The efficacy of sensory neural entrainment on acute and chronic pain: A systematic review and meta-analysis

Aim

The use of sensory entrainment methods may provide non-pharmacological interventions for pain management.^16,23–24 One previous meta-analysis focused on binaural beats on cognitive and psychological states, including pain. ²⁶ However, there are currently no systematic reviews or meta-analyses synthesising and critiquing the evidence about the effectiveness of sensory entrainment methods for pain perception. Our study addresses this gap by critically appraising the current evidence on the efficacy of sensory entrainment in reducing adults’ subjective pain intensity.

Search strategy

In November 2020, five electronic databases were systematically searched with no date limits applied: Medline, Embase, PsycInfo, Web of Science and Scopus. Keywords and their synonyms were used with ‘AND’ and ‘OR’ to narrow or broaden the search depending on the search strategy for each database. The electronic searches were followed by hand citation searches and reference list review to identify additional studies. The search was limited to the English language and only included published studies.

Study eligibility

The population, intervention, control, outcomes and studies (PICOS) framework ²⁸ was used to specify the inclusion and exclusion criteria and inform the key search terms used. The PICOS components are presented in Table 1.

OPEN IN VIEWER

Table 1.

The PICOS components of the research question and respective search terms used.

PICOS component	Description	Key search terms
Population	Human adults experiencing and not experiencing pain	—
Intervention	Sensory entrainment including auditory, visual and tactile	‘Sensory entrainment’ OR ‘auditory entrainment’ OR ‘sound entrainment’ OR ‘acoustic entrainment’ OR ‘audiovisual entrainment’ OR ‘music entrainment’ OR ‘visual entrainment’ OR ‘photic entrainment’ OR ‘light entrainment’ OR ‘tactile entrainment’ OR ‘touch entrainment’ OR ‘vibration entrainment’ OR ‘monaural beats’ OR ‘binaural beats’ OR ‘isochronic beats’ OR ‘isochronic tones’ OR ‘binaural tones’ OR ‘monaural tones’ OR ‘binaural sounds’ OR ‘monaural sounds’ OR ‘isochronic sounds’ OR ‘flashing light’ OR ‘flashing screen’ OR ‘LED glasses’ OR ‘flashes of light’ OR ‘light flashes’ OR ‘pitch panning’ OR ‘slow brushing’ OR ‘fast brushing’ OR ‘brainwave entrainment’ OR ‘oscillatory entrainment’ OR ‘neural entrainment’ OR ‘light pulses’ OR ‘gamma entrainment’ OR ‘alpha entrainment’ OR ‘beta entrainment’ or ‘theta entrainment’ OR ‘delta entrainment’ OR ‘hemi-sync’ OR ‘hemisync’ OR ‘hemispheric synchronised’ OR ‘auditory beats’ OR ‘vibration therapy’ OR ‘rhythmic stimulation’ OR ‘rhythmic sensory stimulation’
Control	Sham stimulation, control stimulation, no stimulation	—
Outcomes	Pain perception and intensity	‘Pain perception’ OR ‘pain threshold’ OR ‘pain measurement’ OR ‘pain management’ OR ‘pain assessment’ OR ‘pain severity’ OR ‘analgesia’ OR ‘analgesic effect’ OR ‘analgesic activity’ OR ‘nociception’ OR ‘antinociceptive’ OR ‘visual analogue scale’ OR ‘numerical rating scale’
Studies	Experimental and observational studies	—

Study selection

One author [RM] undertook the initial searching and screening of abstracts. All other authors were consulted when there was doubt of eligibility. The full texts were read and screened by at least two authors [RM and HN/IO/QV]. Data abstraction and quality assessment were undertaken independently by two authors [RM and HN], and a third was consulted where discrepancies arose [QV/IO].

Data extraction

A data extraction sheet for included studies captured: study citation (author, year of publication); location; study setting; study design; number of participants; characteristics of participants; entrainment method; frequency of entrainment; time of exposure; comparison/control group; outcome measures; timing of measurement; primary outcome; secondary outcome and final conclusions.

