The short answer: on the current evidence, proton pump inhibitors probably do not cause dementia. The headline that frightened everyone came from a single German claims-data study in 2016, and the better-designed research that followed has largely failed to reproduce it.
An umbrella review published in 2026 — a review of all the previous reviews — concluded that most analyses find no link, and that a genuine causal relationship appears unlikely.
But I don’t think “don’t worry about it” is the right takeaway either, and here’s why. If you have silent reflux and you’re on a PPI that isn’t actually controlling your throat symptoms, you’re carrying whatever small uncertainty exists for no benefit at all. That’s the real problem — not dementia risk, but risk without reward.
Key Takeaways
- The 2016 study that started the scare found a 44% increased dementia risk — but it was based on prescription claims data with limited control for confounding factors.
- A large prospective study of 18,934 older adults published in 2023 found no association at all (hazard ratio 0.88, meaning slightly lower risk, not statistically significant).
- Pooled analyses of prospective studies have not found a significant link, including in people using PPIs for more than five years.
- A 2026 umbrella review concluded a true causal relationship appears unlikely.
- The most likely explanation for the early findings is confounding by indication — the people prescribed PPIs long-term were sicker and older to begin with.
- None of this makes indefinite PPI use sensible, particularly for LPR, where these drugs frequently don’t work well in the first place.
- The better question isn’t “will this give me dementia” but “is this drug doing anything for me, and what’s the plan?”
Where the dementia scare came from
In February 2016, researchers published an analysis of German health insurance data covering 73,679 people aged 75 and over, none of whom had dementia at the start. They compared those on regular PPI prescriptions against those who weren’t, and found the PPI group had a significantly higher rate of incident dementia — a hazard ratio of 1.44, meaning roughly a 44% relative increase [Gomm et al., JAMA Neurology, 2016].
The paper went around the world. It’s still the study behind most of what you’ll read about PPIs and your brain, nearly a decade later.
The proposed mechanism sounded plausible, too. PPIs cross the blood-brain barrier. Mouse work had suggested they might increase beta-amyloid levels. And long-term acid suppression reduces vitamin B12 absorption, and B12 deficiency genuinely does cause cognitive problems. Three separate threads all pointing the same way.
So it wasn’t an unreasonable alarm. It just hasn’t held up.
What the better studies found
Here’s the thing about claims-data studies: they tell you what was prescribed, not what was swallowed, and they’re poor at capturing the reasons behind a prescription. That matters enormously here.
In 2023, researchers analyzed data from ASPREE, a large trial of 18,934 community-dwelling adults aged 65 and over in the US and Australia. Crucially, medication use was verified at annual in-person visits rather than inferred from prescription records, and dementia was diagnosed against formal criteria rather than read off a billing code. They found no association between PPI use and incident dementia — a hazard ratio of 0.88, with a confidence interval spanning 1 — no association with cognitive impairment, and no difference in how cognitive test scores changed over time [Mehta et al., Gastroenterology, 2023].
That’s a considerably better study design than the one that raised the alarm, and it found nothing.
Pooling helps too. A meta-analysis restricted to prospective studies — six studies, 308,249 people, average follow-up five years — found a pooled hazard ratio of 1.16 with a confidence interval crossing 1, meaning no statistically significant effect. Restricting to studies with more than five years of PPI use didn’t change it, and the figure for Alzheimer’s disease specifically was 1.06 [Desai et al., Diseases of the Esophagus, 2020].
A separate meta-analysis with formal bias assessment reached a similar place: a pooled risk ratio of 1.16 with the lower bound sitting right on 1.00, and no clear evidence of an association — though the authors were careful to say some risk couldn’t be entirely ruled out [Ahn et al., British Journal of Clinical Pharmacology, 2023].
And most recently, the umbrella review: eleven systematic reviews and meta-analyses published between 2016 and 2024, covering 21 studies between them. Only one of the high-confidence reviews found a significant association with dementia, and none of the six that looked specifically at Alzheimer’s disease found one. The conclusion was that a true causal relationship appears unlikely [Emad et al., Digestive Diseases and Sciences, 2026].
So why did the first study find something?
Most likely confounding by indication — the trap that catches observational drug research over and over.
