Sleeping pills, insomnia and dementia: what the research says

All sleep articles

By Small Hours · Last reviewed

Two people seen from behind walking side by side along a forest path in autumn morning light, fallen leaves underfoot

These are two different questions, and the research answers them differently. Poor sleep and later dementia go together in large pooled studies, but when insomnia is measured carefully the link weakens or disappears, and the leading expert review does not count sleep as a proven risk factor. [1][4][7] Long-term use of benzodiazepine sleeping pills is linked with dementia in several studies and not in others, and the same expert review concludes that these drugs "do not appear to cause dementia". [7][8][11] Nothing here proves that either insomnia or a sleeping pill causes dementia. Please do not stop or change a sleeping medicine on your own. [16]

If you have lain awake worrying that your insomnia is damaging your brain, this article is for you. It is also for anyone who has read a headline about sleeping pills and dementia and wants to know what the studies behind it actually found. Every study here watched people over time and counted who later developed dementia. That design can show that two things go together. It cannot show that one caused the other, and the authors of these studies say so themselves. [6][7][8]

What the research says

Pooled studies find that people who report poor sleep are somewhat more likely to be diagnosed with dementia years later, but the link is modest and inconsistent for insomnia itself. A meta-analysis of 18 follow-up studies (246,786 people, average follow-up 9.5 years) found that sleep disturbance in general was associated with later dementia, and that insomnia specifically was associated with Alzheimer's disease but not with vascular dementia or dementia overall. [1] Two large studies that measured insomnia carefully found no association at all: 7,492 Norwegians followed for 11 years, and 6,851 older French adults followed for 12 years. [4][5] What the research says about pills is similar in shape. A Quebec study found Alzheimer's disease was about 1.5 times as likely in older people who had used benzodiazepines, with a stronger link after more than 180 daily doses, but a Seattle study found the heaviest users had no higher rate than non-users, and a Danish study of 235,465 patients found no link for benzodiazepines or Z-drugs. [8][10][11] The Lancet Commission on dementia, reviewing all of this in 2024, left sleep off its list of 14 modifiable risk factors and wrote that benzodiazepines "do not appear to cause dementia". [7] In plain terms, both links are associations, not proven causes, and the pill link is as unsettled as the insomnia one.

