Available via license: CC BY 4.0
Content may be subject to copyright.
Contents lists available at ScienceDirect
Preventive Medicine Reports
journal homepage: www.elsevier.com/locate/pmedr
Associations between screen time and lower psychological well-being
among children and adolescents: Evidence from a population-based study
Jean M. Twenge
a,⁎
, W. Keith Campbell
b
a
San Diego State University, United States of America
b
University of Georgia, United States of America
ABSTRACT
Previous research on associations between screen time and psychological well-being among children and adolescents has been conflicting, leading some researchers
to question the limits on screen time suggested by physician organizations. We examined a large (n= 40,337) national random sample of 2- to 17-year-old children
and adolescents in the U.S. in 2016 that included comprehensive measures of screen time (including cell phones, computers, electronic devices, electronic games, and
TV) and an array of psychological well-being measures. After 1 h/day of use, more hours of daily screen time were associated with lower psychological well-being,
including less curiosity, lower self-control, more distractibility, more difficulty making friends, less emotional stability, being more difficult to care for, and inability
to finish tasks. Among 14- to 17-year-olds, high users of screens (7+ h/day vs. low users of 1 h/day) were more than twice as likely to ever have been diagnosed with
depression (RR 2.39, 95% CI 1.54, 3.70), ever diagnosed with anxiety (RR 2.26, CI 1.59, 3.22), treated by a mental health professional (RR 2.22, CI 1.62, 3.03) or
have taken medication for a psychological or behavioral issue (RR 2.99, CI 1.94, 4.62) in the last 12 months. Moderate use of screens (4h/day) was also associated
with lower psychological well-being. Non-users and low users of screens generally did not differ in well-being. Associations between screen time and lower psy-
chological well-being were larger among adolescents than younger children.
1. Introduction
A growing proportion of children and adolescents' leisure time is
spent with screens including smartphones, tablets, gaming consoles,
and televisions (Common Sense Media, 2015;Twenge et al., 2019),
raising concerns about the effect of screen time on well-being among
parents, health professionals, and educators (e.g., Kardaras, 2017).
These concerns have prompted physician organizations such as the
American Academy of Pediatrics (AAP) to recommend that parents
limit children's daily screen time, with specific time limits for preschool
children and a general suggestion of limiting time on screens for older
children and adolescents (Radesky and Christakis, 2016). In addition,
the World Health Organization recently decided to include gaming
disorder in the 11th revision of the International Classification of Dis-
eases (WHO, 2018).
Associations between screen time and poor health outcomes such
as obesity and lack of exercise have been well-documented (e.g.,
Chiasson et al., 2016;de Jong et al., 2013;Dumuid et al., 2017;
Poitras et al., 2017). However, research exploring associations be-
tween screen time and more psychological aspects of well-being
among children and adolescents has been inconsistent. Some studies
find significant associations between screen time and low well-being
(Babic et al., 2017;Page et al., 2010;Romer et al., 2013;Rosen et al.,
2014;Twenge et al., 2018a,b;Yang et al., 2013), while others find null
effects or even benefits with greater screen time (Granic et al., 2014;
Odgers, 2018;Przybylski and Weinstein, 2018;Valkenburg and Peter,
2009). Thus, some have suggested that more research is needed before
concluding that screen time limits are justified, arguing that valuable
physician appointment time should not be devoted to discussing
screen time without sufficient evidence for significant associations
with well-being (Przybylski and Weinstein, 2017, 2018). Some re-
searchers have made similar statements about the WHO characterizing
gaming disorder as a mental health issue, maintaining that associa-
tions between gaming and psychological well-being are not sub-
stantial or consistent enough to justify such a classification (Davis,
2018;van Rooij et al., 2018).
Theories and research on psychological well-being support the
notion of a broad concept including emotional stability, positive in-
https://doi.org/10.1016/j.pmedr.2018.10.003
Received 9 September 2018; Received in revised form 30 September 2018; Accepted 14 October 2018
⁎
Corresponding author at: Department of Psychology, San Diego State University, 5500 Campanile Drive, San Diego, CA 92182-4611, United States of America.
E-mail address: jtwenge@mail.sdsu.edu (J.M. Twenge).
Preventive Medicine Reports 12 (2018) 271–283
Available online 18 October 2018
2211-3355/ © 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/BY/4.0/).
terpersonal relationships, self-control, and indicators of flourishing
(Diener et al., 1999;Ryff, 1995) as well as diagnoses of mood dis-
orders such as anxiety or depression (Manderscheid et al., 2010). Low
emotional stability, disrupted relationships, and low self-control have
all been implicated in greater morbidity and mortality (Graham et al.,
2017;Shipley et al., 2007;Shor et al., 2013;Turiano et al., 2015), and
mental health issues such as mood disorders are a significant risk
factor for morbidity and mortality, including via non-suicidal self-
harm behaviors, suicide attempts, and completed suicides (Hawton
et al., 2013;Murray et al., 2012).
In terms of prevention, establishing possible causes and outcomes of
low psychological well-being is especially important for child and
adolescent populations. Half of mental health problems develop by
adolescence (Erskine et al., 2015). Thus, there is an acute need to
identify factors linked to mental health issues that are amenable to
intervention in this population, as most antecedents (e.g., genetic pre-
disposition, trauma, poverty) are difficult or impossible to influence.
Compared to these more intractable antecedents of mental health, how
children and adolescents spend their leisure time is more amenable to
change.
To our knowledge, few if any previous studies have examined a
broad array of psychological well-being items in relation to screen time.
Moreover, although other studies have examined associations between
screen time and symptoms of anxiety and depression, no previous study
we know of has examined associations between screen time and actual
diagnoses of anxiety or depression or reports of professional treatment
for mental health issues. Furthermore, it is crucial for measures of
screen time to include not just television but more recently introduced
digital media including electronic gaming, smartphones, tablets, and
computers. In addition, studies using the same items to assess several
age groups of children and adolescents are rare, which is unfortunate as
Table 1
Approximate hours a day of screen use by age group, U.S., 2016.
