What decades of sleep research say about dementia, apnea and sleeping pills

You’re 68. You’ve read that poor sleep raises the risk of dementia, and you’re awake at 3 a.m. most nights. The fix seems obvious: a nighttime pain reliever from the drugstore shelf, the kind labeled PM, to make sure you get your seven hours.

The sleep ingredient in many of those products is diphenhydramine or a related first-generation antihistamine. The American Geriatrics Society’s Beers Criteria, which flag medicines that adults 65 and older are generally best off avoiding, name exactly that group of sleep aids.

That gap between the worry and the remedy runs through this whole subject. The evidence linking sleep and dementia across decades is real but mostly observational. The trials of fixing sleep are humbler than the headlines. And the most practical decisions concern the medicines people take to sleep. How sleep helps store memories, and the contested idea that it rinses waste from the brain, belong to the coverage of sleep and memory; this article is about decades.

What 25 years of sleep data show about dementia

The strongest long-term evidence comes from Whitehall II, a study of British civil servants. In 2021 a team led by Séverine Sabia at Inserm in Paris and University College London reported on 7,959 participants who answered the same question across the decades: how many hours do you sleep on an average weeknight? Over a mean follow-up of nearly 25 years, 521 developed dementia, typically diagnosed in their late 70s.

Compared with seven hours, sleeping six hours or less was linked with higher dementia risk:

• At age 50: a hazard ratio of 1.22, meaning about 22% higher risk over the follow-up.
• At age 60: 1.37, or about 37% higher.
• At age 70: 1.24, but the range of plausible values included no difference, so the link at that age wasn’t statistically clear.

People who reported short sleep at 50, 60 and 70 alike had about 30% higher risk than consistent seven-hour sleepers. That number travels well in headlines. Its 95% confidence interval runs from 1.00 to 1.69, and a lower bound of exactly 1.00 sits on the line of no effect.

Self-reports are a weak measure, so a subgroup of 3,888 participants aged 60 to 83 wore a wrist accelerometer for nine days. What people said and what the device recorded correlated only modestly (0.41 on a scale where 1 is perfect agreement). Even so, those in the lowest third, sleeping less than about 6 hours 13 minutes, had a hazard ratio of 1.63.

Long sleep, often flagged as risky, showed no clear link in this study. The authors explain why earlier findings may have misled. Dementia’s brain changes begin years before diagnosis and can themselves alter sleep, a problem called reverse causation. They cite a Swedish cohort of 28,775 people 65 and older in which a link between long sleep and dementia disappeared once cases diagnosed in the first five years were set aside.

The study’s limits are plain. It is observational, so other differences between short and normal sleepers could explain part of the link. Participants were employed at recruitment and healthier than the general population. And because dementia develops so slowly, the authors note, it is hard to study with randomized trials. On the library’s evidence scale, short sleep as a dementia risk sits at positive evidence, contested interpretation. None of the guidelines discussed in this article presents better sleep as a way to prevent dementia.

Sleep apnea: a clearer link, modest trial results

In obstructive sleep apnea, the airway repeatedly narrows or closes during sleep, so breathing stops and starts and blood oxygen dips. The US Preventive Services Task Force (USPSTF) estimated from 2007-2010 data that moderate to severe sleep apnea affected 13% of men and 6% of women aged 30 to 70. It lists cognitive impairment among the outcomes associated with untreated sleep apnea, alongside heart disease, type 2 diabetes and motor vehicle crashes.

What the observational studies found

In a 2011 JAMA study led by Kristine Yaffe at the University of California, San Francisco, 298 women with an average age of 82 had overnight sleep studies. About five years later, 44.8% of those with sleep-disordered breathing had developed mild cognitive impairment, a measurable decline short of dementia, or dementia itself, compared with 31.1% of those without it. Measures of low oxygen tracked the risk. Measures of broken-up sleep and total sleep time did not, which points toward the breathing problem itself rather than sleep loss. The study included only older women.

A 2025 meta-analysis in GeroScience pooled 39 cohort studies. Sleep apnea was associated with about 33% higher risk of all-cause dementia, and insomnia with about 36% higher risk. The paper’s title says sleep disorders increase the risk, a stronger claim than cohort data can carry.

What happened when sleep apnea was treated

The trials are humbler. The standard treatment, continuous positive airway pressure (CPAP), uses a bedside machine and mask to keep the airway open with pressurized air. In the APPLES trial, published in 2012, 1,098 people with sleep apnea at five US centers used either real CPAP or a sham device for six months. A sham device looks and feels like the real thing but doesn’t deliver therapeutic pressure. The real CPAP group did better on one measure of executive function at two months, but not at six, and showed no benefit for attention or memory. What CPAP clearly improved was sleepiness, measured both by questionnaire and by how long people could stay awake in a lab test.

