The same songs, with and without the beats: 64 care-home residents and their sleep

The hard question about binaural beats is never whether the audio helps. It is whether the beats do anything the music alone would not. This trial is built to answer exactly that, which is why it carries more weight here than its sixty-four participants would suggest.

Lin and colleagues recruited 64 residents with poor sleep quality from a long-term care institution in Taiwan and randomised them to two groups. Both heard the same thing: twenty minutes of Taiwanese Hokkien oldies, twice a day, three days a week, for fourteen days. In one group the recordings had binaural beats embedded underneath; in the other they did not. Participants were not told which they were getting. Sleep quality and depressive symptoms were measured by questionnaire, and autonomic function by heart rate variability.

Both groups got better, which is the first thing worth saying. Plain familiar music three times a week improved some measures of autonomic regulation and significantly reduced depression severity, and its effect on sleep quality was inconsistent rather than absent. Choosing music people grew up with, in a care home, is already an intervention.

On top of that, the group with the beats did better than the group without on two measures: sleep quality improved significantly more, and sympathetic nervous activity fell significantly further. Within that group there were also gains in mean heart rate and normal sinus beats, a drop in low-frequency normalised units, and reduced depression severity.

That between-group comparison is the finding. It is, as far as we can tell, the best available evidence that embedded binaural beats add something to the music they are embedded in for sleep, in a population that actually sleeps badly. We grade it moderate: a randomised trial with a properly matched active control, not yet replicated.

Limitations. One institution, one language, one musical tradition, and 64 people, so the result needs repeating elsewhere before it carries far. Fourteen days is short, and nobody was followed afterwards. Sleep quality is self-reported, which matters especially here: the Cochrane review of music for insomnia found that subjective sleep improves when measured sleep does not, and this trial recorded no polysomnography that could tell those apart. The heart rate variability result is objective and points the same way, which is the main reason we do not read the whole thing as expectation. We could not obtain the full text, so the funding statement, the exact beat frequencies and the questionnaire used are not confirmed here; the request is logged in docs/reports/fulltext-requests.md.