– Twitter is a great place for consumers to vent their anger about not getting a hold of the latest holiday Starbucks cup, or to see some cheeky replies from their favorite fast-food chains. It also happens to be a useful place for public health researchers to learn about food security, health disparities, and the social determinants of health, according to a new study from the University of Michigan Medical School.
Specifically, Twitter helped public health researchers learn more about the types of foods certain neighborhoods consumed, individuals’ attitudes around certain types of foods, and the health outcomes they might experience due to diet.
“We wondered whether Twitter-based analysis could help us understand communities better,” V.G. Vinod Vydiswaran, PhD, assistant professor in the Department of Learning Health Sciences at the U-M Medical School, said in a statement.
Using community-based demographic and health status surveys, Vydiswaran and team assessed how different members of regional groups discussed food on social media platforms.
They began by making a list of food-related keywords, types of food, methods for food preparation, and popular restaurants. Each item was scored based on its nutritional value, including its level of trans fats or processed sugars.
Next, the team tapped the Twitter API, which allowed them to sift through over 800,000 food-related Tweets generated within the Detroit area. Tweets came from nearly 62,000 unique users and excluded Tweets from celebrities and marketing accounts.
Using that data, the researchers were able to determine which types of foods or restaurants were popular in certain neighborhoods.
“We found differences in the kinds of foods people tweet about in neighborhoods,” said Vydiswaran, who is also assistant professor of information at the U-M School of Information and who is lead author on the new paper in the Journal of the American Medical Informatics Association.
For example, Starbucks was more popular in affluent neighborhoods, as were vegan food items and other coffee items. In poorer neighborhoods, food terms like pizza, tacos, and bacon were common.
When looking at net healthiness scores, the researchers saw vast health disparities. Less affluent neighborhoods saw higher social media discussion about items that received lower nutrition scores, suggesting that individuals living in — or at least Tweeting from — these neighborhoods favor less healthy food.
This may be because they have easier access to less nutritious food. After all, the analysis also revealed that fast food or chain restaurant density was also tied to more Tweets about less nutritious food keywords.
And this could have health impacts, the study showed. A higher occurrence of food-related Tweeting and Tweets about less nutritious foods were tied to one key indicator of obesity-related mortality.
“In the multivariate regression analysis, accounting for other neighborhood measures such as affluence and percentage African American, our Twitter-based food healthiness measure was still significantly correlated with one obesity-linked condition, heart failure,” Vydiswaran explained.
These results suggest that social media, specifically Twitter, could help public health officials shape policy. Understanding how certain demographics approach their food intake could help policymakers develop new regulations about building fast-food restaurants or offering more nutritious school lunches in specific neighborhoods.
Of course, Twitter is imperfect, the researchers acknowledged. Just because a user Tweets something positive about their fast food dinner from a certain neighborhood does not mean she lives there, which could skew the data. Tweets are not always tied to the food that’s actually consumed, either, the authors said.
Nonetheless, Twitter and social media analysis can serve as another useful tool in a large arsenal for public health strategists.
“We are proposing using social media as an additional source of input for policymakers to understand neighbors,” Vydiswaran concluded. “Social determinants of health, such as neighborhood walkability, mentions of food deserts, safety concerns — these kinds of references show up in tweets and could act as additional input for researchers wanting to understand how they affect health more generally.”
This content was originally published here.