Twitter Can Predict Regional Heart Disease Risk
Feeling #angry, #stressed, or #tired? Such negative tweets within a given region can indicate increased heart disease risk, according to University of Pennsylvania Medical School researchers.
Feeling #angry, #stressed, or #tired? Such negative tweets within a given region can indicate increased heart disease risk, according to University of Pennsylvania Medical School (@PennMedicine) researchers.
“Psychological states have long been thought to have an effect on coronary heart disease,” said study co-author Margaret Kern, associate professor at the University of Melbourne in Australia, in a press release. “For example, hostility and depression have been linked with heart disease at the individual level through biological effects. But negative emotions can also trigger behavioral and social responses; you are also more likely to drink, eat poorly, and be isolated from other people, which can indirectly lead to heart disease.”
Negative emotional language and topics strongly correlated with heart disease mortality—a relationship that remained after the researchers accounted for income, education, and other variables. Positive language had the opposite correlation, suggesting that optimism and other positive experiences may protect against heart disease.
“The relationship between language and mortality is particularly surprising, since the people tweeting angry words and topics are in general not the ones dying of heart disease,” said H. Andrew Schwartz, visiting assistant professor in the University of Pennsylvania’s School of Engineering and Applied Sciences department, in a press release. “But that means if many of your neighbors are angry, you are more likely to die of heart disease.”
Tweets can reveal more information about heart disease risk than many traditional factors because they capture the community’s psychological atmosphere, researchers said.
Researchers analyzed a random sampling of public tweets containing location data and studied established emotional dictionaries and automatically generated word clusters reflecting behaviors and attitudes. Their sample included enough tweets and health data for about 1300 counties, which is approximately 88% of the country’s population.
The study supports existing research regarding combined community characteristics and physical healthy predictions. Despite several caveats, the results could provide insights into the effectiveness of community-level public health interventions.
“Twitter seems to capture a lot of the same information that you get from health and demographic indicators, but it also adds something extra,” said Gregory Park, a postdoctoral fellow in the University of Pennsylvania’s School of Arts and Science’s Department of Psychology, in a press release. “So predictions from Twitter can actually be more accurate than using a set of traditional variables.”