[Paper Review] #FoodPorn: Obesity Patterns in Culinary Interactions

Yelena Mejova, Hamed Haddadi+2 more|arXiv (Cornell University)|Mar 5, 2015

Culinary Culture and Tourism15 references17 citations

TL;DR

This study analyzes 164,753 U.S. restaurant visits from Instagram and Foursquare to examine links between food environments, dietary perceptions, and social media engagement. It finds a stronger correlation (0.424) between fast food prevalence and county-level obesity than official data, reveals that local restaurants receive more attention in low-obesity areas, and shows social approval on social media disproportionately favors unhealthy, high-sugar foods like donuts despite users associating #foodporn with healthier cuisines.

ABSTRACT

We present a large-scale analysis of Instagram pictures taken at 164,753 restaurants by millions of users. Motivated by the obesity epidemic in the United States, our aim is three-fold: (i) to assess the relationship between fast food and chain restaurants and obesity, (ii) to better understand people's thoughts on and perceptions of their daily dining experiences, and (iii) to reveal the nature of social reinforcement and approval in the context of dietary health on social media. When we correlate the prominence of fast food restaurants in US counties with obesity, we find the Foursquare data to show a greater correlation at 0.424 than official survey data from the County Health Rankings would show. Our analysis further reveals a relationship between small businesses and local foods with better dietary health, with such restaurants getting more attention in areas of lower obesity. However, even in such areas, social approval favors the unhealthy foods high in sugar, with donut shops producing the most liked photos. Thus, the dietary landscape our study reveals is a complex ecosystem, with fast food playing a role alongside social interactions and personal perceptions, which often may be at odds.

Motivation & Objective

• To investigate the relationship between fast food and chain restaurants and obesity rates at the county level in the U.S.
• To understand individuals’ perceptions and emotional associations with their dining experiences through social media content.
• To analyze social reinforcement patterns—such as likes and comments—on food-related posts to assess how social approval influences dietary behavior.
• To compare social media data with official health statistics and evaluate the predictive power of online behavioral data for obesity trends.

Proposed method

• Collected and analyzed Instagram and Foursquare data from 164,753 U.S. restaurants across all 3,143 counties.
• Correlated the share of fast food and chain restaurants per county with official obesity rates from the County Health Rankings (CHR) and Foursquare data.
• Used hashtag frequency (e.g., #smallbiz, #eatlocal, #foodporn) as a proxy for user perception and dietary preferences.
• Measured social approval via likes and comments on Instagram posts to assess platform-level reinforcement of food types.
• Built a linear regression model to predict county obesity rates using restaurant type, social approval, and hashtag features, though R² was low (0.013).
• Employed content analysis of user tags and image metadata to infer food type, health perception, and social behavior.

Experimental results

Research questions

• RQ1 What is the correlation between the prevalence of fast food and chain restaurants in U.S. counties and local obesity rates, and how does it compare to official survey data?
• RQ2 How do users perceive and tag their dining experiences, particularly in relation to healthy versus unhealthy food choices?
• RQ3 To what extent is social approval on Instagram—measured by likes and comments—linked to the healthiness of the food being shared?
• RQ4 How do local and chain restaurants differ in terms of user engagement and social media visibility across regions with varying obesity levels?

Key findings

• The correlation between fast food restaurant prevalence and obesity in U.S. counties was 0.424 using Foursquare data, exceeding the correlation found in official County Health Rankings data.
• Users in low-obesity counties used hashtags like #smallbiz and #eatlocal significantly more often than in high-obesity counties, indicating stronger local food engagement in healthier regions.
• Despite perceiving fast food as unhealthy, users tagged pictures from fast food venues with twice as many unhealthy-related tags compared to other restaurants.
• Social approval on Instagram was highest for donut shops, cupcake shops, and burger places—foods high in sugar and fat—indicating strong platform-level reinforcement of unhealthy choices.
• The linear regression model using only picture metadata (restaurant type, likes, comments, hashtags) had a very low R² of 0.013, indicating that such data alone is insufficient to predict county-level obesity without deeper demographic context.
• Local restaurants received more user engagement (likes, comments, unique users) than chains, and were more frequently associated with health-conscious hashtags in low-obesity areas.