Interview with Husseini Manji, M.D.

Husseini, Manji M.D. is Chief of the Laboratory of Molecular Pathophysiology at the National Institute of Mental Health (NIMH). He was born in KenyaPicture of Dr. Manji and moved to Vancouver as a teenager. In 2002, Dr. Manji joined the Professional Advisory Council of The Balanced Mind Parent Network, and often answers science questions from our members. Dr. Manji edits several journals and has received numerous prestigious awards, summarized here.

The Laboratory of Molecular Pathophysiology is part of the Mood and Anxiety Disorders Program (MAP) at NIMH, the world's largest research program focused on mood and anxiety disorders.

The Balanced Mind Parent Network interviewed Dr. Manji in December 2002.

Note: Medical terms are defined within the text.

The Balanced Mind Parent Network: Until recently, it was believed that bipolar disorder was a "mood" issue only, and that patients whose emotional swings were stabilized fared quite well. How has your research called this assumption into question?

MANJI: What our research and other peoples' research has recently shown is that mood disorders, while they do entail mood swings or emotional changes, are also associated with changes in important structural areas of the brain. In other fields of neuroscience, we are finding out that what the brain does for a living is adapt to its environment: When an animal learns something, the brain's structure changes. Some of the biochemical pathways that help brains function moment-to-moment are also important for keeping cells alive, growing properly, fortifying them against stress, and making long-term structural changes. Brain imaging studies and studies on the brains of deceased patients (called postmortem studies) are showing us that in many patients with recurrent mood disorders (whether bipolar or unipolar), parts of the brain seem to be atrophied, or shrunken. It is a bad news/good news scenario: The bad news is that we don't want to see structural changes. The good news is that most of the change seems to be cell atrophy, or shrinkage, rather than cell death. At least in some cases you can restore them back to health, which is different than with Alzheimer's or Huntington's, where you can't.

If you do an MRI [brain scan] of areas of the brain such as the frontal cortex, hippocampus, or caudate, you see that these areas are smaller in patients with recurrent mood disorders. If you do postmortem studies and count the number of cells and the number of cellular connections, you find that the cells of these patients are more shrunken. In some cases, they are fewer in number. Both neurons [cells in the nervous system, including brain cells] and glia [specialized nervous system cells that support, insulate, nourish, and connect neurons, as well as regulating the surrounding fluid] seem to be affected in mood disorders. Glia may show a greater degree of loss, while the neurons are more atrophied. But this is a sweeping generalization, because this is currently under study.

Picture of atrophied brain cells





Atrophied brain cells, as seen in the brain of a depressed patient
Reproduced with permission of Husseini Manji, MD

Picture of healthy brain cells





Healthy brain cells, with cellular resilience and connectivity restored after treatment
Reproduced with permission of Husseini Manji, MD

The Balanced Mind Parent Network: Do the changes you see have an identifiable disease pattern, or are they unique to each individual?

MANJI: It is too early to say. If you look at the MRI findings in bipolar disorder and compare them to those in schizophrenia, in many cases they would look similar. In the past, researchers have disregarded those similarities and concluded just that the illness is bad for the brain. Now they are asking, what is the meaning of atrophy in brains of deceased bipolar and schizophrenic patients? When researchers count specific cells and connections, they find two things going on. One, there is some overlap between depression, bipolar disorder, and schizophrenia. Some of the changes may have to do with activation of overall stress pathways that can damage the brain. But other changes, in some layers of the brain, seem more specific for each illness. I think that a pattern specific for bipolar disorder will emerge. There will be some overlap, in that some chemical pathways are overactive, and it doesn't matter what disease you have, they cause the same damage, while other types of changes are disease-specific.

The Balanced Mind Parent Network: Do we know if the structural differences found in bipolar brains existed prior to symptom onset, or if they were the direct result of episodes of illness? And is there any evidence that either mania or depression is more toxic to the brain in this regard?

MANJI: We don't have clear-cut answers yet. In my opinion (because the definitive studies are still lacking), bipolar depression may be different from unipolar depression. There are 3 studies suggesting that even at first onset, patients with bipolar disorder may show some atrophy (but it's quite possible that they've had "mini-episodes" before they have the full-blown episode which results in the diagnosis being made). But there are also data from unipolar depression that the hippocampal size may be negatively related to how long the person has been depressed over the course of a lifetime. In my opinion, some bipolar individuals may start out with an impairment of cellular resilience [ability of cells to withstand stress], so that even normal day-to-day neuronal demands may be excessive and cause dendritic atrophy [shrinkage of dendrites, threadlike extensions of neurons that provide most of their receptive surfaces], etc. However, it is also very likely that each episode takes an additional toll.

I don't think we currently have reason to believe that manic episodes would be less harmful than depressive episodes. We will have more definitive answers when more long-term studies are conducted.

The Balanced Mind Parent Network: Is there evidence that unipolar depression and bipolar depression are actually caused by different disease processes?

