Long-Term Mortality With Multiple Treadmill Exercise Test Abnormalities: Comparison Between Patients With and Without Cardiovascular Disease

Bilal Aijaz, MD; Luciano Babuin, MD; Ray W. Squires, PhD; Stephen L. Kopecky, MD; Bruce D. Johnson, PhD; Randal J. Thomas, MD; Thomas G. Allison, PhD


Am Heart J. 2008;156(4):783-789. 

In This Article


In this large cohort of patients referred for TMET, poor exercise capacity, limited heart rate reserve, and exercise-induced ECG abnormalities were independent predictors of all-cause mortality (after adjustment for age, gender, β-blocker use, and standard coronary risk factors) in patients without baseline CVD. This study confirms previous findings and extends them by demonstrating an incremental increase in mortality with the increase in the number of abnormal TMET results. A direct comparison between those with and without CVD shows that patients without CVD having 2 or 3 abnormal TMET responses had a similar mortality to those with CVD but no abnormal results. This simple combinatorial approach to interpretation of multiple abnormal TMET results provides clinically relevant information not described previously.

All 3 abnormal TMET responses (poor exercise capacity, limited heart rate reserve, and abnormal exercise ECG) were significantly associated with increased mortality in the no-CVD group; however, in the CVD group, poor exercise capacity and limited heart rate reserve, but not abnormal exercise ECG, were associated with increased risk of mortality. This may be due to our inability to evaluate post-TMET interventions and cardiovascular mortality. Patients with ischemic ECG changes may be more likely to have interventions for ischemia after an abnormal TMET. Furthermore, a high number of patients with baseline ECG abnormalities in this group may have affected the interpretation of the exercise ECG response.[23]

Previous studies relating TMET results to long-term outcomes have reported exercise capacity both in relative (percentage of age- and gender-predicted value)[24] and absolute (METs) terms.[12,25] We repeated our analyses (not shown) using METs rather than percentage of age- and gender-predicted exercise time and found essentially the same results, with the exception that gender adjustment became important in the multiple regression analysis when exercise capacity was expressed in METs without adjusting for gender. Although at least 1 article combined multiple treadmill test abnormalities into a single normogram to estimate mortality, the analysis was only for patients without baseline CVD.[26] The ultimate form of the normogram using 5 exercise variables and 7 clinical variables may be more complex compared with our simplified analysis based on only 3 exercise variables. Several previous studies have focused on prognostic information provided by TMET;[7,9,11,15,24,27] however, none provides a comparison of mortality in groups with and without CVD (in both men and women) with the increase in number of abnormal TMET results. Erikssen et al[7] included 2,014 healthy men and compared the accuracy of CVD risk assessment based on classic risk factors with several TMET variables. Goraya et al[11] (3,107 patients) included multiple TMET variables to establish prognosis in elderly persons. Lauer et al[15] and Cole et al[9] used only heart rate variables to establish prognosis. Myers et al[12] established exercise capacity as an independent predictor of mortality in 6,213 men. Poor exercise capacity has been shown to be a strong and independent predictor of all-cause mortality.[12,24,28] Exercise capacity, a modifiable risk factor, is a direct reflection of the functioning of cardiorespiratory system. Because physical inactivity is a major public health problem[29] and improvement in physical fitness improves prognosis,[30,31] physicians should encourage patients to be physically active.

Abnormal heart rate responses during TMET including exercise chronotropic incompetence, slow heart rate recovery, and poor heart rate reserve are well-known predictors of mortality.[9,25,32,33] Heart rate reserve, which includes both resting heart rate and maximal heart rate, was available in our cohort and therefore used.

Several important limitations must be considered. This is an observational, single-center experience. The study population included 95% whites. Whether results may be generalized to other populations cannot be ascertained. We did not examine post-abnormal-TMET interventions such as β-blocker use or coronary bypass surgery that might reduce mortality. This would, however, produce a negative bias because the abnormal TMET may have led to more of these interventions than in the normal-TMET group. All-cause mortality was chosen as a hard end point that is considered a better predictor than cardiovascular death[20,21] and is at lowest risk for ascertainment bias.[21,22] Disease-specific mortality (myocardial infarction, heart failure, stroke, etc) may provide additional information. Exercise capacity was determined from the speed and grade of the treadmill. Directly measuring the oxygen consumption is more accurate. Heart rate recovery is an important predictor for mortality;[9] however, at the time of database design, heart rate recovery was not widely used and hence not stored in our database. Although results will probably be comparable if heart rate reserve was substituted for by heart rate recovery, we cannot rule out any difference between these 2 parameters. There are several differences from the Duke treadmill score, which was not available for comparison in this study. The Duke treadmill score was formulated and validated in people with chest pain only, while using cardiovascular death and infarction as end points, not all-cause mortality as used in this study.[34] Chest pain was a common feature in this study population but not a universal feature. The size of the study population (>10,000) is larger and the duration of follow-up (>10 years) is longer in the current study. The Duke treadmill score also lacks an important parameter (heart rate), the importance of which has been shown in several studies.

Treadmill exercise test is a commonly available, noninvasive, and relatively low-cost test, increasingly being used by noncardiologists. Its role as a risk-stratification tool, without necessarily requiring further invasive testing, is of particular importance to internists. This study confirms the important prognostic information provided by 3 important TMET variables: exercise capacity, heart rate reserve, and ECG abnormalities. A simple additive score provides important information in predicting an incremental increase in mortality with the number of abnormal TMET results. Patients having at least 2 abnormal exercise test variables had similar mortality to those with only CVD but a normal test. This will help clinicians stratify patients without known CVD but with multiple abnormal TMET results into a high-risk group (with mortality comparable to those with CVD), requiring both lifestyle modifications and aggressive pharmacologic treatment.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.