2 to 5 6 to 10 11 to 13 14 to 17 d
TV and video games 1.46 (1.09) 1.53 (1.10) 1.80 (1.39) 1.89 (1.39) 0.34
Electronic devices 0.82 (0.96) 1.25 (1.11) 2.00 (1.40) 2.70 (1.53) 1.46
Total screen time 2.28 (1.72) 2.78 (1.95) 3.80 (2.36) 4.59 (2.50) 1.06
Note: 1. SDs in parentheses.
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Approximate hours a day
Age in years
Total screen
!me
Electronic
devices
TV & video
games
Fig. 1. Hours per day spent on all screens, electronic devices, and TV and video games by individual years of age, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
272
Table 2
Well-being means by hours per day of total screen time (with controls) among age groups and d's comparing cells, U.S., 2016.
None (0) < 1 h 1 h 2 h 3 h 4 h 5 h 6 h 7+ h d0 vs. 1 h d1 h vs. 4 h d1 h vs. 7+ h
Easy child (3 items)
2 to 5 (preschool) 4.42 (0.53) 320 4.33 (0.61) 749 4.37 (0.57) 1707 4.32 (0.61) 2687 4.29 (0.62) 1843 4.28 (0.63) 912 4.26 (0.66) 280 4.14 (0.73) 243 4.18 (0.67) 263 −0.09 −0.15* −0.33*
6 to 10 (elem.) 4.28 (0.65) 215 4.33 (0.63) 348 4.36 (0.63) 1457 4.36 (0.62) 3203 4.32 (0.63) 2187 4.33 (0.64) 1483 4.27 (0.65) 397 4.26 (0.63) 431 4.18 (0.69) 496 0.13 −0.05 −0.28*
11 to 13 (middle) 4.37 (0.65) 46 4.28 (0.72) 104 4.38 (0.66) 477 4.38 (0.60) 1621 4.35 (0.62) 1537 4.33 (0.63) 1464 4.33 (0.60) 525 4.21 (0.70) 566 4.15 (0.73) 895 0.02 −0.08 −0.33*
14 to 17 (h.s.) 4.36 (0.99) 24 4.49 (0.58) 80 4.54 (0.51) 370 4.46 (0.57) 1679 4.40 (0.57) 2488 4.35 (0.60) 2468 4.30 (0.66) 1118 4.20 (0.71) 1370 4.09 (0.77) 2547 0.33 −0.32* −0.61*
Curious
2 to 5 (preschool) 2.98 (0.12) 319 2.98 (0.14) 752 2.98 (0.12) 1716 2.98 (0.15) 2705 2.98 (0.16) 1853 2.96 (0.21) 919 2.96 (0.22) 280 2.95 (0.25) 245 2.96 (0.22) 266 0.00 −0.13* −0.15
6 to 10 (elem.) 2.94 (0.22) 215 2.94 (0.23) 349 2.95 (0.19) 1462 2.95 (0.22) 3225 2.93 (0.26) 2200 2.93 (0.29) 1488 2.88 (0.34) 401 2.88 (0.35) 434 2.88 (0.37) 504 0.05 −0.08* −0.30*
11 to 13 (middle) 2.88 (0.32) 47 2.91 (0.25) 105 2.91 (0.28) 480 2.88 (0.32) 1631 2.89 (0.30) 1537 2.86 (0.36) 1474 2.86 (0.36) 533 2.79 (0.46) 571 2.76 (0.47) 901 0.11 −0.15* −0.37*
14 to 17 (h.s.) 2.64 (0.58) 24 2.90 (0.26) 82 2.90 (0.30) 375 2.88 (0.31) 1691 2.86 (0.34) 2501 2.82 (0.39) 2485 2.79 (0.43) 1132 2.74 (0.48) 1371 2.71 (0.49) 2583 0.78* −0.21* −0.41*
No difficulty making friends
2 to 5 (preschool) 2.87 (0.35) 181 2.89 (0.31) 400 2.90 (0.30) 1112 2.91 (0.29) 1980 2.90 (0.32) 1405 2.86 (0.38) 707 2.81 (0.41) 204 2.85 (0.38) 191 2.84 (0.39) 209 0.10 −0.12* −0.19*
6 to 10 (elem.) 2.81 (0.42) 211 2.83 (0.40) 346 2.85 (0.38) 1454 2.85 (0.38) 3189 2.83 (0.41) 2176 2.82 (0.43) 1470 2.78 (0.48) 396 2.77 (0.45) 429 2.77 (0.47) 492 0.10 −0.07* −0.20*
11 to 13 (middle) 2.74 (0.53) 47 2.80 (0.45) 104 2.78 (0.47) 473 2.82 (0.43) 1620 2.78 (0.48) 1525 2.79 (0.44) 1464 2.82 (0.43) 522 2.70 (0.52) 569 2.66 (0.58) 888 0.08 0.02 −0.22*
14 to 17 (h.s.) 2.81 (0.49) 23 2.76 (0.48) 81 2.88 (0.36) 367 2.85 (0.40) 1678 2.82 (0.42) 2476 2.79 (0.46) 2464 2.76 (0.49) 1113 2.72 (0.51) 1355 2.66 (0.58) 2521 0.19 −0.20* −0.40*
Calm when challenged
6 to 10 (elem.) 2.46 (0.60) 211 2.42 (0.58) 344 2.48 (0.54) 1450 2.45 (0.56) 3190 2.39 (0.57) 2176 2.37 (0.60) 1465 2.32 (0.60) 394 2.35 (0.59) 428 2.32 (0.62) 491 0.04 −0.19* −0.29*
11 to 13 (middle) 2.62 (0.54) 47 2.54 (0.54) 104 2.56 (0.56) 475 2.57 (0.53) 1619 2.53 (0.56) 1521 2.51 (0.56) 1459 2.51 (0.56) 524 2.43 (0.60) 567 2.35 (0.62) 887 −0.11 0.09 −0.35*
