Stimulating Brain Waves During Deep Sleep

Overview

This study aims to better delineate profiles of insomnia subtypes in people with and without depression or PTSD across simultaneous EEG, heart rate, and body temperature monitoring over multiple days in the natural sleeping environment. Using ambulatory EEG headbands, we also aim to compare the influence of auditory stimulation on slow waves and related objective and subjective sleep measures, as well as mental well-being, daytime fatigue, and cognitive performance in healthy sleepers and people with symptoms of insomnia (with and without psychiatric comorbidities). This is a double-blind randomized control trial. The overall protocol includes a web-based screening interview and home-based data collection spanning over 5 weeks. A subset of participants will be invited for in-lab monitoring via 3 overnight visits.

1. Background and Rationale Insomnia is estimated to be the most prevalent sleep disorder, with nearly 10% of the general population meeting diagnostic criteria. Extensive work is still being deployed to better delineate the multiple subtypes of insomnia, and major gains are to be made using multi-domain physiological monitoring to deepen our understanding of this heterogeneous disease. Some individuals with insomnia show low levels of slow wave sleep (SWS). Novel devices which can be used in the field over multiple days now enable longer term monitoring of EEG features such as sleep slow waves to assess not only inter-individual variability, but also intra-individual variability, a phenomenon of high importance to assess the stability of slow wave abnormalities in insomnia subtypes. Recent findings show that carefully timed auditory stimulation can increase sleep slow waves. Whether such auditory stimulation can potentiate slow waves in people with insomnia symptoms remains to be tested. Furthermore, sleep architecture and quantitative EEG measures undergo significant alterations in many people with insomnia. Insomnia is notably linked to prolonged sleep onset latency, short total sleep time, numerous nocturnal awakenings, poor sleep efficiency and low amounts of SWS. For instance, it is estimated that people with chronic insomnia spend 20 minutes less in SWS per night compared to good sleepers. Individuals with insomnia have also been reported to have less slow wave activity (SWA; lower spectral power in the delta frequency range), with a shift towards elevated high frequency EEG activity, a potential physiological marker of increased cortical arousal. SWS and SWA are thought to be reflective of sleep depth and to play a central role in sleep maintenance and restorative sleep functions. SWS/SWA alteration may thus represent one of the factors contributing to the lower awakening threshold of people with insomnia and to some of their daytime symptoms, including cognitive difficulties and significant daytime fatigue. Of note, it has been proposed that SWS abnormalities in insomnia may result in part from dysfunctions in thermoregulatory processes, especially around sleep onset. Autonomic alterations are also visible through elevated heart rate and abnormally low heart rate variability in people with insomnia. Yet very little is known about how daily changes in body temperature and heart rate are related to SWS/SWA dynamics in people with insomnia. There is thus a need for simultaneous ambulatory monitoring across potentially interacting physiological systems. Enhancement of Slow Oscillations Recent studies demonstrated that slow oscillations (SO), a SWA component in lower frequencies (\<1Hz), can also be increased with auditory stimulation. For instance, 50ms pink noise pulses administered during SWS, specifically during the ascending phase of the SO detected in the frontal cortex, increased SO activity. The ascending phase of the SO reflects the depolarization of the cell membrane and increased excitability of the neural network. Further studies confirmed these findings, notably reporting that this may translate across the SWA spectrum. EEG headbands are ambulatory devices for EEG recording capable of stimulation that is phase-locked to SOs. A recent double-blind randomized sham-controlled study using a beta version of these devices in health adults reported a 0.70 specificity and 0.90 sensitivity to automatically detect N3 sleep (as compared to expert human scoring) and rather effective real-time closed-loop detection and stimulation in the SO ascending phase. This yielded an 11-43% increase in spectral power within the delta frequency band during the 4 seconds following stimulations. Importantly, this effect persisted over a sustained period of 10 consecutive nights of stimulation, suggesting no major sign of desensitization over this period. Whether such auditory stimulation can also potentiate slow waves in people with insomnia symptoms remains to be tested. Potential Modulators of Insomnia Symptoms and Slow wave sleep Insomnia is not a homogeneous syndrome and can be characterized by a range of different physiological and psychological profiles. For instance, evidence suggests that people with insomnia who underestimate their sleep durations ("subjective" insomnia subtype) have more SWA reductions than those with objective short sleep durations ("objective" insomnia subtype). In addition, SWS potentiation may not have noticeable impacts in individuals with insomnia with more prominent sleep initiation difficulties (initial insomnia), and there may be little effects to those with early morning awakenings (terminal insomnia) since SWS occurs mostly in the first portion of the night. Importantly, there are indications that the symptomatic profiles defining some of these insomnia subtypes may vary across time. While this type of work