This week, Mad in America examines four new studies around harms linked to benzodiazepine use. The first finds that benzodiazepines are linked to increased risk of heart failure and heart failure related hospitalization. The second reports that long-term benzodiazepine use (60 days or longer) is associated with worse cognitive function in people diagnosed with schizophrenia. The third finds that taking these drugs during pregnancy is linked to an increased risk of miscarriage. The fourth links benzodiazepines to increased risk of falls and fractures in older adults.
Benzodiazepines Linked to Increased Risk of Heart Failure
A new study published in Cardiovascular Toxicology finds that benzodiazepine use is linked to higher risk of heart failure and heart failure related hospitalization in people prescribed these drugs to treat insomnia. This research, led by Yi-Sin Huang from Taipei Medical University in Taiwan, also suggests that benzodiazepines are associated with a higher risk of heart failure hospitalization than other drugs used to treat insomnia.
The goal of this study was to examine the link between heart failure and benzodiapepine use in people with insomnia. The authors also wanted to compare heart failure outcomes between benzodiapepines and other drugs prescribed for insomnia. They conducted a systematic review and meta-analysis of previous research on this topic. In total, they used data from 11 studies.
Benzodiazepine use was linked to a 41% increased risk of heart failure and a 29% increased risk of heart failure related hospitalization compared to non-use. The risk of heart failure related hospitalizations in people taking these drugs to treat insomnia was also elevated compared to other drugs, including Z-drugs, suvorexant,and antipsychotics. These increased risks were particularly elevated in people that had a history of heart failure and those at high-risk.
This research had three main limitations. Most of the included studies were observational, meaning the findings can speak to a link between bezodiazepines and heart failure, but cannot definitively say that these drugs cause heart failure. Only 11 studies were included in the current work. The included studies varied significantly in follow-up range, methods, and patient populations.
Benzodiazepines Linked to Worse Cognitive Functioning
A new study published in BMC Psychiatry finds that long-term benzodiazepine use is linked to worse cognitive functioning in people diagnosed with schizophrenia. This research, led by Caiping Liu from the Shanghai Jiao Tong University School of Medicine in China, also reports that there was no significant difference in terms of cognition between low- and high-dose users. This means that while any long-term benzodiazepine use was associated with poorer cognition, the current work did not find that higher doses were linked to poorer cognitive function.
The authors used data from the SALT-C study in China, a research project around atypical antipsychotic use in people diagnosed with schizophrenia. They used data collected from 114 SALT-C participants recruited at the Shanghai Mental Health Center between 2016 and 2019. For the purposes of analysis the authors divided participants into two groups, people that had never taken benzodiazepines before and those that had taken a stable dose for at least 60 consecutive days. Each of these groups included 57 participants. The groups were similar in terms of sex, age, duration of illness, and types of atypical antipsychochotics prescribed. There were also no significant differences in overall psychiatric symptoms severity, “illness” severity, or psychological functioning.
Long-term benzodiazepine users had an average score of 16.02 on the Montreal Cognitive Assessment compared to 19.33 for non-users. This indicated significantly poorer cognitive performance for participants taking benzodiazepines for at least 60 consecutive days. This deficit was observed mostly in the visuospatial and executive cognitive domain, meaning benzodiazepine users had greater difficulty understanding visual/spatial information and planning, organizing, and solving problems. The authors note that there was no clear dose-response relationship between benzodiazepine use and cognitive impairment. This means similar cognitive impairment was observed in both low- and high-dose users.
This research had four main limitations. Cognitive ability was assessed at one point in time and the authors were not able to track cognitive changes before and after exposure to benzodiazepines. As a result, the findings cannot establish that benzodiazepine use caused poorer cognition, only that these factors are linked. Unmeasured factors may have affected the outcomes. All participants came from a single treatment facility in China, limiting generalizability. The relatively low number of participants further limits generalizability.