The studies behind this
  • Meta-analysis of sleep disturbance and dementia (Shi, 2018): 18 follow-up studies, 246,786 people, 25,847 dementia cases after an average of 9.49 years. People who reported sleep disturbances had a higher rate of all-cause dementia, Alzheimer's disease and vascular dementia. In the subgroup analysis, insomnia was associated with Alzheimer's disease but not with vascular or all-cause dementia, whereas sleep-disordered breathing was associated with all three. Sleep was mostly self-reported, and the authors said the results should be further validated. [1]
  • Updated meta-analysis (Xu, 2020): 51 follow-up studies covering 15 kinds of sleep problem. Insomnia, fragmented sleep, daytime dysfunction, long time to fall asleep, REM sleep behaviour disorder and excessive time in bed were linked to higher rates of cognitive decline or dementia, with moderate-to-high levels of evidence. Both short and long sleep carried higher risk than mid-range sleep, a U-shape. The authors wrote that sleep management "might serve as a promising target", not that it prevents dementia. [2]
  • Whitehall II (Sabia, 2021): 7,959 British civil servants, 521 dementia diagnoses, 25 years of follow-up. Sleeping six hours or less at age 50 or 60, compared with seven hours, was associated with a higher dementia rate (hazard ratios 1.22 and 1.37). Persistent short sleep at 50, 60 and 70 was associated with a 30% higher rate, independent of social, behavioural, cardiometabolic and mental-health factors. This measured sleep duration, not insomnia. [3]
  • The HUNT study (Selbaek-Tungevåg, 2023): 7,492 Norwegians followed for 11 years, with insomnia defined by diagnostic criteria rather than a single question. Probable insomnia disorder was not associated with all-cause dementia (odds ratio 1.03, 95% CI 0.74 to 1.43), Alzheimer's disease (1.07, 0.71 to 1.60) or cognitive test scores. [4]
  • The Three-City study (Cavaillès, 2022): 6,851 French adults aged 65 and over, free of dementia at the start, followed for 12 years. Insomnia complaints and poor sleep quality were not associated with later dementia. Persistent use of sleeping pills was associated with Alzheimer's disease (hazard ratio 1.28, 1.04 to 1.58), and daytime sleepiness with dementia that had a vascular component. [5]
  • Lancet Commission (Livingston, 2020 and 2024): the most-cited expert review of dementia prevention. In 2020 it listed 12 modifiable risk factors and described sleep as "putative", noting that both more sleep and less sleep are associated with dementia, which does not fit a simple story, and that new sleep problems a few years before a dementia diagnosis "might be part of the natural history of the dementia syndrome". In 2024 it listed 14 factors, said sleep was considered but did not meet its bar of consistent evidence, and concluded: "We are unable to make recommendations on sleep as a risk factor." [6][7]
  • Quebec benzodiazepine study (Billioti de Gage, 2014): 1,796 people aged over 66 with a first Alzheimer's diagnosis, matched with 7,184 controls, using Quebec health insurance records. Benzodiazepine use started at least five years before diagnosis was associated with higher odds of Alzheimer's (odds ratio 1.51, 1.36 to 1.69). There was no association below 91 prescribed daily doses, and the association was strongest above 180 daily doses (1.84) and with long-acting drugs (1.70). The authors wrote that the drug "might also be an early marker of a condition associated with an increased risk of dementia". [8]
  • PAQUID (Billioti de Gage, 2012): 1,063 older French adults followed for 15 years. Starting benzodiazepines during the study was associated with a higher dementia rate (hazard ratio 1.60, 1.08 to 2.38). [9]
  • Adult Changes in Thought (Gray, 2016): 3,434 Seattle-area adults aged 65 and over, followed for an average of 7.3 years, with pharmacy records of ten years of use. The lightest users had a slightly higher dementia rate (1.25), but the heaviest users did not (1.07, 0.82 to 1.39), and heavier use was not associated with faster cognitive decline. The authors concluded the results "do not support a causal association". [10]
  • Danish national registers (Osler and Jørgensen, 2020): 235,465 patients who had been seen in hospital for depression or bipolar disorder, three quarters of whom had used benzodiazepines or Z-drugs, followed for a median of 6.1 years. Use of either drug family showed no association with dementia after adjustment, and the heaviest users had the lowest odds (0.83). This is the largest study in the set and the one that included Z-drugs explicitly, but everyone in it had a mood disorder. [11]
  • Meta-analysis with lag times (Penninkilampi and Eslick, 2018): 15 studies, 159,090 dementia cases. Ever use of benzodiazepines was associated with higher odds of dementia (1.39, 1.21 to 1.59). In the studies that ignored prescriptions from the five years before diagnosis, the estimate shrank only slightly (1.30, 1.14 to 1.48), which the authors read as evidence that the link is "not purely an artefact" of early-symptom prescribing. [12]
  • Meta-analysis with evidence grading (Lucchetta, 2018): 12 studies, about 980,000 people. Pooled odds ratio 1.38 (1.07 to 1.77), with 98% heterogeneity between studies and a prediction interval of 0.58 to 3.25, so wide that it includes no effect. The authors rated the certainty of the evidence as very low and said it could not separate Alzheimer's from vascular dementia. [13]
  • Finnish national study (Tapiainen, 2018): 70,719 people with a verified Alzheimer's diagnosis and 282,862 matched controls, with a five-year gap between exposure and diagnosis. Benzodiazepines and related drugs together carried only 6% higher odds (1.06, 1.04 to 1.08), with no major difference between benzodiazepines and Z-drugs. A dose-response pattern disappeared once other psychiatric medicines were taken into account. [14]
  • Anticholinergic drugs (Gray, 2015): the same Seattle population of 3,434 older adults, looking at strong anticholinergic medicines, a group whose most-used members were tricyclic antidepressants, first-generation antihistamines and bladder medicines. Cumulative use showed a dose-related association with dementia; more than 1,095 standardized daily doses over ten years, roughly three years of daily use, carried a hazard ratio of 1.54 (1.21 to 1.96). The most recent year of use was excluded to avoid counting prescriptions for early symptoms. [15]
  • Canadian deprescribing guideline (Pottie, 2018): lists cognitive impairment and memory disorders among the harms associated with benzodiazepines and Z-drugs, on observational evidence mostly from older adults, and concludes that their use for insomnia in aging adults "has been associated with falls, dementia, motor vehicle accidents, and physical addiction". It recommends offering a slow, supervised taper to adults 65 and over. [16]
  • CBT-I and cognition (Siengsukon, 2026): 200 cognitively normal older adults with insomnia symptoms were randomised to CBT-I or a control condition and followed for a year. CBT-I did not improve cognitive test scores or brain amyloid. The authors said longer follow-up is needed. [17]

Question one: does insomnia raise the risk of dementia?