was mostly based on subjective self-reported measures, few studies have reported data about the stability of objective sleep EEG and other physiological features across insomnia subtypes, reporting relative short term stability but possible inter-subtype variability. This limits our understanding of the multiple pathophysiological factors involved in the maintenance of insomnia, with central implications for the daily management of this chronic illness. New wearable technologies are creating an unprecedented opportunity to fill in this knowledge gap. Slow Oscillations in Insomnia and Comorbid Mental Disorders Insomnia is highly comorbid with mental disorders. In fact, studies have found that those suffering from sleep disturbances are more psychologically aroused, perceive their lives as more stressful, and use emotion-based coping strategies more than do good sleepers. Since there are reports of SWS/SWA dysregulation in people with depression or post-traumatic stress disorder (PTSD), there may be some interactions between these mental disorders and insomnia on these sleep EEG features. One common feature of insomnia, depression and PTSD is the deregulation of the stress system leading to increased autonomic activation. This is notably reflected by elevated blood pressure and heart rate, reduced heart rate variability, faster breathing rate, increased skin conductance and abnormal temperature regulation. Possible Benefits of Slow Oscillation Potentiation for Cognitive Functions SWS/SWA and SO are known to play an important role in cognition, notably in terms of memory functions. For instance, low SWS has been shown to disrupt memory consolidation. It has been shown that SWS enhancement during sleep can facilitate declarative memory consolidation. Considering the memory difficulties commonly reported by individuals with insomnia and the observed alterations in sleep-related memory consolidation in this population, there is a need to investigate whether SO potentiation during sleep may impact daily cognitive performance in people with and without insomnia. 2. Study Objectives A) To better delineate profiles of insomnia subtypes in people with and without depression or PTSD across simultaneous EEG, heart rate, skin conductance, and body temperature monitoring over multiple days in the natural sleeping environment. B) To compare the influence of auditory stimulation on SO and related objective and subjective sleep measures, as well as mental well-being, daytime fatigue, and cognitive performance in healthy sleepers and people with symptoms of insomnia (with and without psychiatric comorbidities). C) To run exploratory analyses to determine if individual SO sensitivity to sound stimulation during sleep relates to auditory event-related potentials in the wake state. 3. Study Design 3.1 General Overview This study integrates 2 design frameworks: a) observational design (Aim A), and b) crossover design (Aims B and C). A crossover design was chosen to reduce the influence of confounding covariates because each participant serves as her/his own control. Optimal crossover designs are statistically efficient and require fewer participants than cross-sectional designs. The overall protocol includes a web-based screening interview and home-based data collection spanning over 5 weeks. A subset of participants will be invited for in-lab monitoring via 3 overnight visits. 3.2 Recruitment, initial consent and screening Participants will be recruited from 3 sites: The Royal Ottawa Mental Health Centre (ROMHC), the Insomnia Clinic of the Center for Advanced Research in Sleep Medicine, and the CERVO Brain Research Centre. Recruitment will also be achieved through institutional 'Permission to contact" registries, as well as through external recruitment with advertisements distributed via web-based platforms (e.g., institutional websites (e.g. www.theroyal.ca), sleeponitcanada.ca, the eMentalHealth Research Study Directory (www.ementalhealth.ca), related web Kijiji, Facebook and twitter). 3.2.1 Contact pathways: A. "Clinician referral" pathway Clinicians from any of the study sites identify patients they would like to refer to the study, get the patients to fill out the institution-specific "Permission to be contacted form", and send this form to the research team. B. "Consent to be contacted" pathway ROMHC programs and external collaborators have an existing list of people who consented to be contacted for research. These individuals will be contacted by the research team by phone or email. C. "Direct contact" pathway Potential participants reach out directly to the research team after seeing advertisements or hearing about the study by word of mouth. For all pathways, these procedures will be followed: 1. A research assistant will contact potential participants by email to send them a summary of the study and a link to complete an anonymous online screening form for the first set of eligibility criteria on a secured web platform (Qualtrics, www.qualtrics.com; license held via the University of Ottawa) and a unique identification code. The first question of this online form will be asking participants to confirm they consent to take part to the screening process. This will be documented in the screening database. The second question of this online screening form will ask participants to insert their unique anonymous research code. This screening form will include the WHO-ASSIST item 2, Sleep Disorders Symptoms Checklist-25, Sleep Condition Indicator (SCI), Primary Care PTSD Screen (PC-PTSD), Patient Health Questionnaire (PHQ-9) and should take between 10 and 15 minutes to complete. 