A Nonexercise Prediction Equation For Cardiorespiratory Fitness Without The Use Of Physical Activity

Abstract

Low cardiorespiratory fitness (CRF) is an independent predictor of morbidity and mortality. The majority of healthcare settings use some type of electronic health record system (EHRs). However, many EHRs do not have CRF data collected, thereby limiting the types of investigations and analyses that can be done for research. PURPOSE: To develop a nonexercise equation to estimate and classify CRF (METs) using variables commonly found in EHRs. METHODS: Participants were 41,861 apparently healthy adults (21.4% women) from the Aerobics Center Longitudinal Study examined from 1974 to 2005. Estimated CRF was based on sex, age, measured body mass index, measured resting heart rate, and smoking status. Actual CRF was measured by a maximal treadmill test. RESULTS: After nonlinear feature augmentation was conducted, separate linear regression models were used for male and female patients to calculate Pearson’s correlation and regression coefficients.Cross-classification of actual and estimated CRF was conducted using the lowest 20th percentile as the low-fit category. Correlation coefficients were 0.68 (MD 1.33) and 0.63 (MD 1.23) for men and women respectively. The models explained 46% (SEE 1.69) and 40% (SEE 1.54) variance in CRF for men and women respectively. Correct category classification was found in 84% of men and 80% of women. CONCLUSION: The regression models developed in the present study provided useful estimation and classification of CRF in a large population of men and women. The models may provide a valid method for conducting investigations using CRF data derived from EHRs. Supported by JSPS KAKENHI Grant 19K19437