PreprintPDF Available

Sleep monitoring and wearables : a systematic review of clinical trials and future applications (Preprint)

Preprints and early-stage research may not have been peer reviewed yet.


BACKGROUND Sleep disorders are a major public health issue. Nearly one in two people will experience sleep disturbances during lifetime with a potential harmful impact on well-being, physical and mental health. The rise of connected objects is bringing new opportunities in sleep monitoring. OBJECTIVE To better understand the clinical value wearables-based sleep monitoring, we conducted a review of the literature, including feasibility studies and clinical trials on this topic. METHODS We searched PubMed, PsycINFO, ScienceDirect, the Cochrane Library, Scopus, and Web of Science up to June 2017. We created the list of keywords based on two domains: wearables and sleep. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses to identify, select, and critically appraise relevant research while minimizing bias. RESULTS The initial research collected 255 articles. 18 articles meeting the inclusion criteria were included in the final analysis. Out of the selected articles, four categories appeared. Feasibility studies propose testing new connected tools during sleep, on small samples of subjects. Population comparison studies propose to evaluate the sleep of patients compared to that of healthy subjects. Several studies evaluated connected objects in comparison with polysomnography, a reference test in sleep assessment. Finally, an article evaluates the impact of sleep disorders in the clinic. CONCLUSIONS We conducted a broad analysis of studies on the clinical and technical aspects of the use of wearables for sleep monitoring. This review of the literature showed that wearables are acceptable and promising monitoring tools in a wide range of clinical applications for sleep monitoring.
Sleep monitoring and wearables : a systematic review of clinical trials and future
Elise Guillodo1, Christophe Lemey1,2,3, Mathieu Simmonet2, Juliette Ropars4,5, Sofian Berrouiguet1,2,3
1. Department of Psychiatry and Emergency, University Hospital, Brest, France.
2. IMT Atlantique, Lab-STICC, F-29238 Brest, France
3. SPURBO EA 7479, Université de Bretagne Occidentale.
4.Laboratoire de Traitement de l'Information Médicale, INSERM, UMR 1101, Brest, France.
5.Department of Child Neurology, University Hospital of Brest, Brest, France
Sleep disorders are a major public health issue. Nearly one in two people will experience sleep
disturbances during lifetime with a potential harmful impact on well-being, physical and mental
health. The rise of connected objects is bringing new opportunities in sleep monitoring.
To better understand the clinical value wearables-based sleep monitoring, we conducted a review
of the literature, including feasibility studies and clinical trials on this topic.
We searched PubMed, PsycINFO, ScienceDirect, the Cochrane Library, Scopus, and Web of Science
up to June 2018. We created the list of keywords based on two domains: wearables and sleep. We
used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses to identify, select,
and critically appraise relevant research while minimizing bias.
The initial research collected 255 articles. 18 articles meeting the inclusion criteria were included
in the final analysis. Out of the selected articles, four categories appeared. Feasibility studies
propose testing new connected tools during sleep, on small samples of subjects. Population
comparison studies propose to evaluate the sleep of patients compared to that of healthy subjects.
Several studies evaluated connected objects in comparison with polysomnography, a reference test
in sleep assessment. Finally, an article evaluates the impact of sleep disorders in the clinic.
We conducted a broad analysis of studies on the use of virtual reality in patients with eating
disorders. This review of the literature showed that virtual reality is an acceptable and promising
monitoring tool in a wide range of clinical applications
Sleep disorders
Sleep disorders are a major public health issue. Nearly one in two people will experience sleep
disturbances during their lifetime (1) with a potential harmful impact on well-being, physical and
mental health (2). The International Classification of Sleep Disorders distinguishes the following
six categories: insomnia, sleep-related breathing disorders, central hypersomnia, circadian rhythm
disorders, parasomnias and sleep-related motor disorders (3). For example, insomnia is
characterized by complaints about the duration and quality of sleep, difficulty falling asleep,
nocturnal awakenings, early awakening and / or non-recuperative sleep (4). This symptomatology
must be present at least three times a week, for at least one month, with negative consequences on
the next day. Sleep and mental health are highly related, with many mental health problems also
being associated with sleeping disorders (5). Traditionally, sleeping disorders have been viewed as
a consequence of mental health disorders, and evidence also suggests that sleeping disorders can
contribute to the development of new mental health problems (6).
Sleep Monitoring
Normal sleep is characterized by a succession of four to six cycles lasting about ninety minutes.
Each of these cycles consists of slow-wave phases and rapid eye movement (REM) sleep, which are
related to slowdown and activation of the central nervous system. During REM sleep, or stage 5,
rapid eye movements are observed and muscle tone is abolished. The early-night cycles are
especially rich in deep, slow sleep and the latter in REM sleep (7). The duration of normal sleep
varies between six and ten hours, depending on several factors, the most important of which are
age and genetics.
Normal and pathological sleep can be explored either subjectively, i.e. by asking the subject, or
objectively, using sensors. An epidemiological study conducted in 2013 with over 1000
participants found a subjective prevalence of insomnia of 15%, while the objective prevalence
measured by polysomnography (PSG) was 32% (8). To date, polysomnography remains the gold
standard to objectively assess sleep characteristics. The polysomnograph plots a hypnogram,
integrating data from several sensors: electroencephalogram (EEG), electromyogram (EMG),
electrooculogram (EOG), thoracic movement (from belts on the chest and abdomen), airflow
measures, oximetry, and electrocardiogram (ECG). The sleep stages are scored according to
standard visual criteria based on the EEG, EOG, and EMG sensors (5). The assessment must be
carried out under controlled conditions in the laboratory for eight to twelve hours. An automated
hypnogram analysis is possible, but still needs manual integration of data (7). Successful recording
of the polysomnography over the course of the recording and the analysis of the results must be
carried out by a clinician with expertise in sleep pathologies in brain disorders. Although
polysomnography is considered the "gold standard", it is an examination with limitations: it can be
cumbersome for the patient, not very accessible, and not being realized in ecological conditions. It
may therefore not be suitable for all populations of interest, such as individuals who are suicidal,
sensitive to the environment or who may require emergency care.
Wearables and eHealth
Increased adoption of features of the internet has increased the possibilities for improved patient
monitoring. Integration of smartphones and wearables tools into medical practice has heralded
the electronic-health (eHealth) and mobile-health (mHealth) era. These tools can be used to self-
monitor or self-assess, allowing individuals to better understand their behavior and body, and
therefore their health. Aspects of daily life are particularly targeted, with measures of diet,
physical activity or sleep. These self-measurements can be tracked and analyzed with the objective
of modifying individual behaviors, including using educational approaches. We therefore observe
an association between the concepts of self-monitoring or self-tracking, and empowerment, with
greater patient involvement and better autonomy. Finally, these devices also allow for the
possibility for clinicians to access and review clinical data in real time (9).
