C). Data are pseudonymized and managed in compliance with the General Data Protection Regulation (GDPR) and Italian privacy legislation. Study data are collected and managed using REDCap electronic data capture tools^23,24 hosted on the IT platform of the ICN, developed in collaboration with the Consortium of Bioengineering and Medical Informatics (CBIM).

Sample size and statistical analysis

Sample size calculation

The trial adopts a 2 × 2 factorial randomized design, allowing simultaneous evaluation of two interventions—genetic and imaging-based CVD risk communication—both independently and in combination. A total of 12 000 participants will be randomized equally across four study groups (control, genetic, imaging, and combined genetic + imaging), corresponding to approximately 3000 individuals per arm (see Supplementary material online, Appendix D). Sample size determination was based on simulation analyses exploring a range of plausible event rates for the CVD incidence co-primary outcome in the control group (2–10%) and assuming relative risks of 0.85 for each intervention. Interaction effects between the two interventions were modelled under different scenarios, including absence of interaction, synergistic, and antagonistic effects. For each condition, data were simulated for 1000 repetitions and analysed using logistic regression models incorporating the main and interaction effects. These simulations confirmed that a total enrolment of 12 000 participants provides at least 80% power at a two-sided significance level of 0.05 to detect clinically meaningful differences in the primary endpoint across intervention groups, while accommodating potential attrition. To ensure that 12 000 participants are randomized while maintaining balanced representation across centres, Phase I enrolment is set at 30 000 individuals.

Approximately 30% of screened participants are expected to fall into the very-high-risk category by SCORE2/SCORE2-OP and thus be excluded from randomization. Of the remaining participants, an additional 10–20% are anticipated to decline genetic or imaging procedures, withdraw consent, or face logistical exclusions. This approach is expected to yield approximately 15 000–18 000 eligible participants, ensuring a sufficient number of individuals for Phase II. The analyses were performed using Stata software (StataCorp, College Station, TX).

Statistical analysis

All data will undergo initial analysis in accordance with recommendations from the STRATOS Initiative.²⁵ This process will include metadata setup, systematic data cleaning, quality screening, and descriptive reporting before formal hypothesis testing. Frequency distributions will be inspected to identify missing data, outliers, and implausible values. Continuous variables will be summarized as mean ± SD or median (interquartile range), depending on distribution; categorical variables will be presented as frequencies and percentages.

The primary analysis will follow the intention-to-treat principle. Frequency of missing outcome data will be compared across intervention groups. Participants with missing follow-up data will be excluded from the primary model (complete case analysis), with sensitivity analyses performed to assess the robustness of findings to missingness. Missing data will primarily be handled by complete-case analysis for the primary model. Multiple imputation will be conducted as part of ancillary sensitivity analyses to assess robustness under missing-at-random assumptions. The effect of the study group (control, genetic, imaging, or combined) on the 12-month differences in estimated CVD risk (SCORE2 or SCORE2-OP) will be assessed using a two-way analysis of covariance model, incorporating the main effects of genetic and imaging information and their interaction and adjusting for centre and baseline assessments of the analysed outcomes.

Where a significant interaction is observed, pairwise comparisons will be estimated as follows: (i) genetic + imaging vs. genetic only—to assess the incremental impact of imaging information; (ii) genetic + imaging vs. Imaging only—to assess the incremental impact of genetic information; and (iii) genetic + imaging vs. control—to assess the combined effect relative to standard care. If no interaction is detected, main effects will be reported separately for genetic and imaging interventions. Subgroup analyses will also be performed by age, sex, and baseline risk to assess potential effect modification by incorporating relevant interaction terms with interventions. Significant interactions (P < 0.05) will prompt stratified reporting by subgroup categories defined by median and quartile cut-offs. Exploratory analyses will compare CAC- vs. carotid ultrasound-based reclassification. When long-term follow-up data will become available, time-to-event analyses for CVD events and mortality will be conducted using Kaplan–Meier curves and Cox proportional hazards models. The proportional hazards assumption will be verified using Schoenfeld residuals and formal test assessing interaction of interventions with follow-up time. Competing risks will be addressed using Fine–Gray subdistribution hazard models, with Gray’s test applied for group comparisons of cumulative incidence functions. Secondary analyses will similarly evaluate differences in individual risk factors (blood pressure, lipid profile, smoking, body weight), adherence to preventive medication, and psychological or behavioural outcomes.

All analyses will be performed using SAS 9.4 (SAS Institute, Cary, NC) and R software (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance will be set at two-sided P ≤ 0.05 for assessment of intervention main effects and P ≤ 0.01 for subgroup analyses.

