Mood Stabilisers in Bipolar Disorder

what they are, how they work & which ones exist

🧠 marks optional specialist deep dives you can skip without losing the thread.

1 What a Mood Stabilizer Is

Before we get into any single drug, it pays to slow down on the label itself, because the term mood stabilizer is one of those words that sounds far more precise than it actually is. this first chapter stays deliberately conceptual, it sketches what the category really means, what it explicitly does not mean & how its meaning has quietly drifted across the last few decades. Think of it as the map worth having in your hand before we climb down into the pharmacology.

1.1 A Clinical Task, Not a Drug Class

The first time you hear the word mood stabilizer, a certain picture tends to form almost on its own: one clearly outlined kind of pill, a clean chemical group, the way you can picture something concrete behind a word like antibiotic or painkiller. And exactly that picture is the one that leads you astray, because a mood stabilizer is not a chemical family at all, it is a job.

That sounds a little odd at first, but it’s the cleanest way into the whole topic, since under this one label you find substances that share almost nothing pharmacologically. There’s lithium, a simple salt, more precisely an element from the very first column of the periodic table, which carries no classic docking site in the body at all in the sense of a receptor. Right next to it sit the anticonvulsants, originally developed against epileptic seizures, so valproate & carbamazepine & lamotrigine & then come the atypical anti psychotics, actually built for treating schizophrenia, things like quetiapine, olanzapine, aripiprazole and a few more. Three completely different worlds, all of them somehow sharing one hat.

And what actually links these substances has nothing to do with how they are built, the connecting thread is purely what they are meant to achieve over the course of the illness, which is catching the pathological swings of bipolar disorder upward into mania & downward into depression while keeping such episodes from coming back. Whoever manages that job earns the mood stabilizer label, no matter where the substance chemically comes from.

How strictly that label gets handed out, though, has been a professional debate for decades now. The strictest definition wants a true mood stabilizer to work in four areas at once, namely acute mania, acute depression, the prevention of manic episodes & the prevention of depressive ones, with the extra condition that none of these four may be worsened by the drug. Hold up that bar & to this day only a single substance clears it fully, which is lithium. The more pragmatic definition that everyday practice actually runs on asks for rather less, two of the four properties plus the principle of doing no harm in any phase & under that softer roof valproate, lamotrigine, carbamazepine & several atypicals find room as well.

This blurriness by the way, is not some cosmetic flaw the field simply forgot to tidy up. It carries a historical reason & that reason is almost embarrassingly mundane, because none of these substances was designed on the basis of a biological theory of bipolar disorder in the first place. Lithium was stumbled upon in 1949, by accident, by the australian psychiatrist John Cade, the anticonvulsants wandered into psychiatry through the detour of Epilepsy research & the atypicals arrived by way of schizophrenia. Put plainly, we found the tools first & only began to understand why they help a good while later & in a few corners we still don’t understand it. None of that is a weakness of the text. It simply reflects where the science stands today & it will keep us busy all the way through chapter 3.

1.2 What It is Not

Barely any term in psychiatry carries around as many misunderstandings as this one and almost all of them trace back to a single fear, that a drug like this dampens your personality, flattens you out, leaves you numb & quietly regulates away the highs & lows of normal life. It’s the image of the emotionally anaesthetized person who only functions & nothing beyond that. The worry is completely understandable, it just doesn’t match what actually happens.

Suppressing normal feelings was never the goal of a mood stabilizer meaning that nobody is meant to lose their joy over a project that finally worked out, or their grief when they lose someone close. The real goal is relapse prevention, so keeping the next pathological episode from arriving & catching the pathological extremes before they take over & the cleanest way to feel that is the gap between a healthy elevated mood & a full mania, where somebody doesn’t sleep for three nights, gives away their savings & feels truly invulnerable, or the gap between ordinary sadness & a depression that pins a person to the bed for weeks. Well-adjusted, an affected person still feels the whole normal range of experience, every bit of it, because what gets stabilized is the course of the illness across time, not the mood of one particular Tuesday afternoon.

There’s a second distinction worth drawing in here too, one that tends to get muddled in everyday talk & it goes like this: yes, some of these drugs make you tired & in the acute treatment of a mania that’s even welcome, since sleep counts among the central stabilizing factors we have. But that sedation is a tool, or at most a side effect & it is never the actual mechanism of action, so a mood stabilizer is far more than a glorified sedative. It reaches deeper & works more slowly & its most important benefit shows up not in the very first night but across months & years.

1.3 How the Class Changed Meaning

Thirty years ago, whoever said mood stabilizer almost always meant lithium, full stop. For decades, it was the drug that practically embodied the whole category & that picture has shifted fundamentally since then, in a genuinely remarkable direction.

Across the last two or three decades the atypical antipsychotics have climbed massively in prescribing practice, while lithium has quietly slipped back & the numbers leave little room for doubt. In the US the lithium share in outpatient bipolar treatment dropped from over 30 percent to under 15, while over the same stretch the atypicals rose from somewhere around 12 to over 51 percent. In Germany the lithium share of mood stabilizer prescriptions fell between 2009 & 2018 from 31.4 to 26.2 percent, while the antipsychotics climbed from 38.4 to 53.1, so the trend points the very same way across Europe & across the Atlantic alike.

