Antidepressants and Bipolar Depression

The role of antidepressants in treating bipolar depression has been hotly debated in psychiatry for over twenty years. During that time many studies involving bipolar patients have been conducted, thousands of articles written, and hundreds if not thousands of experts in bipolar disorder have expressed their opinion on the matter. The most common view is that antidepressants should be used as little as possible in patients with bipolar disorder, and in those not known to have it but who might have it (based on family history, personal history, response to medications in the past).

The concerns go beyond the possibility of causing a switch from depression into mania. It is commonly stated that antidepressants can cause mixed episodes (a combination of manic and depressive symptoms), cycle acceleration (more episodes in a given period of time than would have occurred without an antidepressant being introduced), treatment resistant depression (a prolonged episode of depression unresponsive to the usual treatment measures), increased risk of suicide, and rapid cycling (four or more distinct episodes of depression or mania in one year). Clinicians have been warned not to use antidepressants, and consequently almost all of us feel bad/guilty to one degree or another because we still use them; they remain the most commonly used treatment in bipolar depression. Their continuing popularity is likely multifactorial. I think part of it is the lack of evidence based alternatives that make intuitive sense. Up until recently the only FDA-approved medications were Symbyax (a combination of olanzapine and fluoxetine that has never caught on in spite of considerable effectiveness and ease of use) and Seroquel (why use a sedating antipsychotic to treat depression?).

Where are we now with over twenty years of study and debate? Which of the many questions about the use of antidepressants in patients with manic-depressive illness can we answer and which guidelines can we confidently follow?

To address this important issue, The International Society for Bipolar Disorders ISBD) Task Force on Antidepressant Use in Bipolar Disorders was convened. These global experts were chosen based upon a search of citations on the specific topic of antidepressant use in bipolar disorder. They represent the big names in bipolar disorder from around the world.

Ten topic areas were researched and the quality of the evidence rated from A (excellent ) to D (poor). Topics covered included the efficacy of antidepressants as monotherapy for bipolar depression, antidepressants used as adjuncts to mood stabilizers such as Lithium and Valproate in bipolar depression, antidepressants  used as part of maintenance treatment and risks of antidepressant use including switches into mania or mixed episodes, cycle acceleration and precipitation of suicidal thoughts and behavior.

The interested reader can study the entire article. It is easy to summarize the findings: We know almost nothing about the use of antidepressants in bipolar disorder in general, and in bipolar depression in particular. Do antidepressants work in bipolar depression? We don’t know. Do antidepressants hurt or cause harm in bipolar disorder? This is unknown. Of the ten questions researched, the data base was rated as poor for seven of them; this is the lowest rating that could be given. Only one out of the ten questions asked had a good data base available for answering the question; none had an excellent data base. The only question that could be answered with some confidence is that Symbyax (the combination of the antidepressant fluoxetine and the second generation antipsychotic olanzapine) works for bipolar depression, and that paroxetine and bupropion in combination with a mood stabilizer probably do not work. It is ironic that experts have for many years stated that bupropion is likely the best antidepressant to use in bipolar depression, and now one of the few things we can say with confidence is that bupropion does not help.

What is the take away message? If we are going to use antidepressants in bipolar disorder, and most of us are (the data is very clear about that), we need to be frank with our patients that we don’t know if they are helpful, or harmful. And that there are alternative treatments, such as quetiapine (Seroquel), fluoxetine/olanzapine (Symbyax), and lurasidone (Latuda) that do have evidence to back up their effectiveness in decreasing symptoms of depression in patients with bipolar disorder. These other medications have more potential side effects, some of which are quite serious, but we know they can work.

The International Society for Bipolar Disorders (ISBD) Task Force Report on Antidepressant Use in Bipolar Disorders

