WIRED: The impact of media and technology use on stress (cortisol)
and inﬂammation (interleukin IL-6) in fast paced families
Tamara D. Aﬁﬁ
, Nicole Zamanzadeh
, Kathryn Harrison
, Micelle Acevedo Callejas
Department of Communication at the University of California Santa Barbara, USA
Department of Communication Studies at the University of Iowa, USA
This study examined how technology and media use affect stress (cortisol) and inﬂammation (interleukin
IL-6) in dual earning parents and their adolescents. Sixty-two families reﬂected on their technology use
the past week and collected saliva on two consecutive days that week. Technology use had the greatest
effect on adolescents. Adolescents with greater phone use, general media exposure, and larger social
networks via Facebook had a greater rise in their cortisol awakening response (CAR) and higher IL-6.
Fathers' phone use and email were also associated with an increase in their CAR and IL-6. When
bedtime technology use was high, greater general media use was associated with an increase in CAR for
adolescents, but a decrease for fathers. Technology use did not signiﬁcantly affect cortisol diurnal rhythm
or mothers’biosocial markers. This study contributes empirical evidence of the physiological conse-
quences of technology use among family members and provides potential theoretical explanations for
Published by Elsevier Ltd.
In 2015, a Pew survey estimated that 68% of Americans owned a
smartphone and 45% owned tablets (Anderson, 2015). Although
people are increasingly dependent on these portable devices that
provide mobility, the majority of media is still consumed in the
home (Common Sense Media, 2013; Nathanson, 2015). Within the
family, parents experience the blurring of work and personal life
through the ﬂexibility that Internet-connected mobile devices offer
and often struggle to navigate these domains (Nam, 2014). Mean-
while, adolescents are exploring identity, completing schoolwork,
and developing and maintaining friendships via technology (boyd,
2015). Indeed, technology has become an inevitable and important
part of the fabric of most American families.
Despite the increased adoption of technology, research suggests
it has mixed effects on well-being. Technology can help family
members communicate efﬁciently, multitask, regulate moods, and
facilitate social support (Carvalho, Francisco, &Relvas, 2015). For
adolescents, developing in the digital age has enabled increased
autonomy through easy access to peer networks, but it can also
increase parental monitoring (boyd, 2015). Similarly, while
technology can expedite communication and information sharing
that enhances positive emotional connections (Carvalho et al.,
2015), it can also contribute to surveillance and diminished face-
to-face communication that can fuel stress in families (McDaniel,
2015). Heavy technology use in particular has been associated
with depression, loneliness, anxiety, and narcissism (Rosen,
Whaling, Rab, Carrier, &Cheever, 2013).
Although scholars have begun to understand how technology
use affects individuals psychologically, much less is known about
how it affects them physiologically, especially within families
where multiple members are using technology simultaneously. The
current study addresses this void in the literature by focusing on
how technology use, and the type of technology used, affect bio-
logical stress responses (measured through cortisol) and immune
systems (measured through the pro-inﬂammatory marker, inter-
leukin IL-6) of dual earning parents and their adolescents (ages
13e18). An affordances approach is used to explore distinctions in
both the role and the effects of technology on family members’
stress. The impact of technology and media on stress in families is a
pressing issue given the increasing number of dual career house-
holds with adolescents in which all members tend to live fast-
paced lives (i.e., busy lifestyles replete with obligations of
balancing work and family and extracurricular activities) and
where everyone is “wired”most of their waking hours. We refer to
The researchers would like to thank the families who participated in the study
and the research assistants who helped with data collection.
E-mail address: taﬁﬁ@comm.ucsb.edu (T.D. Aﬁﬁ).
Contents lists available at ScienceDirect
Computers in Human Behavior
journal homepage: www.elsevier.com/locate/comphumbeh
0747-5632/Published by Elsevier Ltd.
Computers in Human Behavior 81 (2018) 265e273
these families as fast-paced families or FPF.
1. Technology and media as tools that can affect stress
Media psychology provides a particularly useful theoretical
approach to technological effects (Rutledge, 2012, pp. 43e61).
Scholars who adopt a media psychology approach often model how
affordances of media and technology constrain and/or enhance
social perceptions, affect cognitive processing, and shape social,
emotional, and psychological outcomes. Importantly, these effects
also include physiological outcomes. Technology and media are
used here as overarching terms that include a multitude of devices,
as well as activities and new media that saturate the environment.
Through the attributes of portability and connection to the Internet,
technological devices afford perpetual contact and co-presence of
social networks and unprecedented amounts of information (Rice
&Katz, 2003). Therefore, these valuable tools can create tremen-
dous demands and ultimately stress that can affect the physiolog-
ical health of their users. In the current study, we focus on the
interconnections between one's social environment (e.g., family
and technology/media use) and biological stress responses (e.g.,
cortisol as a marker of stress and IL-6 as a marker of inﬂammation
or immune strength) or what researchers often refer to as “bioso-
cial markers”of stress.
Technologies, such as the mobile phone, have become essential
tools for organizing and managing information, social coordination,
and maintaining relationships (Rice &Hagen, 2010, pp. 2e39). This
is likely particularly true for FPF who require management and co-
ordination of everyone's busy schedules. Portable devices connected
to the Internet (i.e., social media and chatting or messaging appli-
cations) can also connect people to their ofﬂine social support net-
works and diverse online social support networks at multiple times
throughout the day. Riva, Ba~
nos, Botella, Wiederhold, and Gaggioli
(2012) also note that there are positive effects of technology on af-
fective quality, engagement/actualization, and connectedness. They
found that these technologies can facilitate positive emotions, allow
people to be more engaged and active in theirinterests via inducing
ﬂow or total involvement, and increase social presence and social
capital. Yet, the multitude of new media and technology, which
afford people constant access to engaging information, entertain-
ment, and social connections, can lead to overload.
The technological advances that facilitate connection can also be
stressful for adults and adolescents if they overwhelm their ca-
pacity and/or desires for information and communication. For
instance, regardless of age, the pressure to be constantly available
to social networks alone can result in communication overload
(Stephens et al., 2017). Reinecke et al. (2016) found that across the
lifespan, people reported experiencing digital stress or digitally
caused strain. Their participants also engaged in communication
due to fears of missing out (FOMO) on information and the social
pressure to be responsive, which predicted anxiety, burnout, and
depression. However, taking an affordances approach, these expe-
riences of stress may also be linked to particular technological use
that are advantageous for people's goals at their particular life-
stages. Lee, Son, and Kim (2016) also found that social
networking site (SNS) use can result in information, communica-
tion, and technological overload, and later SNS fatigue. In both
studies, fatigue and stress existed across a range of ages, but was
highest among youth (29 or below). Adults, however, often report
feeling a “technostress”due to new technologies embedded in
work life, including virtual ofﬁces and increased ﬂexibility to work
from home (Jarvenpaa &Lang, 2005).