14 to 17 (h.s.) 2.70 (0.58) 23 2.68 (0.56) 81 2.75 (0.45) 367 2.70 (0.49) 1677 2.66 (0.50) 2472 2.60 (0.54) 2462 2.55 (0.57) 1110 2.48 (0.60) 1352 2.45 (0.62) 2523 0.11 −0.29* −0.50*
Works to finish tasks started
6 to 10 (elem.) 2.71 (0.48) 211 2.66 (0.49) 345 2.72 (0.46) 1450 2.70 (0.48) 3182 2.65 (0.50) 2175 2.64 (0.52) 1465 2.58 (0.55) 392 2.61 (0.55) 430 2.57 (0.56) 491 0.02 −0.16* −0.31*
11 to 13 (middle) 2.75 (0.50) 47 2.79 (0.39) 104 2.72 (0.46) 474 2.72 (0.46) 1625 2.70 (0.47) 1522 2.67 (0.50) 1461 2.67 (0.50) 525 2.55 (0.57) 566 2.51 (0.59) 887 0.06 −0.10* −0.39*
14 to 17 (h.s.) 2.67 (0.49) 24 2.78 (0.45) 81 2.83 (0.37) 366 2.81 (0.39) 1675 2.76 (0.43) 2468 2.71 (0.47) 2455 2.66 (0.52) 1114 2.60 (0.57) 1352 2.54 (0.58) 2523 0.38 −0.26* −0.52*
Does not argue too much
6 to 10 (elem.) 2.66 (0.51) 209 2.64 (0.57) 346 2.67 (0.55) 1452 2.64 (0.58) 3193 2.60 (0.60) 2178 2.58 (0.61) 1467 2.56 (0.63) 393 2.58 (0.59) 430 2.48 (0.67) 490 0.04 0.16* −0.33*
11 to 13 (middle) 2.69 (0.56) 47 2.54 (0.55) 104 2.68 (0.55) 476 2.69 (0.54) 1621 2.63 (0.58) 1524 2.62 (0.59) 1465 2.61 (0.59) 526 2.54 (0.65) 569 2.47 (0.68) 887 −0.02 −0.10* −0.33*
14 to 17 (h.s.) 2.60 (0.66) 23 2.71 (0.55) 80 2.81 (0.46) 366 2.79 (0.46) 1681 2.73 (0.50) 2477 2.71 (0.53) 2461 2.68 (0.57) 1114 2.61 (0.60) 1354 2.52 (0.67) 2530 0.45 −0.19* −0.45*
Ever diagnosed with anxiety
11 to 13 (middle) 9.6% (0.29) 47 6.8% (0.25) 105 9.9% (0.30) 481 7.6% (0.26) 1634 10.0% (0.30)
1540
8.5% (0.28) 1477 9.3% (0.29) 532 11.2% (0.32) 573 12.2% (0.33) 904 0.01 0.05 0.07
14 to 17 (h.s.) 11.5% (0.32)
24
12.0% (0.33) 80 7.9% (0.26) 374 8.4% (0.28) 1698 9.7% (0.30) 2504 12.2% (0.33)
2489
13.4% (0.34)
1131
17.7% (0.38)
1374
18.1% (0.39)
2578
−0.13 0.13* 0.27*
Ever diagnosed with depression
11 to 13 (middle) 4.6% (0.21) 47 1.6% (0.12) 105 3.7% (0.19) 481 1.9% (0.14) 1629 4.1% (0.19) 1543 3.8% (0.19) 1479 4.3% (0.21) 534 5.4% (0.23) 573 7.2% (0.26) 906 −0.05 0.05 0.15*
14 to 17 (h.s.) 10.2% (0.30)
24
8.3% (0.28) 82 5.3% (0.23) 376 5.1% (0.23) 1700 6.3% (0.24) 2508 8.6% (0.28) 2493 8.8% (0.29) 1131 11.6% (0.32)
1379
12.7% (0.33)
2582
−0.20 0.12* 0.23*
Treated or needed to be treated by mental health professional, past 12 months
11 to 13 (middle) 7.6% (0.25) 47 10.4% (0.30)
104
13.5% (0.34) 480 10.5% (0.30)
1633
12.6% (0.33)
1539
12.2% (0.33)
1474
9.8% (0.30) 532 14.8% (0.36) 573 18.1% (0.39) 904 0.18 −0.04 0.12*
14 to 17 (h.s.) 25.8% (0.42)
24
15.9% (0.37) 82 9.8% (0.29) 374 11.5% (0.32)
1693
12.8% (0.34)
2496
14.1% (0.35)
2487
17.0% (0.38)
1128
20.7% (0.41)
1373
21.9% (0.41)
2578
−0.53 0.13* 0.30*
Took medication for psychological issue, past 12 months
11 to 13 (middle) 9.0% (0.29) 46 6.2% (0.24) 105 8.6% (0.28) 479 6.9% (0.25) 1623 8.5% (0.28) 1529 9.4% (0.29) 1473 9.1% (0.29) 529 12.4% (0.33) 571 13.3% (0.34) 894 −0.01 0.03 0.15*
14 to 17 (h.s.) 11.7% (0.32)
23
11.7% (0.32) 82 5.5% (0.22) 372 8.2% (0.28) 1686 8.6% (0.28) 2492 9.9% (0.30) 2481 12.1% (0.33)
1116
14.9% (0.36)
1366
16.1% (0.37)
2562
−0.27 0.15* 0.30*
Notes: 1. Within levels of screen time, numbers in each cell are: well-being means, SDs in parentheses, and n's. 2. d= effect size corresponding to difference in standard deviations. 3. * = t-test comparing cells significant
at p< .05. 4. For diagnoses, treatment, and medication, base rates were high enough for reliable comparisons only among the two older age groups.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
273
age may be a significant moderator of associations between screen time
and psychological well-being.