The honest answer is: poor sleep and dementia are associated in large pooled studies, the association is modest, and it is not clear which way it runs. Three things make the question harder than a headline suggests.

  • "Sleep problems" is not the same as insomnia. The pooled studies lump together short sleep, long sleep, poor sleep quality, breathing problems during sleep and insomnia. [1][6] When the pieces are separated, breathing problems carry the most consistent link, and insomnia the least. The 2018 meta-analysis found insomnia associated with Alzheimer's disease only, not with dementia overall. [1] The two large studies that measured insomnia carefully, in Norway and France, found no association at all. [4][5]
  • Dementia may cause the poor sleep, rather than the reverse. Dementia develops over many years before it is diagnosed, and disturbed sleep is a common early feature. A study that measures sleep in the few years before diagnosis may be watching an early symptom and mistaking it for a cause. Both Lancet Commission reports say this plainly: the 2020 report wrote that new sleep disturbance in late life "might be part of the natural history of the dementia syndrome, appearing to be a risk factor", and the 2024 report notes a Swedish study in which the link between long sleep and dementia disappeared entirely once cases from the first five years were excluded. [6][7]
  • The best evidence against that explanation is about sleep length, not insomnia. The Whitehall II study measured sleep at age 50 and found a link with dementia diagnosed up to 25 years later, too long a gap for early symptoms to explain. But it measured how long people said they slept, not whether they had insomnia. [3] Short self-reported sleep is also not the same thing as lying awake: a person with insomnia may be in bed for eight hours and sleep for six, and these studies asked about hours of sleep, not time awake in bed.

The 2024 Lancet Commission, which sets a deliberately high bar before naming a risk factor, considered sleep and did not include it, alongside diet and infections. Its summary reads: short sleep "may be associated with a small, increased risk of dementia", prolonged sleep is not a risk factor and may be a consequence of early dementia, and "People should not curtail their sleep to reduce dementia risk." [7]

Question two: do sleeping pills raise the risk of dementia?

Here the picture is a genuine disagreement between well-run studies, and it is worth seeing both sides.

Studies that found a link. The Quebec study is the most-quoted, partly because the link grew with the amount taken: no association under 91 daily doses, and the odds of Alzheimer's 1.84 times higher above 180. [8] The earlier French study found a 60% higher dementia rate in people who started benzodiazepines. [9] The meta-analyses that pool these and others land at odds ratios of about 1.4, and one of them showed the link shrinks only a little when prescriptions from the five years before diagnosis are ignored. [12][13] In the French Three-City study, persistent use of sleeping pills was associated with Alzheimer's disease even though insomnia complaints were not. [5]

Studies that did not. The Seattle study had pharmacy records for a full decade and found that the heaviest users had no higher dementia rate than non-users. If the drug were the cause, more of it should mean more risk, and in that study more of it did not mean more risk. [10] The Danish study, the largest, found no association for either benzodiazepines or Z-drugs, and the heaviest users had the lowest odds. [11] The Finnish study found an association so small (6% higher odds) that its dose-response pattern vanished once antidepressants and antipsychotics were accounted for; the authors suggested those other drugs, or the conditions they treat, might explain part of the link. [14]

Why they might disagree. Several authors name the same problem. Sleeping pills are prescribed for poor sleep and anxiety, and both are early features of dementia. So a person in the early, undiagnosed stage of dementia is more likely to be given a sleeping pill, and the pill then looks like a cause. The Quebec authors wrote that the drug "might also be an early marker of a condition associated with an increased risk of dementia". [8] The Seattle researchers excluded the last year of use for exactly this reason. [10] One meta-analysis argues the link survives that correction; another rates the whole body of evidence as very low certainty. [12][13] Reading all of it, the 2024 Lancet Commission concluded: "Benzodiazepines do not appear to cause dementia." [7]

What about Z-drugs and over-the-counter sleep aids? The evidence is thinner. Z-drugs such as zopiclone and zolpidem were included in the Danish study, which found no link, and in the Finnish study, which found no difference between them and benzodiazepines. [11][14] We found no study of zopiclone on its own. First-generation antihistamines, the drowsy ingredient in many non-prescription sleep aids, belong to a family of medicines with strong anticholinergic effects, and in the Seattle population about three years of daily use of that family was associated with a 54% higher dementia rate. [15] That study reported the family as a whole, not the sleep aids on their own.

