Estimated age based on exercise stress testing performance outperforms chronological age in predicting mortality

Aims: We sought to estimate patients' age based on their stress testing exercise performance (A-BEST), and evaluate whether A-BEST would be a better predictor of mortality when compared to chronological age.

Methods: We included 126,356 consecutive patients referred for exercise (electrocardiography, echocardiography or myocardial perfusion imaging) stress testing at our institution from January 1st, 1991 to February 27th, 2015. Estimated age was computed based on exercise capacity (number of peak estimated metabolic equivalents of task), chronotropic reserve index and heart rate recovery, taking into account patient's gender and medications that affect heart rate. Uni and multivariable Cox models were used to determine the association of A-BEST with mortality. Improvement in predicting mortality using A-BEST compared to chronological age was evaluated with the use of net reclassification improvement and C statistic.

Results: Mean age was 53.5 ± 12.6 years and 59% were men. At follow-up (mean duration was 8.7 years), 9929 (8%) died. After adjustment for clinical comorbidities, higher metabolic equivalents of task (adjusted hazard ratio (HR) for mortality 0.71, 95% confidence interval (CI) 0.70-0.72, P < 0.001) and higher chronotropic reserve index (adjusted HR for mortality 0.97, 95% CI 0.96-0.99, P = 0.0135) were associated with improved survival, whereas abnormal heart rate recovery (adjusted HR for mortality 1.53, 95% CI 1.46-1.61, P < 0.001) and higher A-BEST (adjusted HR for mortality 1.05, 95% CI 1.04-1.05, P < 0.001) were associated with higher mortality. When comparing prediction models using A-BEST versus chronological age, a significant increase in the area under the curve was demonstrated if A-BEST was used (0.82 vs. 0.79, P < 0.001). The overall net reclassification improvement was 0.30 (P < 0.001).

Conclusion: Estimated age based on exercise stress testing performance is a better predictor of mortality when compared to chronological age.