Abstract 36: Arm Exercise Stress Test Predictors of Outcome Over 12 Years

Abstract

For patients with lower extremity disabilities precluding treadmill exercise, pharmacologic nuclear imaging stress tests necessitate administration of ionizing radiation, require much greater equipment and personnel resources, and are several times more expensive and time-consuming than exercise electrocardiography (ECG). In addition, they do not provide powerful prognostic and clinically relevant information on exercise capacity or on symptomatic, hemodynamic and ECG responses to exercise. In lower extremity disabled individuals, we have reported that arm exercise (AXT) capacity in METs and delta (peak exercise - rest) heart rate ([[Unable to Display Character: ∆]]HR) prognosticate survival and an abnormal exercise ECG predicts all-cause mortality. However, these data included follow-up of only 5.3 years and did not provide information on heart rate recovery (HRR), an important prognosticator of survival for leg exercise. Furthermore, the predictive role of HRR has not been well-characterized in high-risk populations or for arm exercise. Thus, the purpose of the current investigation was to determine whether arm exercise HRR, AXT METs, [[Unable to Display Character: ∆]]HR and an abnormal ECG response independently prognosticate all-cause mortality, subsequent myocardial infarction (MI) or coronary revascularization (CVASC) over a longer follow-up period in patients who cannot perform leg exercise. A total of 446 veterans, aged 64 +/- 11 yrs. underwent AXT between 1997 and 2002 and were followed to an endpoint of death or at least 10 years. Average annual mortality, MI, CVASC and combined event rates were 7.0%, 1.9%, 2.8% and 12.1%, respectively. Over an average follow-up of 12.0 +/- 1.3 years for survivors, HRR at both 1 and 2 minutes post-exercise was highly and inversely related to mortality (p < 0.0001) by univariate analysis and remained highly significant after adjustment for demographic and clinical clinical variables (p< 0.0001 for HRR 1 minute post-exercise; hazard ratio (HR)/bpm = 0.95 for death; confidence interval (CI ) = 0.94-0.97). AXT METs was independently related to survival by univariate and multivariate analyses (both p < 0.0001; HR/MET = 0.55 for death; CI=0.46-0.66). Delta HR also predicted survival by both univariate and multivariate analyses (p < 0.0001; HR/bpm = 0.98 for death; CI = 0.98-0.99). An abnormal AXT ECG was predictive of mortality by univariate (p = 0.03) but not multivariate analysis. None of these variables predicted MI but an exercise endpoint of typical chest pain and an abnormal AXT ECG prognosticated subsequent CVASC (multivariate p < 0.0001; HR = 4.70; CI = 1.81-12.22 and p = 0.0002; HR = 2.64; CI = 1.61-4.33, respectively). We conclude that in high risk patients who cannot perform lower extremity exercise, HRR, AXT METs, and [[Unable to Display Character: ∆]]HR are independently associated with survival, an abnormal AXT ECG predicts univariate mortality, and an AXT endpoint of typical chest pain and abnormal AXT ECG portend subsequent CVASC.