Chronic pain represents a significant unmet global health need. It is highly prevalent, affecting one-fifth to one-third of adults (1, 2) and represents one of the greatest contributors to disability globally (3). Conventional analgesic medications have poor efficacy and unfavourable side effect profiles when used for chronic pain (4, 5) and many individuals regard their pain as inadequately controlled (6). New paradigms of treatment are required, which is why the interplay between sleep and chronic pain has received increased attention in recent years as a priority area for research (7, 8).

Sleep problems are very common in people living with chronic painful conditions (9, 10), and almost universal in those with fibromyalgia (11). The relationship between pain and sleep is bidirectional (12), and is seen to operate on both short (13) and long (14, 15) time scales. Observationally, non-restorative sleep has been found to be a strong independent predictor of new onset widespread pain (16), and experimental sleep deprivation and sleep fragmentation increases pain (17). There is a strong rationale for linking novel approaches to the intractable problem of chronic pain with sleep disturbance, given this close relationship.

Alpha entrainment is a neuromodulatory approach that has the potential to help people living with chronic pain. Entrainment is when an oscillating system becomes synchronised in phase to an external periodic force. Cortical electrical oscillations, or “brainwaves”, demonstrate this phenomenon in response to rhythmic stimuli, which can be sensory or direct electric or magnetic stimulation (18). Alpha entrainment refers to modulation of cortical activity in the alpha band (8–12 Hz). The overall role of the alpha rhythm has been proposed as providing inhibitory gating between brain regions (19), thereby re-routing resources and information to task-relevant areas. It is involved in pain expectation (20), attention to pain (21) and expectation of pain relief (22) and therefore provides a promising avenue for novel treatments, considering its entrainment is technologically achievable through non-invasive sensory stimulation. Alpha entrainment has been found to reduce experimental laser-induced pain in healthy participants (23) including with the electrophysiological correlate of reduced amplitude of the laser evoked potential (24). In laboratory studies with individuals with chronic musculoskeletal pain, 4 min of 10 Hz sensory stimulation has been shown to successfully entrain alpha and decrease pain (25) and the degree of frontal alpha power increase is found to correlate with the reduction in pain (moderate strength correlation, Pearson r 0.33 for pain intensity, 0.40 for pain unpleasantness) (26). To our knowledge alpha entrainment has not previously been used pre-sleep in those with chronic pain, but other forms of non-invasive brainwave entrainment have been investigated for use at night in the home environment in this group. Audio-visual stimulation decreasing in frequency from 8 to 2 Hz was used by participants with chronic pain and insomnia aiming to aid sleep onset. It was shown that entrainment to the stimuli (in this case delta power) was successful (27), and in both younger (28) and older (29) participants who had chronic pain and insomnia improvements were seen in both symptoms. In a pilot randomised controlled trial, adherence to the intervention was 99% and participants found it easy to use (30), indicating that the concept of audio-visual stimulation at night may be acceptable in this population. Outside of chronic pain, alpha entrainment has been used by healthy participants as one part of a stimulation programme aiming to optimise sleep quality (31). In this pilot trial, the first quarter of a 90 min sensory stimulation programme used 8 Hz (lower range of the alpha frequency) stimulation, and subsequently lower frequencies, and found improvements in participant reported sleep quality. Alpha entrainment therefore holds promise for the treatment of the closely linked problems of chronic pain and sleep disturbance. We designed a smartphone application called home-based Brainwave Entrainment Technology (hBET) to deliver 10 Hz alpha stimulation. This is the first time this modality has been used pre-sleep by those with chronic pain.

The aim of this study is to explore the feasibility of pre-sleep use of hBET with individuals with chronic pain and sleep disturbance and give an indication of the effect of the treatment.

This was an uncontrolled feasibility study, comparing pre- and post-intervention measures. There was no randomisation and participants were not blinded to the intervention received. Participants were 28 individuals living with chronic pain and sleep disturbance, recruited from NHS clinics dealing with chronic pain in two regions in the north of England (Leeds and Manchester) and via online publicity materials. Inclusion criteria were: age 18–80, non-cancer pain of over 3 months' duration including nocturnal pain of at least 4/10 (on 0–10 NRS), and self-reported sleep difficulties (defined as trouble falling asleep, difficulty staying asleep, waking up too early, or waking up unrefreshed on 3 or more nights per week during the past month). Individuals were excluded from participating if they had: any seizure disorder, photosensitivity, planned pain intervention during the study period, hearing and vision problems causing inability to use the stimulation, or inability to consent.

Twenty eight individuals (22 female, 6 male) participated and their demographic and background data are shown in Table 1.

To inform the feasibility assessment, data completeness and attrition are displayed in Figure 3.

Overall, participants chose to use Audio hBET on 458 nights and Visual hBET on 187 nights. This gives an adherence rate of 92% in participants for whom diary records were available (hBET used on 645 of 700 available nights). Five participants tried Visual hBET for fewer than 5 nights before choosing to discontinue it and return to Audio. These quickly aborted periods of Visual use are not included in the analysis, leaving 175 nights of Visual hBET use.

