Accepted Article

DR S N GHAEMI (Orcid ID : 0000-0001-9259-0885)

Article type : From Research to Clinical Practice

Digital Depression:

A New Disease of the Millennium?

S. Nassir Ghaemi MD

Professor of Psychiatry, Tufts University School of Medicine, Boston, MA; Lecturer on

Psychiatry, Harvard Medical School, Boston, MA

Address correspondence to Dr. Ghaemi, NIBR, 220 Mass Ave, 2nd floor, Cambridge MA 02138.
Phone: 617-871-8000 Email: [email protected]

Short title: Digital depression

Funding Source: No funding was secured for this review.

Financial Disclosure: Nassir Ghaemi is employed by Novartis Institutes for Biomedical Research,
Cambridge MA.

This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/ACPS.13151
This article is protected by copyright. All rights reserved
 Data availability statement: Data sharing is not applicable to this article as no new data were created
Accepted Article
or analyzed in this study.

Keywords: Social media, depression, anxiety, suicide, adolescence, digital technology, smartphones

Abstract

Objective: In the past decade, since the innovation of the smartphone, there has been an
increase in depression, anxiety and suicidality among teenagers and young adults. The objective of
this article is to review the current evidence for these associations, and to provide initial clinical
guidance. Methods: A narrative review of the available literature on digital technology, social media,
and psychiatric outcomes in adolescents. Results: Psychiatric outcomes have worsened in adolescents
in the past decade, correlating with the invention of the smartphone and the rise of social media.
Depressive symptoms among American teenagers rose rapidly around 2012, and now are reported in
22% of adolescents, which is at least double the rate in adults. Suicide rates have risen, especially
among teenage girls in the US, with a doubling of completed suicide in the past decade. have doubled
in a decade. A causal relationship between social media use and these harmful psychiatric outcomes
is supported by emerging randomized data showing reduced depressive symptoms associated with a
decrease in social media use in college students. Conclusions: Social media and digital technology
correlates with harmful psychiatric outcomes in adolescents and young adults. Clinical
recommendations should include limitations in social media use.

Clinical Recommendations:

This article is protected by copyright. All rights reserved

 1. If clinically meaningful depression and/or anxiety is present, social media use should be
Accepted Article limited to less than one hour daily or stopped altogether.
2. If suicidality is present, social media use should be restricted even further, or stopped
altogether.
3. Children and adolescents should not sleep with their smartphones.

Limitations:

1. Much of the available psychological outcomes are based on self-report evaluations.

2. Most of the available evidence is observational or epidemiological, and not randomized.
3. A causal relationship regarding the associations reported may or may not be present

This article is protected by copyright. All rights reserved

 Introduction
Accepted Article In the last decade, the innovation of the smartphone, and the rise of social media, has
revolutionized daily life all over the world. Besides its many benefits, many harms have been noted,
including the high prevalence of pornography and associated sexual uses to these new technologies.
While such uses among consenting adults might be defended in a liberal society, the effects of such
uses in adolescents and children is more debatable.

This article adds to a nascent literature (1) on the psychiatric effects of social media and the
role of pediatric clinicians in understanding and intervening in this process. Unlike prior reviews, this
article focuses on depression and suicidality, examines causal factors such as substance abuse and
sexual trauma, and provides some preliminary clinical guidelines. Most of this article focuses on
children and adolescents, but an extension to young adults also is provided.

State-of-the-art

Prevalence of social media use among adolescents (ages 13-17)

The first smartphone was introduced in 2007.. In 2018, almost all American teenagers (95%)
used smartphones, and 45% reported that they are online “almost constantly” (2). In the same survey,
use of social media was high, with most frequent use for YouTube (85%), Instagram (72%), Snapchat
(69%), with fewer using Facebook (51%) or Twitter (32%).

The frequency of use of social media among children and adolescents is intense. In a large
national survey of 2600 adolescents and children (3), the typical adolescent is exposed to an average
of about nine hours daily of media screen time (meaning social media, television, and video games).
The average teen spends 2 hours and 42 minutes on a smartphone daily, along with 1 hour and 37
minutes on a computer, and 45 minutes on a tablet. When these and other forms of digital technology
are combined, teens spend a total of 6 hours and 40 minutes of digital technology. Less time is spent
watching television (1 hour and 31 minutes).

