The American College of Sports Medicine recently published new guidelines to classify individuals at increased risk for coronary heart disease (CHD) before starting a vigorous (>60% maximum oxygen consumption) exercise program. We compared the prognostic value of the new guidelines to the earlier guidelines.
Subjects included men and women over 30 years of age, free from known CHD according to the Lipid Research Clinic Follow-up Cohort. Endpoints included an abnormal exercise test result (± 1 mm ST, anginal chest pain, or exercise hypotension) and future death of CHD ascertained during an average 12.2 years of follow-up. Data analysis included 2 × 2 contingency tables to compare the test characteristics of the old and new guidelines.
Among 4,074 subjects, 219 (5.4%) had abnormal exercise test results and 65 (1.6%) eventually died of CHD. The new guidelines classified 75.7% of individuals as high risk versus 58.7% using the old guidelines. Comparing the new guidelines with the old, test sensitivity was significantly (P = 0.007) higher for patients with an abnormal exercise test result (87% versus 77%) than for patients who died of CHD (100% versus 99%). Test specificity of the new guidelines was significantly (P < 0.001) lower than the old guidelines for individuals with abnormal exercise test results (25% versus 42%) and those who died of CHD (25% versus 42%). The new guidelines also showed significantly lower overall diagnostic accuracy and positive likelihood ratios compared with the old guidelines for both patients with abnormal exercise test results and those who died of CHD. The positive predictive value for the new and old guidelines for both patients with abnormal exercise test results and those who died of CHD were similar.
Although the new guidelines are more sensitive in the abnormal exercise test result endpoint, they are less specific and overall less accurate than the old guidelines given the low prevalence of CHD in this asymptomatic population. The current guidelines should be modified to better target high-risk adults.
From the *Centre for Cardiovascular Risk Assessment, and the †Divisions of General Internal Medicine and ‡Clinical Epidemiology, Montreal General Hospital; and Departments of §Medicine and ∥Epidemiology & Biostatistics, McGill University, Montreal, Quebec.
Address for correspondence: Steven Grover, MD, MPA, FRCPC, Centre for Cardiovascular Risk Assessment, The Montreal General Hospital, 1650 Cedar Avenue, Montreal, Quebec, H3G 1A4.