Common Heartburn Drugs Linked to 66% Higher Dementia Risk in Massive New Study

Proton pump inhibitors are so common they barely register as medication anymore: omeprazole and pantoprazole sit in medicine cabinets worldwide, prescribed for heartburn, ulcers, and acid reflux. But a large new cohort study published in Scientific Reports by Haoquan Liu and colleagues suggests that long-term, high-dose use of these drugs may be associated with a meaningfully higher risk of developing dementia, with the signal strongest in people under 65.

PPI Use and Dementia Risk: What 2.4 Million Adults Revealed

The numbers in this study are hard to dismiss on scale alone. Researchers drew on Taiwan's Adult Preventive Health Services and National Health Insurance databases to follow 2,444,999 adults aged 40 and older, none of whom had dementia at the start. Over the follow-up period, 137,126 participants, about 5.6%, developed dementia. The mean age at dementia diagnosis was 72.7 years, and 58.6% of those affected were women.

The key finding is a dose-response relationship: the more cumulative PPI exposure a person had, the higher their associated dementia risk. At high exposure levels, the hazard ratio reached 1.66 (95% CI, 1.64 to 1.69), meaning high-exposure PPI users were roughly 66% more likely to develop dementia than non-users, after adjustment for age, sex, BMI, lifestyle factors, and comorbidities like hypertension and diabetes.

That association was even sharper in younger adults. For people under 65, high-exposure PPI use carried a hazard ratio of 1.97 (95% CI, 1.89 to 2.06), suggesting the drug's potential effect on the brain may be more consequential when exposure begins earlier in life.

Why This Matters for Millions of Patients Taking Acid Suppressants

PPIs are not a niche medication. They are among the most prescribed drug classes globally, and up to 60% of PPI prescriptions are estimated to lack a valid clinical indication, meaning many people are taking them long-term without a clear medical reason. Separate research has found that roughly 38% of older long-term PPI users lack a disease- or drug-related justification for continued use.

If a dose-dependent relationship between PPI use and dementia holds up under further scrutiny, the public health implications would be substantial. The drug class is used across all age groups, and the Taiwanese data suggest middle-aged adults may face disproportionate risk. Dementia prevention is already one of the most consequential challenges in aging research; identifying modifiable pharmaceutical exposures is a meaningful piece of that puzzle.

The biological mechanism remains unresolved. PPIs suppress gastric acid by inhibiting the H+/K+ ATPase pump in gastric parietal cells, but how that peripheral action might translate into neurodegeneration is not yet clear. Proposed pathways in the literature include effects on the gut microbiome and potential direct accumulation in brain tissue, though neither has been definitively confirmed.

How the Taiwan Cohort Study Isolated the PPI Signal

The study's methodological strength lies in its scale and its use of cumulative defined daily doses to quantify exposure, rather than simply asking whether someone ever used a PPI. That granularity enabled the researchers to model a dose-response curve, which is one of the more persuasive epidemiological signals for a causal relationship.

Multivariate Cox proportional hazards models were used throughout, adjusting for an extensive list of potential confounders. Participants who had dementia at baseline were excluded, and cohort study designs like this one are well-suited to assessing temporal relationships between an exposure and a later outcome, since participants are dementia-free at enrollment.

Interestingly, the study found that the association between PPI use and dementia risk attenuated with advancing age. In the oldest age groups, the hazard ratios were smaller, while the signal was most pronounced in those under 65. This age-dependent pattern adds a layer of nuance: the concern may be less about people already in their 70s and more about the cumulative burden of starting PPI use in middle age.

Why the Science Is Still Contested

This study lands in a genuinely contested scientific space. A 2019 meta-analysis of six cohort studies found no statistically significant association between PPI use and dementia or Alzheimer's disease. A large prospective cohort study known as ASPREE-XT similarly found no link after adjusting for confounders. And a major Welsh cohort study found PPI use was actually associated with roughly a 30% reduced risk of dementia, a result pointing in the opposite direction.

A focused review of the pantoprazole literature, among the most commonly studied individual PPIs, found prior Taiwanese data showing a 22% increased dementia risk in long-term users, while U.S. and Australian prospective data showed no such signal. The current study, with nearly 2.5 million participants and a dose-quantified exposure measure, offers the most granular look yet, but it does not resolve the contradiction.

As an observational study, it cannot establish causation. Unmeasured confounders remain a concern: people who take PPIs long-term may differ from non-users in ways that are difficult to fully capture, including the underlying conditions that led to PPI use in the first place. A randomized controlled trial would be needed to isolate the drug's effect, and such a trial would be logistically and ethically complex to design.

What Comes Next for PPI Safety Research

The Liu et al. findings add meaningful weight to a literature that has been running in circles for a decade. The dose-response pattern and the especially elevated risk in younger adults are the two signals most worth watching in follow-up work. Future research will need to disentangle whether the dementia risk, if real, is driven by specific PPI molecules, cumulative acid suppression, microbiome disruption, or something else entirely.

For now, the practical message is not to stop medically necessary PPI therapy without consulting a physician. But the study does reinforce what deprescribing advocates have argued for years: long-term PPI use should be periodically reassessed, and patients taking these drugs without a clear ongoing indication may be accepting a risk they do not need to take.

Evidence rating

EMERGING — 73/100

Caveats:

• Observational design cannot establish causation — unmeasured confounders (e.g., underlying GI disease severity, diet, or polypharmacy) may explain part of the association.
• Results are inconsistent with prior large cohort studies, including the ASPREE-XT trial and a Welsh cohort that found no association or a reduced dementia risk with PPI use.
• The causal mechanism linking PPI use to dementia has not been established; the authors acknowledge this is unresolved.
• The study population is drawn entirely from Taiwan's National Health Insurance database, which may limit generalizability to other ethnic or healthcare populations.
• Funding was not disclosed by the authors.

Large retrospective cohort study (n=2,444,999) using Taiwanese national health insurance data. The design can identify associations but cannot establish causation, and unmeasured confounding remains a material risk. Results are inconsistent with other large-scale cohort studies in this area, and the study has not yet been independently replicated.

References

1. Hsing-Hui Liu, Tzu-I Chen, Yong-Chen Chen, et al. (2026). Proton pump inhibitor use and risk of dementia in a population-based cohort study. Scientific Reports. https://doi.org/10.1038/s41598-026-55683-7
2. Monica Andrawes, Wessam Andrawes, Abhishek Das, Keith Siau (2025). Proton Pump Inhibitors (PPIs)—An Evidence-Based Review of Indications, Efficacy, Harms, and Deprescribing. Medicina. https://doi.org/10.3390/medicina61091569
3. Susanna M. Wallerstedt, Johan Fastbom, Johannes Linke, Sigurd Vitols (2016). Long‐term use of proton pump inhibitors and prevalence of disease‐ and drug‐related reasons for gastroprotection—a cross‐sectional population‐based study. Pharmacoepidemiology and Drug Safety. https://doi.org/10.1002/pds.4135
4. Overview: Cohort Study Designs – PMC – NIH. NIH. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6032483/
5. Hoong Chen Teo, Ahimsa Campos-Arceiz, Binbin V. Li, Mingquan Wu, Alex Mark Lechner (2020). Building a green Belt and Road: A systematic review and comparative assessment of the Chinese and English-language literature. PLOS ONE. https://doi.org/10.1371/journal.pone.0239009
6. Sreeja Koundinya, Rumitha Chakilam, Gayatri Brahmandam, et al. (2025). Long-term proton pump inhibitor use and risk of dementia: a focused review on pantoprazole. Annals of Medicine & Surgery. https://doi.org/10.1097/ms9.0000000000004022