Benzodiazepines Use During Pregnancy Linked to Miscarriage
A new study published in BMC Pregnancy and Childbirth links benzodiazepine use during pregnancy to miscarriage. This research, led by Cyndie Picot from the Hospices Civils de Lyon in France, also presents some evidence that miscarriage risk increases with dose.
The aim of this study was to investigate the link between benzodiazepine use during pregnancy and miscarriage. The authors also wanted to examine whether miscarriage risk varied between individual benzodiazepines. They conducted a systematic review and meta-analysis of previous research around benzodiazepines and miscarriage. To be included in the current work, studies had to compare miscarriage risk between a group that used benzodiazepines during pregnancy and one that did not. The researchers used data from 10 studies.
Overall, benzodiazepine use during pregnancy was linked to a 68% increased risk of miscarriage. All seven individual benzodiazepines with at least 100 exposed pregnancies were linked to greater risk of miscarriage:
-
Clonazepam – 82% increased risk
-
Diazepam – 70%
-
Lorazepam – 48%
-
Alprazolam – 48%
-
Oxazepam – 48%
-
Bromazepam – 55%
-
Chlordiazepoxide – 42%
While the meta-analysis found increased risk for each of these individual benzodiazepines, the results did not reach statistical significance for bromazepam and chlordiazepoxide.
Three included studies examined the dose-response relationship between benzodiazepine use and miscarriage, with all three reporting that larger doses translated to higher risk. For example, one study reported an 84% increased risk of miscarriage for low dose clonazepam compared to a more than quadrupled risk (4.5 times higher odds) for high doses. The authors note that they were unable to combine results from the these three studies into a single does-response estimate due to significant differences in definitions and methods.
This research had four main limitations. Five of ten included studies had a high risk of bias due to the possibility of unmeasured factors affecting the findings. Included studies were mostly observational, meaning these results cannot definitively say that bezodiazepines caused miscarriage. Studies differed in their definition of miscarriage. Early miscarriages are likely under-reported, which could have affected the results.
Benzodiazepines Linked to Increased Risk of Falls and Fractures in Older Adults
A new literature review published in Medical Science reports that benzodiazepine use is linked to greater risk of falls and fractures in older patients. This review, led by Szymon Barczak from Mazovian Brodnowski Hospital in Poland, also reports that these risks may be greatest in the 24 hours after starting benzodiazepines.
The goal of this review was to examine the link between benzodiazepine use and fall/fracture risk in older adults. The authors reviewed 21 previously published studies on fall/fracture risk and benzodiazepine use in geriatric patients. Included studies had to be peer-reviewed and published in English.
Most included studies found a link between benzodiazepine use, falls, and fractures in older adults. Included studies found between a 40% to a more than tripled (3.128 times higher odds) increased risk of falls, and between 20% to nearly quadrupled (3.73) increased risk of fractures linked to benzodiazepine use. One study presented evidence that the risk of falls was nearly quadrupled (3.79 times higher odds) in the first 24 hours following the start of benzodiazepine use. Some included studies also reported that higher doses, longer duration of use, and polypharmacy were all linked to increased fall and fracture risk. The authors note that not all included studies linked benzodiazepine use to greater fall risk. Studies conducted with hospitalized participants were less likely to find a link between benzodiazepine use, falls, and fractures.
This study had 3 main limitations. Only English language studies were included, limiting generalizability. Included studies varied significantly in design, definitions of falls and fractures, and exposure periods. Most included studies were observational, meaning these results cannot say for certain that benzodiazepine use causes falls and fractures.
