Q:  This is just a simple question. Can a person continue to build a tolerance for the medications perscribed for depresion and or bipolar disorder?

Dear L'S --

Unfortunately, this very important question turns out not to be simple. Think about it: how could you tell the difference between developing "tolerance" to a medication treatment for depression or bipolar disorder, versus the possibility that the underlying illness was worsening over time and  thus "escaping" control by that medication?

For some of our medications, "tolerance" is well known. The group of medications known as "benzodiazepines", which includes Valium and Klonopin (clonazapam) and lorazepam (Ativan) and Xanax (Alprazolam), all demonstrate a potential for the development of tolerance. If the medication is taken every day for 2-3 weeks, the body adapts. The medication is often no longer as effective. Then, to get the same initial benefit, the dose must be increased. Within a few weeks, that dose as well will be less effective. This process leads to what we call "dose escalation" which must be avoided, as ultimately, when one attempts to stop the medication, there is also (if tolerance has developed) a withdrawal phase in which agitation, anxiety, and insomnia are superimposed on whatever initial symptoms were present.

Withdrawal symptoms thus mark physiologic tolerance. Antidepressants also have a withdrawal syndrome, so in one respect one could say that the body develops a tolerance of sorts to this class of medications as well. But a loss of the initial improvement is not directly linked to this withdrawal phenomenon. That is, people whose benefits were sustained can still get the withdrawal symptoms (flu-like symptoms, electric shock-type sensations, agitation and irritability -- quite a wide variety of experiences). And vice versa: people who lost their initial benefit do not always have withdrawal symptoms when they stop the medication. Therefore, with antidepressants, "tolerance" is quite different than the tolerance we recognize for benzodiazepines. 

However, "loss of response" to an antidepressant-- a loss of the initial benefit -- is a well recognized phenomenon.  ndeed, loss of response to an antidepressant (so called "Prozac poop-out") has been shown to be a weak but statistically significant marker for bipolar disorder and people whose symptoms have not otherwise suggested a bipolar diagnosis. (This is one of many such "bipolar soft signs". See Ghaemi and colleagues, 2002; their article is described and then linked from my website). 

By contrast, the "mood stabilizers" prescribed for bipolar disorder have not been associated with a withdrawal pattern; nor a "loss of response", at least not a loss of response such as we see commonly with antidepressant medications (some of which is sure to be the loss of a placebo response, not a loss of true drug response). However, a return of symptoms even while someone continues to adhere completely to a previously effective combination of medication and non-medication treatments -- this is unfortunately extremely common. Indeed, it is almost more the rule than the exception in many research studies, where by the end of a long-term study (1-5 years), far more than half of the patients have experienced relapse.

Therefore, most unfortunately, one could say that an outcome which looks like "tolerance" (what was working before is no longer working, without any changes) is almost the norm in the treatment of severe mood conditions, at least bipolar disorder (on the whole I think this is probably less common in really solid good responses to antidepressant treatments for depression).

Finally, I think it is important to know that relapse is more common when symptoms are not completely resolved. In other words, a true "remission", such that no symptoms remain, has a better prognosis, a better chance of remaining stable, then a partial response to treatment in which some symptoms remain active. This is the reason why some of us push so hard to get all symptoms completely controlled: it improves the likelihood that the treatment will "stick". Of course, there is often a trade-off between trying to drive every last symptom into the ground, versus considerations of risk and side effects which one must potentially endure in order to get that kind of an outcome. For many patients, there is a relatively active juggling back-and-forth between trying to get complete symptom control and trying to drive side effects to zero. No wonder I find myself so busy.

Thank you for an important question.

Dr. Phelps