Tirzepatide, Semaglutide, and Dementia Risk — What New Real-World Data Actually Shows

Bottom Line

A large real-world study published in January 2026 found that people with type 2 diabetes on tirzepatide had a lower rate of newly diagnosed dementia over two years than people on semaglutide or on SGLT2 inhibitors, and that semaglutide also outperformed SGLT2 inhibitors. The differences are real and statistically significant, but this is an observational comparison of medical records, not a randomized trial designed to answer this question — so it should inform curiosity and conversation with a physician, not a decision to switch medications for brain-health reasons alone.

Background

GLP-1 receptor agonists cross into the brain and interact with insulin signaling and inflammation pathways, which has driven a wave of interest in whether they affect dementia risk independent of weight loss. Most of the prior evidence has come from post-hoc analyses of diabetes and cardiovascular trials not designed to study cognition. This new paper by Younis and colleagues, published online January 14, 2026 in Diabetes Research and Clinical Practice, takes a different approach: three separate “target trial emulations” — a method that uses real-world medical record data to approximate what a randomized trial would show by carefully matching patients on their baseline characteristics before comparing outcomes.

What the Study Found

• Data source: TriNetX, a federated network drawing on more than 150 million patient records across more than 140 healthcare organizations.
• Population: Adults with type 2 diabetes who had no dementia diagnosis at baseline, followed for two years after starting one of the compared medications.
• Tirzepatide vs. semaglutide: 12,246 matched patients per arm. Tirzepatide was associated with a 31% lower hazard of incident dementia (HR 0.69, 95% CI 0.48–0.99, p=0.04).
• Tirzepatide vs. SGLT2 inhibitors: 14,462 matched patients per arm. Tirzepatide showed a 34% lower hazard (HR 0.66, 95% CI 0.47–0.93, p=0.02).
• Semaglutide vs. SGLT2 inhibitors: 57,959 matched patients per arm — by far the largest comparison. Semaglutide showed a 29% lower hazard (HR 0.71, 95% CI 0.63–0.78, p<0.01).
• All-cause mortality: Tirzepatide and semaglutide were both associated with lower mortality than SGLT2 inhibitors, and tirzepatide lower than semaglutide (HR 0.72, 95% CI 0.58–0.90) — though the authors specifically flagged the tirzepatide-vs-SGLT2i mortality comparison as one where their own sensitivity analysis (E-values) suggests unmeasured confounding could plausibly explain some or all of the effect size.

Limitations

• This is an observational, retrospective comparison of electronic health records, not a randomized controlled trial. Target trial emulation is a rigorous way to minimize bias in that kind of data, but it cannot fully rule out confounding — for example, if healthier or more health-engaged patients are more likely to be prescribed tirzepatide in the first place, that alone could produce a result like this.
• Two years of follow-up is short for a disease like dementia that typically develops over a decade or more; the result shows what happened in that window, not necessarily the medication’s long-term effect on dementia risk.
• The population is adults with type 2 diabetes, not people using these medications solely for weight management without diabetes — the results may not generalize to BeLite’s non-diabetic weight-loss patients.
• The authors describe their own findings as hypothesis-generating and explicitly call for confirmation in randomized controlled trials before any clinical recommendation should change.

BeLite Clinical Perspective

This study is a reason for cautious optimism, not a reason to change anyone’s medication. Patients already being counseled on tirzepatide vs. semaglutide — for example using the comparative efficacy data in our head-to-head comparison of the two drugs — may reasonably ask about this result, and it’s fair to tell them the directional signal is encouraging and consistent with what’s known about GLP-1 receptor agonists and neuroinflammation. It is not, on its own, grounds to switch medications purely for brain-health reasons, particularly for patients doing well on their current regimen. For patients also asking about long-term cardiovascular outcomes on tirzepatide, our 3-year SURMOUNT-1 cardiovascular risk analysis is a useful companion read — both pieces point toward the same theme, that tirzepatide’s benefits may extend meaningfully beyond the number on the scale, while being clear that neither is a substitute for an individualized conversation with your prescribing physician.

FAQs

Q: Does this mean tirzepatide prevents dementia?

A: No. The study shows an association — people taking tirzepatide had fewer new dementia diagnoses than comparison groups over two years — but it cannot prove the medication caused that difference. Dementia risk is influenced by many factors the study could not fully control for.

Q: Should I switch from semaglutide to tirzepatide because of this study?

A: Not based on this alone. The result is interesting but preliminary, and switching medications involves other considerations — tolerability, cost, insurance coverage, and how well you’re responding to your current treatment. Talk to your prescribing physician before making any change.

Q: Does this apply to people who don’t have diabetes?

A: The study population was adults with type 2 diabetes. Many BeLite patients use these medications for weight management without diabetes, and the results haven’t been specifically studied in that group, so they should be read as suggestive rather than directly applicable.

Q: Why did SGLT2 inhibitors come out worse in this comparison?

A: The study compared the three drug classes to each other, not to no treatment. SGLT2 inhibitors are effective, well-studied diabetes medications in their own right; this result says the two GLP-1-based therapies were associated with a lower dementia rate than SGLT2 inhibitors in this particular dataset, not that SGLT2 inhibitors are unsafe.

Q: Is this the first study to look at GLP-1 drugs and dementia?

A: No — it adds to a growing body of research, including prior analyses suggesting lower stroke and dementia-related outcomes with semaglutide and tirzepatide in people with diabetes and obesity. This study’s target trial emulation design is a methodological step up from many earlier post-hoc analyses, but all of this research is still observational.

References

1. Younis A, Henney AE, Riley DR, et al. Target trial emulations for tirzepatide, semaglutide and SGLT2-inhibitors for dementia in patients with type 2 diabetes: real-world evidence from a retrospective cohort study. Diabetes Res Clin Pract. 2026;235:113083.
2. BeLite Medical Center. Tirzepatide vs. Semaglutide: Which Weight Loss Drug Works Better? [May 2025 NEJM Study Explained]
3. BeLite Medical Center. Does Tirzepatide Lower Cardiovascular Risk? What the 3-Year SURMOUNT-1 Analysis Actually Shows.
4. BeLite Medical Center. Our Medications.