No abstract

Objective To investigate the relationship between maximal exercise capacity measured before severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and hospitalization due to coronavirus disease 2019 (COVID-19). Patients and Methods We identified patients (≥18 years) who completed a clinically indicated exercise stress test between 01 January 2016 and 29 February 2020 and had a test for SARS-CoV-2 (i.e., real-time reverse transcriptase polymerase chain reaction test) between 29 February 2020 and 31 May 2020. Maximal exercise capacity was quantified in metabolic equivalents of task (METs). Logistic regression was used to evaluate the likelihood that hospitalization secondary to COVID-19 is related to peak METs, with adjustment for 13 covariates previously identified as associated with higher risk for severe illness from COVID-19. Results We identified 246 patients (age= 59±12 years; 42% male; 75% black race) who had an exercise test and tested positive for SARS-CoV-2. Among these, 89 (36%) were hospitalized. Peak METs were significantly lower (P <.001) among patients who were hospitalized (6.7±2.8) compared to those not hospitalized (8.0±2.4). Peak METs were inversely associated with the likelihood of hospitalization in unadjusted (OR= 0.83, 95% CI= 0.74, 0.92) and adjusted models (OR= 0.87, 95% CI= 0.76, 0.99). Conclusion Maximal exercise capacity is independently and inversely associated with the likelihood of hospitalization due to COVID-19. These data further support the important relationship between cardiorespiratory fitness and health outcomes. Future studies are needed to determine if improving maximal exercise capacity is associated with lower risk of complications due to viral infections, such as COVID-19.

Background Claudication is a common and disabling symptom of peripheral artery disease that can be treated with medication, supervised exercise or stent revascularization. Methods We randomly assigned 111 patients with aortoiliac peripheral artery disease to receive one of three treatments: optimal medical care [OMC], OMC plus supervised exercise [(SE], or OMC plus stent revascularization [ST]. The primary endpoint was the change in peak walking time (PWT) on a graded treadmill test at 6 months as compared with baseline. Secondary endpoints included free-living step activity, quality of life (QOL) using the Walking Impairment Questionnaire (WIQ) and Peripheral Artery Questionnaire (PAQ), and cardiovascular risk factors. Results At six month follow-up, change in PWT (the primary endpoint) was greatest for SE, intermediate for ST, and least with OMC (mean change vs. baseline 5.8±4.6, 3.7±4.9, and 1.2±2.6 minutes, respectively; p<0.001 for the comparison of SE vs. OMC; p=0.02 for ST vs. OMC; and p=0.04 for SE vs. ST). Although disease-specific quality of life as assessed by the WIQ and PAQ also improved with both SE and ST compared with OMC, for most scales the extent of improvement was greater with ST than SE. Free-living step activity increased more with ST than with either SE or OMC alone (114±274 vs. 73±139 vs. −6±109 steps/hour) but these differences were not statistically significant. Conclusions Supervised exercise treatment results in superior treadmill walking performance than stent placement, even for those with aortoiliac PAD. The contrast between better walking performance for SE and better patient-reported QOL for ST warrants further study.

Machine learning is becoming a popular and important approach in the field of medical research. In this study, we investigate the relative performance of various machine learning methods such as Decision Tree, Naïve Bayes, Logistic Regression, Logistic Model Tree and Random Forests for predicting incident diabetes using medical records of cardiorespiratory fitness. In addition, we apply different techniques to uncover potential predictors of diabetes. This FIT project study used data of 32,555 patients who are free of any known coronary artery disease or heart failure who underwent clinician-referred exercise treadmill stress testing at Henry Ford Health Systems between 1991 and 2009 and had a complete 5-year follow-up. At the completion of the fifth year, 5,099 of those patients have developed diabetes. The dataset contained 62 attributes classified into four categories: demographic characteristics, disease history, medication use history, and stress test vital signs. We developed an Ensembling-based predictive model using 13 attributes that were selected based on their clinical importance, Multiple Linear Regression, and Information Gain Ranking methods. The negative effect of the imbalance class of the constructed model was handled by Synthetic Minority Oversampling Technique (SMOTE). The overall performance of the predictive model classifier was improved by the Ensemble machine learning approach using the Vote method with three Decision Trees (Naïve Bayes Tree, Random Forest, and Logistic Model Tree) and achieved high accuracy of prediction (AUC = 0.92). The study shows the potential of ensembling and SMOTE approaches for predicting incident diabetes using cardiorespiratory fitness data.

Background