Do Celebs Freely Shill Unhealthy Eating on Social Media?

Celebrities commonly endorse unhealthy foods in traditional advertising, but they're also promoting these foods -- mostly without compensation -- on Instagram, researchers reported.

An analysis of 181 highly-followed celebrity Instagram accounts, based on the British Nutrient Profile Index (NPI), found that 87% earned a food score and nearly 90% a beverage score so low they would fail legal youth advertising limits in the U.K., according to Bradley Turnwald, PhD, of the University of Chicago, and colleagues.

In addition, only a small fraction of posts (4.8%) were sponsored by food and beverage companies, they reported in JAMA Network Open.

Turnwald's group also noted that posts that featured foods with healthier nutrition scores were associated with significantly fewer likes (b -0.003, 95% CI -0.006 to 0.000, P=0.04) and comments (b -0.006, 95% CI -0.009 to -0.003, P<0.001) from followers. However, this trend was not observed for posts that featured beverages.

"These results suggest that influential depictions of consumption of unhealthy foods and beverages on social media are a sociocultural problem that extends beyond advertisements and sponsorships," they wrote.

"Celebrities are, of course, entitled to post foods and beverages as they wish on their personal social media," the authors added. "They themselves are individuals existing in societies that value and normalize unhealthy consumption, and it is possible that social media posts by the general public are similarly unhealthy."

The observed association between unhealthy food posts and increased follower engagement suggests a potential incentive to post about unhealthy foods, they explained. "However, given celebrities' broad following, there is potential to shape their followers' perceptions that healthy eating is normative and valued if celebrities commit to posting a healthier profile of foods and beverages. It is also important for followers to remember that social media likely represents a curated, incomplete window into what celebrities actually consume," they said.

The cross-sectional study included the Instagram accounts of popular athletes, musicians, actors, and television personalities. Slightly more than half (56%) were male and 44% were female. Their median age was 32. At the time data was collected, April 2012 to March 2020, the 181 celebrities had 5.7 billion total followers.

For each account, the researchers identified up to 30 of the most recent posts depicting a food or beverage. A total of 3,065 posts containing 5,180 foods and beverages were included in the study.

The researchers graded the healthfulness of each food and beverage depicted using NPI, an established tool that uses the macronutrient and micronutrient content of foods and beverages to generate a grade from 0 (least healthy) to 100 (most healthy). Foods with scores less than 64 and beverages with scores less than 70 are rated as "less healthy," and there are limits to advertising these to young people in the U.K., they explained.

The researchers used mixed-effects regression models to generate an overall NPI score for each celebrity account.

In an accompanying editorial, Ellen Selkie, MD, of the University of Wisconsin School of Medicine and Public Health in Madison, noted that "Although the finding that almost 90% of celebrity social media accounts had posts of foods and beverages with average scores in the less healthy range is striking and may be alarming to readers, there are some caveats."

"The median food NPI score was in the low 50s, and the median beverage NPI score was between 68 and 69. This suggests that even the celebrity profiles that were less healthy overall likely had some healthy content," Selkie explained. In addition, "Only 60.5% of posted foods and 54.8% of posted beverages had less healthy scores. Understanding the data at the level of posts adds depth to the study's findings."

Turnwald's group noted other study limitations including the fact that it did not include Instagram videos. Also, the study did not report racial or ethnic information on the celebrities, but whites were likely overrepresented, they said.

"In addition, we chose to focus on traditional celebrities rather than social media influencers because celebrities compose a majority of the most followed social media accounts, and doing so allowed for comparison of this study's findings with the literature on celebrity endorsements," the authors explained. "Additional research is needed to understand whether celebrities and social media influencers differentially affect followers' health behaviors. We did not explore effects on followers' eating and drinking behavior."

Comment