Do benzodiazepines increase the risk of dementia?

Published in Health News and Evidence

Date published: About this date

Clinical content may change after this date. This information is not intended as a substitute for medical advice from a qualified health professional. Health professionals should rely on their own expertise and enquiries when providing medical advice or treatment.

Practice points | New findings suggest a link | Chicken or the egg? | How reliable was the study? | Benzodiazepine use in Australia | Long acting vs short acting | What does this mean for clinical practice? | References


  • Long term use of benzodiazepines is associated with an increased incidence of dementia and cognitive impairment.
  • A case-control study of older Quebec residents has found that use of benzodiazepines daily for more than three months is associated with increased incidence of Alzheimer’s disease, even after adjusting for anxiety, depression and insomnia.
  • The link became stronger as exposure to benzodiazepines increased, with the highest incidence occurring in those who took benzodiazepines for more than six months.
  • To date, research demonstrates a link, but cannot prove cause or effect.
  • It remains unclear whether benzodiazepines increase the risk of dementia or are prescribed to combat pre-clinical symptoms.

Practice points

  • Prescribe the lowest dose of benzodiazepines for the shortest amount of time.
  • Review older patients’ medications regularly, particularly when they transition from hospital or aged-care facilities.
  •  In patients who have been taking benzodiazepines long term,  gradually reduce the dose over several months.

New findings suggest a link

In 2012, a prospective cohort study of more than 1000 elderly French people reported that those who had taken benzodiazepines experienced a 50% increased relative risk of dementia within 15 years of starting the medicine, compared with controls who had never taken benzodiazepines.1

Mounting observational evidence has also linked use of benzodiazepines to adverse cognitive effects.2 But limitations in study design and sample size have hampered the ability to draw reliable conclusions, and it has been unclear whether negative cognitive effects have a long term impact or are reversible when the drugs are stopped.

Now a Canadian case-control study published in the British Medical Journal in September 20143 by the same researchers as the French study has reinforced previous concerns1 of a link between benzodiazepines and dementia, suggesting a more permanent impact.

Using the Quebec health insurance program administrative claims database, researchers compared 1796 people aged over 66 years with a first diagnosis of Alzheimer’s disease with 7184 controls. The cases were all living in the community in Quebec, and each had been followed up for at least six years before diagnosis. Each case was matched by age, sex and length of follow up with four control patients who had no history of dementia. The researchers then used claims data to compare benzodiazepine use between the two groups.3

Benzodiazepine 'ever use' was associated with an increased risk of Alzheimer’s disease. The association became stronger as exposure increased. Thus when cumulative exposure was less than three months (< 91 prescribed daily doses, PDD), no association was found. But exposure between three months and six months (91–180 PDD) was associated with increased risk (OR 1.32, 95% CI 1.01 to 1.74), and a strong association was found with more than six months exposure (> 180 PDD; OR 1.84, 95% CI 1.62 to 2.08).3

What does it mean?

These findings translate to an increase in risk of:

  • 32% for people using benzodiazepines daily for three to six months
  • 84% for people using benzodiazepines daily for more than six months.3

Long-acting benzodiazepines with a half-life of 20 hours or more were also associated with higher risk than short-acting benzodiazepines.3 The authors say this adds to the evidence of a dose–response effect since daily use of long-acting benzodiazepines results in 'more or less constant blood and brain active concentrations'.3

Chicken or the egg?

To control for the risk that the drugs had been prescribed to combat early symptoms of Alzheimer’s disease (prodromal symptoms such as anxiety, depression, insomnia) only patients whose treatment began at least five years before diagnosis were included. In the sensitivity analysis this was pushed back to six years, but the association remained unchanged. There was also no significant interaction for anxiety, depression, or insomnia, and additional adjustment in the sensitivity analysis did not change the conclusions.3

Still, it cannot be completely ruled out that the anxiety or insomnia that prompted the use of benzodiazepines were not early symptoms of dementia. For example, anxiety and sleep disorders could be associated with early β amyloid lesions in the brain.4,5 The accompanying editorial calls for better monitoring in older adults to help clarify this.2

How reliable was the study?

The methodology was rigorous and the authors took care to minimise bias:2,3

  • Both the cases and matched control subjects were randomly selected from a well-defined and representative cohort living in the community in Quebec (Canada).
  • The authors used exposure windows of at least five years before diagnosis to reduce the likelihood that the benzodiazepines were prescribed to treat pre-clinical symptoms of dementia.
  • The authors adjusted for potential confounders (including underlying depression, anxiety and insomnia, as well as many other comorbidities).
  • Diagnoses were always made by a physician, and in many cases would have been confirmed in order for the patient to have received reimbursement for repeat prescriptions of benzodiazepines.
  • The study’s sponsor had no influence on the design and the conduct of the study.

However as with any case-control study, there is potential that other confounders may be present that have not been accounted for and reverse causation could still be possible.

