3 Reasons Why Medicines and Anxiety Disorders are a Tricky Mix

Anxiety disorders can be treated without the use of medications. For all of the anxiety disorders listed in DSM-IV, the psychiatrist’s guide for making diagnoses, expert consensus panels have placed cognitive-behavioral therapy as a first-line treatment, meaning that this can be the first (and only) treatment given for any anxiety disorder. What is it that makes anxiety disorders different than other disorders, such as schizophrenia and bipolar disorder, which are seen as requiring medication?


Anxiety disorders always involved learned behaviors. They always involve the amygdala, a part of the brain whose function is to detect threats. The amygdala is a necessary component of the experience of anxiety, which is the emotion experienced in response to some perceived threat. The association between the thing one fears and the fear itself is learned, and so can also be unlearned, even without the use of medicines.


Medications make the amygdala less hyperactive, so that it is less likely to make and use the “threat” label. In some cases, medications alone can rid them of the symptoms of their anxiety disorder.
The difficulty is that the effect of medicines only lasts while the person is taking them. Should one run out of the medicine, or have side effects that make them want to stop, or become pregnant, the anxiety will most likely return. Why is that?
The root of every anxiety disorder is an association with a given experience and the “threat” label of the amygdala. This association is strengthened every time the person avoids the threatening experience. Since the amygdala is continually modifying its labels based on what it observes: if a person avoids something, that confirms the label for the amygdala, and ensures that it will spark the anxiety response next time that thing is experienced; if the person approaches something, even in the face of the anxiety triggered by the amygdala, the amygdala modifies its threat label. With enough exposure to the feared object, the amygdala re-learns that the object is not a threat, and stops having that object trigger the anxiety reaction.


The tricky part comes when a person takes medicine for an anxiety disorder, has relief from anxiety when taking it, and then decides to stop. Three things will occur:

  1. Withdrawal.
  2. Loss of effect from the medicine.
  3. Momentary loss of “non-threat” labels.

Each of these three effects make life difficult for the person stopping the medicine.


Withdrawal from SSRIs (e.g. Zoloft, Paxil, Lexapro, etc.) very commonly has anxiety as its primary symptom. The only exception is Prozac, which has a very long half-life and so typically does not induce withdrawal symptoms (the drug leaves the system not in hours or days, but gradually over several weeks). So the withdrawal the brain experiences from stopping a drug can cause immediate anxiety, independent of other factors.
The second problem, loss of drug effect, only applies to those cases when the drug really was having a positive effect. If a person is stopping because the drug didn’t work for them, then this will not be a problem; but the drugs to help perhaps the majority of people with anxiety disorders that take them, and so, in most cases, the loss-of-effect will be noticed as an increase in anxiety.
The third problem is the most tricky of all, and can occur in anyone who took the medicine, regardless of whether it helped them or not. It is based on the fact that when the amygdala makes the assessment that something is safe, it does so with a number of provisos: it is safe in this place, under these conditions. Should those safety conditions change, the amygdala will stop generalizing the “safe” label, and the person will experience anxiety again. There is a clear example of this taken from animal research: if a mouse is kept in one cage, let’s say with a blue floor, and a strange object (e.g. a Coke can) is placed in the cage, the mouse will initially be afraid of the Coke can (mice have very active amygdalas, which work the same as in humans); but after some time, it will habituate to the Coke can, and begin to treat it like any other part of the cage. If you then take the mouse and put it in a red cage and let it get comfortable, and then put the old Coke can into that new cage, the mouse will be afraid of it again. Why?
The answer scientists give is that amygdalar learning is “milieu dependent” — that is, the amygdala learns that things are threats or non-threats in a way that depends on the whole environment in which the exposure took place. If the mouse’s amygdala learns that the Coke can is “safe” while in the blue cage, it will not necessarily generalize that learning to the new environment or milieu of the red cage.
Unfortunately, our brains consider medicines as a milieu; and so they also consider stopping a medicine as a change in milieu. This means that what was learned as “safe” while on a medicine may not necessarily be considered “safe” while off a medicine; the person will experience this as an inexplicable return of anxiety in the face of triggers that it learned, perhaps long ago (but while on the medicine) were safe.
The good news here is that the mouse in the new red cage will, when exposed again the once-familiar Coke can, re-learn that the can is safe, and will do so much faster than it took the first time. If the mouse were capable of reflection, when it saw the Coke can come to the new cage and noticed that its anxiety had come back, it may have thought, “Here I go again — back to square one!” But it would have been wrong: this was not exactly a relapse, it was simply a break in generalization. After the initial anxiety, the amygdala again learns that the can is safe, and the mouse’s anxiety would cease.


In the case of humans, when a person stops a medicine, especially if it is too abrupt, they may get withdrawal anxiety and conclude, “I guess that I need the medicine — as soon as I stop it, the anxiety returns.” This is an incorrect assessment: what they really needed was a more gradual taper.
The anxiety that comes from losing the good effect of the medicine is usually not immediate, and occurs over the course of a few weeks, not the first few days. Noticing the gradual return of the old feelings of anxiety, the person may conclude, “This proves that I need to be on a medicine.” This too would be mistaken: just because medicines help with anxiety does not mean that only medicines will help.
The anxiety that is felt when in the presence of prior triggers which one had learned were safe while on the medicine can be similarly misleading. An increase in anxiety in the face of given triggers does not necessarily mean that a relapse is occuring; it is best understood as one’s amygdala being extra cautious, ensuring that even in the new environment (i.e., not being on the medicine) the old trigger is still safe. To get past this hurdle, one must simply carry on with not avoiding triggers, just as one did while on the medicine. The anxiety will generally go away faster than before — and at the very least, one has the benefit of the knowledge that this trigger does not necessarily a true threat. As they prove this to their amygdalas, those making a try for behavioral interventions instead of pharmacological ones will then be able to experience the relief of getting better — without all the tricky provisions.


Click here to get help for anxiety. To learn more about Dr. Majeres, click here.