The current research aims to examine associations between screen
time and a diverse array of measures of psychological well-being (in-
cluding emotional stability, relationships with caregivers, self-control,
diagnoses of mood disorders, and treatment of mental health issues)
among a large population-based survey of the caregivers of children
and adolescents ages 2 to 17 collected in 2016 in the U.S.
2. Method
2.1. Participants
Participants were the caregivers of 44,734 children and adoles-
cents 2 years of age and older in the U.S. in the National Survey of
Children's Health (NSCH) conducted in 2016 by the U.S. Census
Bureau. As many items on psychological well-being were asked only of
caregivers with children ages 2 and over, we restricted our analyses to
children ages 2 to 17.
Households were contacted by mail at random to identify those with
children or adolescents 17 years old or younger. In every household,
one child was randomly selected to be the subject of the survey. The
survey was administered either online or on paper, with an over-
sampling of children with special health care needs. The response rate
was 40.7%. Data are publicly available on the NSCH website.
We excluded children and adolescents with at least one of 8 major
conditions that might affect their day-to-day functioning: Autism,
blindness, cerebral palsy, deafness, Down Syndrome, developmental
delay, epilepsy, or intellectual disability (mental retardation), as these
were related to both well-being and to screen time. For example, among
14- to 17-year-olds, 33% of those with zero daily use of screens had one
of these conditions, compared to an overall 10.1% base rate. The sur-
vey's comprehensive screener for special health care needs could not be
used as it also includes those who receive any mental health care ser-
vices, which is a variable of interest. These exclusions resulted in a
sample nof 40,337.
Fig. 2. Mean score on the easy child index (1–5), by age and level of screen time, with controls, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
274
In the final sample, the children and adolescents were 49.8% male
and 50.2% female and were 71% White, 16% Hispanic, 6% Black, and
7% other. Family income was widely distributed, with 9% of children
below the 100% poverty level and 44% with family incomes at the
400% level or more. The sample was designed to be nationally re-
presentative of all U.S. children at these ages but underrepresents some
groups due to lower response rates.
We grouped children and adolescents into four categories based on
age that roughly correspond to educational levels: Preschoolers 2 to
5 years old (n= 9361), elementary schoolers 6 to 10 years old
(n= 10,668), middle schoolers 11 to 13 years old (n= 7555), and high
schoolers 14 to 17 years old (n= 12,753). These categories also cor-
respond to the structure of the survey, with some questions only asked
of the caregivers of preschool children and others asked only of children
6 years of age and up.
2.2. Measures
The survey asked two items about screen time. First, “On an average
weekday, about how much time does [child's name] spend in front of a
TV watching TV programs, videos, or playing video games?” Second,
“On an average weekday, about how much time does [child's name]
spend with computers, cell phones, handheld video games, and other
electronic devices, doing things other than schoolwork?” For both, re-
sponse choices were recoded to none = 0, less than an hour =0.5, an
hour = 1, 2 h = 2, 3 h = 3, and 4 or more hours = 5. For means, see
Table 1.
We added together the estimated number of hours spent on TV/
video games and on digital media devices to create a measure of total
screen time and recoded the results into 8 categories: None (no screen
time), < 1 h (0.01 to 0.99), 1 h (1.00 to 1.49), 2 h (1.50 to 2.49), 3 h
(2.50 to 3.49), 4 h (3.50 to 4.49), 5 h (4.50 to 5.49), 6 h (5.50 to 6.49)
and 7 h or more (6.50 and higher). Among the two older groups, very
few reported no screen time at all (n= 46 for 11- to 13-year-olds and
n= 24 for 14- to 17-year-olds), so these cells should be interpreted with
caution.
We examined all items in the NSCH survey that measured psycho-
logical well-being, broadly construed (see Supplemental material for
Fig. 3. Percentage not curious or interested in learning new things, by age and level of screen time, with controls, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
275
item wording including response choices). Most items did not correlate
highly enough to be combined into scales and are thus analyzed as
single items. The exceptions were three items measuring how easy the
child is and four items measuring self-control. We coded all items so
that higher scores indicated higher well-being.
2.3. Analysis plan
Analyses included controls for possible confounding variables: child
race (dummy variables for Black, Hispanic, and Other, with non-
Hispanic White as the comparison group), child sex, child age, house-
hold adults' highest grade completed (continuous, using the detailed
item including college education), family poverty ratio (a measure of
family income), and family structure (living with two biological/
adoptive parents vs. not). We did not weight analyses and did not re-
place missing data.
For items on a continuum, we report means in tables and percent
low in well-being in figures; categorical items (e.g., yes or no, such as
diagnoses of anxiety or depression) are reported as percentages in both.
The tables report effect sizes (d, or difference in terms of standard de-
viations) as well as p-values for t-tests comparing means at different
levels of use. The text reports relative risk (RR) with 95% confidence
intervals (CIs) for dichotomized items.
We first examine items asked of caregivers of several ages of chil-
dren and then those asked only of caregivers of preschool children.
Given the curvilinear relationship between screen time and well-being
found in previous research (Przybylski and Weinstein, 2017;Twenge
et al., 2018b), we identified the inflection point at which the trend in
well-being moved from positive to negative to inform our analyses
(Simonsohn, 2017). Thus, we compare no use to low levels of use, low
use to moderate use, and low use to high use.
Fig. 4. Percentage who do not stay calm when challenged, by age and level of screen time, with controls, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
276
3. Results
3.1. Age differences in screen time
Total screen time averaged 3.20 h a day (SD = 2.40) and was pro-
gressively higher among older children, primarily driven by more time
spent on electronic devices (see Table 1 and Fig. 1). The largest increase
in screen time occurred between elementary school and middle school.
By high school (ages 14 to 17), adolescents spent 4 h and 35 min a day
with screens according to caregivers' reports.
