Preliminary research has found that drinking coffee may have a hand in reducing risks of heart failure or stroke.
Presented at the 2017 Scientific Sessions conference of the American Heart Association, the research used machine learning for data analysis using information from the Framingham Heart Study, which includes the cardiovascular health of participants and the food they eat. Based on results, coffee drinkers had 7 and 8 percent lower risks of developing heart failure and suffering a stroke respectively, for every extra cup of coffee they had every week.
The research was also published in the Journal of the American Heart Association.
Validating Machine Learning
To ensure result validity and identify risk direction, the researchers carried out a traditional analysis involving two studies with similar data sets: the Atherosclerosis Risk in Communities Study and the Cardiovascular Health Study. They discovered that the results produced via machine learning were consistent with their findings for the two other studies.
"Our findings suggest that machine learning could help us identify additional factors to improve existing risk assessment models," said Laura M. Stevens, B.S., the study's first author, pointing out that existing methods of assessment are excellent but not 100% accurate.
Many risk factors associated with heart failure and stroke are known but the researchers believe that there is lot more that have yet to be identified.
Other Risk Factors For Heart Disease, Stroke
Aside from coffee drinking, machine learning analysis also identified red meat consumption as a potential risk factor, although, the relationship between eating red meat and heart failure or stroke is not as clear. In the Framingham Heart Study, red meat consumption was associated with lower risks of heart failure and stroke but validating those results in similar studies have been challenging because there is not one definition of what red meat is.
The American Heart Association, however, suggests limiting eating red meat because of its high levels of saturated fat, recommending a diet rich in vegetables, fruits, whole grains, fish, poultry, and low-fat dairy products.
In the meantime, the researchers came up as well with a predictive model utilizing known risk factors gathered from the Framingham Risk Score. By adding coffee to risk factors such as age, blood pressure, and other cardiovascular disease-associated patient characteristics, the model's prediction accuracy was bumped up to 4 percent.
The research received funding support from the University of Colorado School of Medicine and the American Heart Association. Other authors include David Kao and Carsten Görg.