2. If the questionnaires indicate that the first set of eligibility criteria are met, the participant will be invited for the second and last screening step: a screening phone interview (based on the M.I.N.I. without the suicidality module). This interview should last for about 15-45 minutes. Before the start of the interview, the research assistant will confirm verbal consent to pursue the screening process. This will also be documented in the screening database. If the potential participant does not meet the last set of eligibility criteria assessed via the phone interview, the study will be terminated. 3.2.2 Ongoing Consent Procedure For participants found to be eligible after the screening phone interview, the research assistant will summarize the next steps of the study verbally and address any questions or concerns that participants may have at the end of the phone interview. Special care will be given to ensure that participants understand that their participation is entirely voluntary and that they can withdraw from the study at any stage (without any consequence on the treatment(s) they receive at the specific study site or their relationship with any of the clinicians at that study site). Eligible participants will be sent an electronic form containing the information and consent form (ICF) and will have as much time as they need to read it at home before enrolling in the study. As part of that same electronic form, they will confirm consent by ticking consent boxes and entering their full name. This information will be stored in a consent database showing the participant's full name and consent date, which will be kept on restricted access folders on the ROMHC's server in separate files than the research data. This online consent form was created based on published guidelines from peer-reviewed articles. Ongoing consent will also be obtained electronically at the follow-up time points. 3.3 Data collection and management 4.3.1. Baseline assessment At the end of the screening interview and after obtaining verbal consent to enroll in the study, eligible participants will be asked additional questions about the course of their mental disorders, including age at onset and duration of longest episode (inspired from the Diagnostic interview for genetic studies), as well as the proportion of time with PTSD and/or depressive symptoms over the past 1, 5 and 10 years. Participants will be given a package with the equipment to start ambulatory measurements. Until the pandemic situation resolves, participants will have the option to come to the ROMHC for curbside pick-up in strict conditions respecting physical distancing (i.e. both participant and staff will be masked, staff will put the package directly in the trunk without coming in close contact with the participant) or to get a courier delivery with all mail and mail back costs covered by the research team. The Royal Logo or name will not appear on the package. Participants will complete 1 week of monitoring at home using the EEG headband device (DREEM2, Dreem, Paris, France; used for monitoring purpose only, no stimulation at baseline), an elastic chest belt to measure heart rate (Zephyr Bioharness), and a small sensor fixed on the abdomen using hypo-allergic tape to monitor skin temperature (i-button). During this period, participants will fill a sleep log (containing the core elements of the ''Consensus Sleep Diary''). Participants will be instructed on how to use this equipment on brief pre-recorded online videos. They will be asked to liaise with the research team if they have any questions. On the seventh (last) day of the ambulatory monitoring period, participants will also be asked to fill out a battery of questionnaires through a secure web-based platform (Qualtrics) using an anonymous research code. This will take about 1 hour to complete and will inquire about general demographic and clinical information, sleep, daytime functioning, mental health. They will also complete a 40-60 minute online neuropsychological battery (Cambridge Brain Sciences; https://www.cambridgebrainsciences.com/science/tasks). They will also complete two 18-minute resting state EEG recordings: i) 30 minutes before sleep on the evening of the seventh day, and ii) 30 minutes after waking up on the following morning. As part of the consent process, participants who are clients from the ROMHC, the Insomnia Clinic of the Center for Advanced Research in Sleep Medicine, or the CERVO Brain Research Centre will have the option of granting the research team access to their medical chart to document diagnostic information and data from their clinical sleep assessments (results from their questionnaires, overnight sleep study data, diagnoses and treatment plan) were available. 3.2.2. Experimental conditions All participants will take part in two experimental conditions separated by one week in a randomized order: seven days of ambulatory monitoring with auditory stimulation, seven days of ambulatory monitoring without auditory stimulation. Participants, care providers, and all researchers involved in data collection or processing will be blinded to the order of these experimental conditions. On the second night of the ambulatory monitoring period, participants will be asked to complete a 15-30 minute online verbal learning and memory test (paired associates learning; Ngo et al, 2013). They will also complete a second 10-15 minute memory recall testing to assess sleep-related memory consolidation 30 minutes after waking up on the following morning. They will complete this on the second day of both intervention arms with alternate forms of the test. On the final day of each experimental condition and before the start of the second experimental condition, participants will be asked to fill out a subset of the online questionnaires about sleep, daytime functioning, and mental health which should take about 