Background Exercise treadmill testing is used to evaluate patients for coronary disease and predict risk of future cardiac events. While it is known that the Duke Treadmill Score (DTS) obtained from this test is a useful composite prognostic index used to determine event-free survival, we are less clear about the impact and interactions of pre-existing coronary risk factors on exercise outcome and ejection fraction.
Objective To evaluate the impact of pre-existing coronary risk factors on exercise outcome during cardiac stress testing and cardiac function.
Methods We analysed a prospective registry of patients referred for exercise stress myocardial perfusion imaging between 1 January 2006 and 31 December 2008 who had no known coronary disease (CAD). Clinical characteristics including symptoms, medical history and pre-existing coronary risk factors were prospectively collected and validated at time of exercise testing. The DTS was computed immediately after exercise testing by trained cardiologists using standard formula. Cardiac ejection fraction post-stress was obtained by nuclear myocardial perfusion imaging. T-test and ANOVA were used to compare mean DTS between groups. A linear regression model was fitted to determine independent predictors of DTS.
Results We looked at 6298 patients (mean age 55 ± 11, 38% females) with diabetes (20%), hypertension (51%), hyperlipidemia (14%), smoking (63%) and family history of CAD (17%). There was a statistically significant correlation between the presence of each risk factor and mean DTS (p < 0.05). Patients who had diabetes had lower mean DTS compared to those who did not have diabetes (DTS = 7.73 vs 9.31, p < 0.001). There were significant differences in DTS between EF categories (EF ≤ 30, median DTS = 6.00; EF = 31–50, median DTS = 8.00; EF > 50, median DTS = 8.50; p = 0.006). The association between DTS and EF with risk factors was cumulative; patients who had 3 risk factors were more likely to have lower DTS (OR 1.419, 95% CI 1.296–1.555, p with EF≤30 vs 18% with EF > 50, p < 0.001); compared to patients with no risk factors. Linear regression analysis showed that female gender (beta = –2.152, p < 0.001) and diabetes (beta = –0.968, p < 0.001) were strong independent predictors of DTS.
Conclusions We found significant associations between pre-existing coronary risk factors and exercise outcome. The impact of risk factors was cumulative; in particular gender and diabetes independently predicted DTS.