The effect of vibration on falling asleep – a pilot study

Insomnia affects many people. Although non-pharmacological therapies, compared to pharmacotherapy, can produce long-term improvement with reduced risk of side effects, options for non-pharmacological treatment of insomnia are limited. Recently, research on the effects of rocking and mechanical whole-body vibration (WBV) on sleep has begun. Inspiration came from the everyday reality of car or train travel or the use of cradles – as is well known, these vibrational environments promote increased levels of sleepiness. In studies, WBV of 0.25 – 4 Hz has been shown to induce sleepiness and reduce sleep latency (delay) and improve sleep quality. However, it is not easy to develop a suitable vibration system with such low frequencies for therapeutic use. The aim of the US researchers, the authors of the presented report, was to investigate the effect of a novel system based on a vibrating bed with superimposed vibrating waves of two different (similar) frequencies on the falling asleep and neural activity of people with moderate insomnia. When applied, the superimposition of two vibrational waves with similar frequencies produces a new wave that oscillates with the average frequency of the input waves, but is amplitude modulated with a frequency equal to the difference between the input waves. This difference frequency is referred to as the beat frequency vibration (BFV).

• The mean delay of unambiguous sleep was (in minutes): for the control session 8.73 ± 2.43, for the SWV session 6.92 ± 1.59 and for the BFV session 5.18 ± 1.00.
• Participants rated the BFV session as more comfortable, helping them fall asleep faster and improving sleep quality. However, these ratings were not statistically significantly different from the control or SWV conditions.

Compiled from:

Study population

Fourteen students (eight women and six men, aged 22.2 ± 3.0 years) were recruited for the study. Most participants reported clinical insomnia of moderate severity. The average sleep latency exceeded 40 minutes.

Study procedure

Study participants completed sleep diaries every morning and evening. In addition, they attended four weekly laboratory sessions. During a single session, participants lay in bed and attempted to fall asleep. They did this for a maximum of two hours or until they awoke from their nap for good. Brain activity was measured using an electroencephalography (EEG) system.

The first session was aimed at familiarising the subjects with the experimental environment. In subsequent sessions, vibration or only the sound made when vibration was generated was used (control group).

Multi-scale sample entropy (MSE {part of the study not described in this paper}) was used as a measure of the complexity of the neuronal signals. 

Use of vibration in the study

A novel vibration system consisting of four vibration generators placed under each corner of the bed was used. Two frequency combinations were used:

1.all four generators operated at 26.75 Hz – standing wave vibration (SWV);

2.the generators at the top of the bed vibrated at 26.5 Hz and the generators at the bottom at 27 Hz, resulting in a differential frequency of 0.5 Hz – a horizontal wave running at the beat frequency vibration (beat frequency vibration, BFV).

The BFV is experienced as a travelling wave causing both tactile stimulation and a vestibular sensation similar to that felt during rocking.

Results

Sleep latency was considered according to 2 definitions: according to the American Academy of Sleep Medicine, (AASM) or according to the definition defined by the authors as ‘unequivocal sleep’.

The mean AASM sleep latency was (in minutes): for the control session 6.95 ± 1.83, for the SWV session 6.80 ± 1.56, and for the BFV session 4.94 ± 0.99.

The mean delay of unambiguous sleep was (in minutes): for the control session 8.73 ± 2.43, for the SWV session 6.92 ± 1.59, and for the BFV session 5.18 ± 1.00.

The above results were not statistically significant. However, statistical analysis showed that for unambiguous sleep, statistical significance would have been reached with a larger sample (n = 24), for comparison of the effects of BFV versus control. For AASM sleep, the sample size should then be 46 subjects. The differences between the BFV therapy and the control group were small, indicating that a larger sample may not affect statistical significance in this case.

Participants rated the BFV session as more comfortable, helping them fall asleep faster and improving sleep quality. However, these ratings were not statistically significantly different from the control or SWV conditions.

Comment

The presented results of this pilot study show that the superimposition of two different vibration waves (BFV) can be a good non-pharmacological treatment for sleep problems. The effects of the novel BFV approach (0.5 Hz) and the traditional SWV (26.75 Hz) on various physiological and subjective metrics were compared. However, due to the small number of participants in the study, conclusions are limited. As mentioned above, statistical analyses showed that in future studies a sample size of n = 30 would lead to statistically significant results for unambiguous sleep latency.

In conclusion, the 0.5 Hz wandering vibration wave BFV showed a tendency to reduce sleep latency, more so than the control or SWV conditions.

More in: