While Twitter is known as an easy, direct and immediate way for people to share what's on their minds, researchers from the University of Pennsylvania think it can also be used to indicate a community's psychological well being and can predict rates of heart disease. The study was published in the journal Psychological Science.
Among the factors that contribute to the risk of heart disease is stress. The researchers showed how Twitter can capture more information about heart disease risk than many traditional factors combined, because it characterizes the psychological atmosphere of a community. Expressions of negative emotions such as anger, stress and fatigue in a county's tweets were associated with higher heart disease risk. Positive emotions like excitement and optimism were associated with lower risk.
Although researchers figured that the psychological well being of communities is important for physical health, it is hard to measure. Using Twitter to gauge a community's collective mental state could be useful tool for epidemiology and measuring the effectiveness of public-health interventions.
Because billions of users are writing about their daily experiences, thoughts and feelings, the world of social media represents a new area for psychological research. Such information could be an invaluable public health barometer if it can be tied to real-world outcomes. With this in mind, the researchers from the World Well-Being Project have tried to determine how well the language people use online correlates with their inner thoughts and feelings.
The team tired to glean information from the words people use when speaking or writing. Linguistic analysis can work as effectively as traditional questionnaires in analyzing an individual's personality.
Having determined correlations between language and emotional states, the researchers wanted to show connections between those emotional states and physical outcomes inherent in them. Coronary heart disease, the leading cause of death worldwide, was deemed to be an ideal candidate.
As a common cause of early death, public health officials carefully count when heart disease is identified as the underlying cause on death certificates. They also collect data about potential risk factors, such as smoking, obesity, hypertension and lack of exercise. Researchers can find this data on a county-by-county level in the United States, so they tried to align this physical epidemiology with their digital Twitter version.
Using public tweets from between 2009 and 2010, the researchers established emotional dictionaries, as well as automatically generated clusters of words reflecting behaviors and attitudes, to analyze a sample of tweets from various individuals who had listed their locations. There were sufficient tweets and health data from about 1,300 counties, representing 88 percent of the country's population.
Researchers determined that negative emotional language and topics, such as words like "hate" or expletives, remained strongly correlated with heart disease mortality, even after variables like income and education were factored into the equation. Positive emotional language appeared to do just the opposite, implying that optimism and positive experiences, words like "wonderful" or "friends," may protect against heart disease, but when many of one's neighbors are angry, that person is more likely to die of heart disease.
Thus, the combined characteristics of communities can predict physical health more effectively than the reports of any one individual. Still, researchers acknowledged that they needed to use caution in determining the social factors that influence what kinds of messages people choose to share on Twitter. If the method proves to be effective, it could be used to determine how well public-health interventions work on the community level, rather than on an individual level. The researchers believe that these tweets are amassing information about people that cannot be readily accessed in other ways.
