Estimating maximal oxygen consumption from heart rate response to submaximal ramped treadmill test

The Cambridge Ramped Treadmill Test (CRTT) is an incremental, multistage exercise test (I: steady-state walk, II: walk ramped speed, III: walk ramped incline, and IV: run ramped speed on flat). It is typically deployed as a submaximal test with flexible test termination criteria, making it an attractive option for population-based studies of cardiorespiratory fitness. We conducted a study in healthy adults to test the validity of maximal oxygen consumption estimates (VO2 max; ml O2{middle dot}kg-1{middle dot}min-1) predicted from CRTT heart rate response using several methods: a heart rate-to-work rate linear regression method across several test termination criteria, either when a percentage of age-predicted maximal heart rate was achieved (50% through 100%) or at the end of distinct CRTT stages (II, III, and IV); and two single-point walk-test calibration methods using data from either CRTT stage I (low-point method) or stage II (high-point method). For estimates from the linear regression method, prediction bias ranged from -3.0 to -1.6 ml O2{middle dot}kg-1{middle dot}min-1 and Pearsons r ranged from 0.57 to 0.79 for endpoints at percentages of age-predicted maximal heart rate; results were similar for stages III and IV endpoints, but predictions using data only up to stage II had poorer agreement. Agreement was moderate when using the low-point (mean bias: -4.3 ml O2{middle dot}kg-1{middle dot}min-1; Pearsons r: 0.71) and high-point (mean bias: -3.5 ml O2{middle dot}kg-1{middle dot}min-1; Pearsons r: 0.69) methods. Heart rate response to the CRTT can be used to predict VO2max with acceptable validity in common epidemiological scenarios.