Brain Stimulation for Sleep: Why It Hasn't Worked Yet and What Comes Next

Sleep medicine has long sought non-pharmacological ways to improve sleep quality, and non-invasive brain stimulation (NIBS) has repeatedly been proposed as a promising candidate. Yet despite a surge of published studies, no NIBS approach has accumulated sufficient evidence to be recommended as a treatment for any sleep disorder. This 2025 perspective review by Krone and colleagues, published in the Journal of Sleep Research, dissects why progress has stalled and charts a credible path forward.

The authors systematically identify three compounding failure modes in existing research. First, study designs are chronically underpowered and lack effective blinding—critically important because stimulation devices generate stronger placebo effects than either cognitive behavioural therapy or placebo pills in insomnia trials. Second, conventional TMS and transcranial electrical stimulation (tES) have inherent technological limits: poor spatial precision, inability to reach subcortical sleep-wake circuits, high interindividual variability in neurophysiological response, and safety-imposed caps on stimulation intensity. Third, the field suffers from heterogeneous outcome measures and an influx of poorly described studies that skew meta-analyses toward false positives.

On the technology front, two emerging modalities offer genuine advances. Transcranial temporal interference stimulation (TIS) uses intersecting high-frequency electric fields to create focal amplitude-modulated stimulation deep in the brain without strongly activating overlying cortex. Transcranial ultrasound stimulation (TUS) delivers focused mechanical waves with high spatial and temporal resolution to subcortical targets. Both extend NIBS reach to thalamic and hypothalamic nuclei that are central to sleep-wake regulation—structures simply out of reach for legacy approaches.

The review also highlights closed-loop auditory stimulation (CLAS) as a particularly mature real-time strategy. By detecting ongoing slow oscillations and delivering precisely timed acoustic pulses, CLAS reproducibly enhances NREM slow waves and associated sleep spindles—the neural signatures most linked to memory consolidation and restorative sleep. This approach sidesteps many blinding and variability problems because stimulation is yoked to the individual's own brain rhythms.

Conceptually, the authors argue that chronic insomnia disorder is a poor initial proving ground for NIBS because subjective complaints and objective polysomnographic measures are poorly correlated, making effect sizes hard to detect and interpret. Instead they propose four more tractable targets: (1) amplifying specific sleep oscillations to boost concrete sleep functions such as memory consolidation or glymphatic clearance; (2) using targeted memory reactivation during REM sleep to modify nightmare content and traumatic memories in PTSD; (3) increasing daytime wake intensity in depression to relieve fatigue and build homeostatic sleep pressure; and (4) disrupting pathological interictal and ictal activity in sleep-dependent epilepsies where NIBS effects on cortical excitability are already better characterised. These targets offer clearer mechanistic rationales and more objective, measurable endpoints.