Wearables and sleep monitoring
In the field of sleep, the rise of connected devices was largely based on actimetry. The actimeter
uses an accelerometer worn on the wrist and thus detects the movements of the limb (10). Its use
has increased because it is easy to use and allows recordings over periods of time longer than a
single night of polysomnography. However, this assessment method has some limitations. Indeed,
according to the numerous comparative studies with polysomnography, it is shown that the
actimeter hardly detects sleepiness, underestimates the latency of falling asleep, and
overestimates the number of micro-awakenings, compared to the reference examination. Finally,
this device does not provide information on the stages of sleep. Actimetry is therefore limited to
subjects with circadian rhythm disturbances and to evaluation of total sleep time (11). Some
devices use electro-encephalographic and electro-oculographic recordings (12). But these devices,
still little accomplished, require the positioning of several electrodes and are therefore impractical
for home use by the patient (11).
Other devices measure heart rate, and rely on the variability of the heart rate to give the stages of
sleep. Indeed, this variability is higher during paradoxical sleep or nocturnal awakenings, and
lower during slow sleep (13), by sympathetic or parasympathetic action modulations of the
autonomic nervous system (11). These devices are available in different forms: watch, chest band,
electrodes or monitoring on the mattress or pillow, but they still have poor results (10).
Overall, wearables are promising sleep monitoring methods and allow for the recording of several
nights , when polysomnography only assesses a single night of recording (10). Three reviews have
examined the potential features of wearable devices for sleep monitoring (14–16). However, none
of these reviews used a systematic review method in order to report recent clinical research
results. Our hypothesis is that connected devices could reliably and objectively measure sleep
ambulatorily. We therefore conducted a review of the literature regarding the use of connected
devices in sleep assessment.
2.Materials and methods
This literature review aims to identify clinical trials of mobile sleep recording in adult populations
in order to gain an overview of existing assessment techniques and their use in sleep disorders.
The literature search was conducted in June 2018 via the Pubmed, PsycINFO, Science Direct and
Cochrane databases. The key words used were chosen from the terms used in health terminology
of the biomedical reference thesaurus or MeSH terms. The search was conducted using "AND" and
"OR" logistic operators in the MeSH terms, titles, summaries (Figure 1). The keywords used were
« sleep AND wearable electronic devices».
Selection criteria
The primary selection criterion was the reporting of clinical trials using portable devices for sleep
recording in adults. We eliminated studies exploring the recording of activity during the day. Our
review therefore focuses solely on sleep recording. In addition, we selected trials in adult
populations, so pediatric studies were excluded.
Data extraction and analysis
The analysis of the articles was conducted in two stages. As a first step, a census based on the
review of titles and abstracts of scientific papers meeting the inclusion criteria was conducted. In a
second step, the information relating to the publication was extracted (author, date of publication,
country, design of the study, population, number, inclusion criteria, exclusion criteria, scales of
evaluation, objectives of study, protocols, results). The full text of all publications that were not
excluded after analysis of titles and abstracts was reviewed. All studies meeting the inclusion
criteria were included.
The steps of the literature review research and analysis are summarized in Figure 1. The initial
search identified 255 articles. No duplicate articles were identified, and screening based on the
titles removed 94 articles. 109 articles were excluded after review of the abstracts. After review of
the full text, 34 additional articles were excluded because they did not meet the inclusion criteria.
Thus, 18 articles meeting the inclusion criteria were included in the final analysis – an overview of
these studies is shown in Table 1.
Studies identified on database
search (n=255 )
Other sources
Records after duplicates removed
Records screened (abstracts)
Record excluded after title
screening (n=94)
Full text addressed for eligibility
Studies included in the review
Full text excluded after abstract
screening (n=109)
Identification Screening Eligibility Included
      
  
Studies identified on database
search (n=255 )
Other sources
Records after duplicates removed
Records screened (abstracts)
Record excluded after title
screening (n=94)
Full text addressed for eligibility
Studies included in the review
Full text excluded after abstract
screening (n=109)
Identification Screening Eligibility Included
      
  
Figure 1 : PRISMA Flow Chart
Year of publication
Of the 18 studies, 1 (5,56%) study was published in 2008, 1 (5,56%) in 2011, 1 (5,56%) in 2012, 3
(16,67%) were published in 2014, 3 (16,67%) in 2015, 5 (27,78%) studies were published in
2016, 2 (11,11%) in 2017, and 2 (11,11%) in 2018.
Country, origin of development
Of the 18 studies identified, 5 (27,78%) were conducted in the United States, 3 (16,67%) in Italy, 2
(11,11%) were conducted in Israel, 2 (11,11%) in the United Kingdom, 1 (5,56%) study was
conducted in Taiwan, 1 (5,56%) in Thailand, 1 (5,56%) in France, 1 (5,56%) in Korea, 1 (5,56%) in
Australia, and finally 1 (5,56%) study was conducted in Finland.
Studied population
Of the 18 studies, 10 (55,56%) studies were conducted in healthy adults, 4 (22,22%) were
conducted in patients with sleep disorders, 1 (5,56%) with people older than 60, 1 (5,56%) in
patients with fibromyalgia, 1 (5,56%) in patients with bipolar disorder, and 1 (5,56%) in patients
with Parkinson's disease.
Design and size
Of the 18 studies in this review, 10 (55,56%) are comparability studies with the gold standard, 4
(22,22%) are feasibility studies, 3 (16,67%) are population comparison studies, and 1 (5,56%)
assesses the impact of sleep disorders in the clinic. The samples are heterogeneous in size, ranging
from 1 to 15839 patients.
Results analysis
On reviewing the full text of the articles in this review, four categories appeared. Feasibility studies
propose testing new connected tools during sleep, on small samples of subjects. Population
comparison studies compare the sleep of patients to that of healthy subjects. There are also
several studies evaluating connected devices in comparison with polysomnography, a reference
test in sleep assessment. Finally, an article evaluates the impact of sleep disorders in the clinic.
Feasibility studies
In their study, Baron et al (17). aimed to outline the theoretical foundation and iterative process of
designing the "Sleep Bunny", a technology-assisted sleep extension intervention including a
mobile phone app, wearable sleep tracker, and brief telephone coaching. The population was made
of 6 adults with short sleep duration (<7 hours), testing the application with the sleep tracker and
the telephone coaching once a week, over 4 weeks. The survey, based on open-ended questions,
asked participants to provide comments on the content, layout and general feedback about the
app. In conclusion, users enjoyed the wearable sleep tracker and found the app visually pleasing,
but suggested improvements to the notification and reminder features.
The team of Castiglioni et al. (18) studied the feasibility of wearing "MagIC-SCG" sternal wrist
device, during high altitude sleep, which is conducive to hypoxia. Their device recorded the
electrocardiogram, respiratory movements, sternal accelerations, and oxygen saturation. The
study demonstrated the feasibility of recording and using the equipment in high altitude
The same device was used by Di Rienzo et al. (19) to study the feasibility of estimating cardiac
functions such as contraction and isovolumic relaxation times, or ventricular ejection time during
sleep. The data were transmitted in real time to an external device via a Bluetooth connection.
