Does Taking Xanax Actually Cause Dementia?

The honest answer is that decades of research have not proven Xanax or other benzodiazepines directly cause dementia, despite persistent headlines suggesting otherwise. Large observational studies do find a statistical association between benzodiazepine use and later dementia diagnoses, but the size of that link shrinks or vanishes when researchers account for the fact that early, undiagnosed dementia itself causes the anxiety and insomnia that lead people to benzodiazepines in the first place. The question is far from settled, but the evidence leans more toward “confounded correlation” than “proven cause.”

What the Numbers Actually Show

Multiple meta-analyses have pooled data from observational studies to estimate how much benzodiazepine use is associated with dementia risk. The numbers vary depending on which studies are included and how the data are analyzed. One meta-analysis of ten studies found that people who had ever used benzodiazepines had roughly a 50 percent higher risk of dementia compared to people who never used them.¹ A large French prospective study followed nearly a thousand older adults and found new benzodiazepine users had about a 60 percent higher hazard of developing dementia over the follow-up period.² And an umbrella review that gathered five meta-analyses encompassing 30 studies found effect sizes ranging from about 1.38 to 1.78, meaning benzodiazepine users were somewhere between 38 and 78 percent more likely to be diagnosed with dementia.³

Those numbers sound alarming. But that same umbrella review concluded that the evidence supporting a real link was “weak” and the methodological quality of the underlying studies was “low.” The problem is not the math. It is what the math cannot account for.

The Chicken-and-Egg Problem

The single biggest challenge in this research is something epidemiologists call protopathic bias, though a simpler term is reverse causation. Dementia does not begin the day someone gets a diagnosis. The disease process starts years, sometimes more than a decade, before noticeable memory loss. During that long pre-diagnosis window, people often develop anxiety, sleep disturbances, and depression. Those are exactly the symptoms that prompt a doctor to prescribe Xanax or a similar drug. So a person may start taking benzodiazepines not because the drug will eventually harm their brain, but because their brain is already changing in ways that produce anxiety as an early symptom.

One study using electronic medical records and machine learning found exactly this pattern: short-acting benzodiazepines appeared to act as a kind of unintentional screening test, with the people who developed dementia soonest after starting the drug likely already having subclinical disease at the time of prescription.⁴ A case-control analysis made this even clearer: when people started benzodiazepines less than a year before an Alzheimer’s diagnosis, the odds ratio was about 2.2, but for those who started two to three years before diagnosis, it dropped to essentially 1.0, meaning no increased risk at all. After accounting for benzodiazepine use during this prodromal phase, long-term use was actually associated with a lower risk of Alzheimer’s.⁵

Researchers have tried to address this by building in “lag times,” excluding benzodiazepine use in the years just before a dementia diagnosis, to avoid counting prescriptions that were likely treating early dementia symptoms. A systematic review and meta-analysis that focused specifically on studies using the longest lag times (five years or more) still found a modest but statistically significant association, with an odds ratio of about 1.30.⁶ So the link does not completely disappear. But a more recent analysis using advanced confounding-control methods went further and found that when you restrict the comparison to people who already have depression or anxiety, the association between benzodiazepines and dementia vanishes entirely. The apparent link appeared to be driven by the higher rates of diabetes, cardiovascular disease, depression, and anxiety among benzodiazepine users, not by the drug itself.⁷

What Brain Imaging Reveals

If benzodiazepines truly caused Alzheimer’s-type damage, you would expect long-term users to show more amyloid plaques in the brain and smaller hippocampi, the hallmark structural changes of the disease. A study from the MEMENTO cohort in France looked at exactly this using PET scans and MRI in older adults who did not yet have dementia. The findings were the opposite of what you would predict if the drugs caused neurodegeneration: chronic benzodiazepine users had lower brain amyloid levels and slightly larger hippocampal volumes than non-users, even after adjusting for age, sex, education, genetic risk factors, and psychiatric history.⁸