Meta-analysis

Inclusion criteria. We used four criteria for including a study in the meta-analysis: (1) was an experimental study; (2) used frequency stimulation in a sensory modality; (3) had control treatment; and (4) provided sufficient information to extract effect sizes. All nine studies met these criteria. We used a plot digitizer (https://automeris.io/WebPlotDigitizer/index.html) to extract the information needed from descriptive data plots for three studies.^20,29,30

Statistical analysis. We used a random-effects model with restricted maximum likelihood estimation (REML) to measure the effect of entrainment on pain perception. We compared the standardised mean difference between the experimental (entrainment) and control treatment, using Hedges’ g as the effect size measure. We report I² and τ² as measures of between-study heterogeneity. For eight of the nine studies, we were able to explore the effect of entrainment on pain perception in different conditions, giving rise to 26 effect sizes. Additionally and for exploratory purposes, we used subgroup analysis for different outcomes measures including (1) fentanyl requirements during surgery; (2) perceived pain intensity based on VAS or numeric rating scale; and (3) Event-Related Potentials (ERPs) time-locked to laser-evoked pain. Finally, we present funnel plots with standard error of effect sizes ³¹ and the Egger regression test ³² to rule out possible publication bias. The meta-analysis was conducted using the meta (version 5.0–1 ³³ ) and metasens (version 1.0–1 ³³ ) packages for R Studio (version 1.4.1106). The data and script for all analyses are available at the Open Science Framework (osf.io/s64vj) or upon request to the authors.

Critical appraisal

Included studies were appraised using the Critical Appraisal Skills Programme checklist for randomised controlled trials, which assesses the methodological quality of the study and the strength of results based on randomisation, blinding, sample size, baseline characteristics, reporting of results and value of the results. There were no ethical issues identified in conducting this review; therefore, no applications for approval were sought.

Study characteristics

Six studies investigated binaural beats as the entrainment method.^28,34–38 Two studies investigated LED goggles/visual stimulation^20,30 and one study investigated both binaural beats and LED goggles. ³⁹

Four studies investigated alpha frequencies.^20,30,38,39 Of these, two investigated 8-, 10- and 12 Hz stimulation frequencies^20,39 and two investigated 10 Hz stimulation frequency.^30,38 Two studies investigated theta frequency entrainment.^28,37 Of these, one investigated 5 Hz frequency stimulation ³⁷ and the other investigated 6 Hz stimulation. ²⁸ One study investigated alpha (10 Hz) and theta (7 Hz) visual stimulation, but the authors considered theta stimulation as an additional control along with a 1 Hz frequency stimulation (see Table 3). ³⁰ The three remaining studies investigated Hemi-Sync® (i.e. hemispheric synchronisation ⁴⁰ ) as the stimulation method.^34–36 This method present different sounds including binaural beats, but the frequencies contained in the hemi-sync sounds were not reported.

Four studies were undertaken in the United Kingdom,^20,30,34,39 three in the United States, ^28,35,36 one in Greece ³⁷ and one in Turkey. ³⁸ Four of the nine studies investigated patients undergoing surgery,^34–36,38 three involved volunteers with chronic pain,^28,30,37 and two involved healthy volunteers.^20,39

Three studies investigated analgesia requirements during general anaesthesia as their primary outcome.^34–36 Fentanyl requirements during surgery in micrograms ³⁴ and micrograms/kilogram/minute ³⁵ and intravenous morphine equivalent in milligram ³⁶ were used as the outcome measures. Fentanyl was administered intravenously if heart rate and/or blood pressure increased more than a fixed percentage of pre-operative baseline (e.g. 15%–20%) until these measurements returned to baseline. The dosage was used as a proxy of nociceptive control.

Four studies investigated pain intensity following entrainment.^28,37–39 Two studies investigated pain intensity following entrainment as a secondary outcome.^30,36 Two studies measured pain using a visual analogue scale.^36,38 Three studies measured pain using a numeric rating scale.^30,37,39 One study measured pain using the West Haven-Yale Multidimensional Pain Inventory. ²⁸ Two studies investigated alpha power following entrainment,^20,30 one of which also investigated changes to the amplitude of the N2 and P2 components of laser-evoked potentials in EEG data. ²⁰ The amplitude of these components are correlated with subjective pain intensity.