Technologies can also create stress among family members by
interfering with their relationships. “Technoference,”which is a
term for the experience relational partners have when technology
interrupts conversations and other important shared experiences,
has been found to produce technology-related conﬂicts between
romantic partners and between parents and children (McDaniel,
2015). Conﬂicts can inhibit one's quality of life and family re-
lationships and ultimately induce physiological stress (Kulhman,
Repetti, Reynolds &Robles, 2016). McDaniel and Coyne (2014)
found that the majority (70%) of romantic partners perceived
computers, mobile phones and television to “sometimes”or “more
often”interfere with their relationship. More frequent “technofer-
ence”has also been associated with lower overall well-being (e.g.,
lower reported relationship satisfaction, greater depressive symp-
toms, and lower life satisfaction). Although largely studied among
romantic partners or parent-child dyads, Turkle (2011) remarks on
the emotional disconnect among family members at the dinner
table when their devices captivate them. Technology can interrupt
essential face-to-face interactions and produce lower life satisfac-
tion and technology-related conﬂict and stress in families if it is not
Moreover, technology can interrupt, distract, or delay in-
dividuals' vital behaviors such as eating, exercising and sleeping,
which can increase stress. One of the most robust effects of tech-
nology and media usage is on sleep duration and sleep disturbances
(Cain &Gradisar, 2010). Sleep and quality of sleep are particularly
important to the body's recovery from daily stress. Yet, research
suggests that people's technology interfere with their very basic
and important functions. Thom
ee, Harenstam and Hadberg (2011)
discovered that high mobile phone usage was related to more
sleep disturbances and depression. Lemola, Perkinson-Gloor,
Brand, Dewald-Kaufmann, and Grob (2014) also discovered that
adolescents who owned a smartphone had later bed times. Further,
adolescents' reporting of late night usage of smartphones in bed
was associated with depressive symptoms. Overall, electronic me-
dia use was negatively correlated with sleep duration and posi-
tively correlated with disturbances and depression. By tampering
with sleep, technology might hinder the body's ability to recover
from stress and prepare for oncoming daily stressors. Technology's
ability to overload people with information and disrupt life,
including sleep, provide evidence for its role as a stressor that can
affect family members' biological stress responses.
2. Technology and physiological stress in families
To understand the impact of technology on physiological stress,
it is ﬁrst important to provide a brief overview of how the mind and
body respond to stress. In the face of a stressor, challenge or threat,
the body responds via the hypothalamic-pituitary-adrenal (HPA)
axis's secretion of glucocorticoids (Juster, McEwen, &Lupien, 2010).
After the brain assesses that something is stressful, the
sympathetic-adrenal-medullary (SAM) releases adrenaline acti-
vating the HPA axis. Stimulating the production of corticotrophin
releasing hormone (CRH) leads the anterior pituitary to release the
adrenocorticotropic hormone (ACTH), which ﬁnally activates the
adrenal cortex leading to cortisol. Cortisol (and other hormones and
enzymes) energizes the body and prepares it to respond to an
impending stressor (e.g., “ﬂight or ﬁght”responses).
Cortisol also has a consistent diurnal pattern or cycle in healthy
children and adults (Saxbe &Repetti, 2010). A healthy diurnal
pattern has a high initial cortisol level that peaks approximately
30 min after waking. People need a certain level of cortisol after
they wake to combat the stress they face during their day (Munck,
Guyre, &Holbrook, 1984). The rise in cortisol that occurs approxi-
mately 30 min after waking is called the cortisol awakening
response or CAR (Adam et al., 2014). Nevertheless, too much or too
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e273266
little of a rise in CAR has been associated with poor physical and
mental health (see Fries, Dettenborn, &Kirschbaum, 2009). The
CAR is followed by a progressive reduction or steep drop in cortisol
throughout the morning with a nadir in the late afternoon, reaching
its lowest point at bedtime (Saxbe &Repetti, 2010). A smaller drop
in cortisol throughout the day or the ﬂatness of people's diurnal
rhythms is indicative of worse health (Doane et al., 2013).
The impact of stress on the HPA axis often depends upon the
nature of the stress and how the physiological stress responses are
measured. In the presence of acute perceived environmental
stressors, individuals often experience elevated levels of cortisol.
Chronic or recurring stress can comprise daily functioning of the
HPA axis (Juster et al., 2010). The HPA axis can become dysregulated
due to allostatic load, making the diurnal cortisol rhythm ﬂatter,
more sensitive to stress and challenge, poorer at recovery from
stressors, or non-responsive (McEwen, 1998).
Allostasis is the body's natural response to stress and is a dy-
namic process through which multiple biological systems respond
and adapt to environmental stressors (McEwen, 1998). This adap-
tation is necessary and important for humans to adapt to their
environment's demands. However, the body's ability to adapt
effectively to stress can be compromised when it is overstimulated.
Allostatic load is the cumulative “wear and tear”of chronic and/or
enduring stress on biological, physiological, and psychological
systems (Repetti, Robles, &Reynolds, 2011). Although we are not
measuring allostatic load in the current study, the overuse of
technology could contribute to acute and chronic stress and anxiety
through communication overload and technoference and manifest
in people's biological stress responses.
Chronic stress can weaken the body's immune system, making it
more susceptible to disease. Research shows that psychological
stress (from a wide range of sources, including conﬂict, relationship
dissatisfaction, and loneliness) can dysregulate people's stress re-
sponses and inhibit their immune functioning (see Graham et al.,
2009). Stress can elevate the productive of proinﬂammatory cyto-
kines, one of which is IL-6. Inﬂammation is an essential way the
body responds to infection or injury, but the over production of it is
thought to be a crucial connection between stress and poor health
(Jaremka, Lindgren, &Kiecolt-Glaser, 2013). Chronic inﬂammation
is predictive of a host of diseases and disorders, including type II
diabetes, obesity, cardiovascular disease, high blood pressure, and
premature aging (McEwen &Wingﬁeld, 2003).
Different types of stress, however, seem to have different effects
on these responses. Environmental stress can affect the acute
functioning of the HPA axis, the diurnal variability of the system,
and the production of proinﬂammatory cytokines (e.g., Desantis
et al., 2007). For the most part, too much stress tends to
contribute to the over production of IL-6, which then contributes to
even greater stress (Jaremka et al., 2013). Acute or more moderate
stress can exacerbate the diurnal variability of cortisol and often
produces a greater rise in CAR, whereas more severe and chronic
stress often produces a blunting of the diurnal variability and
cortisol awakening response (CAR) (Stetler &Miller, 2005).
Chronically high and chronically low cortisol levels have been
associated with psychological problems such as anxiety and
depression, whereas moderate levels are more indicative of adap-
tive functioning (Gordis, Granger, Susman, &Trickett, 2006).
Among anxious individuals, however, morning cortisol levels
should be exceedingly high and remain somewhat higher in the
evening (Doane et al., 2013). Although there is very little research
connecting technology use to cortisol or IL-6, Wallenius et al. (2010)
found that school age children (39 ten year olds and 3313 year olds)
who used greater amounts of technology the preceeding day (3 or
more hours) had a diminished or lower CAR the next morning. As
these authors speculated, too much technology use likely
overwhelmed the children's physiological stress system and its
ability to function properly. Long hours of technology use day after
day might function as a natural stressor that builds over time.
Correspondingly, technology can be used to facilitate connection
in physical proximity (Carvalho et al., 2015) or act as a tool for
isolation and separation even while in physical proximity (Turkle,
2011). In particular, the general shared experience of high tech-
nology use could produce communication overload, which should
be associated with higher stress among family members. But,
because of the lack of research, the exact nature of how technology
use affects cortisol patterns and IL-6 is unclear. Thus, we propose
the following research question:
: In what ways does the amount of overall technology used
by parents and adolescents during the week affect their CAR, IL-
6, and cortisol diurnal rhythms?
In addition, the types of technology and media parents and
adolescents use could produce different levels of stress. An affor-
dances theoretical framework provides a systematic approach to
exploring the characteristics of media that contribute to their social
and personal uses and effects. Speciﬁcally, work-related email use
on mobile devices by parents could be most disruptive by dis-
placing parents’time and attention and most predictive of
technology-related stress (McDaniel, 2015). This might represent
communicative overload as an outcome of a lack of boundaries
between work and home (Jaavrenpa &Lang, 2005). Furthermore,
higher social media use (such as Facebook) affords higher social
connectivity, visibility, accessibility, edibility, persistence (of mes-
sages), searchability and social feedback (Treem &Leonardi, 2012,
pp. 143e189). Although these affordances have beneﬁts, they can
also be stressors (Fox &Moreland, 2015; Morin-Major et al., 2016).
