Errors in predicting maximal oxygen consumption in pregnant women.

This study was designed to determine the accuracy of estimated values of maximal heart rate (HRmax) and oxygen consumption (VO2) during pregnancy. We measured HR and maximal VO2 (VO2max) at rest and during cycle (CE) and treadmill exercise (TE) tests with rapidly increasing exercise intensities during gestation and after delivery. Pregnancy was found to affect the linear relationship of HR and %VO2max so that the intercept increases with advancing gestation and the slope decreases. Estimated maximal HR (HRmax, est), 220 - age (yr) x beats/min, overestimated measured HRmax by 8% (CE) and 5% (TE). For VO2max estimated by Astrand's nomogram (VO2max, est1) and by linear extrapolation of submaximal values of HR and VO2 to HRmax, est (VO2max, est2), individual errors were large (SD 17-28%). Mean VO2max, est1 overestimated measured VO2max by 20% during CE but not during TE (-2%) and elicited the erroneous impression that VO2max decreases during CE in pregnancy. Mean VO2max, est2 values were not significantly different from measured VO2max values. This apparent accuracy resulted from two opposing errors: 1) HRmax, est overestimated HRmax, and 2) above 70% VO2max the slope of the HR-%VO2max relationship was significantly reduced. Therefore neither method to estimate VO2max can replace the measurement of VO2max.