top of page

Does Gabapentin Cause Dementia?

3 days ago
7 min read



When Medical Research Becomes a Source of Fear (Understanding Correlation, Causation, and the Nocebo Effect)


Not long ago, I was talking with someone who had been taking gabapentin while tapering a benzodiazepine. He'd been making progress, working hard on his recovery, and recently had a pretty good vacation. He could go hiking, eat good food, and have big, full, surprisingly low-symptom days.

 

Then something happened. When he picked up his gabapentin prescription, his pharmacist asked whether he'd seen the recent studies linking gabapentin to dementia. Now, imagine hearing that when you're already struggling with medication withdrawal, worrying about your health, and doing your best to get through each day. It's a pretty frightening thing to hear.

 

Naturally, my student began ruminating on the medication and was planning on holding his benzodiazepine taper so he could taper the gabapentin. The problem is that it’s usually not a good idea to taper two meds simultaneously. Even though he planned to hold the benzo while tapering the gabapentin, he was already over 70% through his taper. It could be a lot on an already sensitized system.

 

But it raises an important question: What did those studies actually find?

 

Because as we all know, correlation isn’t always causality.

 

And perhaps more importantly, how do we distinguish what research tells us from what our fears tell us it means?

 

What the Research Found

 

In 2025, researchers published a study examining gabapentin prescriptions and dementia diagnoses among more than 26,000 patients with chronic lower-back pain. They reported that patients receiving six or more gabapentin prescriptions had a 29% higher relative risk of being diagnosed with dementia, compared with patients who weren't prescribed gabapentin.

 

Initially, that sounds alarming and is certainly worth investigating. But there's something important to understand. The study was observational. It identified a correlation, not a demonstrated causal relationship. In other words, researchers found that gabapentin prescriptions and dementia diagnoses were associated with one another. They did not establish that gabapentin actually caused dementia.

 

Those are two very different conclusions.

 

Correlation Is Not Causation

 

Let's use a simple example. Imagine researchers discover that people who eat more ice cream are also more likely to drown. Should we conclude that ice cream somehow makes people more susceptible to drowning?

Of course not.

 

Both ice cream consumption and drowning incidents tend to increase during the summer. Warmer weather is a third variable that helps explain the relationship. The correlation is real, but our initial interpretation of what it means could be completely wrong.

 

Another example might be a correlation between sunscreen and shark attacks.

 

Studies show that people wearing sunscreen are more likely to be involved in a shark attack. But not because sunscreen attracts sharks, but because people often wear sunscreen while swimming in the ocean where sharks are.

Medical research is considerably more complicated, but the underlying principle is the same.

 

For example, researchers have found associations between certain medications and serious illnesses. Sometimes those medications contribute to the problem. Other times, the people taking them already have medical conditions that increase their risk.

 

This is called confounding.

 

And it's one of the biggest challenges researchers face when interpreting observational studies.

 

What Else Could Explain the Gabapentin Findings?

 

Consider the population of people who are prescribed gabapentin. Many experience chronic pain, neurological conditions, sleep disturbances, or other medical problems. These conditions may independently be associated with cognitive difficulties or dementia risk. Age is another consideration. Older adults are more likely to be prescribed certain medications and are already at higher risk for dementia.

 

Researchers attempt to control for these differences statistically, and the 2025 gabapentin study did account for several important factors. But statistical adjustments aren't perfect. There may still be differences between groups that researchers couldn't measure or account for.

 

There's also something called reverse causation.

 

What if early neurological changes sometimes contribute to symptoms that lead a person to seek treatment in the first place? In that scenario, the developing illness might help explain why someone was prescribed medication, rather than the medication necessarily causing the illness. We can't assume that's what happened in the gabapentin studies. But it's a plausible alternative explanation researchers must consider.

 

What Happens When Other Studies Disagree?

 

Here's where things become especially interesting. A 2024 study involving more than 200,000 adults with chronic pain found no statistically significant increase in dementia risk associated with gabapentin use. That matters because the study population was nearly 10x larger than the study reporting a correlation between gabapentin and dementia.

 

And a 2023 meta-analysis examining anti-seizure medications found that an initially significant association between these medications and dementia disappeared when researchers accounted for the conditions being treated.

 

That doesn't prove gabapentin is harmless.

 

In fact, other studies have reported an increased association, including a more recent 2026 meta-analysis. But that analysis also found substantial differences between studies, and its overall evidence was rated as low certainty. When researchers restricted their analysis to studies with a lower risk of bias, the association was no longer statistically significant.

 

So what should we conclude?

 

That the question remains unresolved. And that's not a weakness of science. It's part of how science works. Scientific understanding evolves as researchers collect evidence, challenge earlier findings, identify limitations, and refine their conclusions. Sometimes the most accurate answer isn't yes or no. Sometimes it's simply that we don't know yet.

 

And while we may not have the full answer yet, it’s clear that, across most studies, when adjusted for other false correlations, there doesn’t seem to be a strong relationship between gabapentin and dementia. Certainly not enough to alarm people and have them racing off their meds while tapering other medications.

 

Understanding Relative Risk

 

Another important distinction is the difference between relative risk and absolute risk. Imagine a study finds that a medication increases the relative risk of some condition by 50%.

 

That sounds enormous. But suppose the original risk was 2%. A 50% relative increase would bring that risk to 3%.