3.2. Screen time and well-being
The association between screen time and well-being was not linear
and showed an inflection point at 1 h/day of use for most measures (see
Table 2 and Figs. 2–6). With one exception (the item on curiosity), well-
being did not differ significantly between those spending no time on
screens and those spending an hour a day. After an hour a day, how-
ever, increasing screen time was generally linked to progressively lower
psychological well-being. In terms of relative risk (RR), high users of
screens (7+ h/day) carried twice the risk of low well-being as low users
(1 h/day), including not staying calm (e.g., among 14- to 17-year-olds,
RR 2.08, 95% CI 1.72, 2.50), not finishing tasks (RR 2.53, CI 2.01,
3.20), not being curious (RR 2.72, CI 2.00, 3.71), and arguing too much
with caregivers (RR 2.34, CI 1.85, 2.97; see Figs. 2–6). High (vs. low)
users of screens were also described as more difficult to care for. Effect
sizes were generally larger among adolescents than among children.
In most cases, moderate users of screens (4 h/day) were also sig-
nificantly lower in well-being than low users (1 h/day), though with
lower effect sizes (see Table 2). Among 14- to 17-year-olds, moderate
users (vs. low users) were 78% more likely to not be curious (RR 1.78,
CI 1.30, 2.43), 60% more likely to not stay calm when challenged (RR
Fig. 5. Percentage who do not finish tasks started, by age and level of screen time, with controls, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
277
1.60, CI 1.32, 1.93), 66% more likely to not finish tasks they started (RR
1.66, CI 1.31, 2.11), and 57% more likely to argue too much with their
caregivers (RR 1.57, CI 1.24, 2.00; see Figs. 2–6). As with the com-
parisons between low and high use, differences in well-being between
low and moderate users were smaller among younger children than
among older adolescents.
3.3. Screen time and diagnoses of anxiety and depression
High users of screens were also significantly more likely to have
been diagnosed with anxiety or depression. Fourteen to 17-year-olds
spending 7+ h/day with screens (vs. 1h/day) were more than twice as
likely to ever have been diagnosed with depression (RR 2.39, 95% CI
1.54, 3.70) or anxiety (RR 2.26, CI 1.59, 3.22; see Fig. 7). High users of
screens were also more likely to have seen or needed to have been seen
by a mental health professional (RR 2.22, CI 1.62, 3.03), and more
likely to have taken medication for a psychological issue (RR 2.99, CI
1.94, 4.62; see Fig. 8) in the last 12 months. Moderate use was also
linked to a greater risk of depression (RR 1.61, CI 1.03, 2.52) and an-
xiety diagnoses (RR 1.52, CI 1.06, 2.18) among 14- to 17-year-olds,
though not among 11- to 13-year-olds.
3.4. Screen time and well-being items asked only of caregivers of preschool
children
We next examined the items asked only of caregivers of preschool
children. High users of screens were more likely to lose their temper,
less likely to calm down when excited, and less likely to switch tasks
without anxiety or anger (see Table 3 and Fig. 9). The largest effect size
appeared for self-control (d= −0.41), which included perseverance,
sitting still, completing simple tasks, and not becoming distracted; both
high and moderate users of screens displayed significantly lower self-
control than low users. In terms of relative risk, high (vs. low) users of
screens were twice as likely to often lose their temper (RR 1.99, CI 1.44,
Fig. 6. Percentage who argue too much with their caregivers, by age and level of screen time, with controls, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
278
2.77) and were 46% more likely to not be able to calm down when
excited (RR 1.46, CI 1.13, 1.88). Preschool children with moderate
screen use were also lower in well-being than those at low levels of use
(see Table 3). In terms of relative risk, moderate users (vs. low users)
were 30% more likely to not bounce back (RR 1.30, CI 1.15, 1.47) and
33% more likely to lose their temper (RR 1.33, CI 1.02, 1.72).
There were few significant associations between screen time and
social interaction items such as showing affection to caregivers, smiling
and laughing, playing well with others, or showing empathy for others
(see Table 3). However, several of these items (particularly showing
affection and smiling and laughing) suffered from an extreme lack of
variance with 95% of caregivers agreeing this was true of the child,
limiting their utility.
4. Discussion
Children and adolescents who spent more time using screen media
were lower in psychological well-being than low users. High users of
screens were significantly more likely to display poor emotion regula-
tion (not staying calm, arguing too much, being difficult to get along
with), an inability to finish tasks, lower curiosity, and more difficulty
making friends. Caregivers also described high users as more difficult to
care for and as lower in self-control. Among adolescents, high (vs. low)
users were also twice as likely to have received diagnoses of depression
or anxiety or needed treatment for mental or behavioral health condi-
tions. Moderate users were also significantly more likely than low users
of screens to be low in well-being and, among 14- to 17-year-olds, to
have been diagnosed with depression or anxiety or need mental health
Fig. 7. Percentage ever diagnosed with anxiety or depression, by age and level of screen time, with controls, U.S., 2016. Error bars are ±1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
279
treatment. Non-users generally did not significantly differ in well-being
from low users of screens.
The AAP recommendations set specific screen time limits only for
children 5 years of age and younger. One set of researchers questioned
these limits based on null correlations between screen time and four
well-being items included in the 2011 NSCH (Przybylski and Weinstein,
2018). However, the more comprehensive set of well-being items in the
2016 NSCH produces significant associations between screen time and
well-being on 18 of 19 indicators, providing substantial support for
screen time limits. Notably, we found that the association between
screen time and low well-being was larger for adolescents than for
younger children, consistent with at least one earlier study (Rosen
et al., 2014). This suggests that the AAP and other organizations fo-
cused on public health might consider extending recommendations for
specific limits on screen time to preteens and teens.
It is worth speculating about why the associations between screen
time and psychological well-being were larger among adolescents.
