TMS and Sleep: What the Research Shows and What Remains Uncertain

"Poor sleep" sounds like one problem, but it can describe many different conditions. A person may have difficulty falling asleep, staying asleep, waking too early, maintaining a stable sleep schedule, breathing normally during sleep, controlling uncomfortable leg sensations, or sleeping through pain, hot flashes, medication effects, anxiety, depression, or neurologic disease.

That matters when evaluating transcranial magnetic stimulation research. A study of primary insomnia does not automatically apply to sleep apnea. A study of sleep problems in fibromyalgia does not prove an effect in uncomplicated chronic insomnia. A change on a sleep questionnaire does not necessarily mean that objective sleep architecture changed.

Researchers are studying whether selected TMS protocols can influence cortical hyperarousal and brain networks involved in sleep regulation. Some randomized trials report encouraging improvements in subjective sleep measures, and a smaller number report selected objective changes. Protocols vary substantially, long-term evidence remains limited, and TMS is not FDA-cleared for insomnia or general sleep improvement.

Sleep is produced by coordinated activity across several brain systems. Circadian timing, sleep pressure, arousal networks, emotional regulation, sensory processing, breathing, movement, pain, and environmental input all contribute.

One model of chronic insomnia involves hyperarousal. In simple terms, the brain and body remain too activated when they should be shifting toward sleep. Research has identified differences in cortical excitability and network connectivity in some people with insomnia. That gives investigators a scientific reason to ask whether inhibitory or otherwise targeted stimulation can influence those patterns.

A plausible mechanism is only the beginning. Sleep complaints can arise from many causes, and a change in cortical excitability may not address the dominant problem for a particular person. This is why careful studies define the diagnosis being investigated rather than treating all poor sleep as interchangeable.

Sleep studies have targeted the right dorsolateral prefrontal cortex, left dorsolateral prefrontal cortex, dorsomedial or medial prefrontal regions, and other sites. Low-frequency rTMS is common because it is generally used to reduce cortical excitability, but investigators have also studied continuous theta burst stimulation and other patterns.

Trials differ in motor-threshold intensity, pulses per session, number of sessions, treatment duration, coil placement, concomitant medication, and whether sleep is measured during treatment, immediately afterward, or months later.

Those differences matter. A 1 Hz right-prefrontal protocol delivered over 20 sessions is not equivalent to a 10-session continuous theta burst protocol directed at another region. A positive finding from one cannot be used as proof for the other.

The most common research outcomes are subjective questionnaires such as the Pittsburgh Sleep Quality Index and the Insomnia Severity Index. These instruments are clinically useful because they measure how people experience their sleep and daytime impairment.

Subjective improvement matters. It is not the same as objective evidence that total sleep time, sleep efficiency, sleep onset latency, wake after sleep onset, rapid-eye-movement sleep, or slow-wave sleep changed.

Objective measures may include polysomnography or actigraphy. Polysomnography records brain waves, eye movements, muscle activity, heart rhythm, breathing, oxygen, and other variables during sleep. Actigraphy estimates sleep and wake patterns using movement over time. Each method answers different questions and has limitations.

The strongest sleep literature would show that patients feel better, function better during the day, demonstrate relevant objective changes, and maintain those improvements after treatment stops.

A 2026 systematic review and meta-analysis included 19 studies, representing 23 trials and 1,690 adults with insomnia disorder. Pooled analyses favored active rTMS over sham on subjective sleep measures, including the Pittsburgh Sleep Quality Index and Insomnia Severity Index.

The review also examined objective sleep variables. Sleep efficiency showed a pooled difference favoring rTMS, but total sleep time, sleep onset latency, wake after sleep onset, and rapid-eye-movement sleep did not show statistically significant pooled differences in the limited studies available for those outcomes.

The subjective findings are encouraging. The interpretation still requires caution. The PSQI analysis had substantial heterogeneity, protocols and targets differed, objective data came from relatively few studies, and only a limited part of the literature addressed follow-up. A pooled average does not identify the optimal protocol or establish durable benefit for an individual.

A 2025 pilot trial randomized 62 participants with chronic insomnia to active or sham medial-prefrontal rTMS. Forty-one completed 20 sessions and the follow-up assessments. Response and remission rates were higher in the active group immediately after treatment, but those between-group differences were no longer statistically significant at the four-week follow-up. Some symptom measures remained different.

That pattern is informative. It suggests a possible short-term signal while preserving the question of durability. It also shows why the immediate end-of-treatment result should not be presented as though it guarantees sustained remission.

A 2024 randomized, sham-controlled study evaluated continuous theta burst stimulation in 46 people with primary insomnia, with 41 completing treatment and follow-up. The active group had lower insomnia severity, selected objective differences, and reported stability through six months. The result is encouraging, but it remains one modest-sized study using a particular protocol. Independent replication is essential before broad conclusions are drawn.

Sleep studies are especially vulnerable to expectation effects because the outcomes often include how a person perceives sleep quality, latency, and daytime function.

A 2019 primary-insomnia study explicitly examined both efficacy and placebo response. The existence of a meaningful sham response does not mean symptoms are imaginary. It means that expectation, attention, study participation, natural fluctuation, and the sensory experience of a procedure can influence reported outcomes.

Credible sham design is difficult in TMS because active stimulation creates sound, scalp sensation, and muscle movement. If participants can guess their assignment, blinding becomes weaker. That is one reason study design matters as much as the size of the reported improvement.

Some TMS studies evaluate primary or chronic insomnia. Others include participants whose sleep problems occur with depression, fibromyalgia, Parkinson's disease, chronic pain, or another condition.

A 2024 double-blind randomized study in 42 patients with fibromyalgia reported differences between real and sham stimulation on subjective sleep measures and polysomnography. That study is useful because it included objective sleep testing and follow-up. It is indirect evidence for general insomnia because fibromyalgia-associated sleep disturbance has its own biology, pain burden, and treatment context.

Similarly, a trial in patients already taking hypnotic medication evaluates an adjunctive strategy, not TMS as a standalone intervention. Results should be described in the population and treatment context in which they were obtained.

TMS research should not collapse all sleep complaints into a single diagnosis.

Poor sleep may be related to:

A neuromodulation protocol studied for insomnia does not diagnose or correct these other causes automatically. Research into TMS and sleep does not replace an appropriate medical and sleep evaluation.

When I read a sleep TMS paper, I would ask:

Those questions help separate a promising signal from a settled clinical conclusion.

The studies below were selected because they contribute different kinds of information: pooled evidence, recent sham-controlled trials, placebo-response context, adjunctive use, and objective sleep measurement. Inclusion does not mean that a study proves effectiveness or applies to every sleep disorder.

Current research does not establish that TMS permanently corrects insomnia, restores normal sleep architecture in every patient, treats every cause of poor sleep, replaces established sleep evaluation or care, or produces a predictable and durable benefit for an individual.

The sleep literature contains an encouraging signal, particularly in subjective insomnia measures. It also contains substantial heterogeneity, relatively few objective sleep studies, limited long-term follow-up, varied targets and pulse patterns, and different clinical populations.

The correct conclusion is not that TMS has been proven to treat insomnia. It is that carefully designed studies have reported findings that justify further research, while the optimal protocol, durability, objective effects, and appropriate population remain uncertain.