Self-supervised learning of accelerometer data provides new insights for sleep and its association with mortality.
Journal
NPJ digital medicine
ISSN: 2398-6352
Titre abrégé: NPJ Digit Med
Pays: England
ID NLM: 101731738
Informations de publication
Date de publication:
20 May 2024
20 May 2024
Historique:
received:
18
08
2023
accepted:
22
02
2024
medline:
21
5
2024
pubmed:
21
5
2024
entrez:
20
5
2024
Statut:
epublish
Résumé
Sleep is essential to life. Accurate measurement and classification of sleep/wake and sleep stages is important in clinical studies for sleep disorder diagnoses and in the interpretation of data from consumer devices for monitoring physical and mental well-being. Existing non-polysomnography sleep classification techniques mainly rely on heuristic methods developed in relatively small cohorts. Thus, we aimed to establish the accuracy of wrist-worn accelerometers for sleep stage classification and subsequently describe the association between sleep duration and efficiency (proportion of total time asleep when in bed) with mortality outcomes. We developed a self-supervised deep neural network for sleep stage classification using concurrent laboratory-based polysomnography and accelerometry. After exclusion, 1448 participant nights of data were used for training. The difference between polysomnography and the model classifications on the external validation was 34.7 min (95% limits of agreement (LoA): -37.8-107.2 min) for total sleep duration, 2.6 min for REM duration (95% LoA: -68.4-73.4 min) and 32.1 min (95% LoA: -54.4-118.5 min) for NREM duration. The sleep classifier was deployed in the UK Biobank with 100,000 participants to study the association of sleep duration and sleep efficiency with all-cause mortality. Among 66,214 UK Biobank participants, 1642 mortality events were observed. Short sleepers (<6 h) had a higher risk of mortality compared to participants with normal sleep duration of 6-7.9 h, regardless of whether they had low sleep efficiency (Hazard ratios (HRs): 1.58; 95% confidence intervals (CIs): 1.19-2.11) or high sleep efficiency (HRs: 1.45; 95% CIs: 1.16-1.81). Deep-learning-based sleep classification using accelerometers has a fair to moderate agreement with polysomnography. Our findings suggest that having short overnight sleep confers mortality risk irrespective of sleep continuity.
Identifiants
pubmed: 38769347
doi: 10.1038/s41746-024-01065-0
pii: 10.1038/s41746-024-01065-0
doi:
Types de publication
Journal Article
Langues
eng
Pagination
86Subventions
Organisme : Wellcome Trust (Wellcome)
ID : 223100/Z/21/Z
Organisme : British Heart Foundation (BHF)
ID : RE/18/3/34214
Organisme : DH | National Institute for Health Research (NIHR)
ID : NIHR203316
Organisme : Wellcome Trust (Wellcome)
ID : 227093/Z/23/Z
Informations de copyright
© 2024. The Author(s).
Références
Meyer, N., Harvey, A. G., Lockley, S. W. & Dijk, D.-J. Circadian rhythms and disorders of the timing of sleep. Lancet 400.10357, 1061–1078 (2022).
Ferrie, J. E., Kumari, M., Salo, P., Singh-Manoux, A. & Kivimäki, M. Sleep epidemiology-a rapidly growing field. Int. J. Epidemiol. 40, 1431–1437 (2011).
doi: 10.1093/ije/dyr203
pubmed: 22158659
Short, M. A., Gradisar, M., Lack, L. C., Wright, H. & Carskadon, M. A. The discrepancy between actigraphic and sleep diary measures of sleep in adolescents. Sleep Med. 13, 378–384 (2012).
doi: 10.1016/j.sleep.2011.11.005
pubmed: 22437142
Wainberg, M. et al. Association of accelerometer-derived sleep measures with lifetime psychiatric diagnoses: a cross-sectional study of 89,205 participants from the UK Biobank. PLoS Med. 18, e1003782 (2021).
doi: 10.1371/journal.pmed.1003782
pubmed: 34637446
pmcid: 8509859
Mantua, J., Gravel, N. & Spencer, R. Reliability of sleep measures from four personal health monitoring devices compared to research-based actigraphy and polysomnography. Sensors 16, 646 (2016).
doi: 10.3390/s16050646
pubmed: 27164110
pmcid: 4883337
Boe, A. J. et al. Automating sleep stage classification using wireless, wearable sensors. NPJ Digit. Med. 2, 1–9 (2019).
doi: 10.1038/s41746-019-0210-1
Devine, J. K., Chinoy, E. D., Markwald, R. R., Schwartz, L. P. & Hursh, S. R. Validation of Zulu watch against polysomnography and actigraphy for on-wrist sleep-wake determination and sleep-depth estimation. Sensors 21, 76 (2020).
doi: 10.3390/s21010076
pubmed: 33375557
pmcid: 7796293
Patterson, M. R. et al. 40 years of actigraphy in sleep medicine and current state of the art algorithms. NPJ Digit. Med. 6, 51 (2023).
doi: 10.1038/s41746-023-00802-1
pubmed: 36964203
pmcid: 10039037
Doherty, A. et al. Gwas identifies 14 loci for device-measured physical activity and sleep duration. Nat. Commun. 9, 1–8 (2018).
doi: 10.1038/s41467-018-07743-4
Jones, S. E. et al. Genetic studies of accelerometer-based sleep measures yield new insights into human sleep behaviour. Nat. Commun. 10, 1–12 (2019).
doi: 10.1038/s41467-019-09576-1
Katori, M., Shi, S., Ode, K. L., Tomita, Y. & Ueda, H. R. The 103,200-arm acceleration dataset in the UK biobank revealed a landscape of human sleep phenotypes. Proc. Natl Acad. Sci. 119, e2116729119 (2022).
doi: 10.1073/pnas.2116729119
pubmed: 35302893
pmcid: 8944865
McHugh, M. L. Interrater reliability: the kappa statistic. Biochem. Med. 22, 276–282 (2012).
doi: 10.11613/BM.2012.031
Sundararajan, K. et al. Sleep classification from wrist-worn accelerometer data using random forests. Sci. Rep. 11, 1–10 (2021).
doi: 10.1038/s41598-020-79217-x
Trevenen, M. L., Turlach, B. A., Eastwood, P. R., Straker, L. M. & Murray, K. Using hidden Markov models with raw, triaxial wrist accelerometry data to determine sleep stages. Aust. N. Z. J. Stat. 61, 273–298 (2019).
doi: 10.1111/anzs.12270
Yin, J. et al. Relationship of sleep duration with all-cause mortality and cardiovascular events: a systematic review and dose-response meta-analysis of prospective cohort studies. J. Am. Heart Assoc. 6, e005947 (2017).
doi: 10.1161/JAHA.117.005947
pubmed: 28889101
pmcid: 5634263
Itani, O., Jike, M., Watanabe, N. & Kaneita, Y. Short sleep duration and health outcomes: a systematic review, meta-analysis, and meta-regression. Sleep Med. 32, 246–256 (2017).
doi: 10.1016/j.sleep.2016.08.006
pubmed: 27743803
Liang, Y. Y. et al. Joint associations of device-measured sleep duration and efficiency with all-cause and cause-specific mortality: a prospective cohort study of 90 398 UK Biobank participants. J. Gerontol. A Biol. Sci. Med. Sci. 78, 1717–1724(2023).
Wang, C. et al. Association of estimated sleep duration and naps with mortality and cardiovascular events: a study of 116,632 people from 21 countries. Eur. Heart J. 40, 1620–1629 (2019).
doi: 10.1093/eurheartj/ehy695
pubmed: 30517670
Taheri, S. Sleep and cardiometabolic health-not so strange bedfellows. Lancet Diabetes Endocrinol. 11.8, 532–534 (2023).
Golbus, J. R., Pescatore, N. A., Nallamothu, B. K., Shah, N. & Kheterpal, S. Wearable device signals and home blood pressure data across age, sex, race, ethnicity, and clinical phenotypes in the michigan predictive activity & clinical trajectories in health (mipact) study: a prospective, community-based observational study.Lancet Digit. Health 3, e707–e715 (2021).
doi: 10.1016/S2589-7500(21)00138-2
pubmed: 34711377
Agnew Jr, H., Webb, W. B. & Williams, R. L. The first night effect: an EEG study of sleep. Psychophysiology 2, 263–266 (1966).
doi: 10.1111/j.1469-8986.1966.tb02650.x
pubmed: 5903579
Doherty, A. et al. Large scale population assessment of physical activity using wrist worn accelerometers: the UK Biobank study. PloS One 12, e0169649 (2017).
doi: 10.1371/journal.pone.0169649
pubmed: 28146576
pmcid: 5287488
Straker, L. et al. Cohort profile: the Western Australian pregnancy cohort (Raine) study–generation 2. Int. J. Epidemiol. 46, 1384–1385j (2017).
pubmed: 28064197
pmcid: 5837608
Dontje, M. L., Eastwood, P. & Straker, L. Western Australian pregnancy cohort (Raine) study: generation 1. BMJ Open 9, e026276 (2019).
doi: 10.1136/bmjopen-2018-026276
pubmed: 31138581
pmcid: 6549642
van Hees, V., Charman, S. & Anderson, K. Newcastle polysomnography and accelerometer data https://doi.org/10.5281/zenodo.1160410 (2018).
Plekhanova, T. et al. Validation of an automated sleep detection algorithm using data from multiple accelerometer brands. J. Sleep Res. 32.3, e13760 (2023).
Byrne, E. M., Gehrman, P. R., Trzaskowski, M., Tiemeier, H. & Pack, A. I. Genetic correlation analysis suggests association between increased self-reported sleep duration in adults and schizophrenia and type 2 diabetes. Sleep 39, 1853–1857 (2016).
doi: 10.5665/sleep.6168
pubmed: 27397570
pmcid: 5020367
Migueles, J. H. et al. Equivalency of four research-grade movement sensors to assess movement behaviors and its implications for population surveillance. Sci. Rep. 12, 1–9 (2022).
doi: 10.1038/s41598-022-09469-2
Willetts, M., Hollowell, S., Aslett, L., Holmes, C. & Doherty, A. Statistical machine learning of sleep and physical activity phenotypes from sensor data in 96,220 UK biobank participants. Sci. Rep. 8, 1–10 (2018).
doi: 10.1038/s41598-018-26174-1
Walmsley, R. et al. Reallocation of time between device-measured movement behaviours and risk of incident cardiovascular disease. Br. J. Sports Med. 56, 1008–1017 (2022).
doi: 10.1136/bjsports-2021-104050
Berry, R. B. et al. Rules for scoring respiratory events in sleep: update of the 2007 AASM manual for the scoring of sleep and associated events: deliberations of the sleep apnea definitions task force of the American Academy of Sleep Medicine. J. Clin. Sleep Med. 8, 597–619 (2012).
doi: 10.5664/jcsm.2172
pubmed: 23066376
pmcid: 3459210
He, K., Zhang, X., Ren, S. & Sun, J. Identity mappings in deep residual networks. in European Conference on Computer Vision, 630–645 (Springer, 2016).
Huang, Z., Xu, W. & Yu, K. Bidirectional LSTM-CRF models for sequence tagging. arXiv preprint arXiv:1508.01991 (2015).
Yuan, H. et al. Self-supervised learning for human activity recognition using 700,000 person-days of wearable data. npj Digit. Med. (in the press).
Creagh, A. P. et al. Digital health technologies and machine learning augment patient reported outcomes to remotely characterise rheumatoid arthritis. npj Digit. Med. 7, 33 (2024).
van Hees, V. T. et al. Estimating sleep parameters using an accelerometer without sleep diary. Sci. Rep. 8, 12975 (2018).
doi: 10.1038/s41598-018-31266-z
pubmed: 30154500
pmcid: 6113241