Ethical considerations

The CVRISK-IT trial has received ethical approval from the Central Ethics Committee and all participating IRCCS institutions. Written informed consent is obtained for both study participation and biobanking in accordance with EU eIDAS and GDPR standards. Oversight is provided by a Trial Steering Committee responsible for scientific and ethical integrity, a Trial Management Group for operational conduct, and an Independent Data Monitoring Committee that regularly reviews recruitment progress, safety data, and study performance.

Given the non-invasive nature of the study, risks to participants are minimal. Potential harms are limited to minor discomfort or anxiety associated with risk disclosure, venipuncture, and imaging procedures. Radiation exposure from CAC scanning will be kept as low as reasonably achievable, following national safety standards. Any adverse or unintended effects related to study participation will be recorded and reviewed by the Independent Data Monitoring Committee.

Results

Based on the trial’s design and planned co-primary and secondary outcomes, several results can be anticipated. In the short-term (12 months), it is expected that providing additional personalized information derived from imaging and/or PRS may lead to reductions in estimated 10-year CVD risk, compared with standard SCORE2/SCORE2-OP–based assessment alone. Participants receiving imaging or genetic risk information—particularly those in the combined arm—might show an improved modification of key risk factors, such as systolic blood pressure, LDL cholesterol levels, smoking prevalence, and/or body weight.

At the behavioural level, personalized CVD risk communication is expected to positively influence participants’ risk perception, motivation for lifestyle change, and engagement with preventive strategies. The digital support platform will aim to further strengthen and enhance these behavioural effects across all intervention groups.

Over long-term follow-up (5 years), the trial aims to demonstrate whether these short-term improvements translate into lower incidence of major cardiovascular events, cardiovascular mortality, and all-cause mortality, especially for the combined imaging + polygenic risk arm.

Overall, the findings are anticipated to provide robust evidence on the feasibility and potential clinical utility of integrating imaging and genetic risk modifiers into routine cardiovascular prevention strategies in apparently healthy adults.

Discussion

Despite major advances in prevention and treatment, population ageing, unhealthy lifestyle habits, and persistent inequalities continue to sustain the CVD burden. Current prevention strategies rely on established algorithms such as SCORE2 and SCORE2-OP, which integrate conventional risk factors including age, sex, blood pressure, and cholesterol. However, these models do not capture the full biological and subclinical heterogeneity of CVD. Evidence suggests that imaging measures of atherosclerosis and genetic susceptibility, quantified by PRS,⁹ have the potential to substantially improve individual risk characterization. Yet, whether this information enhance preventive behaviours or clinical outcomes remains uncertain.

The CVRISK-IT trial is designed to address this evidence gap by testing whether integrating imaging and/or genetic information into CVD risk communication produces meaningful changes in estimated risk, preventive behaviours, and long-term outcomes. By combining a large-scale, randomized design with a national infrastructure, the study aims to provide robust evidence on the clinical utility of precision-enhanced prevention strategies.

The trial leverages two validated and complementary risk modifiers: imaging to quantify subclinical atherosclerosis and PRS to estimate inherited CVD susceptibility. Together, these tools bridge phenotypic and genetic determinants of risk, supporting a more comprehensive and personalized approach to prevention. The factorial randomized design allows assessment of both the independent and combined effects of imaging and genetic information, providing insight into the potential synergistic benefit of multimodal risk communication.

Furthermore, the study uses of a dedicated digital platform, ‘MyCardioSpace’,²¹ to deliver results, monitor engagement, and support lifestyle counselling. During follow-up, this platform enables continuous participant interaction through personalized recommendations (e.g. digital notifications and phone calls), remote data collection, ensuring consistency and active communication while improving trial efficiency. Such a hybrid digital–clinical model aligns with current trends in CVD prevention and may serve as a template for future large-scale studies. Samples form CVRISK-IT will also contribute to the BBDCARDIO biobank and the Italian Cardiology Network data infrastructure providing lasting value beyond the trial itself, and supporting future research on the determinants of CVD [e.g. lipoprotein(a), C-reactive protein] and other chronic conditions in collaboration with other Italian initiatives.²⁰

If successful, CVRISK-IT could have substantial implications for both clinical practice and public health policy. Demonstrating that precision tools such as imaging and PRS improve adherence, treatment uptake, or estimated CVD risk would provide strong justification for their integration into European and national prevention guidelines. Furthermore, evidence from a large, real-world population will inform cost-effectiveness analyses, supporting decisions on resource allocation in preventive cardiology. The study also has the potential to refine risk communication strategies. Understanding how individuals respond to imaging and genetic information—emotionally, cognitively, and behaviourally—will provide insight into how personalized risk communication can most effectively promote lasting behavioural change.