Given the affordances literature or different functions of technol-
ogy, it is logical to assume that the forms of technology and media
use matter in their impact on cortisol and IL-6.
H1. Technology used (i.e., mobile phone usage, email usage, gen-
eral media use, or social network size on Facebook) by parents and
adolescents during the week will differ in their effects on cortisol
diurnal rhythms, CAR, and IL-6.
Finally, because research indicates that high technology use is
related with late night usage, sleep disturbances and decreased
sleep quality (Cain &Gradisar, 2010; Lemola et al., 2014), we hy-
pothesize the following:
H2. Overall technology/media use will interact with nighttime
screen/cell phone use to predict CAR, such that those with the
greatest technology/media use and nighttime screen/cell phone
use will have the highest CAR, and those with the lowest technol-
ogy/media use and nighttime screen/cell phone use will have the
Families reﬂected on their technology use the past week. They
also reported on their night time technology use, sleep, and
collected four saliva samples throughout the day on two consecu-
tive days in the middle of that week to test for cortisol and IL-6.
Sixty-two heterosexual couples (Mage ¼44.68) and one of their
adolescent children (ages 13e18 ; M¼14.8; 31 sons and 31
daughters) participated in this study. All the couples were married
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e27 3 267
(94%) except for four, who were cohabitating. Eight (13%) of the
couples were remarried. Most adolescents and parents were
Caucasian (adolescent n¼55 or 89%; parents ¼115 or 93%). The
median household income of the families was $113,000. Nine of the
parents had a high school degree (7%), 24 (19%) had an Associates'
Degree or some college, 47 (38%) had a Bachelor's degree, and 41
(33%) had an advanced degree (MA, Ph.D., MD). The parents had an
average of three children and worked an average of 44.7 h per
week. The adolescents spent an average of 11 h a week in extra-
curricular activities and had 4.07 social media accounts. Parents
reported 2.07 social media accounts.
The families resided in the Midwest part of the U.S. They were
recruited through a university-wide email listserv at a research
university, advertisements sent to employees at large businesses,
Craigslist, and network sampling. Once the families called the
researcher, the researcher explained the study, screened members,
and gained verbal consent. Both parents needed to be employed full
time (min. of 40 h a week) and have two or more children still living
in their home. All parents needed to be in a romantic relationship
and living together and the adolescent needed to be currently
residing with them. Family members were excluded if they had a
health condition or were taking medications that could have
altered their hormones.
Data collection for each family lasted one week. Each family
member completed an initial survey on the same weekend, daily
diary logs the following Monday through Friday, and an exit survey
at the end of the week on the weekend. In this exit survey, par-
ticipants reported about their technology and media use during the
last week. All links to the online surveys and diary logs, as well as
reminders, were emailed to the family members individually each
day. The families choose two consecutive days in the middle of the
week to collect their saliva samples.
Upon consent, families were mailed details of the study, saliva
collection procedures, collection materials marked by days and
times, medical forms, and cold packets. Parents and adolescents
collected saliva by passively drooling into a small vial. Families
were asked not to drink alcohol or visit the dentist within 48 h
before their saliva collections. They were also instructed not to
exercise rigorously, consume caffeine, use illegal drugs, or smoke
on the days of collection. They were told not to brush their teeth,
eat, or drink anything an hour before their saliva collections. Saliva
samples were collected immediately upon waking in the morning
before their feet touched the ﬂoor, 30 min after waking, at noon (in
a private room, before lunch), and right before bedtime (resulting in
1488 saliva samples). The family was instructed to freeze their
saliva samples in their home freezer immediately after collection. If
they were at work or school, participants were asked to put their
sample in a freezer or use the frozen packet to keep it cold.
Several procedures were established to help ensure compliance
and accuracy in the data collection. Before the family began the
study, a researcher either called the family using regular voice
dialing or FaceTime (allowing the researcher to visually demon-
strate how to collect the saliva). The instructions were also written
in a letter, where the saliva collection was also visually depicted. A
wake-up text reminder was also sent to each family member the
morning of the saliva collections using their self-reported expected
wake-up times. Immediately after collecting every saliva sample,
the participants texted the researcher “collected,”providing a time
stamp. Each participant was compensated $35. The data were
collected during the academic year and not during a break or family
vacation. Upon retrieval from the family's home, the saliva samples
were immediately transferred to a freezer in the laboratory (20
C). The samples were then shipped overnight to Salimetrics where
they were frozen at 80
C until analyzed. All of the samples were
assayed for salivary cortisol in duplicate by enzyme immunoassay
(Granger et al., 2012). Intra- and inter-assay coefﬁcients of variation
should be less than 10 and 15%, respectively. Cortisol is reported in
micrograms per deciliter (
3.3.1. Media and Technology Use Scale
An adapted version of the Media and Technology Use and Atti-
tudes scale (Rosen, Whaling, Carrier, Cheever &Rokkum, 2013)was
completed by the parents and adolescents at the end of the study.
We asked participants about the frequency in which they engaged
in each of the following activities in the last week: emailing (on all
devices; 4 items; adolescent
¼0.86), phone use (including all smart phone use, calling, texting,
music, pictures, internet, etc.; 14 items; adolescent
¼0.93), general media use (i.e., TV viewing,
gaming, media sharing, internet searching; 13 items; adolescent
¼0.75), and social media use
(e.g., “checked Facebook or other social networks,”“post status
updates”“post photos”“read postings”; 9 items; adolescent
¼0.95). These subscale items
were averaged and varied along the following anchors: 1 (never), 2
(once this past week), 3 (several times this past week), 4 (once a
day), 5 (several times a day), 6 (once an hour), 7 (several times an
hour), and 8 (always). Finally, we included the “Facebook”subscale
that more speciﬁcally asked about participants’networks on
Facebook. Participants were ﬁrst asked if they had a Facebook ac-
count (yes ¼1; no ¼0). They were then asked about their social
network on Facebook that ranged on a scale from 0 (none) to8 (751
or more) (e.g., “How many total friends do you have on Facebook,
including friends you may not know?”“How many Facebook
friends do you know in person?”adolescent
3.3.2. Night-Time Technology Use
Five items were created for this study to assess participants’
reports of their typical technology use at bedtime in the initial,
entry survey. Participants indicated whether or not (i.e., indicated
yes ¼1orno¼0) they usually “fall asleep at night looking at a
screen (or right after looking at a screen while still in bed) of some
sort (e.g., TV, laptop, cell phone, iPod, etc.),”“use your cell phone as
an alarm clock,”“silence your phone when you sleep (reverse
coded),”“check your cell phone in the middle of the night”and
“receive instant notiﬁcations on your cell phone (for email, texts,
apps, etc …) in general?”
3.3.3. Hours of Sleep
Hours of sleep was a signiﬁcant control variable in the models.
Participants were asked in the diary logs the two days of their saliva
collections how many hours of sleep they got the night before. The
two evenings were averaged.