One possibility is that adolescents, compared to younger children, are
considerably more likely to have social media accounts and to spend
more time online. Peer relationships are particularly important for
adolescents (Fuligni and Eccles, 1993), and if social media replaces
face-to-face interaction, that may have a negative impact on well-
being and mental health. Time spent on social media, gaming, and
online is more strongly correlated with low well-being than watching
TV/videos, and TV/videos are the more common screen activity for
younger children (Rosen et al., 2014). Unfortunately, associations
with well-being for TV vs. other screen activities cannot be de-
termined in this dataset as time spent on TV and electronic gaming
was included in the same item.
Adolescents are also more likely than younger children to have their
own smartphone (Rosen et al., 2014), which allows the use of tech-
nology in more situations. This may increase the possibility of Internet
addiction, excessive gaming, or problematic social media use, which
has been linked to low well-being (Satici and Uysal, 2015). It may also
increase the impact on sleep, as smartphones may be brought into the
bedroom or even the bed, with negative impacts on sleep duration and/
Fig. 8. Percentage receiving mental health treatment and percentage taking medication for psychological issues during the last 12 months, by age and level of screen
time, with controls, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
280
or sleep quality (Twenge et al., 2017). Smartphones may also be used
during face-to-face social interactions, which may negatively impact
those interactions and blunt their usually positive impact on well-being
(Dwyer et al., 2018).
Due to the cross-sectional design of the study, it is not possible to
determine if screen time leads to low well-being, low well-being leads
to screen time, or both. However, several longitudinal studies have
found that increases in recreational screen time precede lower psy-
chological well-being among children and adolescents (Allen and Vella,
2015;Babic et al., 2017;Hinkley et al., 2014;Kim, 2017) as well as
among adults (Kross et al., 2013;Schmiedeberg and Schröder, 2017;
Shakya and Christakis, 2017). In addition, experiments have shown that
the presence of smartphones can lower enjoyment during social inter-
actions (e.g., Dwyer et al., 2018;Kushlev et al., 2017) and that ab-
staining from social media use for one week can increase well-being
(Tromholt, 2016). In other studies, the relationship appears to be re-
ciprocal, with screen time and well-being each causing the other
(Gunnell et al., 2016). These studies suggest that at least some of the
causation moves from screen time to lower well-being. Regardless of
the direction of causation, however, these associations have meaningful
clinical implications for screening and intervention. For example, an
assessment of screen time may help providers identify children and
adolescents at higher risk for mental health issues and broach the topic
of the possible role of screen time in mental health among these in-
dividuals.
These data are limited by several factors. First, screen time was
reported by caregivers and not the children or adolescents themselves.
This likely resulted in underestimates of screen time and may have
unknown interactions with reports of well-being. The well-being mea-
sures may be influenced by caregivers' perceptions and may under-
report issues that children do not disclose to their parents. This is likely
to be less of an issue for the items on diagnoses of anxiety and de-
pression and reports of taking medication. In addition, informant re-
ports are often considered a strength in study design, as in some cases
observers can provide more accurate information than is possible via
self-report (Connelly and Ones, 2010); that is especially true for
younger children. Second, the survey assessed only weekday screen
time, and screen time may be higher on weekends. However, previous
research found similar associations with well-being for weekday and
weekend use of screen media (Przybylski and Weinstein, 2017).
Weekday screen time is also likely to vary less and thus may produce a
more reliable estimate. Third, the survey items combined TV and
electronic games into one question, which allowed only an analysis of
total screen time and not any differentiation between legacy media (TV)
and digital media (electronic games, Internet, social media, etc.).
Fourth, although the Census Bureau attempted to recruit a re-
presentative sample, the response rate was not 100% and some groups
(such as Black Americans) are under-represented relative to their per-
centage of the total U.S. population in the final sample.
In summary, these results show a negative association between
screen time and psychological well-being among children and ado-
lescents. Across a diverse array of well-being measures, including
measures of self-control, relationships with caregivers, emotional
stability, diagnoses of anxiety and depression, and mental health
treatment, psychological well-being was progressively lower from 1 h
a day of screen time to 7 or more hours a day of screen time, parti-
cularly among adolescents. The significant association between screen
time and well-being may have important clinical implications for the
mental and physical health of children and adolescents and for de-
veloping guidelines for specific screen time limits for older children
and adolescents.
Funding
This study received no funding.
Table 3
Well-being means on items asked only of caregivers of preschool children among 2- to 5-year-olds by hours per day of total screen time (with controls) and d's comparing cells, U.S., 2016.
