ORIGINAL ARTICLE: PDF OnlyMcConnell Timothy R. PhD; Foster, Carl PhD.†; Conlin, Nancy C. MS†; Thompson, Nancy N. MSJournal of Cardiopulmonary Rehabilitation: July 1991 - p 255-260 Buy Abstract McConnell and Clark (1987) and Foster et al. (1984) have developed equations for predicting functional capacity (VO2peak) from maximal treadmill performance. McConnell investigated the effects of handrail support (HHR). while Foster did not allow HRS. On the basis of the known reduction in the aerobic demands of walking by HRS, a right shift in the treadmill time (TT)/VO2peak carve would be expected when HRS is allowed. Instead, the McConnell equation intersects the Foster equation and predicts higher VO2peak at test durations of less than 7 minutes on the Bruce protocol (a left shift). To resolve this inconsistency, we studied 41 patients during treadmill exercise both with and without HRS. VO2peak (31.0 \pm 12.0 vs 31.9 \pm 11.4 ml/min/kg) and peak heart rate (157 \pm 34 vs 158 \pm 31 beats/min) were not different, although TT was significantly longer with HRS (9.9 \pm 4.1 vs 8.0 \pm 2.9 min). Both submaximal VO2 and heart rate were significantly reduced at matched work loads by HRS. The McConnell equation predicted VO2peak accurately (mean error = 0.2 \pm 5.1 ml/ min/kg) during HRS exercise. The Foster equation tended systematically to underestimate VO2peak during non-HRS exercise (mean error = −4.0 \pm 3.4 ml/min/kg). The mean absolute error in predicting VO2peak was similar for the McConnell (4.2 \pm 2.8 ml/min/kg) and Foster (4.4 \pm 3.4 ml/min/kg) equations. Curve-fitting analysis and reanalysis of the original data form the Foster equation suggested that the unexpected relationship of the McConnell and the Foster equations was due to small sampling errors, and suggested that the limits of accuracy of predicting VO2peak may have been achieved by the existing equations. © Lippincott-Raven Publishers.