Treatment Concepts

Nassir Ghaemi MD

After reading the diagnostic concepts summary, we can turn to the relative pros and cons of using dopamine blockers and “mood stabilizers” for depression and/or manic symptoms. We will also assess harms of antidepressants and amphetamines.  

First, I prefer to use the phrase “dopamine blockers” instead of “antipsychotics” because these agents are not just for psychosis.  Also, I prefer to use the phrase “second messenger modifiers” (which I’ll define below) instead of “mood stabilizers” because the latter term is clinically misleading and scientifically meaningless. Also the term “antidepressant” is false because these drugs do not treat many kinds of depression. I prefer using the phrase “serotonin reuptake inhibitor” instead.

Dopamine blockers

These drugs block dopamine receptors, among other effects. They have been shown to improve depression and mania in many studies. Some have weight gain and can cause or worsen diabetes and cardiovascular risk factors (like high cholesterol).  Others do not have those risks, or have less of them.  All drugs in this class have the most common shared risk of “akathisia”, which means physical restlessness and feeling like you need to jump out of your skin. This effect can be accompanied by suicidal thinking, so it is important to recognize it and stop the medication if it happens immediately. If it is repeated, but these medications are still used or tried, the side effect can often be treated with propranolol, a beta-blocker.

Drugs in this class that have the least side effects (no weight gain or diabetes or cardiovascular risks) are abilify, Geodon, latuda, saphris, and fanapt.  Risperdione has some of those risks, but less than Seroquel or zyprexa which have the most of those side effects.  Again, all agents have akathisia, which occurs in about 20% of people, and is worse at a higher dose.

Second messenger modifiers

These drugs directly affect proteins inside the neurons, called second messengers, that transmit messages or information from the synapse where neurotransmitters act (like dopamine or serotonin).  The messages are transmitted inside the neurons by these proteins to the genes in the nucleus, and those messages change what the neuron produces and/or its connections to other neurons. These complex changes actually change the anatomy and function of the brain, and we think this is why this class of medications seems to have the most long-term effects and benefits for mood illnesses.

Clinically, these drugs PREVENT mood episodes of ANY kind, depressive or manic.  THIS IS THE KEY TO TREATING MOOD ILLNESSES: PREVENTION, NOT JUST TREATMENT OF CURRENT SYMPTOMS.   So, dopamine blockers or “antidepressants” may improve current manic or depressive symptoms, but they may not prevent manic or depressive episodes. In contrast, second messenger modifiers may have less effects for current manic or depressive symptoms, but appear to be very effective in prevent depressive or manic episodes in the future. 

Specific drugs in this class are lithium, divalproex (Depakote), carbamazepine (Tegretol), and lamotrigine (Lamictal).  They each have some risks and benefits.  

Many people are concerned about weight gain. Of these four second messenger modifiers, two do NOT cause any weight gain: lamotrigine and carbamazepine.  

Weighing Benefits and Risks 

Lithium is widely proven for over five decades to have the most proof of benefit for prevention of mood episodes, not just manic, but equally well in prevention of depressive episodes.  Lithium keeps neurons alive, and is the only drug proven to have that kind of neuroprotective benefit not only in animals but in humans with many replicated studies. It appears to reduce the risk of dementia, which is 2-4 fold higher in depression or bipolar illness, back to the population norm.  It is the ONLY drug proven to prevent suicide and to reduce overall mortality in ANY psychiatric illness; it thus extends the life span of the average patient by about ten years.  These are all its major benefits, which need to be kept in mind when assessing the side effects below.  Those side effects happen in about one-third of persons overall, while most people can take lithium with little or no side effects. Some people, a minority, have many side effects and may then just stop lithium, without any long-term harm.

Lamotrigine has much fewer side effects, but is not effective at all short-term, unlike lithium or the other agents, and it is less effective for prevention of manic symptoms or episodes than the other agents. It is quite good in many people in prevention of depressive episodes. If you have an autoimmune illness or if you have rashes to other medications, you may be at increased risk of a serious, potentially fatal, rash reaction to lamotrigine, however, which makes this medication not a first-line choice in those settings.

Divalproex and carbamazepine are both effective for depression and mania, acutely and in prevention, but have more side effects than lamotrigine. 

Divalproex causes the most weight gain of the second messenger modifiers, but again this side effect happens mainly in about 1/3 of persons or less.  Many people can take it without weight gain.  About 10% of women can develop polycystic ovarian syndrome with it, and it is harmful in pregnancy. For these reasons, it is a lower line choice for young women if other agents have not been used.  

Carbamazepine has fewer overall side effects than depakote, but it can cause some cognitive symptoms (memory, attention problems), or nuisance side effects as below. These side effects are lessened with the slow-release formulation, and again they do not occur in more than about 1/3 of persons clinically. It also has some risk of rash, but less than lamotrigine, and less likely to be potentially fatal.  Its main issue is that it has drug interactions, reducing blood levels of other medications, and these drug interactions can thus reduce the benefits with other medications. If you are younger and not on other medications or dont need them, then this can be a useful medication.  Again it has no weight gain. 

With these considerations in mind, summary side effects are as follows:



Nuisance side effects:  sedation, nausea, vomiting, diarrhea, tremor, acne, increased thirst and urination, weight gain, cognitive impairment

Medical risks:  potential for hypothyroidism and chronic renal insufficiency, occurs in a small number of patients, can be prevented by regular labs followed by change in medication

Based on studies that involve decades of prospective follow-up and kidney biopsies, which is available in the publications link, 60% of persons had zero kidney effects from lithium, even after decades of treatment.  About 40% had reversible changes, demonstrated clinically as increased urination (decreased renal concentration capacity), and 5% had long-term kidney impairment.  That kind of long-term effect can be identified with every 6 month blood tests, and if it begins to happen, lithium can be stopped and progression of kidney impairment can be stopped. The main risk factors for this long-term kidney impairment, based on that above studies, is multiple daily dosing and acute renal failure. Thus, the long-term risks can be further reduced by dosing lithium once daily,  and by dosing it lower than the upper end of the therapeutic range (meaning no higher than 1.0), so as to reduce the risk of acute lithium toxicity by dehydration or other causes.  Further, this is a reason to keep lithium doses as low as is effective; frequently "subtherapeutic" (according to a laboratory range) levels are clinically therapeutic, and at those lower doses, lithium will have even less risk of causing long-term kidney impairment. In short, with appropriate usage and medical supervision, lithium has a very low risk of causing long-term kidney impairment.  Another recent study, also in the publications link, identified 1.2% of long-term lithium-treated patients as developing long-term kidney disease.  Those patients had been on lithium for an average of 23 years.

In summary:   95-99% of lithium-treated persons will not suffer from long-term kidney disease, even with decades of treatment.   

Lithium blood levels are not frequent:  Every 6 months or so is sufficient for long-term treatment.  

Valproic acid/Divalproex (Depakote)

Nuisance side effects:  sedation, nausea, vomiting, diarrhea, tremor, hair loss, weight gain, cognitive impairment

Medical risks:  potential for hepatitis which in rare cases in adults can be severe or life-threatening, rare cases of pancreatitis which can be life-threatening if undetected.  Decreased clotting factors can lead to bleeding tendency though rarely associated with serious medical risk.

Carbamazepine ER (Tegretol)


Nuisance side effects:  sedation, nausea, cognitive impairment, double vision, ataxia

Medical risks:  potential for hepatitis which in rare cases in adults can be severe or life-threatening; rare cases of blood abnormalities which can severe and life-threatening in rare cases; potentially serious rash in < 1 in 10000 persons at slow dose titration, mostly in the first 6 months of dose titration, which  can be life-threatening.  Major drug interactions which can decrease the efficacy of other agents including birth control pills.

Lamotrigine (Lamictal)

Nuisance side effects:  occasionally nausea, non-serious rash in 5-20% of persons

Medical risks:  potentially serious rash in 1 in 5000 persons at slow dose titration, mostly in the first 6 months of dose titration, which  can be life-threatening; drug interaction with depakote increased risk of rash (nonserious or serious) and thus dose should be halved and titration even slower with depakote 

Harms and Inefficacy of "Antidepressants"

Antidepressants,  in my opinion and in the consensus judgment of the APA treatment guidelines, are best used as a second-line or later choice for the acute major depressive episode, if necessary. The only studies in which they have been compared to mood stabilizers find that mood stabilizers alone are as effective as antidepressants for the acute major depressive episode in bipolar disorder. Further, repeated studies show that antidepressants are ineffective in prevention of future depression.  Thus, if after using one or two mood stabilizers, an antidepressant is added for an acute major depressive episode in bipolar disordre, it should be tapered off after improvement from the acute depression, usually about 2 months later.  Only if someone repeatedly relapses into depression when antidepressants are stopped, or is severely suicidal, should antidepressants be maintained indefinitely. 

Antidepressants are thus not effective in long-term prevention of bipolar disorder, and only marginally beneficial for short-term acute treatment.  They should not be used long-term in most people; even in the studies which support their use, only about 20% of patients were reported to have benefit at one year. Thus 80% of patients should not receive antidepressants long-term.

Besides this lack of efficacy, antidepressants have two risks:  causing acute mania, and causing long-term rapid cycling or mood destabilization. In terms of acute mania, this risk is minimized with concomitant mood stabilizer or antipyschotic treatment, and it is lower with certain antidepressants, as discussed below. The overall risk is abouot 10-50% depending on which antidepressant is used in bipolar disorder type I. The risk is somewhat lower in type II, and <1% in unipolar depression. Of the available agents, only paroxetine, sertraline,m and bupropion have been shown to have a relatively low risk of causing acute mania, and therefore I prefer to use those agents.  Citalopram also appears to have a low risk.  Effexor has over twice the risk of inducing mania as those other agents, and TCAs are higher risk. I believe that all stronly noradrenergic (including Cymbalta and Strattera) or dopaminergic agents (including amphetamine stimulants) should not be used. Mania induction is also dose related, so the lower the dose the better, and I tend to use about half the dose of antidepressants in bipolar disorder compared with unipolar depression.

A second risk with antidepressants is long-term mood destabilization, with more and more mood episodes (of both depression and mania) over time. About one-third of patients treated with antidepressants develop rapid-cycling bipolar disorder, based on the only randomized study of the topic.  In that study, antidepressant discontinuation was effective in slowing down rapid cycling.  In fact, this is the only proven intervention for rapid cycling.  Lithium is ineffective, as is lamictal (see below), and, contrary to popular belief, anticonvulsants have never been shown to be more effective than lithium for rapid cycling.  Depakote is only slighlty more effective than lithium for rapid cycling according to the only randomized comparison.  Rapid cycling is a severe form of bipolar disorder, therefore, in which no single agent is effective, and the only proven intervention is to stop antidepressants. Beyond that, we usually combine multiple mood stabilizers in rapid cycling.

In sum, antidepressants  can act as mood destabilizers, counteracting the benefits of mood stabilizers.  Thus, no patient has received a therapeutic trial of a mood stabilizer, unless it occurs in the absence of antidepressants.  Frequently, past trials of mood stabilizers “failed” with concomitant antidepressant use, but later the same mood stabilizers will have efficacy in the absence of antidepressant use.

Harms of Amphetamines

My views on amphetamines are supportable by research, though not mainstream among many psychiatrists, so I am providing some references along with the statements I make below and I've provided a full paper on the topic spelled out in more detail with all available citations (1). (See publications link to download freely)

Amphetamines, including all versions of methylphenidate, are antidepressants, and thus share all of the risks described above.  In the largest study of this topic, conducted by us, on amphetamines in adults with bipolar disorder, we found a 40% manic switch rate, even with mood stabilizers.  These are highly mood destabilizing agents.   I also will comment that their use for purported adult ADHD is often, in my view, incorrect, because the manic symptom of distractibility or the depressive symptom of poor concentration is then being mistakenly converted into its own diagnosis.  Indeed in the largest epidemiological study of adult ADHD, the National Comorbidity study, about 80% of those persons who meet adult ADHD criteria also meet bipolar or MDD criteria, which either means that such persons are highly unlucky, and every time they have one disease they get two, or it means that the adult ADHD definition really overlaps almost completely with mood disorder definitions.  Since mood disorders are accepted as primary and higher on the hierarchy than ADHD (even adult ADHD proponents accept this view), one should not diagnosed adult ADHD in the context of active mood symptoms, nor even if diagnosed when patients are euthymic, should it be treated extensively, especially with amphetamines, in my view, given these mood destabilization data.    

Amphetamines improve attention short-term, in everyone, not just patients with ADHD, which is why they are addictive drugs.  Thus, symptomatic improvement with them is not at all diagnostic.  There are risks that need to be weighed against this short term benefit.  (All the following discussion about amphetamines includes methylphenidate, and all its derivatives).

It is important to also note that amphetamine stimulants are far from benign.  A number of animal studies have shown that amphetamine stimulants are harmful to the developing brain in rats:  they lead to decreased hippocampal size (2) decreased dopaminergic activity (3) and increased corticosteroid response to stress (4).  All these effects, it should be noted, are the opposite of the neurobiological effects of antidepressants and lithium, which appear to be neuroprotective, leading to larger hippocampal size and decreased corticosteroid activity (5).  In other words, amphetamine stimulants, neurobiologically, appear harmful to the brain, like drugs of abuse (6), and unlike most prescribed psychotropic medications. Further amphetamine use in adolescent animals has been associated with higher amounts of depressive and anxiety behavior in adulthood (4, 7),  perhaps due to this hippocampal atrophy (8, 9).

Obviously these animal findings may not translate into humans, but we cannot assume that they will not, and these results, so at odds with most medications that are beneficial for mental illness, should give us pause, at least in terms of long-term treatment for years.

Further, amphetamines cause mania, or worsen mania, and like antidepressants, likely cause mood destabilization in persons with bipolar disorder.  These risks would also be present in this case since he has a neurological susceptibility to manic-like symptoms.  Studies in children indicate that the frequency of amphetamine induced mania ranges from 10-50%, less when given with a mood stabilizer, and especially if given after mood symptoms are initially stabilized with a mood stabilizer in the absence of antidepressants.  However, in general, when amphetamines are used chronically, mood stabilizers appear to be ineffective, in my view, only because they are given with mood destabilizers – the amphetamines.  It is not a therapeutic trial, in my view, of the mood stabilizer unless the amphetamines are stopped first, and then, in my experience, the attentional symptoms usually improve along with the mood symptoms.  Sometimes they may not; and it may then take years of avoiding amphetamines, and using lithium to help with long-term neuronal regrowth, to gradually improve cognitive function. 

Also, the longest term randomized study of children with amphetamines, the MTA study, found that after one year, functional outcomes were equal with intensive behavior therapy versus amphetamines.  The medications are not more effective than psychosocial interventions for functional outcomes (meaning school, academic, and interpersonal function).  The question is what kind of psychosocial interventions to institute. These can range from individual psychotherapies including behavior modification and social skills training aimed at helping with attentional impairment and impulsivity, and programs for parents and teachers aimed at increasing the kind of structure that helps such children. One such program is called The Incredible Years,, and it found that 2/3 of children with ADHD had completely improved within one year of follow up in a randomized trial. 

A good summary of available psychosocial treatment for ADHD-like symptoms can be found here:

1.          Vergne DE WE, Barroilhet S, Fradkin Y, Ghaemi SN: Adult ADHD and amphetamines: a new paradigm. Neuropsychiatry 2011; 1(6):591-598

 2.           Lagace DC, Yee JK, Bolanos CA, Eisch AJ: Juvenile administration of methylphenidate attenuates adult hippocampal neurogenesis. Biol Psychiatry 2006; 60(10):1121-30

3.             Brandon CL, Marinelli M, White FJ: Adolescent exposure to methylphenidate alters the activity of rat midbrain dopamine neurons. Biol Psychiatry 2003; 54(12):1338-44

4.             Bolanos CA, Barrot M, Berton O, Wallace-Black D, Nestler EJ: Methylphenidate treatment during pre- and periadolescence alters behavioral responses to emotional stimuli at adulthood. Biol Psychiatry 2003; 54(12):1317-29

5.             Duman RS, Malberg J, Nakagawa S: Regulation of adult neurogenesis by psychotropic drugs and stress. J Pharmacol Exp Ther 2001; 299(2):401-7

6.             Brandon CL, Steiner H: Repeated methylphenidate treatment in adolescent rats alters gene regulation in the striatum. Eur J Neurosci 2003; 18(6):1584-92

7.             Carlezon WA, Jr., Mague SD, Andersen SL: Enduring behavioral effects of early exposure to methylphenidate in rats. Biol Psychiatry 2003; 54(12):1330-7

8.          Lagace DC, Noonan MA, Eisch AJ: Hippocampal neurogenesis: a matter of survival. Am J Psychiatry 2007; 164(2):205

9.          Duman RS: Depression: a case of neuronal life and death? Biol Psychiatry 2004; 56(3):140-5