California Is Healthier Because of Tomatoes, Says Twitter

One of the effects of social media is the ability to gather vast amounts of information in a very short period of time. (As opposed to standing with a clipboard on the side of the street and flagging folks down.) This changes the game of marketing surveys. It also changes the game of scientific research.

Using information offered up by social media -- specifically, from Twitter -- in order to examine our eating habits is one of the ideas behind the fascinating project The Lexicocalorimeter, from the University of Vermont's CompStoryLab.

The project analyzed 49 different states and examined two things. (No Hawaii, sorry.) It looked for mentions of foods and physical activities on Twitter. Researchers put a value on the state's "caloric input" and "caloric output," and crunched those two numbers to get a "caloric balance."

"We see the instrument as more of a real-time diagnostic tool, complementary to traditional methods such as food and activity diaries," wrote Peter Dodds, one of the study authors, in an email. "It's very hard to gain a lot of information on a population's health-related behaviors and our instrument gives a fast way to gain a real-time estimate."

You can take a gander at the study if you want to see the other 48 states in the study, but for our purposes, let's focus on California.

As far as calorie expenditure goes, California ranks 10th out of 49, with the most popular activities being, in order, dancing, running, and hiking. This all checks out on an anecdotal level, but the "caloric input," the food-based portion of the proceedings, is where things get interesting.

California consumes an average of 267.32 calories "per food phrase caloric shift." This is a tad less than the average American (267.92 calories), and the state ranks 22nd out of 49 states. The number one food mentioned? Donuts. After that, it was noodles and bacon. After that, it was chocolate, cookies, and peanut butter. Right about now you're probably wondering how California ended up not physically breaking off and falling into the ocean. (Or, at the very least, ranking dead last in the study.)

The reason has largely to do with the next item on the list: Tomatoes. "That's helping drive California's score down by being eaten more frequently," wrote Dodds.

Among the other foods Californians are eating that leads to the relatively modest ranking in caloric input? Mushrooms, strawberries, frozen yogurt, pasta, and avocados. By being huge consumers of these foods, California is partially countering its terrible junk food obsession.

That leads to the question: Just what are we supposed to do with this information?

"We can connect food and activity with health problems (e.g., obesity and diabetes) and estimate how these are changing over time," wrote Dodds. "The output from our instrument should be useful for public policy makers and for the general populace. And over time, we'll be able to see how a population's tweets about food and activity change, and then be able to ask what's going on (e.g., if people or marketers are talking more about unhealthy food)."

And although the current model is admittedly simplistic (and there are also glitches in the program: ice cream comes up as a low calorie food, for instance), this is the first time that a team is trying to accomplish something like this and they're contining to tweak the formula and design of the program. Besides, it offers policymakers and consumers alike the ability to access a massive store of information.

At the very least, it reminds us to eat a few more tomatoes with dinner.