Missing Script

Get updates

Evidence Explorer · MS-N251 · Mind–body and sleep · insomnia in breast-cancer survivorship · Appraised November 6, 2026

A six-week voice-interactive CBT-I program delivered through a smart speaker improved insomnia severity and several sleep measures more than sleep education in a small randomized trial of breast-cancer survivors.

Evidence at a glance

  • Benefit signal: Large short-term insomnia benefit; improvements in sleep quality, wake time after sleep onset and sleep latency, but not total sleep time.
  • Certainty: Low to moderate. The randomized result is promising, but the trial was small, short and device-specific.
  • Risk context: Behavioral sleep treatment is generally low risk; temporary sleepiness, privacy and accessibility matter with a voice-activated platform.
  • Cancer-control relevance: No tumor-control, recurrence or survival outcome was tested.

What was studied

Seventy-six women with a history of breast cancer were randomized equally to six weeks of voice-interactive CBT-I or an educational control; 70 completed the study.

The primary endpoint was insomnia severity. Secondary measures included sleep quality, sleep latency, wake after sleep onset, sleep efficiency and total sleep time.

What the study found

Insomnia Severity Index scores fell 8.4 points with voice CBT-I and 2.6 with education (p<.001; d=1.41, 95% CI 0.87 to 1.94).

Sleep quality favored the intervention by 0.56 units (95% CI 0.39 to 0.74). Wake after sleep onset improved by 9.54 minutes (95% CI 1.93 to 17.10) and sleep latency by 8.32 minutes (95% CI 1.91 to 14.70). Total sleep time did not differ (0.01 hours; 95% CI −0.27 to 0.29).

What this does not establish

The study cannot tell whether the benefit came from CBT-I content, voice interaction, novelty or engagement. It was not powered for uncommon harms or long-term outcomes.

A proprietary or research platform may not be equivalent to consumer smart-speaker sleep skills.

Safety and practical context

Voice platforms introduce privacy and data-handling questions in addition to usual CBT-I considerations. Users should understand what is recorded, stored or shared.

Sleep scheduling should be modified for severe fatigue, driving risk, seizure disorders, mania risk or unstable medical symptoms.

Questions for the care team

  • Is the device’s privacy policy acceptable?
  • Does the program provide genuine CBT-I rather than generic sleep advice?
  • Who can adjust the schedule if daytime sleepiness worsens?

Sources and identifiers

Funding and conflicts: One author reported a grant from the Sleep Health and Wellness Center; another reported NCI support during the study and other outside funding. Readers should consult the full disclosure statement.

Source access and dates: Primary PubMed record and registry were appraised. Sources checked November 6, 2026. Last literature search November 6, 2026.

Educational evidence summary, not individualized medical advice. Psychological and sleep interventions should complement—not replace—oncology, palliative, mental-health or sleep-medicine care.