Isabella Pacchiarotti, M.D., Ph.D.; David J. Bond, M.D., Ph.D.; Ross J. Baldessarini, M.D.; Willem A. Nolen, M.D., Ph.D.; Heinz Grunze, M.D.; Rasmus W. Licht, M.D., Ph.D.; Robert M. Post, M.D.; Michael Berk, M.D., Ph.D.; Guy M. Goodwin, F.Med.Sci.; Gary S. Sachs, M.D.; Leonardo Tondo, M.D.; Robert L. Findling, M.D.; Eric A. Youngstrom, Psy.D., Ph.D.; Mauricio Tohen, M.D., Dr.P.H.; Juan Undurraga, M.D.; Ana González-Pinto, M.D., Ph.D.; Joseph F. Goldberg, M.D.; Ayşegül Yildiz, M.D.; Lori L. Altshuler, M.D.; Joseph R. Calabrese, M.D.; Philip B. Mitchell, M.B.B.S., M.D.; Michael E. Thase, M.D.; Athanasios Koukopoulos, M.D.; Francesc Colom, Psy.D., Ph.D.; Mark A. Frye, M.D.; Gin S. Malhi, M.D.; Konstantinos N. Fountoulakis, M.D., Ph.D.; Gustavo Vázquez, M.D., Ph.D.; Roy H. Perlis, M.D.; Terence A. Ketter, M.D.; Frederick Cassidy, M.D.; Hagop Akiskal, M.D.; Jean-Michel Azorin, M.D.; Marc Valentí, M.D., Ph.D.; Diego Hidalgo Mazzei, M.D.; Beny Lafer, M.D.; Tadafumi Kato, M.D., Ph.D.; Lorenzo Mazzarini, M.D.; Anabel Martínez-Aran, Psy.D., Ph.D.; Gordon Parker, M.D., Ph.D.; Daniel Souery, M.D., Ph.D.; Ayşegül Özerdem, M.D., Ph.D.; Susan L. McElroy, M.D.; Paolo Girardi, M.D.; Michael Bauer, M.D., Ph.D.; Lakshmi N. Yatham, M.D.; Carlos A. Zarate, M.D.; Andrew A. Nierenberg, M.D.; Boris Birmaher, M.D.; Shigenobu Kanba, M.D., Ph.D.; Rif S. El-Mallakh, M.D.; Alessandro Serretti, M.D., Ph.D.; Zoltan Rihmer, M.D., Ph.D.; Allan H. Young, M.D., Ph.D.; Georgios D. Kotzalidis, M.D.; Glenda M. MacQueen, M.D., Ph.D.N.; Charles L. Bowden, M.D.; S. Nassir Ghaemi, M.D., M.P.H.; Carlos Lopez-Jaramillo, M.D., Ph.D.; Janusz Rybakowski, M.D., Ph.D.; Kyooseob Ha, M.D.; Giulio Perugi, M.D.; Siegfried Kasper, M.D.; Jay D. Amsterdam, M.D.; Robert M. Hirschfeld, M.D.; Flávio Kapczinski, M.D., Ph.D.; Eduard Vieta, M.D., Ph.D.

Am J Psychiatry 2013;170:1249-1262. doi:10.1176/appi.ajp.2013.13020185


Stop Antidepressants Slowly

People commonly decide to stop taking an antidepressant. Does it matter how this is done? If a patient is pregnant, isn’t it best to stop the medication immediately? If you or your patient is taking one of the antidepressants that is rarely associated with the SSRI withdrawal syndrome, it doesn’t matter whether the medication is stopped quickly or slowly, does it? What about the patient who has been stable for a year, or two years; if they are asymptomatic, happy, and at a good place in their life, surely it doesn’t matter if they stop citalopram 20mg without tapering.

An article published in the American Journal of Psychiatry in 2010 indicates it does matter. Though this article is three years old, its significance leads me to write about it now. It is my impression that many clinicians are unaware of the risk of not tapering antidepressants, or neglect to educate their patients about the importance of going off medication gradually in almost all situations.

It is well known in psychiatry that rapid cessation of mood stabilizers increases the risk of a mood episode. In fact, previous studies of maintenance treatment have been criticized for including a treatment arm randomizing patients to abruptly stopping their medication. This is because it is widely understood that by stopping the drug suddenly, the group randomized to placebo is going to have an increased rate of relapse. This unnecessarily puts patients at higher risk for an episode of their illness and makes the drug being studied look more effective at preventing relapse than it really is. It is no longer considered ethical or valid to stop medication suddenly in trials involving bipolar disorder.

Ross Baldessarini and his colleagues have been publishing studies for many years involving the inhabitants of the island of Sardinia. They have previously documented the risks of rapidly stopping mood stabilizers and antipsychotics. The Sardinian studies have focused on mood, anxiety and psychotic disorders, particularly the longitudinal course of these illnesses, using a naturalistic design.  In the current study they looked at the consequences of the abrupt withdrawal of antidepressants in patients with major depression, bipolar 1 and bipolar 2 disorder, and panic disorder.

