Problematic Social Media Use and Health Among Adolescents

Abstract

(1) Background: The use of social media has become an integral part of adolescents’ daily lives. However, the intensive use of social media can develop into a health-threatening addiction, but unfavourable health consequences can occur even with less use. Social media user groups categorized as no-risk, moderate risk (of developing problematic behaviour), and problematic use were examined with reference to their prevalence, their associations with individual determinants and health, and the increased health risk between groups. (2) Methods: The Finnish nationally representative HBSC data (persons aged 11, 13, and 15, n = 3408) and descriptive and binary logistic regression analysis were applied. (3) Results: Problematic social media use (9.4%) was most common among older age groups, and among persons with moderate/low school achievement, low health literacy, and low parental monitoring. Belonging to a moderate risk group (33.5%) was most frequent among girls, and among adolescents with low/moderate parental monitoring and health literacy. All the negative health indicators systematically increased if the respondent belonged to a moderate risk or problematic use group. (4) Conclusions: The study confirmed the association between problematic social media use and negative health outcomes and highlighted the need to pay close attention to adolescents at moderate risk who exhibited negative health outcomes.
Keywords: social media; problematic social media use; adolescents; health

Full text

Int. J. Environ. Res. Public Health 2021, 18, 1885. https://doi.org/10.3390/ijerph18041885 www.mdpi.com/journal/ijerph
Article
Problematic Social Media Use and Health Among Adolescents
Leena Paakkari *, Jorma Tynjälä, Henri Lahti, Kristiina Ojala and Nelli Lyyra
Faculty of Sport and Health Sciences, University of Jyväskylä, P.O. Box 35 (L), 40014 Jyväskylä, Finland;
jorma.a.tynjala@jyu.fi (J.T.); henri.o.lahti@jyu.fi (H.L.); kristiina.ojala@jyu.fi (K.O.); nelli.lyyra@jyu.fi (N.L.)
* Correspondence: leena.paakkari@jyu.fi
Abstract: (1) Background: The use of social media has become an integral part of adolescents’ daily
lives. However, the intensive use of social media can develop into a health-threatening addiction,
but unfavourable health consequences can occur even with less use. Social media user groups cate-
gorized as no-risk, moderate risk (of developing problematic behaviour), and problematic use were ex-
amined with reference to their prevalence, their associations with individual determinants and
health, and the increased health risk between groups. (2) Methods: The Finnish nationally repre-
sentative HBSC data (persons aged 11, 13, and 15, n = 3408) and descriptive and binary logistic
regression analysis were applied. (3) Results: Problematic social media use (9.4%) was most com-
mon among older age groups, and among persons with moderate/low school achievement, low
health literacy, and low parental monitoring. Belonging to a moderate risk group (33.5%) was most
frequent among girls, and among adolescents with low/moderate parental monitoring and health
literacy. All the negative health indicators systematically increased if the respondent belonged to a
moderate risk or problematic use group. (4) Conclusions: The study confirmed the association be-
tween problematic social media use and negative health outcomes and highlighted the need to pay
close attention to adolescents at moderate risk who exhibited negative health outcomes.
Keywords: social media; problematic social media use; adolescents; health
1. Introduction
The recently published international report on Health Behaviour in School-aged
Children [1] covered the health and health behaviour of 11-, 13-, and 15-year-old adoles-
cents in 45 countries. The findings included a figure of 7% of adolescents reporting prob-
lematic social media use. Furthermore, 35% of adolescents could be characterized as in-
tensive electronic media users, having online contact with people almost all the time, and
throughout the day [1]. One may thus agree with the notion of Griffiths and Kuss [2] that
“teenagers particularly appear to have subscribed to the cultural norm of continual online
networking”. Social media refers to various internet-based semi-public and public sites
and services (e.g., social network sites, blogging, and video sharing), and related tools that
provide spaces for user-profiles and for user-generated content, socializing, and sharing
[3,4]. Undoubtedly, for today’s young people, who have grown up with instant access to
social media [5], the social media offer spaces and possibilities to find friends, reduce so-
cial isolation, and gain social support [6], with opportunities also for learning, creativity,
and self-fulfilment. However, if the use develops into problematic behaviour, it may bring
unfavourable consequences with it.
Research on problematic social media use, disorder, or addiction among adolescents
has increased during the last decade. However, many scholars do not refer to problematic
social media use as an addiction or disorder, as the phenomenon has not yet been officially
classified as such, e.g., [7]. Though a growing body of research has shown a link between
problematic social media use and unfavourable health, the data leading to results and
conclusions have mainly been self-reported, as have the data leading to classification as a
Citation: Paakkari, L.; Tynjälä, J.;
Lahti, H.; Ojala, K.; Lyyra, N.
