Acquiring rich longitudinal passive sleep data across childhood and adolescence (8-18yrs)-the AMBIENT sleep study
Lead Research Organisation:
University of Edinburgh
Department Name: Centre for Clinical Brain Sciences
Abstract
Sleep is an essential part of health, allowing individuals to function at their best. It is widely reported that many children and teens do not get enough sleep, putting them at greater risk of various negative health and educational outcomes. However, these links have yet to be fully investigated. For example, during the key developmental stages of childhood and adolescence (8-18years), sleep behaviours change significantly. Despite this, we do not know what constitutes normal changes in sleep patterns, what are potentially problematic, or indicate underlying current or future problems, and what external factors may be influencing changes.
Current data collection methods have various problems. Sleep can be measured in detail using polysomnography, however this is not possible to use this long-term and at-scale. Subjective self-report questionnaire data, or sleep diaries, though possible to conduct at scale, have been shown to be unreliable. Measures derived from research quality sleep-wake watches (actigraphs), are a more objective way of gathering data, however, they are not always well tolerated. Based on our own experience of actigraphy in young people in our award (Sleep, circadian rhythms and mental health in schools (SCRAMS) project), this technique presented two main issues (i) low tolerance, especially in people with hyper touch sensitivity or movement limitation and (ii) limited by battery life and data storage, restricting its period of use. Difficulties are also reported in other studies: both UK Biobank and ABCD (large studies on 10s of thousands of people) administered actigraphy, but only short time durations (a few weeks at most). Such short recordings are not appropriate for understanding sleep behavior in depth, as the effects of weekends, school holidays or seasonal changes are not represented.
During our SCRAMS focus groups, pupils identified other problems with actigraphy: (i) a few days were required to get used to it, (ii) uncomfortable/bulky to wear, (iii) strap sometimes caused dry skin/rash, and (iv) easy to forget to put back on after bathing/swimming. Thus, 18% of our participants did not wear the actigraph for the study duration. Teacher feedback indicated dropping out was higher in pupils from vulnerable groups or with lower academic performance. Similar bias in terms of those who did not complete data collection are reported in other studies like UB Biobank. There is therefore an urgent need for new ways to collect sleep data that are easier to use and over longer periods of time and that can lessen these biases.
We propose to develop and evaluate a new way of collecting sleep data that requires no active participation once installed. We will test a non-wearable sleep tracker that uses radar-based sleep monitoring in the home. Somnofy (VitalThings, Norway) is a commercially available sleep monitor that has been validated against PSG and undergone relevant safety approvals. The system can be used over relatively long periods of time and has the potential to be used at scale. We will optimize methods across the ages 8-18yrs to test feasibility and performance against standard methods of sleep data collection. We will co-produce materials with young people in order that research can be made accessible, engaging, and relevant across the different age ranges and backgrounds. We will invite them to either join our product review group, or our citizen science group, to engage them in the research.
We will therefore develop protocols and analysis approaches that facilitate the use of these new generation of sleep data collection methods at scale in future longitudinal adolescent cohort studies. These methods can then be used to address questions around how patterns change over critical developmental windows, what reflects normal sleep behaviours, what potentially predicts negative outcomes, and what may be protective for sleep health
Current data collection methods have various problems. Sleep can be measured in detail using polysomnography, however this is not possible to use this long-term and at-scale. Subjective self-report questionnaire data, or sleep diaries, though possible to conduct at scale, have been shown to be unreliable. Measures derived from research quality sleep-wake watches (actigraphs), are a more objective way of gathering data, however, they are not always well tolerated. Based on our own experience of actigraphy in young people in our award (Sleep, circadian rhythms and mental health in schools (SCRAMS) project), this technique presented two main issues (i) low tolerance, especially in people with hyper touch sensitivity or movement limitation and (ii) limited by battery life and data storage, restricting its period of use. Difficulties are also reported in other studies: both UK Biobank and ABCD (large studies on 10s of thousands of people) administered actigraphy, but only short time durations (a few weeks at most). Such short recordings are not appropriate for understanding sleep behavior in depth, as the effects of weekends, school holidays or seasonal changes are not represented.
During our SCRAMS focus groups, pupils identified other problems with actigraphy: (i) a few days were required to get used to it, (ii) uncomfortable/bulky to wear, (iii) strap sometimes caused dry skin/rash, and (iv) easy to forget to put back on after bathing/swimming. Thus, 18% of our participants did not wear the actigraph for the study duration. Teacher feedback indicated dropping out was higher in pupils from vulnerable groups or with lower academic performance. Similar bias in terms of those who did not complete data collection are reported in other studies like UB Biobank. There is therefore an urgent need for new ways to collect sleep data that are easier to use and over longer periods of time and that can lessen these biases.
We propose to develop and evaluate a new way of collecting sleep data that requires no active participation once installed. We will test a non-wearable sleep tracker that uses radar-based sleep monitoring in the home. Somnofy (VitalThings, Norway) is a commercially available sleep monitor that has been validated against PSG and undergone relevant safety approvals. The system can be used over relatively long periods of time and has the potential to be used at scale. We will optimize methods across the ages 8-18yrs to test feasibility and performance against standard methods of sleep data collection. We will co-produce materials with young people in order that research can be made accessible, engaging, and relevant across the different age ranges and backgrounds. We will invite them to either join our product review group, or our citizen science group, to engage them in the research.
We will therefore develop protocols and analysis approaches that facilitate the use of these new generation of sleep data collection methods at scale in future longitudinal adolescent cohort studies. These methods can then be used to address questions around how patterns change over critical developmental windows, what reflects normal sleep behaviours, what potentially predicts negative outcomes, and what may be protective for sleep health
Technical Summary
This project will use a co-designed research approach, tailored across this age range, to develop and evaluate an innovative passive sleep data collection method that has potential to be used at scale in future studies of adolescent health. Existing methods of assessing sleep have limitations regarding self-report biases, acceptability and scalability. Here we make use of new-generation of ambient and passive remote data capture methods that are unobtrusive for participants, which can be deployed over long periods (months to years).
A Young Persons Advisory Group (YPAG), N=15, will be used to inform research practice and to develop age-appropriate resources. We will also recruit N=45 young people (8-18years), from existing infrastructure and conduct passive sleep data collection over a two-month period per participant, with simultaneous measures of actigraphy. We will target children from culturally and socially diverse backgrounds through our links with 'SHINE'.
Feasibility will be assessed by examining data quality and by feedback on acceptability. Validity will be assessed by comparison with standard objective approaches of measuring sleep (actigraphy). To engage and share skills with the young people, we will invite them to one of two groups. One group will be 'product reviewers' and the other 'citizen scientists', who will co-develop research around the importance of environmental measures associated with sleep health (e.g. bedroom noise, light).
In addition, we will split the sample into equally sized age-bins of 8-11, 12-14, 15-18 years. Comparisons will then be possible between these age bins in terms of feasibility, acceptability, validity to explore potential differences across the age ranges, important for future work.
Overall, this project has the potential to establish new innovative contactless data collection methods for sleep health across childhood and adolescence (8-18yrs), appropriate to the different ages.
A Young Persons Advisory Group (YPAG), N=15, will be used to inform research practice and to develop age-appropriate resources. We will also recruit N=45 young people (8-18years), from existing infrastructure and conduct passive sleep data collection over a two-month period per participant, with simultaneous measures of actigraphy. We will target children from culturally and socially diverse backgrounds through our links with 'SHINE'.
Feasibility will be assessed by examining data quality and by feedback on acceptability. Validity will be assessed by comparison with standard objective approaches of measuring sleep (actigraphy). To engage and share skills with the young people, we will invite them to one of two groups. One group will be 'product reviewers' and the other 'citizen scientists', who will co-develop research around the importance of environmental measures associated with sleep health (e.g. bedroom noise, light).
In addition, we will split the sample into equally sized age-bins of 8-11, 12-14, 15-18 years. Comparisons will then be possible between these age bins in terms of feasibility, acceptability, validity to explore potential differences across the age ranges, important for future work.
Overall, this project has the potential to establish new innovative contactless data collection methods for sleep health across childhood and adolescence (8-18yrs), appropriate to the different ages.
