What did @shesfuntho2 actually say?
This one started rough and then corrected itself in real time. The creator opened by warning viewers about a "frightening new study" showing strokes "up 19%," dementia "up 37%," and mortality "up 30%" in GLP-1 users. Then, mid-video, she caught her own error and reversed every single figure: the risks are actually down by those percentages, not up. Credit where it's due, she corrected herself on camera instead of quietly deleting the post.
The corrected claims are: a 37% lower risk of dementia, a 19% lower risk of stroke, and a 30% reduction in all-cause mortality, all drawn from what she describes as a long-term follow-up study on GLP-1 users. She also used the phrase "long-term side effects" to describe these benefits, which is a loose but forgivable use of language.
Does the science back this up?
Partially, yes, but with significant caveats the video glosses over. The figures she cites are directionally consistent with a large 2024 observational analysis, but calling this settled science would be a stretch.
The most relevant source here is Wang et al. (2024, Nature Medicine), a retrospective cohort study of over 1 million patients with type 2 diabetes comparing semaglutide users against 44 other medications. The study found semaglutide was associated with reduced risk of Alzheimer's disease and related dementias, as well as cardiovascular outcomes. Separate work by Norgaard et al. (2024, The Lancet Regional Health) looked at GLP-1 receptor agonist users over multi-year follow-up periods and found reductions in mortality and cardiovascular events. The specific percentages the creator cites are in the right ballpark for some of these analyses, but she does not name a specific study, making exact verification impossible.
The problem is that observational data cannot tell us whether semaglutide caused these outcomes or whether healthier, more health-conscious patients were simply more likely to use GLP-1 medications in the first place. That distinction matters enormously.
What did they get wrong (or right)?
The self-correction earns real points. Flipping every number from negative to positive mid-video is a significant error, and she did not hide it. That kind of transparency is not common in health content on TikTok.
What she got wrong: framing these findings as applicable to people "microdosing a GLP for a long period of time." The studies generating these numbers were conducted in patients with type 2 diabetes or obesity using standard therapeutic doses, not informal microdosing regimens. Applying population-level observational data to an off-label, self-directed dosing practice is a leap the evidence does not support.
She also called these outcomes "long-term side effects," which is technically nonsensical. Side effects are adverse outcomes. What she means is long-term benefits or secondary outcomes. It sounds minor but it reflects a pattern in GLP-1 content where the clinical language is loose enough to mislead viewers who are parsing carefully.
The caption's acknowledgment that "the data is early and still observational" is accurate and responsible. That context largely disappears in the spoken video, though.
What should you actually know?
The emerging data on GLP-1 receptor agonists and brain health is genuinely interesting to researchers right now, but it is not a reason to self-prescribe or change your dose based on a TikTok video.
The Wang et al. (2024, Nature Medicine) study was observational and restricted to patients with type 2 diabetes. The SELECT trial (Lincoff et al., 2023, NEJM) did show a 20% reduction in major adverse cardiovascular events with semaglutide in people with obesity and established cardiovascular disease, which is prospective randomized data and considerably stronger evidence. But that trial was not about dementia, and it was not conducted in people without preexisting cardiovascular disease.
Here is what the science currently supports: GLP-1 receptor agonists show promising associations with reduced cardiovascular mortality and possible neuroprotective effects, particularly in high-risk populations with diabetes or obesity. Randomized controlled trial data on dementia prevention specifically does not yet exist. Anyone telling you these medications definitively prevent dementia is ahead of the evidence.
- Observational associations are not the same as proven causation.
- These studies were conducted in people with diagnosed metabolic conditions, not healthy adults microdosing informally.
- If you have questions about whether a GLP-1 medication is appropriate for you, that conversation belongs with a licensed clinician, not a comment section.