Problematic Social Media Use and
Health Among Adolescents. Int. J.
Environ. Res. Public Health 2021, 18,
1885. https://doi.org/10.3390/ijerph
18041885
Academic Editor: Paul B.
Tchounwou
Received: 20 December 2020
Accepted: 10 February 2021
Published: 15 February 2021
Publisher’s Note: MDPI stays neu-
tral with regard to jurisdictional
claims in published maps and insti-
tutional affiliations.
Copyright: © 2021 by the authors. Li-
censee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and con-
ditions of the Creative Commons At-
tribution (CC BY) license (http://crea-
tivecommons.org/licenses/by/4.0/).

Int. J. Environ. Res. Public Health 2021, 18, 1885 2 of 11
problematic user. According to Zendle and Bowden-Jones [8], given the lack of diagnostic
criteria and the limited amount of high-quality longitudinal research, not much can be
said about the nature of excessive social media use and its similarities with known addic-
tions. It is for this reason that somewhat broad terms, such as problematic social media
use, have been adopted.
Similarly, because there is no shared understanding of what such problematic behav-
io ur means or how it s hould be me asured , there are wide va riatio ns in de fining the curren t
level of problematic use in the population [9,10]. Furthermore, if intensive engagement
with social media and networking is seen as the status quo, it can be hard to define when
the use becomes problematic [2].
Some scholars [9] have characterized problematic social media use via components
that apply to all addictions [11], while others [10] have used criteria based on the Diag-
nostic and Statistical Manual of Mental Disorders (DSM-V) [12]. These have been applied
to internet gaming disorder, with possibilities for defining social media disorder more
generally. Following this approach, the following nine components could be used to de-
fine and measure social media disorder: preoccupation (“substantial time used to think
about the activity”), withdrawal, tolerance, problems in reducing/stopping (“unsuccessful
attempts to stop”), giving up other activities, continuance despite problems, deception or
covering up, use to escape from or relieve negative moods, and indications of risk or loss
regarding relationships or opportunities [13]. These nine components formed the basis of
the HBSC study referred to above, which used the Social Media Disorder Scale with nine
yes/no items [10]. Respondents with six or more “yes” answers were categorized as prob-
lematic users [14].
Adolescents’ problematic social media use has been viewed as a complex phenome-
non with several explanatory factors. Problematic use has been explained via individual-
level factors (e.g., fear of missing out [15,16]), family-level factors (e.g., childhood mal-
treatment [17], lessened family support [18], lower family functioning [19]), and/or friend-
level factors (lower levels of friend support [18]). Macro-level factors have also been pro-
posed, including the normalization of the surveillance culture, involving online compari-
son, and possibilities to find out what people are doing [20]. Gender and age have often
been regarded as explanatory factors for disordered social media use. Girls are more likely
to report problematic social media use than boys [1,9,21–24], and older girls more likely
than younger girls to report such use [1]. Boys have been found to be more likely to report
disordered internet gaming [23]. Across countries, no consistency has been found regard-
ing the association between social media use and family affluence [1].
The fairly extensive use of social media among adolescents has raised concerns about
the consequences for adolescents’ health and wellbeing e.g., [25,26]. Most research has
focused on the link between social media use (including social network site addiction) and
various psychosomatic complaints or disorders. Following a longitudinal study, Vannucci
and Ohannessian [27] reported that “social media patterns appear to differentially predict
psychosocial adjustment during early adolescence, with high social media use being the
most problematic”. Problematic social media use has been linked to a greater likelihood
of depressive symptoms [5,9,22,28–32], anxiety [28,32], lower self-esteem [9,28,32], social
isolation [29], lower life-satisfaction [18], poorer sleep quality [21,28,32], disordered eating
[33], and higher body image dissatisfaction [34]. Considering the associations with the
various psychosomatic complaints, it is easy to understand why assessment of social me-
dia use has been proposed as a standard element in psychiatric assessment [35]. Further-
more, problematic social media use has been associated with unfavourable behaviours
such as lower levels of physical activity [19] and impaired ability to regulate daily respon-
sibilities [28], while higher social media use has been associated with the initiation of risky
behaviours (e.g., substance use and risky sexual behaviours, [36]). For their part, digital
stimuli, notably the rapid shifts in attention caused by high use of social media have been
associated with cognitive factors such as a momentarily shortened attention span and low-
ered information processing [37]. Overall, there are indications that problematic and high

Int. J. Environ. Res. Public Health 2021, 18, 1885 3 of 11
social media use may predict a heightened risk for an array of negative health outcomes
among adolescents.