The researchers suggested this could reflect a protective mechanism: by calming overactive brain circuits, benzodiazepines might reduce neural activity and neuroinflammation in ways that slow amyloid buildup. This is a single study and far from definitive, but it stands in stark contrast to the narrative that these drugs are eating away at the brain. A large population-based study found a similar pattern at the epidemiological level, with overall benzodiazepine use showing no association with dementia risk compared to never-use.⁹

Proposed Biological Mechanisms

Even though the epidemiological case for causation is shaky, researchers have explored how benzodiazepines could theoretically help or harm the brain over time. Three mechanisms come up repeatedly in the literature. First, benzodiazepines can reduce the activity of enzymes involved in producing amyloid-beta, which could theoretically slow plaque formation, though this potential benefit has never been confirmed in clinical studies. Second, because brain cells near existing amyloid plaques may already be releasing extra GABA (the calming neurotransmitter that benzodiazepines enhance), people with preclinical dementia could be especially sensitive to the cognitive-dulling effects of these drugs, even at normal doses. Third, some people compensate for early brain damage by recruiting alternative neural networks. Benzodiazepines, by generally lowering brain activation, could undermine that compensatory ability and make cognitive problems surface sooner than they otherwise would.¹⁰

None of these mechanisms have been definitively established in humans. The first would suggest a protective effect, while the second and third describe ways the drug could reveal or accelerate existing disease rather than cause it from scratch. This fits the broader picture: benzodiazepines probably make early cognitive problems worse and more visible, which is different from triggering dementia in an otherwise healthy brain.

Does Duration or Dose Make It Worse?

If benzodiazepines directly caused dementia, you would expect a clear dose-response relationship: more drug, more risk, in a stepwise fashion. The evidence on this point is mixed. The meta-analysis of long-term users found that longer use was associated with modestly higher risk compared to short-term use, with about a 21 percent increase.¹ But a large retrospective cohort study of U.S. veterans found something that undermines the dose-response argument: veterans with low benzodiazepine exposure had essentially the same slightly elevated risk of dementia as veterans with high exposure, both around 5 to 6 percent above baseline. The risk did not climb with higher doses.¹¹

A population-based Dutch study similarly found no dose-dependent pattern overall, though one subgroup stood out: people who used the highest cumulative doses of anxiolytic benzodiazepines (drugs used for anxiety, as opposed to those used mainly for sleep) did show a modest, statistically significant increase in dementia risk.⁹ Whether that reflects something specific about anxiolytic benzodiazepines or simply the fact that people with severe, chronic anxiety tend to have more dementia risk factors is an open question.

Short-Acting Versus Long-Acting Drugs

Xanax (alprazolam) is a short-acting benzodiazepine, meaning it hits the brain quickly and wears off relatively fast. You might assume that longer-acting drugs like diazepam would be worse for cognition because they linger in the body, but research points in the opposite direction. One large Taiwanese study found that short-acting benzodiazepine users had roughly double the odds of dementia compared to long-acting users, even after adjusting for age, sex, depression, and other medications.¹² The MEMENTO brain-imaging study also noted that short-acting benzodiazepines were more strongly associated with larger hippocampal volumes than long-acting ones.⁸

One explanation for the short-acting pattern is that drugs like Xanax produce sharper peaks and troughs in brain chemistry, leading to more rebound anxiety and more frequent dosing, which keeps the cycle going. Another possibility is that the protopathic bias story is especially strong for short-acting drugs: when someone develops new anxiety from early dementia, a fast-acting drug is a natural first prescription. The elevated risk may say more about who gets prescribed these drugs than about what the drugs do.

Benzodiazepines Versus Z-Drugs for Sleep

People who take Xanax for insomnia sometimes wonder whether switching to a “Z-drug” like zolpidem or zopiclone would be safer for their brain. A study of middle-aged and older adults with chronic insomnia found that benzodiazepine exposure was an independent risk factor for cognitive impairment, while Z-drug use showed no correlation with global cognitive decline. Interestingly, Z-drug use was positively associated with attention performance.¹³ That said, the Taiwanese study mentioned above found that Z-drug users also had higher dementia odds than long-acting benzodiazepine users, so the picture is not straightforward.¹² Neither class of drug is considered a good long-term solution for sleep in older adults.

