Mapping obesity from Facebook

Tell me what you like* on Facebook and I’ll tell you who you are. Or, at least, what you usually eat. A team of researchers from Boston Children’s Hospital reached that conclusion. If the percentage of people who like pages related to active and healthy lifestyles is high, then the obesity rate in that community is lower. Likewise, if they like pages related, for example, to television, then it is likely that this group of people tends to be obese. These researchers assure that by comparing the information of different Facebook users through geotagging - a process that fixes geographic information as metadata to different file formats - with information from various health studies in the United States, it is possible to help predict, track and map obesity rates. Even at a level as small as neighborhoods or colonies. Thus offering a way to intervene online to combat such a pressing problem. This is possible thanks to the enormous amount of information available on social networks. In addition, it is much cheaper and more easily accessible than the usual methods. The correlation between Facebook users’ interests and obesity rates in their communities is statistically significant. This analysis can be very useful for estimating obesity levels in different areas and in real time. Apart from directing public health campaigns that promote healthier behaviors.

Methodology

What the scientists did was acquire information from people’s posts, likes, and shares. Particularly in New York. Then they compared the percentage of people who move and lead a healthy life (or at least that’s what their timeline indicates) and of people who spend their time glued to the television (come on, if according to their Facebook profile they like like 100 different shows) with various health surveys. Very close results were obtained in both cases. This confirms the geographical relationship between Facebook activity and the obesity rate. The importance of this study is not limited to facilitating the improvement of public health services. It is a very effective way to obtain information at the population level that would be very difficult to obtain in any other way.