The outpouring of daily gripes on Twitter could be used to predict heart disease rates, according to research published in Psychological Science this January. The team, led by scientists from the US and Australia looked at tweets containing negative language, and compared this to known rates of heart disease across 1,347 counties in the north east of the US.
“Hostility and chronic stress are known risk factors for heart disease, but they are costly to assess on a large scale,” explain the researchers in their latest study. So they turned to Twitter. The challenge facing many scientists now is how to collect accurate information, and analyse it effectively, when there is so much data available. Trying to predict heart disease rates accurately costs a lot of money and time. This study shows that incorporating big data from sources like Twitter into science can cheaply and quickly provide accurate information that otherwise would be expensive to collect.
Analysing 148 million tweets, the researchers suggest that the wealth of information that can be collected in this way can help policymakers more effectively target public health campaigns in the communities that would benefit most.
The scientists point out that there is a discrepancy in age between the Twitter users and people who are suffering heart diseases. The average age of a Twitter user is about 35, compared to people with heart disease most commonly aged over 60. The researchers added; “it is not obvious why Twitter language should track heart disease mortality. The people tweeting are not the people dying.” Instead this suggests that the information gathered from Twitter paints a picture of the community’s overall environment. Twitter users talk about the overall state of their neighbourhood, describing the quality of health care, social and economic prospects in their community. It’s these factors that contribute to heart disease rates.
The researchers suggest that negative tweets could reflect happiness, health and wellbeing in a community in general. Areas with poor public health are likely also to have poor prospects in other areas, like employment, quality of housing and social support, which all contribute to the health of citizens. The language used in tweets can tell you how happy people are with the outlook in that area, and even the health of the parents and grandparents of Twitter users.
Image: Researchers’ own
The Untweetable Truth 
One thought on “Can Twitter be used to predict heart disease?”