One of the strengths of this study, and of the other studies from this group, is the comprehensive nature and duration of follow up. Meticulous records have been kept for over forty years. There may not be a comparable database of such a large number of individuals, followed so closely and for so long, anywhere else in the world. The studies published have been of great interest and usefulness to psychiatry.

The principal finding of this study was that “the abrupt or rapid discontinuation of clinically effective antidepressant treatment was associated with a significantly shorter time to a first new episode of major depression or panic.” The patients were clinically well at the time the medication was stopped, without even mild degrees of depression or panic. In other words, there was no evidence that these patients were experiencing a relapse or recurrence prior to going off medication. The decision to stop medication rapidly was made by the patient in a large majority of cases. The increased risk of recurrence was found with both newer and older antidepressants, though the benefits of a slow taper of the medication was more pronounced with older medications.

Based on the findings, the authors made the following recommendations:

1) Gradual dose tapering should be done in all situations unless a patient is experiencing a severe adverse medical effect or a severe manic reaction.

2) Patients should be warned that abrupt discontinuation of an antidepressant can lead to early withdrawal reactions and, over a period of several months, a return of the illness being treated.

3) Primary care clinicians, who prescribe most of the antidepressants in the US, should be aware that it is the rate of dose tapering and discontinuation that plays an important role in determining the risk of stopping medication.

Though they do not say this unequivocally, the authors imply that even in the case of pregnancy the risks of stopping antidepressant medication outweigh any potential benefits. Based on the current data available to clinicians, I agree. There is a very high risk of recurrence of illness if a woman stops medication for schizophrenia, bipolar disorder or depression while pregnant. The evidence does not indicate any benefit from suddenly or abruptly stopping medication during pregnancy. The fetus has already been exposed to the medication, and rapidly discontinuing it puts the woman and her baby at the added risk of an episode of the illness she suffers from. The more appropriate course of action is to get an expert in mental illness and an obstetrician involved, so that the risks and benefits of continuing or stopping pharmacological treatment can be examined and discussed with the patient. The knee jerk reaction of stopping the antidepressant right away can create a much more severe situation and cause more harm.

Summing up: Stopping antidepressants suddenly increases the risk that symptoms will return. This applies to depression (both in unipolar and bipolar disorders) and panic disorder. Slowly tapering medication rather than stopping it rapidly is almost always the right thing to do. If treating patients with antidepressants, educate them about this. Many of them will choose to stop medication on their own in the future, and it will decrease the risk of recurrence if they do so slowly.

Illness Risk Following Rapid Versus Gradual Discontinuation of Antidepressants                                                                 Ross J. Baldessarini, M.D.; Leonardo Tondo, M.D., M.S.; Carmen Ghiani, M.D.; Beatrice Lepri, Psy.D.                           Am J Psychiatry 2010;167:934-941. 

Do antidepressants have any role in treating mild depression?

In recent years there has been a lot of attention given to the question: “Do antidepressants really work”? There is general agreement that in cases of severe depression, the evidence supports their efficacy. However, there is ongoing debate about the usefulness and role of pharmacotherapy for milder depressions.

There have been at least two meta-analyses of antidepressant trials that found no benefit from taking an antidepressant when an individual has mild to moderate depression. I am not an expert on statistical methods, and am not qualified to judge the merits of these meta-analyses (in fact, about all I can say about a meta-analysis is that it involves lumping the results of a number of studies together, giving some more weight than others depending on the size and quality of the study, and then running statistical tests on the combined results). There are many critiques of these two studies by other experts in psychiatry and statistics, pointing out the various ways in which the studies chosen to be included in the meta-analyses, and the statistical methods used to analyze them, lead to erroneous conclusions. Also, there are other analyses, using a larger database, that find that antidepressants are effective in treating mild depression.

There are points on which I am in complete agreement with the medication skeptics. Psychotherapy is underused, is often undervalued, and sometimes has more lasting benefits. Insurance coverage for psychotherapy should be equal to that of coverage for treatment with an antidepressant. Medication has been over-utilized and relied on to do things which it is unable to accomplish. Pharmaceutical companies have had much too great an influence on training and education of physicians, and have proven to be unethical on so many occasions that almost anything that comes from them is suspect. Even diagnostic categories have been influenced by them. Their impact has been so great that I wonder if much of the research done over the past couple of decades is reliable, even some not directly supported by drug company funding.

As an example, during the late 1990’s and into the 2000’s, there was an enormous push to get clinicians to suspect bipolar disorder under every rock or bush on the clinical landscape, with warnings of the dire consequences that comes from using an antidepressant rather than a mood stabilizer, preferably a second-generation antipsychotic. Much, probably most, of the research done, studies published, conferences held and speakers paid to talk about this were funded by pharmaceutical companies with a financial incentive for clinicians to prescribe a newer antipsychotic (such as Zyprexa, Seroquel, Risperdal, Abilify, Geodon). The list of soft signs of bipolar disorder, and clues from the history, that supposedly supported a diagnosis of bipolar disorder became quite extensive. One of the consequences of Big Pharma money was an explosion in the rates of diagnosis of bipolar disorder, which continues today.There are good reasons to screen for bipolar disorder when a patient presents with depression, in particular the facts that antidepressants don’t work very well for bipolar depression and that some patients can be harmed by them. The exact role of antidepressants in bipolar depression continues to be debated, in part because we still don’t have enough information from good studies.

Getting back to the main issue I wanted to address: are antidepressants ever helpful for chronic mild depression, or recurring forms of mild-moderate depression? Putting aside the research, which can be interpreted either way depending on what you want to believe, my experience leads me to say yes, most definitely. But not always, and not for everyone. Both professionally and in my private life, I have observed them help some individuals, and do nothing for others. Most convincing to me is the antidepressant effect they have produced, of many years duration, for people I know well. Without an antidepressant, they would have been forced to endure years of depression. And psychotherapy had been engaged in, sometimes for long periods of time. If clinicians were to follow the current trend in some circles and not prescribe for mild or moderate depression, it would consign some people to experience unnecessary suffering, impaired functioning and a lower quality of life. And of course, the impact on family can be severe.

I am not arguing that every depression should be treated with medication. I believe that medication should be one of a number of treatment options, and that clinicians should not be the ones to decide what is the right treatment. Rather, they should educate their patients and be guided by their patients’ preferences. Psychotherapy can be effective for depression, but it does not work for everyone, and some people don’t want psychotherapy. Also, most psychotherapy studies used well-trained therapists following a manualized treatment approach, such as Interpersonal Psychotherapy or Cognitive Behavior Therapy. In the real world, therapists tend not to follow evidence-based therapy guidelines, opting for an “eclectic” approach, which has little or no evidence to support its effectiveness. Having personal experience with both traditional psycho-analysis and a humanistic/eclectic approach, I know that psychotherapy can work and be very helpful. However, I also know from professional and personal experience that psychotherapy can be a waste of time and money; it can turn people off to getting help of any kind if they have the misfortune of seeing an incompetent therapist, and it occasionally makes some people more ill or worse off than before they engaged in therapy. Harm from psychotherapy may not be common, but it is incorrect to assume that only medication has the potential to have adverse effects.

Antidepressants are one of a number of treatment options that should be offered to someone with depression, even if it is mild. Exercise, behavioral activation, cognitive therapy, dynamic therapies and medication are some of the evidence-based treatments to be discussed with patients, so that they can make an informed decision.

Folic Acid, Vitamin B12 and Negative Symptoms of Schizophrenia

[In Brief: Folic acid and vitamin B12 help a subset of individuals with schizophrenia by decreasing negative symptoms. Side effects are no different than with placebo. Adding folate and vitamin B12 is a low risk strategy for symptoms that are very difficult to treat.]


Negative symptoms play an important role in the disability caused by schizophrenia. Symptoms in schizophrenia are divided into 4 main types: positive, negative, cognitive (impaired concentration, poor planning and problem solving, etc.) and mood. Use of the terms positive and negative has nothing to do with whether the symptoms are good or bad. Positive symptoms refers to hallucinations, delusions and disorganization. Negative symptoms are some of the most disabling, and play a larger role in impaired functioning than positive symptoms. Negative symptoms include loss of interest (apathy), loss of pleasure (anhedonia), decreased motivation (avolition), decreased emotional expression (affective flattening) and decreased speech (alogia). These are sometimes referred to as the 5 A’s.

Negative symptoms are hard to treat. When the newer antipsychotics (Risperdal, Zyprexa, Seroquel, Abilify, and others) became available in the mid-90’s and later, they were promoted as superior to older antipsychotics, in part because of a purported greater ability to decrease negative symptoms. It took quite awhile, but we now know that these second-generation antipsychotics are no better at decreasing negative symptoms than first-generation antipsychotics. It is possible that the older antipsychotics are more likely to increase negative symptoms, particularly when used in high doses.

An article published in the May, 2013 issue of JAMA Psychiatry studied the effect of adding folate and vitamin B12 on negative symptoms, taking into account the role of genetic variation in several genes involved in folate absorption and metabolism. The study was double-blind, placebo controlled and enrolled 140 patients with chronic schizophrenia. Participants were stable but still symptomatic. Doses were 2mg/day of folate and 400micrograms/day of vitamin B12.

The authors were particularly interested in studying folate because of the epidemiological, biochemical and genetic association studies that have identified folate deficiency as a risk factor for schizophrenia. Findings of prior investigations include the following:

* After famines in China and The Netherlands, rates of schizophrenia doubled, presumably as a result of dietary deficiencies.

* Low folate levels in the 3rd trimester of pregnancy are associated with more than a 2 fold increased risk in the development of schizophrenia in offspring.

* Low folate levels have been found in patients with schizophrenia.

* In patients with schizophrenia, low folate levels are associated with more severe negative symptoms, but do not correlate with positive symptoms.

* 2 small studies found that adding a folate supplement benefited patients with schizophrenia, if they had low folate levels.

* In patients with prominent negative symptoms, folate supplements were helpful in the subgroup with a specific genotype.

The author’s hypothesis was that folate and vitamin B12 supplementation would decrease negative symptoms, but only in those patients with hypofunctioning (low functioning) genes involved in the absorption and metabolism of folate. In other words, only some people would have more emotional expression, talk more, and be more motivated, and it would be those with genes that were less effective at absorbing and metabolizing folate.

I should have mentioned earlier that this is a difficult paper. I do not recommend reading the whole thing, unless you like struggling with genotypes, alleles and vague statements, while in search of limited conclusions, of uncertain value.

That being said, the study found (as predicted) that folate and vitamin B12 supplementation did help negative symptoms, but only “modestly”, and only in a subset of patients. Other symptoms of schizophrenia did not benefit. There was no difference in side effects between those taking the vitamins and those taking placebo. When genotype was taken into account, there was a 27% difference in negative symptoms between those receiving folate and B12 and those receiving placebo. Is there any practical significance of this finding?                                                                                                                                                                                                                       

In the authors words:  “Even small effects of folate and vitamin B12 supplementation could be clinically meaningful, though, given the disability associated with negative symptoms, the lack of available treatments,35 and the minimal apparent adverse effects of vitamin supplementation.”

In my slightly optimistic interpretation: negative symptoms interfere with the ability to function, they are hard to treat, these doses of vitamins are safe, and even a small improvement in negative symptoms might mean a lot to the individual and their family. There is little to lose and possibly something small but important to gain.                                                                                                                        I could not find anything in the article that would allow one to estimate the proportion of patients likely to benefit from supplementation.


For those looking for a punishing read, the full article can be found in the May, 2013 issue of JAMA Psychiatry.





Randomized Multicenter Investigation of Folate Plus Vitamin B12 Supplementation in Schizophrenia


Joshua L. Roffman, MD, MMSc; J. Steven Lamberti, MD; Eric Achtyes, MD, MS; Eric A. Macklin, PhD; Gail C. Galendez, BS; Lisa H. Raeke, MA; Noah J. Silverstein, BA; Jordan W. Smoller, MD, ScD; Michele Hill, MD; Donald C. Goff, MD


JAMA Psychiatry. 2013;70(5):481-489. doi:10.1001/jamapsychiatry.2013.9

Prazosin for Nightmares/PTSD

[ In Brief: Prazosin decreases nightmares, improves quality of sleep and decreases overall PTSD symptoms in active-duty soldiers. It is well tolerated.]

Nightmares related to the trauma are experienced by most people with PTSD. There are no medications approved for treating nightmares, though trazodone, prazosin, clonidine and sedating antipsychotics are frequently prescribed. Of these, prazosin is the best studied and has the most support.

In the September, 2013 issue of The American Journal of Psychiatry, prazosin is compared to placebo in the treatment of nightmares and PTSD in active-duty combat veterans. The trial was double-blind and randomized. Primary outcome measures were severity of nightmares, quality of sleep and overall PTSD symptom levels. Prazosin was superior to placebo on all three measures, to a significant extent. Nightmare severity decreased by about 50% in those taking prazosin, and it was associated with marked or moderate improvement in global illness severity in 64% (versus 27% of those on placebo) an NNT of 3 (Number Needed to Treat). This mean only 3 people needed to be treated for 1 person to have a moderate or marked response, which is pretty good for psychiatric medications (and pharmacotherapy in the rest of medicine). However, only 3 patients on prazosin went into remission, versus 0 on placebo; so most patients were still symptomatic. Prazosin was not an effective treatment for most people in and of itself.

There were no differences overall in the number of patients experiencing side effects, which were generally mild and transient. There was 1 episode of brief syncope (passing out due to insufficient blood supply to the brain). This is a well known side effect of prazosin; if it occurs, it is usually with the first dose or very early in treatment. Other side effects included (with the percentage of those on prazosin listed first): lightheadedness (25% vs 20%), nasal congestion (22% vs 11%), lack of energy (0% vs 3%), drowsiness (3% vs 9%), depression (0% vs 6%). There were no changes in blood pressure. Headaches were significantly less common with prazosin.


I have prescribed prazosin a few times, without much success. After reading this article, I think the most likely reason for that is the much lower dose at which I was prescribing it. In this study, prazosin was given twice a day; For men, final doses were 4 mg mid morning and 15.6 mg at night; for women, final doses were 1.7 mg mid morning and 7 mg at night. The authors point out that women should be treated with lower doses due to greater sensitivity to the medication, and that the drug’s short half-life indicates it is likely to be more effective if taken twice a day. The medication was titrated over the course of 5 weeks to get to the final dose, to decrease the risk of syncope.

Conclusion: prazosin is an effective medication for decreasing nightmares, improving sleep and lessening PTSD symptoms. It is well tolerated when titrated slowly. Most patients still have substantial symptoms after a trial of prazosin, and should also be offered evidence-based treatment, such as exposure therapy. It is unknown how long prazosin should be continued. Side effects are likely to decrease over time, and the medication is generic and relatively inexpensive.


A Trial of Prazosin for Combat Trauma PTSD With Nightmares in Active-Duty Soldiers Returned From Iraq and Afghanistan

Murray A. Raskind, M.D.: et al

Am J Psychiatry 2013;170:1003-1010. doi:10.1176/appi.ajp.2013.12081133


Can Antipsychotics help Depression?

[Short take: Not first-line for most people with depression. Newer antipsychotics help 1 out of 10 takers achieve remission. They have a number of potentially serious side effects.]

Major depression affects between 10% and 20% of people in the US at some point in their lifetime. Though there are many antidepressants available, from a number of different classes of medication, the majority of people that take an antidepressant do not get a complete response to the first, or even second medication they take. As each new treatment is tried, and fails, the likelihood of a good response decreases. As a result, many with depression are left with no response or only a partial response  and are unsure what to do next. Psychotherapy, particularly the evidence based forms (such s Cognitive Behavior Therapy (CBT), Commitment and Acceptance Therapy (ACT), Interpersonal Therapy (IPT), and probably mindfulness based therapies founded on CBT), are successful as first-line treatments and good alternatives to medication. They should also be tried in treatment resistant depression. Finding a well-trained therapist can sometimes be difficult, though perhaps no more difficult than finding someone up to date in psycho-pharmacology. Both medications and psychotherapies have benefits and risks.

In recent years, in an effort to expand the market for their medications, pharmaceutical companies have funded a number of large, well-done studies to assess the usefulness of adding the newer antipsychotics to antidepressants for treating major depression. The FDA has approved 2 of these medications, quetiapine and aripiprazole, for major depression that is not responding adequately to antidepressants (olanzapine has been approved in a fixed-dose combination with fluoxetine, but not as an add-on to any antidepressant). Unlike T3 and Lithium, which have been studied primarily with older antidepressants, the second generation* antipsychotics have been studied as an add-on to SSRI’s (sertaline, fluoxetine, paroxetine, citalopram, escitalopram) and SNRI’s (venlafaxine, duloxetine).

Do the atypical* anti-psychotics help decrease symptoms of depression when added to an antidepressant? Given the size of the studies and their statistical power, it is clear that they do help. When compared to adding a placebo, quetiapine and aripiprazole help more people respond (meaning that the symptoms decrease by at least 50%) and achieve remission. And the experts all agree, remission is what counts. When a person is in remission from depression they are back to normal, or at least almost completely back to their non-depressed baseline. And this is very important, since even mild residual symptoms of depression increase the risk of depression coming back, and each episode of depression increases the risk of another episode and the risk of entering a state of chronic depression.

How well do they work? Simply put, not very well, particularly given the potential side effect burden associated with them. The NNT (number needed to treat) is 10. This means that compared to placebo, the addition of an SGA to an antidepressant will help 1 in 10 people achieve remission. NNT is one of simplest ways of looking at the clinical meaningfulness of a treatment. Studies frequently report results that are statistically significant, but statistically significant is not the same as clinically meaningful. A large study may have a very strong finding in a statistical sense, but the treatment may not help very many people, or not help them very much. An NNT of 2 or 3 indicates a very useful treatment. NNT’s of 4-7 are helpful.

Partial or non-response to antidepressants is a big problem, and reasonable and effective treatment options are badly needed. The SGA’s are effective, but helping only 1 in 10 is on the borderline of usefulness. An NNT of 10 is considered marginal, and though a completely benign treatment might be looked at favorably, these medications are not free of side effects.

What are the side effects of the atypical antipsychotics? The side effects vary somewhat between medications, but can be substantial. Aripiprazole (Abilify) is less likely to cause sedation, weight gain, increased triglycerides and increased blood sugar than quetiapine (Seroquel XR), but more likely to cause restlessness and inability to stay still (akathisia). Long-term side effects have not been systematically studied for either of them, and it is unclear how much risk of tardive dyskinesia (abnormal muscle movements that can be permanent) is associated with their use. Since these medications can cause increased weight, blood sugar and lipids, they likely put people at risk of developing diabetes and cardiovascular disease.

Should SGA’s be used in treating depression? Yes, they definitely have a role. Following the recommendations of Connolly and Thase as stated in Management Of Treatment-Resistant Major Psychiatric Disorders, SGA’s should be considered as first-line add-on options when the antidepressant is well-tolerated enough to be continued, and there is some urgency in decreasing symptoms. The evidence supports them as having a fairly rapid effect, with some improvement seen within 1-2 weeks. The other situation in which they are first-line and clearly indicated is when psychotic symptoms are present along with depression.

To sum up: Lack of a complete response to an antidepressant is the norm. There are many options, but not a lot of information to guide the selection of treatments, or the best sequence to use them in. If a person has taken just one antidepressant and not had a good response, and the situation is not urgent, then adding an antipsychotic is not likely to be the next step. On the other hand, if someone has had several antidepressants, and their depression is not getting better, then an antipsychotic could be a very helpful option. Unanswered is the question of how long the antipsychotic should be taken, and monitoring of side effects is important, particularly weight gain, metabolic and muscular ones.

* Several terms are used in referring to a group of medications that were initially used in treating schizophrenia, and which now are commonly used in treating Bipolar Disorder, Major Depression, and other conditions; these include atypical antipsychotics, second-generation antipsychotics (SGA’s), and newer antipsychotics. This is to distinguish them from the older, or typical, antipsychotics such as Thorazine, Haldol, Stelazine and Prolixin. Though the SGA’s all block dopamine, and hence share the ability to decrease psychotic symptoms, there is a lot of variability among them, and antipsychotic may not be a very good term for them. They are effective at treating manic episodes, preventing recurrence of manic episodes and decreasing depressive symptoms, among other things.

References for Treatment Resistant Depression

1) Connolly,K.R., Thase,M.E. (2012) Systematic Approach to Treatment-Resistant Depression: Psychopharmacology and Psychotherapy. In C.B.Nemereoff, Management Of Treatment-Resistant Major Psychiatric Disorders(pp. 23-48). New York: Oxford University Press.