3.3.4. Physiological Measures
IL6 was measured with the wake-up saliva on day two of the
collection. For cortisol, the average across each of the time points
for the two collection days was calculated. CAR was measured as
the “30 min after waking up score”minus the “wakeup score.”The
diurnal rhythm of cortisol was assessed across the four time points
(wake up, 30 min after waking up, noon, and bedtimedcoded 0, 1,
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e273268
Means and standard deviations for the variables are provided in
Table 1. On average, the parents and adolescents reported moderate
levels of technology and media use. Facebook networks averaged
between “51e100”to “101e17 5 ”friends. Most notably, mothers
reported using email and Facebook more than fathers and adoles-
cents and had a larger friendship network on Facebook. Fifty-eight
percent (n¼36) of adolescents, forty-eight percent (n¼30) of
fathers, and eighty-ﬁve percent (n¼53) of mothers indicated they
had a Facebook account. The cortisol and IL6 scores were positively
skewed and were transformed with a natural log transformation.
Multilevel modeling (using SPSS MIXED) with a random inter-
cept was used to account for the nested nature of the data and the
growth curve for the diurnal rhythm of cortisol. This statistical
approach allowed for a detailed analysis of the effects of variables at
the level of the individual family member (father, mother, adoles-
cent–level 1), nested within a family (level 2). For each outcome, we
ﬁrst tested an unconditional or intercepts only model to obtain an
estimate of the intraclass correlation (ICC). If there was signiﬁcant
variance in the outcome variable, we then proceeded to build the
models by ﬁrst testing control variables and their interaction with
family role. Numerous control variables were examined (e.g.,
caffeine intake, number of children's activities, number of hours
working, mental health, income, education), but the only one that
was signiﬁcant was participants' averaged hours of sleep the nights
before their saliva collections. Therefore, this was the only control
variable included in the models. Finally, predictor variables were
entered, and interactions with family role were tested.
Separate intercepts for mothers and fathers were created by
inserting dummy codes for mothers and fathers into the model
instead of including a traditional intercept. The adolescent was the
omitted group in the dummy coding. The models were then run
again, making mothers the omitted group to determine all possible
combinations of results for family role. After these analyses were
complete, mothers, fathers and adolescents were then combined
into one variable with the intercept included in the ﬁxed effect (and
the random effect for the growth curve models). For the growth
curve models, separate slopes for mothers, fathers, and children
were also created by crossing them with time. The predictors were
grand mean centered.
4.1. Results for CAR
In the ﬁnal model for email use and CAR, there was a signiﬁcant
interaction for email use and fathers when compared to mothers
B¼0.01, 95%CI [0.001, 0.02], t(102.47) ¼2.24, p<0.05, but no
signiﬁcant interaction for email use and mothers when compared
to adolescents B¼0.01, 95%CI [-0.02, 0.01], t(61.26) ¼0.67, ns,or
email use for fathers when compared to adolescents B¼0.01, 95%CI
[-0.004, 0.02], t(59.15) ¼1.25, ns. This indicates that after control-
ling for the average number of hours slept at night, fathers who
reported greater email use had a greater rise in their CAR compared
to mothers. The random intercept variance remained signiﬁcant in
the models, Var ¼0.25e26, p<0.01. The test for whether these
family roles and email use were signiﬁcantly different from each
other was not signiﬁcant B¼0.007, 95%CI [-0.007, 0.008],
t(101.28) ¼0.20, ns.
In the ﬁnal model for phone use and CAR, there were no sig-
niﬁcant interaction for phone use and fathers when compared to
mothers B¼0.004, 95%CI [-0.01, 0.01], t(95.54) ¼0.06, ns, no sig-
niﬁcant interaction for phone use and mothers compared to ado-
lescents B¼0.005, 95%CI [-0.02, 0.01], t(60.54) ¼0.61, ns, and no
signiﬁcant interaction for phone use for fathers compared to ado-
lescents B¼0.01, 95%CI [-0.02, 0.003], t(108.39) ¼1.43, ns.
However, there was a main effect for adolescents for phone use at
p¼0.05, B¼0.01, 95%CI [-0.0003, 0.02], t(59.94) ¼1.95, indicating
that (controlling for sleep) as adolescents increased their phone
use, it corresponded with a slight rise in CAR. The random intercept
variance was signiﬁcant, Var ¼0.22-0.24, p<0.05.
In the ﬁnal model for general media use and CAR, there was a
signiﬁcant interaction for general media use for fathers when
compared to adolescents B¼0.02, 95%CI [-0.03, 0.002],
t(110.82) ¼2.31, p<0.05, a signiﬁcant interaction for general
media use for mothers compared to adolescents B¼0.02, 95%CI
[-0.03, -.0001], t(101.40) ¼2.02, p<0.05, but no signiﬁcant
interaction for general media use for fathers compared to mothers
B¼0.001, 95%CI [-0.02, 0.01], t(104.85) ¼0.13, ns. This indicates
that after controlling for the number of hours slept at night, ado-
lescents who reported greater general media use had a greater rise
in their CAR compared to mothers and fathers. There was also a
main effect for adolescents for general use B¼0.09, 95%CI [0.002,
0.17], t(59.49) ¼2.04, p<0.05, suggesting that as adolescents
increased their general media use, it corresponded with a rise in
CAR. The random intercept variance was signiﬁcant, Var ¼0.27e29,
p<0.001. The test for whether these family roles were signiﬁcantly
different than each other was signiﬁcant B¼0.01, 95%CI -0.02,
-.003], t(127.91) ¼2.65, p<0.01.
In the ﬁnal model for Facebook use and CAR, there was a sig-
niﬁcant interaction for Facebook and mothers compared to ado-
lescents B¼0.02, 95%CI [-0.04, -.01], t(84.16) ¼2.70, p<0.01, but
no signiﬁcant interaction for Facebook for fathers compared to
adolescents B¼0.01, 95%CI [-0.03, 0.002], t(96.13) ¼1.65, ns,
and no signiﬁcant interaction for Facebook for fathers compared to
mothers B¼0.01, 95%CI [-0.01, 0.03], t(92.46) ¼1.17, ns. This in-
dicates that after controlling for the number of hours slept at night,
adolescents who reported greater Facebook use had a greater rise in
their CAR compared to mothers. There was also a main effect for
adolescents for Facebook B¼0.01, 95%CI [0.002, 0.02],
t(58.45) ¼2.49, p<0.05, suggesting that (controlling for sleep), as
adolescents increased their Facebook use, it corresponded with a
rise in CAR. The random intercept variance was signiﬁcant,
Var ¼0.27, p<0.01. The test for whether these family roles were
signiﬁcantly different than each other was signiﬁcant B¼0.01,
95%CI [-0.02, -.0009], t(142.02) ¼2.20, p<0.05.
Descriptive statistics for predictor variables and IL6 and cortisol by family member.
Mothers Fathers Adolescent
Mean Hours Slept 6.99a 0.87 6.67b 0.98 7.44ba 1.03
Email Use 5.25ab 1.64 4.35bc 1.55 3.80ac 1.33
Phone Use 4.02 1.28 3.72a 1.28 4.30a 1.20
General Media Use 2.63a 1.11 2.66b 1.05 3.31ab 1.08
Social Media 3.56a 1.54 2.41ab 1.48 3.79b 1.49
Facebook 2.54ab 0.93 2.17b 1.22 2.04a 1.38
IL6 28.11 26.00 48.78 93.34 62.61 129.08
CAR 0.16þ0.10 0.10þ0.01 0.19þ0.04
Cortisol Wake 0.29 0.12 0.29 0.15 0.29 0.12
Cortisol 30 Min after Wake 0.45þ0.22 0.39aþ0.14 0.48a 0.16
Cortisol Noon 0.11a 0.07 0.14 þa 0.07 0.12þ0.06
Cortisol Bedtime 0.08 0.10 0.07 0.10 0.07 0.07
Note: IL6 and Cortisol scores are the raw scores before being transformed. Scale
anchors for email use to social media range from 1e8 [1 (never), 2 (once this past
week), 3 (several times this past week), 4 (once a day), 5 (several times a day), 6
(once an hour), 7 (several times an hour, and 8 (always)]. The speciﬁc“Facebook”
scale asked participants about their network on Facebook a scale from 0 (none) to 8
(751 or more) (e.g., “How many total friends do you have on Facebook, including
friends you may not know?”“How many Facebook friends do you know in person?”
