Twitter predicts rates of heart disease

Smoking, obesity and income levels are standard measures of heart disease rates, but what about Twitter? Yes, those 140-character social media blasts are actually incredibly accurate at predicting whether people get sick or not.

A team of researchers from the University of Pennsylvania discovered this surprising correlation between county-by-county rates of heart disease and language used on Twitter.


As noted by Psychological Science, tweets that express negative emotions – anger, stress, fatigue, etc.  – were linked with higher risk of heart disease in any given county. The opposite was also true; the more positive tweets a county had, the lower its risk of heart disease.

Tweets, it turns out, offer unprecedented insight into the psychological well-being of a community, and therefore whether its citizens stand a greater risk of dying from heart disease.

Researchers were perplexed why Twitter and its typically youthful users could so accurately predict rates of heart disease mortality.


“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 study co-author H. Andrew Schwartz. “But that means if many of your neighbors are angry, you are more likely to die of heart disease.”

The team analyzed 148 million tweets from 1,347 US counties between June 2009 and March 2010, according to Pacific Standard, enough to account for 88 percent of the US population.


Looking at common emotional terms and word clusters reflecting attitudes and topics like hostility, boredom, and fatigue, researchers compared tweets to county-level rates of heart disease mortality. The results were undeniable; even after accounting for factors like education and income levels, negative tweets remained strongly linked to heart disease deaths.Positive tweets were associated with lower rates of heart disease deaths, so perhaps think twice before engaging in your next Twitter rant.