Health misinformation by social media influencers: Encouraging critical reflection

Research | Media | Publications

In a recent study at the University of Vienna, communication scientists Jaroslava Kaňková and Jörg Matthes explore the effects of two types of misleading communication by social media influencers, misinformation and overgeneralized messaging, and test an intervention designed to help critical evaluation of this content.

Polina Zimmerman

Through a mixed-design online experiment with Gen Z participants (18-25 years old) using realistic TikTok videos, the study found that misinformation and overgeneralized content undermined the accuracy of health beliefs, message credibility, source trustworthiness, and parasocial relationships. The intervention largely reduced susceptibility to misinformation, but the effect declined over time. For overgeneralized messaging, the intervention's effects were minimal and short-lived.
 

 

by Aline Vianne Barr


For Gen Z, social media influencers (SMIs) are a valid and trusted source of health information. SMIs shape online opinions and attitudes toward a variety of topics, gaining trust through relatability and perceived credibility. They are also identified as a growing source of health misinformation, sharing content that often contradicts scientific consensus. Beyond misinformation, SMIs have been found to participate in overgeneralized messaging – a more indirect form of misleading health communication. Overgeneralized messaging occurs when an influencer shares a factually true personal anecdote but presents it as universally applicable advice. 

These overgeneralized messages might be inappropriate or even harmful for audiences due to varying health profiles. Because of the parasocial relationship and sense of relatability often created by SMIs, audience members could fail to realize the limited applicability of these messages. In turn, assuming their relevance for their own lives could potentially lead to detrimental outcomes similar to those of misinformation.

In order to counter the negative and compelling impact of both health misinformation and overgeneralized messaging by SMIs, the authors test an intervention designed to encourage critical reflection. The intervention strives to encourage individuals to engage in deliberate evaluation and reflection of the health content they see on social media. The basis of this approach is previous evidence from literature that shows critical thinking can reduce trust in and engagement with social media misinformation while not influencing the perception of precise and accurate content. A lack of these critical reflection skills is strongly associated with increased susceptibility to false or misleading content. Data for this study was collected through a 3x2 mixed-design online experiment in two waves – one with immediate responses and a follow-up two weeks later – in May and June 2025. The 562 participants were 18-25 years old and from the United States.

  • Misinformation is not the only form of potentially harmful health content circulating on social media. More subtle messages based on personal anecdotes can also mislead audiences, and our findings show they may be harder to counter.
    Study author Jaroslava Kaňková

During the experiment, participants were shown realistic TikTok-style videos of a fictitious influencer discussing either the carnivore diet (a diet in which only animal products are consumed) or "dry scooping" pre-workout powder. Both of these practices have been promoted by real SMIs despite warnings from medical professionals about their potential health risks. The videos were manipulated to contain either misinformation, overgeneralized claims, or correct information. A randomly selected half of the participants viewed an animated video explaining how to critically evaluate social media health content using three perspectives: information-related, source-related, and user-related. This animation was shown before the video of the fictitious SMI.

The study tracked the following outcomes: belief inaccuracy, message credibility, source trustworthiness, and parasocial relationship strength. For belief inaccuracy, participants were presented with health-related statements and had to indicate whether each was correct or false based on their own knowledge. Message credibility, source trustworthiness, and parasocial relationship strength were ranked by participants on a scale. For example, for "message credibility" participants were given the question: "How would you rate the message presented in the TikTok video you watched?" and could answer with the following options: "1. Inaccurate or accurate; 2. Inauthentic or authentic; 3. Unbelievable or believable".

Results show that exposure to both misinformation and overgeneralized messaging from social media influencers significantly impacted participants compared to correct messaging. Both types of misleading content led to greater inaccuracy in participants’ health-related beliefs. Participants rated misinformation and overgeneralized messages as significantly less credible than correct messages. Encountering misleading claims reduced the perceived trustworthiness of the influencer. Exposure to both types of misleading content actively diminished the strength of the parasocial relationships participants felt toward the influencer.

The intervention through critical reflection was found to be highly effective at reducing susceptibility to explicit misinformation in the short term. Contrary to the success with misinformation, the intervention had minimal to no effect on participants’ ability to identify and resist the subtler overgeneralized claims. 

By the two-week follow-up, the protective benefits of the intervention against misinformation had significantly diminished. This highlights that while encouraging critical thinking is a useful "shield" against blatant misinformation, it is less effective against anecdote-driven messaging that appears factually "true" but is contextually misleading.


Study author Jaroslava Kaňková concludes: Misinformation is not the only form of potentially harmful health content circulating on social media. More subtle (and factually accurate) messages based on personal anecdotes can also mislead audiences, and our findings show they may be harder to counter. As policymakers and platform designers work to address health misinformation, they should also account for these harder-to-detect forms of misleading health content on social media.