MANJI: Unipolar depression is a much broader category than bipolar depression. In some ways, unipolar depression is somewhat less specific, somewhat like fever. With fever, you can get there many ways-from viral infection, bacterial infection, a reaction to a drug, etc. Depression may be similar. In my opinion, a subgroup of what we see as unipolar depression probably belongs in the bipolar family. Most of the data suggest that in bipolar disorder, multiple genes are susceptible, and if you get enough of them, you have bipolar illness. It is not like Huntington's disease, which arises from one gene. The older European studies are suggesting that in recurrent unipolar depression and bipolar disorder that run in families, treatment response is similar (i.e., they respond to lithium treatment), and this suggests that you have in recurrent unipolar depression many of the same genes as in bipolar disorder. More and more data are confirming this, including brain imaging studies, such as those done by Wayne Drevets.

There are certainly differences between bipolar disorder and unipolar depression. Bipolar overall is a worse illness than major depression. It is more recurrent and it has an earlier age of onset and a greater likelihood of suicide. More substance abuse and disability is associated with it, but a subgroup of unipolar depressives belong in this family.

The Balanced Mind Parent Network: Until recently, it was believed that bipolar disorder results from an imbalance in brain chemicals called neurotransmitters. These neurotransmitters attach to receptors on the cell membrane [outside wall of the cell], and there their role more or less ends. Your research and others' have called this assumption into question.

MANJI: We are in the midst of a molecular medicine revolution. Some people think we have reached a more fundamental understanding of the human body than we did in the antibiotic era. Probably the biggest advance has been in understanding how much goes on inside the cell. Whether in the pancreas or blood or brain, every cell talks to other cells through chemical messengers (called neurotransmitters in the brain). For many years, we have focused on the possibility that it is the chemical sending the message between cells that is overabundant or underavailable. Recently, we have discovered that the chemical sending the message between cells is just the starting point. Once the chemical binds to the cell membrane, there is a dynamic reaction: Hundreds and thousands of reactions go on inside the cell, and they seem to be much more important to the cell's functioning than the binding of the neurotransmitter to the outside.

Understand that one nerve cell receives connections from something like 10,000 other nerve cells. It has to make sense of all that input. Inside the cell, in the signaling pathways, is where a lot of the sense is made. All these various inputs are integrated into one coherent message. If the signaling pathways involved in making sense are haywire, you are in trouble. If one neurotransmitter is off, you can deal with it, you have dozens of others. In Parkinson's disease, for example, patients can lose 50% of their dopamine neurons and still cope. But if what all the signals converge into has problems, you will have lots of trouble. It also looks like signals in the cell are very important in terms of amplifying the signal sent by the neurotransmitter. One molecule of adrenaline [or epinephrine, a neurotransmitter that speeds physical reactions as part of the "fight or flight" response] on the outside of the cell will cause the cell to make 100,000 molecules of a second messenger inside the cell. [Second messengers are molecules produced or released inside the cell wall in response to the binding of a neurotransmitter to a receptor on the outside of the cell wall.] Even small changes, small problems inside the cell are almost certain to cause disease. You just cannot adapt to those changes inside the cell.

In almost every field of medicine-diabetes, cancer, endocrine research, etc. -- most of the research is moving inside the cell. What the chemical does inside the cell is more important than what it does in the synapse [the space between two neurons]. That shift has been going on for now about a decade. Of all psychiatric diseases, bipolar disorder is the one in which there is the most evidence that the abnormalities are primarily inside the cells. If you measure serotonin, norepinephrine, or dopamine [different neurotransmitters], you do find abnormalities, but more of the abnormality is likely inside the cell -- after the neurotransmitter binds. I can't tell you precisely which of the molecules are the problem -- we have clues but we don't know yet. It is how the cell reacts to the neurotransmitter which is more important than the neurotransmitter itself. And that is where the mood stabilizer seems to work. We think that is one reason why mood stabilizers are effective in bipolar disorder. When you treat with antidepressants, you treat far away from where the problem is, because antidepressants affect neurotransmitter levels and action out in the synapse. If the machinery within the cell is all working fine in a patient, then the antidepressant will have the expected effect. The antidepressant starts to have its effect way outside the cell and lots of things have to change en route before you can turn on the gene in the cell nucleus to achieve the desired effect. If the inner cell signaling machinery is not working fine (and we believe that in bipolar patients, that is where the fundamental problem is), then it is possible you can mess with a neurotransmitter all you want, but it won't be able to convey its information to the cell nucleus and turn on and off the gene that will make the difference.

In some cases, correcting that intracellular signaling abnormality is enough to restore mood stability. In others, you restore the function of the signaling pathway in the cell back to normal, but find you still need some "oomph" to turn on the pathway after it is restored. An antidepressant can do that, so you see benefits -- but only after you have given the mood stabilizer a chance to fix the internal cell signaling problem first; otherwise you run the serious risk of overshooting and triggering manic episodes.