This article is protected by copyright. All rights reserved

 The same survey found that for children ages 8-12 (“tweens”), media screen time absorbs
Accepted Article
about 6 hours daily. They spend more time with television (1 hour and 29 minutes daily) than on any
other individual form of digital technology. But when tablets, smartphone use, computer use, and
video games are added up, tweens spend 4 hours and 36 minutes on digital technology. About 15% of
tweens are active users of social media.

Psychiatric symptoms in children and adolescents: Changes in the past decade

Prevalence studies of psychiatric symptoms in children identify notable changes in the past
decade. One prominent study in the United States involves yearly surveys of adolescents conducted
from 1991 to 2016, measuring a total sample of 1.1 million persons. Self-reported mental health
outcomes were self-esteem, life satisfaction, and perceived happiness. These factors took a sudden
shift downwards in 2012 (4). Currently, about 22 percent of teenagers exhibit multiple symptoms of
depression (5). These depressive states are highest in those with three or more hours per day of
digital technology usage. These data are self-reported.

In 2010, a Nielsen survey found that the average adolescent sends 3339 texts monthly, or
about one hundred daily (6). This activity does not include equal or more frequent messaging on
social media, such as Snapchat or Instagram. Many adolescents sleep with their phones, and this
constant interaction goes on often past midnight and into their sleeping hours. This usage of digital
technology in the bedroom may explain an observed 35-43% decrease in sleep time among
adolescents (7).

According to the Centers for Disease Control 2017 annual youth survey, 17.2% of adolescents
had seriously considered attempting suicide in the prior year, a decrease from 29% in 1991. This rate
had decreased in the 1990s until about 2007, and has gradually increased since then. 7.4% of
adolescents reported making a suicide attempt in the prior year (8), similar to prior decades. However,
this adolescent suicide attempt rate is markedly higher than adults, which was only 0.8% in the past
year (8). This difference is nearly ten-fold higher in adolescence than adulthood.

This article is protected by copyright. All rights reserved

 Further, the completed suicide rate among adolescents increased by almost 20% in a recent
Accepted Article
five year period (from 2010 to 2015). This increase is consistent with, and faster than, an overall
general population increase in completed suicide of 33% over two decades (1999 to 2017), according
to the CDC’s National Vital Statistics System (9).

Importantly, there is a gender difference in recent adolescent completed suicide rates, with
much greater harm in females than males (10). In adolescent girls and young women (age 15-24),
from 1999 to 2017, completed suicide rates doubled (5.8 per thousand in 2017 vs 3.0 per thousand in
1999; a 93.3% increase). In adolescent boys and young men, a smaller increase was seen (22.7 per
thousand in 2017 vs 16.8 per thousand in 1999; a 35.1% increase).

Association of social media with depression and anxiety

A 2016 systematic review of mental health and social media (11), found that 29 of 30 studies
report an association with worsened mood and anxiety.. Yet there may be a differential harm effect
among social media platforms, with some being relatively worse than others. In a survey in the UK
(12), of about 1,500 teens and young adults, it was found that 91% used social media. 70% reported
cyber-bullying. The respondents reported increased depression and anxiety symptoms after using
four of the five main social media networks, in order from worst to best: Instagram, Snapchat,
Facebook, Twitter, and YouTube. YouTube was the only network that was associated with overall
positive results for mental health outcomes. Everything else lowered the sense of well-being and
mental health of teenagers, with Snapchat and Instagram being the worst.

These results in teenagers are consistent with social media effects in adults. Facebook, which
is more commonly used by adults, was found to be associated with negative effects on self-reported
mental health (13). In 5208 adults surveyed over three years (2013-2015), one standard deviation of
increased usage of Facebook (status updates, likes clicked, friends clicked) was associated with a 5-
8% decline in self-reported mental health measures. The average adult Facebook user spending an
hour daily on the social media site.

This article is protected by copyright. All rights reserved

 A causal relationship?
Accepted Article There is a correlation of the invention of smartphones and rise of social media, and intense
usage of such digital technology by adolescents and some children, with a major rise of anxiety and
depression among those adolescents and children. The question remains whether this association is
causal.