According to the authors, this is the first study of its kind, but one that would require more
investigation and more subjects to obtain actionable results.
The team of Kayyali et al. (20), in the United States, investigated the feasibility of portable sleep
recording at the patient's home. Their device, "PSG @ home", was placed in the thoracic region and
recorded respiratory movements, oxygen saturation, airflow, snoring, body position, and the
electrocardiogram. Their study demonstrated the feasibility of using this discreet device overnight
at the subject's home.
 !"
! #
$! %! & !" ' !"
(! $!!! !"
 
 
0! #
 
34 /-"!"!""
5" "
6*" *"
 !!*"
)!" "
 15"8"
2" "!"
"#;" *"
 %!"
> ?!2"""
6#" 
 %!
/ .B1/>"#""
 %!"DE"! "
! "
! %6"
 "
8!"! ""
)! %!""
F"@ #F""
 *""!"
25" *"4"
 6#
,B C"H!#"
")! ""
"6 #"
*" "
 **"
) 
 H"
/ I!  ! 8!"6 "2
%!" *"
% 
*"% ""
 %!
."!", L#"
,-4 I!  ! 0!!"
2" !1
 !"!*"
! 
. ""
! "!"
,B I!  ! 8!"""
 ! #"
 H",/"
. I!  ! L#"!)""1
") 
! *""
@ !!"
&!6! 
I 
&!6! 
B"! "
 
 !
0 !"6*"
 ""
)!""! 
! 8!"%)#""
"S #""
 
 %!
E 8!""5"!)"
)S #"
"#! *"
6 "
 +""
 %!"
' 6#
 "
$T *""
U #"&'"
 *""
6! ","
*")S "
 *"
 #"!)"
2" *!*"
 "
. 8"!%!"
*! """
)S #"!)"
" "
.- I!  ! L#"!)"!"
!" *"
 *"6!""
 "
".-" "
2" ) "
A !""
B. >/"2"*!!""
! 8!"6!""
 
 ! "$("
 "
6! ""
 "
 "$("!*
' 
* 
! 
! ""
 "
4 I!  !#"!)"
 !!*
"!" #"!"
(" ! #
( "4>
 %"
 "
#! *"
 
4- .B13-"# !#"!)"
 !!*
L'"*"$ 
%2" 
#+"" "
)! "
 "!)"
 H
.B14>"# $#"
 *"""2!"
! %!
6 %!" *"
"""*! "!)"
#" H"
.""4  "I"
6! 
.- I!  ! 8!"6 ""
 #"!)""
$ G"$G"
! "2"
@!#* ","@(,"
2"2" ""
. $G"!#"
 "
 *"
!* 
!! "
!6" "
6 "6"
 *"!!"
 *"!! 
 #"!!+"@#"
8  
! %"
. &"""""
! %!")!"
 !"!""
 
.> I!  ! 8!"""
7%!" *"
"("" %1
!! !*"
! 
. (") "
2") "
 
#! *"
., 1!)!"
8!"6 ""
 *"!"
! 
! 
2!""! "
 !!*
" "
 *" 
8!"S %%6#"
!"! "
! ""!"
 5"""
 "
 %!"
 "
! "!)"
. &"!"
 *"
! "
"! 
Population comparison
The purpose of the study by Fagherazzi et al. (21) was to highlight the determinants involved in
poor sleep. For this, they calculated the 7-consecutive-night deep sleep/total sleep ratio of a large
number of users of Withings connected devices. They used an algorithm using data provided from
both the accelerometer and temperature sensor. A ratio indicating poor sleep was defined as
below 0,40. Their findings show that young men with elevated heart rate and high blood pressure
were at higher risk for poor sleep quality.
Migliorini et al. (22) compared sleep records between a healthy adult population and a patient
with bipolar disorder. The monitoring was done by the "Smartex" T-shirt equipped with sensors.
The data collected were the electrocardiogram, the respiratory activity and the movements via an
accelerometer, allowing the stages of the sleep and an estimation of the percentage of paradoxical
sleep to be obtained. The results showed a variability of the reduced heart rate in the individual
with bipolar disorder, as well as an increase in the percentage of paradoxical sleep. These results
would need to be confirmed by a larger sample, but seem to be an interesting way of identifying
thymic decompensations.
The study by Sringean et al. (23) compared sleep in the homes of individuals with Parkinson's
disease, with that of their spouses or partners, as healthy controls, to provide a quantitative
analysis of nocturnal hypokinesia. Portable sensors were worn at the trunk and limbs. Records
included number, speed, acceleration, degree and duration of movement / turnarounds, number of
bed exits, and limb movements. They conclude by pointing out the effectiveness of their system to
record nocturnal movements, demonstrating the significant presence of nocturnal hypokinesia.
Comparison with polysomnography
The American team of De Zambotti et al. (24) compared data from the Jawbone UP with
polysomnography data collected simultaneously. The Jawbone UP is a wristband that records
accelerometer data in its first version. Comparisons were made between total sleep time, bedtime,
sleep latency and nighttime awakenings. It has been shown that the estimates of these parameters
are in good agreement with polysomnography, a reference examination in sleep pathologies.
Dafna et al. (25) studied the use of a portable respiratory sound recording tool, with the aim of
estimating respiratory rate by analyzing the audio signal. The data were compared to those of
polysomnography. The authors concluded that their method was reliable and robust in estimating
the respiratory rate, especially since the device is not very intrusive and does not interfere with
the subject's sleep.
Kang et al. (26) compared the accuracy of detecting sleep epochs of the commercial Fitbit Flex
device with polysomnography. They studied a population of 41 individuals with insomnia disorder
and 21 good sleepers. Participants wore the wearable electronic device while undergoing
polysomnography, during 1 night. The measures of interest in this study were Total Sleep Time
(TST), Sleep Efficiency (SE), Sleep Onset Length (SOL), Wake After Sleep Onset (WASO). They
conclude that the frequency of agreement was high in good sleepers but significantly lower in
those with insomnia.
The team of Kuo et al. (27) developed and evaluated a hand-held wrist-based sleep recording tool
based on actimetry. The developed device was judged to be energy efficient and highly accurate in
measuring sleep efficiency, total sleep time, sleep time, and nighttime awakenings.
Polysomnography measurements were taken simultaneously. It was shown that the different
variables were concordant, and significantly so for total sleep time and sleep efficiency. According
to the authors, this system is interesting for obtaining objective sleep data at the patient’s home.
In their pilot study, Looney et al. (28) compare electroencephalographic recordings recorded by
standard electrodes at the level of the scalp, and by an intra-auricular device, simultaneously
during sleep. The lines were read blindly by an expert. The results showed a significant
concordance of the two recordings.
Mantua et al. (29) compared the data from five portable connected devices recording sleep with
those of the gold standard, polysomnography. The devices under study were Actiwatch, Basis,
Misfit Shine, Fitbit Flex, and Withing Pulse O2. The recordings were made simultaneously at the
participant's home, with participants wearing the five devices on the wrists, and
polysomnography was installed. Significant data loss was reported by "Fitbit Flex" and "Misfit
Shine". The correlation analysis allows them to conclude that there is no significant difference in
estimating total sleep time between polysomnography and each of the five devices. In addition,
only "Actiwatch" had concordant data with the baseline sleep efficiency test. The light sleep time
differed for all devices. Finally, a correlation of deep sleep time was significant only for "Basis".
The team of Parak et al. (30) compared the nightly heart rate recording of the connected watch
"PulseOn" with the reference test, the electrocardiogram. The study, conducted at home, shows
that the device correctly detects 99.57% of heartbeats, making it an accurate method during sleep.
In their sleep laboratory study, Rodriguez-Villegas et al. (12) compared the effectiveness of a
wireless system in the detection of apnea and hypopnea with that of polysomnography. The 17-
gram apparatus was placed on the skin of the anterior aspect of the neck. It recorded turbulence in
the trachea using an acoustic chamber. Data were analyzed blind. The tolerance of the device was
greater than that of polysomnography. However, the results did not agree with the gold standard
regarding the correct detection of hypopneas. In conclusion, this tool could be an adequate
solution for the monitoring of apneas in ecological conditions, but would not replace a complete
recording in the sleep laboratory.
Sano et al. (31) studied the comparison of electroencephalogram data with the Q Sensor Affectiva,
concerning the detection of waking and sleeping phases in a specialized hospital laboratory. The Q
Sensor Affectiva is a watch that records skin temperature, cutaneous conductance and
acceleration. In their conclusion, it appears that the combination of acceleration and skin
temperature is most effective in the sleep / wake classification.
The Australian team of Sargent et al. (32) evaluated the validity of a commercial wearable device,
the Fitbit HR Charge, for measuring Total Sleep Time (TST). To do this, they compared the
commercial wearable device with the gold standard, the polysomnography, during 30 nights and
20 naps. They chose a population of 12 well-trained young athletes. This study showed that the
Fitbit HR Charge overestimated TST for night-time sleep periods and for daytime naps.
Impact of sleep in the clinic
In the study by Agmon et al. (33), the impact of sleep on walking performance in institutionalized
elderly was measured using a connected watch using an accelerometer. Sleep efficiency, sleep
latency, total sleep time, and nocturnal awakenings were taken into consideration. The team
demonstrated that a decrease in recovery sleep was significantly associated with a decrease in
start-up speed and a greater variability of walking during double tasks.
This review of the literature shows an increasing interest in the use of connected health devices
for sleep assessment. Their use is proposed in a broad spectrum of pathologies as well as in
healthy individuals.
First of all, this review and other recent studies (34–37) indicate that health scientists no longer
consider wearables as “toys”. Nowadays, they are tools to help patients, doctors and scientists.
Thanks to new technologies, on the one hand we are now able to wear small and non-intrusive
devices during daily and nightly life (38) and on the other hand we are also capable to deeply
analyse big datasets (39). In other words, wearables and systems to collect and analyse
physiological sleep data are now acceptable and powerful. Even if this new perspective is
promising, it is only the beginning as previous studies often included only small samples. In this
review of the literature, only 7 studies among 18 included 30 participants or more. One could
expect this will improve thanks to the increase of the acceptance (40,41). Recent findings also
emphasize that little is still known about physiological monitoring in ecological situations (42).
This could be explained because we could not easily record wide datasets before. The most
common methodology identified in this review was motion sensing via accelerometry, a well-
studied method in the field of sleep research which has been used for decades (43). The research
literature consistently shows that wrist accelerometry, even in healthy adults, has high sensitivity
for sleep, while having much lower specificity for sleep.
Another interesting finding among the four feasibility studies identified in this review is that
outpatients did not only experience worn wrists devices but also slept with electrodes and
cutaneous sensors. Despite of this, conclusions still pointed out the high feasibility level of studies
with such an equipment. Even if these studies should be proceeded with larger samples, one could
guess that on-board polysomnography-tools are on the way.
Limitations of the review’s method
Although the studies selected in this review are recent, the rapid evolution of technologies in this
area makes it difficult to adjust research to keep pace with commercial releases. Since starting the
review process, additional articles may have been published. Moreover, while the conclusions are
encouraging, most of them are pilot studies with small samples. In addition, the tools tested are
commercial, and therefore have no health device label. Limits on scientific validity mean that these
devices are not usable in a current clinic, and it may be premature to recommend them. Given the
rapid progression of technologies, it does not seem unrealistic to think that more complete and
validated devices will be available soon. Moreover, the heterogeneity of the population studied in
this review makes it difficult to issue a general conclusion. In return, this reflects the broad
spectrum of usability of connected devices in the field of sleep. Finally, it should be noted that this
review of the literature does not provide any data on the use of these objects in the long term,
because they are mostly short clinical trials, with devices that may have defects in the collection of
data due to limited battery for example.
One of the most significant issue related to wearables use in clinical monitoring is to establish the
congruency between the well-proven gold standard and the new emerging wearables. Indeed, the
results from polysomnography comparisons revealed that sleep experts validated the data from
the mobile devices. Since these experts have experience working with polysomnography, it
reinforces the statement. From a methodological point of view, the study from Looney et al (25)
highlighted the importance of a blind interpretation process. By this respect, one could suggest
that if the same expert could work indifferently with both system, wearables definitely are
efficient. However, further studies with more participants is required to generalize these results,
this review demonstrates the potential for wearables to get generate more knowledge about sleep.
Ethical aspects
The expansion of connected devices has raised ethical concerns. On the one hand, there is a risk of
instrumentalisation of the human body and health. The purely technical reading of a physical or
somatic disorder is a reduction of a more complex reality. Quantification of the self would lead to a
reduction of social representations of illness and health in general: numerical data, compared to a
norm, could impoverish the variety of feelings and the expression of the patient's experience, as
well as the clinic associated with it. In addition, the lack of contextualization engendered could
devalue the knowledge of practitioners who value the experience. On the other hand, there are
questions about reliability and data protection. Patients may be apprehensive about the risk of
loss of control over data and privacy. It is therefore essential to establish a climate of trust around
the use of these tools in medical practice. Finally, connected devices are also sometimes presented
as a threat to social solidarity, on which the French health system is based. We are talking about a
"digital divide", and the "unconnected" or those whose psychological state would not benefit from
these tools, could then be victims of social, cultural and economic marginalization.
Moreover, it has been shown that individuals lose interest in connected objects after a few months
of use. The reasons given are a lack of adaptation to the subject's needs, or adjustment of use
through the developmental process.