The authors address several limitations:2,3

  • Potential for mis-classification since cases were classified according to diagnoses recorded in claims data.
  • Possible delay between the actual onset of Alzheimer’s disease and the date the diagnosis was recorded.
  • Benzodiazepines could possibly confound the clinical diagnosis of dementia by impairing cognitive function.
  • Neuropsychiatric symptoms used for adjustment – such as anxiety, depression, and sleep disorders – could have been under-reported if the physician did not consider them a main diagnosis.
  • The authors did not have access to information on socioeconomic status, education level, smoking or alcohol consumption, and so could not adjust for those factors.
  • Benzodiazepine use might be an early marker of a condition associated with increased risk of dementia, rather than the cause.2,3

Benzodiazepine use in Australia

Australian and international guidelines recommend short-term use of benzodiazepines, generally for no longer than four weeks.4,5

Despite this, benzodiazepines are widely prescribed for long-term disorders.6 About 15% of Australians over 65 use benzodiazepines,7 primarily to treat anxiety or insomnia. Females are twice as likely to be prescribed a benzodiazepine, and prevalence increases with age.7

A 2012 systematic review of 68 randomised double-blind placebo-controlled trials found benzodiazepines consistently induced both amnestic and non-amnestic cognitive impairments. As dose and duration of benzodiazepine use increased, incidence of both forms of cognitive impairment also went up. Notably, adults over 60 were more sensitive to low doses of benzodiazepines than younger participants.8

Long acting vs short acting

In a 2013 submission to the Therapeutic Goods Administration about a proposed schedule change, the Royal Australian College of Physicians wrote that 'short-acting benzodiazepines are more problematic than long-acting benzodiazepines owing to their fast acting component, which renders their potential to be more addictive and more toxic'.9

Australian guidelines also state that 'If long-term use is contemplated, specialist review and advice should be sought first and a drug with a long half-life should be chosen and the lowest effective dose should be prescribed'.10

However this most recent study found a stronger association with Alzheimer’s disease for long-acting benzodiazepines (OR 1.70; 95% CI 1.46 to 1.98) than for short-acting ones (OR 1.43; 95% CI 1.27 to 1.61).3

The authors conclude that 'benzodiazepine prescription in older people should comply with good practice guidelines – that is, the shortest duration with a preference for formulations with a short half-life'.3

What does this mean for clinical practice?

Current guidelines already advise caution in using benzodiazepines in older people since they are more prone to adverse effects and toxicity, and have increased risk of ataxia, falls and injuries, memory loss and cognitive impairment from benzodiazepines.10

Likewise, due to the risk of dependence it is already generally recommended that benzodiazepines be used only for short periods of less than four weeks.4

This study adds weight to the evidence that benzodiazepines could be a modifiable risk factor for dementia, but it still does not close the debate. However it highlights the importance of following the current recommendations, including regularly reviewing patients' drug regimens.

Benzodiazepines are often started in hospital or at aged-care facilities where non-pharmacological behavioural approaches can be more challenging to implement. They may then be continued inadvertently. Review medicines regularly, especially when a patient transitions back into the community. In patients who have been taking benzodiazepines long term, gradually reduce doses over several months.10

For more information on 'deprescribing' drugs in older patients, see: Deprescribing (Australian Prescriber)

  1. Billioti de Gage S, Bégaud B, Bazin F, Verdoux H, Dartigues JF, Pérès K, Kurth T, Pariente A. Benzodiazepine use and risk of dementia: prospective population based study. BMJ 2012;345. [Online]
  2. Yaffe K and Boustani M. Benzodiazepines and risk of Alzheimer's disease. BMJ 2014;349:g5312. [PubMed]
  3. 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. [PubMed]
  4. Baldwin DS, Aitchison K, Bateson A, et al. Benzodiazepines: risks and benefits. A reconsideration. J Psychopharmacol 2013;27:967–71. [PubMed]
  5. Therapeutic Guidelines Limited. Chapter Psychotropic: Psychotropic treatment in older people. 2013. [Online] (accessed 1 October 2014)
  6. Stephenson CP, Karanges E and McGregor IS. Trends in the utilisation of psychotropic medications in Australia from 2000 to 2011. Aust N Z J Psychiatry 2013;47:74–87. [PubMed]
  7. Windle A, Elliot E, Duszynski K, et al. Benzodiazepine prescribing in elderly Australian general practice patients. Aust N Z J Public Health 2007;31:379–81. [PubMed]
  8. Tannenbaum C, Paquette A, Hilmer S, et al. A systematic review of amnestic and non-amnestic mild cognitive impairment induced by anticholinergic, antihistamine, GABAergic and opioid drugs. Drugs Aging 2012;29:639–58. [PubMed]
  9. The Royal Australian College of Physicians. Rescheduling of Benzodiazepines from Schedule 4 to Schedule 8 2013. [Online] (accessed 3 October 2014)
  10. Therapeutic Guidelines Limited. Psychotropic treatment in older people. 2013. [Online] (accessed 1 October 2014)