Development and validation of decision rules models to stratify coronary artery disease, diabetes, and hypertension risk in preventive care: cohort study of returning UK Biobank participants

Background A wide range of predictive models exist that predict risk of common lifestyle conditions. However, these have not focused on identifying pre-clinical higher risk groups that would benefit from lifestyle interventions and do not include genetic risk scores. Objective To develop, validate, and compare the performance of three decision rule algorithms including biomarkers, physical measurements and genetic risk scores for incident coronary artery disease (CAD), diabetes (T2D), and hypertension in the general population against commonly used clinical risk scoring tools. Methods We identified 60782 individuals in the UK Biobank study with available follow-up data. Three decision rules models were developed and tested for an association with incident disease. Hazard ratios (with 95% confidence interval) for incident CAD, T2D, and hypertension were calculated from survival models. Model performance in discriminating between higher risk individuals suitable for lifestyle intervention and individuals at low risk was assessed using the area under the receiver operating characteristic curve (AUROC). Results We ascertained 500 incident CAD cases, 1005 incident T2D cases, and 2379 incident cases of hypertension. The higher risk group in the decision rules model had a 40-, 40.9-, and 21.6-fold increase in risk of CAD, T2D, and hypertension, respectively ( P < 0.001 for all). Risk increased significantly between the three strata for all three conditions ( P < 0.05). Risk stratification based on decision rules identified both a low-risk group (only 1.3% incident disease across all models), as well as a high-risk group where at least 72% of those developing disease within 8 years would have been recommended lifestyle intervention. Based on genetic risk alone, we identified not only a high-risk group, but also a group at elevated risk for all health conditions. Conclusion We found that decision rule models comprising blood biomarkers, physical measurements, and polygenic risk scores are superior at identifying individuals likely to benefit from lifestyle intervention for three of the most common lifestyle-related chronic health conditions compared to commonly used clinical risk scores. Their utility as part of digital data or digital therapeutics platforms to support the implementation of lifestyle interventions in preventive and primary care should be further validated.