****
Barczak, S., Hertz, Z., & Barczak, J. (2026). The risk of falls and fractures in geriatric benzodiazepine users: A review. Medical Science, 30(167), 1–11. (Link)
Huang, Y.-S., Prawesti, G. N., Chiang, Y.-C., Chen, Y.-C., & Chen, H.-Y. (2026). Effects of benzodiazepines on heart failure outcomes and adverse events: A systematic review and meta-analysis. Cardiovascular Toxicology, 26(10). (Link)
Liu, C., Zhang, L., Pan, Q., Shi, Y., Yu, W., Li, G., & Li, H. (2026). Long-term benzodiazepine use is associated with poorer cognitive function in schizophrenia: Findings from the Salt-C Cohort. BMC Psychiatry, 26(1). (Link)
Picot, C., Piroux, Y., Pleau, J., Bérard, A., Cucherat, M., & Cottin, J. (2026). Risk of miscarriage after benzodiazepine use during pregnancy: Updated systematic review and meta-analysis. BMC Pregnancy and Childbirth, 26(1). (Link)
Yeesh. This all feels like it’s reaching to me.
But of course, I would think so because I’ve been using a small dose of a medium-length half-life benzo for 30 years, and I survived a 10-minute sudden cardiac arrest due to hemorrhagic shock (e.g., not due to benzos) and made a full recovery, which… is pretty damn rare.
I’m totally biased. I feel like benzos helped me stop drinking and manage a lot of symptoms from my physical health condition. There’s no way I can prove that’s true, and I try to be very skeptical of my instincts on this subject.
I’m also disgustingly fit for a guy pushing 70, and I work with college students at an elite school; one of the reasons I’ve stayed there is that I can keep up with the students. I did fall recently, but… that was early in the day, before I’d taken a benzo, and it’s probably the first time I’ve fallen in a quarter century. And it’s probably a bad idea for anyone to take any psychoactive drug of any kind during pregnancy– and many other kinds of drugs as well.
Are any of these studies controlling for comorbid alcohol or illicit drug use or other drugs, say, that might change the way benzos are metabolized?
I sometimes wonder about meta-analyses and what they can really tell us. I don’t think they are useless by any means… and it makes sense that you’d want to avoid prescribing these drugs to patients who already had heart problems, or who drank heavily, or who were on many other medications, or who were pregnant. Polypharmacy has bad outcomes generally, and it’s important to identify specifics drug combinations where it’s more dangerous.
What bothers me is the idea that we’re still prescribing SSRIs to patients for anxiety when they are both less effective and obviously more dangerous than benzos, at least for patients who do not drink or use other drugs that depress the CNS. As a psychotherapist, this is very frustrating, because my own experience is that virtually every patient in my caseload responds poorly to SSRIs, and even those who don’t seem to experience harms also seem very slow to realize gains.
I’ve also only had two or three patients who told me, “Yeah, SSRIs completely changed my life for the better.” And on the rare occasions they felt that way, I couldn’t find any evidence for it. I have no idea if you could say the same thing for benzos, because no one seems to prescribe them anymore– maybe they’re still being used in China.
What I’d like to see is a study that uses CURES and the EHR to follow non-drinking patients who take benzos for an extended period of time and compares them to patients who are prescribed SSRIs, and that includes both empirical data and patient reports. But even then, I kind of despair at finding meaningful answers. How do you compare the harms, say, from emotional blunting and anhedonia, which are enormous and much harder to measure empirically, with the harms of addiction, dependence, and over-sedation, which are easier to measure, but rarer, particularly now that patient compliance for Schedule II drugs can be monitored by a PDMP?
As is so often the case, I just can’t even imagine a study design that would be useful, which is incredibly frustrating.
Report comment
I agree. This is a lot of emotional overreaction by those who may have actually been harmed.
Report comment
I take Valium for sudden severe vertigo. I was diagnosed with Menieres Syndrome. The Valium stops it quickly. I only take 30 tablets every 6 months. Hardly an addiction. These scare tactics have to stop. People who actually need the medication are being denied due to fearful doctors. Please stop!
Report comment
Oh, yeah. That is a crazy low dose– wish I could get by with that!
I’m glad I’ve never felt I wanted to take more. Benzos are not that seductive to me, they just… put me to sleep eventually. Why would I want to take a second pill or a third?! And when would I take them, during the day? I have to work and drive, and I enjoy and am good at doing both those things, so that won’t be happening!