One line of evidence for causality is a recent randomized trial (14) where reduction in social
media use was associated with improvement in psychiatric symptoms. 143 undergraduate students
were randomized to limit social media use (Facebook, Instagram, Snapchat) to 10 minutes per
platform per day for 4 weeks, or to social media usage as usual. The mean daily social media use at
baseline was about 300 minutes daily, or 5 hours per day. The experimental group social media use
fell to about 150 minutes daily. The experimental group had a lower score on the UCLA loneliness
scale versus the control group. The Beck Depression Inventory (BDI) overall score was not provided
in the paper, but it was noted that since most college students were not appreciably depressed, it
would be hard to detect a decline. However, when students with high baseline BDI scores were
analyzed (above the clinical cut-off of 14), a decline in depression was seen in the experimental group
(baseline BDI 23.0 versus week 4 BDI 14.5, versus no change in the control group (baseline and 4
week BDI was 22.8). The percentage improvement seen in depressive symptoms (37%) was
appreciable. A comparable randomized study (15) using the BDI to assess standard antidepressant
medications found a similar effect size (in a comparable subgroup with baseline BDI of 29.5 ± 8.4,
improvement was seen with the serotonin reuptake inhibitor citalopram to 17.0 ± 10.1, a 42.4%
relative overall improvement). In other words, reducing social media use in a depressed young adult
can have a similar benefit as treatment with an antidepressant.

Possible causal factors

The use of social media itself need not be harmful, unless it is a means for harm. Among
adolescents in particular, negative social media activity often appears to involve interpersonal
bullying, sexual harassment, and communication for the purposes of substance use or abuse.

This article is protected by copyright. All rights reserved

 The psychological context of adolescence is important, with biological and sexual maturation
Accepted Article
occurring at a rapid pace, including partially developed brain maturation, especially in the frontal
lobe, leading to increased impulsivity and poor judgment. Social factors also are relevant. In the
West, since the work of Erikson (16), it has been accepted that adolescence often involves an identity
crisis, which can include rebellion against parental figures and social norms, leading to
experimentation with sex and drugs/alcohol. In recent years in the United States, a liberalization in
parenting styles has been observed, especially among affluent white suburban communities. The
legalization of marijuana in some states, and the availability of nicotine via e-cigarettes, has provided
adolescents with easier access to some addictive substances. Estimates from the federal government
suggest that marijuana use is increasing and beginning to approach frequency of alcohol use in
adolescence (about 40 percent frequency for marijuana and 60 percent for alcohol by the end of high
school) (17).

Along with drug use, sexual activity is increasing. In its 2017 survey,7 the Centers for Disease
Control has documented that by age 18, 45.3 of adolescents have had sexual intercourse, and 34.2%
are currently sexually active. 11.3% of adolescent girls and 3.5% of boys reported having been forced
to have sexual intercourse. About 15 percent are reported to have four or more sexual partners before
age eighteen. Given the availability of smartphone technology, much of this sexual activity is
facilitated by texting and social media, especially Snapchat and Instagram. “Sexting” is common,
even though illegal. About 20 percent of adolescents engage in sexting (18). District attorneys tend to
decline to prosecute such adolescents, leaving little risk to such behavior.

Some studies indicate that the new internet-based sex culture influences boys to demand porn-
like sex, and causes girls to allow it (19). Those studies find that much of social media directs
teenagers towards a hypersexual culture, with 93% of boys and 62% of girls exposed to pornography
(20). It is reported that 74% of Snapchat photos are selfies, often by girls in sexualized poses. 36% of
selfies are altered. About 1000 selfies are posted on Instagram every ten seconds (21). Such new
cultural norms may lead to lowered self-esteem, and anxiety and depression related to how one’s body
looks, or how successfully one conforms to the new norms of an overly sexualized culture.

