David N. Osser, MD

Do you prescribe antidepressants for your patients with bipolar depression? If so, think again, and consider results from the Systematic Treatment Enhancement Program–Bipolar Disorder (

) studies. This was the largest federally funded group of studies of bipolar disorder treatment, and they included important randomized clinical trials involving subsamples of the 4360 patients enrolled.

Among the significant findings were the following:

1. Antidepressants (bupropion, paroxetine) are not more effective than placebo for bipolar depression (24% for the antidepressants versus 27% for the placebo in a 6-month trial).1 The antidepressants, when added to a mood stabilizer, did not induce more switches into mania (10% versus 11%), but the patients who participated in the study were probably at very low risk for switching. However, patients with bipolar depression with mixed features (defined as having 2 or more manic symptoms [eg, agitation and racing thoughts] with their depression) developed manias over the next several months that were more severe as compared with placebo-treated patients.

2. A group of 86 patients with bipolar disorder who responded to an antidepressant when depressed were identified. Some were rapid cyclers (4 or more episodes per year). All of the patients were also prescribed mood stabilizers such as lithium, valproate, or a second-generation antipsychotic. These patients were randomized to stay on the antidepressants or discontinue them. The patients with rapid cycling who were continued on their antidepressant had triple the number of depressions per year compared with those without rapid cycling. Yet, there was no difference in the rate of depressions between rapid and non-rapid cycling in the antidepressant discontinuation group. Thus, depressive morbidity and cycling were worsened by continuation of an antidepressant in patients with rapid cycling.

 were collected on the first 1500 patients looking for evidence of the antidepressant-associated chronic irritable dysphoria (ACID) syndrome described in previous case reports.

 It had been proposed that ACID was caused by treating patients who had bipolar depression with an antidepressant. The patient sample included 63% bipolar I patients with depression; most of the remaining patients had bipolar II. The researchers identified 83 patients in whom a depressive episode developed during the study period and for whom there was at least 1 year of follow-up information. Those who received an antidepressant for their depression had a 10-fold higher risk for ACID than the patients who were antidepressant free. ACID did not seem to occur in the natural course of bipolar disorder; it only occurred in those who received antidepressants.


Findings from STEP-BD studies indicate that antidepressants should be avoided in most cases when treating acute depression in patients with bipolar disorder. Rather, one should choose from among 5 medications with a reasonable evidence-base for effectiveness and safety in treating or at least preventing bipolar depressions, namely: lithium, quetiapine, lamotrigine, lurasidone, and cariprazine—or combinations of these.

 is Associate Professor of Psychiatry, Harvard Medical School, and Consulting Psychiatrist, US Department of Veterans Affairs, National Telemental Health Center, Bipolar Disorders Telehealth Program, Brockton, MA.

1. Sachs GS, Neirenberg AA, Calabrese JR, et al. 

Effectiveness of adjunctive antidepressant treatment for bipolar depression

2007;356:1711-1722.
2. El-Mallakh RH, VÃ¶hringer PA, Ostacher MM, et al. 

Antidepressants worsen rapid cycling course in bipolar depression: a STEP-BD randomized clinical trial

2015;184:318-321.
3 El-Mallakh RH, Ghaemi SN, Sagduyu K, et al.

 Antidepressant-associated chronic irritable dysphoria (ACID) in STEP-BD patients

This article was originally posted on April 24, 2020, and has since been updated. -Ed

Articles

Do you prescribe antidepressants for your patients with bipolar depression? If so, think again, and consider results from these studies.

The Case Against Antidepressants for Bipolar Depression: Findings from STEP-BD

We frequently encounter patients who use cannabis. More than 50 indications for medical marijuana are approved by various state governments, although no indications are FDA-approved. Marijuana is legal in 36 states-and 10 states allow recreational use-but regulation of the quality or purity of these products is minimal. Some patients are convinced that marijuana calms them, uplifts them, blunts their anxiety, enables them to sleep, and has other benefits like helping with pain. Many report that it is the only thing that is reliably helpful for neuropsychiatric symptoms, compared with prescribed medications. Notably, scores of nicotine users say the same thing: it is their coping strategy of choice for just about every stress or even their “only pleasure in life.” Others say the same things about benzodiazepines. Rhetorical question: What do these substances have in common?

Regardless of whatever short-term benefit patients perceive from cannabis, the evidence points clearly to an association between usage and worsening course of bipolar disorder over time. In a study of 4915 participants, Henquet et al.

 found a strong increased risk of manic symptoms associated with cannabis over a three year follow-up (after controlling for possible covariates). They also saw an earlier age of onset of bipolar disorder, greater overall illness severity, more rapid cycling, poorer life functioning, and poorer adherence with prescribed treatments.

 evaluated the subsequent course of patients with bipolar disorder who stopped cannabis use after an illness episode and compared their outcome with patients with bipolar disorder who had never used cannabis and a group that continued to use. The total sample included 1922 patients. In a two year period, the continued users had significantly lower rates of recovery, greater work impairment, and fewer were living with a partner. The data were based on patient reports; given that fact, there was likely under-reporting and probably an underestimate of the association between cannabis use and lives worsened. A systematic review of the effects of cannabis on mood and anxiety disorders confirmed a negative association between cannabis use and long-term outcomes.