Cognitive Fog While on Benzos and Whether It Clears

Separate from the dementia question, benzodiazepines clearly impair memory, attention, and processing speed while you are taking them. This is not controversial. The drug works by enhancing GABA activity in the brain, which slows neural firing broadly. For someone taking Xanax daily, that slowdown can mimic early dementia symptoms: forgetting names, losing track of conversations, struggling with word retrieval. A memory clinic cohort study found that patients with moderate to high cumulative exposure to sedative and anticholinergic drugs (a category that overlaps heavily with benzodiazepine use) experienced steeper cognitive decline on standardized tests than patients with no exposure.¹⁴

The more hopeful finding is that at least some of this impairment reverses when the drug is stopped. A controlled study of elderly nursing home residents found that after gradual benzodiazepine discontinuation, memory and cognitive function improved compared to residents who stayed on the medication.¹⁵ However, another study tracking people for six months after withdrawal found only modest recovery of certain deficits, and cognitive performance remained below that of matched controls who had never used benzodiazepines.¹⁶ The picture that emerges is that stopping does help, but if you have been on benzodiazepines for years, you may not get all the way back to your cognitive baseline. Whether that residual gap reflects lasting drug effects or pre-existing differences between people who end up on long-term benzodiazepines and those who do not is, once again, unclear.

Why Doctors Are Cautious About Benzos in Older Adults

Even without definitive proof of a dementia link, there are strong reasons why prescribing guidelines discourage benzodiazepine use in people over 65. The American Geriatrics Society’s Beers Criteria, an influential list of medications considered potentially inappropriate for older adults, includes benzodiazepines.¹⁷ The reasons go beyond dementia risk and include increased fall risk, hip fractures, excessive sedation, respiratory depression (especially when combined with opioids), and a well-documented tendency toward physical dependence that makes stopping difficult.

For anxiety in older adults, clinical guidelines recommend antidepressants as first-line treatment. SSRIs and SNRIs with favorable drug-interaction profiles are generally preferred. Buspirone, a non-benzodiazepine anti-anxiety medication, is sometimes used, though there is less research on it in elderly populations specifically.¹⁸ Cognitive behavioral therapy, when accessible, is effective for anxiety and carries no drug-related risks at all.

How Tapering Works if You Want to Stop

If you have been on Xanax or another benzodiazepine for more than a few weeks, stopping abruptly is dangerous. Benzodiazepine withdrawal can cause seizures, severe rebound anxiety, insomnia, and other serious symptoms. A 2025 clinical practice guideline recommends gradual dose reductions, starting with cuts of 5 to 10 percent per step, with the pace generally not exceeding 25 percent reduction every two weeks. Some clinicians transition patients from a short-acting drug like Xanax to a longer-acting one like diazepam, which produces smoother blood levels and can make the taper more comfortable. The guideline emphasizes that tapering should be individualized and that patients need regular check-ins with each dose reduction.¹⁹

The timeline for a full taper can range from a few months to well over a year, depending on how long you have been on the drug, your dose, and how you tolerate the reductions. Rushing the process increases the chance of withdrawal symptoms and relapse. If you are considering tapering, the first step is a conversation with your prescribing doctor, not a sudden decision to skip doses.

The Anticholinergic Overlap

One complication in the research is that many people who take benzodiazepines also take other medications with anticholinergic properties, drugs that block the neurotransmitter acetylcholine. These include certain antihistamines, bladder medications, tricyclic antidepressants, and antipsychotics. Anticholinergic drugs have their own well-documented association with cognitive decline. The Irish Longitudinal Study on Ageing found that new use of strongly anticholinergic medications was associated with declines in recall scores over two years, while benzodiazepine and Z-drug use were not independently associated with cognitive decline in their models after adjusting for confounders.²⁰ The veterans study also found that once cumulative anticholinergic exposure was accounted for, the benzodiazepine-dementia link was minimal and did not increase with dose.¹¹

This matters practically because someone worried about their cognitive risk from Xanax might be on two or three other medications that individually contribute more to the problem. A medication review with a pharmacist or geriatrician that looks at total anticholinergic and sedative burden is more useful than fixating on any single drug.