This article is protected by copyright. All rights reserved

 From Research to Clinical Practice
Accepted Article Prior to the last decade, most psychiatric clinicians were faced with childhood anxiety and
depression in the context of biological diseases, such as bipolar illness or unipolar depression. These
conditions tended to be genetic in origin, and required medication treatment long-term. They were,
however, less common than in adulthood. Now, it appears that there may be a new illness of anxiety
and depression that is much more common in adolescence than in adulthood, a digital depression that
is driven by negative feelings engendered by cyber-bullying, sexual harassment, and increased drug
use, all abetted by social media platforms.

The high rate of depressive state prevalence in adolescents, about 22%, is concerning if
accurate. The lifetime prevalence rate of a full clinical depressive episode in the overall U.S.
population is 5 to 10 percent (22). In adolescents, that number has now doubled in a generation. This
high rate of depression has no biological explanation. A cultural correlation, such as the invention of
the smart phone and digital technology, is plausible. Whether such a link is causal can neither be
proven nor disproven at present, but alternative explanations are not apparent.

Clinical recommendations

At present, any recommendations would be preliminary, based on a rapidly changing

empirical base. With that limitation, the table suggests a possible decision tree approach to managing
social media usage in teenagers.

One approach would be to teach parents about the psychiatric risks and harms of social media,
and to encourage a major restriction of use of smartphones and social media in children and
adolescents Adolescents grow and mature by late adolescence, although even then they can be quite
impulsive. Certainly in middle adolescence years, ages fourteen to sixteen, they need more
supervision than is given typically in terms of their use of social media if the resulting harms are to be
avoided. Delaying smartphone use and social media until high school, preferably after age sixteen,
may be a simple and wise recommendation in the interests of public health.

This article is protected by copyright. All rights reserved

 Experts in social media and mental health (23) are beginning to recommend that some limits
Accepted Article
should exist in use of social media, as opposed to completely uncontrolled usage. Based on the
currently available studies, a conservative recommendation for the general population of teenagers
would be to limit social media use to two hours or less daily. In general, a basic recommendation
could be made as well to remove smartphones from the bedroom, to improve sleep hygiene. In
teenagers who experience anxiety or depression, a differentiation could be made based on severity.
For those teens with mild to moderate anxiety or depressive symptoms, such that they experience such
symptoms but are not clinically diagnosed or treated, it might make sense to reduce social media
usage to one hour or less per day, along with no availability in the bedroom at night. Further, the types
of social media that are most associated with negative psychiatric effects, namely Snapchat and
Instagram, could be prohibited. For teens with severe anxiety or depressive symptoms, such that they
either experience disabling symptoms or are clinically diagnosed or treated, it would be reasonable to
stop all social media use altogether.

Smartphones could be treated like alcohol, something that is best managed in adulthood, with
limitations beforehand.

In addition to tighter limits on social media and smartphone usage, individual psychotherapy
as soon as possible in adolescents with any depression or mood swings, which can help them make
the changes needed in their use of digital technology, and in exploring their attitudes toward drugs
and sexuality.

Limitations

An important limitation is that the psychiatric outcomes here examined relate to symptoms,
not diagnoses. Persons with depressive symptoms do not necessarily meet the clinical threshold for
official clinical diagnoses of “major depressive disorder.” Similarly with anxiety symptoms and
anxiety disorders. However, we know that depressive states are not all-or-nothing. Rather there is a
spectrum, and there is good scientific evidence that medical harms, such as cardiovascular disease and

This article is protected by copyright. All rights reserved

 other kinds of morbidity and mortality, are present even in persons who may not meet clinical
Accepted Article
thresholds of diagnosis (24).

This review is based mostly on correlation between prevalence data on psychiatric symptoms
and social media use. Only two studies with randomized data were available to establish a causal
connection. As is well known, observational correlations may not be causal, due to other confounding
factors. Further, the validity of the self-reported psychiatric symptom data should be assessed with
studies with formal rating scales. The frequency of depression in these studies, for instance, may be
more or less than can be shown by self-report survey data, which was the most commonly used
method. Even self-reported suicide attempt rates in adolescence should be assessed with clinical
validation studies. In the absence of such studies to date, one cannot be certain that other major social
changes in recent years might not be important for adverse psychiatric outcomes.