What this does and does not mean for you

  • If you have insomnia: the research does not show that it is giving you dementia. The careful studies found no link, and the leading expert body declines to call sleep a risk factor. [4][5][7] Chronic insomnia is worth treating because of what it does to your days, not out of fear for your brain decades from now. CBT-I is the first-line treatment in major guidelines.
  • If you take a sleeping pill: the research does not show that it is giving you dementia either. It does show that long-term use of benzodiazepines and Z-drugs carries recognised harms in the present, including memory problems, falls and car accidents, mostly documented in older adults, which is why Canada's family-medicine guideline recommends that people aged 65 and over be offered a supervised, gradual reduction. [16] Whether and how to do that is a conversation with your prescriber; see coming off sleeping pills.
  • Does treating insomnia protect the brain? Nobody knows yet. The one randomised trial built to find out gave CBT-I to 200 older adults and found no change in cognitive tests or brain amyloid after a year; its authors say longer follow-up is needed. [17]
  • Worry about this is itself a sleep problem. Lying awake at 3 a.m. rehearsing a frightening headline keeps the alarm system on. If this fear is a nightly companion, it belongs in the conversation with your doctor or therapist as much as the insomnia does.

For clinicians

The two questions a patient conflates have different evidence bases. For insomnia disorder, the two studies with the longest follow-up and the most careful case definition (HUNT, Three-City) are null, the pooled estimates are driven by broadly defined sleep disturbance and sleep-disordered breathing, and the 2024 Lancet Commission explicitly declines to list sleep. [1][4][5][7] For benzodiazepine receptor agonists, the association is present in case-control and some prospective designs and absent in the two largest prospective datasets with the best exposure measurement; protopathic bias and confounding by indication are the explanations the authors themselves favour, and the lag-time meta-analysis is the strongest counter-argument. [8][10][11][12] The Canadian deprescribing guideline's recommendation for adults 65 and over rests on documented present-day harms, not on the dementia question. [16] Small Hours has a page for clinicians whose patients use the program.

Where the evidence is thin

  • Every study here is observational. None is a randomised trial, so none can show cause. [6][7]
  • "Sleep disturbance" in the meta-analyses was mostly self-reported and defined differently from study to study. [1][2]
  • Two careful studies of insomnia itself are null, but they are two studies, and their authors call for longer follow-up. [4][5]
  • The Danish study, the largest on sleeping pills, was entirely in people with mood disorders. [11]
  • Z-drugs have far less evidence than benzodiazepines, and we found no study of zopiclone on its own. [11][14]
  • The antihistamine finding is for a whole family of anticholinergic drugs, not for over-the-counter sleep aids specifically. [15]
  • One year is too short to see a dementia outcome, so the CBT-I trial answers a narrower question than the one people ask. [17]

Where does this leave you?

If insomnia has lasted months, the reason to treat it is the life you are living now, and the first-line treatment in major guidelines is CBT-I. Small Hours turns a nightly sleep diary into a six-week CBT-I program built around your own sleep. It does not give medication advice, and it works alongside the care you already have. You can sign up at trysmallhours.com.

When should you talk to a doctor?

  • Before changing any sleeping medicine. Stopping some of them suddenly can be dangerous, and the decision belongs with your prescriber. [16]
  • If you or your family notice memory or thinking changes that are new, regardless of how you sleep.
  • Loud snoring, breathing pauses or heavy daytime sleepiness, which point to sleep apnea rather than insomnia.
  • Low mood, anxiety or fear about your health that is keeping you awake. If you are in distress in Canada, call or text 9-8-8 at any hour.

General education, not medical advice.

