Newcastle PSG+Accelerometer study 2015

This data set contains 55 .bin files, 28 .txt files, and one .csv file, which were collected in Newcastle upon Tyne (UK) to evaluate an accelerometer-based algorithm for sleep classification. The data come form a a single night polysomnography recording in 28 sleep clinic patients. A description of the experimental protocol can be found in this open access PLoSONE paper from 2015: https://doi.org/10.1371/journal.pone.0142533.

Polysomnography

Sleep scores derived from polysomnography are stored in the .txt files. Each file represents a time series (one night) of one participant. The resolution of the scoring is 30 seconds. Participants are numbered. The participant number is included in the file names as “mecsleep01_...”. pariticpants_info.csv is a dictionary of participant number, diagnosis, age, and sex.

Accelerometer data

Accelerometer data from brand GENEActiv (https://www.activinsights.com) are stored in .bin files. Per participant two accelerometers were used: One accelerometer on each wrist (left and right). The right wrist from participant 10 is missing, hence the total number of 55 bin files. The tri-axial (three axis) accelerometers were configured to record at 85.7 Hertz. The accelerometer data can be read with R package GENEAread https://cran.r-project.org/web/packages/GENEAread/index.html. Additional information on the accelerometer can be found on the manufacturers product website: https://www.activinsights.com/resources-support/geneactiv/downloads-software/, including a description of the binary file structure on page 27 of this (pdf) file: https://49wvycy00mv416l561vrj345-wpengine.netdna-ssl.com/wp-content/uploads/2014/03/geneactiv_instruction_manual_v1.2.pdf. The participant number and the body side on which the accelerometer is worn are included in the file names as “MECSLEEP01_left wrist...”.

Participant information

The .csv file as included in this dataset contains a dictionary of the participant numbers, sleep disorder diagnosis, participant age at the time of measurement, and sex.

Example processing

The code we used ourselves to process this data can be found in this GitHub repository: https://github.com/wadpac/psg-ncl-acc-spt-detection-eval. Note that we use R package GGIR: https://cran.r-project.org/web/packages/GGIR/, which calls R package GENEAread for reading the binary data.

Performance (mean with 95% CI) of bAHI at PSG-AHI cutoffs of 5, 15, and 30 (all subjects, N = 78).

Performance (mean with 95% CI) of STOP-Bang score of 5 at PSG-AHI cutoffs of 5, 15, and 30 (N = 78).

Dream EEG and Mentation (DREAM) data set

---Data set information---

Common name: Aamodt_evening_sleep Full name: (Optional, unwieldly full name) Authors: Aamodt, A., Sevenius Nilsen, A., Markhus, R., Kusztor, A., HasanzadehMoghadam, F., Kauppi, N., Thürer, B., Storm, J.F. and Juel, B.E. Location: Oslo, Norway Year: 2022 Set ID: [SET BY DATABASE] Amendment: [SET BY DATABASE] Corresponding author ID: [SET BY DATABASE] Download URL: [SET BY DATABASE]

Previous publications: 1) Aamodt, A., Nilsen, A.S., Thürer, B., Moghadam, F.H., Kauppi, N., Juel, B.E. and Storm, J.F. (2021) ‘EEG Signal Diversity Varies With Sleep Stage and Aspects of Dream Experience’, Frontiers in psychology, 12, p. 655884. 2) Aamodt, A., Sevenius Nilsen, A., Markhus, R., Kusztor, A., HasanzadehMoghadam, F., Kauppi, N., Thürer, B., Storm, J.F. and Juel, B.E. (2023) ‘EEG Lempel-Ziv complexity varies with sleep stage, but does not seem to track dream experience’, Frontiers in human neuroscience, 16. Available at:

Correspondence: Bjørn E Juel (b.e.juel@medisin.uio.no)

---Metadata---

Key ID: [SET BY DATABASE] Date entered: [SET BY DATABASE] Number of samples: [INFERRED BY DATABASE] Number of subjects: [INFERRED BY DATABASE] Proportion REM: [INFERRED BY DATABASE] Proportion N1: [INFERRED BY DATABASE] Proportion N2: [INFERRED BY DATABASE] Proportion experience: [INFERRED BY DATABASE] Proportion no-experience: [INFERRED BY DATABASE] Proportion healthy: [INFERRED BY DATABASE] Provoked awakening: (Whether the study protocol instated provoked awakenings; choose Yes, No, or Some) Time of awakening: (General time of day when the awakenings occurred; choose Morning, Day, Evening, Night, or Mixed) Form of response: (The form of response given by subjects, used to infer their experience; choose Free, Structured, Categorical, or Other) Date approved: [SET BY DATABASE]

---How to decode data files---

The files in the PSG directory are named according to the subjectID and the CaseNumber for each awakening. For example the file 10_0103.edf is associated with the third awakening (case 0103) from subject 10.

--Treatment group codes--

No treatment groups

---Experimental description---

See previous publications

--DREAM categorization procedure--

The original experiment categorized the dream reports according to the DREAM definitions.

---Technical details---

The data include a single EMG channel (called EMG1), which is a bipolar derivation of the standard LAT and RAT electrodes placed on the chin of the sleeping subject. The data include two EOG channels (called EOGu and EOGl), which were placed in the American Academy of Sleep Medicine (AASM) recommended E1 and E2 positions below (l for lower) and above (u for upper) the lateral canthi.

--Data acquisition--

See published manuscripts

--Data preprocessing--

Data were rereferenced to a common average reference and zero centered to adjust for the DC offset. Raw data can be requested in accordance with statements in published manuscripts.

ObjectivesObstructive sleep apnea (OSA) may increase the risk of Alzheimer's disease (AD). However, potential associations among sleep-disordered breathing, hypoxia, and OSA-induced arousal responses should be investigated. This study determined differences in sleep parameters and investigated the relationship between such parameters and the risk of AD.MethodsPatients with suspected OSA were recruited and underwent in-lab polysomnography (PSG). Subsequently, blood samples were collected from participants. Patients' plasma levels of total tau (T-Tau) and amyloid beta-peptide 42 (Aβ42) were measured using an ultrasensitive immunomagnetic reduction assay. Next, the participants were categorized into low- and high-risk groups on the basis of the computed product (Aβ42 × T-Tau, the cutoff for AD risk). PSG parameters were analyzed and compared.ResultsWe included 36 patients in this study, of whom 18 and 18 were assigned to the low- and high-risk groups, respectively. The average apnea–hypopnea index (AHI), apnea, hypopnea index [during rapid eye movement (REM) and non-REM (NREM) sleep], and oxygen desaturation index (≥3%, ODI-3%) values of the high-risk group were significantly higher than those of the low-risk group. Similarly, the mean arousal index and respiratory arousal index (R-ArI) of the high-risk group were significantly higher than those of the low-risk group. Sleep-disordered breathing indices, oxygen desaturation, and arousal responses were significantly associated with an increased risk of AD. Positive associations were observed among the AHI, ODI-3%, R-ArI, and computed product.ConclusionsRecurrent sleep-disordered breathing, intermittent hypoxia, and arousal responses, including those occurring during the NREM stage, were associated with AD risk. However, a longitudinal study should be conducted to investigate the causal relationships among these factors.

Contingency table of OSA severity measured by PSG and BSP (all subjects, N = 78).

Summary of subject characteristics and PSG results.

Performance (mean with 95% CI) of bAHI at PSG-AHI cutoff of 15 for subjects taking heart-rate affecting medicines and subjects with comorbidities.