Yet these considerations are not new, nor specific to this subject. They are classic questions
about the nature of science, and were central to the debate about cigarette smoking and lung cancer.
Half a century ago, the famed founder of clinical epidemiology, A. Bradford Hill, addressed these
concerns and provided guidelines for establishing causality in such settings (25). These guidelines
acknowledge the limitations of observational research, but they also reject the claim that such data are
invalid, and provide means to proceed with careful work along multiple lines of evidence, until more
confidence can be obtained in drawing causal inferences. If we apply Bradford Hill’s nine guidelines
for an association being causation, the available evidence for social media and psychiatric harm begin
to show evidence for seven: strength, consistency, temporality, dose-response, plausibility,
coherence, and experiment (24, 25). Such an approach is now accepted in scientific epidemiology
(26, 27), and are reflected in the basic principles of evidence-based medicine (24, 28, 29).

Just as with lung cancer, which arose in correlation with cigarette smoking as well as the
introduction of the automobile (and related air pollution), proponents and opponents of a new cultural
change can find blame everywhere (30). It will take concerted research on multiple avenues to
establish the facts. It took over half a century for cigarettes to be accepted as causal for lung cancer,
in the absence of randomized studies (which are not feasible for such long-term harms of cultural
behavior) (30). In the meantime, millions of persons perished. For decades, there was no knowledge

This article is protected by copyright. All rights reserved

 about causal mechanisms between cigarette smoking and lung cancer; the relevant carcinogens were
Accepted Article
simply unknown. Similarly, this review can be criticized for not identifying specific mechanisms by
which social media might cause psychiatric harm. That is another line of research which needs to be
explored. These comments are not meant to assert definitively that social media is as harmful to
mental health as cigarette smoking was to lung health. Some analogy is not completely unreasonable
though.

Scientific researchers naturally are cautious in asserting caution, as they should be. But this
caution should be combined with a recognition of the need for physicians, who are not laboratory
scientists, to make decisions in the daily health care of the general population. The rejection of any
such inferences would amount to the rejection of the entire field of clinical epidemiology, which is
essential to medical science and practice (27).

This article is protected by copyright. All rights reserved

Accepted Article
Table. Clinical Recommendations for children and adolescents

All Children with Children with

clinically suicidal ideation
meaningful or attempts
depression/anxiety

Tweens (age 8- No social NA NA

12) media.
Do not sleep
with
smartphones

Teenagers (age Limit social If mild/moderate: No social media.

13-17) media to < 2 Limit social media
hours daily. to <1 hour daily
Do not sleep If severe: No social
with media
smartphones.

This article is protected by copyright. All rights reserved

Accepted Article
Bibliography

1. HOGE E, BICKHAM D, CANTOR J. Digital Media, Anxiety, and Depression in Children. Pediatrics.
2017;140:S76-S80.
2. ANDERSON M, JIANG J. Teens, social media & technology 2018. Pew Research Center; 2018 [updated
2018; cited 2019 June 5]; Available from: http://www.pewinternet.org/2018/05/31/teens-social-media-
technology-2018/.
3. RIDEOUT V. The Common Sense Census: Media Use by Tweens and Teens. . San Francisco: Common
Sense Media; 2015.
4. TWENGE JM, MARTIN GN, CAMPBELL WK. Decreases in psychological well-being among American
adolescents after 2012 and links to screen time during the rise of smartphone technology. Emotion.
2018;18:765-80.
5. TWENGE J, JOINER T, ROGERS M, MARTIN G. 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. 2018;6:3-17.
6. NIELSEN.COM. US teen mobile report calling yesterday, texting today, using apps tomorrow. 2010
[updated 2010; cited]; Available from: https://www.nielsen.com/us/en/insights/news/2010/u-s-teen-mobile-
report-calling-yesterday-texting-today-using-apps-tomorrow.html.
7. TWENGE JM, KRIZAN Z, HISLER G. Decreases in self-reported sleep duration among U.S. adolescents
2009-2015 and association with new media screen time. Sleep Med. 2017;39:47-53.
8. KANN L, MCMANUS T, HARRIS WA, et al. Youth Risk Behavior Surveillance - United States, 2017.
MMWR Surveill Summ. 2018 ;67:1-114.
9. HEDEGAARD H, CURTIN SC, WARNER M. Suicide Rates in the United States Continue to Increase.
NCHS Data Brief. 2018 Jun:1-8.
10. CDC. QuickStats: Suicide rates*,† for teens aged 15–19 years, by sex — United States, 1975–2015.
Atlanta: MMWR Morb Mortal Wkly Rep; 2017. p. 816.
11. SEABROOK EM, KERN ML, RICKARD NS. Social Networking Sites, Depression, and Anxiety: A
Systematic Review. JMIR Ment Health. 2016;3:e50.