A 2024 meta-analysis in Sleep Medicine pooled 11 randomized trials with 923 patients, most treated for three months or less. Only one test improved, a timed task of switching between numbers and letters that measures mental flexibility. The authors called for longer studies.

A different kind of evidence looks more encouraging and proves less. A 2021 analysis of Medicare claims by Galit L. Dunietz and colleagues at the University of Michigan followed 53,321 beneficiaries 65 and older with diagnosed sleep apnea. Those who used positive airway pressure had 22% lower odds of a new Alzheimer’s diagnosis. People who get, and keep using, a breathing machine may differ in many ways from those who don’t, so claims data can’t separate the treatment from the patient.

Screening has the same gap. In 2022 the USPSTF issued an I statement, its term for insufficient evidence, on screening adults who have no symptoms, concluding it could not weigh the benefits against the harms. The statement covers screening, not the evaluation of people who report symptoms. Verdict: CPAP to protect thinking is measured, but the gains were small and short-lived; CPAP to prevent dementia hasn’t been tested in a trial.

The sleep medicines that work against the goal

The Beers Criteria‘s 2023 update, summarized for the public by the American Geriatrics Society’s Health in Aging Foundation, lists three groups of sleep medicines among the potentially inappropriate choices for adults 65 and older: the so-called Z-drugs zolpidem, zaleplon and eszopiclone; benzodiazepines such as lorazepam, alprazolam and temazepam; and over-the-counter sleep aids and cold remedies containing diphenhydramine or other first-generation antihistamines. What the list’s avoid label means, and why it isn’t a ban, is spelled out in the library’s guide to having a single pharmacist review all your medicines.

The case against these drugs in older adults starts with a poor trade. A 2005 meta-analysis in BMJ by Jennifer Glass and colleagues at the University of Toronto pooled 24 randomized trials of sleeping pills in 2,417 people aged 60 or older. Sleep improved by a small margin: about 25 minutes more per night and fewer awakenings. Adverse cognitive events were 4.78 times as common as on placebo, and daytime fatigue 3.82 times as common. The authors concluded that in people over 60 the benefits may not justify the risk.

Other documented harms are specific to drug groups:

• Z-drugs. In 2019 the Food and Drug Administration added a boxed warning to the labels of eszopiclone, zaleplon and zolpidem. It covers complex sleep behaviors, such as sleepwalking and sleep-driving, that can happen while a person isn’t fully awake. These are rare, but serious injuries and deaths have occurred. The eszopiclone label adds that next-day impairment can occur even when the person doesn’t feel impaired.
• Benzodiazepines and Z-drugs alike. A 2019 patient guide from the Health in Aging Foundation says older adults taking either group are about twice as likely to have falls and fractures, and twice as likely to have a car crash. Drowsiness behind the wheel is covered more fully in the article on fatigue.
• First-generation antihistamines. These block acetylcholine, a chemical messenger in the nervous system. A large 2019 English study in JAMA Internal Medicine, led by Carol A.C. Coupland, notes that such drugs can cause short-term confusion and memory loss in older people.

Sleep medicine guidelines line up with this. The American Academy of Sleep Medicine’s 2017 drug guideline rated every one of its recommendations weak, and suggested clinicians not use diphenhydramine for chronic insomnia. The 2025 joint guideline from the Department of Veterans Affairs and the Department of Defense lists benzodiazepines, diphenhydramine, trazodone and antipsychotics as not suggested for chronic insomnia, and tells clinicians to check for cognitive impairment and other sleep disorders before prescribing; its summary card doesn’t state how strong those recommendations are. Melatonin’s separate story is told in its own entry.

Do sleeping pills cause dementia? The question is still open

Here the evidence genuinely conflicts.

The case for concern. A 2014 BMJ study by Sophie Billioti de Gage and colleagues compared 1,796 people in Quebec diagnosed with Alzheimer’s disease with 7,184 similar people without it, all over 66. Benzodiazepine use that began at least five years before diagnosis was linked with 51% higher odds, and 43% after adjusting for anxiety, depression and insomnia. The link grew with exposure: 84% higher odds beyond 180 daily doses, and none below 91. Even these authors allowed that benzodiazepine use might partly be an early marker of a condition that leads to dementia.