Weakness of analysis system
This review did not identify any studies describing the use of data mining techniques. Data mining
is a set of techniques that can be used to explore clinical questions in large databases. The data
mining process includes several steps, including data selection, data processing, and machine
learning to identify which factors may influence results. For example, data mining techniques have
been described to find relationships and patterns between electronic data and neurobiological
data (44). Machine learning techniques allow processing of real-time observational information by
continuously learning from data to build understanding and uncover previously unexpected
associations and patterns in data (45). Commercial wearables devices implicitly use these
techniques, at least during development process of the product. These techniques may help
clinicians to better interpret recordings provided by wearables. Unsupervised machine learning is
the machine learning task of inferring a function to describe hidden structure from "unlabeled"
data . These techniques should be also described in making sense of enormous wearables datasets
is a common challenge in the Big Data era that is best overcome using data mining techniques.
The current landscape of marketing claims that stretch beyond the scant validation literature may
contribute to the discrepancy between consumer and health care community adoption of sleep
monitoring devices. Clinicians and researchers, more familiar with the extensive experience
polysomnography (PSG) and actigraphy, still seem to be reluctant to integrate wearables into
clinical practice, pending rigorous validation. Wellness claims are not necessarily adapted to
clinical practice setting, and as a result, has delayed the adoption of this devices in clinical practice.
In this review, we described several clinical applications of consumer sleep monitoring technology.
Consumer sleep devices contribute to the blurry boundary between sleep as a medical concept
and sleep as “wellness” and the need for a framework to interpret consumer sleep device outputs.
Future applications
Wearables are able to perform passive (or autonomous) data gathering, i.e., to extract information
about the users without their intervention. Actigraphy, geolocation, and communication activity
are usual features of current smartphones and may be useful indicators, if properly processed, of
the individual’s condition or state (46). Advances in sensor technologies, and novel textile-
electronic integration technics also draw new perspectives for behavioral ecological assessment
(BEA). Moreover, it is currently possible to find commercially available wearable sensing
technologies for several wellness and clinical purposes, ranging from simple heart rate monitors
(HR) (47) to monitors of physical activity, rehabilitation after surgical intervention or sleep quality
assessment. Overall, an extensive panel of physical and mental conditions (e.g., insomnia, elderly,
diabetic, cardiac, or respiratory problems) can be remotely monitored by the appropriate health
care professional (e.g., physician, doctor, nurses). These devices are often bound to a smartphone,
increasing the networking capabilities and the user experience. Collected data can be processed
and transferred over the Internet to a remote clinical back-end server for further analysis,
assessment, and decision making and intervention.
This review of the literature on connected devices in health shows the growing interest for these
new technologies, as well as the wide applications that can be made of them. Indeed, it is observed
that the studies can reflect different specialties of medicine and that the populations studied are
varied. In addition, this interest is recent, the vast majority of studies dating from 2014.
Qualitatively, the majority of devices are considered comfortable (30), easy to use (20), and
preserve the natural sleep of the user (25,28), making them good candidates for home care
(12,22,27). In addition, they have an economic advantage and the preliminary results of this study
show a good correlation with the reference examination (24). Given the many benefits, we must
consider connected devices as promising tools for the modernization of care.
1. Tan M, Marra C. The cost of sleep disorders: no snoring matter. Sleep. mars 2006;29(3):282-3.
2. Li SX, Lam SP, Zhang J, Yu MWM, Chan JWY, Chan CSY, et al. Sleep Disturbances and
Suicide Risk in an 8-Year Longitudinal Study of Schizophrenia-Spectrum Disorders. Sleep. 1 juin
3. World Health Organization, éditeur. International statistical classification of diseases and related
health problems. 10th revision, Fifth edition. Geneva, Switzerland: World Health Organization; 2016. 3
4. Kerkhofs M. Troubles du sommeil de l’adulte et de l’adolescent. avr 2015;12(2).
5. Baglioni C, Nanovska S, Regen W, Spiegelhalder K, Feige B, Nissen C, et al. Sleep and mental
disorders: A meta-analysis of polysomnographic research. Psychol Bull. sept 2016;142(9):969-90.
6. Baglioni C, Battagliese G, Feige B, Spiegelhalder K, Nissen C, Voderholzer U, et al. Insomnia as
a predictor of depression: a meta-analytic evaluation of longitudinal epidemiological studies. J Affect
Disord. déc 2011;135(1-3):10-9.
7. Kushida CA, Littner MR, Morgenthaler T, Alessi CA, Bailey D, Coleman J, et al. Practice
parameters for the indications for polysomnography and related procedures: an update for 2005. Sleep.
avr 2005;28(4):499-521.
8. Castro LS, Poyares D, Leger D, Bittencourt L, Tufik S. Objective prevalence of insomnia in the
São Paulo, Brazil epidemiologic sleep study. Ann Neurol. oct 2013;74(4):537-46.
9. Baig MM, GholamHosseini H, Moqeem AA, Mirza F, Lindén M. A Systematic Review of
Wearable Patient Monitoring Systems - Current Challenges and Opportunities for Clinical Adoption. J
Med Syst. juill 2017;41(7):115.
10. Kelly JM, Strecker RE, Bianchi MT. Recent developments in home sleep-monitoring devices.
ISRN Neurol. 2012;2012:768794.
11. Muzet A, Werner S, Fuchs G, Roth T, Saoud JB, Viola AU, et al. Assessing sleep architecture
and continuity measures through the analysis of heart rate and wrist movement recordings in healthy
subjects: comparison with results based on polysomnography. Sleep Med. mai 2016;21:47-56.
12. Rodriguez-Villegas E, Chen G, Radcliffe J, Duncan J. A pilot study of a wearable apnoea
detection device. BMJ Open. 3 oct 2014;4(10):e005299.
13. Viola AU, Simon C, Ehrhart J, Geny B, Piquard F, Muzet A, et al. Sleep processes exert a
predominant influence on the 24-h profile of heart rate variability. J Biol Rhythms. déc
14. Sadeh A. The role and validity of actigraphy in sleep medicine: an update. Sleep Med Rev. août
15. Russo K, Goparaju B, Bianchi MT. Consumer sleep monitors: is there a baby in the bathwater?
Nat Sci Sleep. 5 nov 2015;7:147-57.
16. Ko P-RT, Kientz JA, Choe EK, Kay M, Landis CA, Watson NF. Consumer Sleep Technologies:
A Review of the Landscape. J Clin Sleep Med. 15 déc 2015;11(12):1455-61.
17. Baron KG, Duffecy J, Reid K, Begale M, Caccamo L. Technology-Assisted Behavioral
Intervention to Extend Sleep Duration: Development and Design of the Sleep Bunny Mobile App. JMIR
Ment Health. 10 janv 2018;5(1):e3.
18. Castiglioni P, Meriggi P, Rizzo F, Vaini E, Faini A, Parati G, et al. Seismocardiography while
sleeping at high altitude. Conf Proc IEEE Eng Med Biol Soc. 2012;2012:3793-6.
19. Di Rienzo M, Vaini E, Castiglioni P, Lombardi P, Parati G, Lombardi C, et al. Wearable
seismocardiography for the beat-to-beat assessment of cardiac intervals during sleep. Conf Proc IEEE
Eng Med Biol Soc. 2014;2014:6089-91.
20. Kayyali HA, Weimer S, Frederick C, Martin C, Basa D, Juguilon JA, et al. Remotely attended
home monitoring of sleep disorders. Telemed J E Health. mai 2008;14(4):371-4.
21. Fagherazzi G, El Fatouhi D, Bellicha A, El Gareh A, Affret A, Dow C, et al. An International
Study on the Determinants of Poor Sleep Amongst 15,000 Users of Connected Devices. J Med Internet
Res [Internet]. 23 oct 2017 [cité 17 juill 2018];19(10). Disponible sur:
22. Migliorini M, Mendez MO, Bianchi AM. Study of Heart Rate Variability in Bipolar Disorder:
Linear and Non-Linear Parameters during Sleep. Front Neuroeng. 2011;4:22.
23. Sringean J, Taechalertpaisarn P, Thanawattano C, Bhidayasiri R. How well do Parkinson’s
disease patients turn in bed? Quantitative analysis of nocturnal hypokinesia using multisite wearable
inertial sensors. Parkinsonism Relat Disord. févr 2016;23:10-6.
24. de Zambotti M, Claudatos S, Inkelis S, Colrain IM, Baker FC. Evaluation of a consumer fitness-
tracking device to assess sleep in adults. Chronobiol Int. 2015;32(7):1024-8.
25. Dafna E, Rosenwein T, Tarasiuk A, Zigel Y. Breathing rate estimation during sleep using audio
signal analysis. Conf Proc IEEE Eng Med Biol Soc. 2015;2015:5981-4.
26. Kang S-G, Kang JM, Ko K-P, Park S-C, Mariani S, Weng J. Validity of a commercial wearable
sleep tracker in adult insomnia disorder patients and good sleepers. J Psychosom Res. 2017;97:38-44.
27. Kuo C-E, Liu Y-C, Chang D-W, Young C-P, Shaw F-Z, Liang S-F. Development and Evaluation
of a Wearable Device for Sleep Quality Assessment. IEEE Trans Biomed Eng. 23 sept 2016;
28. Looney D, Goverdovsky V, Rosenzweig I, Morrell MJ, Mandic DP. Wearable In-Ear
Encephalography Sensor for Monitoring Sleep. Preliminary Observations from Nap Studies. Ann Am
Thorac Soc. déc 2016;13(12):2229-33.
29. Mantua J, Gravel N, Spencer RMC. Reliability of Sleep Measures from Four Personal Health
Monitoring Devices Compared to Research-Based Actigraphy and Polysomnography. Sensors (Basel). 5
mai 2016;16(5).
30. Parak J, Tarniceriu A, Renevey P, Bertschi M, Delgado-Gonzalo R, Korhonen I. Evaluation of
the beat-to-beat detection accuracy of PulseOn wearable optical heart rate monitor. Conf Proc IEEE Eng
Med Biol Soc. août 2015;2015:8099-102.
31. Sano A, Picard RW. Comparison of sleep-wake classification using electroencephalogram and
wrist-worn multi-modal sensor data. Conf Proc IEEE Eng Med Biol Soc. 2014;2014:930-3.
32. Sargent C, Lastella M, Romyn G, Versey N, Miller DJ, Roach GD. How well does a
commercially available wearable device measure sleep in young athletes? Chronobiol Int. 11 mai
33. Agmon M, Shochat T, Kizony R. Sleep quality is associated with walking under dual-task, but
not single-task performance. Gait Posture. sept 2016;49:127-31.
34. Min YH, Lee JW, Shin Y-W, Jo M-W, Sohn G, Lee J-H, et al. Daily collection of self-reporting
sleep disturbance data via a smartphone app in breast cancer patients receiving chemotherapy: a
feasibility study. J Med Internet Res. 23 mai 2014;16(5):e135.
35. Min YH, Lee JW, Shin Y-W, Jo M-W, Sohn G, Lee J-H, et al. Daily collection of self-reporting
sleep disturbance data via a smartphone app in breast cancer patients receiving chemotherapy: a
feasibility study. J Med Internet Res. 23 mai 2014;16(5):e135.
36. Wu W, Dasgupta S, Ramirez EE, Peterson C, Norman GJ. Classification accuracies of physical
activities using smartphone motion sensors. J Med Internet Res. 5 oct 2012;14(5):e130.
37. Poirier J, Bennett WL, Jerome GJ, Shah NG, Lazo M, Yeh H-C, et al. Effectiveness of an
Activity Tracker- and Internet-Based Adaptive Walking Program for Adults: A Randomized Controlled
Trial. J Med Internet Res. 9 févr 2016;18(2):e34.
38. Mercer K, Giangregorio L, Schneider E, Chilana P, Li M, Grindrod K. Acceptance of
Commercially Available Wearable Activity Trackers Among Adults Aged Over 50 and With Chronic
Illness: A Mixed-Methods Evaluation. JMIR Mhealth Uhealth. 27 janv 2016;4(1):e7.
39. Sathyanarayana A, Joty S, Fernandez-Luque L, Ofli F, Srivastava J, Elmagarmid A, et al. Sleep
Quality Prediction From Wearable Data Using Deep Learning. JMIR Mhealth Uhealth. 4 nov
40. Puri A, Kim B, Nguyen O, Stolee P, Tung J, Lee J. User Acceptance of Wrist-Worn Activity
Trackers Among Community-Dwelling Older Adults: Mixed Method Study. JMIR Mhealth Uhealth. 15
nov 2017;5(11):e173.
41. Ridgers ND, Timperio A, Brown H, Ball K, Macfarlane S, Lai SK, et al. Wearable Activity
Tracker Use Among Australian Adolescents: Usability and Acceptability Study. JMIR Mhealth Uhealth.
11 avr 2018;6(4):e86.
42. Brinton JE, Keating MD, Ortiz AM, Evenson KR, Furberg RD. Establishing Linkages Between
Distributed Survey Responses and Consumer Wearable Device Datasets: A Pilot Protocol. JMIR Res
Protoc. 27 avr 2017;6(4):e66.
43. Russo K, Goparaju B, Bianchi MT. Consumer sleep monitors: is there a baby in the bathwater?
Nat Sci Sleep. 2015;7:147-57.
44. van Os J, Verhagen S, Marsman A, Peeters F, Bak M, Marcelis M, et al. The experience
sampling method as an mHealth tool to support self-monitoring, self-insight, and personalized health
care in clinical practice. Depress Anxiety. 2017;34(6):481-93.
45. Berrouiguet S, Perez-Rodriguez MM, Larsen M, Baca-García E, Courtet P, Oquendo M. From
eHealth to iHealth: Transition to Participatory and Personalized Medicine in Mental Health. J Med
Internet Res. 3 janv 2018;20(1):e2.