None (0) < 1 h 1 h 2 h 3 h 4 h 5 h 6h 7+ h d0 vs. 1 h d1 h vs. 4 h d1 h vs. 7+ h
Affectionate 2.98 (0.14)
321
2.95 (0.21)
754
2.96 (0.19)
1714
2.96 (0.20)
2704
2.96 (0.20)
1857
2.94 (0.26)
917
2.93 (0.27)
281
2.95 (0.23)
243
2.93 (0.34)
266
−0.11* −0.10* −0.14
Smiles and laughs 2.98 (0.11)
322
2.98 (0.14)
755
2.98 (0.12)
1715
2.98 (0.15)
2705
2.99 (0.12)
1858
2.98 (0.18)
919
2.96 (0.24)
280
2.97 (0.19)
246
2.98 (0.18)
266
0.00 0.00 0.00
Bounces back 2.74 (0.44)
321
2.70 (0.49)
751
2.73 (0.46)
1708
2.72 (0.48)
2701
2.72 (0.49)
1857
2.64 (0.56)
915
2.63 (0.57)
281
2.68 (0.52)
246
2.68 (0.55)
265
−0.01 −0.18* −0.11
Does not lose temper 3.05 (0.53)
181
3.05 (0.48)
400
3.05 (0.53)
1113
3.03 (0.51)
1987
2.99 (0.53)
1406
2.96 (0.57)
709
2.89 (0.59)
205
2.82 (0.68)
190
2.89 (0.68)
210
0.00 −0.16* −0.29*
Can calm down when excited 3.09 (0.60)
180
3.00 (0.57)
397
3.00 (0.61)
1112
3.02 (0.61)
1984
2.99 (0.62)
1403
2.98 (0.62)
710
2.85 (0.65)
205
2.81 (0.68)
191
2.86 (0.68)
210
−0.15 −0.03 −0.23*
Switch tasks without anxiety or
anger
3.49 (0.56)
182
3.44 (0.54)
400
3.49 (0.55)
1114
3.48 (0.56)
1986
3.41 (0.58)
1404
3.40 (0.60)
711
3.38 (0.57)
205
3.32 (0.57)
190
3.39 (0.63)
211
0.00 −0.16* −0.20*
Task self-control (4 items) 3.16 (0.37)
177
3.10 (0.42)
394
3.08 (0.40)
1107
3.06 (0.40)
1980
3.00 (0.42)
1396
2.98 (0.40)
704
2.96 (0.47)
201
2.86 (0.49)
189
2.91 (0.46)
207
−0.21* −0.25* −0.41*
Plays well with others 3.35 (0.54)
180
3.39 (0.53)
399
3.41 (0.54)
1114
3.39 (0.55)
1985
3.35 (0.55)
1407
3.35 (0.57)
711
3.27 (0.66)
203
3.34 (0.61)
190
3.40 (0.61)
210
0.11 −0.11* −0.01
Empathy 3.24 (0.70)
182
3.32 (0.66)
399
3.32 (0.66)
1115
3.30 (0.67)
1989
3.27 (0.70)
1409
3.25 (0.70)
711
3.23 (0.74)
205
3.27 (0.68)
191
3.31 (0.71)
211
0.12 −0.10* −0.01
Notes: 1. Within levels of screen time, numbers in each cell are: well-being means, SDs in parentheses, and n's. 2. d= effect size corresponding to difference in standard deviations. 3. * = t-test comparing cells significant
at p< .05.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
281
Conflict of interest statement
The authors declare that there are no conflicts of interest.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.pmedr.2018.10.003.
References
Allen, M.S., Vella, S.A., 2015. Screen-based sedentary behaviour and psychosocial well-
being in childhood: cross-sectional and longitudinal associations. Ment. Health and
Phys. Act. 9, 41–47.
Babic, M.J., Smith, J.J., Morgan, P.J., Eather, N., Plotnikoff, R.C., Lubans, D.R., 2017.
Longitudinal associations between changes in screen-time and mental health out-
comes in adolescents. Ment. Health and Phys. Act. 12, 124–131.
Chiasson, M.A., Scheinmann, R., Hartel, D., et al., 2016. Predictors of obesity in a cohort
of children enrolled in WIC as infants and retained to 3 years of age. J. Community
Health 41, 127–133.
Common Sense Media, 2015. Landmark report: U.S. teens use an average of nine hours of
media per day, tweens use six hours. https://www.commonsensemedia.org/about-
us/news/press-releases/landmark-report-us-teens-use-an-average-of-nine-hours-of-
media-per-day.
Connelly, B.S., Ones, D.S., 2010. An other perspective on personality: meta-analytic in-
tegration of observers' accuracy and predictive validity. Psychol. Bull. 136,
1092–1122.
Davis, N., 2018. Screen Time Harm to Children is Unproven, Say Experts. The Guardian.
https://www.theguardian.com/science/2018/jun/21/screen-time-harm-to-children-
is-unproven-say-experts (June 21).
de Jong, E., Visscher, T.L.S., HiraSing, R.A., Heymans, M.W., Seidell, J.C., Renders, C.M.,
2013. Association between TV viewing, computer use and overweight, determinants
and competing activities of screen time in 4- to 13-year-old children. Int. J. Obes. 37,
47–53.
Diener, E., Suh, E.M., Lucas, R.E., Smith, H.L., 1999. Subjective well-being: three decades
of progress. Psychol. Bull. 125, 276–302.
Dumuid, D., Olds, T., Lewis, L.K., et al., 2017. Health-related quality of life and lifestyle
behavior clusters in school-aged children from 12 countries. J. Pediatr. 183, 178–183.
Dwyer, R., Kushlev, K., Dunn, E., 2018. Smartphone use undermines enjoyment of face-to-
face social interactions. J. Exp. Soc. Psychol. 78, 233–239.
Erskine, H., Moffitt, T., Copeland, W., et al., 2015. A heavy burden on young minds: the
global burden of mental and substance use disorders in children and youth. Psychol.
Med. 45, 1551–1563.
Fuligni, A.J., Eccles, J.S., 1993. Perceived parent-child relationships and early adoles-
cents' orientation toward peers. Dev. Psychol. 29, 622–632.
Graham, E.K., Rutsohn, J.P., Turiano, N.A., et al., 2017. Personality predicts mortality
risk: an integrative data analysis of 15 international longitudinal studies. J. Res. Pers.
70, 174–186.
Granic, I., Lobel, A., Engels, R.E., 2014. The benefits of playing video games. Am. Psychol.
69, 66–78.
Gunnell, K.E., Flament, M.F., Buchholz, A., et al., 2016. Examining the bidirectional re-
lationship between physical activity, screen time, and symptoms of anxiety and de-
pression over time during adolescence. Prev. Med. 88, 147–152.
Hawton, K., Casanas, I., Conabella, C., Haw, C., Saunders, K., 2013. Risk factors for
suicide in individuals with depression: a systematic review. J. Affect. Disord. 147,
17–28.
Hinkley, T., Verbestel, V., Ahrens, W., et al., 2014. Early childhood electronic media use
as a predictor of poorer well-being: a prospective cohort study. JAMA Pediatr. 168,
485–492.
Kardaras, N., 2017. Glow Kids: How Screen Addition is Hijacking Our Kids – And How to
Break the Trance. St. Martin's Griffin, New York.
Kim, H.H., 2017. The impact of online social networking on adolescent psychological
well-being (WB): a population-level analysis of Korean school-aged children. Int. J.
Adolesc. Youth 22, 364–376.
