Bipolar Therapy Treatment: What Drug and Therapy Trials Show

Bipolar therapy treatment combines medication with psychotherapy. NIMH names mood stabilizers and atypical antipsychotics as the most commonly prescribed drugs, and lists several talk therapies that can be effective when used with medication^[1]. Its 2025 guide states that bipolar disorder usually requires lifelong treatment and does not go away on its own^[1].

How many people live with bipolar disorder is covered on our condition page. This guide covers treating bipolar disorder: what the options are and what randomized trials show for each phase of the illness.

Those phases are acute mania, bipolar depression and long-term maintenance. NIMH notes that drugs used for an acute mood episode can differ from those used to manage symptoms over the long term, and that some people try several medications before finding the ones that work best^[1].

The labels vary, but the approach does not: bipolar mood disorder treatment, and treatment for the older terms manic-depressive disorder and manic depression, refer to the same illness^[1].

Diagnosis shapes the plan. Bipolar I disorder is defined by manic episodes lasting at least 7 days or severe enough to need hospital care, while bipolar II disorder involves depressive and hypomanic episodes without full mania^[1]. How clinicians make that distinction is covered in our guide to how bipolar disorder is diagnosed.

In bipolar disorder treatment, psychotherapy is studied as an add-on to medication. The largest synthesis, a 2021 JAMA Psychiatry network meta-analysis, pooled 39 randomized trials with 3,863 participants comparing medication plus a manualized therapy against medication plus a control condition^[2].

Across 20 two-group trials with usable data, structured therapies were associated with lower recurrence rates than control treatments (odds ratio 0.56)^[2]. Psychoeducation with guided practice of illness-management skills was associated with fewer recurrences in a family or group format than in an individual one (odds ratio 0.12, with a wide confidence interval)^[2].

For depressive symptoms, cognitive behavioral therapy (CBT) was associated with stabilization compared with usual treatment; estimates for family or conjoint therapy and interpersonal therapy pointed the same way with less certainty^[2]. The authors caution that differences in populations, treatment length and follow-up temper these conclusions^[2].

The STEP-BD psychotherapy trial

The Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) tested intensive therapy directly. Among 293 outpatients with bipolar I or II depression, those assigned to up to 30 sessions of family-focused therapy, interpersonal and social rhythm therapy or CBT had a 64.4% year-end recovery rate, compared with 51.5% for a 3-session collaborative care control^[3].

Patients in intensive therapy were 1.58 times more likely to be clinically well in any study month, and the three therapies did not differ significantly from one another^[3].

NIMH describes the main approaches this way: interpersonal and social rhythm therapy works with a person's biological and social rhythms, family-focused therapy has helped adolescents and adults when started with mood-stabilizing medication after an episode, and CBT adapted for insomnia can help as part of treatment for bipolar depression^[1]. CBT's wider evidence base is reviewed in our guide to cognitive behavioral therapy for depression.

Bipolar I disorder treatments start with bringing mania under control. NIMH lists mood stabilizers such as lithium or valproate and atypical antipsychotics as the most common drug types^[1]. Condition-level data are on our bipolar I disorder page.

The broadest head-to-head comparison is a 2011 Lancet network meta-analysis of 68 randomized trials with 16,073 adults in acute mania^[8]. Haloperidol, risperidone, olanzapine, lithium, quetiapine, aripiprazole, carbamazepine, asenapine, valproate and ziprasidone were all more effective than placebo; gabapentin, lamotrigine and topiramate were not^[8].

Antipsychotic drugs as a group were significantly more effective than mood stabilizers in that analysis, and olanzapine, risperidone and quetiapine led to fewer discontinuations than lithium^[8]. The search ended in November 2010, so drugs studied later are not part of the comparison^[8].

Lamotrigine shows why phase matters. It did not beat placebo for mania^[8], yet it outperformed placebo for bipolar depression and for preventing depressive relapse in later analyses^[4][7].

Bipolar depression is often treated with a mood stabilizer or an atypical antipsychotic. NIMH states that antidepressants are not used alone because they can trigger a manic episode or rapid cycling^[1].

The largest drug comparison to date, a 2023 Lancet Psychiatry network meta-analysis, analyzed 101 randomized trials covering 20,081 adults and 68 medications^[4]. With moderate confidence, olanzapine plus fluoxetine, quetiapine, olanzapine, lurasidone, lumateperone, cariprazine and lamotrigine reduced depressive symptoms more than placebo^[4].

Effect sizes were modest, from a standardized mean difference of 0.41 for olanzapine plus fluoxetine to 0.16 for lamotrigine^[4]. Antidepressants as a class appeared effective but carried a higher risk of manic switch than antipsychotics^[4].

Adding an antidepressant to a mood stabilizer has been tested directly. In the 2007 STEP-BD trial of 366 patients, durable recovery was 23.5% with an adjunctive antidepressant and 27.3% with placebo, with no significant difference and similar rates of switching into mania^[5].

The STEP-BD psychotherapy results above also apply to this phase, since that trial enrolled people during an episode of bipolar depression^[3]. Depression often dominates bipolar II disorder; NIMH notes that many people with it spend extended periods in a persistent, low-grade depressive state^[1].

Maintenance treatment aims to prevent the next episode. A 2021 Molecular Psychiatry network meta-analysis pooled 41 randomized trials with 9,821 participants followed for an average of 70.5 weeks^[7].

Every active treatment except carbamazepine and paliperidone outperformed placebo for preventing any mood episode^[7]. Asenapine, lithium, olanzapine, quetiapine and valproate also had lower all-cause discontinuation than placebo^[7].

Lithium has the most specific evidence on suicide. A 2013 BMJ meta-analysis of 48 randomized trials with 6,674 participants found lithium was associated with fewer suicides than placebo (odds ratio 0.13) and fewer deaths from any cause (odds ratio 0.38)^[6]. It found no clear benefit for deliberate self-harm^[6].

Lithium also requires monitoring. MedlinePlus notes that doctors order lab tests to check the response, that the full benefit can take 1 to 3 weeks or longer, and that NSAIDs such as ibuprofen may interact with it^[9]. That drug page was last revised in 2017^[9].

Electroconvulsive therapy (ECT) can relieve severe bipolar symptoms and is usually considered only when other treatments have not worked or when a rapid response is needed, such as with suicide risk or catatonia^[1]. How the procedure works and what studies show is covered in our ECT explainer.

Transcranial magnetic stimulation (TMS) uses magnetic waves to relieve depression over a series of sessions. NIMH describes it as less powerful than ECT, but it needs no general anesthesia and carries little risk of memory effects^[1].

Light therapy is the best evidence-based treatment for seasonal affective disorder, and NIMH says it could also be considered for seasonal worsening of bipolar depression^[1].

FDA has not approved any complementary health approach for bipolar disorder. NIMH notes that supplements, yoga, acupuncture and similar approaches have not been comprehensively studied for it^[1].