Some patients exhibit psychotic symptoms after cannabis use. A recent study found that schizophrenia or bipolar disorder developed in many of these patients.

 This happened in 47% of patients who became psychotic on cannabis over the next four years. This was the highest incidence of conversion after initial psychosis following use of various substances: the second highest was with amphetamine, at 32%. More patients converted to schizophrenia than to bipolar. As for any possible benefits, a recent review concluded that the evidence supporting use of cannabis for psychotic (or other psychiatric) disorders is “very low” and “inadequate.”

The non-psychiatric medical benefits of cannabis are also very thinly evidenced despite outsized claims to the contrary.

 It is said to be “irresponsible” to encourage patients addicted to opiates to switch to cannabis for their problems with pain.

Thus, it seems that patients with bipolar disorder should stay away from cannabis in all its forms. Quitting cannabis should be on the short list of interventions if patients are not doing well. This is a tough sell in today’s political environment regarding cannabis legalization. Many newspaper editorials and politicians support its beneficial effects and use. Clinicians should not back down and accept patients’ insistence on using this product; rather, they should continue efforts to educate and to consider the problem to be a serious one that potentially interferes with otherwise appropriate and effective bipolar treatments that may be offered.

1. Henquet C, Krabbendam L, de Graaf R, et al. Cannabis use and expression of mania in the general population. J Affect Disord. 2006;95:103-110.

2. Zorrilla I, Aguado J, Jaro JM, et al. Cannabis and bipolar disorder: does quitting cannabis use during manic/mixed episodes improve clinical/functional outcomes? Acta Psychiatrica Scand. 2015;131:100-110.
	3. Mammen G, Rueda S, Roerecke M, et al. Association of cannabis with long-term clinical symptoms in anxiety and mood disorders: a systematic review of prospective studies. J Clin Psychiatry. 2018;79:17r11839.

4. Starzer MSK, Nordentoft M, Hjorthoj C. Rates and predictors of conversion to schizophrenia or bipolar disorder following substance-induced psychosis. Am J Psychiatry. 2018;175:343-350.
	5. Radhakrishnan R, Ranganathan M, D’Souza DC. Medical marijuana: what physicians need to know. J Clin Psychiatry. 2019;80:45-47.

6. Hill JH. Medical use of cannabis in 2019. JAMA. 2019;322:974-975.

7. Humphreys K, Saitz R. Should physicians recommend replacing opioids with cannabis? JAMA 2019; 321(7): 639-640.

Regardless of whatever short-term benefit patients perceive from cannabis, the evidence points clearly to an association between usage and worsening course of bipolar disorder over time.

Cannabis: Patients With Bipolar Should Avoid Use

Do you need help persuading patients to try lithium? A major but often underemphasized benefit of lithium is its unique neuroprotective effects. Bipolar disorder seems to be a condition characterized by gradual progressive shrinkage in some areas of cortical gray matter as well as some tracks of white matter. These losses are associated with increased neurocognitive impairment. However, in patients who have been treated with long-term lithium, there is better preservation of white matter structural integrity.

 Reduction of gray matter volume in bipolar patients has been shown to be arrested, or even reversed to some degree, by lithium. No other mood stabilizers have this effect. Some ambivalent patients have been persuaded to accept lithium after reading the article by Giakoumatos and colleagues.

At the microcellular level (in rats), lithium is protective against neuronal apoptosis (destruction) and glutamate-induced cytotoxicity, and it promotes dendritic length and number in the hippocampus.

 The preservation of cognitive function in humans with bipolar disorder seems to occur through the brain-derived neurotrophic factor (BDNF) system: lithium markedly increases secretion of BDNF from cortical and hippocampal neurons and upregulates intracellular production of relevant proteins.

The evidence for lithium’s neuroprotection even extends to the prevention, treatment, or slowing of mild cognitive impairment, Alzheimer disease, Parkinson disease, and amyotrophic lateral sclerosis in non-bipolar patients.

 Even low-dose lithium has been found in preliminary studies to slow the progression of Alzheimer disease.

Given this information, it should not be surprising that the benefits of lithium treatment seem to exist for patients across the life span.

 However, there is no reason to wait: the long-term benefits seem greatest when lithium is started early in the course of the illness,

 perhaps because some of the brain structural losses may be irreversible. These mechanisms of lithium action also help explain the evidence that of all medication regimens used for bipolar treatment, lithium is the medication most likely to be maintained as a monotherapy over time. This has been demonstrated in surveys of bipolar patients receiving maintenance treatment in the US and in two European studies.

Dr Osser is Associate Professor of Psychiatry, Harvard Medical School, and Consulting Psychiatrist, US Department of Veterans Affairs, National Telemental Health Center, Bipolar Disorders Telehealth Program, Brockton, MA.

 Giakoumatos CI, Nanda P, Mathew IT, et al. Effects of lithium on cortical thickness and hippocampal subfield volumes in psychotic bipolar disorder. 

Nunes MA, Schowe NM, Monteiro-Silva KC, et al. Chronic microdose lithium treatment prevented memory loss and neurohistopathological changes in a transgenic mouse model of Alzheimer disease. 

De-Paula VJ, Gattaz WF, Forlenza OV. Long-term lithium treatment increases intracellular and extracellular brain-derived neurotrophic factor in cortical and hippocampal neurons at subtherapeutic concentrations. 

Forlenza OV, De-Paula VJ, Diniz BS. Neuroprotective effects of lithium: implications for the treatment of Alzheimer disease and related neurodegenerative disorders. 

Young RC, Mulsant BH, Sajatovic M, et al. GERI-BD: A randomized double-blind controlled trial of lithium and divalproex in the treatment of mania in older patients with bipolar disorder. 

Kessing LF, Vradi E, Andersen PK. Starting lithium prophylaxis early v. late in bipolar disorder. 

Hayes JF, Marston L, Walters K, et al. Lithium vs. valproate vs. olanzapine vs. quetiapine as maintenance monotherapy for bipolar disorder: a population-based UK cohort study using electronic health records. 

Do you need help persuading patients to try lithium?

The Neuroprotective Effects of Lithium

There appears to be some confusion regarding the efficacy of aripiprazole in bipolar mania and depression and for preventing bipolar episodes. Aripiprazole is approved by the US Food and Drug Administration as a treatment for acute mania. It is not one of the more robustly efficacious agents for mania, however; haloperidol, risperidone, olanzapine, and quetiapine (in that order) top the list in a meta-analysis by Cipriani and collegues.

Haloperidol is not recommended as a first-line agent because of the risk of tardive dyskinesia (more common in mood disorder patients versus schizophrenia). Moreover, evidence indicates that it is the antimanic agent associated with the highest risk switching to a depressive episode after resolution of the mania.

 Aripiprazole had two early trials in which it did not differ from placebo in mania, which lowered its effect size compared with other options. Efficacy was found in studies using 15 to 30 mg doses.

In bipolar depression, the situation is a little more complicated, but in short, aripiprazole has not been found to be effective. Two large manufacturer-sponsored trials failed to show any difference from placebo after eight weeks of treatment.

 It is not FDA-approved for bipolar depression. What is confusing for clinicians is that aripiprazole is well-known for its antidepressant effect as an augmenter of selective serotonin reuptake inhibitors in unipolar depression. It was heavily advertised for that purpose (at least, until it recently became generic). It has never been studied as an augmentation for antidepressants in bipolar depression.

The use of antidepressants (with or without augmenters) in bipolar depression remains controversial at best. The exception is the combination of olanzapine and fluoxetine, which is effective and FDA-approved for bipolar depression, although one usually wants to avoid it because olanzapine has severe metabolic adverse effects and induces insulin-resistance.

In the 8-week aripiprazole studies, there was separation from placebo in some of the earlier weeks, but by week 8 there was no difference (number needed to treat = 44).

 Dosage at endpoint reached a mean of 16.5 mg daily. Some have speculated that if they had used lower doses (eg, starting at 2 to 5 mg and titrating to 5 to 10 mg), the researchers might have found a more sustained antidepressant effect. The proposed mechanism is that at lower doses there could be less akathisia and it might have kept working on the depression over the full eight weeks. This is possible and deserves study.

However, in the one maintenance study of aripiprazole in bipolar disorder (which resulted in FDA-approval as a maintenance agent in bipolar disorder), there was significant benefit for reducing manic episodes over six months compared with placebo, but no efficacy in preventing depressions or mixed states with depressive symptoms.

Other antipsychotics used as monotherapy have been found effective for bipolar depression. These include quetiapine, lurasidone, and cariprazine (all FDA-approved). Lamotrigine and lithium have efficacy and FDA-approval as maintenance treatments in bipolar depression. Thus, for many patients with bipolar depressions, these five medications would be preferred over aripiprazole.

Dr Osser is Associate Professor of Psychiatry, Harvard Medical School, and Consulting Psychiatrist, US Department of Veterans Affairs, National Telemental Health Center, Bipolar Disorders Telehealth Program, Brockton, Mass.

1. Cipriani A, Barbui C, Slanti G, et al. Comparative efficacy and acceptability of antimanic drugs in acute mania: a multiple treatments meta-analysis. Lancet. 2011; 378:1306-1315.

2. Goikolea JM, Colom F, Torres I, et al. Lower rate of depressive switch following antimanic treatment with second-generation antipsychotics versus haloperidol. J Affect Disord. 2013;144:191-198.

3. Thase ME, Jonas A, Khan A, et al. Aripiprazole monotherapy in nonpsychotic bipolar I depression: results of 2 randomized, placebo-controlled studies. J Clin Psychopharm. 2008;28:13-20.

4. Keck PE Jr, Calabrese JR, McQuade RD, et al. A randomized, placebo-controlled 26-week trial of aripiprazole in recently manic patients with bipolar I disorder. J Clin Psychiatry. 2006;67:626-637.â

There appears to be some confusion regarding the efficacy of aripiprazole in bipolar mania and depression and for preventing bipolar episodes.