And here is the remarkable part, because the shift cannot simply be put down to the newer drugs being better. Lithium is still the substance with the broadest & most consistent evidence for long term prevention & it remains the only one that carries a robust signal against suicidality. Exactly this tension has earned its own name in the literature, the lithium paradox, where the gold standard gets prescribed less and less right at the moment it actually stands on the firmest evidence.

The reasons behind it are less scientific than practical & very human so lithium demands regular blood checks and a structured monitoring of kidney & thyroid & that, plainly, makes work. The fear of kidney & thyroid damage is real enough, though in its actual extent it gets overestimated more often than not & layered on top come the aggressive marketing of the new patent protected atypicals, the thin profit margin of an old generic, the lingering stigma of the old drug and, not least, real gaps in training, since lithium pharmacology carries a reputation for being demanding. All of that together has quietly pushed a very effective substance off toward the sidelines.

Whether the shift is actually justified isn’t something you can settle with a plain yes or no & that’s precisely what makes the topic interesting. For single phases, above all acute mania & acute bipolar depression, the atypicals do carry solid advantages that we will get into properly later on. But lithium’s real crown sits somewhere else entirely, in the long term phase prophylaxis across both poles, bundled together with the anti-suicidal & the neuroprotective effect.

2 The Basic Principle: Two Poles & Maintenance

To make sense of why these substances behave so differently, you first need the basic architecture of the illness they are aimed at. bipolar disorder organises itself around two opposite poles & a long quieter stretch in between & each of those asks something different of a drug. this short chapter lays out that frame, the manic pole, the depressive pole & the phase prophylaxis that runs underneath the whole course, because nearly everything in the later chapters hangs off exactly this structure.

2.1 Manic Pole, Depressive Pole, Phase Prophylaxis

At its core, bipolar disorder is an illness of mood swinging between two opposite poles, at the one end stands mania, with its elevated or irritable mood, an excess of drive & energy, a clearly reduced need for sleep, racing thoughts, an inflated sense of self-worth and, frequently, genuinely risky behaviour, while at the other end stands depression, in a way the mirror image, all leaden heaviness, joylessness, missing drive & hopelessness. Plenty of people also live through mixed states, where parts of both poles show up at the same time & these mixed pictures are especially tormenting & count as especially high risk.

ICD-11 files bipolar disorder as an episodic affective disorder & it essentially separates bipolar I, with its pronounced manic episodes, from bipolar II, with hypomania & severe depressions. The finer diagnostic questions I went through in detail in my bipolar text, so the basic frame is plenty here and what really matters is the consequence for treatment, because out of the two poles plus the longitudinal course there grow three quite different clinical tasks.

The first task is treating an acute mania, which is essentially putting out a fire that is already burning. The second is treating an acute bipolar depression, so catching the fall as it drops downward and the third one, which is really the beating heart of long term treatment, is phase prophylaxis, keeping the fire from breaking out again in the first place. an acute episode runs for weeks to months, while the prophylaxis quietly accompanies a person across years and exactly this three way split is why the term mood stabilizer stays so stubbornly blurry, since a drug can be brilliant at one of these tasks & weak at another.

2.2 Why Substances Work Asymmetrically

The central asymmetry of the whole topic shows itself right here, barely any substance is equally strong at both poles & in prevention all at once. The drugs work asymmetrically & almost all of them lean to one side or the other.

Put in everyday terms, lithium is strong in mania & above all in prevention, while only moderate in acute depression. Lamotrigine is practically useless against acute mania, yet a reliable guard at the depressive pole, valproate does well against mania and poorly against bipolar depression, whereas quetiapine belongs to the small handful that achieve something across all three tasks. Most antipsychotics, meanwhile, lean toward the manic side & only a few of them reach over to the depressive pole at all.

So why the asymmetry? Because mania & depression, neurobiologically, are no clean mirror images of each other. It would be lovely & convenient to picture bipolar disorder as one single switch, sometimes shoved too far up & sometimes too far down, but the brain just doesn’t run that way. At the two poles you’ll find partly different circuits, different transmitter dynamics, the sleep wake rhythm and different processes inside the nerve cells, so a drug that turns down an overdriven manic state, say by dampening the dopamine tone, does little against a depression & can even make it worse.

The practical fallout of all this is large, because treatment gets tailored to pole & phase, very often in combinations & there simply is no single stabilizer that fits everyone. Once that clicks, it also becomes clear why a therapy gets switched the moment the phase changes & why two people carrying the very same diagnosis can end up on quite different drugs.

3 How They (Probably) Work

This chapter is one of the two cores of the text & it opens with a plain admission so: We actually know fairly precisely what these substances do on the molecular level, which channels & receptors & enzymes they act on, the real trouble being that we know far less well why exactly those molecular effects end up stabilizing a mood. Between the single molecule & the finished clinical effect sits a wide gap in knowledge & whoever keeps that in mind will read the following mechanisms the right way, as well-founded hypotheses rather than settled truth. Throughout all of it a modern, multifactorial understanding applies, one that long ago left behind the old idea of the single derailed transmitter.

3.1 Lithium, the Great Puzzle

Lithium has been in use for over 75 years now and to this day there is no real consensus on how it actually works. There’s no lithium receptor you could pin the effect on neatly, instead the effect spreads itself across several points of attack deep inside the nerve cell and exactly that variety is what makes the whole thing so hard to grasp.

The first big candidate is the inhibition of an enzyme with the clunky name GSK-3, short for glycogen synthase kinase 3. You can picture GSK-3 as a kind of central switching node inside the cell, one that influences well over a hundred different target molecules, among them the signalling of dopamine & glutamate, the adaptability of synapses, the inner clock, inflammatory processes & the cells own energy supply. Lithium throttles this enzyme, partly in a direct way by displacing magnesium at the enzymes working site & partly through more roundabout detours, but because GSK-3 is steering so many functions at once, it’s almost impossible to say which of these effects is really the mood stabilizing one.

Candidate number two is the so-called inositol depletion, known in the field, after the man who proposed it, as the Berridge hypothesis. Lithium slows the recycling of inositol, a building block of an important intracellular signalling cascade and through that it dampens the transmission of certain signals. The idea is elegant, it does carry one weak spot though, because if you lower the inositol level in mice by purely genetic means, the lithium effects can’t be reproduced in full, so the simple depletion story, on its own, falls a little short.

The third candidate is maybe the most exciting of the three, precisely because it points beyond mood altogether because lithium acts neurotrophic & neuroprotective, in that it raises the growth factor BDNF, imaging studies show an increase in grey matter under long term treatment & at the cellular level it works against programmed cell death, encourages the cells own self-cleaning & shifts the processing of the tau protein, which in turn throws a bridge across to the dementia hypothesis. On top of that comes a clear effect on the inner clock, since lithium lengthens the circadian period in a whole range of models from insect to human & that fits neatly with the disturbed day night rhythms you see in bipolar courses.

Pulled together, a unifying narrative starts to take shape & it says that lithium does not act on a single transmitter at all. It works down into the intracellular signaling pathways whose shared endpoint is neuroplasticity & neuroprotection, so the capacity of nerve cells to adapt, to connect & to survive & that view sits beautifully with the multifactorial model of bipolar disorder, with its components of disturbed mitochondrial energy production, disturbed rhythm & impaired synaptic plasticity. And yet one limit stays standing, because the one single mechanism of lithium, even now, we cannot name.

3.2 The Anticonvulsants

The second family grew out of epilepsy treatment, where anticonvulsants calm overexcitable nerve cells, so it seemed almost obvious to suspect that the very same dampening effect would also smooth out the mood. That suspicion turns out to be wrong though & it’s an instructive kind of wrong, because topiramate, gabapentin & levetiracetam are all strong anticonvulsants & yet practically useless for bipolar mood stabilization, which means the antiepileptic effect on its own simply doesnt explain the mood stabilizing one. What each individual agent achieves hangs instead on its own particular molecular profile.

Valproate is easily the most versatile of the three. it strengthens the dampening transmitter system GABA, blocks sodium channels, leans on certain calcium channels, reaches through the inhibition of so-called histone deacetylases all the way into the steering of genes & probably throttles GSK-3 on top. Which of these effects actually carries the antimanic & preventive action isn’t settled & its exactly this breadth that makes valproate clinically effective & mechanistically messy at the very same time.

Lamotrigine, by contrast, carries a clearly different profile & that difference explains its special role. it blocks sodium channels preferentially where too much is going on right at that moment, namely at the endings that release the excitatory transmitter glutamate, so it ends up dampening an overshooting glutamate release. this glutamatergic profile sets lamotrigine apart from both valproate & carbamazepine & fits the hypothesis that a glutamatergic component is involved in bipolar depression, which makes it understandable why, of all of them, lamotrigine guards the depressive pole rather than the mania.

Carbamazepine, finally, blocks mostly voltage dependent sodium channels, with effects on adenosine & on the NMDA receptor discussed on top of that. These days it sits more in the role of a reserve agent and that has less to do with its efficacy than with its difficult side effect & interaction profile, which well come back to.

3.3 The Antipsychotics

The third family pulls on a different lever altogether, mainly on dopamine its put simply, a mania goes hand in hand with an overactive dopamine signaling and the atypical antipsychotics step in right there, either by blocking the dopamine receptors of the D2 type, the way olanzapine, risperidone & quetiapine do, or by acting as partial agonists that neither switch the tone fully off nor let it run fully loose, the way aripiprazole & the D3 weighted cariprazine do. Either route lands you with an antimanic effect.

So that though, only explains the one side of it. The same substances also reach into the serotonin system, where a blockade of the 5-HT2A receptor, as quetiapine, lurasidone, olanzapine & lumateperone all show, together with a partial agonism at the 5-HT1A receptor, as in quetiapine, lurasidone & cariprazine, comes out antidepressant & anxiolytic. With quetiapine a small peculiarity rides along on top, because its active metabolite norquetiapine inhibits the reuptake of noradrenaline, so it behaves rather like a classic antidepressant & that’s one of the reasons why, of all of them, quetiapine reaches as far as the depressive pole.

And then there’s sleep, because a good many of these substances block histamine & alpha receptors & make you tired through exactly that, which in a mania is a real therapeutic gain, since restored sleep is one of the most effective stabilizing factors there is. so the mood stabilizing effect of these drugs ends up assembled out of several building blocks at once, a dynamic modulation of the balance between dopamine & serotonin, the regulation of the sleep wake cycle & glutamatergic effects layered in & the closer you look, frankly, the more the line between antipsychotic & mood stabilizer blurs.

3.4 🧠 The Speculative Thread: Intracellular Signaling & Neuroplasticity

This section is an optional deep dive, so whoever wants to can skip it without losing the thread, while whoever likes to climb one floor further down will find here maybe the most interesting thought of the whole chapter.

Three substance families, three completely different molecular stories and yet you almost have to ask whether there’s a common denominator hiding underneath all of it. Possibly there is & the answer would sit not up in the synapse but down in the cell. The older picture of psychiatry painted mood as a balance of transmitters in the synaptic cleft, a question of more or less serotonin & dopamine, while the newer picture digs a good deal deeper, because below the receptors, inside the nerve cell, nested signalling cascades are at work with players like GSK-3, secondary messengers, transcription factors and growth factors like BDNF. These networks decide how well neurons adapt, connect & keep themselves alive, so really they decide the cells plasticity & its resilience.

So that though, only explains the one side of it. the same substances also reach into the serotonin system, where a blockade of the 5-HT2A receptor, as quetiapine, lurasidone, olanzapine & lumateperone all show, together with a partial agonism at the 5-HT1A receptor, as in quetiapine, lurasidone & cariprazine, comes out antidepressant & anxiolytic. with quetiapine a small peculiarity rides along on top, because its active metabolite norquetiapine inhibits the reuptake of noradrenaline, so it behaves rather like a classic antidepressant & thats one of the reasons why, of all of them, quetiapine reaches as far as the depressive pole.

And then there’s sleep, because a good many of these substances block histamine & alpha receptors & make you tired through exactly that, which in a mania is a real therapeutic gain, since restored sleep is one of the most effective stabilizing factors there is. So the mood stabilizing effect of these drugs ends up assembled out of several building blocks at once, a dynamic modulation of the balance between dopamine & serotonin, the regulation of the sleep-wake cycle & glutamatergic effects layered in & the closer you look, frankly, the more the line between antipsychotic & mood stabilizer blurs.

4 The Individual Agents

now that we have a rough sense of how these drugs work, we can finally turn to the individual agents. so that the whole thing doesnt collapse into a dry stack of profile cards, a small reading aid up front is worth the little detour.

4.1 How to Read an Agent

With each of these drugs, three questions are worth asking. the first is which pole & which phase it actually covers, so whether it works in acute mania, in acute depression, in prevention, or only in one of those fields. The second is how safe it is, what side effects it drags along & whether there are particular risk situations to keep an eye on & the third is what monitoring it needs, meaning which checks have to run for it to stay safe over time.

In the German context a fourth question pushes its way in & its more important than you might first assume: what is the substance even approved for over here? with the antipsychotics a clear gap yawns open between the US american & the European approval picture & several drugs that form a firm part of bipolar therapy in the States carry no German approval for this indication at all, which is why ill flag it explicitly at the spots where it matters.

The defining property of lithium is its narrow therapeutic window, because the distance between an effective level & a toxic one is small. for prevention you aim, in Germany, at a serum level between 0.6 & 0.8 mmol per litre, lower again in older people & the measurement gets taken twelve hours after the last intake. that narrow window is precisely why the regular blood checks exist & its those checks that have earned lithium its reputation as the high maintenance drug.

Let the level climb too high & the signs of poisoning start to loom, first a tremor, nausea, and grogginess, then at clearly higher values confusion, seizures and, in the extreme, coma. A rise like that often gets set off by everyday situations you wouldnt immediately tie to lithium, things like fluid loss, vomiting or diarrhoea, a low salt diet, plus a handful of drugs, since anti inflammatories like ibuprofen, diuretics of the thiazide type & common blood pressure drugs from the ACE inhibitor & sartan groups all push the lithium level upward. in plain practice that means anyone on lithium should know that a harmless looking ibuprofen tablet, taken in the middle of a stomach bug, can quietly turn dangerous.

The monitoring more or less follows from all that on its own. Before starting you check kidney values, thyroid, calcium and, depending on the risk, an ECG, then the level itself gets checked after starting & after every dose change & at wider intervals thereafter, while kidney, thyroid and calcium carry on getting checked roughly every six months for good. Over the years lithium can bring on an underactive thyroid, which is well treatable, occasionally a raised calcium level by way of the parathyroids and, in rare cases stretched across very long timespans, a slowly declining kidney function. these checks are a bit of work & no more than that. they buy you the broadest evidence base any mood stabilizer has to offer.

4.2 Lithium

lithium is the reference substance & for one fairly simple reason, that out of all the drugs it covers the most ground. in acute mania it lands moderately to strongly effective, though with a slow onset and in acute bipolar depression it stays moderate, while its real strength sits in long term prevention across both poles, something no other drug pulls off in this breadth and with this depth of evidence. layered on top come two specialties that lithium alone offers, an anti suicidal effect that ranks among the most robust single pieces of evidence for medication based suicide prevention anywhere in psychiatry, even while it stays under discussion, plus a neuroprotective signal that surfaces, among other places, in lower dementia rates across large observational studies.

The defining property of lithium is its narrow therapeutic window, because the distance between an effective level & a toxic one is small. For prevention you aim, in Germany, at a serum level between 0.6 & 0.8 mmol per litre, lower again in older people & the measurement gets taken twelve hours after the last intake. that narrow window is precisely why the regular blood checks exist & its those checks that have earned lithium its reputation as the high maintenance drug.

Let the level climb too high & the signs of poisoning start to loom, first a tremor, nausea and grogginess, then at clearly higher values confusion, seizures and, in the extreme, coma. A rise like that often gets set off by everyday situations you wouldn’t immediately tie to lithium, things like fluid loss, vomiting or diarrhoea, a low salt diet, plus a handful of drugs, since anti inflammatories like ibuprofen, diuretics of the thiazide type & common blood pressure drugs from the ACE inhibitor & sartan groups all push the lithium level upward. In plain practice that means anyone on lithium should know that a harmless looking ibuprofen tablet, taken in the middle of a stomach bug, can quietly turn dangerous.

The monitoring more or less follows from all that on its own. Before starting you check kidney values, thyroid, calcium and, depending on the risk, an ECG, then the level itself gets checked after starting & after every dose change & at wider intervals thereafter, while kidney, thyroid and calcium carry on getting checked roughly every six months for good. Over the years lithium can bring on an underactive thyroid, which is well treatable, occasionally a raised calcium level by way of the parathyroids and, in rare cases stretched across very long timespans, a slowly declining kidney function. these checks are a bit of work & no more than that. they buy you the broadest evidence base any mood stabilizer has to offer.

4.3 Valproate

Valproate is a dependable workhorse against the manic side of things. in acute mania it does well, in mania prevention it stays solid & against bipolar depression it comes out weak, which on its own would still leave it an uncomplicated enough choice. What dominates the whole picture though is something else entirely, namely the by far heaviest reproductive restriction of any mood stabilizer & the red line is worth stating up front: in people who can become pregnant, valproate is in effect no longer an option in the bipolar indication, permissible at all only under the strict conditions of a dedicated pregnancy prevention programme.

Even well away from pregnancy, valproate carries a side effect profile worth respecting. Weight gain, sometimes considerable, is common & tremor, hair loss and, in younger women, hormonal changes along the lines of a polycystic ovary syndrome ride along with it, while rarer but serious are a liver injury, a pancreatitis, a drop in platelets & an ammonia driven brain dysfunction. Its real place today is mostly with men, with people who carry no pregnancy potential & wherever the alternatives have already failed or a coexisting migraine or addiction tilts the decision in its favour.

4.4 Lamotrigine

Lamotrigine is the guardian of the depressive pole, against acute mania it does nothing, in acute depression it earns at best a limited value as an add on, yet as a preventer against depressive episodes it is genuinely reliable & that is exactly where its main approval sits & exactly where it becomes most valuable, above all in depression dominant courses & in bipolar II.

the one catch you really have to respect is the titration & it pays to keep two very different things apart here. harmless skin rashes are actually fairly common under lamotrigine, somewhere around three to ten percent & the large majority of them stay completely benign. the genuinely dangerous reaction is a different beast, the Stevens-Johnson syndrome & that one, by contrast, is rare, in the order of four hundredths of a percent under monotherapy with proper slow titration. the real trouble is that you cannot reliably tell a harmless early rash apart from the start of a dangerous one just by looking & exactly that is why lamotrigine has to be eased in slowly across roughly six weeks. rushing the dose drives the risk up & so does valproate given alongside, since valproate lifts the lamotrigine level & pushes the risk of a serious reaction up steeply. the clinical rule that falls out of all this is simple & worth holding onto, patience while dosing up & at any new skin rash in the early weeks, pause the drug at once & get medical advice.

Once lamotrigine is properly established, it sits among the best tolerated drugs there are. it brings no notable weight gain, no sexual dysfunction & lands somewhere between cognitively neutral & slightly favourable, so for a lot of people with a depression heavy course it turns into a very comfortable long term option.

4.5 Carbamazepine

carbamazepine these days is more of a second choice agent, both against acute mania & in prevention & that owes less to any lack of efficacy than to its awkward accompanying profile, where one point in particular stands out: carbamazepine is a strong enzyme inducer. it cranks up the breakdown of a whole range of other drugs & drives their levels down through that, which gets especially treacherous with the contraceptive pill, whose effect can fail outright & further with blood thinners, other psychotropics & immunosuppressants & on top of everything it even speeds up its own breakdown, which makes finding the right dose more of a moving target.

Further quirks pile on from there, a tendency toward low sodium values in the blood, very rarely serious disturbances of the blood count & in people of certain asian descent a genetically raised risk of severe skin reactions that justifies a screening beforehand. A close relative, oxcarbazepine, carries a milder enzyme effect, though it leans harder toward low sodium & stands on weaker evidence in the bipolar indication. Bottom line, carbamazepine is a niche agent, the kind you reach for mostly when the other options are off the table.