The case against causation. A 2016 BMJ study by Shelly L. Gray and colleagues followed 3,434 people 65 and older in Seattle for an average of 7.3 years, using pharmacy records. Risk was modestly higher at low use but not at the highest use, and heavier users didn’t decline faster on cognitive tests. The authors wrote: “These results do not support a causal association between benzodiazepine use and dementia.”

Design explains much of the gap. One compared cases with controls and looked back at prescriptions; the other followed people forward and counted pharmacy fills. The pattern Gray’s team found, more risk at low doses, fits a familiar trap: anxiety, insomnia and low mood can be early symptoms of dementia, and those symptoms get prescriptions. Coupland’s team noted that some early symptoms, such as depression, can start up to 10 years before diagnosis.

The anticholinergic evidence is similarly mixed for sleep aids. Gray’s team reported in 2015 that the highest cumulative use of strong anticholinergics, with first-generation antihistamines among the most common, went with 54% higher dementia risk. The English study, with 58,769 people diagnosed with dementia and 225,574 controls, found clear associations for anticholinergic antidepressants, bladder drugs and several others, but no significant increase for antihistamines. Its authors suggested patients not taking their prescriptions might explain that. A prescription database also can’t see over-the-counter purchases, a limit worth keeping in mind for sleep aids.

Verdict: positive evidence, contested interpretation. No randomized trial has tested it, and the dispute is unresolved. The documented harms described above, from confusion to falls to next-day impairment, don’t depend on settling it.

CBT-I: the guideline first-line treatment for insomnia

If the pills are a poor trade, what do the guidelines name instead? The American College of Physicians was blunt in 2016: “ACP recommends that all adult patients receive cognitive behavioral therapy for insomnia (CBT-I) as the initial treatment for chronic insomnia disorder.” It rated that a strong recommendation based on moderate-quality evidence. Its second recommendation, rated weak and based on low-quality evidence, covers people for whom CBT-I alone hasn’t worked: a shared decision with the clinician about whether to add short-term medicine. The American Academy of Sleep Medicine gave multicomponent CBT-I a strong recommendation in 2021.

CBT-I isn’t a list of bedroom tips. The VA/DoD guideline describes its core parts: limiting time in bed to the time spent asleep, which builds sleep drive; re-training the bed as a cue for sleep rather than wakefulness; and working on a racing mind and mistaken beliefs about sleep. The AASM suggests clinicians not rely on sleep hygiene alone as a treatment.

One trial suggests a benefit beyond sleep. In a 2022 JAMA Psychiatry study at UCLA led by Michael R. Irwin, 291 adults aged 60 and older with insomnia but no current depression received two months of either CBT-I or sleep education. Over up to three years, 12.2% of the CBT-I group developed major depression, against 25.9% of the comparison group, roughly half the risk. The trial ran at a single site, more people dropped out of follow-up in the CBT-I group, and its outcome was depression, not memory or dementia. If depression brings thoughts of suicide, call or text the 988 Suicide and Crisis Lifeline at 988.

A 2025 trial from Concordia University in Montreal looked at older adults coming off sedative sleep medicines under a supervised withdrawal plan. Among 47 participants, both groups, with and without added CBT-I, succeeded in stopping and reported less severe insomnia; CBT-I added better-rated sleep quality. Neither group’s scores on cognitive tests changed. The trial was small, and no group stayed on the drugs for comparison.

Verdict: for chronic insomnia itself, US guidelines name CBT-I as the treatment of choice. CBT-I to protect memory or prevent dementia hasn’t been tested. The Health in Aging Foundation notes it can be delivered by a trained therapist or through self-guided apps and websites.

Questions to bring to a prescriber or pharmacist

None of the research above is a reason to change a sleep medicine on your own. The FDA makes one exception for a rare reaction to three specific drugs, described under the warning signs below, and even then it pairs stopping with an immediate call to a health care professional. The Health in Aging Foundation says never to stop a medication without first talking with your health care provider, and its guidance on benzodiazepines and Z-drugs warns that stopping suddenly can cause withdrawal, so reductions are made gradually. The decision about keeping, reducing or switching belongs to you and your prescriber together. These questions help make it a real conversation:

1. What is this medicine treating, and was it meant to be short-term?
2. It’s on the Beers list for my age group. What makes it the right choice for me anyway?
3. Could it explain daytime grogginess, unsteadiness or memory slips I’ve noticed?
4. Do any of my over-the-counter products, such as nighttime pain relievers or cold remedies, contain a sedating antihistamine?
5. Should my snoring or daytime sleepiness be checked with a sleep study before we treat the insomnia?
6. Is CBT-I available to me, in person or online, and does my insurance cover it?
7. If we change or reduce the medicine, what’s the plan, and which withdrawal effects should I watch for?

