Rosamond, Wayne

Introduction: Deep neural networks are increasingly used to generate diagnostic algorithms in cardiovascular medicine. A biomarker of cardiovascular aging, derived from a deep-learning algorithm applied to digitized 12-lead electrocardiograms (ECGs), has recently been introduced. This biomarker, delta age (δ-age), is defined as the difference between predicted ECG age and chronological age. Hypthesis: We hypothesized that δ-age will improve the prediction of incident fatal and non-fatal myocardial infarction (MI) and stroke on top of what contemporary CVD prediction tools would do. Methods: In this cohort study, we included 7,111 men and women from the Norwegian Tromsø Study conducted in 2015-16, with follow-up through 2021 for incident fatal and non-fatal myocardial infarction (MI) and hemorrhagic or cerebral stroke. We used Cox proportional hazards regression models to assess the independent effect of δ-age on MI and stroke. Discrimination was evaluated using Harrell’s concordance statistic (C-index) and the net reclassification improvement (NRI). Results: During a median follow-up of 5.9 years, we observed 155 incident cases of MI, 141 cases stroke and 290 cases with either MI or stroke. δ-age was observed with a mean of 0 and standard deviation of 6.2 years. In men and women combined, hazard ratios (HRs) per standard deviation increase in δ-age, after adjustment for traditional risk factors, were 1.24 (95% confidence interval (CI) 1.09, 1.41) for the combined outcome, 1.26 (1.06, 1.49) for MI and 1.25 (1.03, 1.50) for stroke. In men, the corresponding HRs were 1.27 (1.09, 1.49), 1.46 (1.18, 1.79) and 1.02 (0.80, 1.31), respectively, and in women, 1.20 (0.97, 1.49), 0.87 (0.64, 1.20) and 1.58 (1.19, 2.11), respectively. The C-index increased modestly when δ-age was added to a model with traditional risk factors. The 95% CI for the C-index increase excluded zero for the combined outcome overall, and for MI in men and for stroke in women. The NRI was 26.0% (13.3%, 38.1%) for the combined outcome, 17.5% (0.6%, 33.5%) for MI and 37.2% (20.1%, 53.0%) for stroke. Conclusions: Incorporating δ-age into primary prevention risk prediction models significantly improved performance beyond traditional cardiovascular risk factors for the combined outcome and separately for MI and stroke.

Background Prior studies have demonstrated a link between the metabolic syndrome and increased risk of cardiovascular mortality. Whether the metabolic syndrome is associated with sudden cardiac death is uncertain. Methods and Results We characterized the relationship between sudden cardiac death and metabolic syndrome status among participants of the ARIC (Atherosclerosis Risk in Communities) Study (1987–2012) free of prevalent coronary heart disease or heart failure. Among 13 168 participants, 357 (2.7%) sudden cardiac deaths occurred during a median follow‐up of 23.6 years. Participants with the metabolic syndrome (n=4444) had a higher cumulative incidence of sudden cardiac death than those without it (n=8724) (4.1% versus 2.3%, P <0.001). After adjustment for participant demographics and clinical factors other than components of the metabolic syndrome, the metabolic syndrome was independently associated with sudden cardiac death (hazard ratio, 1.70, 95% confidence interval, 1.37–2.12, P <0.001). This relationship was not modified by sex (interaction P =0.10) or race (interaction P =0.62) and was mediated by the metabolic syndrome criteria components. The risk of sudden cardiac death varied according to the number of metabolic syndrome components (hazard ratio 1.31 per additional component of the metabolic syndrome, 95% confidence interval, 1.19–1.44, P <0.001). Of the 5 components, elevated blood pressure, impaired fasting glucose, and low high‐density lipoprotein were independently associated with sudden cardiac death. Conclusions We observed that the metabolic syndrome was associated with a significantly increased risk of sudden cardiac death irrespective of sex or race. The risk of sudden cardiac death was proportional to the number of metabolic syndrome components.