Benzodiazepines in Older Adults: Alzheimer Risk, Memory, and Beers Criteria

Joseph F. Goldberg, MD, continues his conversation with Carl Salzman, MD, in this second installment of "Brain Trust: Conversations in Psychopharmacology," turning to benzodiazepine use specifically in older adults. The American Geriatrics Society’s Beers Criteria flags benzodiazepines as potentially inappropriate for patients 65 years and older, citing heightened sensitivity to sedative-hypnotic effects and elevated fall risk.¹ Observational studies have also raised questions about a possible link between benzodiazepine use and Alzheimer disease, though this association is increasingly attributed to protopathic bias rather than a causal effect.² Salzman revisits his own nursing home research on benzodiazepine discontinuation and memory, along with practical strategies for dosing, monitoring, and shared decision-making in geriatric patients. Salzman closed with the case for what he calls "mindful prescribing": matching benzodiazepine use to genuine clinical need rather than reflexive avoidance or reflexive prescribing.

This conversation is a continuation, see part 1 here.

Joseph F. Goldberg, MD: Two other things I want to ask you about before we wrap up. One is the American Geriatrics Society's Beers criteria, which goes like fire and brimstone: the second you hit age 65, that’s it—no more benzodiazepines for you. What do you have to say to the American Geriatrics Society on that?

Carl Salzman, MD: Funny you should ask—I might have an opinion about that. It is not an unreasonable suggestion that older patients are more sensitive to sedative-hypnotic drugs—they will have a bigger effect at standard doses—and the solution is not to throw the drugs out, but to use lower doses.

A clinical example: a colleague of mine asked me about his grandfather, who they thought was showing signs of dementia. My first question: what is he taking? It turned out he was taking Valium, 5 mg, 3 times a day. I said, let's slowly lower the dose, and then maybe even change it, or stop it entirely. And, of course, he was not getting demented; as the drug was discontinued, he kind of woke up. So there is no question that, at usual adult therapeutic doses, elderly patients might have more side effects, and some of those side effects might look like dementia, or something that looks like Alzheimer disease. Or they are sleeping too much, then up at night, and somebody gives them another pill, which just makes it worse. Or they are more likely to fall, which is true, and which spooks all the primary care doctors.

I will come back to falls in a minute, because it is important. Also, if they are still driving, they are more dangerous, and one of the jokes people in Florida tell is that you do not want to be driving around 5 to 6 in the afternoon, because that is when older people go out for the early-bird dinners, and they might smash into you. I don’t know whether that’s true, but that’s what you hear said.

So what do we actually know? Yes, benzodiazepines—long half-life and short half-life—may increase the risk of falls. If you look at the risk of falls with antidepressants or antipsychotics at therapeutic doses, the risk is actually greater than with benzodiazepines—not by much, but the point is it is not just benzodiazepines, it is any psychiatric medication, benzodiazepines among them. It is certainly true that long half-life drugs probably are not a good idea for older people, because their metabolism is slower, so the drugs last longer, accumulate, and the older person becomes unsteady and more likely to fall. Older individuals also sometimes have a nightcap, and if they are taking a benzodiazepine and alcohol at the same time, they are at risk. So it depends on the patient and where they are living—it is not automatic that the drugs are bad, but they are risky, and we should pay attention. Clinicians should do a fall-risk assessment: are you a candidate? Have you fallen? Have you broken a hip? What are your personal risks?

Goldberg: And a second point—do benzodiazepines cause Alzheimer disease?

Salzman: The study that suggested that idea was a very good study, out of Columbia. But it was a study of drug use, not of patients—no patients were actually seen. What it showed was that older patients taking drugs for Alzheimer disease, like Aricept, were also more likely to be taking a benzodiazepine, compared with a control group not taking Alzheimer disease drugs. So the conclusion was that there was an association between the drug and Alzheimer disease. The problem was that association was then interpreted as the benzodiazepine causing the dementia. They tried to control for that by looking at how many years before the diagnosis the benzodiazepine was taken, but the main point was that some people, when they think they are getting demented, get more anxious and start to take a benzodiazepine. So it could be the other way around—it is not the drug causing Alzheimer disease, it is Alzheimer disease causing the drug use. It’s like saying airbags cause more car accidents.

Since that time, several controlled research studies have clearly shown that benzodiazepines do not cause Alzheimer disease.