Clinical Updates From Bipolar Telehealth

Lamotrigine monotherapy has emerged as a possible treatment for bipolar depression. It is not associated with weight gain and is less likely to cause neurocognitive adverse effects and sedation. Most patients tolerate it well. It does carry a risk of dangerous rashes, but the risk is minimized by slow dosage titration and alert monitoring. The incidence of serious rash may also be less than previously reported (in particular, when faster dosage titration was used). The rate now appears to be 0.1% or less.

The evidence supporting lamotrigine usage for acute bipolar depression, however, is mixed. On the positive side, one double-blind, placebo-controlled study of lamotrigine (50 mg/d, 200 mg/d, or placebo) in bipolar I depression (n = 195) showed favorable results.

 At 50 mg, 41% of patients improved;on 200 mg, 51% responded; and 26% improved on placebo.

These positive findings about lamotrigine had been countered, however, by four negative studies that are large, industry-supported, double-blind, placebo-controlled clinical trials studying lamotrigine treatment in acutely depressed bipolar I and II patients. None of the four studies found a statistical difference between lamotrigine and placebo. As a result, lamotrigine did not receive Food and Drug Administation approval for acute bipolar depression. A meta-analysis of these five studies, however found a small benefit with an overall effect size of 0.27.

 In more severely ill patients (Hamilton 24 or more), lamotrigine had a greater separation from placebo (0.47) mostly because placebo effect was lower in this group. Lamotrigine was not better than placebo if the baseline Hamilton was lower than 24 (0.07 effect size). In another approach to predicting response, a small observational study showed better results at lamotrigine blood levels around 4 ng/mL.

The efficacy of lamotrigine as maintenance therapy is fairly robust. Two large, 18-month studies showed efficacy, enabling lamotrigine to obtain FDA approval for maintenance use.

 It had no efficacy for preventing mania, but at least it did not increase the risk of mania compared with placebo.

Lamotrigine has shown no efficacy in treating acute mania, which makes it less desirable than lithium, quetiapine, or cariprazine in that there will be no coverage for the manic/hypomanic phases. It also has no apparent benefit for suicidal ideation or behaviors. In fact, like all anticonvulsants, it carries a warning about possible increased risk of suicidality. On the other hand, its relatively benign side effect profile might make it a first choice for some patients, especially if the past hypomanias have been mild.

Lamotrigine might also be worth considering in women of childbearing potential. The teratogenic risk of lamotrigine is approximately 1% to 4%, the lowest among the antiepileptic medications. A large observational study found an odds ratio compared with controls of 1.3 for orofacial cleft abnormalities.

 If, after carefully weighing the options, lamotrigine is selected for these women, close monitoring of serum level is prudent as dosing and serum concentration dramatically differ depending on the stage of pregnancy.

 Calabrese JR, Bowden CL, Sachs GS, et al, for the Lamictal 602 Study Group. Double-blind placebo-controlled study of lamotrigine monotherapy in outpatients with bipolar I depression. 

 Geddes JR, Burgess S, Hawton K, et al. Long-term lithium therapy for bipolar disorder: systematic review and meta-analysis of randomized controlled trials. 

 Kesebir S, Akdeniz F, Demir A, Bilici M. A comparison before and after lamotrigine use in long-term continuation treatment: effect of blood level. 

 Bowden CL, Calabrese JR, Sachs G, et al. A placebo-controlled 18-month trial of lamotrigine and lithium maintenance treatment in recently manic or hypomanic patients with bipolar I disorder. 

 Calabrese JR, Bowden CL, Sachs, G, et al. A placebo-controlled 18-month trial of lamotrigine and lithium maintenance treatment in recently depressed patients with bipolar I disorder. 

 Dolk H, Wang H, Loane M, et al. Lamotrigine use in pregnancy and risk of orofacial cleft and other congenital anomalies. 

The good, the bad, and the efficacy of lamotrigine.

Lamotrigine: Its Role in Bipolar Disorder

An international task force of experts in the diagnosis and treatment of bipolar disorder was recently convened to address concerns about DSM-5 criteria, particularly the definitions of manic episodes. This initiative was funded by the National Health and Medical Research Council of Australia; 64 individuals from 14 countries spanning five continents responded to a questionnaire probing different aspects of the criteria.

 The findings are pertinent for clinicians struggling to make the bipolar I and II diagnoses, which have been found to be both underdiagnosed and overdiagnosed in different studies. The consequences for patient outcomes with these missed diagnoses can be severe.

The DSM’s duration criteria is an area of concern. Currently, it includes seven days for mania and four days for hypomania. Patients with only brief episodes who meet phenotypic criteria are excluded from the diagnosis, and a lot of time is spent with patients struggling to find if they have had the syndrome for four or more days. Some task force members felt that any duration of this syndromal criteria should be sufficient. However, most responders (75%) thought that ultra-short episodes have many possible explanations, and some limit should be imposed. Of those nominating a finite period, the modal recommendation was two days for both mania and hypomania.

Thus, for patients who have a sufficient number of symptoms to meet criteria, it may be that [hypo]mania durations of as little as two days could be a reasonable basis for concluding that the diagnosis is probably bipolar disorder and treatments should be chosen accordingly.