Despite a recent increase in research on social media use and health, there have been
calls for more research “to protect children and young people from the negative effects
social media can have on their risk-taking behaviours, mental health, and wellbeing” [6].
This underlines the utility of obtaining a nationally representative sample to determine
how social media use is associated with a comprehensive set of health indicators (exam-
ined in the same sample), and whether an association with negative health indicators ex-
ists among persons adjudged to have a heightened (albeit moderate) risk for problematic
use. The study by Bányai and her colleagues [9] using a nationally representative Hun-
garian sample is one of the few to reveal that the likelihood of poorer health (i.e., lower
self-esteem and a higher level of depressive symptoms) is already increased among ado-
lescents with a heightened risk of problematic social media use, as compared to those with
no such risk. Hence, the rationale for this study is not only the need to examine in depth
the use of social media in relation to health, but also to use nationally representative data
encompassing various individual determinants and health indicators. So far, there has
been no such research on Finnish school-aged children.
This study aimed to examine the association between adolescents’ social media use
and health plus health behaviour, using nationally representative Finnish data from the
Health Behaviour in School-aged Children (HBSC) survey. The specific research questions
for the study were:
• What are the comparative prevalence figures for groups categorized as no-risk (i.e.,
at no risk for problematic social media use), moderate risk (i.e., at heightened risk of
problematic social media use), and problematic in their social media use? (RQ1)
• How are various individual determinants (gender, age, family affluence, parental
monitoring, health literacy, academic aspiration, school achievement) associated
with social media use? (RQ2)
• How are various social media user groups (no-risk, moderate risk, and problematic
use) associated with various health indicators (health complaints, physical inactiv-
ity, loneliness, low self-rated health, morning tiredness, short sleep) (RQ3), and,
how much does the risk of negative health indicators increase between groups?
(RQ4)
2. Materials and Methods
2.1. Design
Nationally representative data were collected from Finnish adolescents as part of the
HBSC study in 2018. Stratification in sampling was based on European Union NUTS clas-
sification (Nomenclature of Territorial Units for Statistics). The sampling was based on
the Finnish national school register and was carried out via a random cluster sampling
method adjusted for the province, municipality, and school size. Probability proportional
to size sampling method (PPS) was applied in school selection. The primary sampling unit
(PSU) was the school, and within the school the class was randomly selected. Participation
was voluntary, and the student response rate was 57%. Anonymous data were collected
online by Webropol software (Webropol Oy, Helsinki, Finland). The Ethics committee of
the University of Jyväskylä gave approval regarding ethical issues, and the school princi-
pals gave school-level approval.
2.2. Participants
The HBSC study focuses on collecting the data from 11-, 13-, and 15-year-olds fol-
lowing the international HBSC protocol [14]. The participants were 3408 Finnish adoles-
cents (11 yrs, n = 993; 13 yrs, n = 1246; 15 yrs, n = 1169). The sample included an almost

Int. J. Environ. Res. Public Health 2021, 18, 1885 4 of 11
equal number of boys (n = 1706) and girls (n = 1702), and gender and age were not associ-
ated (χ²(2) = 0.027, p = 0.987).
2.3. Measures
Self-reported gender and age were measured by asking participants to select the cor-
rect alternative.
Problematic social media use was measured with the 9-item Social Media Disorder Scale
(SMD scale) using a dichotomous (No/Yes) answer scale [8]. The items covered preoccu-
pation, tolerance, withdrawal, displacement, escape, problems, deception, displacement,
and conflict. Based on the values obtained, the respondents were categorized into three
groups: a no-risk group, a moderate risk group (i.e., at a heightened risk of developing
problematic use), and a problematic use group. The cut-off value for the problematic use
group was 6 or more yes answers; for the moderate risk group the corresponding values
were 2 to 5, and for the no-risk group 0 to 1 [38].
Family affluence scale III [39] was used to measure the self-reported socioeconomic
position, via six items: ownership of a car, having one’s own bedroom, number of family
computers, number of bathrooms, ownership of a dishwasher at home, number of family
holidays during the past 12 months. The computed scores were recoded into three cate-
gories to indicate levels of relative affluence: low family affluence (lowest 20%), medium
family affluence (middle 60%), and high family affluence (highest 20%) [40]. Parental mon-
itoring was measured by six items, via a 3-point scale focusing on adolescents’ perceptions
of parental monitoring and awareness regarding friends, spending money, after-school
and free-time activity, internet activity, and going out at night [41]. The responses for
monitoring by both mother and father were computed to obtain a sum score that was
recoded into three categories: low parental monitoring (lowest 33.3%), medium parental
monitoring (middle 33.3%), and high parental monitoring (highest 33.3%).
