Edith is a 55 yo woman with hypertension and generalized anxiety disorder (GAD). She is establishing primary care with you because her PCP retired. She first received treatment for GAD 20 years ago. She took fluoxetine 40mg for a couple of months – it was helpful for anxiety but caused troublesome sexual side effects. She then tried paroxetine 20mg – it was similarly helpful for anxiety, but caused anorgasmia and constipation. She was also in psychotherapy for a year with a psychiatrist who started her on clonazepam 1mg BID. The clonazepam did not cause side effects and was very helpful for her anxiety. The intensity of her worries decreased, allowing her to focus better. She believes her career success over the last 20 years would not have happened without the medication. She also thinks she was a better mother because she could give her daughter more space during adolescence. Edith continues to take clonazepam 1mg BID every day and asks you to refill it. She has not had side effects from clonazepam, specifically no sedation, imbalance, or cognitive problems. She has never drank above healthy limits for alcohol and has never used any unprescribed drugs. She has always taken the clonazepam as prescribed.
Should you continue benzodiazepines for a stable patient with generalized anxiety disorder?
- Yes
- No
- It depends on the balance between the medication’s efficacy, side effects, and risks in a particular patient, and the benefits, side effects, and risks of alternative treatments, including no treatment.
This PsychSnap may be controversial. I know of large primary care practices and some psychiatrists who specialize in anxiety disorders who have made the decision to categorically not prescribe benzodiazepines. That approach is not my practice. I ask you to keep an open mind as you read this review on using benzodiazepines in anxiety disorders, and then email your reactions to psychsnaps@ucsf.edu.
The guidelines from many countries for the treatment of anxiety disorders recommend CBT and SSRIs/SNRIs as first line treatments for anxiety disorders. Those are also my first line treatments. Some guidelines state that BZDs are well-tolerated, safe, and effective for short-term use. Long-term use of BZDs for anxiety disorders is commonly cautioned against due to the risk of side effects, including the potential for dependence, sedation, and cognitive impairment.
New patients with anxiety disorders commonly present on chronic benzodiazepines. Do you continue the prescription? Do you recommend a taper? How strongly do you make that recommendation? Do you force a taper? Do BZDs even work in the long-term? Concretely, should you recommend that Edith taper the clonazepam and start an SSRI/SNRI or something else for GAD?
Are benzodiazepines as effective as antidepressants for the short-term treatment of anxiety disorders?
Yes. They act sooner and have at least equivalent efficacy to serotonergic antidepressants (SSRI/SNRIs) for patients with anxiety disorders at 4-8 weeks according to a 2013 meta-analysis (Offidani 2013). These data apply to GAD, panic disorder, and social anxiety disorder, not to OCD or PTSD.
Another systematic review and meta-analysis of 56 randomized placebo controlled trials (N=12,655) of BZDs or SSRIs/SNRIs for GAD showed a larger effect size for BZDs (Hedges’ g= 0.50, medium effect) than for SSRIs/SNRIs (Hedges’ g=0.34, small to medium effect) (Gomez 2018). An open-label 8 week randomized trial compared clonazepam 2mg and paroxetine 40mg daily in patients with panic disorder. Both treatments were equally helpful for panic disorder (Nardi 2012).
Are benzodiazepines as effective as antidepressants for the long-term treatment of anxiety disorders? Yes, but the data is limited.
Don’t people develop tolerance to benzodiazepines? People rarely develop tolerance to the anxiolytic effects of BZDs.
Patients with panic disorder who responded to their initial treatment (paroxetine 40mg or clonazepam 2mg) in the Nardi trial (n=84) continued their treatment for a total of 3 years. Both paroxetine and clonazepam remained effective in treating panic disorder over 3 years without dose escalation or loss of efficacy in either group. People did not need a higher dose of the BZD to achieve the same anxiolytic effect (Nardi 2012).
A statewide study of patients with Medicaid looked at the frequency of BZD dose escalation over 2 years among patients who were initially prescribed appropriate dosages of BZDs (defined as <20 diazepam milligram equivalents (DMEs) for younger patients and less than <10 DMEs for older patients). The aim was to measure the probability of developing tolerance or a substance use disorder in patients using long-term BZDs, defined as an escalation to high dosages over 2 years. 2,440 patients were taking benzodiazepines for any diagnosis. Of those on chronic BZDs, the median dose remained constant at 10 DMEs (clonazepam 1mg, lorazepam 2mg). The incidence of escalation to high dosages of BZDs was 1.6% (40/2,440 patients). High dosages of BZDs were defined as above 40 DMEs in adults or 20 DMEs in older adults (Soumerai 2003, dose conversions from Borrelli 2022).