Think about who ends up on a PPI indefinitely at age 80. People on multiple other medications. People with heart disease taking aspirin or anticoagulants who need stomach protection. People who’ve been unwell enough to see doctors often. People with poorer diets and less mobility.
Every one of those things is independently associated with dementia risk. So when you compare long-term PPI users against non-users, you aren’t really comparing a drug against no drug — you’re comparing sicker people against healthier people, and the drug takes the blame for the difference.
This is exactly what you’d expect to see wash out when researchers move to prospective designs with proper in-person assessment. And that’s what happened.
The part that actually matters if you have silent reflux
I’ve laid out the reassuring evidence honestly, because I think you deserve it straight. Now the uncomfortable part.
Very few people with LPR are helped much by PPIs in the first place. The placebo-controlled trials in laryngopharyngeal reflux are genuinely disappointing, and I’ve gone through why PPIs underperform in LPR in detail elsewhere. The short version: reducing acid production doesn’t stop reflux happening. It changes the pH of what comes up. Pepsin still travels to your throat, still binds to the tissue, and still reactivates whenever something acidic passes through. Non-acid reflux carries on entirely unaffected.
Which is why so many people take omeprazole for six months, feel no real change in their throat, and stay on it anyway — because stopping feels risky and nobody has given them an alternative plan.
That’s the position worth worrying about. Not because PPIs will give you dementia; the evidence says they probably won’t. But because indefinite medication with no measurable benefit is a bad deal at any level of risk. And the documented downsides of long-term use — low magnesium, B12 depletion, and the various other omeprazole side effects — are far better established than the dementia link ever was.
When a PPI is genuinely worth taking
I don’t want to talk anyone out of a drug they need, so let me be clear about where PPIs earn their place:
- Erosive esophagitis. Visible damage to the esophageal lining seen on endoscopy. PPIs heal it, reliably, and this is what they were designed for.
- Barrett’s esophagus. Ongoing acid suppression is part of standard management.
- Stomach protection. If you’re on long-term NSAIDs or certain anticoagulants, a PPI is preventing bleeding that could kill you. Don’t stop it over a headline.
- Confirmed acid-predominant GERD with a good response. If it works and your symptoms return without it, it’s doing its job.
What’s much harder to justify is the very common scenario: a PPI started years ago for throat symptoms, never formally reviewed, still being repeat-prescribed, with the person unable to say whether it helps. If you can’t name what your PPI is achieving, that’s worth raising with your doctor — not stopping unilaterally, but asking.
What to do if you want off them
The important thing: don’t stop abruptly. Your stomach responds to acid suppression by upregulating acid production, and when you remove the drug suddenly you get rebound hypersecretion — a surge of acid that can be worse than what you started with. It typically peaks around two weeks after stopping and settles over four to eight weeks.
People routinely misread this as proof they needed the drug all along, and go straight back on. I’ve covered the whole mechanism in getting off PPIs and acid rebound, and it’s the single most useful thing to understand before you try.
A sensible approach involves tapering rather than stopping, having something in place to cover the rebound period — alginate rafts are the most useful tool here because they work mechanically rather than chemically — and having your diet already sorted before you start, not after. Doing it the other way round is why most attempts fail. The Wipeout Diet Plan is the framework I’d want in place first, because coming off acid suppression while still eating and drinking the things that drive reflux is asking to fail.
And do it with your doctor, particularly if you have any of the conditions listed above.
Conclusion
If you came here worried, you can reasonably set the dementia question down. The alarming 2016 finding hasn’t survived better research, the largest prospective study found nothing, and the most recent synthesis of all the evidence concludes a causal link is unlikely. That’s about as reassuring as observational medicine gets.
The question that deserves your attention instead is whether your PPI is doing anything. If you have silent reflux, there’s a fair chance it isn’t — and the answer to that isn’t a different drug, it’s addressing what’s actually driving the reflux. That’s what the Wipeout Diet Plan is built to do. It’s the complete protocol: which foods and drinks reactivate pepsin in your throat and why, the pH thresholds that matter, meal timing, the night-time routine, and a structured way to work through it rather than a list of things to avoid. It was designed first around LPR — the throat-based form where acid suppression disappoints most — but because it targets the same underlying mechanisms it works just as well for GERD and ordinary heartburn. It’s also, for most people, the thing that makes coming off a PPI survivable.
