Abstract MP15: Developing and Validating VO 2 Prediction from 6-Minute Walk Test (6MWT) and Association with Cardiovascular Health and Risk Scores: Coronary Artery Risk Development in Young Adults (CARDIA)

Introduction: Cardiorespiratory fitness (CRF) is a valuable diagnostic and prognostic measure. Yet, assessing VO 2 max (indicator of CRF) in population-based studies can be challenging due to requiring sophisticated equipment and trained personnel. Thus, we built and validated a model to estimate VO 2 max from a 6MWT, which is less burdensome for participants (submaximal effort). We tested the hypothesis that our model provides valid estimates of VO 2 max that are associated with cardiovascular health - AHA’s Life’s Simple 7 (LS7) and ASCVD scores. Methods: Data are from 576 CARDIA participants [60% female; 40% Black; mean ± SD age = 61.5 ± 3.6; BMI = 28.6 ± 5.7 kg/m 2 ] who attended the Year 35 follow-up exam (2020-22) and completed VO 2 max and 6MWT tests. The study sample was split randomly into a training (2/3) or test (1/3) set. The prediction equation was developed using stepwise multiple linear regression models in the training set. We evaluated accuracy of the model using student's t-test and Pearson correlation coefficients (r) by examining the difference between the observed and predicted VO 2 max in the test dataset. Linear regression models adjusted for age, and sex were used to examine associations of VO 2 max estimates with cardiovascular health and risk indices. Results: The prediction equation included 6MWT distance, sex, age, BMI, physical activity status, and smoking status (R 2 adj = 0.55, RMSE = 4.84). No significant difference was observed between predicted and measured VO 2 max (26.2 ± 5.4 vs 26.0 ± 7.5 mL/kg/min; p =0.76) and the measures were strongly correlated (r=0.80, p <.001) with Bland-Altman plot is presented (Figure). Predicted VO 2 max explained more variance in LS7 score (R 2 adj = 0.55) compared to measured (R 2 adj = 0.37), while explaining similar variance in ASCVD risk (R 2 adj = 0.40 vs 0.38). Conclusions: This prediction equation provides a convenient and useful estimate of CRF estimate of CRF from 6MWT and may provide utility in clinical practice to screen and monitor CRF when direct measurement is not feasible.