The group also came up with a new proposed definition for hypomania and mania:

A distinct period of either an abnormally elevated and expansive mood, or an irritable and quick-tempered mood, characterized by increased activity and cognition such that the individual feels unusually energized, hyper, or wired. Such changes are perceived at the time or on later reflection as excessive and an unambiguous change in functioning from the individual’s usual state.

The consensus was that this characterized the disorder better than the current DSM definition.

In differentiating bipolar I from bipolar II, the majority voted to continue to require marked disturbance in functioning at work or socially for bipolar I (mania). However, the requirement that hospitalization would automatically qualify the patient for bipolar I was strongly opposed. This requirement was considered arbitrary and too dependent on local health services practices, insurance, or other external factors. However, if the patient was hospitalized, that certainly could be an indicator of severity and likely mania. The group thought psychosis should be retained as a criterion indicating mania.

The group has studies underway to test these proposed changes and others to see how well they describe a consecutive series of patients in clinical practice. They are also looking at 78 possible symptoms of mania to see if any are better than the seven in DSM-5.

Dr Osser is Associate Professor of Psychiatry, Harvard Medical School, and Consulting Psychiatrist, US Department of Veterans Affairs, National Telemental Health Center, Bipolar Disorders Telehealth Program, Brockton, MA. He reports no conflicts of interest concerning the subject matter of this article.

1. Parker G, Tavella G, Macqueen G, et al. Revising DSM-5 criteria for the bipolar disorders: Phase 1 of the AREDOC project. 

Bipolar II Disorder: Modelling, Measuring, and Managing

, 3rd ed. New York, NY: Cambridge University Press; 2019

An international task force of experts in the diagnosis and treatment of bipolar disorder was recently convened to address concerns about DSM-5 criteria, particularly the definitions of manic episodes. The DSM’s duration criteria is an area of concern.

Mania and Hypomania: The Latest Thinking on Duration of Episodes and Other Features

In the past several years at the VA National Tele-Bipolar Disorder Expert Consultation Program, my colleagues Drs Eric Smith, Laszlo Gyulai, and I have been writing monthly educational newsletters for clinicians (nationwide) who have been requesting telemedicine consultation on diagnostic and treatment problems of patients with bipolar disorder or possible bipolar disorder.

 Over 2500 consultation requests have been received over more than 8 years and 2100 consultation notes have been delivered, so our group has developed considerable experience with this challenging disorder and the wide variety and complexity of the clinical scenarios and comorbidities that are encountered.

The newsletters are designed for easy and quick reading: all are one typed page in length, with one to three reference citations. The Editor and Board of 

 learned of these newsletters and requested to have them published (after review) and made available to their readership. The first one can be read 

. The content of these newsletters does not reflect the position of the Department of Veterans Affairs and is solely the product of the authors.

As the coordinator of these publications, I should introduce myself. I have an appointment as Associate Professor of Psychiatry at Harvard Medical School at the VA Boston Healthcare System, Brockton Division. My area of clinical and academic expertise is in the development of evidence-derived algorithms for the psychopharmacology of psychiatric disorders, an endeavor I have been dedicated to since 1985. These algorithms have been published and republished in updated versions and (since 1995) converted to web-based applications that enable rapid and international access to the recommendations through the (free) website of the Psychopharmacology Algorithm Project at the Harvard South Shore Residency Training Program (psychopharm.mobi).

Dr Robert Patterson, a psychiatrist affiliated with McLean Hospital, created the technology for translating the algorithms into their web versions. 

 has published articles over the years about the algorithms.

 There are algorithms for 10 disorders. There is no commercial activity associated with the website nor have we received any funding for our work from pharmaceutical firms or others: it has been a labor of love for Dr Patterson and me and about 25 collaborators on the different algorithms. It was a great privilege in 2011 to be offered the opportunity to do these consultations and see these complex patients and refine my thinking about their needs and the optimal application of the evidence-base to their care.

To introduce the columns, I begin with a few brief comments about the challenges of diagnosing bipolar disorder. On this, the evidence is clear: the condition is frequently both underdiagnosed and overdiagnosed.

 Patients having manic episodes may be functioning fairly well (or, as they see it, very well indeed) when manic and may not recognize or want to acknowledge the abnormal mood state, and may not seek care. Erroneous diagnoses and incorrect treatment can continue for many years.

The rate of suicide in bipolar disorder is quite high. Hence, the most important initial advice we can give readers of this column is to please pull out your DSM-5 and review the criteria for mania and hypomania. Take the time to carefully assess the patient using these criteria, looking particularly for a history of discreet episodes lasting at least 3 to 4 days. It can help to have a significant other who knows the patient well give input as to whether these discreet episodes have been occurring. Avoid making a positive diagnosis of mania or hypomania after hearing about or observing just a couple of symptoms (eg, racing thoughts, rapid speech, tangential speech). Quick diagnosis on a hunch is a very error-prone process with bipolar disorder. Patients with rapid or overly productive speech and thoughts could have ADHD, PTSD, substance use disorders, or other problems-or all of them at once. If so, each needs to be treated separately-prioritized in order of importance.

 readers. There will be some surprises coming in future issues!

Dr Osser is a Consulting Psychiatrist, US Department of Veterans Affairs, National Telemental Health Center, Bipolar Disorders Telehealth Program, Brockton, MA.

