Factors Associated With Cardiorespiratory Fitness at Completion of Cardiac Rehabilitation: Identification of Specific Patient Features Requiring Attention

Abstract

BackgroundWe aimed to determine and compare predictors of postcardiac rehabilitation (CR) cardiorespiratory fitness (CRF), improvements in a large cohort of subjects with varying baseline CRF levels completing CR for ischemic heart disease and to refine prediction models further by baseline CRF. MethodsThe Alberta Provincial Project for Outcomes Assessment in Coronary Heart disease (APPROACH) and TotalCardiology (TotalCardiology, Inc, Calgary, Alberta, Canada) databases were used retrospectively to obtain information on 10,732 (1955 [18.2%] female; mean age 60.4, standard deviation [SD] 10.5 years) subjects who completed the 12-week comprehensive CR program between 1996 and 2016. Peak metabolic equivalents (METs) were determined at program start and completion and identified patients at baseline with low fitness (L-Fit) (< 5 METs), moderate fitness (M-Fit, 5-8 METs), or high fitness (H-Fit, > 8 METs). Multivariable linear regression models were developed to predict METs at completion of the program. ResultsAcross all fitness groups, mean baseline METs was the strongest predictor of CRF at completion of CR. Other factors—including sex, age, current smoking status, obesity, and diabetes—were highly predictive of post-CR CRF (all P < 0.05). Compared with H-fit patients, coronary artery bypass graft and chronic obstructive pulmonary disease in L-Fit patients, and cerebrovascular disease in M-Fit patients had an additional negative effect on the overall model variance in post-CR CRF. ConclusionExpected CRF at the end of CR is highly predictable, with several key patient factors being clear determinants of CRF. Although most identified patient factors are not modifiable, our analysis highlights populations that may require extra attention over the course of CR to attain maximal benefit.