The authors say that fitness checks should be joined with other more commonly measured risk factors, given their strong predictive value.
A new artificial intelligence (AI) study shows that cardiac and respiratory fitness and age are the two most important predictors of all-cause mortality in coronary patients.
“The models need to be simple if we are to apply them in clinical practice,” lead researcher Cristina de Souza e Silva, MD (CLINIMEX, Rio de Janeiro, Brazil) told TCTMD. “If we build a model that is too complex, no one will use it. It happens with a lot of different models that we have today. We were able to create a model with aerobic fitness and age that has good accuracy for predicting mortality. Adding one or the other variable doesn’t seem to add a lot of information.”
De Souza e Silva said that while predictive models to identify high-risk patients are frequently used for heart disease, cardiorespiratory fitness is often not one of the studied variables. This is despite the fact that the American Heart Association considers cardiorespiratory fitness a Clinical vital sign And that since 2002, Jonathan Myers, Ph.D. (Stanford University, California), lead author of the AI study, has shown that the ability to exercise has been a The strongest predictor of mortality Even when compared to other cardiovascular risk factors.
Machine learning and black boxes
Speaking with TCTMD, de Souza e Silva said that the goal of the current study is to develop a predictive model for survival in CAD patients using variables that bypass traditional risk factors. To do this, the researchers turned to a Canadian database of patients undergoing cardiac rehabilitation after coronary angiography. As part of rehabilitation, all patients had a complete baseline clinical evaluation, including a symptom-limited graded exercise test, often performed on a treadmill, to determine maximal exercise capacity.
From the data set, there were 260 variants identified from 23,215 patients. These variables were eventually narrowed down to 25 of the more than 13,000 patients and their predictive significance for all-cause mortality was assessed. Importantly, the researchers avoided a “black box” approach with machine learning, a model created through input data that is beyond human understanding.
“You have a number of different technologies with machine learning,” de Souza e Silva said. “With a black box, you put everything in it and then you get a result. The problem is that you don’t know how the algorithm got there. There is a lot of criticism about the black box. [approach] and how it is used, particularly in medicine.”
For a variety of different reasons, I don’t think clinicians always value aerobic fitness and don’t really know how to measure it in clinical practice. Cristina de Souza e Silva, MD, PhD
Next, the researchers developed a “life tree” machine learning model which allowed them to build a predictive model by adding different variables. In prognostic modeling, the largest increase in C-index (0.729) was observed when maximum exercise capacity and age were added. There were increases when gender, smoking status, presence of malignant diseases, and chronic heart failure were added to the model, but these increases were small.
Given the findings, as well as previous research showing the importance of aerobic fitness for predicting future outcomes, de Souza e Silva said clinicians need to take exercise capacity seriously in the clinic. “Just as we measure blood pressure, blood glucose, cholesterol, or height and weight, we should also assess aerobic fitness,” she said. “But for a variety of different reasons, I don’t think clinicians always value aerobic fitness and don’t really know how to measure it in clinical practice.”
She added that this should not be a barrier, as validated questionnaires are available to help clinicians assess aerobic fitness in the clinic. “In a few minutes, you can estimate their ability to exercise.”
Claudio Gil Araujo, MD, PhD (CLINIMEX), a fellow at de Souza e Silva but not involved in the AI study, said the new paper adds to existing research using new and advanced machine learning techniques for confirmation. And the assertion that aerobic fitness “is a very strong marker for clinical diagnosis.”
He stressed that given that cardiorespiratory fitness is an important variable for predicting all-cause mortality in coronary heart disease patients, modifying exercise habits in these patients should be a priority. else Search this week Given the effect of exercise on coronary heart patients, he emphasized the importance of being physically active, or maintaining current exercise habits, even after a diagnosis of MI.
