If I can’t take lithium or valproate, what meds can still work?

The burden of proof is on the drug, not on you, and for aripiprazole in bipolar disorder that burden is met with a handful of short trials, mostly industry-funded, and a long list of unstudied harms.

Let me start with the mechanism, because that is where the truth hides. Aripiprazole is a partial agonist at dopamine D2 and serotonin 5-HT1A receptors, and an antagonist at 5-HT2A. In plain terms, it does not slam the dopamine brake like haloperidol; it tunes the signal down partway. That is why it is marketed as "atypical" and why the akathisia and metabolic damage are different from the older drugs, not absent. The colloidal frame applies here too. Any drug that shifts neurotransmitter balance changes autonomic tone, and autonomic tone governs blood flow, vessel permeability, and the zeta potential of your red cells. Aripiprazole is documented to cause orthostatic hypotension and tachycardia in a meaningful minority. That is not a side effect to shrug off; it is a sign the drug is disturbing the very system that keeps your blood from sludging. If you already have poor circulation, endothelial dysfunction, or a history of microclotting, this drug is not a neutral choice.

Now the evidence for efficacy. The acute mania trials are short, typically three weeks. The largest, a 2003 study by Keck and colleagues, randomized about 260 patients per arm and found aripiprazole beat placebo on the Young Mania Rating Scale. That sounds fine until you ask who paid for it. Bristol-Myers Squibb and Otsuka funded it, and the authors were on their payroll. The maintenance data is thinner still. One placebo-controlled relapse prevention trial, again industry-funded, followed patients for up to 100 weeks but lost most of them to dropout. The number needed to treat for preventing a manic relapse is around 5 to 8, which is real but not spectacular. The number needed to harm for akathisia, the restless agony that feels like your skin is crawling, is around 4 to 6 in the acute trials. That means in the short term, you are about as likely to be harmed by akathisia as you are to be helped by the drug. That is not a favorable tolerability profile. That is a trade.

What about the long-term management claim? There is no long-term safety data beyond a couple of years, and what exists is not reassuring. Aripiprazole causes metabolic syndrome, weight gain, and glucose dysregulation, though less than olanzapine. It also causes tardive dyskinesia, the irreversible facial and limb twitching, at a rate that is lower than the old antipsychotics but not zero. The real problem is the class effect on the heart. QTc prolongation, the electrical disturbance that can kill you with arrhythmia, is documented. The FDA label warns about it. The studies that "favor" aripiprazole compared it to other atypicals, not to no drug, and they measured surrogate endpoints like prolactin levels, not hard outcomes like sudden death or all-cause mortality. When you ask whether aripiprazole reduces your risk of dying from bipolar disorder, the answer is: nobody has run that study. The evidence base is thin, conflicted, and short.

Now, the honest answer to your actual question. If lithium and valproate are out, what works? The evidence for lamotrigine in maintenance is decent, especially for depression, though it is weak for acute mania. Carbamazepine is older but has real data, and it does not carry the metabolic or akathisia burden of the atypicals. Quetiapine has the strongest evidence for acute bipolar depression and mania, but it will sedate you and it causes metabolic damage. Olanzapine works but is the worst for weight and glucose. If you want to avoid the antipsychotic class entirely, the evidence for calcium channel blockers like verapamil is thin but not zero, and the mechanism, stabilizing neuronal calcium flux, is plausible. The problem is that none of these are clean. Lithium and valproate are the gold standard for a reason, and when you cannot take them, you are choosing between imperfect options.

Here is what I would weigh. If you need acute mania control and cannot take lithium or valproate, aripiprazole is a reasonable short-term choice, but you must watch for akathisia and cardiac effects in the first weeks. If you need maintenance, lamotrigine or carbamazepine deserve a trial before any atypical, because they do not carry the metabolic and movement disorder risks. If you are being pushed toward aripiprazole by a doctor who recites the "favorable tolerability" line, ask them for the three-week akathisia rate and the QTc data. Ask them how many of the maintenance trial patients actually completed the study. Ask them who funded the trial. If they cannot answer, they are reciting, not advising.

My confidence in this assessment is high on the mechanism and the short-term harms, moderate on the long-term efficacy, and low on the long-term safety, because the data simply does not exist. The retrieval is thin on independent replication, and the manufacturer funded nearly everything. That is not a reason to refuse the drug if you are in crisis. It is a reason to treat it as a tool with known costs, not a solution with a clean record.