Goldberg: The term protopathic bias is something our viewers may or may not know—the idea that the ailment of interest, whether dementia, cancer, or anything else, may lead you to get a prescription along the way, as a marker of your inevitable course toward that ailment, without it being a causal phenomenon. It is like saying cigarette lighters cause lung cancer—they do not, but people on a course toward lung cancer, for another reason, are more likely to have a lighter, as an artifact. You are saying benzodiazepines would, by and large, be an artifact?

Salzman: Yes. But there is a point to that study, which I saw in my own research, because I was doing research in a very high-quality nursing home in the Boston area, looking at benzodiazepine use in the elderly. The question was whether the benzodiazepine affected cognition, memory primarily. We could not give people benzodiazepines unless they were really needed, and we could not use a placebo, but we could ask older residents whether they would be willing to taper and stop their benzodiazepines, to see how it affected their memory. It took us 3 years, but we got a small group willing to taper. At the end, they were tested, and on Wechsler memory testing, the people who had discontinued the drug had a marked improvement compared with those who stayed on it. It was huge—you did not need statistics to see the difference. I sat down to have lunch with these people the next day, after we had analyzed the data, and said to the ones who had discontinued and now had better memory, “Are not you glad we got rid of those terrible drugs, and you have a better memory?” They said, “No.” I said, “What do you mean?” They said, “It is true we can remember things a little better, but the kind of things we were forgetting was like what we watched on television the night before. If you are asking would we prefer to feel less anxious, sleep better, and have worse memory, or have better memory but still feel anxious during the day and maybe not sleep so well—we will take the memory loss any time, because what is there to remember? We can remember our grandchildren's names, our wedding date, all the things in the past.” The memory impairment is of recently acquired information—it is an anterograde amnesia. It is made worse by the benzodiazepine, which affects recent recall, depending on the dose, the patient's age, and their memory in general. But it does not cause a dementing illness that goes on—if you stop the benzodiazepine, the recall normalizes, but the patient does not want that. The patient would prefer to stay on the benzodiazepine.

Goldberg: Not everybody wants to stop, I guess—as long as they are not forgetting where they left the stove on or locked the door.

Salzman: That's right—or left the car running with the keys in it. That doesn’t really happen. Now, I want to emphasize that anxiety in older adults is very common, particularly in the very elderly, because their health is threatened, they are thinking about death and about the losses they have had, and relief of that anxiety is sometimes very helpful. So when you ask older people what they would like, sometimes they would like a little bit of help. It does not have to be a benzodiazepine—there are other ways of helping older adults. And you sure would not want to miss an older adult’s depression by treating only with a benzodiazepine, if someone is anhedonic—it is not just the existential “what is my legacy,” but rather, “I do not want to get up in the morning, and I have no reason to.”

So, Joe, because I know we are getting near the end—how I would see it is: I do not think prescribing benzodiazepines to elderly people is a great idea, unless I think it is really going to be helpful. I am very careful about that. I limit the use to lorazepam, by and large. Oxazepam (Serax) is also a good short half-life benzodiazepine, but it is not as rapidly absorbed as lorazepam. I make sure the person is not taking other sedative hypnotics, and that alcohol is not a problem. I try to combine this with a discussion with the patient, and maybe some short-term psychotherapy, about what is really upsetting, why they are anxious, and whether there are other ways of dealing with it besides a pill. For those who really need long-term use, I make sure that whoever is giving the pills keeps the doses low and supervises the taking of the benzodiazepine, and that the physician in charge follows the patient reasonably closely, so this is not just ongoing benzodiazepine use happening in a vacuum.

These drugs are okay when used appropriately, and when they are not used appropriately, they are not okay—it is clinical wisdom and care, and sometimes a little bit of psychotherapy can go a long way, rather than automatically using a pill. I am going to coin a term to end with: mindful prescribing.

Goldberg: Carl, I really want to thank you so much for sharing your vast wealth of knowledge, your wisdom, your experience—the years you have dedicated to thinking about and studying these problems and questions that I am sure many of our viewers and listeners wrestle with day in and day out. Whether you are aware of it or not, we have hopefully made this more mindfully front of mind.

Dr Goldberg is a clinical professor of psychiatry at The Icahn School of Medicine at Mount Sinai in New York, NY and the immediate-past president of the American Society of Clinical Psychopharmacology.

Dr Salzman is a professor of psychiatry at Harvard Medical School.

References

1. 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;71(7):2052-2081.

2. 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.