Educational aspiration was measured by a question covering seven educational/voca-
tional tracks after comprehensive school. The response options upper secondary school
and double examination (conducted at both upper secondary school and vocational
school) were combined and labelled academic plans, while the others were labelled non-
academic plans. Health literacy was measured by the Health Literacy for School-Aged Chil-
dren (HLSAC) instrument [42,43]. The scale includes 10 items measuring adolescents’
knowledge and competencies to promote health. The responses were computed to a sum
score, which was then recoded into three categories: low health literacy (values 10–25),
medium health literacy (values 26–35), and high health literacy (values 36–40) [44]. Aca-
demic achievement was measured by asking students to indicate their most recent mark in
the school subjects first language and mathematics. The mean value for the two marks
was calculated and recoded into three categories: low school achievement (mean value 4–
7), medium school achievement (mean value 7.5–8.5), and high school achievement (mean
value 9–10).
Health complaints were measured by the HBSC symptom checklist (HBSC-SCL) [45].
Respondents evaluated the frequency of various complaints, both somatic (headache,
neck, and shoulder pain) and psychological (feeling low, nervousness, irritability) over
the last six months. The response options about every week, more than once a week, and
about every day were combined to indicate frequent and regularly experienced psycho-
somatic symptoms. Multiple health complaints were defined as experiencing three or
more health complaints weekly or more often. Physical inactivity was measured using
Moderate-to-Vigorous-Physical-Activity (MVPA) questions [46]. Persons meeting the
MVPA guidelines on only 0 to 2 days per week were categorized as physically inactive.
Loneliness was measured by one item: Do you ever feel lonely? The response options very
often and quite often were combined to indicate frequently experienced loneliness. Self-
rated health (SRH) was evaluated by a single question measuring the individual’s percep-
tion and evaluation of their health [47]. The response options fair and poor were combined
to indicate low SRH.

Int. J. Environ. Res. Public Health 2021, 18, 1885 5 of 11
Morning tiredness was measured with a single item: How often do you feel tired when
you get up on school mornings? The response options varied from rarely or never to 4 or
more times a week. Being tired 4 or more times a week was categorized as a risk for ado-
lescent health. Sleep duration was calculated as the difference between bedtime and wake-
up time on school days. Bedtimes were inquired about as follows: When do you usually
go to bed if the next morning is a school day? Wake-up times were asked via the question:
When do you usually wake-up on school mornings? The response alternatives for both
questions were set at half-hour intervals. Sleep duration of at most 7 hours a night was
categorized as a risk for adolescent health and referred to as short sleep.
2.4. Analyses
The descriptive analysis included frequencies, percentages, confidence intervals
(CIs), and associations between social media use and background variables (χ² test).
The associations between social media addiction and health indicators were analysed
by binary logistic regression. The analyses were adjusted for PSU. Odds ratio (OR) values
were calculated, indicating the strength of the association and the increased risk level.
Missing data were handled via a listwise deletion procedure. The significance level for all
the analyses was set at p < 0.05. All the analyses were conducted by Stata (version 16)
(StataCorp LLC, Texas, USA).
3. Results
3.1. Prevalence of No-Risk, Moderate Risk, and Problematic Social Media Use (RQ1), and Asso-
ciated Individual Determinants (RQ2)
For the total sample, the problematic social media prevalence was 9.4% and the mod-
erate risk prevalence 33.5% (Table 1). Overall, no gender differences were found regarding
problematic use, but girls more than boys belonged to the moderate risk group. Problem-
atic use was more prevalent among 13- and 15-year-old adolescents than among 11-year-
olds (11.2% for the 13- and 15-year-olds versus 5.9% for the 11-year-olds). Furthermore,
higher parental monitoring and higher health literacy were statistically significantly asso-
ciated with a lower prevalence of moderate risk and problematic social media use. Low
academic achievement was linked to a higher prevalence of problematic social media use.
There was no association between family affluence or educational aspiration and the prev-
alence of no-risk, moderate risk, or problematic social media use.

Int. J. Environ. Res. Public Health 2021, 18, 1885 6 of 11
Table 1. The prevalence of no-risk, moderate risk, and problematic social media user groups: totals, and proportions by
gender, age, family affluence, parental monitoring, educational aspiration, health literacy, and academic achievement.
Variable No-Risk Social Media Use Moderate Social Media Use Problematic Social Media Use Total χ²(df); p-Value
% (95% CI) % (95% CI) % (95% CI) % (n)
All 57.1 (54.8–59.3) 33.5 (31.6–35.5) 9.4 (8.3–10.8) 100 (3077)
Gender χ²(2) = 76.98; <0.001
Boys 65.1 (62.2–67.8) 26.6 (24.1–29.3) 8.3 (6.9–10.1) 100 (1496)
Girls 49.6 (46.5–52.7) 40.0 (37.2–42.8) 10.5 (8.8–12.4) 100 (1581)
Age χ²(4) = 27.06; <0.001
11 years 60.5 (56.7–64.2) 33.6 (30.3–37.2) 5.9 (4.5–7.6) 100 (912)
13 years 53.5 (49.6–57.4) 35.3 (31.6–39.1) 11.2 (9.0–14.0) 100 (1122)
15 years 57.1 (53.5–60.6) 31.7 (28.9–34.7) 11.2 (9.2–13.5) 100 (1043)
FAS χ²(4) = 4.89; <0.353
Low 57.3 (53.0–61.4) 32.2 (28.4–36.1) 10.6 (8.3–13.4) 100 (663)
Medium 56.1 (53.2–58.9) 34.8 (32.4–37.2) 9.2 (7.7–10.8) 100 (1782)
High 60.2 (55.7–64.5) 31.6 (27.4–36.0) 8.2 (6.1–11.0) 100 (552)
Parental monitoring χ²(4) = 93.30; <0.001
Low 46.1 (41.8–50.4) 37.3 (33.6–41.1) 16.7 (13.8–19.9) 100 (644)
Moderate 51.4 (46.9–55.9) 38.9 (34.5–43.5) 9.7 (7.6–12.1) 100 (669)
High 68.7 (64.9–72.3) 25.5 (22.0–29.4) 5.8 (4.1–8.0) 100 (681)
Educational aspiration χ²(2) = 16.66; <0.080
High school 58.3 (54.3–62.3) 32.3 (29.0–35.9) 9.4 (7.2–12.1) 100 (700)
Vocational 55.2 (49.3–61.1) 30.6 (25.7–36.0) 14.2(10.7–18.6) 100 (321)
Health literacy χ²(4) = 94.80; <0.001
Low 42.8 (33.9–52.2) 34.7 (27.0–43.3) 22.4 (15.8–30.8) 100 (180)
Moderate 50.4 (47.3–53.5) 38.2 (35.1–43.3) 11.2 (9.2–14.0) 100 (1175)
High 65.2 (61.2–68.9) 25.9 (22.7–29.4) 9.0 (7.0–11.4) 100 (739)
Academic achievement χ²(4) = 54.74; <0.001
Low 53.2 (47.8–58.5) 30.3 (26.1–34.8) 16.6 (12.9–20.9) 100 (535)
Moderate 55.2 (51.5–59.0) 33.0 (29.5–36.7) 11.8 (10.0–14.3) 100 (970)
High 57.7 (52.8–62.5) 37.0 (32.4–41.7) 5.4 (3.7–7.9) 100 (585)
3.2. Prevalence of Negative Health Indicators among No-Risk, Moderate Risk, and Problematic
Users (RQ3)
No-risk users experienced the fewest health complaints or other negative health in-
dicators, and problematic users reported the most. In the total sample, the prevalence of
multiple health complaints within the moderate risk group was 33.0%, and in the prob-
lematic group 41.4% (Table 2). Overall, among the problematic users the most common
negative health indicators were irritability and nervousness.
Table 2. Prevalence of negative health indicators among no-risk, moderate risk, and problematic user groups.
Health indicators No-risk social media use Moderate risk social me-
dia use
Problematic social media
use χ²(df); p-value
% (95% CI) % (95% CI) % (95% CI)
Health complaints a
Headache 32.0 (29.5–34.6) 46.3 (42.7–49.8) 51.9 (46.0–57.8) χ²(2) = 79.4; <0.001
Neck & shoulder pain 26.4 (24.1–28.9) 37.4(34.1–40.9) 47.5 (41.2–53.9) χ²(2) = 71.1; <0.001
Feeling low 26.1 (23.6–28.8) 49.2 (45.5–52.9) 53.5 (47.2–59.6) χ²(2) = 188.4; <0.001
Nervousness 36.4 (33.8–39.1) 59.4 (55.7–62.9) 65.6 (59.7–71.0) χ²(4) = 182.4; <0.001
Irritability 42.4 (39.6–45.1) 64.8 (61.4–68.1) 70.7 (65.1–75.8) χ²(2) = 173.4; <0.001
Multiple health complaints 14.6 (12.8–16.6) 33.0 (29.7–36.6) 41.4 (35.3–47.9) χ²(2) = 181.8; <0.001
Other health indicators b
Physical inactivity 10.2 (8.5–12.2) 12.1 (10.1–14.8) 23.4 (19.1–28.3) χ²(2) = 40.6; <0.001
Loneliness 9.2 (7.7–10.9) 18.4 (15.9–21.2) 27.4 (22.2–33.2) χ²(4) = 93.6; <0.001
Low SRH 12.6 (10.9–14.6) 18.8 (16.2–21.7) 29.7 (24.3–358) χ²(2) = 60.1; <0.001
Morning tiredness 25.3 (23.0–27.7) 38.0(34.7–41.4) 55.9 (49.3–61.8) χ²(2) = 126.2; <0.001
Short sleep 17.7 (15.2–20.5) 23.9 (20.4–28.0) 39.3 (32.9–46.0) χ²(2) = 55.7; <0.001
Note: a percentage of respondents in each social media use group reporting health complaints weekly or more often. b
percentage of respondents in each social media use group reporting other negative health indicators.

Int. J. Environ. Res. Public Health 2021, 18, 1885 7 of 11
3.3. Associations between Social Media User Groups and Negative Health Indicators (RQ4)
Adolescents in the problematic social media user group were three times more likely,
and adolescents in the moderate risk group twice as likely to experience irritability, nerv-
ousness, and feeling low. In addition, problematic users were twice as likely to suffer from
neck and shoulder pain and headache as compared to the no-risk group. All the OR values
for the moderate risk and problematic group were statistically significant at level p < 0.001.
(Table 3).
Table 3. Binary logistic regression on the associations between social media user groups and psychosomatic health com-
plaints.
Groups
Health complaints
Headache Neck and shoulder
pain Feeling low Nervousness Irritability Multiple health
complaints
OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
Addiction group
No-risk 1 1 1 1 1 1
Moderate risk 1.68 (1.40–2.02) *** 1.57 (1.30–1.89) *** 2.46 (2.05–2.96) *** 2.35 (1.96–2.81) *** 2.31 (1.95–2.73) *** 2.59 (2.10–3.20) ***
Problematic 2.13 (1.65–2.75) *** 2.35 (1.76–3.12) *** 2.94 (2.22–3.88) *** 3.11 (2.37–4.07) *** 3.05 (2.34–3.98) *** 3.77 (2.77–5.13) ***
Gender
Boy 1 1 1 1 1 1
Girl 1.76 (1.47–2.10) *** 1.54(1.30–1.83) *** 2.71 (2.25–3.26) *** 1.79 (1.51–2.11) *** 1.86 (1.59–2.17) *** 2.52 (2.06–3.07) ***
Age
11 years 1 1 1 1 1 1
13 years 1.09 (0.85‒1.39) 1.11 (0.87–1.40) 1.19 (0.91–1.56) 1.19 (0.94–1.50) 1.14 (0.89–1.46) 1.12 (.85–1.47)
15 years 1.25 (1.01–1.54) * 1.48 (1.19–1.84) ** 1.73 (1.34–2.21) *** 1.11 (0.89–1.39) 1.25 (1.00–1.57) 1.65 (1.27–2.15) ***
Note: * p < 0.05, ** p < 0.01, *** p < 0.001.
The prevalence of morning tiredness, short sleep, and loneliness was over three times
as likely, and inactivity twice as likely among the problematic social media users as com-
pared to the no-risk users. Furthermore, loneliness was twice as likely to be experienced
by the moderate risk group as compared to the no-risk group (Table 4).
Table 4. Binary logistic regression on the associations between social media user groups and other negative health indi-
cators.
Groups
Health indicators
Physical inactivity Loneliness Low SRH Morning tiredness Short sleep
OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
Addiction group
No-risk 1 1 1 1 1
Moderate risk 1.25 (0.96–1.65) 2.05 (1.62–2.59) *** 1.60 (1.28–1.99) *** 1.76 (1.47–2.09) *** 1.60 (1.25–2.06) ***
Problematic 2.46 (1.81–3.35) *** 3.37 (2.42–4.70) *** 2.68 (1.99–3.61) *** 3.43 (2.61–4.51) *** 3.23 (2.32–4.50) ***
Gender
Boy 1 1 1 1 1
Girl 1.00 (0.77–1.31) 1.94(1.54–2.46) *** 1.02 (0.82–1.26) 1.26 (1.07–1.50) ** 0.73 (0.59–0.92) **
Age
11 years 1 1 1 1 NA
13 years 2.11 (1.43–3.11) *** 1.18 (0.85–1.64) 2.12 (1.54–2.93) *** 1.44 (1.16–1.78) ** 1
15 years 4.84 (3.42–6.85) *** 1.95 (1.43–2.65) *** 2.31 (1.64–3.26) *** 1.85 (1.51–2.27) *** 1.60 (1.22–2.09) **
Note: ** p < 0.01, *** p < 0.001.