At times, my HMO has really put the screws to me trying to get me to go off them. But after the cardiac arrest and 5 months in the ICU, they seem to be more understanding. I have a very weird disease that causes my physical and emotional energy levels so change radically, even after exercise, meditation 2x or more per day every day without fail, etc.
When someone I know starts SSRIs, sometimes it’s just like losing a friend. Not always, of course, but it’s a complete personality change.
Yeah, can’t think of a great study design– or if I did, who would fund it!
Report comment
I’m thinking much of the benzo hyper addiction reaction has to do with where you are geographically. I’ve had a pain syndrome for 20 years, permanent and incurable. Rx’d ativan and Vicoden by Scripps Hospital physicians and i was taking very low doses of each. Moved north, detoxed myself off the Vicoden taking just tylenol and ativan, 4-6mg/day. Pain helped, able to exercise, sleep. OMG, no benzos! Completely clueless about benzos and knew nothing about how to lower or stop, not even sharing with me. I’ve been an RN for over 50 years, it takes 3 weeks of inpatient or very long outpatient detox to come off benzos…..and exactly why if it’s not being abused and helps whatever condition it’s prescribed for. Was inpt for 4 days with heart failure secondary to acute benzo withdrawal thanks to provider (quite frankly) stupidity. In ER 2 more times. Hello!! Just because I take a benzo responsibly doesn’t mean I’m an addict,,,,esp. when prescribed by excellent providers.
Report comment
Very interesting, Nettie! I am prescribed 1.5 mg. Lorazepam 1x per night, but take a bit less– usually 1.25, never more than 1.4 mg. Yes, I did get a digital scale to weigh the pill halves and quarters, that is how paranoid I am about them!
That timeline sounds about right. I am on less than half that, and I estimate it would take at least five or six weeks to come off them on a taper. An abrupt discontinuation would likely out me out of commission, or barely functional, for 60 to 90 days, just my back of the envelope.
15 months ago, I was in the ICU for a ruptured pancreatico-duoenal pseudo-aneurysm, as I noted above, my heart stopped for 10 minutes. I have no idea why CPR even worked, how I retained cognitive function. I was in the ICU for 5 weeks on all kinds of narcotics, of course — had an emergency ex-lap, it’s astounding I survived, all kinds of trouble with multiple organs, etc — but NO Ativan. After 10 to 12 days, I thought, “Dang, I’m off this stuff, and I’m doing fine even in this totally extreme medical environment. I’m calm, rational, totally on top of my own care despite being in moderate to severe pain. Maybe I’ll never need to go back on them. Cool!”
About day 16, the delusional episodes started– only when I was sleeping, I was alert and oriented x4 whenever I was awake, but they were getting much worse every day, and soon, I was only sleeping two or three hours a night. After a few days I realized, “Crap. I’m in withdrawal. Late to the party, but here it is.”
But then– convincing my treatment team to reinstate! Fortunately, being a health care provider and knowing the system, I knew the right plays, refused SSRIs (my poor reaction to them is well documented, my wife and colleagues visited and had my back) and got palliative care involved. Eventually, one of the residents finally said, “Hmm. Okay, I looked into it, that is remotely possible. The differential diagnosis is tachycardia upon awakening.”
And I was like, “Holy crap, I forgot all about that, but that rings a bell from clinical training. Let’s make sure to monitor that next time I ‘fall asleep’ (e.g., lapse into some delusional twilight state) and see what happens.’
Hours later I drifted off into another waking nightmare, woke up with my heart pounding.
They reinstated the Ativan the next day, albeit at a weird dose — .5 mgs 2x per day. (How does that restore a normal sleep cycle?!) Anyway, it was good enough. I got back up to 5 hours of sleep, recovered, and eventually got outta there.
But it was way, way too much work having to convince folks.
I am so glad you got through all that, Nettie, and yes– responsible providers! We have CURES and PDMPs now, let’s use them– and use them respectfully and appropriately.
Report comment