By: Timothy Hall, MD, PhD

Anxiety–whether social anxiety, generalized anxiety, or panic attacks–is a common problem in primary care. It also shows up months to years after the initial presentation when the patient needs to be referred for dependence or misuse of the benzodiazepines they were prescribed. As our understanding has advanced regarding the physiology of anxiety and learning, and also the long-term effects of benzos, psychiatrists and addiction specialists increasingly see benzos as appropriately being used mainly in limited circumstances.

First a bit about anxiety and learning: The classic panic attack manifests as an abnormal internal sensation (enteroception), often associated with something fairly benign, like mitral valve prolapse or an isolated PVC, or simple overreactivity of the autonomic nervous system. This leads to systemic release of epinephrine from the adrenal glands, causing tachypnea, tachycardia, sweating, trembling, Shallow, rapid breathing blows off CO2, causing lightheadedness, tingling in face and extremities, a (misleading) sensation of not being able to get enough air, and of tunnel vision and “things closing in”. This is very frightening for the patient and is quite uncomfortable even after education on its relatively benign nature (which is the first step in treating panic attacks). However, the adrenal surge of epinephrine will be tapped out after 10-15 minutes. This means that any orally delivered benzo will have its main effect after the peak panic is already resolving. Administering a benzo once the panic attack is already in progress merely reinforces the patient’s expectation that the panic is a) dangerous and b) requires a prn medication to make it go away.

Anxiety disorders in general are associated with avoidances–a pattern of avoiding potential triggers for anxiety and panic, such as crowds, enclosed spaces, heat, intense physical activity, or stressful images and thoughts. The avoidances are often progressive, with the patient avoiding more and more potential triggers that are ever less closely associated with the original trauma or stressor. The gold standard treatment for most anxiety disorders is cognitive-behavioral therapy with graduated, progressive exposure to triggers and stressful situations, often accompanied by skills training and repeated practice. (The “Treatments That Work®” series is a good reference, mostly based on work from Penn and UCLA.)

Benzodiazepines actually hinder progress in several ways. Patients tend to escalate doses over time, requiring higher or more frequent dosing to treat their anxiety. They also find themselves chasing the benzo withdrawal, which leaves them at a more anxious baseline than they would have had in the absence of the benzo. Unfortunately our brains are not well-equipped to make this association. We feel the immediate benefit of the benzo in reducing our anxiety–in fact, we often credit it for the benefit that is actually the natural resolution of the epinephrine surge of panic, or the placebo of being able to take a pill and regain the illusion of control. Only the most observant patients will see that they experience increased anxiety hours later. Benzos also interfere with the growth of new inhibitory synapses in the brain, which are needed to suppress the learned anxious associations. Thus benzos not only cause a higher-anxiety baseline, but also interfere with the natural recovery process from traumas and stressors.

First-line treatment for anxiety should generally be an SSRI or SNRI antidepressant–something serotonergic. I warn patients that this can increase anxiety for a few weeks before their anxiety should come down. Bridge prescriptions of antihistamines (diphenhydramine or hydroxyzine) or gabapentin, or possibly a long-acting benzo such as diazepam (with education on benzos) can all be used in the meantime for more anxious patients while waiting for the SSRI/SNRI to kick in. (Antihistamines and gabapentin are of course being used off-label here.) Standing doses are often preferable to prn dosing, in order to avoid the reinforcement of “anxiety –> needs a prn med”.

So when are benzos appropriate? The best use is in highly predictable, time-limited stressors: infrequent anxiety caused by public speaking, airplane travel, visits with family (that don’t involve alcohol). Some patients will carry their Xanax as a talisman; they rarely take it, but feel much better knowing they have the bottle. I am very happy to refill an expired, half-used bottle of a benzo that’s used in this fashion. The sad fact is that most anxiety disorders are more chronic, and require more chronic treatment in the form of SSRI/SNRI and/or CBT; prn treatment is rarely optimal except in the above-mentioned situations.

If a patient is already on a benzodiazepine, most other meds will seem second-best, because nothing else acts as quickly, and our brains are not built to detect the medium- and long-term problems as readily as we see the short-term relief. It’s therefore best not to start. Getting a patient off benzos once problems have developed takes time, effort, and persistence on the part of both patient and providers.

For more information, check out the Ashton Clinic’s website at


CBAM is a multidisciplinary center that seeks to advance the prevention and treatment of chronic illnesses, especially in communities with health disparities. As part of the UCLA Department of Family Medicine, CBAM works at the intersection of academia and community with a focus on treating addictions and preventing the spread of HIV.

© The UCLA Center for Behavioral & Addiction Medicine (CBAM). All Rights Reserved.