4. Discussion
Using a nationally representative sample from Finland, this study sheds new light
on adolescents’ social media use and related individual determinants and health indica-
tors. It represents one of the few to offer information on how being a moderate or prob-
lematic social media user increases the likelihood of reporting various unfavourable
health indicators; hence, it highlights the need to examine not only problematic versus

Int. J. Environ. Res. Public Health 2021, 18, 1885 8 of 11
non-problematic users, but also persons at moderate risk of developing problematic so-
cial media use.
Problematic social media use is of concern for 9.4% of Finnish adolescents. This is
close to the European average (7.4%) [18], while a further 33.5% could be characterized as
at moderate risk of developing problematic behaviour. Problematic social media use was
most common among the older age groups, and among persons with moderate or low
school achievement, low health literacy, and low parental monitoring. Furthermore, be-
longing to a moderate risk group was most frequent among girls and among persons with
low to moderate parental monitoring. Family wealth was not linked to social media use
in any of the examined groups. One reason for this may be the inexpensive cost of internet
access in Finland, as reported by the European Commission [48], which notes that “Fin-
land has developed an equitable and inclusive information society”.
Our findings did not support recent findings on girls as more likely than boys to
report problematic social media use [9,21]. However, a larger proportion of girls (40%)
than boys (28%) belonged to the group with a moderate risk for problematic use, indicat-
ing a need to pay special attention to girls in attempts to prevent problematic social media
behaviour. Research has shown that several social media behaviours, such as selfie behav-
iour, are typical of both genders, but that for instance appearance concerns explain prob-
lematic social media use only among boys [49]. This highlights the need to study in detail
the gender differences in social media use and the patterns explaining them. In addition,
since this is the first time that Finnish adolescents’ social media use has been measured
via a specific problematic-use instrument, we do not know how gender differences have
developed over time.
Problematic social media use by adolescents has widely been recognized as a possible
indicator for negative health outcomes e.g., [18,21,50]. The findings clearly indicated that
among the moderate risk users and the problematic users, there was a higher frequency
of various psychosomatic complaints and other negative health indicators than in the no-
risk group. For example, among the adolescents identified as belonging to the no-risk
group, 15% reported having multiple health complaints and 36% nervousness on a weekly
or daily basis. In the moderate risk group, the corresponding figures were 33% (multiple
health complaints) and 59% (nervousness), and in the problematic use group 41% (multi-
ple health complaints) and 66% (nervousness). Furthermore, the prevalence of weekly
morning tiredness increased from 25% (no-risk group) to 56% (problematic use group),
and short sleep from 18% to 39%. One alarming finding was that the risk of loneliness was
double for the moderate risk group and triple for the problematic use group. There have
been mixed findings on the effects of social media on loneliness, and the differences have
been explained via the individual’s orientation to social comparison (such that persons
with a higher orientation towards social comparison have an increased risk of social me-
dia increasing loneliness) [51].
Given that the risk for all the measured negative health indicators systematically in-
creased if the respondent belonged to a moderate risk or a problematic use group, there
is a need to broaden the scope of investigation to persons at moderate risk of developing
problematic behaviour instead of studying only problematic versus no-risk users. We
found an increased risk for 9 out of 11 measured unfavourable health indicators among
the moderate risk group, and often the risk was over twice as large as that for the no-risk
group. Moreover, in relation to 5 out of 11 measured unfavourable health indicators
(headache, feeling low, nervousness, irritability, and having multiple complaints), the risk
increased if the respondent belonged to the moderate risk group as compared to the no-
risk group, with no detectable difference between moderate risk and problematic users.
In addition, given that 43% of the Finnish adolescents belonged to either the moderate risk
(34%) or the problematic group (9%), with only small age differences, the magnitude of
the findings on the associations between social media use and health should not be un-
derestimated. As Clark et al. [6] have argued, “the health and societal costs of problematic
use of the internet across the lifespan are unknown, but they could be huge”. The critical

Int. J. Environ. Res. Public Health 2021, 18, 1885 9 of 11
question is what “moderate risk” actually means. In a positive sense, one could say that
at this stage adolescents still have control over their social media use. However, it is diffi-
cult to say how near persons in this group are to developing a problematic behaviour.