Facebook networks averaged between “51e100”to “101e175”friends). Corre-
sponding letters across rows indicate signiﬁcant differences between family
members for that variable at a minimum of p<0.05. þ¼p<0.10.
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e27 3 269
In the ﬁnal model for social media use and CAR, there were no
signiﬁcant interaction for social media use and fathers when
compared to mothers B¼0.003, 95%CI [-0.01, 0.01], t(99.81) ¼0.51,
ns, no signiﬁcant interaction for social media use and mothers
compared to adolescents B¼0.006, 95%CI [-0.01, 0.006],
t(108.04) ¼1. 0 0 , ns, and no signiﬁcant interaction for social media
use for fathers compared to adolescents B¼0.003, 95%CI [-0.01,
0.008], t(105.89) ¼0.49, ns. The random intercept variance was
signiﬁcant, Var ¼0.23e25, p<0.01.
4.2. Results for IL-6
In the ﬁnal model for email use and IL-6, there was no signiﬁcant
interaction for email use for fathers compared to mothers B¼0.09,
95%CI [-0.03, 0.20], t(103.32) ¼1.51, ns, no signiﬁcant interaction for
email use for mothers compared to adolescents B¼0.0007, 95%CI
[-0.13, 0.13], t(88.01) ¼0.01, ns, or email use for fathers compared to
adolescents B¼0.10, 95%CI [-0.05, 0.24], t(101.16) ¼1.40, ns.
However, there was a main effect for fathers and email use, indi-
cating that (controlling for sleep) the more fathers used their email,
the more their IL-6 increased. The random intercept variance was
not signiﬁcant, Var ¼0.05, ns.
In the ﬁnal model for phone use and IL-6, there was a signiﬁcant
interaction for phone use for fathers compared to adolescents
B¼0.18, 95%CI [0.01, 0.34], t(103.86) ¼2.15, p<0.05, no signiﬁcant
interactions for phone use for mothers compared to adolescents
B¼0.03, 95%CI [-0.12, 0.19], t(94.70) ¼0.44, ns, and the interaction
for phone use for fathers compared to mothers was approaching
signiﬁcance B¼0.13, 95%CI [-0.006, 0.28], t(95.31) ¼1.90, p¼0.06.
After controlling for sleep, adolescents who reported greater phone
use had a greater increase in their IL-6 compared to fathers. Fathers
who also reported greater phone use had a slightly greater increase
in their IL-6 compared to mothers. The random intercept variance
was not signiﬁcant, Var ¼0.07, ns. The test for whether these family
roles were signiﬁcantly different than each other was not signiﬁ-
cant B¼0.05, 95%CI [-0.04, 0.15], t(93.74) ¼1.01, ns.
In the ﬁnal model for general media use and IL-6, there was no
signiﬁcant interaction for general media use for fathers when
compared to mothers B¼0.07, 95%CI [-0.10, 0.25], t(103.21) ¼0.86,
ns, no signiﬁcant interaction for general media use and mothers
when compared to adolescents B¼0.04, 95%CI [-0.13, 0.22],
t(95.81) ¼0.47, ns, and no signiﬁcant interaction for general media
use and fathers when compared to adolescents B¼0.13, 95%CI
[-0.07, 0.32], t(105.93) ¼1.26, ns. Therefore, there were no signiﬁ-
cant effects for general media use on IL-6. The random intercept
variance was not signiﬁcant, Var ¼0.03, ns.
In the ﬁnal model for Facebook use and IL-6, there was a sig-
niﬁcant interaction for Facebook for mothers compared to adoles-
cents B¼0.16, 95%CI [-0.32, 0.006], t(108.88) ¼2.06, p<0.05, an
interaction for Facebook and fathers compared to adolescents that
was approaching signiﬁcance B¼0.15, 95%CI [-0.32, 0.03],
t(91.14) ¼1.68, p¼0.09, but no signiﬁcant interaction for Facebook
for fathers compared to mothers B¼0.03, 95%CI [-0.22, 0.15],
t(91.77) ¼0.36, ns. This indicates that after controlling for the
number of hours slept at night, adolescents who reported greater
Facebook use had a greater increase in their IL6 compared to
mothers and fathers. There was also a main effect for adolescents
for Facebook B¼0.12, 95%CI [0.01, 0.23], t(54.21) ¼2.25, p<0.05,
suggesting that (controlling for sleep) as adolescents increased
their Facebook use, it corresponded with an increase in IL-6. The
random intercept variance was not signiﬁcant, Var ¼0.09-0.13, ns.
The test for whether these family roles were signiﬁcantly different
than each other was signiﬁcant B¼0.11, 95%CI [-0.21, -.01],
t(134.98) ¼2.35, p<0.05.
In the ﬁnal model for social media use and IL-6, there was no
signiﬁcant interaction for social media use for fathers when
compared to mothers B¼0.03, 95%CI [-0.10, 0.16], t(96.61) ¼0.40,
ns, no signiﬁcant interaction for social media use and mothers
when compared to adolescents B¼0.04, 95%CI [-0.17, 0.08],
t(100.39) ¼0.69, ns, and no signiﬁcant interaction for social media
use and fathers when compared to adolescents B¼0.002, 95%CI
[-0.14, 0.14], t(106.93) ¼0.03, ns. Therefore, there were no signiﬁ-
cant effects for social media use on IL-6. The random intercept
variance was not signiﬁcant, Var ¼0.04-0.05, ns.
4.3. Results for CAR and nighttime technology use and general
The results also revealed a signiﬁcant three-way interaction for
general media use, night time technology use and fathers
compared to adolescents B¼0.01, 95%CI [-0.02, -.006],
t(54.42) ¼2.06, p<0.05, but no signiﬁcant interaction for general
media use, night time technology use and mothers compared to
adolescents B¼0.003, 95%CI [-0.009, 0.02], t(57.86) ¼0.49, ns. The
three-way interaction for general media use, night time technology
use and fathers compared to mothers was approaching signiﬁcance
B¼0.01, 95%CI [-0.03, 0.001], t(105.10) ¼1.83, p¼0.07. The
random intercept variance remained signiﬁcant in the ﬁnal models,
Var ¼0.26-0.27, p<0.01. Follow-up regression analyses, controlling
for sleep, were then conducted to examine the nature of the as-
sociation between general media use and CAR for fathers, mothers,
and adolescents. Night time use was broken up into low and high
levels based upon a mean split of the data. These analyses revealed
the when night time technology use was high, general media use
was associated with a signiﬁcant rise in CAR for adolescents,
¼0.34, t ¼2.39, p<0.05, but a signiﬁcant and sharp decrease in
CAR for fathers,
¼0.72, t ¼2.73, p<0.05, and no signiﬁcant
association for mothers,
¼0.06, t ¼0.25, ns. The associations
between general media use and CAR for adolescents,
t¼0.56, ns, fathers
¼0.06, t ¼0.42, ns, and mothers
t¼0.10, ns, were not statistically signiﬁcant when night time use
4.4. Results for diurnal rhythm
The results for the unconditional growth curve model indicated
that there was not sufﬁcient variance across family members'
cortisol over time. Adding in technology as predictors also showed
a lack of convergence, suggesting that there was not enough dif-
ference or change in the family members’diurnal rhythms.
