Can you trust TikTok for nutrition advice? Study exposes #WhatIEatinaDay

A deep dive into how TikTok’s algorithm amplifies engaging but misleading nutrition advice—while dietitians’ evidence-based content gets buried. What does this mean for young people’s health?

Study: #WhatIEatinaDay: The Quality, Accuracy, and Engagement of Nutrition Content on TikTok. Image Credit: AnnaStills / Shutterstock

In a recent study published in the journal Nutrients, researchers at the University of Sydney, Australia, assessed the performance of nutrition-related posts on a well-known social media platform, categorizing them based on engagement, quality, and accuracy.

Their findings indicate that the platform prioritizes high levels of engagement or ‘virality’ over the accuracy of content, leading to concerns about misinformation that can have severe negative effects on the dietary and health behaviors of adolescents and young adults. However, the study found that different levels of accuracy did not result in statistically significant differences in engagement metrics.

Background

Social media is increasingly important as a source of nutrition and health information for younger people. Its popularity among these demographics makes it a critical tool for disseminating health information.

With over one billion active users each month, 63% of users on the platform assessed in this research are aged 10 to 29, and its algorithms prioritize short and engaging video content rather than long-form or curated posts.

The nature of social media posts has increased concerns about misinformation that can spread rapidly if an inaccurate post becomes viral, especially as a significant proportion of posts related to nutrition and health information are not uploaded by experts such as nutritionists or dietitians.

Biased, weight-normative, inaccurate, and other potentially harmful messaging can increase body dissatisfaction and disordered eating habits among young women and adolescents of color, who are more likely to rely on social media for health information. Additionally, low-income populations may face greater exposure to misinformation due to barriers to accessing professional healthcare.

Previous research from other platforms indicates that misleading content influences the food preferences of adolescents, increasing the risk of developing obesity or adopting poor dietary behaviors. Understanding the exposure of vulnerable age groups to content about weight loss or diet culture from non-expert sources is key to tackling misinformation on social media.

About the study

In this study, the research team examined the engagement metrics, prevalence, and characteristics of nutritional posts uploaded on the social media platform from September 2023 to March 2024.

Relevant posts were identified using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) protocol. A pilot study was used to identify hashtags, and the posts were accessed through new accounts to reduce bias due to the algorithm. The screening process identified 1054 relevant posts, of which 250 were randomly selected for analysis.

The research team collected information about engagement by examining how many times a post was liked, commented on, shared, or saved, as well as the number of subscribers the associated content creators had. Details about the post’s nutritional relevance, language, and accessibility were also collected. The posts were grouped into nine groups based on their nutritional focus and into eight groups based on the type of content creator (with one category, “content farms,” identified as producing large volumes of low-quality content that exploit engagement algorithms).

The quality of the posts was assessed based on criteria such as the level of engagement (to assess misinformation spread potential), transparency (reliability and clarity of information), financial disclosure (whether sponsorships, promotional intent, or affiliations were disclosed), accuracy (defined as whether a post was aligned with evidence-based dietary guidelines such as the Australian Dietary Guidelines), and authorship (disclosure of the creator’s credentials).

Posts were classified as ‘completely accurate’ to ‘completely inaccurate’ based on accuracy and graded A-D, where A indicated the strongest alignment with evidence. The dataset was then analyzed to generate descriptive statistics and compare inaccurate and accurate posts based on engagement.

Quality of nutrition-related TikTok posts as defined by the Social Media Evaluation Checklist

Findings

Health and wellness influencers accounted for 32% of the posts, fitness-related content creators posted 18%, and lifestyle content creators or other creators not focused on health or nutrition accounted for 18%. In contrast, nutritionists, dietitians, or other health experts were represented in only 5% of posts for dietitians and 4% for nutritionists, making expert-driven content a minority.

Regarding content, approximately 34% of the posts were about weight loss, meal ideas, or recipes. In 32% of the videos, influencers shared “What I Eat in a Day” posts, which the study identified as a major source of weight-normative messaging and misinformation. Nutritional supplements were discussed in 10% of posts, while nutrition to meet specific goals was covered in 7%.

In nearly half of applicable posts, the creator used testimonials to promote their own business or product. Only 18% included transparent advertising, and only 13% identified and disclosed sources of conflict of interest. Approximately 63% of posts were found to promote stereotypical attitudes, and 55% did not provide evidence-based information. Only 10% of posts highlighted potential risks associated with the behaviors and products they promoted despite the potential for harm.

The research team found that 36% of the posts were completely accurate, with 29% being mostly accurate. In contrast, 19% of posts contained utterly inaccurate information, and 18% were mostly inaccurate. Only 12% of posts were given a grade of A, 12% a grade of B, 20% a grade of C, and 15% a grade of D; 41% could not be assessed. Dietitians produced the most accurate posts, with 42% of their content classified as entirely accurate.

Notably, mostly accurate and grade-A posts had the highest engagement on average, but the differences in engagement metrics between accurate and inaccurate posts were not statistically significant. This suggests that while credibility matters, TikTok’s algorithm does not necessarily favor accurate content over misleading content.

Distribution of nutrition topics in nutrition-related TikTok posts by levels of (a) accuracy and (b) evidence.

Conclusions

Social media platforms, including the one analyzed in this study, are dominated primarily by non-expert content creators. While popular, these creators may share information lacking in scientific credibility.

The findings highlight an urgent need for stricter moderation of nutrition misinformation, particularly for posts that promote extreme dieting, unverified weight loss claims, or products with undisclosed sponsorships.

Posts from qualified professionals, particularly those on weight loss, need to be given more attention to minimize harm to vulnerable groups such as young adults.

Concerningly, “What I Eat in a Day” videos were highlighted as a key driver of misinformation, reinforcing unrealistic dietary standards. This, along with the platform’s lack of regulation over misleading content, suggests a need for more proactive interventions.

The study calls for policies that encourage platforms to verify health content creators’ credentials and promote evidence-based nutrition guidance over viral but misleading content. Effort is also needed to educate users about misinformation on social media to minimize the effects of exposure to misleading content.