TL;DR
Get health and wellness essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
Researchers reviewing online nutrition advice for people with breast cancer found that 12.5% of 151 claims were supported by reliable evidence or established scientific facts. The study, based on searches conducted from January to April 2025, found the highest proportions of unsupported or contradicted claims on TikTok and X; it did not test whether people followed the advice or how it affected them.
A review of 151 online nutrition claims aimed at people diagnosed with breast cancer found that just 12.5% were supported by reliable evidence or established scientific facts, according to a study published in Patient Education and Counseling. Claims on TikTok and X had the largest shares classified as lacking evidence or contradicted by research, highlighting the difficulty of finding dependable diet guidance online.
The researchers collected material between January and April 2025 from Google, Facebook, Instagram, Pinterest, YouTube, YouTube Shorts, TikTok, X and ChatGPT-4. They used five everyday searches, including “breast cancer diet” and “cancer nutrition,” and assessed prominent search results and chatbot responses. A multidisciplinary expert panel compared the claims with clinical guidelines and peer-reviewed systematic reviews.
Of the claims, 8.6% were supported by clinical guidelines or systematic reviews, while about 4% reflected established scientific facts. Researchers classified around 63% as having no evidence, 21% as having weak evidence and 4% as contradicted by established literature. The report grouped the figures for claims with weak or no support and those contradicted by research as 87.5% of the total.
Platform results varied. The study classified 76.3% of TikTok claims and 72.9% of X claims as lacking support or contradicted. Google had the highest share of scientifically supported information at 33.3%, followed by ChatGPT at 30%. Only five ChatGPT-4 responses were evaluated, and the researchers said the answers were similar and general across the five prompts, limiting conclusions about AI advice.
Why Diet Claims Can Affect Care
People may look online for food guidance while managing treatment effects, seeking to support their general well-being or wondering whether diet can affect the chance of recurrence. In that setting, the study’s findings matter because confident-sounding claims can be difficult to distinguish from evidence-based guidance, particularly when advice is shared through short videos or personal accounts.
The results do not show that every unsupported claim is false, or that any particular diet causes harm. They show that most claims in the selected search results did not have strong backing from the guidelines and research used by the reviewers. That distinction matters: a lack of evidence is not proof that a claim is wrong, but it is not a sound basis for presenting it as established fact.
The study also found that claims posted by health professionals were more often evidence-supported than those from personal accounts or organizations, but the differences were not statistically significant. The researchers said professionals produced about one-third of the non-AI material reviewed. Readers making decisions about diet during or after treatment can discuss questions with their oncology team or a qualified oncology dietitian rather than relying on a single online post.
As an affiliate, we earn on qualifying purchases.
How Researchers Checked Online Advice
The study addressed online nutrition advice for people affected by breast cancer, a population that includes patients seeking help with treatment-related side effects and survivors looking for lifestyle information. The source report cites World Health Organization estimates of 2.3 million new breast cancer cases and 670,000 deaths worldwide in 2022. Wider access to information online makes advice easy to find, while its quality can be difficult to judge.
Researchers searched using non-academic wording intended to reflect how patients might look for answers. They reviewed top results on social platforms and Google’s first page, along with five ChatGPT-4 responses. Human reviewers classified claims as supported, contradicted or backed by little evidence, using guidance from the European Society for Clinical Nutrition and Metabolism, the American Society of Clinical Oncology and the Clinical Oncology Society of Australia, as well as recent systematic reviews.
Frequently discussed topics included cruciferous vegetables, alcohol, plant-based diets and the claim that sugar “feeds” cancer. These were counts of how often topics appeared in the collected material; they do not indicate how common the beliefs are among patients or whether the claims changed anyone’s choices.
“Only 12.5% of the 151 post-diagnosis nutrition claims were supported by reliable evidence or established scientific facts.”
— Study authors, as summarized in the News-Medical report
evidence-based nutrition for breast cancer
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Limits of the Search Results
The findings describe a specific sample of search results and responses, gathered over four months in 2025. They do not establish the quality of every post, website or AI answer available to patients, and search results may change over time. The chatbot assessment was especially limited: only five ChatGPT-4 responses were reviewed.
The researchers did not assess whether patients followed the advice, how it affected their diets or health, or whether encountering a claim changed their treatment decisions. Nor do the platform percentages show the likelihood that a typical user will encounter misinformation: the study examined selected results rather than all platform content. The source report does not provide enough detail here to establish how search personalization or location may have shaped results.
It also remains unclear whether the proportions found in this review apply to other cancers, languages, search terms or later versions of AI tools. The results identify weaknesses in the material sampled, not a clinical assessment of individual dietary choices.
As an affiliate, we earn on qualifying purchases.
What Patients and Researchers Can Do
The study does not set out a follow-up timetable or a new clinical recommendation. For readers, its practical implication is to treat broad online diet claims cautiously and ask a medical team or qualified oncology dietitian how they apply to an individual’s treatment and needs. A social media post or chatbot response should not be treated as a substitute for personalized clinical guidance.
Further research could examine a larger number of AI responses, repeat searches over time and test whether people act on the claims they encounter. The researchers also call attention to the need for clearer, accessible nutrition information from health professionals and organizations. Until those questions are studied, the reported figures should be read as a snapshot of selected online results, not a complete measure of all breast cancer diet advice.
supportive foods for breast cancer recovery
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did the study find about online breast cancer diet advice?
Among 151 claims reviewed, 12.5% were supported by reliable evidence or established scientific facts. The rest were classified as having no or weak evidence or as contradicted by established research.
Which sources had the highest shares of unsupported claims?
In the selected material, TikTok had a 76.3% share and X a 72.9% share of claims classified as lacking support or contradicted. Those figures apply to the study’s sampled results, not everything posted on either platform.
Did the study show that unsupported diet claims are false?
No. The authors’ distinction was that lack of supporting evidence does not necessarily mean a claim is false. The review assessed how claims matched available guidelines and research; it did not test every claim directly.
How reliable was the ChatGPT comparison?
The researchers reviewed only five ChatGPT-4 responses. Although 30% of the chatbot claims were classified as scientifically supported, the small sample and similar, general responses limit what can be concluded about AI-generated nutrition advice.
Where can someone get diet guidance after a breast cancer diagnosis?
The study did not evaluate individual care options. People with questions about food during or after treatment can discuss them with their oncology team or a qualified oncology dietitian, who can account for their circumstances.
Source: rss
Cold & flu season Picks
humidifiers
As an affiliate, we earn on qualifying purchases.