4.6 The Antipsychotics: The Shared Principle

Before we walk through the individual antipsychotics, a quick word on where they belong, so- the modern atypicals are by now an integral part of bipolar therapy & in no way merely schizophrenia drugs & in the bipolar context the term antipsychotic is honestly a little misleading, since it suggests the job is treating psychoses, when really these drugs are doing something quite different here.

the shared principle we already met back in chapter 3, a dampening, or rather a retuning, of dopamine against the mania, with a modulation of serotonin & sleep on top that, in some substances, reaches all the way to the depressive pole. along exactly that line they split into two camps, the one group with its strength sitting clearly on the manic side and the other reaching over into depression & of all things its with that second group where the German approval reality turns into the decisive question.

4.7 The Mania-Directed Ones

this group shines on the manic side & in mania prevention, while against the depressive pole, in monotherapy, it comes out weak or simply unproven.

olanzapine is a very strong antimanic agent that also holds up in maintenance and in Germany it carries approval for mania & for phase prophylaxis, its big drawback being the metabolic load, since out of all the atypicals it brings the highest risk of weight gain together with blood sugar and lipid disturbances. aripiprazole, for its part, works antimanic & is approved for the prevention above all of manic episodes while staying metabolically friendly, its signature side effect being akathisia, that tormenting inner restlessness with an urge to keep moving & as a sole agent against bipolar depression it has flatly failed in studies. risperidone is approved orally against mania & works well, though it lifts the prolactin level noticeably, which can pull sexual & hormonal side effects along behind it & as a depot injection it gets used for prevention, in Germany off the approval in the bipolar indication.

For completeness, two further atypicals, asenapine & ziprasidone, are approved for mania in Germany while playing a smaller everyday role. asenapine gets taken up as a melt tablet across the oral mucosa & ziprasidone wants some attention because of its effect on cardiac repolarization & has to be taken with a meal.

4.8 The Depression-Directed Ones

This group reaches all the way to the depressive pole, the single hardest target there is & its precisely here that the gap between the american & the German approval picture grows widest, which is also precisely where you, as a reader, have to watch your step, because so much of the information floating around online quietly mirrors the US reality instead of ours.

quetiapine is the all rounder of the group & at the same time the only atypical approved in Germany across all three phases of bipolar disorder, so for mania, depression & maintenance alike. In bipolar depression it works at a dose around 300 milligrams, with sedation & a noticeable metabolic load as its drawbacks & for the depressive pole that makes quetiapine the pharmacological workhorse of German practice.

Now to the decisive German reality because cariprazine, lurasidone & lumateperone are, unlike in the US, not approved in Germany for the bipolar indication and cariprazine in particular is approved in the EU exclusively for schizophrenia, while in the States it gets used for bipolar mania and depression too. its a D3 weighted partial agonist, leans toward akathisia & carries an active metabolite with a very long half life that drags out the onset, so in Germany any use in the bipolar indication would land off label.

lurasidone, a first choice against bipolar depression in the US, is in the EU approved only for schizophrenia & in Germany it hasnt even been on the market since March 2015, after the manufacturer pulled it following the benefit assessment procedure. one pharmacological quirk on the side, it has to be taken with a meal of at least 350 kilocalories or it barely gets absorbed at all & while its metabolically friendly, for patients in Germany its in practice obtainable only through international pharmacy import.

lumateperone, finally, has been approved in the US since 2021 for bipolar depression & carries a very favourable profile when it comes to weight & movement disorders, yet in Europe there is no approval & in Germany its neither approved nor regularly available, so i list it here purely for completeness, since a German option it is not.

the bottom line for Germany, then, comes out soberingly clear. for the depressive pole what actually stands available, pharmacologically, is above all quetiapine as an approved option, lamotrigine for prevention, lithium & the free, off label used combination of olanzapine & fluoxetine, while the shiny US options largely never make it onto the table over here.

5 Safety With Conception & Pregnancy

this chapter is a delicate one & it deserves a calm, careful treatment, One thing belongs right at the front before anything else, a pregnancy does not put bipolar disorder on pause. an untreated episode in pregnancy or in the postpartum stretch is itself a considerable risk, reaching all the way to postpartum psychosis, which is a psychiatric emergency, so this is never about a blanket stop everything. its always about an individual weighing up, made together with the treating team & obstetrics & ideally begun long before conception.

the clear red line is called valproate! It carries by far the highest malformation risk, with severe malformations in the order of around ten percent, among them neural tube defects like spina bifida as well as heart & facial malformations & on top of that a lower mean IQ in the children & an autism risk raised roughly threefold after exposure in the womb. for that reason a european pregnancy prevention programme has been in force since 2018, under which valproate, in the bipolar indication, is contraindicated in people who can become pregnant, unless the strict conditions of that programme are met, so a reliable contraception, a yearly specialist review & a signed risk acknowledgement, which in Germany runs through the red hand letters of the BfArM. in 2024 a precautionary recommendation was even stretched to men, built on a statistical signal rather than proven causation & for every person with pregnancy potential valproate is, in bipolar treatment, in effect off the table.