The cross-sectional data do not allow causal inferences and arguments on whether
various negative health outcomes predispose towards or are consequences of problematic
so cial media use [5] . In additi on, in t his study we examined the associations between prob-
lematic social media use and health indicators but did not assess the different types of
social media used by the participants. Hence, caution is needed in generalizing from the
results, given that the associations of different types of media can vary greatly. One should
also note the role of probable mediating factors between social media use and health. For
instance, Viner et al. [52] found that a lack of sleep or of physical activity mediated the
association between social media use and various mental health problems, and they con-
sidered direct causalities between extensive social media use and health problems to be
very unlikely. However, in an experimental study among university students conducted
by Hunt et al. [53] it was observed that limitations on social media use may decrease lone-
liness and depressive symptoms. Heffer et al. [54] took the view that that social media use
did not in itself predict mental health problems; in fact, causality operated the other way
round. Undoubtedly, further research is needed to avoid a situation in which various
health problems are explained purely via social media use, with technology regarded as
“an illness that society must address” [3]. According to Boyd [3], “It is easier for adults to
blame technology for undesirable outcomes than to consider other social, cultural, and
personal factors that may be at play.” More broadly, research is needed to explore prob-
lematic social media use and its associations with health in a more comprehensive man-
ner, using for instance an ecological approach to identify explanatory factors across sev-
eral domains (i.e., individual, family, peer, school and community) in the same sample,
and at the same time.
5. Conclusions
Based on the findings of this study, more research is needed on the role of health
literacy and school achievement in empowering adolescents to have control over their
social media use, with attention given also to the role of parental monitoring in creating a
supportive environment for more controllable social media use. However, one can sug-
gest that preventative interventions should be comprehensive in nature, i.e., they should
focus on individuals and their environments, and be targeted at the whole population,
albeit with a special focus on those at moderate risk and those with problematic habits.
One must accept that the internet and social media play a significant role in people’s lives,
and that prohibitions of it are practically impossible [55]. Hence, interventions should fo-
cus on factors that support adolescents’ skills in regulating their own behaviour [56].
Author Contributions: Conceptualization, L.P., J.T., and N.L.; methodology, J.T, N.L, and L.P.; for-
mal analysis, J.T and N.L.; investigation, L.P., J.T., H.L., K.O., and N.L..; resources, L.P.; writing
(preparation of original draft) L.P., J.T., H.L., K.O., and N.L.; writing (review and editing) L.P., J.T.,
H.L., K.O., and N.L.; visualization, N.L.; funding acquisition, L.P. All authors have read and agreed
to the published version of the manuscript.
Funding: This research was funded by Juho Vainio Foundation (grant) and Ministry of Social Af-
fairs and Health.
Institutional Review Board Statement: The study was conducted according to the guidelines of the
Declaration of Helsinki, and approved by the Ethical Committee of the University of Jyväskylä.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the
study.
Data Availability Statement: The data presented in this study are available on reasonable request
from the corresponding author. The data are not publicly available due to ethical requirements.

Int. J. Environ. Res. Public Health 2021, 18, 1885 10 of 11
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the
design of the study; in the collection, analyses, or interpretation of data; in the writing of the manu-
script, or in the decision to publish the results.
References
1. Inchley, J.; Currie, D.; Budisavljevic, S.; Torsheim, T.; Jåstad, A.; Cosma, A.; Kelly, C.; Már Arnarsson, A.; Barnekow, V.; Weber,
M.M. Spotlight on Adolescent Health and Well-Being. Findings from the 2017/2018 Health Behaviour in School-Aged Children (HBSC)
Survey in Europe and Canada; International Report; WHO Regional Office for Europe: Copenhagen, Denmark, 2020; Volume 1.
2. Griffiths, M.D.; Kuss, D. Adolescent social media addiction (revisited). Educ. Health 2007, 35, 49–52.
3. Boyd, D. It′s Complicated: The Social Lives of Networked Teens; Yale University Press: New Haven, U.S.A., 2014.
4. Obar, J.A.; Wildman, S.S. Social media definition and the governance challenge-an introduction to the special issue. Telecommun.
Policy 2015, 39, 745–750.
5. Ivie, E.J.; Pettitt, A.; Moses, L.J.; Allen, N.B. A meta-analysis of the association between adolescent social media use and depres-
sive symptoms. J. