More than one in three posts by major health and wellness influencers promoted a product or service in a new analysis of Instagram, TikTok and YouTube content. Lifestyle tips were more common than discussion of medical issues.
Pew Research Center examined 535,512 posts published from Feb. 1 through April 30 by 6,334 influencers who address U.S. audiences, according to its new report on wellness posts. The study measured what appeared in the posts, not whether their health claims were accurate or whether a product worked.
Pew classified 38% of the posts as promotions, a category that included disclosed sponsorships and other suggestions to buy a product or subscribe to a service. About 92% of the influencers made at least occasional promotional appeals, and roughly a third did so in at least half their posts. The measure also counted appeals for free products and services, such as workout or nutrition guides; researchers said it may miss subtle or undisclosed sponsorships.
Promotion was especially common in posts about nonprescription medications and supplements, appearing in 66% of that group, and in posts about physical appearance, appearing in 56%. Those figures describe the content Pew analyzed, not the share of all social media health posts or the quality of any promoted product.
Lifestyle advice outweighed medical subjects
Just over half the posts discussed healthy living, while about a third addressed medical issues, Pew found. Fitness appeared in 25% of posts, tips for healthy habits in 20%, nutrition in 19% and physical appearance in 14%.
Among nutrition posts, 42% mentioned protein directly. Medical conditions appeared in 19% of all posts; treatments requiring a prescription or medical professional appeared in 14%; and personal experiences with a condition, treatment or the health care system appeared in 11%.
The categories overlapped. Pew said 45% of posts mentioned multiple topics, and 19% combined a medical topic with healthy-living material. Another 30% of posts did not discuss health or wellness at all, instead covering subjects such as an influencer’s personal life.
Among posts about medical conditions, cancer appeared in 8.5%. Diabetes, anxiety and depression, substance use and attention-deficit/hyperactivity disorder each appeared in at least 5% of that subset. Those are shares of condition-related posts, not of the full sample.
Professional background shaped what appeared
Influencers who described themselves as health care professionals discussed conditions, mental health and supervised treatments more often than those without a stated medical background, according to Pew’s related analysis of influencer backgrounds. Self-described nutritionists focused more on nutrition, while athletes and coaches tended toward fitness.
Influencers whose biographies emphasized personal experience with illness posted about medical conditions at about the same rate as conventional medical professionals. Among the 16% of influencers who listed no particular expertise, medical conditions still appeared in about one in 10 posts.
Pew also found that 11% of the posts explicitly warned of a particular health harm and 9% referred to an outside source while making a health claim. Self-described researchers cited outside sources in 30% of their posts, the highest rate among the professional groups discussed in the report. Pew did not check whether cited sources were valid or whether the claims built on them were true.
The researchers’ definition of a health warning was narrow: it required a direct statement that a practice caused a specific physical or mental harm. Posts discussing symptoms or health advice without such a warning were not counted in that category.
What the sample can and cannot show
The methodology limited the analysis to individual influencers with at least 100,000 followers on one of the three platforms, primarily English-language content, a U.S. audience and posts about health or wellness during the study period. It does not describe smaller accounts, organizations or all content Americans encounter online.
Pew collected captions, audio and available image text, then used automated tools to classify topics and other features. Human reviewers built a 600-post validation set sampled across the three platforms; the classification model achieved an F1 score of at least 0.6 for each label. YouTube audio transcription was limited to the first five minutes of each video, and image-text analysis was not conducted for YouTube posts.
The study describes the supply of content from a defined group of large accounts. It cannot establish what any particular audience saw, whether a medical claim was sound or whether viewing a post changed someone’s health decisions.
FingerLakes1.com is the region’s leading all-digital news publication. The company was founded in 1998 and has been keeping residents informed for more than two decades. Have a lead? Send it to news@fingerlakes1.com.
