This is what Johannes C. Eichstaed, from the University of Pennsylvania and a team of collaborators who have carried out research on the language used on Twitter and how this can predict mortality due to heart disease , specifically of the atherosclerotic type, have shown (by atherosclerotic plaques on the walls of blood vessels).
The results of their study have recently been published in the journal Psychological Science (2015) and highlight the possibility of using social networks in disease prevention and health care.
In Psychology, the importance of certain factors in cardiovascular diseases was described a long time ago. Among these psychological factors we find emotions such as hostility and stress , two risk factors that have been widely studied in this discipline. As the authors describe, in social networks like Twitter we express ourselves through language, and through it we can show negative emotions such as anger or a bad social adjustment. In their work, the researchers analyzed a total of 148 million localized tweets from 1,347 counties, searching for keywords and sequences of two or three words, including negative ones such as’ hate ‘or’ jealous’ and positive ones such as’ friends. ‘, ‘great’.
The results showed that those language patterns that reflected more anger and worse social relationships were risk factors for cardiovascular disease and mortality, regardless of age. On the other hand, those tweet patterns that reflected positive emotions and psychological adjustment turned out to be protective factors against atherosclerotic-type heart diseases. Importantly, the authors highlight that all relationships remained statistically significant after controlling for other variables such as educational level and income.
Finally, statistical analyzes showed that the language used on Twitter predicted this type of cardiovascular disease better independently than when other health risk factors such as diabetes, hypertension and other socioeconomic factors were included.
It is interesting to see how certain important health problems such as cardiovascular ones can be predicted through something as frequent and simple as our tweets. Perhaps studies like this one that use what we know as Big Data , encourage health institutions and companies to review what type of language we use to find out if we show language patterns that may pose a risk to the population in the medium and long term. Of course, what does seem clear is that social networks serve much more than we initially believed.
Marisa Fernández is a Senior Neuropsychologist at Unobrain