 Bauer MS, Krawczyk L, Tuozzo K, et al. Implementing and sustaining team-based telecare for bipolar disorder: lessons learned from a model-guided, mixed methods analysis. 

 Osser DN, Patterson RD. Enhancing clinical decision-making. 

 Osser DN. The role of guidelines and algorithms for psychopharmacology in 2007. 

 McIntyre RS, Zimmerman M, Goldberg JF, First MB. Differential diagnosis of major depressive disorder versus bipolar disorder: current status and best clinical practices. 

The evidence is clear: bipolar disorder is frequently both underdiagnosed and overdiagnosed.

Fasten Your Seatbelts

Current estimates are that 4% to 5% of the population is at risk for a disorder on the bipolar spectrum. Among the patients in the so-called soft portion of that spectrum are those with a disturbance of temperament in the direction of hypomania.

The concept of temperament is a product of German nosological research from a century ago starting with Kraepelin. In the US, the concept has been championed by Hagop Akiskal, MD and his colleagues. Akiskal is now the editor emeritus of the 

. The notion of depressive temperament has been incorporated into DSM-5 nosology in the form of “persistent depressive disorder” (previously called dysthymia). The other pole was called hyperthymia by the Germans. DSM committees have considered adding hyperthymia but have not done so. The research base on it is still, to many, unconvincing. However, it seems that in clinical practice one encounters individuals who have chronic low-grade hypomanic symptoms-high energy, need for less sleep than others, chronic optimism, chronic risk taking. These individuals can be prone to major depressions and can become severely suicidal.

Akiskal and colleagues have been describing these patients for almost 40 years. Their research criteria for hyperthymic temperament include onset before age 21, habitual sleep of less than 6 hours even on weekends, excessive use of denial, and traits (described originally by Schneider et al) that include being overoptimistic, self-assured, grandiose, overtalkative, warm and people-seeking, uninhibited, promiscuous, and meddlesome.

 Neurobiological studies have suggested the individuals have dopaminergic dysregulation.

Treatment issues have focused on what medications to use when hyperthymic individuals become depressed. The studies have all been uncontrolled. However, it seems that antidepressants are ineffective for these depressions and often trigger a mixed state or frank mania at times. Mood stabilizers and medications effective for bipolar depression may be more appropriate for the depressions in these patients.Usually their sunny temperament itself doesn’t require treatment and may, in fact, foster excellent productivity and creativity during much of their lifespan.

 Akiskal HS, Mallya G. Criteria for the “soft” bipolar spectrum: treatment implications. 

 Rihmer Z, Akiskal KK, Rihmer A, Akiskal HS. Current research on affective temperaments. 

Current estimates are that 4% to 5% of the population is at risk for a disorder on the bipolar spectrum. Among the patients of that spectrum are those with a disturbance of temperament in the direction of hypomania.

Hyperthymic Temperament

Dr Pies is Professor in the psychiatry departments of SUNY Upstate Medical University, Syracuse, NY and Tufts University School of Medicine, Boston. He is Editor in Chief Emeritus of 

 (2007 to 2010). Dr Osser is Associate Professor of Psychiatry, Harvard Medical School, Boston. He is an Editorial Board member of 

Consider the following vignettes and what they might teach us regarding so-called “antidepressant withdrawal.”

Mr A is a 40-year-old technician who has a diagnosis of MDD. He has been successfully treated with supportive therapy and 75-mg sertraline daily for the past 9 months. He has been in full remission for the past 6 months and asks his psychiatrist about “getting off the meds.” Since this is Mr A’s first episode of MDD, he and his psychiatrist decide that a cautious trial off the sertraline is reasonable.

The dose is reduced to 50 mg daily for 2 weeks; then 25 mg daily for 2 weeks; then 12.5 mg daily (1/2 of a scored 25-mg tablet) for 2 weeks. The patient is seen twice during this period and monitored carefully for any discomfort. Mr A tolerates the taper well, although he does report some mild nausea, occasional headaches, and hypersensitivity to sound, on 12.5 mg daily. Accordingly, the tapering is extended another 2 weeks (for a total of 8 weeks). The medication is then discontinued without difficulty.

Ms B is a 30-year-old financial manager with a history of three episodes of severe MDD over the past 10 years. She has been stable and doing well on paroxetine (60 mg/d) for the past 3 years, prescribed by her family physician, and is now strongly considering having a child. After discussing the risks of continuing an antidepressant during pregnancy, she and her doctor decide to discontinue the paroxetine. The dose is decreased to 40 mg daily for 2 weeks; then 20 mg daily for 1 week; then discontinued. During the first 2 weeks, the patient complains of occasional mild nausea, occasional abdominal cramping, and mild dizziness. However, Ms B is determined to “push ahead” with the tapering process and the schedule is maintained. On 20 mg daily, the patient reports no new or bothersome symptoms, and the paroxetine is stopped after a week.

Within 3 days, the patient calls her physician and complains of intense anxiety, insomnia, restless legs, diarrhea, dizziness, and what she describes as “brain zaps” (“like an electric shock to my head”). The paroxetine is immediately re-started at 20 mg daily, and the patient’s symptoms abated slowly over the next month.

These two vignettes-based on many patients we have encountered over a combined 70+ years of clinical experience-represent two possible outcomes of discontinuing serotonergic antidepressants (SSRIs and SNRIs). In our experience, the first scenario is far more common, and represents sound medical management of the tapering process. The second scenario is not one we have seen in our practice but is 

 among the “layperson withdrawal community.”

 Anecdotal reports of the second type-some with much worse outcomes-have led to 

 in the lay media that SSRIs and SNRIs are addictive, and ought to be grouped with known drugs of abuse.

 run by a psychiatrist (who will not prescribe antidepressants) declares that these medications “. . . are habit forming, so withdrawal can be excruciating.”

While we do not deny that severe reactions can and do occur when antidepressants are stopped suddenly (or the dose reduced too rapidly), we also believe that fears of such “excruciating” experiences are greatly overstated, in the context of proper psychiatric care. At the same time, we acknowledge that many “prescribers” of antidepressants-

 of whom are primary care physicians-discontinue antidepressants much too rapidly.

 Moreover, as critics of these drugs rightly point out, it is very hard to find detailed, professionally approved guidelines for tapering and discontinuation of antidepressants.

What do we know about antidepressant “withdrawal” syndromes? »

What do we know about antidepressant withdrawal syndromes?

Among the most thorough recent reviews of antidepressant discontinuation is that of 

 who comprehensively analyzed 18 clinical trials of SSRI discontinuation. (Note that these authors consistently use the term “discontinuation symptoms” rather than “withdrawal symptoms”-a point of great contention within the “layperson withdrawal community”

 but consistent with DSM-5, which describes “Antidepressant Discontinuation Syndrome” in terms quite similar to those of Jha and colleagues.) We would summarize the main conclusions of Jha and colleagues as follows:

1 SSRI discontinuation can lead to a wide variety of systemic and neuropsychiatric manifestations, including but not limited to flu-like symptoms; cardiovascular, gastrointestinal, and musculoskeletal complaints; worsening of mood, exacerbation of anxiety, perceptual abnormalities; and insomnia/nightmares.

2 SSRI discontinuation typically has its onset 1 to 10 days after drug discontinuation, and typically resolves spontaneously within 2 to 3 weeks. However, some reports document persistence of SSRI discontinuation (“withdrawal”) symptoms for up to 1 year.

3 The actual incidence of SSRI discontinuation symptoms is difficult to ascertain, because they are not systematically assessed in clinical practice; however, 

 found that up to 40% of patients reported new-onset of untoward symptoms after abrupt discontinuation of SSRIs (note the key term: abrupt).

4 For more than 60% of patients, SSRI discontinuation will not be associated with any significant symptoms. When present, symptoms range from mild to moderate-to-severe in intensity and, in some cases, may affect multiple organ systems.

5 Paroxetine has a short half-life (approximately 21-24 h), has the 

 of discontinuation symptoms, compared with other SSRIs.

 Paroxetine is also moderately anticholinergic; has no active metabolites; and undergoes non-linear pharmacokinetics, all of which may contribute to its propensity for discontinuation/withdrawal effects. Fluoxetine has a long half-life and probably has the lowest rates. Although not reviewed by Jha and colleagues, the short half-life SNRI, venlafaxine, is also 

 in discontinuation syndromes, apparently more so than other SNRIs.

6 The differential diagnosis of apparent SSRI discontinuation/withdrawal symptoms includes various medical conditions; discontinuation of other psychotropic medications; and, importantly, re-emergence of the initial depressive episode. 

 suggests that gradual tapering of the antidepressant does not seem to lower relapse rates, and patients with three or more prior episodes or a chronic course have a much higher relapse risk, after discontinuation of antidepressants.

7 The pathophysiology of SSRI discontinuation/withdrawal is not completely understood, but probably involves (among other mechanisms) reduced levels of serotonin in the CNS.

8 SSRI discontinuation/withdrawal can usually be resolved within 2 to 3 days by re-institution of the discontinued agent, or introduction of another SSRI. Substituting fluoxetine as a long-acting “buffering” agent is often recommended.

9 Antidepressants may need to be taken for at least 4 to 6 weeks, before the development of SSRI discontinuation/withdrawal becomes a serious concern. There is a general consensus that the longer the patient has taken an SSRI, the more likely or more serious the syndrome, following abrupt discontinuation.

10 SSRI discontinuation/withdrawal can be managed via prevention and treatment strategies, including educating patients regarding risk of discontinuation, using gradual discontinuation, and by resumption of an SSRI for burdensome symptoms.

Fava and colleagues disagree with the “mainstream” view of Jha and colleagues.

Clinicians are familiar with the withdrawal phenomena that may occur from alcohol, benzodiazepines, barbiturates, opioids, and stimulants. The results of this review indicate that they need to add SSRI to the list of drugs potentially inducing withdrawal phenomena. The term “discontinuation syndrome” minimizes the vulnerabilities induced by SSRI and should be replaced by “withdrawal syndrome.”

