TMS is not FDA-cleared for bipolar depression and the evidence is mixed: a pooled analysis found higher response but not remission versus sham, and the one Deep TMS trial missed its primary endpoint. We consider it off-label for stable patients on a mood stabilizer, with psychiatrist review and weekly monitoring for a switch into mania. Insurance rarely covers it.
This page explains what the trials have found, why the results are less consistent than in unipolar depression, how real the mania risk is, which patients we might consider, and what the course looks like when we do proceed. If you have bipolar disorder and have been told TMS is simply unavailable to you, the truth is more nuanced; if you have been told it works just as well as for unipolar depression, that is not what the data say.
Who it is for
The patient we might consider is an adult with bipolar I or bipolar II disorder in a current depressive episode of at least moderate severity, who is stable on a mood stabilizer such as lithium, lamotrigine, valproate or an atypical antipsychotic, who has not had a manic, hypomanic or mixed episode in at least several months, and who has not responded to the medications with the best evidence for bipolar depression, typically quetiapine, lurasidone, lamotrigine or lithium, alone or in combination. Bipolar II patients, whose highs are less severe, are generally lower risk for a switch. Patients with rapid cycling, a current mixed state, psychotic features, a manic episode in the past six months or a history of antidepressant-induced mania are not candidates, because the switch risk in those groups is higher and the circuit we would be stimulating is the one most implicated in switching. Patients who are not willing to continue a mood stabilizer throughout the course are not candidates either.
How it works
The TMS protocol for bipolar depression in the research literature is mostly the same as for unipolar depression: high-frequency stimulation of the left dorsolateral prefrontal cortex. The rationale is the same too, since the prefrontal underactivity seen in unipolar depression is also present in the depressive phase of bipolar disorder. The complication is that bipolar brains switch, and activating a mood circuit in a person whose mood regulation is unstable carries a theoretical risk of overshooting into hypomania, the same concern that applies to antidepressants in bipolar patients. That is why a mood stabilizer is a precondition. Mood stabilizers, particularly anticonvulsants, also raise the motor threshold, which means the stimulation intensity needed to reach the cortex is often higher in these patients, and that is measured at the first session rather than assumed. Some research has used low-frequency right-sided stimulation instead, on the theory that it is less activating, but the data are too thin to say it is safer.
People with bipolar I disorder spend roughly three times as many weeks depressed as manic, and in bipolar II the ratio is closer to 40 to one, making the depressive phase the main source of disability.
Source: Judd et al., Archives of General Psychiatry, 2002 and 2003
A meta-analysis of 19 randomized sham-controlled trials including 181 patients with bipolar depression found that active TMS produced significantly higher response rates than sham, with an odds ratio of about 2.7, but not significantly higher remission, and no increase in treatment-emergent mania.
Source: McGirr et al., World Psychiatry, 2016
In a randomized sham-controlled trial of Deep TMS in 50 patients with bipolar depression on mood stabilizers, the active group improved faster at week four but the difference was not significant at the eight-week primary endpoint; no manic switches occurred in the active group.
Source: Tavares et al., Neuropsychopharmacology, 2017
A review of TMS trials in unipolar and bipolar depression found treatment-emergent mania in 0.84 percent of patients receiving active stimulation and 0.73 percent receiving sham, a difference that was not statistically significant.
Source: Xia et al., International Journal of Neuropsychopharmacology, 2008
What to expect
The evaluation can take place at any of our 13 offices or by video, and for bipolar patients it is more detailed. The psychiatrist reviews the full mood history, including every manic, hypomanic and mixed episode and what triggered it, scores the current depression with the PHQ-9 and HAM-D, scores current manic symptoms with the Young Mania Rating Scale, confirms mood stabilizer adherence and often checks a lithium or valproate level. If we proceed, treatment takes place at our TMS treatment centers, with Palm Beach Gardens as the primary center, on the standard 36-session schedule. You complete the PHQ-9 and a mania screen every week, and we ask a family member or partner to tell us if they notice reduced sleep, pressured speech or unusual energy. Any sign of a switch stops the course for reassessment.
Benefits
- A non-drug option for a phase of bipolar illness where antidepressants carry switch risk and limited evidence
- Pooled trial data show higher response rates than sham, with no excess of treatment-emergent mania
- No weight gain, sedation or metabolic effects, which limit many bipolar medications
- Can be delivered alongside lithium, lamotrigine or other mood stabilizers without drug interactions
- Weekly mood and mania monitoring built into the course, so any switch is caught early
Risks and side effects
Beyond the standard TMS side effects, two risks are specific to bipolar patients. The first is a switch into hypomania or mania. Pooled analyses of trials in unipolar and bipolar patients put the rate of treatment-emergent mania under 1 percent and not significantly higher than sham, but these were patients on mood stabilizers and under close observation, and case reports exist. We treat the risk as real and monitor for it. The second is seizure threshold: bipolar patients are more likely to be on medications that lower it, such as bupropion, or that raise the motor threshold, such as anticonvulsants, which changes dosing. Lithium does not raise seizure risk at therapeutic levels. We also screen more carefully for sleep deprivation, since poor sleep both lowers seizure threshold and can precipitate mania.
Cost and insurance
Most insurers, including Medicare, cover TMS for major depressive disorder and specifically exclude bipolar disorder, because the FDA clearance is for unipolar depression and the trials in bipolar depression are small. A few commercial plans have begun to cover bipolar depression under specific criteria, and our TMS coordinator will check your plan. When coverage is denied, an appeal with a letter of medical necessity documenting failed trials of quetiapine, lurasidone, lamotrigine or lithium occasionally succeeds. Otherwise a course is self-pay, and we quote the current rate before you decide. We will not code bipolar depression as major depressive disorder to obtain coverage.
Why bipolar depression is a different problem
People with bipolar disorder spend far more time depressed than manic: roughly three weeks depressed for every week manic in bipolar I, and closer to 40 to one in bipolar II. The depressive phase drives most of the disability, the lost work and the suicide risk. Yet the treatment toolbox for bipolar depression is smaller than for unipolar depression. Antidepressants have modest efficacy and can trigger mania. The medications with the best evidence, quetiapine, lurasidone, lamotrigine, lithium and the olanzapine-fluoxetine combination, carry weight gain, sedation or slow titration. Many patients have tried several and remain depressed.
That gap is why TMS is of interest, and also why it needs to be approached with care: the same circuit activation that lifts depression is, in theory, the mechanism that could push a bipolar patient past euthymia.
What the trials show
The pooled picture
A 2016 meta-analysis gathered 19 randomized sham-controlled trials that included bipolar patients, 181 in total. Active TMS was associated with significantly higher response rates than sham, with an odds ratio of about 2.7. Remission rates were not significantly different, and the trials were small enough that a real remission benefit could have been missed. Treatment-emergent mania was not more common with active stimulation. The authors concluded that TMS appears effective for the acute depressive phase and that larger trials are needed. That conclusion still stands.
The Deep TMS trial
The only randomized sham-controlled trial of Deep TMS in bipolar depression enrolled 50 patients on stable mood stabilizers and gave them 20 sessions over four weeks. The active group improved faster, with a significant difference at week four, but by the eight-week primary endpoint the sham group had caught up and the difference was no longer significant. No patient in the active group switched into mania. A trial of 50 patients that misses its primary endpoint does not prove the treatment fails; it proves the trial was too small to be sure either way. But it is the best controlled Deep TMS data in this population and we do not spin it.