Signs that call for a quicker conversation

• Doing things while not fully awake after a Z-drug. The FDA’s 2019 alert tells patients who sleepwalk, sleep-drive or can’t remember activities done while on eszopiclone, zaleplon or zolpidem to stop taking it and contact their health care professional right away, and it tells clinicians not to prescribe these drugs to anyone who has had such an episode. Whether anything replaces the drug is a decision to make with the prescriber.
• Loud snoring, gasping or pauses in breathing that a bed partner notices, with daytime sleepiness. The National Heart, Lung, and Blood Institute advises raising these symptoms with a health care provider, since a sleep study may be needed to diagnose sleep apnea.
• New confusion, falls or unsteadiness in an older relative who takes a sleep aid. That combination is worth reporting to their prescriber the same day rather than saving for the next routine visit.
• Insomnia lasting months. Whether CBT-I fits, and where to get it, is a conversation to have with your clinician. If memory changes are what worry you, when they call for formal assessment is the subject of a later entry in this series.

The takeaway

The sleep-dementia link is real enough to take seriously and too observational to promise that any sleep fix protects the brain. What the evidence does support is concrete. Snoring with daytime sleepiness deserves a sleep study conversation. Chronic insomnia has a first-line treatment that isn’t a pill. And for adults 65 and older, the common sleep medicines, including the ones sold next to the aspirin, carry documented risks that a prescriber or pharmacist can weigh with you at the next medication review.

