The Relation of Arm Exercise Peak Heart Rate to Stress Test Results and Outcome.

ABSTRACT

PURPOSE: Arm exercise is an alternative to pharmacologic stress testing for >50% of patients unable to perform treadmill exercise, but no data exist regarding the effect of attained peak arm exercise heart rate on test sensitivity. Thus, the purpose of this investigation was to characterize the relationship of peak arm exercise heart rate responses to abnormal stress test findings, coronary revascularization, and mortality in patients unable to perform leg exercise. METHODS: From 1997 until 2002, arm cycle ergometer stress tests were performed in 443 consecutive veterans age 64.1 yr (11.0 yr) (mean (SD)), of whom 253 also underwent myocardial perfusion imaging (MPI). Patients were categorized by frequency distributions of quartiles of percentage age-predicted peak heart rate (APPHR), heart rate reserve (HRR), and peak heart rate-systolic blood pressure product (PRPP). Exercise-induced ST-segment depression, abnormal MPI findings, coronary revascularization, and 12.0-yr (1.3 yr) Kaplan-Meier all-cause and cardiovascular mortality plots were then characterized by quartiles of APPHR, HRR, and PRPP. RESULTS: A reduced frequency of abnormal arm exercise ECG results was associated only with the lowest quartile of APPHR (≤69%) and HRR (≤43%), whereas higher frequency of abnormal MPI findings exhibited an inverse relationship trend with lower APPHR (P = 0.10) and HRR (P = 0.12). There was a strong inverse association of APPHR, HRR, and PRPP with all-cause (all P ≤ 0.01) and cardiovascular (P < 0.05) mortality. The frequency of coronary revascularization was unrelated to APPHR or HRR. CONCLUSIONS: Arm exercise ECG stress test sensitivity is only reduced at ≤69% APPHR or ≤43% HRR, whereas arm exercise MPI sensitivity and referral for coronary revascularization after arm exercise stress testing are not adversely affected by even a severely blunted peak heart rate. However, both all-cause mortality and cardiovascular mortality are strongly and inversely related to APPHR and HRR.