References

  1. Shi L, Chen SJ, Ma MY, et al. Sleep disturbances increase the risk of dementia: a systematic review and meta-analysis. Sleep Med Rev. 2018;40:4-16. Source for reference 1 (opens in a new tab)
  2. Xu W, Tan CC, Zou JJ, Cao XP, Tan L. Sleep problems and risk of all-cause cognitive decline or dementia: an updated systematic review and meta-analysis. J Neurol Neurosurg Psychiatry. 2020;91(3):236-244. Source for reference 2 (opens in a new tab)
  3. Sabia S, Fayosse A, Dumurgier J, et al. Association of sleep duration in middle and old age with incidence of dementia. Nat Commun. 2021;12(1):2289. Source for reference 3 (opens in a new tab)
  4. Selbaek-Tungevåg S, Selbaek G, Strand BH, et al. Insomnia and risk of dementia in a large population-based study with 11-year follow-up: the HUNT study. J Sleep Res. 2023;32(4):e13820. Source for reference 4 (opens in a new tab)
  5. Cavaillès C, Berr C, Helmer C, et al. Complaints of daytime sleepiness, insomnia, hypnotic use, and risk of dementia: a prospective cohort study in the elderly. Alzheimers Res Ther. 2022;14(1):12. Source for reference 5 (opens in a new tab)
  6. Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413-446. Source for reference 6 (opens in a new tab)
  7. Livingston G, Huntley J, Liu KY, et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. Lancet. 2024;404(10452):572-628. Source for reference 7 (opens in a new tab)
  8. Billioti de Gage S, Moride Y, Ducruet T, et al. Benzodiazepine use and risk of Alzheimer's disease: case-control study. BMJ. 2014;349:g5205. Source for reference 8 (opens in a new tab)
  9. Billioti de Gage S, Bégaud B, Bazin F, et al. Benzodiazepine use and risk of dementia: prospective population based study. BMJ. 2012;345:e6231. Source for reference 9 (opens in a new tab)
  10. Gray SL, Dublin S, Yu O, et al. Benzodiazepine use and risk of incident dementia or cognitive decline: prospective population based study. BMJ. 2016;352:i90. Source for reference 10 (opens in a new tab)
  11. Osler M, Jørgensen MB. Associations of benzodiazepines, Z-drugs, and other anxiolytics with subsequent dementia in patients with affective disorders: a nationwide cohort and nested case-control study. Am J Psychiatry. 2020;177(6):497-505. Source for reference 11 (opens in a new tab)
  12. Penninkilampi R, Eslick GD. A systematic review and meta-analysis of the risk of dementia associated with benzodiazepine use, after controlling for protopathic bias. CNS Drugs. 2018;32(6):485-497. Source for reference 12 (opens in a new tab)
  13. Lucchetta RC, da Mata BPM, Mastroianni PC. Association between development of dementia and use of benzodiazepines: a systematic review and meta-analysis. Pharmacotherapy. 2018;38(10):1010-1020. Source for reference 13 (opens in a new tab)
  14. Tapiainen V, Taipale H, Tanskanen A, et al. The risk of Alzheimer's disease associated with benzodiazepines and related drugs: a nested case-control study. Acta Psychiatr Scand. 2018;138(2):91-100. Source for reference 14 (opens in a new tab)
  15. Gray SL, Anderson ML, Dublin S, et al. Cumulative use of strong anticholinergics and incident dementia: a prospective cohort study. JAMA Intern Med. 2015;175(3):401-407. Source for reference 15 (opens in a new tab)
  16. Pottie K, Thompson W, Davies S, et al. Deprescribing benzodiazepine receptor agonists: evidence-based clinical practice guideline. Can Fam Physician. 2018;64(5):339-351. Source for reference 16 (opens in a new tab)
  17. Siengsukon CF, Hand LK, Nelson E, et al. The impact of cognitive behavioral therapy for insomnia on cognitive performance and amyloid beta in older adults: a randomized controlled trial. Alzheimers Dement. 2026;22(6):e71591. Source for reference 17 (opens in a new tab)

This article is general education, not medical advice. It cannot diagnose anything or replace a conversation with your doctor. Read how we choose and check sources on the how we write page.

Share this article

Know someone who is lying awake? Send it to them.

Make the small hours smaller

Small Hours is a free sleep diary with a six-week CBT-I program behind it, built on the same research. Sign up tonight, and fill in your first two-minute diary tomorrow morning.

Create an account

Small Hours is a self-guided program for adults. It is not an emergency service. If you are in crisis, call or text 9-8-8 in Canada or the US at any hour.