Open-label and naturalistic data
Several case series and clinic registries have reported response rates in bipolar depression in the range seen for unipolar depression, often 50 percent or higher, in patients maintained on mood stabilizers. Without a sham arm these numbers include placebo response, which is substantial in bipolar depression, so they are encouraging rather than conclusive.
| Evidence | Finding | Weight |
|---|---|---|
| McGirr 2016 meta-analysis, 181 patients | Higher response than sham, not higher remission; no excess mania | Moderate |
| Tavares 2017 Deep TMS RCT, 50 patients | Faster improvement at week 4; primary endpoint at week 8 not met; no switches | Moderate, small sample |
| Open-label series | Response rates similar to unipolar depression on mood stabilizers | Low, no sham control |
| Xia 2008 safety review | Treatment-emergent mania 0.84% active vs 0.73% sham | Moderate |
The mania switch risk, honestly
Case reports of TMS-induced hypomania and mania exist, dating back to the 1990s, mostly in bipolar patients and a few in people with no prior bipolar history. In pooled trial data the rate is under 1 percent and no higher than sham. Both of those facts are true. The reconciliation is that trial patients were selected for stability, were on mood stabilizers and were watched weekly, which kept a real but small risk small. Patients who are off mood stabilizers, who are rapid cycling, who are in a mixed state or who have switched on antidepressants before are at higher risk and are the patients we do not treat with TMS.
Switches, when they happen, usually occur within the first few weeks and present as reduced need for sleep, pressured speech, irritability, unusual productivity or spending. Caught early, they typically resolve with stopping stimulation and adjusting the mood stabilizer.
Who we might consider
- Adults with bipolar I or II disorder in a current depressive episode of at least moderate severity
- Stable on a mood stabilizer, with a therapeutic level where applicable, and committed to continuing it
- No manic, hypomanic or mixed episode in at least the past several months
- No rapid cycling, no current mixed features, no psychotic symptoms
- Inadequate response to or intolerance of at least two guideline-recommended treatments for bipolar depression
- A family member or partner willing to help monitor for early signs of a switch
- Able to attend daily sessions for six weeks and to protect sleep during the course
Who we would not treat
- Anyone not on a mood stabilizer, or unwilling to stay on one
- Current or recent mania, hypomania or mixed state
- Rapid cycling, defined as four or more episodes in a year
- A history of antidepressant-induced mania, which signals a brain prone to switching
- Psychotic features in the current episode
- Active heavy alcohol or stimulant use, which both raise seizure risk and destabilize mood
How the course differs from unipolar TMS
The stimulation protocol is the same: high-frequency Deep TMS over the left prefrontal cortex, 36 sessions over about nine weeks. What differs is the surround. The motor threshold is often higher because anticonvulsants raise it, so your dose is individually measured rather than assumed. Monitoring is heavier: a PHQ-9 and a mania rating scale every week, a check-in with the psychiatrist at least every two weeks, and a named family member or partner who has agreed to call us if sleep drops or energy spikes. Sleep hygiene is non-negotiable, because sleep loss both lowers the seizure threshold and precipitates mania. Any sign of a switch pauses the course for reassessment.
Alternatives to weigh first
Before TMS, the medications with controlled evidence in bipolar depression should have been tried: quetiapine, lurasidone, lamotrigine, lithium, cariprazine, lumateperone or the olanzapine-fluoxetine combination. Psychotherapy designed for bipolar disorder, including interpersonal and social rhythm therapy and family-focused therapy, reduces relapse. ECT remains the most effective treatment for severe bipolar depression, particularly with psychotic features or catatonia. Spravato is in the same off-label position as TMS for bipolar depression, with rapid short-term benefit and questions about durability. Your psychiatrist will place TMS in this sequence based on your history rather than as a default.
If you have bipolar depression that has not responded to standard treatment and want an honest assessment of whether off-label TMS makes sense for you, Stedfast Psychiatry's TMS program starts with a psychiatrist evaluation at any of our 13 offices or by video. Treatment sessions take place at our TMS treatment centers, with Palm Beach Gardens, 11641 Kew Gardens Ave, Suite 209, as the primary center. Call (561) 437-2199 to schedule a TMS evaluation.