We are finding out that the same signaling pathways that are the targets of mood stabilizers and seem to be involved in bipolar disorder are the ones that are involved in keeping the neurons healthy and surviving and resilient against stressors and insults. We are starting to think it is a continuum -- the signaling pathways that are impaired when someone has a bipolar illness are the same pathways that are involved in keeping the neuron and the glia healthy and surviving. So in bipolar disorder, the overall health of the brain cell is compromised, and that is where you see the cell loss and atrophy. By targeting these pathways for treatment, not only might you be able to target the symptoms of the disease, but you also might help the cell atrophy and death.

The Balanced Mind Parent Network: When you speak of a flaw in cell signaling in bipolar patients, does this refer to all cells in the body or only to brain neurons (or some other specific set of cells)?

MANJI: Once again, there are not a lot of data. We think that the manifestations of the abnormal signaling appear mainly in specific brain cell circuits. However, this could arise in two ways: (1) the signaling protein abnormality is present only in specific neurons (i.e., some proteins are produced only in certain neurons, so if they are abnormal, it's purely a neuronal phenomenon); or (2) the signaling protein abnormality is more widespread, but only becomes a problem in certain areas of the body (brain).

There are abnormalities in some signaling pathways in peripheral cells [cells in the peripheral nervous system, outside the brain] in patients with bipolar disorder. Are these the same as what's going on in the brain? Unclear. And the peripheral cell studies need to be interpreted very cautiously because lots of things, even very minor things such as whether the patient is standing or sitting, can affect the results. There are very few studies in medication-free euthymic [mood-stable] bipolar patients. In any case, it's quite possible that some of the abnormalities (e.g. calcium elevations) are also present in cells like platelets, and may play a role in the enhanced cardiovascular mortality seen in mood disorders.

The Balanced Mind Parent Network: How far off are we from having a diagnostic test for bipolar disorder? And until then, how can parents avoid the wrong diagnosis?

MANJI: We are still pretty far off. Unfortunately, bipolar disorder is going to turn out to be due to different genes, and different bipolar families may have different genes. I think we are 5 to10 years away from finding a bipolar signature, whether it is one from brain imaging or genes. And even that would not be diagnostic across the board, but only suggestive. More than one test will likely be involved. But the genetic signature could be very predictive, such that if you have it, there is a 90% chance you will respond to this drug or get this weight gain effect. Almost certainly we are headed there. Science can be very difficult to predict because tomorrow someone doing cancer research could come up with a finding that has amazing implications for those of us doing bipolar research. There is no question we will get there eventually.

From a treatment standpoint, it is very important to correctly determine if depression is unipolar or bipolar. There are data that antidepressants might trigger highs and rapid cycling etc. So from a diagnostic standpoint, recurrent unipolar disorder is to be considered the same as bipolar disorder because it would guide you to a safer treatment. You try mood stabilizers rather than jumping to an antidepressant.

I am deviating a little, but the age of onset of bipolar disorder in European countries may be higher than it is in North America. One possible reason is that maybe in North America we are using antidepressants earlier than in Europe. Antidepressants don't make anyone bipolar. But if you are one of these people who have these risk factors or susceptibility genes, and your illness is showing up mainly as depression, an antidepressant could trigger the high and induce your first manic episode earlier. Odds are you were going to have it anyway. It is possible that antidepressants used in someone who belongs in the bipolar family is one of the things that could trigger disease onset earlier. Substance abuse is another. We see the first episode in association with substance abuse in adolescence, which was the stressor that unmasked the bipolar disorder.

The Balanced Mind Parent Network: If you have a child presenting with just depressive symptoms, how do you know whether he or she is going to develop bipolar disorder?

MANJI: Until we learn more about how to distinguish the two, family history would be one way. This is a young person who is depressed now but possibly has some bipolar gene(s). On average, people who present young with the illness have a greater likelihood of being bipolar. Another loose criterion is sleeping excessively, although it is not proven and hard to prove. Also, psychomotor retardation looks more like a feature of bipolar depression than of unipolar depression. The last two criteria are very vague. Clinically, however, they are something I pay attention to. This is the safer thing to do.

There is almost no question that when bipolar is suspected, it is safer to treat with a mood stabilizer first and only after to try adding an antidepressant. The downsides of such an approach are: (1) On average, mood stabilizers aren't as effective as just an antidepressant. Antidepressants have a greater wallop. A mood stabilizer may take longer or not be quite as robust in terms of treating the depression. If someone is catatonically depressed, you don't want to have them waiting for weeks for it to kick in, although it is almost the better thing to do. (2) If you have got someone on a mood stabilizer and they haven't responded and you add an antidepressant and they get better, then you are faced with the dilemma, did you need the mood stabilizer? (3) The last thing is that no one likes to take medication. If you expose someone to a medication and they haven't gotten better and have only experienced side effects, they might be biased against taking more medication in the future, even though trying the mood stabilizer first is the correct thing to do.

So clinically, there is not a straightforward answer. But from an academic standpoint, always start with mood stabilizer to treat a child with depression and a family history of bipolar disorder.