46. Heo JS, Eom J, Kim Y-H, Park SK. Recent Progress of Textile-Based Wearable Electronics: A
Comprehensive Review of Materials, Devices, and Applications. Small. janv 2018;14(3).
47. Varnfield M, Karunanithi M, Lee C-K, Honeyman E, Arnold D, Ding H, et al. Smartphone-based
home care model improved use of cardiac rehabilitation in postmyocardial infarction patients: results
from a randomised controlled trial. Heart. nov 2014;100(22):1770-9.
... It is now common for physiological parameters such as HR, temperature, and physical activity levels to be measured, often with the aim of improving health outcomes ranging from self-management of diabetes to seizure monitoring fast pace of development [17]. Similarly, sleep trackers are increasingly common place and are a cost effective tool for monitoring sleep, though they appear to vary in reliability [18,19]. In the context of eSports, there is a natural applicability of these devices in training, especially considering the setting of a gaming house; an arrangement in which players cohabit a tailored facility with access to high speed internet and quality computing equipment. ...
Full-text available
eSports is a rapidly growing industry with increasing investment and large-scale international tournaments offering significant prizes. This has led to an increased focus on individual and team performance with factors such as communication, concentration, and team intelligence identified as important to success. Over a similar period of time, personal physiological monitoring technologies have become commonplace with clinical grade assessment available across a range of parameters that have evidenced utility. The use of physiological data to assess concentration is an area of growing interest in eSports. However, body-worn devices, typically used for physiological data collection, may constitute a distraction and/or discomfort for the subjects. To this end, in this work we devise a novel “invisible” sensing approach, exploring new materials, and proposing a proof-of-concept data collection system in the form of a keyboard armrest and mouse. These enable measurements as an extension of the interaction with the computer. In order to evaluate the proposed approach, measurements were performed using our system and a gold standard device, involving 7 healthy subjects. A particularly advantageous characteristic of our setup is the use of conductive nappa leather, as it preserves the standard look and feel of the keyboard and mouse. According to the results obtained, this approach shows 3–15% signal loss, with a mean difference in heart rate between the reference and experimental device of −1.778 ± 4.654 beats per minute (BPM); in terms of ECG waveform morphology, the best cases show a Pearson correlation coefficient above 0.99.
Full-text available
The validity of a commercially available wearable device for measuring total sleep time was examined in a sample of well-trained young athletes during night-time sleep periods and daytime naps. Participants wore a FitBit HR Charge on their non-dominant wrist and had electrodes attached to their face and scalp to enable polysomnographic recordings of sleep in the laboratory. The FitBit automatically detected 24/30 night-time sleep periods but only 6/20 daytime naps. Compared with polysomnography, the FitBit overestimated total sleep time by an average of 52 ± 152 min for night-time sleep periods, and by 4 ± 8 min for daytime naps. It is important for athletes and practitioners to be aware of the limitations of wearable devices that automatically detect sleep duration.
Full-text available
Background: Wearable activity trackers have the potential to be integrated into physical activity interventions, yet little is known about how adolescents use these devices or perceive their acceptability. Objective: The aim of this study was to examine the usability and acceptability of a wearable activity tracker among adolescents. A secondary aim was to determine adolescents' awareness and use of the different functions and features in the wearable activity tracker and accompanying app. Methods: Sixty adolescents (aged 13-14 years) in year 8 from 3 secondary schools in Melbourne, Australia, were provided with a wrist-worn Fitbit Flex and accompanying app, and were asked to use it for 6 weeks. Demographic data (age, sex) were collected via a Web-based survey completed during week 1 of the study. At the conclusion of the 6-week period, all adolescents participated in focus groups that explored their perceptions of the usability and acceptability of the Fitbit Flex, accompanying app, and Web-based Fitbit profile. Qualitative data were analyzed using pen profiles, which were constructed from verbatim transcripts. Results: Adolescents typically found the Fitbit Flex easy to use for activity tracking, though greater difficulties were reported for monitoring sleep. The Fitbit Flex was perceived to be useful for tracking daily activities, and adolescents used a range of features and functions available through the device and the app. Barriers to use included the comfort and design of the Fitbit Flex, a lack of specific feedback about activity levels, and the inability to wear the wearable activity tracker for water-based sports. Conclusions: Adolescents reported that the Fitbit Flex was easy to use and that it was a useful tool for tracking daily activities. A number of functions and features were used, including the device's visual display to track and self-monitor activity, goal-setting in the accompanying app, and undertaking challenges against friends. However, several barriers to use were identified, which may impact on sustained use over time. Overall, wearable activity trackers have the potential to be integrated into physical activity interventions targeted at adolescents, but both the functionality and wearability of the monitor should be considered.
Full-text available
Background: Despite the high prevalence of short sleep duration (29.2% of adults sleep <6 hours on weekdays), there are no existing theory-based behavioral interventions to extend sleep duration. The popularity of wearable sleep trackers provides an opportunity to engage users in interventions. Objective: The objective of this study was to outline the theoretical foundation and iterative process of designing the "Sleep Bunny," a technology-assisted sleep extension intervention including a mobile phone app, wearable sleep tracker, and brief telephone coaching. We conducted a two-step process in the development of this intervention, which was as follows: (1) user testing of the app and (2) a field trial that was completed by 2 participants with short sleep duration and a cardiovascular disease risk factor linked to short sleep duration (body mass index [BMI] >25). Methods: All participants had habitual sleep duration <6.5 hours verified by 7 days of actigraphy. A total of 6 individuals completed initial user testing in the development phase, and 2 participants completed field testing. Participants in the user testing and field testing responded to open-ended surveys about the design and utility of the app. Participants in the field testing completed the Epworth Sleepiness Scale and also wore an actigraph for a 1-week baseline period and during the 4-week intervention period. Results: The feedback suggests that users enjoyed the wearable sleep tracker and found the app visually pleasing, but they suggested improvements to the notification and reminder features of the app. The 2 participants who completed the field test demonstrated significant improvements in sleep duration and daytime sleepiness. Conclusions: Further testing is needed to determine effects of this intervention in populations at risk for the mental and physical consequences of sleep loss.
Full-text available
Clinical assessment in psychiatry is commonly based on findings from brief, regularly scheduled in-person appointments. Although critically important, this approach reduces assessment to cross-sectional observations that miss essential information about disease course. The mental health provider makes all medical decisions based on this limited information. Thanks to recent technological advances such as mobile phones and other personal devices, electronic health (eHealth) data collection strategies now can provide access to real-time patient self-report data during the interval between visits. Since mobile phones are generally kept on at all times and carried everywhere, they are an ideal platform for the broad implementation of ecological momentary assessment technology. Integration of these tools into medical practice has heralded the eHealth era. Intelligent health (iHealth) further builds on and expands eHealth by adding novel built-in data analysis approaches based on (1) incorporation of new technologies into clinical practice to enhance real-time self-monitoring, (2) extension of assessment to the patient's environment including caregivers, and (3) data processing using data mining to support medical decision making and personalized medicine. This will shift mental health care from a reactive to a proactive and personalized discipline.
Full-text available
Background Wearable activity trackers are newly emerging technologies with the anticipation for successfully supporting aging-in-place. Consumer-grade wearable activity trackers are increasingly ubiquitous in the market, but the attitudes toward, as well as acceptance and voluntary use of, these trackers in older population are poorly understood. Objective The aim of this study was to assess acceptance and usage of wearable activity trackers in Canadian community-dwelling older adults, using the potentially influential factors as identified in literature and technology acceptance model. MethodsA mixed methods design was used. A total of 20 older adults aged 55 years and older were recruited from Southwestern Ontario. Participants used 2 different wearable activity trackers (Xiaomi Mi Band and Microsoft Band) separately for each segment in the crossover design study for 21 days (ie, 42 days total). A questionnaire was developed to capture acceptance and experience at the end of each segment, representing 2 different devices. Semistructured interviews were conducted with 4 participants, and a content analysis was performed. ResultsParticipants ranged in age from 55 years to 84 years (mean age: 64 years). The Mi Band gained higher levels of acceptance (16/20, 80%) compared with the Microsoft Band (10/20, 50%). The equipment characteristics dimension scored significantly higher for the Mi Band (P
Full-text available
Background Sleep is a modifiable lifestyle factor that can be a target for efficient intervention studies to improve the quality of life and decrease the risk or burden of some chronic conditions. Knowing the profiles of individuals with poor sleep patterns is therefore a prerequisite. Wearable devices have recently opened new areas in medical research as potential efficient tools to measure lifestyle factors such as sleep quantity and quality. Objectives The goal of our research is to identify the determinants of poor sleep based on data from a large population of users of connected devices. Methods We analyzed data from 15,839 individuals (13,658 males and 2181 females) considered highly connected customers having purchased and used at least 3 connected devices from the consumer electronics company Withings (now Nokia). Total and deep sleep durations as well as the ratio of deep/total sleep as a proxy of sleep quality were analyzed in association with available data on age, sex, weight, heart rate, steps, and diastolic and systolic blood pressures. Results With respect to the deep/total sleep duration ratio used as a proxy of sleep quality, we have observed that those at risk of having a poor ratio (≤0.40) were more frequently males (odds ratio [OR]female vs male=0.45, 95% CI 0.38-0.54), younger individuals (OR>60 years vs 18-30 years=0.47, 95% CI 0.35-0.63), and those with elevated heart rate (OR>78 bpm vs ≤61 bpm=1.18, 95% CI 1.04-1.34) and high systolic blood pressure (OR>133 mm Hg vs ≤116 mm Hg=1.22, 95% CI 1.04-1.43). A direct association with weight was observed for total sleep duration exclusively. Conclusions Wearables can provide useful information to target individuals at risk of poor sleep. Future alert or mobile phone notification systems based on poor sleep determinants measured with wearables could be tested in intervention studies to evaluate the benefits.
Full-text available
Background: The experience sampling method (ESM) builds an intensive time series of experiences and contexts in the flow of daily life, typically consisting of around 70 reports, collected at 8-10 random time points per day over a period of up to 10 days. Methods: With the advent of widespread smartphone use, ESM can be used in routine clinical practice. Multiple examples of ESM data collections across different patient groups and settings are shown and discussed, varying from an ESM evaluation of a 6-week randomized trial of mindfulness, to a twin study on emotion dynamics in daily life. Results: Research shows that ESM-based self-monitoring and feedback can enhance resilience by strengthening the capacity to use natural rewards. Personalized trajectories of starting or stopping medication can be more easily initiated and predicted if sensitive feedback data are available in real time. In addition, personalized trajectories of symptoms, cognitive abilities, symptoms impacting on other symptoms, the capacity of the dynamic system of mental health to "bounce back" from disturbance, and patterns of environmental reactivity yield uniquely personal data to support shared decision making and prediction in clinical practice. Finally, ESM makes it possible to develop insight into previous implicit patterns of thought, experience, and behavior, particularly if rapid personalized feedback is available. Conclusions: ESM enhances clinical practice and research. It is empowering, providing co-ownership of the process of diagnosis, treatment evaluation, and routine outcome measurement. Blended care, based on a mix of face-to-face and ESM-based outside-the-office treatment, may reduce costs and improve outcomes.
Wearable electronics are emerging as a platform for next-generation, human-friendly, electronic devices. A new class of devices with various functionality and amenability for the human body is essential. These new conceptual devices are likely to be a set of various functional devices such as displays, sensors, batteries, etc., which have quite different working conditions, on or in the human body. In these aspects, electronic textiles seem to be a highly suitable possibility, due to the unique characteristics of textiles such as being light weight and flexible and their inherent warmth and the property to conform. Therefore, e-textiles have evolved into fiber-based electronic apparel or body attachable types in order to foster significant industrialization of the key components with adaptable formats. Although the advances are noteworthy, their electrical performance and device features are still unsatisfactory for consumer level e-textile systems. To solve these issues, innovative structural and material designs, and novel processing technologies have been introduced into e-textile systems. Recently reported and significantly developed functional materials and devices are summarized, including their enhanced optoelectrical and mechanical properties. Furthermore, the remaining challenges are discussed, and effective strategies to facilitate the full realization of e-textile systems are suggested.
This paper presents the comparison of sleep-wake classification using electroencephalogram (EEG) and multi-modal data from a wrist wearable sensor. We collected physiological data while participants were in bed: EEG, skin conductance (SC), skin temperature (ST), and acceleration (ACC) data, from 15 college students, computed the features and compared the intra-/inter-subject classification results. As results, EEG features showed 83% while features from a wrist wearable sensor showed 74% and the combination of ACC and ST played more important roles in sleep/wake classification.
Background As technology increasingly becomes an integral part of everyday life, many individuals are choosing to use wearable technology such as activity trackers to monitor their daily physical activity and other health-related goals. Researchers would benefit from learning more about the health of these individuals remotely, without meeting face-to-face with participants and avoiding the high cost of providing consumer wearables to participants for the study duration. Objective The present study seeks to develop the methods to collect data remotely and establish a linkage between self-reported survey responses and consumer wearable device biometric data, ultimately producing a de-identified and linked dataset. Establishing an effective protocol will allow for future studies of large-scale deployment and participant management. Methods A total of 30 participants who use a Fitbit will be recruited on Mechanical Turk Prime and asked to complete a short online self-administered questionnaire. They will also be asked to connect their personal Fitbit activity tracker to an online third-party software system, called Fitabase, which will allow access to 1 month’s retrospective data and 1 month’s prospective data, both from the date of consent. Results The protocol will be used to create and refine methods to establish linkages between remotely sourced and de-identified survey responses on health status and consumer wearable device data. Conclusions The refinement of the protocol will inform collection and linkage of similar datasets at scale, enabling the integration of consumer wearable device data collection in cross-sectional and prospective cohort studies.