Aren’t benzodiazepines harmful?
There are population based risks and individual harms from BZDs, as with all psychotropic medications.
Benzodiazepines are associated with sedation, reduced cognitive functioning (while someone is taking them) and impaired driving ability. In the elderly, there is an increased risk of falls, fractures and delirium. BZDs also carry a risk of physiologic dependence and withdrawal when used chronically, and a low risk of developing a BZD use disorder in people without substance use disorders. When BZDs are combined with other CNS depressants like opioids and alcohol, they increase the risk of fatal overdose. Recent concerns about BZDs causing dementia have not been convincing, but this question is an area of active research.
SSRIs and SNRIs are associated with sexual dysfunction, weight gain, disturbed sleep, and hyponatremia, among other side effects. When they are stopped, many people experience a significant withdrawal syndrome. Rarely, SSRIs/SNRIs precipitate mania.
Atypical antipsychotics like quetiapine, olanzapine, and risperidone are sometimes used as augmentation agents in anxiety disorders. They are associated with weight gain, metabolic disorder, extrapyramidal side effects including tardive dyskinesia, and orthostasis, among others. They also commonly cause withdrawal symptoms when stopped.
Psychotherapy for anxiety disorders takes time and maybe money. It is unfortunately harder to access than prescription medications, but therapy does not cause cognitive impairment, sexual dysfunction or metabolic disorder.
Let me reframe the question. For whom are benzodiazepines more likely to be harmful?
People with a current or past substance use disorder
People with a substance use disorder (SUD) who are prescribed BZDs are more likely to develop a BZD use disorder than people with a SUD who are not prescribed BZDs (15% v. 6%) (Ghaemi, 2018). In a survey of the general US population (2015-16), only 1.5% (95% CI, 1.26-1.72%) of people who used BZDs had a BZD use disorder (Blanco 2018).
People at higher risk of negative side effects
Children and adolescents, people with cognitive impairments, and older adults (65 years or older) are at higher risk of adverse effects from BZDs. BZDs are also more risky in people who take opioids, drink alcohol heavily, have COPD, or are at a high risk for falls. There is some limited data to suggest that people with impulse control problems associated with borderline personality disorder, ADHD, or traumatic brain injury are at an elevated risk of BZD misuse.
But this goes against everything I’m being taught. Aren’t antidepressants better than BZDs?
In the 1960s, benzodiazepines became the treatment of choice for anxiety disorders and anything else that Mother’s Little Helper could help. They were likely overprescribed. Fluoxetine was released in 1988 and was marketed as a safer alternative to BZDs for treating anxiety disorders. Eli Lilly promoted the story that BZDs were addictive with a serious withdrawal syndrome, while fluoxetine was not addictive and did not have a withdrawal syndrome.* For the last 35+ years, Pharma has aggressively marketed new antidepressants and the second generation antipsychotic drugs. All BZDs are generic, leaving no financial incentive to study or market them and a limited research base to support their use.
Yes, BZDs have real risks associated with them, particularly among older adults and adolescents. Yes, BZDs are associated with a small risk of developing a benzodiazepine use disorder, though in someone without another substance use disorder, that risk is small. These risks must be weighed against the benefits and risks of other available treatments or of untreated illness.
Putting it all together
If you are thinking about prescribing a benzodiazepine:
- Clarify the diagnosis. Are BZDs effective for this diagnosis? Bereavement, recent trauma or PTSD, and OCD are not treated with BZDs.
- Screen for risk factors and pay attention to escalating or high doses of BZDs. Consider urine screening before writing the first prescription.
- Regularly re-evaluate the benefits and risks of ongoing benzodiazepine use. Aim for the lowest effective dose.
- If the risks or side effects outweigh the benefits, collaborate with the patient around a personalized plan for tapering and an alternative plan for treating the anxiety disorder (Hirschtritt, 2021).
Back to Edith: Should you continue the clonazepam 1mg BID for Edith who is 55 years old and whose symptoms of generalized anxiety disorder have been fully treated on benzodiazepine monotherapy for 20 years?
Maybe. Talk with her.
Edith has not developed tolerance to the anxiolytic effect of the clonazepam over 20 years. You ask about potential side effects of BZDs, and she does not have them. You discuss how as people get older, they may start to experience new side effects.
Edith has never tried dose reduction before, and you ask if she might be interested in finding the minimum effective dose at this time. She says not now, but is interested in the future – maybe during the summer when work is less hectic.
You check the state database for controlled substances. You order a urine test that is gas chromatography/mass spectrometry so that it will detect entities like benzodiazepines accurately (some BZDs don’t show up on screening/immunoassay tests), confirming with Edith that she last took clonazepam this morning. You tell Edith that you’ll send in the clonazepam prescription as soon as you get the results from the urine test. You confirm that she has enough pills for the next couple of weeks.
Final Note: I am not suggesting the BZDs should be first or second line treatment for chronic anxiety disorders. I generally recommend therapy and/oruse antidepressants as first line treatment. I generally use BZDs at low-doses for short-term treatment. If patients have been stable on chronic BZDs without side effects, however, there is generally not an urgency to taper the BZD. Attempts to switch from a BZD to an antidepressant in stable patients may do more harm than good.
*The withdrawal syndrome of fluoxetine (and the other SSRIs that followed) was first ignored and then later called a “discontinuation syndrome” by Eli Lilly.
Look out for a future PsychSnap on increasing the safety of prescribed benzodiazepines.
Key Points
- Benzodiazepines are effective treatments for anxiety disorders in the short and long term. When treating anxiety disorders in low risk populations, the risk of dose escalation or tolerance to the anxiolytic effects of BZDs is very low.
- BZDs have risks associated with them, particularly among older adults, adolescents, and people with substance use disorders. Those risks must be weighed against the benefits and risks of other available treatments or of untreated illness.
Related PsychSnaps:
“How do you counsel an older patient about the risks of benzodiazepines?” Zoë Kopp, October 2023.
“How do you taper long-term benzodiazepines?” Emma Samelson-Jones, Nov 2023.
References:
Blanco, Carlos, et al. “Prevalence and correlates of benzodiazepine use, misuse, and use disorders among adults in the United States.” The Journal of clinical psychiatry 79.6 (2018): 1865.
Borrelli, Eric P., et al. “Application of a diazepam milligram equivalency algorithm to assess benzodiazepine dose intensity in Rhode Island in 2018.” Journal of Managed Care & Specialty Pharmacy 28.1 (2022): 58-68.
Chang, Yiheng, et al. “Exploring clinical applications and long-term effectiveness of benzodiazepines: An integrated perspective on mechanisms, imaging, and personalized medicine.” Biomedicine & Pharmacotherapy 173 (2024): 116329.
Dubovsky, Steven L., and Dori Marshall. “Benzodiazepines remain important therapeutic options in psychiatric practice.” Psychotherapy and Psychosomatics 91.5 (2022): 307-334.
Ghaemi, S. Nassir. Clinical psychopharmacology: principles and practice. Oxford University Press, USA, 2018.
Gomez, Angelina F., Abigail L. Barthel, and Stefan G. Hofmann. “Comparing the efficacy of benzodiazepines and serotonergic anti-depressants for adults with generalized anxiety disorder: a meta-analytic review.” Expert opinion on pharmacotherapy 19.8 (2018): 883-894.
Hirschtritt, Matthew E., Mark Olfson, and Kurt Kroenke. “Balancing the risks and benefits of benzodiazepines.” Jama 325.4 (2021): 347-348.
Nardi, Antonio E., et al. “A randomized, naturalistic, parallel-group study for the long-term treatment of panic disorder with clonazepam or paroxetine.” Journal of Clinical Psychopharmacology 32.1 (2012): 120-126.
Offidani, Emanuela, et al. “Efficacy and tolerability of benzodiazepines versus antidepressants in anxiety disorders: a systematic review and meta-analysis.” Psychotherapy and psychosomatics 82.6 (2013): 355-362.
Shinfuku, Masaki, et al. “Effectiveness and safety of long-term benzodiazepine use in anxiety disorders: a systematic review and meta-analysis.” International clinical psychopharmacology 34.5 (2019): 211-221.
Soumerai, Stephen B., et al. “Lack of relationship between long-term use of benzodiazepines and escalation to high dosages.” Psychiatric Services 54.7 (2003): 1006-1011.
Send a Comment