This article is protected by copyright. All rights reserved

 12. CRAMER S, INSKTER B. #Status of Mind: Social Media and Young People’s Mental Health and Well-
Accepted Article
Being. London: Royal Society for Public Health; 2017 Contract No.: Document Number|.
13. SHAKYA HB, CHRISTAKIS NA. Association of Facebook Use With Compromised Well-Being: A
Longitudinal Study. Am J Epidemiol. 2017;185:203-11.
14. HUNT M, MARX R, LIPSON C, YOUNG J. No more fomo: Limiting social media decreases depression
and loneliness. J Soc Clin Psychol 2018;37:751–68.
15. WILES NJ, MULLIGAN J, PETERS TJ, et al. Severity of depression and response to antidepressants:
GENPOD randomised controlled trial. Br J Psychiatry. 2012;200:130-6.
16. ERIKSON E. Childhood and society. New York: W. W. Norton; 1950.
17. NIDA. Monitoring the Future Survey: High School and Youth Trends. Washington DC: NIDA; 2017.
18. HINDUJA S, PATCHIN J. Sexting: A Brief Guide for Educators and Parents. Cyberbullying Research
Center; 2010 [updated 2010; cited 2019 June 5]; Available from: https://cyberbullying.org/sexting-a-brief-
guide-for-educators-and-parents.
19. BRAITHWAITE SR, COULSON G, KEDDINGTON K, FINCHAM FD. The influence of pornography on sexual
scripts and hooking up among emerging adults in college. Arch Sex Behav. 2015;44:111-23.
20. SABINA C, WOLAK J, FINKELHOR D. The nature and dynamics of internet pornography exposure for
youth. Cyberpsychol Behav. 2008;11:691-3.
21. RAWHIDE T. Selfie obsession: The rise of social media narcissism. 2015 [updated 2015; cited 2019
June 5]; Available from: https://www.rawhide.org/blog/infographics/selfie-obsession-the-rise-of-social-
media-narcissism.
22. GOODWIN FK, JAMISON KR, GHAEMI SN. Manic-depressive illness : bipolar disorders and recurrent
depression. 2nd ed. New York, N.Y.: Oxford University Press; 2007.
23. WALTERS A. Solution to social media overuse? Put the phones away! Brown Univ. Brown Univ Child
Adolesc Behav Letter. 2017:8.
24. GHAEMI SN. A clinician's guide to statistics and epidemiology in mental health : measuring truth and
uncertainty. Cambridge, UK ; New York: Cambridge University Press; 2009.
25. HILL AB. The Environment and Disease: Association or Causation? Proc R Soc Med. 1965;58:295-300.
26. PHILLIPS CV, GOODMAN KJ. The missed lessons of Sir Austin Bradford Hill. Epidemiol Perspect Innov.
2004;1:3.

This article is protected by copyright. All rights reserved

 27. HOWICK J, GLASZIOU P, ARONSON JK. The evolution of evidence hierarchies: what can Bradford Hill's
Accepted Article
'guidelines for causation' contribute? J R Soc Med. 2009;102:186-94.
28. SACKETT D. Evidence-based medicine. Lancet. 1995 Oct 28;346:1171.
29. SACKETT DL, ROSENBERG WM, GRAY JA, HAYNES RB, RICHARDSON WS. Evidence based medicine:
what it is and what it isn't. BMJ. 1996;312:71-2.
30. HILL G, MILLAR W, CONNELLY J. "The great debate" 1: Smoking, lung cancer, and cancer
epidemiology. Can Bull Med Hist. 2003;20:367-86.

This article is protected by copyright. All rights reserved