Jesse Koskey, MD. Associate Clinical Professor, Department of Psychiatry and Behavioral Sciences, UC Davis Health, Sacramento, CA.
Dr. Koskey has no financial relationships with companies related to this material.
The first known lullaby was recorded on a cuneiform tablet in Babylonia about 4,000 years ago, and the use of music to aid sleep has continued right up through the age of Spotify. Music therapy is cheaper, easier, and safer than standard insomnia treatments, but is it a sound recommendation to make for our patients?
A look at the data
Music therapy for insomnia is not supported by the kind of large clinical trials that would pass muster at the FDA, but it does have empiric support. The best of these studies compared music to an active control, such as sleep hygiene (avoiding daytime naps), acupuncture, relaxing activity, or listening to audiobooks. A network meta-analysis compared these varied treatment arms with music therapy using data from 1,339 subjects in 20 controlled trials. Music was the most effective intervention both for the overall severity of insomnia (measured by the patient-rated Pittsburgh Sleep Quality Index [PSQI]) as well as sleep quality and initiation. Similar results were obtained from a Cochrane meta-analysis, which used a higher bar for study quality, including 13 randomized controlled trials that compared music therapy to usual care in 1,007 subjects (Jespersen KV et al, Cochrane Database Syst Rev 2022;8(8):CD010459).
Surprisingly, the effect size for music therapy was in the moderate to large range in both of those meta-analyses, which is comparable to the effect we see with z-hypnotics and benzodiazepines (Nowell PD et al, JAMA 1997;278(24):2170–2177). The caveat is that most of the studies enrolled people with only mild sleep problems. When tested in more severe insomnia, music therapy was no more effective than audiobooks or waitlist control (Jespersen KV et al, J Sleep Res 2019;28(4):e12817).
There are other problems with this research. Although most studies used blinded raters, it wasn’t possible to blind the subjects (or “deafen” them, as it were) to the intervention. The studies were also small and depended on subjective reporting. When they did include objective measures of sleep like polysomnography, the effects were usually not significant. In contrast, objective measures improve with hypnotics and cognitive behavioral therapy for insomnia (CBT-I), although these changes are usually mild.
Music can, of course, have observable physiological effects on listeners. Blood pressure, respiration, and heart rate can change tempo in response to musical cues, a phenomenon called entrainment (Bernardi L et al, Circulation 2009;119(25):3171–3180). Musical crescendos are associated with vasoconstriction, and relaxing music correlates with reduced blood pressure and vasodilation in listeners. The ambient music trio Marconi Union deliberately attempted to entrain relaxation through their composition “Weightless,” which reduced blood pressure and anxiety in a small study of healthy volunteers. The piece starts at 60 beats per minute (bpm) and gradually slows over the next eight minutes to finish at 50 bpm.
A musical protocol
There is no right way to apply music for sleep. In most studies, subjects listened to relaxing music for 30–60 minutes before bed. Headphones worked as well as speakers, and music worked regardless of whether subjects were simply instructed to listen to it or were given more complicated relaxation instructions. Sometimes the music cut off after a set time and sometimes it played through the night.
Although most studies involved relaxing music, there was no secret formula to the tunes, and the intervention worked just as well when the researchers selected the tracks as when the subjects did. Familiar tunes without lyrics were preferred, as were slow tempos, regular rhythms, bass tones, and tranquil melodies. Successful trials drew from Western, Indian, and Chinese classical; new age; Gregorian chant; Celtic; and electronic music (such as binaural beats in the 5–7 Hz theta range). One piece that rose to the top in a comparative study was “Weightless,” the ambient track by Marconi Union referenced above.
Even nonrelaxing music can be beneficial, as long as it’s enjoyable. Simply enjoying music of any sort may be helpful (Dickson GT and Schubert E, Sleep Med 2019;63:142–150). A 2020 randomized controlled trial pitted an album called The Most Relaxing Classical Music against sleep music self-selected by study subjects. A total of 95 university students with self-reported difficulty sleeping were randomized to the Most Relaxing group, the bring-your-own group, or a no-music control for four weeks. The prescribed album featured slow, calm pieces such as Pachelbel’s “Canon” and Debussy’s “Clair de Lune.” The self-selected music was not necessarily slow or relaxing. In fact, over 75% of subjects chose Western and Japanese pop, anime, and video game tunes. After a month, both music groups showed similar improvements on the PSQI, while sleep did not change in the nonmusic controls (Yamasato A et al, Tokai J Exp Clin Med 2020;45(4):207–213).
In addition to acting through relaxation, entrainment, and enjoyment, music may also benefit sleepers by masking unwanted environmental sounds, the way white noise does (Dickson and Schubert, 2019). Music can provide a helpful source of distraction for patients who struggle with depressive rumination or posttraumatic anxiety at night, and it is much better than watching TV in the bedroom with its sleep-disrupting blue light.
Distraction may become discordant for some, however. One paper reported that listening to music regularly was associated with “involuntary musical imagery,” aka earworms, as well as a 54% worsening of PSQI when the earworms occurred at night. Female sex and instrumental music were more associated with earworms (Scullin MK et al, Psychol Science 2021;32(7):985–997).
The heterogeneity of subjects, methods, and mechanisms linking bedtime music and improved sleep indicate that music is a flexible and helpful intervention for patients with mild to moderate insomnia. For most, slow, relaxing music produces the best effects. Today’s digital music players have sleep timers that can turn the music off after a set time period. Based on the results above, patients could start by setting a timer for 45 minutes, but those who struggle with middle-of-the-night awakenings may want to keep music accessible from bed. Patients who don’t respond to music or have more severe insomnia may need CBT-I or hypnotics.
Music improves subjective complaints of insomnia. It is safer than sleep medication, and a healthy alternative to sleeping with the TV on.
PO Box 626, Newburyport MA 01950