Outcome of meta-analysis

Figure 2 summarises the results of the meta-analysis in a forest plot. Overall, sensory entrainment decreased pain perception (effect size: mean = −0.69, 95% confidence interval = [−0.91–0.46]). As indicated by the subgroup analysis, this decrease was found for pain intensity rating;^{28,30,36–39} and fentanyl requirement during surgery,^34–37 but not ERP response to laser-evoked pain. ³⁰ A χ² test showed no significant subgroup effect under the random-effects model (χ² = 4.39, p = 0.11; but there was one under the fixed-effect model, χ² = 8.82, p = 0.01). There are indications that some studies or conditions may not show a change in pain perception with sensory entrainment compared to a control treatment. Arendsen et al. ³⁰ did not find a decrease in pain intensity rating while participants were standing after receiving visual alpha entrainment or sitting after receiving visual theta entrainment relative to 1 Hz visual stimulation. Lewis et al. ³⁵ did not observe any differences between listening to binaural beats and listening to a blank tape (silence) for patients undergoing lumbar disk surgery.OPEN IN VIEWER

Although there are entrainment effects, the heterogeneity was generally large. This is the case across all studies (I² = 73% and τ² = 0.25, p < 0.01), and for the fentanyl (I² = 79% and τ² = 0.46, p < 0.01) and intensity (I² = 72% and τ² = 0.22, p < 0.01) subgroups. There were too few effect sizes for the ERP subgroup (see Figure 2). Finally, the funnel plot in Figure 3 indicates that there was no evidence of publication bias (Egger test: t(24) = −0.39, p = 0.70).OPEN IN VIEWER

Critical appraisal

The included studies adopted parallel group randomised controlled trials^34–38 and repeated measures crossover design.^28–30,39 The studies were generally of high quality and were therefore included in the meta-analysis. The most common limitation was the lack of detail around blinding of participants and researchers.^{28–30,34,38,39}

Effectiveness of alpha (8–13 Hz) neural entrainment in acute pain

A key finding of our review suggests that neural entrainment of alpha frequencies by visual and auditory stimulation methods was effective in reducing acute pain. In particular, visual or auditory entrainment at 8, 10 or 12 Hz resulted in lower pain intensity ratings for acute pain induced by laser stimulation or during surgery relative to a control sensory stimulation.^38,39 These results complement EEG studies which found a reduction of alpha power during painful stimulation.^6,9,42,43 That is in line with our hypothesis, we expect that increasing power in alpha frequencies by sensory entrainment would lead to a corresponding reduction in pain intensity.

The subgroup analysis of visual alpha entrainment on the ERP response revealed no consistent evidence for an entrainment effect when pooled over the P1 and P2 components. ²⁰ Interestingly, Ecsy et al. ²⁰ found that the 10 Hz visual entrainment was shown to result in a reduction of the P2 peak amplitude, which may reflect cortical responses to nociception. ⁴⁴ This was observed in the insular cortex where the magnitude of perceived pain from a noxious stimulus is believed to be encoded, ⁴⁵ as well as in the posterior cingulate cortex and the precuneus. Reductions in P2 peak amplitudes in these brain structures indicate that they may be involved in alpha entrainment induced pain suppression, possibly via a mechanism in which increased alpha power leads to a change in attention towards the painful stimulation. Indeed, the reduction in P2 amplitude is consistent with evidence of alpha oscillations involvement in top-down inhibition of activity in sensory networks^9,46 and the correlation between an increase in alpha power and deactivation in cortical areas.^19,47 We note that ERP responses may reflect attentional capture by nociceptive stimulation (i.e. stimulus saliency) rather than perceived pain intensity. ⁴⁸ Thus, more research is needed to determine whether ERP response can be used as an outcome measure for pain perception, for example, by investigating different components of the ERP response.