2)Phillip,S.P.,Pharmacologic Approaches to Treatment Resistant Depression:A re-examination for the Modern Era. Expert Opin Pharmacother. 2010 April;11(5):709-722. doi:1517/1465651003614781.

Chantix (varenicline) for Smoking Cessation

For any psychiatrist, the following is a very common scenario.

I am seeing Frank, a 45 year old man, who is afflicted with schizophrenia. He has been smoking heavily since he was a teenager. Though Frank has heard many times that smoking is bad for him, in all kinds of ways, he has never given much thought to quitting, until now. It might be his recent health changes, the death of a relative, the high cost of a pack of cigarettes, the hassles of finding a place where he can smoke in peace, or some combination of these. Whatever the cause, he is willing to discuss smoking cessation, and smoking cessation aids, for the first time. We talk about the options (one of my favorite words): going cold turkey (commonly done, not very effective for most people), bupropion, nicotine replacement in all it’s forms (gum, patches, lozenges, inhaler, nasal spray), and Chantix. As we go over the pros and cons, I point out that Chantix is the most effective at helping people quit (I use the brand name rather than the generic name varenicline since most people know it as Chantix and it is not yet available as a generic). He is clearly interested in going with the most effective medication, and his case manager and I are excited at his intent to give it a try. But wait, doesn’t Chantix cause people to get depressed, even become suicidal and to try and kill themselves, or to hear voices? Isn’t it contraindicated for people with mental illness?

Soon after Chantix was released it became clear that it was very effective for some individuals, and was generally superior to the medications already being used. Nausea was the biggest problem.

However, not long after it’s release there were reports that it might cause depression, suicidal thoughts, suicide attempts, psychosis and other neuropsychiatric side effects. That  presented smokers and their clinicians with a dilemma.  Smoking kills and disables very large numbers of people. Chantix was the most effective treatment available. Most people do not succeed with their first, second or even third attempt at quitting. It is hard to quit and many people get discouraged.  How dangerous is Chantix? Do the benefits outweigh the risks? Is it dangerous for everybody, or just some people?

These questions generated considerable interest and study, and not just from the makers of Chantix (who have such a strong vested interest that they cannot be trusted to study the issues objectively). We have learned, among other things, that smokers as a group are more prone to depression, and are more likely to attempt and commit suicide than non-smokers. Also, that the act of quitting smoking puts a person at risk of developing depression and suicidal thoughts. The question is not is Chantix dangerous, but is taking Chantix to quit smoking more dangerous for your mental health than other ways of quitting, such as cold turkey, or with nicotine replacement therapy, or with bupropion. And we now have evidence to answer that question, and I think the answer is good news for smokers.

In a study published online on Sept. 13, 2013 in AJP in Advance, researchers from the University of Chicago and at Columbia University performed the largest analyses to date, by combining data from a number of studies that in total included more than 40,000 smokers.

The data showed that varenicline was highly associated with inducing nausea among patients, but not suicide events, depression, or aggression. Current or past psychiatric illness increased the risk of neuropsychiatric events equally among the varenicline and placebo groups. In the drug-comparison studies, the rate of neuropsychiatric events in the varenicline cohort was significantly less than in those receiving nicotine-replacement therapy. Overall, varenicline was more successful in achieving smoking abstinence than placebo or nicotine-replacement therapy.”

In plain language: Chantix was more effective than nicotine replacement and bupropion, and it was no more likely than placebo to be associated with depression, suicidal thoughts, and other neuropsychiatric effects such as hallucinations.

Getting back to the questions asked earlier: Chantix is not contra-indicated in those with a current or past history of mental illness. The evidence at this point, and there is now good data to guide treatment decisions, is that Chantix is more effective than nicotine replacement and bupropion, and possibly even safer. And, it is no more dangerous than taking no medication when trying to quit smoking and a lot more effective than going cold turkey.

Varenicline, Smoking Cessation, and Neuropsychiatric Adverse Events  Robert D. Gibbons, Ph.D.; J. John Mann, M.D.                Am J Psychiatry 2013;:. doi:10.1176/appi.ajp.2013.12121599

Declaration of interest: I have no financial interest in Chantix or an other product used in helping people quit smoking.