40 minutes to complete. They will also complete a 10-minute resting state EEG recording 30 minutes before their habitual bedtime and after their habitual wake times (to allow for the dissipation of morning sleep inertia (56)). The randomization process will consist of a computer-generated random listing of the two potential sequential orders of the experimental conditions, stratified by sex and diagnostic group. This will be managed by a research collaborator who will not have any contact with the study participants and who will not be directly involved in data processing. This research collaborator will label the placebo and active devices with the participant ID codes and order of use for blinded distribution. Since there are no potential risks for the active condition, we do not expect having to break the blinding codes in this study. In the active experimental condition, participants will complete 1 week of SO stimulation with the EEG headband monitor (DREEM2, Dreem, Paris, France) in the form of auditory stimuli (100ms pink noise pulses, i.e., below the waking threshold as established in prior work (36)) sent on the ascending phase of the SO during N3 sleep (36). In the placebo condition, participants will complete 1 week of EEG headband monitoring without auditory stimulations. Characteristics and robustness of event-related potentials triggered by the auditory stimulations during sleep will be collected, in addition to the spectral power in the SO and delta frequency ranges in the 4 seconds following the initial stimulation. 3.2.3 Follow-up measures One week after the last experimental condition, participants will fill out questionnaires (10 to 20 minutes) to track potential changes in sleep and mental health. 3.2.4. In-lab overnight recordings (when allowed by Public Health Directives) A subgroup of participants (n=40, 10 per group as defined in section 5) will be invited for 3 nights of in-lab standard polysomnography. This will enable to compare the detection of N3 sleep by the ambulatory EEG headband with the gold standard of sleep measurements and to monitor the effects of the stimulation on more extensive brain regions than the ones covered by the ambulatory EEG headband. Prior to the optional overnight in-lab recordings, participants will be asked not to refrain from: * Taking any naps or drinking alcohol on the day of their visit * Having coffee, tea, energy drinks Coke/Pepsi or chocolate (i.e. anything that contains caffeine) as of midday on the day of their visit * Consuming any recreational drugs until the end of their last appointment The first night will be conducted at baseline and will serve as an adaptation night in order to get rid of the "first night effect". At the start of each experimental condition, the same subgroup will be re-invited to the sleep laboratory for a night of standard polysomnography concurrent with active/placebo auditory stimulations delivered with the same device used in the experimental condition. For all these in lab visits, participants will arrive about 2 hours before habitual sleep time and leave about 45 minutes after their usual wake-up time. In addition to the use of an EEG headband device (active or placebo depending on the experimental condition), electrodes will be placed on the scalp according to the 10-20 system with: 11 scalp EEG channels (Fp1-2, F3-4, C3-Z-4, P3-4, O1-2), ground and reference channels, right and left EOG, 2 chin EMG and 2 ECG channels. On the first night only, to document the presence of sleep-related respiratory disorders and periodic leg movements. Respiration will be monitored with an airflow cannula (pressure transducer), nasal-oral thermistor, respiratory effort via thoracic and abdominal respiratory belts, and blood oxygen saturation via an oximeter probe on the finger. On the experimental nights (second and third nights), skin conductance will be continuously recorded using a monitor attached on the finger and body temperature will be monitored with five small sensors (i-buttons) fixed on the skin with hypoallergenic sticky tape (located on the neck, hand, abdomen, thigh, and foot). Before and after the sleep episode, in the wake state, participants will undergo a 10-minute resting state EEG recording and event-related potentials (ERP) recordings. To characterise wake state brain responses to auditory stimuli similar to the stimuli used in the sleep experimental manipulation, ERP recordings will involve: i) a \~20-minute auditory oddball task using deviant stimuli with the same physical properties as the 100 ms pink noise used for sleep-based SO potentiation, and ii) a \~20-minute Fast/Slow task with the same stimuli. The Fast/Slow paradigm counts two conditions, 1) long pink noise presented with fast inter-stimuli intervals (ISI), 2) short pink noise presented with slow ISI. In this paradigm, the slower condition taps into frontal lobe functions while the fast condition does not, therefore enabling the assessment of frontal functions, a cortical region of high relevance for sleep slow waves. All tasks will involve the passive listening of sounds through headphones. Participants will then be asked to go to sleep and be woken up at their habitual wake times. Sleep stage scoring will be manually performed by sleep technologists in accordance with the clinical scoring guidelines established by the American Academy of Sleep Medicine. 3.2.3. Debriefing At the end of the last experimental condition, participants will be asked to fill out questionnaires about expectations surrounding the experimental conditions. Once this is completed, they will receive a debriefing email about the order of the two experimental conditions (sent by the research assistant responsible for randomization to keep other research staff blind).