Kross, E., Verduyn, P., Demiralp, E., et al., 2013. Facebook use predicts declines in sub-
jective well-being in young adults. PLoS One 8, e69841.
Kushlev, K., Proulx, J.E., Dunn, E.W., 2017. Digitally connected, socially disconnected:
the effects of relying on technology rather than other people. Comput. Hum. Behav.
76, 68–74.
Manderscheid, R.W., Ryff, C.D., Freeman, E.J., McKnight-Eily, L.R., Dhingra, S., Strine,
T.W., 2010. Evolving definitions of mental illness and wellness. Prev. Chronic Dis. 7,
A19.
Murray, C.J.L., Vos, T., Lozano, R., et al., 2012. Disability-adjusted life years (DALYs) for
291 diseases and injuries in 21 regions, 1990–2010: a systematic analysis for the
Global Burden of Disease Study 2010. Lancet 380, 2197–2223.
Odgers, C., 2018. Smartphones are bad for some teens, not all. Nature 554, 432–434
(February 22).
Page, A.S., Cooper, A.R., Griew, P., Jago, R.P., 2010. Children's screen viewing is related
Fig. 9. Percentage who do not bounce back, often lose their temper, or cannot calm down when excited or wound up, by total screen time, 2- to 5-year-old children,
with controls, U.S., 2016. Error bars are ± 1 SE.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
282
to physiological difficulties irrespective of physical activity. Pediatrics 126,
1011–1017.
Poitras, V.J., Gray, C.E., Janssen, X., et al., 2017. Systematic review of the relationships
between sedentary behaviour and health indicators in the early years (0–4 years).
BMC Public Health 17, 868.
Przybylski, A.K., Weinstein, N., 2017. A large-scale test of the Goldilocks hypothesis:
quantifying the relations between digital-screen use and the mental well-being of
adolescents. Psychol. Sci. 28, 204–215.
Przybylski, A.K., Weinstein, N., 2018. Digital screen time limits and young children's
psychological well-being: evidence from a population-based study. Child Dev.
Radesky, J., Christakis, D., 2016. Media and young minds. Policy statement of the
American Academy of Pediatrics. Pediatrics 138.
Romer, D., Bagdasarov, Z., More, E., 2013. Older versus newer media and the well-being
of United States youth: results from a national longitudinal panel. J. Adolesc. Health
52, 613–619.
Rosen, L.D., Lim, A.F., Felt, J., et al., 2014. Media and technology use predicts ill-being
among children, preteens and teenagers independent of the negative health impacts
of exercise and eating habits. Comput. Hum. Behav. 35, 364–375.
Ryff, C.D., 1995. Psychological well-being in adult life. Curr. Dir. Psychol. Sci. 4, 99–104.
Satici, S.A., Uysal, R., 2015. Well-being and problematic Facebook use. Comput. Hum.
Behav. 49, 185–190.
Schmiedeberg, C., Schröder, J., 2017. Leisure activities and life satisfaction: an analysis
with German panel data. Appl. Res. Qual. Life 12, 137–151.
Shakya, H.B., Christakis, N.A., 2017. Association of Facebook use with compromised well-
being: a longitudinal study. Am. J. Epidemiol. 185, 203–211.
Shipley, B.A., Weiss, A., Der, G., Taylor, M.D., Deary, I.J., 2007. Neuroticism, extraver-
sion, and mortality in the UK Health and Lifestyle Survey: a 21-year prospective
cohort study. Psychosom. Med. 69, 923–931.
Shor, E., Roelfs, D.J., Yogev, T., 2013. The strength of family ties: a meta-analysis and
meta-regression of self-reported social support and mortality. Soc. Networks 35,
626–638.
Simonsohn, U., 2017. Two-lines: the first valid test of U-shaped relationships. http://
urisohn.com/papers/u_shape.pdf.
Tromholt, M., 2016. The Facebook experiment: quitting Facebook leads to higher levels of
well-being. Cyberpsychol. Behav. Soc. Netw. 19, 661–666.
Turiano, N.A., Chapman, B.P., Gruenewald, T.L., Mroczek, D.K., 2015. Personality and
the leading behavioral contributors of mortality. Health Psychol. 34, 51–60.
Twenge, J.M., Krizan, Z., Hisler, G., 2017. Decreases in self-reported sleep duration
among U.S. adolescents 2009–2015 and links to new media screen time. Sleep Med.
39, 47–53.
Twenge, J.M., Joiner, T.E., Rogers, M.L., Martin, G.N., 2018a. Increases in depressive
symptoms, suicide-related outcomes, and suicide rates among U.S. adolescents after
2010 and links to increased new media screen time. Clin. Psychol. Sci. 6, 3–17.
Twenge, J.M., Martin, G.N., Campbell, W.K., 2018b. Decreases in psychological well-
being among American adolescents after 2012 and links to screen time during the rise
of smartphone technology. Emotion 18, 765–780.
Twenge, J.M., Martin, G.N., Spitzberg, B.H., 2019. Trends in U.S. adolescents' media use,
1976–2016: the rise of digital media, the decline of TV, and the (near) demise of
print. Psychol. Pop. Media Cult.
Valkenburg, P.M., Peter, J., 2009. Social consequences of the Internet for adolescents: a
decade of research. Curr. Dir. Psychol. Sci. 18, 1–5.
van Rooij, A.J., Ferguson, C.J., Colder Carras, M., et al., 2018. A weak scientific basis for
gaming disorder: let us err on the side of caution. J. Behav. Addict. 7, 1–9.
WHO: World Health Organization, 2018. Gaming disorder: online Q&A. http://www.
who.int/features/qa/gaming-disorder/en/ (January).
Yang, F., Helgason, A.R., Sigfusdottir, I.D., Kristjansson, A.L., 2013. Electronic screen use
and mental well-being of 10–12-year-old children. Eur. J. Pub. Health 23, 492–498.
J.M. Twenge, W.K. Campbell Preventive Medicine Reports 12 (2018) 271–283
283