It can be postulated that alpha entrainment leads to the inhibition or deactivation of cortical areas involved in pain processing and a subsequent reduction in pain perception. Alpha oscillations are also involved in an attentional suppression mechanism. ⁴⁹ Attention is known to influence pain perception and pain is perceived as more intense when focused on. ⁵⁰ The increase in alpha power following entrainment could suppress attention towards the painful stimulus which is then perceived as less intense.

In contrast to acute pain, the findings by Arensden et al. ³⁰ suggest that 10 Hz visual entrainment may not be effective in reducing pain intensity in patients with chronic musculoskeletal pain. This study may be underpowered (see ⁴¹ ). It may also be possible that chronic-pain-induced neural changes can influence the efficacy of alpha entrainment in reducing pain.^45,51 Moreover, changes in the oscillatory activity observed in chronic-pain patients such as increases in theta oscillations^12,52 may also contribute to the lack of efficacy observed by Arensden et al. ³⁰

Effectiveness of theta (4–7 Hz) neural entrainment in chronic pain

There was no consistent evidence that visual entrainment reduced pain intensity for chronic pain. ³⁰ By comparison, auditory theta entrainment was shown to be effective in reducing chronic pain.^28,37 The results from these studies are surprising as chronic pain has been characterised by an increase in theta power due to altered thalamo-cortical activity.^12,53 Moreover, in participants with no chronic-pain conditions, painful laser stimulation resulted in significantly higher theta power than intensity-matched innocuous tactile or electrical stimulation.^11,54 As entrainment increases theta power, it would not be expected to reduce pain intensity ratings for those with chronic pain. Gkolias et al. ³⁷ speculated that a more distributed network of brain structures entrained to theta frequency might enhance overall neural inhibition that is reduced during thalamo-cortical dysrhythmia resulting in pain reduction. The reduction in pain intensity following theta entrainment despite the observations of increased theta activity in acute and chronic pain demonstrates the need for a better understanding of the neurological processes associated with chronic pain.

Importantly from a clinical perspective, a significant reduction in analgesic medication use was observed when theta entrainment was used on demand over several days.^28,37 This is an important finding given the lack of evidence for the efficacy of long-term use of medications as well as potential adverse effects associated with them.^1,2,55 Three studies used the Hemi-Sync® method ⁴⁰ to present different sounds including binaural beats, but the specific sounds and frequencies were not reported.^34–36 That said, listening to hemi-sync sounds reduced post-operative pain ratings and time until discharge following administration 30 min before surgery and during general anaesthesia. ³⁶ Listening to hemi-sync sounds also decreased the amount of fentanyl patients needed during surgery.^34–36 These studies were also analysed by Peng and Tang ²⁵ who focused on fentanyl requirements and did not consider pain intensity, nor did the researchers consider other sensory modalities. The reduction in analgesic requirements would improve patient safety and so this, along with the pain reductions, demonstrate the potential clinical usefulness of sensory theta entrainment.

Limitations of the evidence

Despite the potential for clinical use of sensory neural entrainment identified in this review, there are several limitations which may influence the conclusions drawn from our study. First, the number of studies included in this study is low and the evidence base is limited. A greater number of studies will be necessary to provide a more robust evaluation of the practical application of sensory entrainment specifically on pain perception.

Secondly, all of the studies with the exception of one, had a modest sample size (<80). Further studies in this area should include a larger sample size to ensure studies are appropriately powered and provide reliable data. A larger sample size may also allow assessment of how psycho-social factors affect the efficacy of sensory entrainment methods in clinical settings in which patients have a large variability of such factors. ⁵⁶ Given this variability in patient cohorts, small sample sizes may be more problematic for clinical studies. For example, Arendsen et al.’s ³⁰ study may have been underpowered to detect a consistent effect size of sensory entrainment for patients with chronic musculoskeletal pain (see also ⁴¹ ).

Lastly, no studies directly compared the effect of sensory entrainment on healthy individuals with individuals suffering from chronic pain. Evidence suggests that chronic pain has an impact on brain function and structure.^45,51 While this review included studies investigating chronic-pain patients and healthy volunteers, there was no study identified which included both groups.