carbamazepine sits in problematic territory too, with a neural tube defect risk around one percent and facial malformations and it drags along that already mentioned trap of rendering the contraceptive pill ineffective, a genuinely double problem in the middle of family planning, so carbamazepine likewise gets avoided in this situation.

lithium occupies a more like a middle position, means that it raises the risk of heart malformations slightly & dose dependently, among them the rare Ebstein anomaly & while the notorious old estimates were wildly exaggerated, modern large cohorts now put the absolute risk of an Ebstein anomaly after first trimester lithium exposure at about one to two per thousand, against roughly one per twenty thousand in the general population. The risk is therefore raised, yet in the individual case it stays low, lower doses are safer, a fetal heart ultrasound around the 18th to 22nd week makes good sense, the level monitoring tightens as things go on and around the birth itself lithium gets briefly paused. for a stable lithium responder, carrying on under careful supervision is, in plenty of cases, the better choice than gambling on a relapse.

lamotrigine, by the large pregnancy registries, counts as no meaningful malformation risk, which is why it belongs among the preferred options & its catch isnt teratogenic at all, its pharmacokinetic, in that the pregnancy hormones speed up the breakdown of lamotrigine so hard that the level can sink across the course by up to ninety percent, with a real danger of losing the effect. so pregnancy here calls for regular level checks & dose adjustments & after the birth a swift reduction, while breastfeeding stays possible in principle under observation of the infant.

quetiapine is the best documented atypical in pregnancy, with reassuring data behind it, which makes it a preferred option for the depressive pole & for maintenance. As a class the atypicals are no strong teratogens & the thing to watch is gestational diabetes, above all with the metabolically loading substances, while for olanzapine & risperidone there are small, not conclusively settled risk signals & quetiapine simply holds the cleanest data.

lined up in order, the risk hierarchy runs from unfavourable to favourable like this, valproate as by far the most problematic, then carbamazepine, then lithium with its modest dose dependent heart risk & after that the atypicals as a class sitting roughly level with lamotrigine, which carries no teratogenic risk but demands an active level management. the decision always stays individual & the worst move of all is almost always the unplanned, abrupt stop the moment a pregnancy becomes known, so anyone planning a pregnancy, or holding it possible, ought to bring the topic up early & actively with the treating psychiatrist.

6 How The Choice Is Made

so how does all of this actually come together at the bedside because once a real person sits across from you & a decision has to be made? this chapter pulls the threads together, the pole & the phase, the tolerability, the safety questions & not least the right diagnosis sitting underneath it all. none of it runs by a rigid algorithm, it is a weighing up & the following sections walk through the factors that genuinely tip the scale.

6.1 Pole, Phase, Tolerability, Safety

after everything weve walked through so far, this much is quickly told, there is no single stabilizer that fits everyone. The choice is a weighing up of several factors at once & which of them moves into the foreground genuinely shifts from person to person.

decisive, to begin with, are the current phase, so whether a mania, a depression, a mixed state or a stable maintenance phase is in front of you, together with the predominant pole across the whole course of the illness, since in depression dominant courses the mind goes rather to lamotrigine & quetiapine & in mania dominant ones rather to lithium, valproate & aripiprazole. tolerability & the side effect profile come in next, because with an existing metabolic syndrome youd rather steer clear of the strongly weight active drugs like olanzapine & quetiapine & lean toward metabolically friendly options like lamotrigine or aripiprazole. & beyond all that, safety around a wish to have children, coexisting illnesses, age, the earlier response, including in close relatives & not least the preference of the affected person themselves, whether thats a willingness to do regular checks or a sensitivity toward weight gain, all carry their own weight as well.

6.2 Diagnosis First

as varied as those factors are, one factor sits above all the others right at the very beginning, the correct diagnosis. it happens with ICD-11 as the primary frame, while DSM-5-TR serves alongside it wherever clinically meaningful differences crop up. That sounds self evident & yet it really isnt, because hypomanias & manias slip past unnoticed in practice all the time, since people mostly come looking for help while in the depression & rarely experience the good phases as anything pathological, which is why, on average, six to ten years pass between the first episode and the correct diagnosis.

why all that matters shows itself in one especially consequential scene, where an antidepressant given on its own, in an unrecognised bipolar depression, can tip a person over into a mania or a mixed state & speed up the switching of phases & its precisely for that reason that handing over an antidepressant alone, in this situation, isnt recommended as first choice in the guidelines. anyone with a family history of bipolar illness, or who recognises phases of unusually heightened energy and mood in themselves, would do well to raise it actively.