When the Risk Gets Closest to a Diagnosis

The case-control analysis of Alzheimer’s disease and vascular dementia offers one of the clearest illustrations of how timing distorts the apparent risk. For Alzheimer’s, the odds ratio for benzodiazepine use beginning less than a year before diagnosis was 2.2. For vascular dementia, it was even higher at 3.3. But as the window between starting benzodiazepines and receiving a diagnosis stretched to two, three, or four years, the numbers fell to around 1.0 for both types of dementia.⁵ This is a strong signal that the drug-first-then-dementia sequence, in many cases, is actually dementia-first-then-drug. People with early unrecognized disease experience anxiety or agitation, get prescribed a benzodiazepine, and then receive a formal dementia diagnosis months later. The drug did not cause the disease; the disease caused the prescription.

This does not mean every instance of the association is explained by reverse causation. Some residual risk persists in studies with longer lag times. But the size of that residual risk is small enough that researchers disagree about whether it reflects a real drug effect, unmeasured confounders like cardiovascular disease and chronic insomnia, or simply the limits of observational study design. Until a randomized controlled trial answers the question definitively, and nobody is going to randomize thousands of people to years of benzodiazepine use to find out, the honest answer remains that we do not know for sure. What we do know is that the risk, if it exists, is considerably smaller than the headlines imply.

1
PubMed Central. Risk of Dementia in Long-Term Benzodiazepine Users: Evidence from a Meta-Analysis of Observational Studies
2
BMJ. Benzodiazepine use and risk of dementia: prospective population based study
3
PubMed Central. Benzodiazepine Use and the Risk of Dementia in the Elderly Population: An Umbrella Review of Meta-Analyses
4
PubMed Central. Benzodiazepine-related dementia risks and protopathic biases revealed by multiple-kernel learning with electronic medical records
5
PubMed. Benzodiazepine Use and Risk of Developing Alzheimer’s Disease or Vascular Dementia: A Case-Control Analysis
6
PubMed. A Systematic Review and Meta-Analysis of the Risk of Dementia Associated with Benzodiazepine Use, After Controlling for Protopathic Bias
7
PubMed. Clearing the confounding confusion: Benzodiazepines and the risk of dementia?
8
PubMed Central. Benzodiazepine use and neuroimaging markers of Alzheimer’s disease in nondemented older individuals: an MRI and 18F Florbetapir PET study in the MEMENTO cohort
9
PubMed Central. Benzodiazepine use in relation to long-term dementia risk and imaging markers of neurodegeneration: a population-based study
10
PubMed. The Benzodiazepine-Dementia Disorders Link: Current State of Knowledge
11
The Journals of Gerontology: Series A. Use of Benzodiazepines and Risk of Incident Dementia: A Retrospective Cohort Study
12
Neurotherapeutics. Benzodiazepines, z-Hypnotics, and Risk of Dementia: Special Considerations of Half-Lives and Concomitant Use
13
PubMed Central. Association Between Z Drugs Use and Risk of Cognitive Impairment in Middle-Aged and Older Patients With Chronic Insomnia
14
PubMed Central. Impact of cumulative exposure to anticholinergic and sedative drugs on cognition in older adults: a memory clinic cohort study
15
International Journal of Geriatric Psychiatry. Cognitive improvement following benzodiazepine discontinuation in elderly nursing home residents
16
PubMed. Lack of cognitive recovery following withdrawal from long-term benzodiazepine use
17
PubMed Central. American Geriatrics Society 2023 updated AGS Beers Criteria® for potentially inappropriate medication use in older adults
18
PubMed Central. Pharmacological Management of Anxiety Disorders in the Elderly
19
PubMed Central. Joint Clinical Practice Guideline on Benzodiazepine Tapering: Considerations When Risks Outweigh Benefits
20
PubMed. Cognitive decline associated with anticholinergics, benzodiazepines and Z-drugs: Findings from The Irish Longitudinal Study on Ageing (TILDA)