Home-based, pre-sleep use of a smartphone programme for alpha entrainment by audio or visual stimulation was feasible for individuals with chronic pain and sleep disturbance. The findings are drawn from domiciliary use of hBET over 4 weeks, in a relatively unselected sample, which gives a level of ecological validity. This study represents the first application of alpha entrainment pre-sleep in those living with chronic pain and sleep disturbance, and the findings will inform the development of further trials in this area.

The hypothesised mechanism of action of hBET is a positive effect on both night time pain and improved sleep, subsequently leading to improved day time pain and related symptoms. Whilst it is well established that pain and sleep interact with each other, the purported inhibitory gating role of alpha could reasonably be expected to modulate both pain perception and sleep onset discretely, but the independent relevance of each in this population requires further study. Caution is needed in the interpretation of the changes in symptoms reported in this uncontrolled study, and conclusions about efficacy cannot be drawn. However, the experiences of users and pattern of response can usefully inform hypothesis generation and the design of future research. There is an indication that the improvements in patient-reported measures of sleep may be of larger magnitude or more readily observed than those of pain. Whilst the BPI pain severity score, which is derived entirely from 0 to 10 NRS pain ratings, did not improve across the whole group, the BPI pain interference score, which takes into account the impact of pain on various activities and functions, did improve. This may reflect different responsiveness of these measures over relatively short time periods in established chronic pain, or reflect the mechanism of action of this neuromodulatory intervention, and is not fully explained by this study. It is also notable that even the improved scores across many domains still remain above clinical thresholds, such as sleep efficiency (remains below 80%), sleep onset latency (remains over 30 min), anxiety and depression scores [remain over the “case” threshold of 8 (46)] and PSQI scores [remain markedly over the threshold of 6 indicating “poor sleepers”(40)]. The responder analysis is exploratory, and does not offer conclusions on effect, but adds weight to the notion that sleep and pain symptomology are acting together, and is consistent with the hypothesised mechanism that targeting both may lead to positively reinforcing benefits to daytime symptoms. It provides a possible approach to pre-specified response definition in future studies.

In this study, diary-reported measures of sleep onset, time and continuity improved whereas actigraphy assessed sleep measures were unchanged. Discrepancy between actigraphy and sleep diary, particularly in chronic pain, has been reported many times and given different interpretations. Systematic review and meta-analysis comparing methods of sleep assessment in chronic pain patients finds the most consistent discrepancy is that actigraphy gives lower estimates of Sleep Onset Latency than diary report, (by 23 min in meta-analysis) (47). This was the case in the current study, with overall 28 min difference found. This has often been referred to as sleep misperception, with the inference that actigraphy provides a superior “objective” measure. Noting that daytime experience plays a large part in how people with chronic pain judge their sleep (48) one interpretation is that sleep diary and actigraphy are measuring different constructs. However, it is also possible that actigraphy may be less accurate in this group. Studies using polysomnography have contested the idea that sleep complaints in fibromyalgia are due to sleep misperception (49) and when systematically reviewed, actigraphy is found to overestimate total sleep time and underestimates sleep onset latency and wake after sleep onset time, compared to polysomnography, in adults with chronic conditions (50). Future studies incorporating electroencephalographic (EEG) monitoring would provide the multiple benefits of more accurate and detailed sleep determination and staging, insight to the validity of actigraphy in this population and direct evaluation of the alpha entrainment effect of this intervention. It is also possible that EEG monitoring could optimise the intervention, either through personalisation to the individual peak alpha, or by using closed-loop stimulation.

Limitations of this study mean it is not possible to conclude that the observed improvements result only from the intervention, as this is an uncontrolled, open-label feasibility study, not designed to determine efficacy. The limitations include the possibility of bias resulting from participant or researcher enthusiasm for the open-label intervention, placebo and secular (time) effects. Participants were aware of the aim of the study and that they were trialling a novel approach to chronic pain which may have appealed and promoted a placebo effect, which is not controlled for. This was mitigated against with transparency from the researchers and study documentation that the effectiveness of the intervention is as yet unknown. A passive effect due to time, such as regression to the mean, is unlikely as the participants had very longstanding symptoms of median 9 years and were not recruited in a way which would enrich the sample with those experiencing acute “flares” of symptoms. Bias in participant reported outcomes was mitigated against by encouraging honest responses. Diary entries may be less susceptible to this bias, being completed iteratively day after day, and it is notable that the findings from diary entries agree with those from baseline and completion questionnaires, although they are generally of smaller effect size. Gender and age were not controlled for at this feasibility stage, but are likely to be relevant covariates in the effect on symptoms, which should be accounted for in future study design. Finally, the study only considers a relatively short time period of four weeks of intervention use, with no longer term follow up. This leaves open the possibility of either a novelty effect exaggerating the benefits, or a more incremental effect being missed due to the short duration of this study. Anecdotal clinical observations tend to suggest that sleep improvements precede improvements in pain, which would require longer a study duration to capture.

In conclusion, hBET appears to be a feasible intervention for the home setting and evaluation under controlled conditions of its clinical effect is warranted.