If you want a lighter starting point, the Wipeout Food Reference Guide is the essential companion — every food and drink that’s safe for acid reflux and LPR with its pH value, so you can see at a glance what’s likely to be causing trouble. The Diet Plan is where the actual work happens; the guide is the reference you’ll keep to hand.
One last thought. Nine years of research spent chasing a signal that probably wasn’t there is a reminder to be careful with single studies, in both directions. The reason to reconsider a long-term PPI isn’t fear of dementia. It’s that a medication should be doing something for you, and if you can’t say what yours is doing, that’s worth a conversation.
Frequently Asked Questions
Should I stop my PPI because of the dementia research?
No — not on that basis. The evidence doesn’t support a causal link, and if you’re taking a PPI for erosive esophagitis, Barrett’s, or stomach protection alongside NSAIDs or anticoagulants, stopping could cause real harm. If you want to come off for other reasons, do it as a planned taper with your doctor.
Does it matter which PPI I take?
Not for this question. The studies looked at omeprazole, pantoprazole, lansoprazole, esomeprazole and rabeprazole as a class, and nothing suggests one is safer than another for cognition. If you’re comparing options for other reasons, medication choices for LPR covers the differences that actually matter.
What about the B12 connection?
That one is real, but it’s a separate issue. Long-term acid suppression genuinely reduces B12 absorption, and severe B12 deficiency can cause cognitive symptoms — which are usually reversible with treatment. That’s an argument for having your B12 checked if you’ve been on a PPI for years, not evidence that PPIs cause dementia.
Is famotidine safer for the brain?
The 2023 ASPREE analysis looked at H2 blockers as well and found no association with dementia or cognitive decline for those either. So neither class shows a signal. Famotidine versus omeprazole compares them on the grounds that do differ.
How long is “too long” to be on a PPI?
There’s no fixed cut-off, and long-term use is appropriate for some conditions indefinitely. The useful rule is that any PPI running beyond about eight weeks without a clear, named reason deserves a review. Most people who’ve been on one for years have never had that review.
Could my brain fog be from my PPI?
Possibly, though reflux itself is a more common culprit — poor sleep from night-time symptoms will fog anyone. Worth checking B12 and magnesium if you’ve been on acid suppression a long time, since deficiencies in both can produce exactly that feeling.
Why do PPIs get blamed for so many things?
Because they’re among the most prescribed drugs in the world, taken mostly by older people with multiple health problems, usually for years. That combination generates enormous observational datasets in which PPI use will correlate with almost every age-related illness — correlations that tend to shrink or vanish when studies are designed to separate the drug from the people taking it.
Research & References
- [Gomm et al., JAMA Neurology, 2016] — Analysis of German health insurance claims covering 73,679 people aged 75 and over found regular PPI users had a significantly increased risk of incident dementia (hazard ratio 1.44). This is the study that generated the original concern.
- [Mehta et al., Gastroenterology, 2023] — Analysis of 18,934 adults aged 65 and over from the ASPREE trial, with medication verified at in-person visits, found no association between PPI use and incident dementia (hazard ratio 0.88), cognitive impairment, or cognitive decline over time.
- [Desai et al., Diseases of the Esophagus, 2020] — Meta-analysis of six prospective studies covering 308,249 subjects found a pooled hazard ratio of 1.16 for dementia, not statistically significant, with results unchanged when restricted to PPI use beyond five years.
- [Ahn et al., British Journal of Clinical Pharmacology, 2023] — Systematic review and meta-analysis with formal bias assessment across nine observational studies found no clear evidence of an association, while noting that some risk could not be entirely excluded.
- [Emad et al., Digestive Diseases and Sciences, 2026] — Umbrella review of eleven systematic reviews and meta-analyses concluded that most find no link between PPI use and dementia, and that given the low certainty of the evidence a true causal relationship appears unlikely.
David Gray
12 years living with LPR · Consultant & researcher
I've lived with LPR for twelve years — the misdiagnoses, the PPI courses that did nothing, the slow work of figuring out what actually helps. Wipeout Reflux is where I translate the research into plain terms for people stuck in the same place. Every claim here is sourced to peer-reviewed work, and I consult one-to-one with LPR sufferers.