Bad Thoughts

Karma may not only be a bitch, it may kill you- or at least break your heart. An intriguing study published in Psychological Science looked at attitude and verbal expression by using a cross-sectional regression model based only on Twitter language and correlated it to coronary artery disease (CAD).

Attitude, stress, emotional stress and mental state have long been correlated to the risk for CAD, but it is often hard and expensive to accurately quantify and measure, particularly on an individual level. Another recent study demonstrated a reduced risk of CAD in those who were generally happy and optimistic.

This current study examined language patterns via Twitter that reflected negative social relationships, disengagement, and negative emotions—especially anger—as risk factors. The researchers studied public tweets from 2009 through 2010. They then established emotional dictionaries, as well as clusters of words that reflected certain behaviors and attitudes. Utilizing information from users who had made their locations available, they collected tweets and health data from about 1,300 U.S. counties; which contains 88% of the country’s population.

After controlling for variables such as income and education Twitter predicted CAD mortality significantly better than did a traditional model. That traditional model used 10 common demographic, socioeconomic, and health risk factors, including smoking, diabetes, hypertension, and obesity. These same risk factors used by doctors and cardiologists and the focus of much intervention and treatment.

What we think and how we feel impacts what we eat, how we act and what is increasingly apparent; our health. So the next time you feel compelled to let loose a social media rant, remember the Karma effect. And remember Bobby McFerrin, “Don’t Worry, Be Happy.”

Eichstaedt, J. C., Schwartz, H. A., Kern, M. L., Park, G., Labarthe, D. R., Merchant, R. M., . . . Seligman, M. E. (2015). Psychological Language on Twitter Predicts County-Level Heart Disease Mortality. Psychological Science, doi: 10.1177/0956797614557867 .