 

That's a one-percentage-point increase in absolute risk.

 

Both statistics are mathematically correct. But they create very different impressions.

 

In the 2025 gabapentin study, a subsequent scientific review reported an absolute difference of approximately one percentage point in dementia diagnoses between comparison groups. This wasn't evidence that gabapentin caused one additional dementia case per hundred people. It was an observed difference between groups. Understanding these distinctions helps us interpret research more accurately without minimizing legitimate concerns.

 

When Research Activates the Bear

 

Now, let's bring this back to recovery. Imagine someone taking medication for years who suddenly encounters a headline suggesting it might cause dementia. What happens next? For someone whose nervous system is already sensitized, that information may be interpreted as an immediate personal threat.

 

The Bear begins asking questions.

 

What if I've damaged my brain?

What if my memory problems are permanent?

What if I develop dementia years from now?

Should I stop taking this medication immediately?

 

And suddenly, a scientific finding about population-level associations becomes a frightening prediction about one's personal future. This is where health anxiety, threat interpretation, and the nocebo effect can enter the picture.

 

The nocebo effect refers to the way negative expectations can contribute to unpleasant symptoms or experiences. That doesn't mean those symptoms are imaginary. It means expectations, attention, and threat perception can influence how the nervous system processes and experiences physical sensations.

 

Someone who becomes frightened about cognitive decline might begin monitoring every forgotten word, every moment of distraction, and every ordinary memory lapse. Things they previously wouldn't have noticed may suddenly feel like evidence of something terrible happening. The research hasn't changed. But the meaning assigned to it has. And that meaning can become a powerful source of distress.

 

Don't Confuse Uncertainty with Danger

 

None of this means we should dismiss research suggesting possible medication risks. Gabapentin has recognized side effects, including dizziness, sedation, and cognitive difficulties in some patients. Like other medications, its benefits and risks deserve periodic evaluation. Naturally, dose, age, and other factors also matter.

 

And because gabapentin can produce withdrawal symptoms, particularly when discontinued abruptly, medication changes should be discussed with a qualified prescriber. But there's a difference between recognizing a potential risk and assuming that risk has already become reality. One is an informed response. The other may be a fear-driven conclusion unsupported by the available evidence.

 

In recovery, we're often learning to tolerate uncertainty without automatically interpreting it as danger. That includes uncertainty about our symptoms, our progress, our future, and even the scientific studies we encounter.

 

Becoming a Better Consumer of Science

 

The next time you encounter a frightening medical headline, consider asking yourself a few questions...

 

Was the study observational or experimental?

Did researchers demonstrate causation, or merely identify an association?

What other variables might explain the findings?

Have other studies reached different conclusions?

 

And perhaps most importantly, what does the evidence actually tell us about our individual situation?

 

We can take medical research seriously without treating every new finding as a personal catastrophe. We can recognize possible medication risks without assuming permanent damage. And we can remain curious, thoughtful, and open to new evidence without surrendering our peace of mind to every frightening headline. The goal isn't to stop listening to science. It's to understand what the science is actually saying.

  

References

Eghrari, N. B., Yazji, I. H., Yavari, B., Van Acker, G. M., & Kim, C. H. (2026). Risk of dementia following gabapentin prescription in chronic low back pain patients. Regional Anesthesia & Pain Medicine, 51(10), 1118–1123.

 

Tsai, S.-E., Yang, S.-F., Wang, Y.-H., & Yeh, C.-B. (2024). Association between gabapentin use and risk of dementia in adults with chronic pain: A nested case-control study. Journal of Affective Disorders, 358, 205–210.

 

Huang, Y.-H., Pan, M.-H., & Yang, H.-I. (2023). The association between gabapentin or pregabalin use and the risk of dementia: An analysis of the National Health Insurance Research Database in Taiwan. Frontiers in Pharmacology, 14, 1128601.

 

Examined 206,802 participants and found a 45% higher relative dementia hazard associated with gabapentin or pregabalin exposure. Importantly, the study combined both medications. 

 

Da Silva, A. M. P., et al. (2026). Gabapentinoid use and risk of Alzheimer's disease-related dementias: A systematic review and meta-analysis of observational studies. Drugs & Aging, 43, 661–673.

 

Zhang, et al. (2023). Anti-seizure medication exposure and the risk of dementia: A meta-analysis of observational studies. Frontiers in Neurology, 14, 1133816.

Colloca, L. (2024). The nocebo effect. Annual Review of Pharmacology and Toxicology, 64, 171–190.

 

Rooney, T., Sharpe, L., Todd, J., Richmond, B., & Colagiuri, B. (2023). The relationship between expectancy, anxiety, and the nocebo effect: A systematic review and meta-analysis with recommendations for future research. Health

Psychology Review, 17(4), 550–577.

 

Tu, Y., Zhang, L., & Kong, J. (2022). Placebo and nocebo effects: From observation to harnessing and clinical application. Translational Psychiatry, 12, 524.

 

U.S. Food and Drug Administration. (2019, December 19). Neurontin, Gralise, Horizant (gabapentin) and Lyrica, Lyrica CR (pregabalin): Drug safety communication—Serious breathing problems.

 

 

 

 

 


 
 
 

Comments


Post: Blog2_Post
  • Facebook
  • YouTube
  • Instagram

©2027 Powers Benzo Coaching LLC

bottom of page