Furthermore, they dispute the often-repeated claim that gradual tapering of SSRIs and SNRIs greatly reduces the likelihood and/or severity of withdrawal syndromes. They state, “Gradual tapering does not eliminate the risk of withdrawal reactions. Indeed, a significant advantage of gradual tapering compared to abrupt discontinuation did not emerge.”

 that many clinicians-including some psychiatrists-have underestimated the potential severity and duration of antidepressant discontinuation/withdrawal syndromes. A widely publicized study generally supported this view, finding that an average of 56.4% of people said they had antidepressant withdrawal symptoms; and of these people, about 46% rated their symptoms as severe.10 The duration of these symptoms ranged from a few days to several months.

However, a commentary on this study by the British National Health Service (NHS) 

 that “. . . the study’s limitations mean we can’t be sure these figures are accurate.”

 For example, the three largest studies used to calculate the percentage of people experiencing withdrawal symptoms utilized self-reports drawn from online surveys. As the NHS commentary pointed out,

 as people are more likely to respond to a survey if they have experienced a problem than if they haven’t. This means the results [of the Davies and Read study] may overestimate the proportion of people who experience antidepressant withdrawal.

Antidepressants are not “addictive” »

 paper spurred some alarming reports in the lay press, which proclaimed, for example, that “

” and implying that antidepressants are addictive drugs.

 We strongly disagree with that characterization and do not believe that SSRI/SNRI discontinuation/withdrawal symptoms should be lumped together with those of clear-cut drugs of abuse, such as alcohol, opioids, and barbiturates.

First of all, there is no conclusive evidence showing that the pathophysiological mechanisms underlying SSRI/SNRI withdrawal symptoms are similar to those in, for example, alcohol, opioid, barbiturate, or benzodiazepine withdrawal. In addition, the usual hallmarks of “addictive” drugs-such as craving for the drug, compulsive use, intentional overuse, “getting high,” diversion for illicit use, etc-are 

 of SSRI/SNRI antidepressants as a class.

Furthermore, the claim by some groups that antidepressants are associated with the development of tolerance requires context and clarification. Unlike genuine substances of abuse, such as alcohol or barbiturates, true tolerance is not commonly observed with antidepressants. The related phenomenon of tachyphylaxis-defined as loss of a previously effective antidepressant treatment response despite staying on the same drug and dosage-is observed in only about 25% of 

 Results from a study by Zimmerman and colleagues

 show that the great majority of relapses attributed to so-called “poop out” of the antidepressant are actually due to the initial response having been a placebo response.

Viewed clinically, there are important, substantial differences between SSRI/SNRI withdrawal symptoms and the syndromes that follow abrupt discontinuation of drugs of abuse. For example, the 

 among patients exhibiting delirium tremens during withdrawal from alcohol is 5% to 25%.

 Barbiturate withdrawal may be accompanied by seizures in up to 75% of subjects, and about 66% may develop a delirium lasting several days. In this context, delirium-related agitation and hyperthermia can lead to exhaustion, cardiovascular collapse, and death.

Potentially lethal withdrawal reactions appear to be exceedingly rare, or non-existent, when SSRI/SNRI antidepressants prescribed in therapeutic doses are abruptly stopped and are virtually unheard of when these antidepressants are tapered gradually. Indeed, we are unaware of any reports of lethal withdrawal reactions when SSRIs/SNRIs are rapidly or even suddenly discontinued. Similarly, seizures following abrupt discontinuation of SSRIs/SNRIs appear to be exceedingly rare, unlike, say, with sudden discontinuation of barbiturates. In fact, we could find only 

 of an SNRI withdrawal-related seizure, in the context of severe hypokalemia due to vomiting.

The vast majority of serious withdrawal symptoms following discontinuation of SSRIs/SNRIs occur when the tapering period is less than 1 to 2 months. This may be particularly the case when the patient has taken the medication for a year or longer.

Based on our extensive experience with antidepressants, we believe that serious withdrawal symptoms are extremely rare when 

 However, we acknowledge that such long tapering periods are probably uncommon in general medical practice, and even in most psychiatric settings. Indeed, our second vignette, involving a rapid, 3-week taper of high-dose (60 mg/d) paroxetine, is an illustration of poor medical management that may be common in some practice settings. A much slower tapering rate would have been appropriate in the case of Ms B.

In contrast, the first vignette illustrates how judicious and extended tapering of SSRIs/SNRIs (over 8 weeks and longer) can avoid serious withdrawal syndromes. Factors associated with higher risk of serious reactions are shown in the 

Even more important than a specific timetable for tapering is the need for careful, systematic, individualized assessment of the patient. Similarly, strategies for cross tapering from one antidepressant to another must be individualized, based on the particular switch that is planned. While attending carefully to the patient’s reported response to treatment is always important, it is especially critical during the vulnerable period of antidepressant tapering.

When well-managed, discontinuation of antidepressants need not pose a significant clinical problem and should not discourage depressed patients from using these beneficial medications.

For an online rejoinder by Read and Davies to this article, 

 and your comment might be considered for posting at the end of this article with a response from the authors.

This article was originally published on 1/16/19 and has since been updated.

With a combined 70+ years of clinical experience, the authors discuss scenarios that represent two possible outcomes of discontinuing serotonergic antidepressants.

Author | David N. Osser, MD