This study examined the physiological consequences of tech-
nology use in fast-paced families. Few studies have tested how
technology use is associated with people's biological stress markers
(i.e., cortisol) (for exceptions, see Brom et al., 2014; Gentile, Bender,
&Anderson, 2017; Heo et al., 2017; Morin-Major et al., 2016;
Wallenius et al., 2010). To our knowledge, this is the ﬁrst study to
assess the association between technology use and inﬂammation.
This study is also important in that it accounts for individual family
members' technology use and biosocial markers as embedded
within larger technology use within the family system. The results
showed that technology affected family members differently,
largely as a function of the types of technology used. The most
evident ﬁnding was that technology had the greatest effect on
adolescents' CAR and IL-6 compared to mothers and fathers. Ado-
lescents with higher phone use, greater general media exposure,
and larger social network sizes via Facebook demonstrated a
greater rise in their CAR and higher rates of IL-6.
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e273270
Even though adolescents' technology use was most connected
to their biosocial stress markers, fathers were also affected. Fathers
who reported greater phone use had a slightly greater increase in
their IL-6 compared to mothers. Fathers who reported greater email
use had a stronger rise in their CAR compared to mothers. In
addition, the more that fathers used their email, the more they
experienced an increase in their IL-6. Finally, when night time
technology use was high, general media use was associated with a
signiﬁcant rise in CAR for adolescents, but a signiﬁcant decrease in
CAR for fathers. Technology use did not signiﬁcantly affect cortisol
diurnal rhythm nor did have any signiﬁcant effect on mother's
biosocial stress markers.
5.1. Interpreting the physiological consequences of technology use
Overwhelmingly, technology use affected adolescents' CAR
more than their parents. Interestingly, this technology use did not
inﬂuence any of the family members' diurnal rhythms, but it
inﬂuenced their CAR. Technology use alone was likely not powerful
enough to signiﬁcantly alter the slope of family members' cortisol
throughout the day, but it did affect their amount of cortisol pro-
duction in the morning. Adolescents' phone, general media use, and
Facebook social network size resulted in a greater rise in CAR for
adolescents than their parents. The only other signiﬁcant predictor
of a greater change in CAR for parents was for fathers, whose email
use was also associated with higher CAR than mother's email use.
Our results also revealed that when fathers used more general
media throughout the day and more technology at bedtime, it
diminished their CAR. But, the same pattern was predictive of a rise
in CAR for adolescents. This ﬁnding might indicate that using
technology right before bed and throughout the day could be
making fathers experience symptoms that emulate depression,
such as extreme fatigue. Adolescents, on the other hand, who
engage in the same patterns may be feeling an increase in anxiety
or stress. A rise in CAR is expected and healthy to combat daily
stressors (Fries et al., 2009). Too high or too low of a CAR, however,
is often predictive of poorer physical and mental health such as
anxiety and depressive symptoms (Fries et al., 2009). Although the
ﬁndings on CAR are still debated, acute or more moderate stress
and anxiety is often predictive of a greater rise in CAR (Stetler &
Miller, 2005), which could be one explanation for rise in CAR for
the adolescents in our sample, whereas more severe and chronic
stress can lead to a blunting of the CAR (Stetler &Miller, 2005),
which might be indicative of fathers’drop in cortisol after waking.
These ﬁndings suggest that the physiological impacts of technology
on families depends on who is using it, how it is used, and why it is
Unlike adults, adolescents have reported using an array of
different types of social media, including Facebook. Adolescents use
Facebook to keep up with social networks (Frison &Eggermont,
2016) and maintain a mediated social identity (Oeldorf-Hirsch,
Birnholtz, &Hancock, 2017). Deﬁning their role in society and
developing an identity are central goals at their age (boyd, 2015).
But, this can produce social stress. Social stress is a pervasive and
common type of stress for adolescents, which can have far reaching
physiological effects (Finnel &Wood, 2016). Because of the
continual access to others' personal information, adolescents might
become overwhelmed with managing the array of emotions from
others’lives, monitoring their place in their social network and
experiencing social comparisons.
Our results suggest that the consequences of being “wired”not
only inﬂuence CAR, but also extend to inﬂammation. Although no
results were found for email and general media use for adolescents
or parents, adolescents who reported greater phone use and
Facebook use experienced greater IL-6. Pro-inﬂammatory
cytokines, such as IL-6, have been associated with poorer physical
and mental health in families (Graham et al., 2009). Over-
production of IL-6 has been correlated with cardiovascular disease,
cancer, psychiatric disorders, and post-traumatic stress disorders
(Carpenter et al., 2010; Cohen, Doyle, &Skoner, 1999). Given that
this is only a one week study with families, however, additional
research is necessary that can track adolescents’technology use
and its association with biosocial markers over the life course.
5.2. Affordances as a theoretical explanation
Even though we did not examine why family members used
certain technology or media, the research on media affordances
(Fox &Moreland, 2015; Rice et al., 2017; Treem &Leonardi, 2012,
pp. 143e189) could shed light on our ﬁndings. Fathers were more
affected physiologically by increased email use than mothers. Fa-
ther's email use may be bringing work into the home and thus
introducing stress. However, adolescents' physiology and technol-
ogy use tells a different story. Even though adolescents were born
with these technologies, overall media consumption generally ap-
pears to be more physiologically stressful for adolescents than their
parents. Furthermore, not all media use had this effect. Unlike
parents, adolescents were unaffected by email. One speculation for
this is that email is not as important in adolescents' lives, especially
for maintaining their friendships, compared to other forms of
technology. Moreover, Facebook social network size, rather than
social media use in general, had an effect on CAR and IL-6 for ad-
olescents. These ﬁndings seem to indicate that it is the reason for
technology use above and beyond technology use itself, which has
an effect on physiological stress.
These ﬁndings contribute to developing theoretical models of
technology effects, such as the affordances framework. For
instance, the mobility of devices (i.e., portability) may extend fa-
thers' capacity to engage in work related media such as email, but
this affordance of availability and accessibility might diminish
work-home boundaries and facilitate communication overload
(Jarvenpaa &Lang, 2005; Reinecke et al., 2016; Schrock, 2015). This
perspective could be applied to the other family members. Mothers'
CAR and IL6 were not signiﬁcantly affected by technology use,
which could indicate that they are better adapted to the volume of
communication that technology have facilitated than fathers and
adolescents. This requires further investigation. Yet, the role of
portability might also explain the greater physiological effects
experienced by adolescents compared to parents, especially as it
relates to social network sizes via Facebook (Morin-Major et al.,
2016; Schrock, 2015). Adolescents’social networks might increase
the number of relationships they have beyond a size that is perhaps
maintainable. This supports other ﬁndings that demonstrate that
people sometimes experience fatigue and stress because of SNS
(Fox &Moreland, 2015; Lee et al., 2016).
Adolescents who are in the process of exploring and developing
their social environment and identity might also be affected by the
content afforded by technology, such as the increased capacity for
selective self-presentation (Rice et al., 2017). Selective self-
presentation may combine with other affordances, such as
increased access and visibility in one's social network, to facilitate
social comparison. Content can be selectively chosen to enhance
self-presentation, creating perceptions that one's peers are happier,
more social, and more attractive. Through social comparisons, ad-
olescents can experience sensations of loneliness (Best, Manktelow,
&Taylor, 2014), dissatisfaction (Tiggemann &Slater, 2013) and
anxiety such as FOMO (Rosen et al., 2013a). The communication
within social networks could facilitate adolescents' social stress by
creating unachievable norms that precipitate adolescent's feelings
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e27 3 271
Similarly, the content of social network messages could produce
stress by facilitating emotional contagion (Fox &Moreland, 2015).
