Normally, when doctors are looking for signs of heart disease in patients, they'll look at the usual suspects: smoking history, exercise and dietary habits, obesity, and so on. Certainly not a person's, um, tweets. However, according to a University of Pennsylvania study, the bite-size posts could be the best predictor of heart disease.
The researchers' goal was to find the relationship between the language used on Twitter and instances of atherosclerotic heart disease (AHD) in a given county. By conducting this experiment, they were also testing the widely accepted idea that a community's overall level of anger or satisfaction is a good predictor of health.
For the study, researchers compared 148 geolocated tweets from 1,347 countries to mortality statistics from the Centers for Disease Control. The emotional language of the tweets was measured in two ways: the use of common terms associated with anger, anxiety, and "positive and negative social relationships" and groups of words reflecting certain attitudes and experiences, including hostility and aggression, boredom and fatigue, optimism, and happy memories.
The surprising result was that Twitter is so good a predictor of a country's susceptibility to heart disease, it actually beats out conventional methods — by a rate of 42 to 36%.
By examining Twitter, the researchers didn't just learn about how people use the platform to express themselves — they learned about the community as a whole. "The language of Twitter," the authors write, "may be a window into the aggregated and powerful effects of the community context."
Moral of the story? Don't tweet angry or you'll die young. Tweet happy instead, and you may very well be protecting your life — at least from AHD, that is.
(h/t Pacific Standard)