Several potential limitations must be acknowledged. The large multicentre structure, involving 17 IRCCS institutes and affiliated centres, introduces potential heterogeneity in participant recruitment, data collection, and clinical practice. Although harmonized protocols and central quality control procedures have been implemented, minor inter-centre variability is inevitable. Imaging assessments represent another potential source of variation. Despite calibration, standardization, and central review, differences in scanner technology and operator expertise may influence measurement precision. The genetic component, being fully centralized, is less susceptible to such variability. Risk communication sessions, which may be performed either in person or remotely, could vary in delivery style and patient engagement. While this pragmatic flexibility reflects real-world practice, it may introduce heterogeneity in risk perception and behavioural responses. Additionally, the requirement for digital access and literacy may lead to a study population that is not fully representative of the broader population, potentially limiting external validity. Maintaining participant engagement and complete data linkage over extended follow-up periods will be essential to ensuring the reliability of long-term outcomes. Finally, extended long-term follow-up is envisioned to complement registry-based outcome tracking.

In conclusion, CVRISK-IT represents a major national initiative to evaluate whether imaging and genetic information can enhance CVD risk assessment and prevention in apparently healthy adults. Its combination of rigorous trial design, digital engagement infrastructure, and large-scale biobanking positions it to provide robust evidence on the clinical and behavioural impact of precision-based risk communication. By contributing to the Italian Cardiology Network and BBDCARDIO, the trial will help clarify the potential value of multimodal risk information for CVD prevention and provide a basis for future research on early detection and personalized health strategies across chronic diseases.

Supplementary material

Supplementary material is available at European Heart Journal—Quality of Care and Clinical Outcomes online.

Author contributions

Emanuele Di Angelantonio (Conceptualization [lead]; Data curation [lead]; Formal analysis [lead]; Investigation [lead]; Methodology [lead]; Validation [lead]; Writing—original draft [lead]; Writing—review & editing [lead]), Paolo G. Camici (Conceptualization [equal]; Data curation [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Pietro Ameri (Conceptualization [supporting]; Data curation [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Luigi Frati (Conceptualization [equal]; Resources [equal]; Validation [equal]), Amalia De Curtis (Methodology [equal]; Resources [supporting]), Marta Rigoni (Data curation [equal]; Formal analysis [equal]; Writing—review & editing [equal]), Maria Carla Roncaglioni (Investigation [supporting]; Methodology [supporting]; Validation [supporting]), Roberto Latini (Conceptualization [equal]; Investigation [equal]; Methodology [equal]; Validation [supporting]), Martino F. Pengo (Conceptualization [equal]; Data curation [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Gianfranco Parati (Conceptualization [equal]; Data curation [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Maria Teresa La Rovere (Conceptualization [supporting]; Investigation [equal]; Methodology [equal]; Validation [equal]), Giuseppe Ferrante (Conceptualization [equal]; Data curation [equal]; Formal analysis [supporting]; Investigation [equal]; Methodology [equal]; Validation [equal]), Anna Severino (Investigation [supporting]; Project administration [supporting]; Validation [supporting]), Stephen Kaptoge (Formal analysis [equal]; Methodology [equal]; Writing—review & editing [equal]), Massimo Volpe (Conceptualization [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]), Giovanna Liuzzo (Conceptualization [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Roberta Baetta (Project administration [supporting]; Writing—review & editing [equal]), Mauro Amato (Data curation [supporting]; Formal analysis [supporting]; Methodology [equal]; Validation [supporting]; Writing—review & editing [equal]), Gualtiero Colombo (Conceptualization [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Damiano Baldassarre (Conceptualization [equal]; Data curation [equal]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Sebastian Guelfi, Federico Ambrogi (Data curation [supporting]; Formal analysis [supporting]; Methodology [supporting]), Rosanna Cardani (Methodology [equal]; Resources [supporting]; Writing—review & editing [equal]), Sara Boveri (Data curation [supporting]; Formal analysis [supporting]; Methodology [supporting]), Ambra Cerri (Conceptualization [supporting]; Funding acquisition [lead]; Project administration [lead]; Resources [equal]), Carmelina Chiarello (Conceptualization [supporting]; Funding acquisition [lead]; Project administration [equal]), Chiara Barberio (Methodology [supporting]; Project administration [lead]; Writing—original draft [supporting]; Writing—review & editing [equal]), Serenella Castelvecchio (Investigation [lead]; Methodology [equal]; Validation [equal]), Massimo Piepoli (Data curation [supporting]; Investigation [equal]; Methodology [equal]; Validation [equal]; Writing—review & editing [equal]), Eloisa Arbustini (Investigation [supporting]; Methodology [supporting]; Writing—review & editing [equal]), and Lorenzo Menicanti (Conceptualization [lead]; Data curation [equal]; Formal analysis [equal]; Funding acquisition [supporting]; Investigation [lead]; Methodology [lead]; Validation [equal]; Writing—original draft [equal])

Funding

This study was supported by Ricerca Corrente Reti funding (RCR-2023-23684267) from the Italian Ministry of Health to IRCCS Policlinico San Donato.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.