Hancock, Gee, Ciaccio, and Lin (2008) found that the emotional
content of the messages shared by people's social networks inﬂu-
ence the mood of messages that people share. This might be
especially true of adolescents who are exploring social networks,
experiencing hormonal changes, and facing the challenges of
school and peers as they prepare for college and enter their ﬁrst
romantic relationships (boyd, 2015). Peer networks' message con-
tent could increase exposure to negative affect and increase ado-
5.3. Final thoughts
The ﬁndings of the current study must be set within its limita-
tions. Because of the intensity and cost of the data collection, the
sample included a small group of families from the Midwest. Our
ﬁndings need to be tested with a larger sample and in more diverse
parts of the United States, and the globe more broadly. Because the
current sample is not randomly generated, we cannot generalize
our ﬁndings within or outside the United States. Different effects of
technology and media use on diurnal rhythm might be realized
with a larger, randomly selected, international sample and one that
is collected over a longer period of time. In addition, our sample
size might have also affected our power to detect signiﬁcant dif-
ferences. We decided to continue to report ﬁndings that were
approaching signiﬁcance (p<0.10) given that signiﬁcance levels are
arbitrarily created by researchers, are likely a reﬂection of sample
size, and are of practical and theoretical importance. We were also
unable to examine differences in technology/media use and
biosocial markers for boys and girls because of our limited sample
size. However, this is an important direction for future research
given that research suggests that technology and media use and
decision making often differ depending upon the sex and age of the
adolescent (e.g., Fedorowicz, Vilvovsky, &Golibersuch, 2010).
Finally, although different technology and media uses were
compared in this study, the reasons for using these technologies are
speculative and based on previous affordances research. Future
research should evaluate distinctions among family members’
needs, uses and gratiﬁcations of media and technology use and how
they affect biosocial stress markers. In addition, gaming was
measured with a couple of items within the general media use
subscale. However, given its popularity with adolescents and adults
alike, researchers might compare its inﬂuence on biosocial markers
to other media and technology. Even with these limitations, the
current study provides foundation information on the effect of
technology use in families on cortisol and immune strength. Our
results suggest that adolescents, in particular, might be most at risk
for the effects of being “wired”in a fast paced world.
Funding for this study was provided by the University of Iowa.
Adam, E. K., Vrshek-Schallhorn, S., Kendall, A. D., Mineka, S., Zinbarg, R. E., &
Craske, M. G. (2014). Prospective associations between the cortisol awakening
response and ﬁrst onsets of anxiety disorders over a six-year follow-up - 2013
Curt Richter award winner. Psychoneuroendocrinology, 44,47e59. https://
Anderson, M. (2015). Technology device ownership: 2015. Pew Research Center.
Best, P., Manktelow, R., & Taylor, B. (2014). Online communication, social media and
adolescent wellbeing. Children and Youth Services Review, 41,27e36.
boyd, d (2015). It's complicated: The social lives of networked teens. New Haven, CT:
Yale University Press.
Brom, C., Buchtov
Sisler, V., D
erenko, F., Palme, R., & Glenk, L. M. (2014).
Flow, social interaction anxiety and salivary cortisol responses in serious
games: A quasi-experimental study. Computers &Education, 79,69e100.
Cain, N., & Gradisar, M. (2010). Electronic media use and sleep in school-aged
children and adolescents: A review. Sleep Medicine, 11,735e742. https://
Carpenter, L. L., Gawugas, C. E., Tyrka, A. R., Lee, J. K., Anderson, G. M., et al. (2010).
Association between plasma IL-6 response to acute stress and early-life
adversity in healthy adults. Neuropsychopharmacology, 35,2617e2623.
Carvalho, J., Francisco, R., & Relvas, A. P. (2015). Family functioning and information
and communication technologies: How do they relate? A literature review.
Computers in Human Behavior, 45,99e108. https://doi.org/10.1016/
Cohen, S., Doyle, W. J., & Skoner, D. P. (1999). Psychological stress, cytokine pro-
duction, and severity of upper respiratory illness. Psychosomatic Medicine, 61,
175 e180 .
Common Sense Media. (2013). Zero to eight: Children's media use in American 2013.
New York, NY: Common Sense Media.
Desantis, A., Adam, E., Doane, L., Mineka, S., Zinbarg, R., & Craske, M. (20 07). Racial/
ethnic differences in cortisol diurnal rhythms in a community sample of ado-
lescents. Journal of Adolescent Health, 41,3e13.
Doane, L. D., Mineka, S., Zinbarg, R. E., Craske, M., Grifﬁth, J. W., & Adam, E. K. (2013).
Are ﬂatter diurnal cortisol rhythms associated with major depression and
anxiety disorders in late adolescence? The role of life stress and daily negative
emotion. Development and Psychopathology, 25, 629e642. https://doi.org/
Fedorowicz, Jane, Vilvovsky, Sonia Gantman, & Golibersuch, Andrew J. (2010).
Gender differences in Teenagers' elective use of computer technology. Com-
munications of the Association for Information Systems, 27. Article 3. Available at:
Finnell, J. E., & Wood, S. K. (2016). Neuroinﬂammation at the interface of depression
and cardiovascular disease: Evidence from rodent models of social stress.
Neurobiology of Stress, 4,1e14. https://doi.org/10.1016/j.ynstr.2016.04.001.
Fox, J., & Moreland, J. J. (2015). The dark side of social networking sites: An
exploration of the relational and psychological stressors associated with Face-
book. Computers in Human Behavior, 45,168e176. https://doi.org/10.1016/
Fries, E., Dettenborn, L., & Kirschbaum, C. (2009). The cortisol awakening response
(CAR): Facts and future directions. International journal of Psychophysiology, 72,
Frison, E., & Eggermont, S. (2016). Exploring the relationships between different
types of Facebook use, perceived online social support and adolescents'
depressed mood. Social Science Computer Review, 34,153e171.
Gentile, D. A., Bender, P. K., & Anderson, C. A. (2017). Violent video game effects on
salivary cortisol, arousal, and aggressive thoughts in children. Computers in
Human Behavior, 70,39e43. https://doi.org/10.1016/j.chb.2016.12.045.
Gordis, E. B., Granger, D. A., Susman, E. J., & Trickett, P. K. (2006) . Asymmetry be-
tween salivary cortisol and
-amylase reactivity to stress: Relation to aggressive
behavior in adolescents. Psychoneuroendocrinology, 31,976e987. https://
Graham, J. E., Glaser, R., Loving, T. J., Malarkey, W. B., Stowell, J. R., & Kiecolt-
Glaser, J. K. (2009). Cognitive word use during marital conﬂict and increases in
proinﬂammatory cytokines. Health Psychology, 28,621e630. https://doi.org/
Granger, D., Fortunato, C., Beltzer, E., Virag, M., Bright, M., & Out, D. (2012). Focus on
methodology: Salivary bioscience and research on adolescence: An integrated
perspective. Journal of Adolescence, 35, 1081e1095. https://doi.org/10.1016/
Hancock, J. T., Gee, K., Ciaccio, K., & Lin, J. M. (2008). I’m sad you’re sad: emotional
contagion in CMC. In Proceedings of CSCW ’08: ACM Conference on computer
supported cooperative work (pp. 295e298). https://doi.org/10.1145/
Heo, J.-Y., Kim, K., Fava, M., Mischoulon, D., Papakostas, G. I., Kim, M.-J., …Jeon, H. J.
(2017). Effects of smartphone use with and without blue light at night in
healthy adults: A randomized, doubleS-blind, cross-over, placebo-controlled
comparison. Journal of Psychiatric Research, 87,61e70. https://doi.org/10.1016/
Jaremka, L., Lindgren, M., & Kiecolt-Glaser, J. (2013). Synergistic relationships among
stress, depression, and troubled relationships: Insights from psychoimmunol-
ogy. Depression and Anxiety, 30, 288e296.