Frequently asked questions
Is TMS FDA-cleared for bipolar depression?
No. TMS devices are cleared for major depressive disorder, which by definition is unipolar, as well as OCD, anxious depression and smoking cessation. Bipolar depression has been studied in small trials with mixed results, and no manufacturer has sought or received clearance for it. Using TMS for bipolar depression is off-label, which is legal and common in medicine but means the evidence and the insurance coverage are both weaker than for unipolar depression.
Does TMS work for bipolar depression?
Sometimes, and less predictably than for unipolar depression. A pooled analysis of 19 small trials found that active TMS roughly doubled or tripled the odds of response compared with sham, but did not significantly increase remission. The one Deep TMS trial in bipolar depression showed faster improvement at four weeks that was no longer significant at eight. Open-label and naturalistic series report response rates similar to unipolar depression. The honest answer is that it helps some patients and the trials are too small to say how many.
Can TMS cause mania?
It can, rarely. Case reports of TMS-induced hypomania and mania exist, mostly in bipolar patients and occasionally in people with no prior history. Pooled trial data put the rate under 1 percent and not significantly different from sham, but those patients were on mood stabilizers and watched closely, which is exactly how we treat. The risk is higher without a mood stabilizer, in rapid cycling, in mixed states and in people who have switched on antidepressants before. Those are the patients we do not treat.
Why do I have to be on a mood stabilizer?
Because it is the single most important protection against a switch. Every trial that found TMS safe in bipolar depression enrolled patients on lithium, an anticonvulsant or an atypical antipsychotic. The mood stabilizer also treats the illness between episodes, which TMS does not. Anticonvulsants raise the motor threshold, so your stimulation dose may be set higher than a unipolar patient's, which we measure on the first day. Stopping a mood stabilizer to pursue TMS would be a mistake, and we will not proceed if you have.
Is TMS safer than antidepressants for bipolar depression?
The switch risk appears comparable or lower, though the data are limited. Antidepressants in bipolar depression carry a well-documented switch risk, especially tricyclics and venlafaxine, and their efficacy in bipolar depression is modest; most guidelines place them low on the list. TMS has no systemic effects and the pooled switch rate is under 1 percent. But TMS has far less evidence in bipolar depression than quetiapine, lurasidone, lamotrigine or lithium, which is why those come first.
Will insurance cover TMS for bipolar depression?
Usually not. Medicare and most commercial TMS policies cover major depressive disorder and exclude bipolar disorder. A small number of plans have added bipolar depression under specific criteria, and appeals with documentation of failed guideline-recommended medications occasionally succeed. Our TMS coordinator checks your specific plan and pursues authorization or an appeal when there is a realistic chance. Otherwise the course is self-pay, and we tell you the cost before you decide. We do not misrepresent the diagnosis to obtain coverage.
What happens if I start to feel too good during the course?
Tell us immediately, and we have asked a family member to tell us too. Needing less sleep, talking faster, feeling unusually productive or confident, spending more or taking risks are the signs we watch for. You complete a mania rating scale every week. If a switch is suspected, we pause TMS, your psychiatrist sees you promptly, and the mood stabilizer may be adjusted. Hypomania caught early resolves quickly in most cases. This is the reason we only treat patients who are already stable on a mood stabilizer.
Should I try Spravato instead?
Spravato (esketamine) is FDA-approved for treatment-resistant unipolar depression, not bipolar depression, so it is in the same off-label position. Ketamine and esketamine have been studied in bipolar depression with rapid short-term benefit and a low switch rate when given with a mood stabilizer, but durability is a concern. Neither TMS nor Spravato is clearly better for bipolar depression; the choice depends on your schedule, history and tolerance for each treatment's risks. A psychiatrist can weigh both with you.
Where this treatment is offered
Conditions this treats
If you are in crisis, call or text 988 or go to the nearest emergency room. Educational content; individual results vary. Reviewed by Dr. David Husted, board-certified psychiatrist.