6.3 What The Evidence Shows Across The Whole Range

Order the study landscape by phase & a differentiated picture comes out, one that does lithium justice without inflating it. In acute mania the atypicals tend to be a touch faster & stronger than the classic mood stabilizers, while lithium & valproate hold their place as solid standards, so in pure acute mania lithium isnt automatically number one. in acute bipolar depression quetiapine and, where its available, lurasidone take the lead, while lithium here works only moderately & that too is not lithiums strongest discipline.

the crown lithium does take home sits in long term prevention. Here it carries the broadest & most consistent evidence for holding off relapses in both directions, into mania as much as into depression & pairing a classic mood stabilizer with an atypical lowers the relapse rate further still, especially on the manic side, with the two extras, the anti suicidal & the neuroprotective effect, riding along on top. exactly this split is what explains the lithium paradox, because judge the drugs only by their speed in the acute phase & the newer substances look more attractive, while judge them by the long game across years & lithium stays the reference.

one note belongs in here as well, theres a tilt built into study funding. a great many atypical studies are manufacturer funded & laid out for short acute phases, while the maintenance data on lithium & lamotrigine come more often out of academic sources & that asymmetry is part of why the newer drugs end up listed so prominently in the guidelines. & one last, frequently forgotten option deserves a mention, because it reaches beyond bipolar disorder altogether, the augmentation of a treatment resistant unipolar depression with lithium, which is well backed & yet, in everyday practice, used far too rarely.

7 Stopping: Why Extra Caution Belongs Right Here

stopping deserves a chapter of its own, because a mistake made right here can turn especially expensive. unlike with a lot of other drugs, ending a mood stabilizer is a therapeutic decision in its own right, with risks of its own, never just a side matter to wave through.

The best studied case is lithium & the finding lands hard. Stop lithium abruptly, inside a few days & you run into a relapse markedly sooner than someone who tapers slowly, since in the classic studies the time to relapse sat at a median of around four months after an abrupt stop, against roughly two years after a slow taper & theres the added phenomenon of a manic rebound that can overshoot well past the original level of illness.

a subtler observation rides along on top of that, much discussed in the field though far from settled, the suspicion that in a portion of those affected lithium might work a touch less well on restarting, after a stop & a relapse, than it did the first time around. i would treat that one with real caution, because it stays contested & the everyday picture is more reassuring than it sounds. A rebound episode can certainly happen when the drug comes off, depressive or manic, that part is real and its the reason you never stop on a whim, but the large majority do respond again on restarting & the situation usually gets brought back under control well. what does look solid underneath all of it is that the prophylactic benefit accumulates across years, so whoever drops the drug the moment stability has finally arrived may be giving away exactly the protection it took so long to earn.

in practice it comes down to this, never stop abruptly. taper slowly instead, medically accompanied, across weeks to months & shape a planned pregnancy in advance rather than slamming on the brakes in the first trimester, a principle that carries over, in spirit, to the other mood stabilizers as well. stopping isnt forbidden, far from it, there can be perfectly good reasons for it, but it is a decision you make carefully, with a plan & together with your medical support, not in the heat of a good phase.

8 What We Still Dont Know

a good explainer doesnt sign off with the impression that everything is sorted, because a great deal of it isnt & that frankly is what keeps this whole field alive.

the mechanism of action of lithium, after over 75 years, is still not conclusively understood. theres no lithium receptor & no confirmed, all unifying hypothesis, only a bundle of plausible mechanisms that together sketch a picture. just as open is the question of who will respond to lithium in the first place, where the best predictive value still lies in clinical features, an episodic course with full recovery between episodes, a classic euphoric mania, a family history of lithium responsive relatives, while genetic risk scores so far add only a little, explaining a very small slice of the differences. somewhere around a third of treated people turn out to be excellent responders & yet predicting that reliably, before you start, simply cant be done.

A conceptual question follows close behind, because substances like cariprazine and lurasidone work clear across all phases, so is the old split between mood stabilizer & antipsychotic even still meaningful & some already talk of polyvalent mood stabilizers. lithiums two special achievements stay subjects of research as well. the anti suicidal effect holds up across the older meta analyses & across large real world cohorts, even though the one big trial built specifically to test it came out negative and lately the guideline bodies have started drawing rather different conclusions from the very same data, so the effect counts, by the majority, as accepted while staying, in truth, contested. the neuroprotective signal runs much the same way, only here the preclinical side has actually grown stronger of late rather than weaker, with newer work pointing toward a role for the brains own endogenous lithium & tying the whole thing neatly back into the GSK-3 story & yet all of it stays preclinical & observational, a distortion through selecting healthier treated patients cannot be ruled out & clinical proof is still missing, with prospective trials running.

& out at the edges of the field, plenty is on the move because substances like Ketamine & Esketamine are being trialled in bipolar depression, so far off label & with an unclear switch risk, endoxifen, an inhibitor of protein kinase C, has reached phase III against acute mania, where it performed comparably to divalproex, cannabidiol is under exploratory study, fresh targeted GSK-3 inhibitors sit in preclinical development & mitochondria supporting strategies like N-acetylcysteine get discussed as add ons. even psilocybin assisted therapy is being tested in bipolar II depression in small studies, with explicit caution because of the switch risk.

Which leaves one closing thought, we hold effective tools that we dont fully understand & they unfold their benefit best when you pick them to fit the person & the phase in front of you, with lithium staying the quiet reference point through all of it, right as the field keeps growing broader. this text replaces no medical consultation, it only means to offer orientation & whoever carries questions about their own treatment is best off talking them through with the treating team that actually knows their course.