Three Doctors on Benzodiazepines and Dementia Risk

Nassir Ghaemi, MD, MPH; Eric B. Larson, MD, MPH; Donovan T. Maust, MD


September 08, 2020

This transcript has been edited for clarity.

Nassir Ghaemi, MD, MPH: Benzodiazepines are commonly used throughout the population in the United States, although in recent years discussions surrounding their safety have mostly focused on older adults.

Eric B. Larson, MD, MPH: Benzodiazepines are used extensively in the elderly population. Depending on what survey you're reading, probably 20%-30% of people are exposed to these drugs. Many of them are taking them on a daily basis or in multiple doses.

Donovan T. Maust, MD: Two significant studies examined the link between benzodiazepine use and dementia in the past several years. First is a Danish study recently published in the American Journal of Psychiatry. This study was unique in that they specifically looked at adults (N = 235,465) with a first-time hospital visit for an affective disorder — things like either major depression or bipolar disorder.

Larson: The Danish analysis confirms some more recent studies, including our own in 2016, which showed that when you look at a community population and try as best as you can to control for confounding by an indication or reverse causality, there is no association.

Maust: I'd say that is the second key analysis, the 2016 Kaiser Permanente study that came out of the state of Washington. It also examined the link between benzodiazepine exposure and cognitive decline and dementia using a very well-characterized population as part of an ongoing longitudinal study. It offered really very good data about cognitive status.

Larson: We essentially did a cohort study following 3434 participants, 65 years of age or older, without dementia upon entry for about 7 years on average. It showed that in our population, although there was a slight increase in the risk for dementia among people who take very low doses of benzodiazepines, there was no association with those people who took higher doses or even the highest doses. It essentially did not offer any support for benzodiazepines as being causative.

Ghaemi: These studies are always large and often observational, which means that we look at how patients do after they receive the medication. The problem with real-life observational data, of course, is that there are lots of other factors that aren't controlled in these studies.

Maust: The challenge with all of these studies is that you have to deal with confounding. For example, people think about dementia as a disease of cognition and memory. But we also know that there are significant changes in behavior that go along with it.

Larson: The main issue is that the indication for which the benzodiazepines would be prescribed could be an early symptom of dementia. And when you prescribe a drug for an early symptom, you're essentially increasing the risk that that drug will appear to be associated with the outcome.

Ghaemi: I honestly don't think these kinds of large studies are very useful. I don't think they add a lot. They have hundreds of thousands of patients and, in the case of Danish study, essentially the whole country. But when you have confounding factors affecting the results, it doesn't matter if you have 10, 10,000, or 10 million patients; the results are still going to be affected by those factors.

Maust: Depending on how you look at it, you could come to the faulty conclusion that it was the medication that led to or increased the risk for dementia, whereas really there were already changes going on in the brain. The anxiety in response to those changes is what led to the prescription of the benzodiazepine. So, essentially, the benzodiazepine is a flag for something bad brewing in this person's brain.

Ghaemi: You really cannot have anything near certainty in deciding whether benzodiazepines do or do not cause dementia based on these kinds of observational studies. It's all a matter of probability.

If we start at a neutral point and say that benzodiazepines may or may not cause dementia, I would think that the most important data are actually from animals. In animal studies, benzodiazepines usually are not harmful, usually do not cause neurotoxicity and don't kill neurons. In fact, they may be protective post-stroke in some studies.

Maust: In general, I would say prescribing benzodiazepines to older adults should be only done with great caution.

Larson: The best evidence that we have is that these drugs do not increase the risk for diseases like Alzheimer's disease and dementia.

Maust: I would say the best evidence in 2020 is that it does not seem that there's a strong link between benzodiazepine use and dementia.

Ghaemi: In general, I would say there's not an association between benzodiazepines and increased risk for dementia that's clinically meaningful or scientifically strong.

Nassir Ghaemi, MD, MPH, is a professor of psychiatry at Tufts Medical Center and a lecturer in psychiatry at Harvard Medical School. He is the author of several general-interest books on psychiatry.

Donovan T. Maust, MD, is a geriatric psychiatrist at the University of Michigan in Ann Arbor.

Eric B. Larson, MD, MPH, is a senior investigator with Kaiser Permanente Washington Health Research Institute and an attending physician in the Department of Internal Medicine at the University of Washington Medical Centers in Seattle.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.