Sources

1. Nature Communications. “Sabia S, Fayosse A, Dumurgier J, et al. "Association of sleep duration in middle and old age with incidence of dementia." Nat Commun, 2021. PubMed. DOI: 10.1038/s41467-021-22354-2.” pubmed.ncbi.nlm.nih.gov
2. JAMA. “Yaffe K, Laffan AM, Harrison SL, et al. "Sleep-disordered breathing, hypoxia, and risk of mild cognitive impairment and dementia in older women." JAMA, 2011. PubMed. DOI: 10.1001/jama.2011.1115.” pubmed.ncbi.nlm.nih.gov
3. Sleep. “Kushida CA, Nichols DA, Holmes TH, et al. "Effects of continuous positive airway pressure on neurocognitive function in obstructive sleep apnea patients: The Apnea Positive Pressure Long-term Efficacy Study (APPLES)." Sleep, 2012. PubMed. DOI: 10.5665/sleep.2226.” pubmed.ncbi.nlm.nih.gov
4. Sleep Medicine. “Durtette A, Dargent B, Gierski F, et al. "Impact of continuous positive airway pressure on cognitive functions in adult patients with obstructive sleep apnea: A systematic review and meta-analysis." Sleep Med, 2024. PubMed. DOI: 10.1016/j.sleep.2024.08.019.” pubmed.ncbi.nlm.nih.gov
5. Sleep. “Dunietz GL, Chervin RD, Burke JF, Conceicao AS, Braley TJ. "Obstructive sleep apnea treatment and dementia risk in older adults." Sleep, 2021. PubMed. DOI: 10.1093/sleep/zsab076.” pubmed.ncbi.nlm.nih.gov
6. GeroScience. “Ungvari Z, Fekete M, Lehoczki A, et al. "Sleep disorders increase the risk of dementia, Alzheimer's disease, and cognitive decline: a meta-analysis." GeroScience, 2025. PubMed. DOI: 10.1007/s11357-025-01637-2.” pubmed.ncbi.nlm.nih.gov
7. US Preventive Services Task Force (JAMA). “"Screening for Obstructive Sleep Apnea in Adults: US Preventive Services Task Force Recommendation Statement." JAMA, 2022;328(19):1945-1950. PubMed. DOI: 10.1001/jama.2022.20304.” pubmed.ncbi.nlm.nih.gov
8. American Geriatrics Society (Journal of the American Geriatrics Society). “2023 American Geriatrics Society Beers Criteria Update Expert Panel. "American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults." J Am Geriatr Soc, 2023. PubMed. DOI: 10.1111/jgs.18372.” pubmed.ncbi.nlm.nih.gov
9. AGS Health in Aging Foundation. “Learn More: Alternatives for Medications Listed in the AGS Beers Criteria for Potentially Inappropriate Medication Use in Older Adults (last updated September 2025).” healthinaging.org
10. AGS Health in Aging Foundation. “Ask the Expert: Sedative-Hypnotic Drugs and Related Medications (July 2019).” healthinaging.org
11. BMJ. “Glass J, Lanctot KL, Herrmann N, Sproule BA, Busto UE. "Sedative hypnotics in older people with insomnia: meta-analysis of risks and benefits." BMJ, 2005. PubMed. DOI: 10.1136/bmj.38623.768588.47.” pubmed.ncbi.nlm.nih.gov
12. U.S. Food and Drug Administration. “Certain Prescription Insomnia Medicines: New Boxed Warning Due to Risk of Serious Injuries Caused by Sleepwalking, Sleep Driving and Engaging in Other Activities While Not Fully Awake (posted April 30, 2019).” fda.gov
13. U.S. Food and Drug Administration. “Eszopiclone tablets, prescribing information (revised 08/2019).” accessdata.fda.gov
14. JAMA Internal Medicine. “Gray SL, Anderson ML, Dublin S, et al. "Cumulative use of strong anticholinergics and incident dementia: a prospective cohort study." JAMA Intern Med, 2015. PubMed. DOI: 10.1001/jamainternmed.2014.7663.” pubmed.ncbi.nlm.nih.gov
15. JAMA Internal Medicine. “Coupland CAC, Hill T, Dening T, Morriss R, Moore M, Hippisley-Cox J. "Anticholinergic Drug Exposure and the Risk of Dementia: A Nested Case-Control Study." JAMA Intern Med, 2019. PubMed. DOI: 10.1001/jamainternmed.2019.0677.” pubmed.ncbi.nlm.nih.gov
16. BMJ. “Billioti de Gage S, Moride Y, Ducruet T, et al. "Benzodiazepine use and risk of Alzheimer's disease: case-control study." BMJ, 2014. PubMed. DOI: 10.1136/bmj.g5205.” pubmed.ncbi.nlm.nih.gov
17. BMJ. “Gray SL, Dublin S, Yu O, et al. "Benzodiazepine use and risk of incident dementia or cognitive decline: prospective population based study." BMJ, 2016. PubMed. DOI: 10.1136/bmj.i90.” pubmed.ncbi.nlm.nih.gov
18. American Academy of Sleep Medicine (Journal of Clinical Sleep Medicine). “Sateia MJ, Buysse DJ, Krystal AD, Neubauer DN, Heald JL. "Clinical Practice Guideline for the Pharmacologic Treatment of Chronic Insomnia in Adults." J Clin Sleep Med, 2017. PubMed. DOI: 10.5664/jcsm.6470.” pubmed.ncbi.nlm.nih.gov
19. U.S. Department of Veterans Affairs and Department of Defense. “VA/DoD Clinical Practice Guideline for the Management of Chronic Insomnia Disorder and Obstructive Sleep Apnea, 2025: Pocket Card.” healthquality.va.gov
20. American College of Physicians (Annals of Internal Medicine). “Qaseem A, Kansagara D, Forciea MA, Cooke M, Denberg TD. "Management of Chronic Insomnia Disorder in Adults: A Clinical Practice Guideline From the American College of Physicians." Ann Intern Med, 2016. PubMed. DOI: 10.7326/M15-2175.” pubmed.ncbi.nlm.nih.gov
21. American Academy of Sleep Medicine (Journal of Clinical Sleep Medicine). “Edinger JD, Arnedt JT, Bertisch SM, et al. "Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline." J Clin Sleep Med, 2021. PubMed. DOI: 10.5664/jcsm.8986.” pubmed.ncbi.nlm.nih.gov
22. JAMA Psychiatry. “Irwin MR, Carrillo C, Sadeghi N, Bjurstrom MF, Breen EC, Olmstead R. "Prevention of Incident and Recurrent Major Depression in Older Adults With Insomnia: A Randomized Clinical Trial." JAMA Psychiatry, 2022. PubMed. DOI: 10.1001/jamapsychiatry.2021.3422.” pubmed.ncbi.nlm.nih.gov
23. Sleep Medicine. “Barbaux L, Cross NE, Perrault AA, et al. "Effects of cognitive-behavioral therapy for insomnia during sedative-hypnotics withdrawal on sleep and cognition in older adults." Sleep Med, 2025. PubMed. DOI: 10.1016/j.sleep.2025.106826.” pubmed.ncbi.nlm.nih.gov
24. National Heart, Lung, and Blood Institute. “Sleep Apnea: Symptoms (last updated January 9, 2025).” nhlbi.nih.gov