The Balanced Mind Parent Network: If a patient's mania is well-controlled on one mood stabilizer, but the patient still has trouble with depression, is it prudent to try an antidepressant?

MANJI: This is also a controversial area. There are two schools of thought. The first says, add a second mood stabilizer to see if that does the trick. The second says, add an antidepressant. The advantage of adding a second mood stabilizer is that it is much less likely to cause rapid cycling or manic episodes. The downside is that the mood stabilizers probably aren't as effective as pure antidepressants at treating depression. Lamotrigine is one of the exceptions.

On the other hand, we may have been guilty of leaving people depressed longer than we should because we have been worried about triggering mania. I don't think it's appropriate, if a person is depressed on a mood stabilizer, to overly handcuff yourself. Low-grade depression can have major impacts on a person's well being, school performance etc. There are currently studies in adults trying two mood stabilizers versus a mood stabilizer + antidepressant to see if there is a big difference. The Europeans tend to use antidepressants more than we do, at least in adults. They feel the risk of cycling and mania has been overblown here. But they do also a very good job with lithium. There was one recent publication from the Emory group where they tried lithium + an SSRI vs. lithium alone, and it suggested that if a patient actually has adequate lithium levels, adding the SSRI doesn't do much. The problem was that lithium levels were too low. It may be that the Europeans are doing a better job than we do with lithium. Here we may tend not to use the antidepressant mood stabilizers (lithium, lamotrigine) as much as we should. We definitely don't use lithium enough and it's possible that when we use it we are not using it well enough -- gradually going up to adequate levels. We are not using it at correct levels.

Now I will contradict myself and say that overall, I think the Europeans have had a tendency, once the mania is under control, to maintain patients on lower levels of lithium than we do. It comes back to the same basic issue: This illness is most susceptible to abrupt changes. So if you keep people stable on one thing, at adequate levels, you can stabilize them. If you have to be constantly switching, you may need more of a sledgehammer. The managed care system here has made it less likely that the gradual increases on one medication to appropriate levels will occur. There is pressure for instant results -- for example, to get people out of hospitals quickly -- so higher doses are used, triggering more extreme swings. This might be why the Europeans overall have had better luck than we have.

It is not inappropriate to use an antidepressant in a depressed bipolar individual whose mania has been well controlled with a mood stabilizer. However, my first choice would be to add another mood stabilizer that is known to have antidepressant effects. Today's data shows that lithium and lamotrigine have the most antidepressant effects.

If you do use an antidepressant, you almost certainly want to aim for a short-term use. Exactly how temporary, we don't know. Make sure the patient isn't getting high. If they are, that warrants earlier discontinuation. In general, once depression lifts, allow perhaps a few weeks and thereafter consider removing the antidepressant. In general, my recommendation would be even if it is a minimal dose, you should still consider tapering it gradually, because it seems like the bipolar system does not react well to abrupt changes. For patients with unipolar depression, we keep them on the antidepressant 6 months to a year after they have responded to it. With bipolar patients, it seems the brain is constantly trying to adapt to what you are doing, and the longer you are on the antidepressant, the longer it becomes likely that the brain is trying to counteradapt to it, so you might not want it there longer than absolutely needed.

Also, it makes a lot of sense to start with a lower than usual dose and increase it very gradually. We don't have data, but in general, avoiding large perturbations of the system is almost always a good idea in bipolar disorder. Same thing with discontinuation (unless there is a true medical reason to discontinue rapidly).

The Balanced Mind Parent Network: What about stimulants? What effect would stimulants, given unopposed to a bipolar child, cause?

MANJI: Everything I said about antidepressants would apply to stimulants and maybe even more so. Enough stimulants seem to be capable of triggering manic-like episodes in anyone. In the lab, most of our animal models of mania are based on using stimulants. That is to say, we use repeated amphetamine administration to make the animal become sensitized so it shows high degrees of motor activity and hedonic [pleasure-seeking] behavior.

Interestingly, you can prevent this manic response to stimulants by pre-treating the animal with lithium. This is how we model human mania for our animal experiments. So if we have a new biochemical pathway that may work, one of the models we use is to treat animals with stimulants to make them hyperactive and then use this drug.

Bipolar people, and perhaps even the general population, respond differently to stimulants. Some people seem to show a euphoric, giddy, and activated response; others feel lousy, an unpleasant revved up. The problem is attention-deficit/hyperactivity disorder (ADHD), where someone who is behaviorally hyperactive is actually calmed down with a stimulant. Clearly, stimulants seem to have a different effect depending on the starting point, whether you are in the ADHD or bipolar category.

From a biochemical standpoint, stimulants seem to be capable of causing dopamine and norephinephrine release in the brain, and those are two things we think might be involved in causing manic episodes. They may also be capable of activating protein kinase C (PKC) [an enzyme in the brain that plays a major role in signaling inside cells], which both lithium and depakote attenuate [weaken, or reduce the activity of]. Anti-manic medications seem to dampen PKC, so it is reasonable to think that stimulants are capable of changing biochemisty in a pro-manic direction. It's quite likely that enough stimulants seem to be capable of making anyone exhibit manic-like behaviors. But someone who is bipolar may be more likely to enter a full-blown manic episode much more rapidly. Our brains have many feedback mechanisms to prevent them from overshooting in any direction. Most people who take a stimulant have something in their brain that prevents it from overshooting. Bipolar patients have a problem regulating this. So things like stimulants or sleep deprivation or depressogenic [depression-causing] factors may have the same effect as in some individuals who are not bipolar, but the brain of the bipolar patient can't self-regulate the effect. That is the problem with stimulants in bipolar illness. It is quite possible that someone carrying the bipolar genes can't keep the stimulant effect in check, thereby making it possible for stimulants to trigger real highs.

As I see it, any time you do something to the brain, the brain tries to counter it and is always trying to restore balance. In bipolar disorder, even the countering is sometimes excessive. After a manic episode, many patients crash and become depressed. There is evidence to believe that the mechanisms their brain use to try and bring them down overshoot and drive them to depression. Eventually, they go back and forth and have these cycles. So one of the things that may be going on with stimulant use in someone who is bipolar is that even if you don't trigger actual visible highs (because their own brains are wired so they can counter the effect), you might be starting the brain's biochemical process to try to counter the effect. By giving the stimulant, you have therefore turned on these biochemical pathways that are driving them to depression.

And so even if you don't see the bipolar patient on a stimulant become manic, you might be hurting them quite a bit with starting these cycles. This might explain some of the difference between Europe and the US-early stimulant use in the US.

That is also why it is very important to treat hypomanic episodes, although not everyone who has them wants treatment. When you are in a hypomanic episode, your own brain may be trying to counter it, and that will cause you to have more depressive episodes, trying to drive the hypomania down. Adequate treatment of hypomania may be very important in the long run.

The Balanced Mind Parent Network: Would a child who has been previously exposed to stimulants become less likely to respond to mood stabilizers later?

MANJI: I think it is possible, but there is just not enough data to know. There is a process called sensitization, which refers to the fact that in animals (and probably humans), you have to give it a certain amount of amphetamine to cause it to become hyperactive. You do this for 2 to 3 weeks repeatedly, making it become hyperactive for a few hours, and find you have made the animal permanently more sensitive to the amphetamine, so that afterwards, even a low dose of amphetamine makes it just as hyperactive. At the beginning, the low dose would not have made it become hyperactive, and now it does. The kindling model in epilepsy is like this -- you stimulate an animal's brain with a small amount of electrical current that doesn't cause a seizure. If you do that over a week or two, what happens is that even that small amount causes seizures. A lot of data suggest that this effect is now permanent. Either with amphetamine stimulation or electric kindling, there is the possibility of producing long-term changes in the brain which have now made it more sensitive.

It's quite possible that with inappropriate stimulant use, maybe you are sensitizing the brain. Does that mean it is less likely to respond to mood stabilizers later? It is quite possible, but we don't have the data to prove it. In animals, you might need higher doses of mood stabilizer later after stimulant exposure to stabilize the animals than you would have before stimulant exposure. In humans, there is only a certain range of lithium a person can tolerate. You can't use double the dose because of side effects. So this remains a theoretical consideration that is a possibility, that early stimulant exposure might make a person nonresponsive to lithium later within the acceptable dose range. The higher dose they would need to stabilize mood would be out of the limits of acceptable toxicity.

The Balanced Mind Parent Network: Given this information about the effects of stimulants, what is the safest course of action to take when a child with a family history of bipolar disorder presents with just ADHD symptoms?

MANJI: I think that there is no question that the safest course in such a situation would be a trial of a mood stabilizer. It's quite possible that some of the "ADHD symptoms" (e.g. hyperactivity, restlessness, distractibility, etc.) are really early manifestations of bipolar disorder. Injudicious use of stimulants runs the risk of triggering a manic episode which can have devastating consequences. It's important to keep in mind that although some individuals show a rapid response to mood stabilizers like lithium, they often need to be taken for several weeks before one sees benefit. Thus, if a mood stabilizer is tried, it's very important not to discontinue it prematurely.

The Balanced Mind Parent Network: Speaking of lithium, your research has uncovered some of its intriguing beneficial properties. Can you highlight the most important ones?

MANJI: Many of the genes that are considered neuroprotective are being remarkably turned on by lithium. Is lithium actually neuroprotective? We hadn't thought this way before. A number of studies have taken animal cells and tried to kill them by causing stroke, etc. These studies have consistently shown that lithium, if administered before you try to do the bad things (such as induce a stroke), protects the animal's neurons. In lithium-treated brains, the size of the resulting stroke is smaller, the number of neurons that die is lower, etc. That was amazing. Since these studies were done in rats, you need to be careful about jumping to conclusions that lithium is neuroprotective in people. Wayne Drevets' group published a finding in Nature about five years ago that in a part of the pre-frontal cortex of bipolar patients or patients with familial recurring unipolar depression, there was almost a 40% reduction in the amount of gray matter. That was a remarkable finding that you have such a reduction in a discrete part of brain. We spoke to him about our lithium findings and asked him to reanalyze the data. He had a small group of patients who had been treated with lithium for a long time and they did not show the brain atrophy compared with the bipolar patients. Interestingly all of the patients with unipolar depression, whether or not they had been treated with antidepressants, still showed the atrophy. That was a suggestion that bipolar treatments might have a protective effect. Valproate (Depakote) in the prefrontal cortex seemed to have the same type of neuroprotective properties. Lithium and depakote do not have identical effects in every brain area, but in this area they did. Brains treated with chronic lithium or valproate seemed not to have the atrophy in the prefrontal cortex. But it was a very small sample and a crossectional study [type of study whose design restricts its findings to association between variables, not proof of cause]. He studied them once. We don't know if it was a cause or effect. Is it the people who don't have the atrophy who responded to the drug in the first place? We did some studies taking bipolar patients off their meds -- they were referred to us because their treatments weren't working. In every case, they either hadn't been on lithium or had been on lithium sparingly -- had started on it, had side effects, switched, and the new med was not working. These are bipolar depressed patients. We did MRI scans and MRS spectroscopy and then put them on lithium in a blinded fashion for 4-6 weeks. Then we did the scans again. We found that almost every single person taking lithium had an increase in N-acetylaspartic acid (NAA) [an amino acid that is viewed as a marker of neuronal health]. And the actual amount of gray matter was going up when they were treated chronically with lithium. This study was done together with Dr. Greg Moore¹. This was happening in areas of the patients' brains that had been atrophied. The increase was not due to swelling from water retention. The increase was seen only in areas where the brain matter had previously atrophied.

Our working hypothesis, and I think it is reasonable, is that lithium is turning on some of these growth signaling pathways and reversing the damage. It seems that the cells are shrunken, not dead, and are capable of going back to their normal sizes and sending normal projections. What lithium seems to be doing is turning on the signaling pathways that produce growth factors in the brain, such as brain-derived neurotrophic factor (BDNF), and where you have atrophy, turning on this pathway seems to be capable of reversing it. And that is a remarkable finding.

A couple of other studies since then have compared lithium-treated patients with untreated patients. They showed that the lithium-treated patients have the increase in gray matter, suggesting that lithium is causing the increase. If that is the case, lithium has a neurotrophic [nourishing to neurons] effect. Would lithium, then, be beneficial in any bipolar patient, even if he or she can't tolerate it or if his or her symptoms respond better to another mood stabilizer? One question is tolerability. We've done a number of animal studies with low-dose lithium. We found that in animals, with a dose of lithium that is one-half or one-third of a usual dose, you still get a large increase in bcl-2, a neuroprotective protein. This suggests that it is quite possible that even low-dose lithium will exert these effects. Many studies are being planned using low-dose lithium as an additional agent in patients being treated with something else. We will add low-dose lithium and follow them for 2-3 years with repeated MRIs and neuropsych testing to see if the addition would be enough to provide neuroprotective effects that would help them long-term with the illness, even if something else is their appropriate treatment for symptoms. That is extremely important. It is a devastating long-term illness, and brain atrophy may be responsible for that. If we can intervene early on and prevent that, there is reason to believe you will have a big impact on the overall course of the illness.

The Balanced Mind Parent Network: Have any of the other mood stabilizers been studied to determine whether they may have similar neuroprotective effects?

MANJI: Most of the others haven't been extensively studied. Almost anything that reduces neuronal excitability (i.e., anticonvulsants) reduces the neuron's energy requirements and thus is likely to have protective effects. However, none of them have been as extensively studied as lithium, and to a lesser extent, valproate. Thus far, all the human data are from lithium. We have looked at the effects of certain other medications on neurotrophic cascades, and we find that carbamazepine [Tegretol] and haloperidol [Haldol] have no effects. We have done a few studies with clozapine [Clozaril], and find that it does increase the levels of bcl-2. Whether other atypical agents like olanzapine [Zyprexa] (which is structurally quite similar to[clozapine]) and risperidone [Risperdal] have similar effects hasn't been studied yet.

The Balanced Mind Parent Network: Since lithium is known to deplete calcium in bones, should bipolar patients on lithium should take calcium supplements?

MANJI: Lithium depletion of calcium in bones is not a very common occurrence. Medication-wise, calcium supplementation is not a big issue.

A small percentage of bipolar patients develop hyperparathyroidism on lithium. In these cases, a more careful investigation is warranted. The hyperparathyroidism should first be corrected; calcium supplementation may then be necessary.

Mood disorders themselves can cause lowering of calcium in bones in older patients, which puts them at risk for osteoporosis because of elevations in cortisol and the activation of the sympathetic nervous system. (Phil Gold at the NIMH has done a lot of good research in this area.) Today we think once you lose that bone, it is cumulative over your lifetime. So almost certainly, unless there is a medical reason not to be taking calcium, you are not doing yourself harm by ensuring adequate calcium levels by diet and exercise. Diet and exercise are the best means, and for most people, they are adequate. If not, then calcium supplements can be taken.

The Balanced Mind Parent Network: How did you get interested in this line of work?

MANJI: During medical school, I was always interested in the functioning of the mind and the brain and specifically in the molecular and cellular level, but I wasn't aware that psychiatry was heading in that direction. During an oral exam, some of my examiners told me that psychiatry was really heading towards trying to understand things at a molecular and cellular level. They captured my interest, and that was what got me hooked into choosing psychiatry. Once I got to the NIH, after I did 3 years of residency training, it became apparent to me that that was the direction where all of medicine is going. That would lead to the biggest breakthroughs and that is where I wanted to go.

I have always been fascinated by bipolar disorder, recognizing how devastating an illness it is. There are times when I think bipolar disorder gets overly romanticized, with an emphasis on the creative aspect. In terms of destigmatizing the disease, that is great, but one has to be careful about not taking away from the fact that these illnesses take a huge toll. I wanted to work on them.

From a scientific standpoint, it fascinated me that this one illness could present with a clinical picture that was so different in the same person. Few illnesses present so dramatically differently. That was one of the things that made it clear to me that the core problem can't be just too much of a single neurotransmitter, it must be something much more fundamental inside the cell which is involved in regulating multiple neurotransmitters, and that's why you see these dramatic differences in presentations. That is why I got hooked on it, wanting to make a difference. And lithium -- I was fascinated by how something that looks so simple -- just like table salt -- could have such a dramatic effect. It seems to me if you could understand how this molecule works, you would be in a much better position to understand the illness. And bipolar disorder is almost certainly an illness that will turn out to be caused by a problem inside the cell.

I have not had one moment of regret about my choice of career. I can't emphasize enough that it is the right time to be working on this disease now. Many of the things that we are doing are things that others before us have wanted to do, but the technological advances have not been there. We are studying bipolar disorder at the right time.

We are finding that almost every field of medicine is looking at the same pathways inside the cell -- whether to learn about the pancreas (to understand diabetes), the breast (to understand breast cancer), or the brain. We quickly trade information, so we can make use of the advances others make. The recognition that these are all illnesses that originate with problems at the cellular level means that advances from one field of medicine can really lead to advances and breakthroughs in ours as well. We also help with the destigmatization. We open eyes at the NIH when we talk to others. They are fascinated to learn that we are studying the same cell signaling pathways as they are.

The Balanced Mind Parent Network: Can you tell us something about the NIH?

MANJI: Everyone should be very proud of the NIH. It is the biggest biomedical organization in the world. On one campus, there are thousands of scientists studying everything you can think of. Side by side, people are studying the cancers, diabetes, ADHD, etc. So there is a clinical center, which is a hospital, where you have people coming in with all the illnesses and coming from around the country and the world, to undergo research studies that involve treatment. Our group will always treat patients in our studies to try and get them better. Based on our work in animals, we try novel and esoteric things in people and in some cases, they respond remarkably well. However, if they do not respond to the investigation "wonder drug," we then attempt to get them better with approved available treatments. Because we don't have the limitation of managed care knocking at the door, we have the time, so we can do things carefully and correctly.

The research takes place in a hospital where each floor houses a different disease. Our research beds are on the 3rd floor. Right next to the beds are labs. There are studies where you get the patient's blood sample and within minutes it is being processed in the lab around the corner. We have outstanding, dedicated clinical researchers (Drs. Zarate, Denicoff, Payne, and Quiroz work closely with me), nurses, and social workers all working together on the research unit.

Recently, a major mood and anxiety program has been developed here, to get people studying these disorders to work more closely together. Dr. Dennis Charney has done a tremendous job putting this together, and I think that will really pay off. We all have our different expertise and we are studying different illnesses or different levels (e.g., molecular level, brain imaging level), but we exchange information. We do a lot of studies to see if there's a medication that is safe in humans that might work for mood disorders. We then do clinical studies with that med, and we have a great team. We try and design the studies of a novel treatment, but we also work with a geneticist. Wouldn't it be nice to know if the person responds to the trial med at the genetic level or not? Then we also look at brain imaging while we study other molecular pathways, so that at the end of the day, we have some very useful information about does this pathway work, for whom, etc. Wouldn't that be great if we had a way of predicting who would respond? So many people work together on the same problem.

Our studies are very intensive. One problem is that in many of the studies, we have to insist on reasonably "pure" patients. We know in all reality that that's not what really happens. People are on multiple medications. What we have to do is to simulate only bipolar illness without any other confound. If we find something, then we can ask, is it also true in someone with comorbidities [coexisting illnesses, such as ADHD]? If we don't start out pure, then you never know. Sometimes people get referred to us and we have to turn them down through no fault of their own, and I am sure as an individual it must be extremely frustrating. But if we are not that selective initially, we will never get answers. The goal is to find things in pure populations and then to expand the studies.

The other thing we do is study a small number of patients (say, 50) very closely. They are on meds that are not working. We take them off meds, study them that way, then study them on a new treatment, and then if it doesn't work we use existing treatments to get them better. We do a few things on a small scale to get an idea, does it work? And then if it looks promising, we take it out to outside research labs where they have larger populations of patients. That way we capitalize on our strength here.

The NIH is an amazingly exciting scientific environment. There is a small group of people who are here permanently and a large group who are here for training and fellowship. Because we are such a unique facility, we have the best people from every place in the world applying to come here. Here, they have the opportunity to do things they can't do in their own countries. This is a unique place where we make a lot of breakthroughs but also train the next generation of scientists.

My group has both a combined basic science and a clinical group. We have a fairly large molecular and clinical lab where we have projects going on all related to mood disorders in some way. We look at other diseases because they provide clues about mood disorders. We do studies in rats, mice, postmortem brain tissues, and immortalized tissue -- you can take blood cells and treat them and then they will grow in the test tube forever. You can study the bipolar person's cells and what makes them different. We do all those things. Brain imaging, brain structure, and so on. Some of these treatments seem to be capable of causing neurons to grow back to normal sizes, and neurogenesis [growth of new nerve cells].

Then, as much as possible, we test our animal findings in humans. If we find a target for the drugs, we ask, is it abnormal in bipolar patients, can we measure it to see if it is abnormal, or if we manipulate that pathway do some patients get better? We take meds that have been developed in other diseases and then try them in bipolar patients. Then we inform our geneticists that these pathway are involved and ask, can a genetic mutation in bipolar disorder account for this? We ask, does this region of this gene regulate this important signaling molecule? If it looks significant and one of our signaling molecules is there, then we start to look at that gene to see if the bipolar person carries a gene in this region.

The Balanced Mind Parent Network: Do you see collaboration with patient advocacy organizations such as The Balanced Mind Parent Network as important for research?

MANJI: I can't overstate the importance of advocacy organizations. I think they are tremendously important. They are vital for helping patients and families on multiple levels, especially in an illness like pediatric bipolar disorder and psychiatric illnesses in general -- not much is known, and there is a lot of stigma, so accurate information may be less accessible than it is in other illnesses.

In terms of research collaboration, we try and enlist the aid of patient advocacy groups for different kinds of research studies. There has to be a partnership between the two. The simpler straightforward studies, such as genetic inquiries that involve interviews and DNA samples, are the types of studies for which the partnership might work best. It may seem at times that we are not doing those types of simple studies as much as we should. One of the problems is that today we still don't have any specific gene to be sure of, so a lot of the studies today are more complex. What is very important is the interviews, and it is hard to ensure quality control for those types of studies, so research groups may not jump on that, even though people are willing to participate. If you don't have adequate quality control, the DNA could be not useful and perhaps even hurt you -- classifying a patient as schizoaffective who should in fact be bipolar II -- the genetic signature will wash out. So the necessity of having things very, very well controlled limits our ability to study large numbers. If the results of a study will not be reliable, then it is better not to do that study at all. If you find a gene that is reliable, then you can export the research to a larger population.

The other obvious area of collaboration is treatment studies, a major area for potential collaboration with advocacy groups. It may seem that we are not doing those enough. I can only speak for my group. But again, the rigors of science demand that quite often, we have to exclude individuals who might, from a psychiatric standpoint, be ideal, but because we are really trying to hone it down to see does it work for people with this particular brain imaging signature, we unfortunately need to exclude those without that signature. So for the treatment studies, the collaboration is absolutely vital, but despite the attractiveness, we have to turn people down, because if we don't, it may set us back.

The last way in which this partnership is valuable is getting ideas and clues. While studies have to be very controlled, quite often it is the isolated incident that triggers a new idea. So if advocacy groups see certain trends that might lead to a study through a simple questionnaire, it can assist us immensely. And there are a lot of ways advocacy groups can help us do better research. The history of medicine is full of examples of someone noting an unexpected observation and making sense out of it. Making good observations is absolutely critical. In the managed care situation, the time necessary to follow individuals and look for changes and get clues is not allowed because of insurance requirements, and so family members who are constantly with the patients may come up with observations and clues that may be very useful to investigators.

I want to conclude by stressing that there is really good reason to be optimistic. Working together, allowing themselves or their children to participate in the research, it is a team effort between scientists and patients. Lots of people are taking this effort seriously and doing very good work. There is ample reason to believe that we will make breakthroughs.

1. Manji HK, Moore GJ, Rajkowska G, and Chen G. Neuroplasticity and cellular resilience in mood disorders. Molecular Psychiatry 2000:5:578-593.

Last updated: July 23, 2010


Join a cyber family of caring and supportive parents. We have a variety of groups designed to meet your needs. Click here to see a list of our online support communities. 

Contact our Family Helpline if you need information and resources specific to your family’s situation and can’t find what you're looking for on our site.