Jarvenpaa, S., & Lang, K. (2005). Managing the paradoxes of mobile technology.
Information Systems Management, 22,7e23.
Juster, R. P., McEwen, B. S., & Lupien, S. J. (2010). Allostatic load biomarkers of
chronic stress and impact on health and cognition. Neuroscience and Biobe-
havioral Reviews, 35,2e16. https://doi.org/10.1016/j.neubiorev.2009.10.002.
Kuhlman, K. R., Repetti, R. L., Reynolds, B. M., & Robles, T. F. (2016). Change in
parent-child conﬂict and the HPA-axis: Where should we be looking and for
how long? Psychoneuroendocrinology, 68,74e81. https://doi.org/10.1016/
Lee, A. R., Son, S. M., & Kim, K. K. (2016). Information and communication tech-
nology overload and social networking service fatigue: A stress perspective.
Computers in Human Behavior, 55,51e
Lemola, S., Perkinson-Gloor, N., Brand, S., Dewald-Kaufmann, J. F., & Grob, A. (2014).
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e273272
Adolescents: Electronic media use at night, sleep disturbance, and depressive
symptoms in the smartphone age. Journal of Youth and Adolescence, 44,
McDaniel, B. T. (2015). “Technoference”: Everyday intrusions and interruptions of
technology in couple and family relationships. In C. J. Bruess (Ed.), Family
communication in the age of digital and social media. New York: Peter Lang
McDaniel, B. T., & Coyne, S. M. (2014). “Technoference”: The interference of tech-
nology in couple relationships and implications for women's personal and
relational well-being. Psychology of Popular Media Culture, 5.https://doi.org/
10.1037/ppm0000 065. Advance online publication.
McEwen, B. S. (1998). Protective and damaging effects of stress media- tors. New
England Journal of Medicine, 338,171e179.
McEwen, B. S., & Wingﬁeld, J. C. (20 03). The concept of allostasis in biology and
biomedicine. Hormones and behavior, 43,2e15.
Morin-Major, J. K., Marin, M. F., Durand, N., Wan, N., Juster, R. P., & Lupien, S. J.
(2016). Facebook behaviors associated with diurnal cortisol in adolescents: Is
befriending stressful? Psychoneuroendocrinology, 63, 238e246. https://doi.org/
Munck, A., Guyre, P. M., & Holbrook, N. J. (1984). Physiological functions of gluco-
corticoids in stress and their relations to pharmacological actions. Endocrine
Review, 5,25e44. https://doi.org/10.1210/edrv-5-1-25.
Nam, T. (2014). Technology use and work-life balance. Applied Research in Quality of
Life, 9,1017e1040. https://doi.org/10.1007/s11482-013-9283-1. article.
Nathanson, A. I. (2015). Media and the family: Reﬂections and future directions.
Journal of Children and Media, 9,133e139. https://doi.org/10.1080/
Oeldorf-Hirsch, A., Birnholt, J., & Hancock, J. (2017). Your post is embarrassing me:
Face threats, identity, and the audience on Facebook. Computers in Human
Behavior, 73,92e99. https://doi.org/10.1016/j.chb.2017.03.030.
Reinecke, L., Aufenanger, S., Beutel, M. E., Dreier, M., Quiring, O.,
Stark, B., …Dreier, M. (2016). Digital stress over the life span: The effects of
communication load and internet multitasking on perceived stress and psy-
chological health impairments in a German probability sample. Media Psy-
chology, 0,1e26. https://doi.org/10.1080/15213269.2015.1121832.
Repetti, R. L., Robles, T. F., & Reynolds, B. (2011). Allostatic processes in the family.
Development and Psychopathology, 23,921e938. https://doi.org/10.1017/
Rice, R. E., Evans, S. K., Pearce, K. E., Sivunen, A., Vitak, J., & Treem, J. W. (2017).
Organizational media affordances: Operationalization and associations with
media use. Journal of Communication, 67,106e130 .
Rice, R. E., & Hagen, I. (2010). Young adults' perpetual contact, social connectivity,
and social control through the internet and mobile phones. In C. Salmon (Ed.),
Communication yearbook, 34. London: Routledge.
Rice, R. E., & Katz, J. E. (2003). Comparing internet and mobile phone usage: Digital
divides of usage, adoption, and dropouts. Telecommunications Policy, 27,
Riva, G., Ba~
nos, R. M., Botella, C., Wiederhold, B. K., & Gaggioli, A. (2012). Positive
Technology: Using interactive technologies to promote positive functioning.
Cyberpsychology, Behavior, and Social Networking, 15,69e77. https://doi.org/
Rosen, L. D., Whaling, K., Carrier, L. M., Cheever, N. A., & Rokkum, J. (2013a). The
media and technology usage and attitudes Scale: An empirical investigation.
Computers in Human Behavior, 29, 2501e2511. https://doi.org/10.1016/
Rosen, L. D., Whaling, K., Rab, S., Carrier, L. M., & Cheever, N. A. (2013b). Is Facebook
creating “iDisorders”? The link between clinical symptoms of psychiatric dis-
orders and technology use, attitudes and anxiety. Computers in Human Behavior,
29, 1243e1254. https://doi.org/10.1016/j.chb.2012.11.012.
Rutledge, P. B. (2012). Arguing for media psychology as a distinct ﬁeld. In K. E. Dill
(Ed.), The Oxford handbook of media psychology. Oxford, UK: Oxford University
Saxbe, D., & Repetti, R. L. (2010). For better or worse? Coregulation of couples'
cortisol levels and mood states. Journal of Personality and Social Psychology, 98,
Schrock, A. R. (2015). Communicative affordances of mobile media: Portability,
availability, locatability and multimediality. International Journal of Communi-
cation, 9, 1229e1246. https://doi.org/10.1177/0094306111425016k.
Stephens, K. K., Mandhana, D. M., Kim, J. J., Li, X., Glowacki, E. M., & Cruz, I. (2017).
Reconceptualizing communication overload and building a theoretical foun-
dation. Communication Theory, 3, 269e289. https://doi.org/10.1111/comt.12116.
Stetler, C., & Miller, G. (20 05). Blunted cortisol response to awakening in mild to
moderate depression: Regulatory inﬂuences of sleep patterns and social con-
tacts. Journal of Abnormal Psychology, 114,697e705.
ee, S., H€
arenstam, A., & Hagberg, M. (2011). Mobile phone use and stress, sleep
disturbances, and symptoms of depression among young adults - a prospective
cohort study. BMC Public Health, 11,1e11. https://doi.org/10.1186/1471-2458-11-
Tiggemann, M., & Slater, A. (2013). NetGirls: The Internet, Facebook, and body image
concern in adolescent girls. International Journal of Eating Disorders, 46,
Treem, J. W., & Leonardi, P. M. (2012). Social media use in organizations: Exploring
the affordances of visibility, editability, persistence, and association. In
Communication Yearbook (Vol. 36)New York, NY: Routledge.
Turkle, S. (2011). Alone together: Why we expect more from technology and less from
each other. New York, NY: Basic Books.
Wallenius, M., Hirvonen, A., Lindholm, H., Rimpela, A., Nygard, C., Saarni, L., et al.
(2010). Salivary cortisol in relation to the use of information and communica-
tion technology (ICT) in school-aged children. Psychology, 1,88e95. https://
T.D. Aﬁﬁ et al. / Computers in Human Behavior 81 (2018) 265e27 3 273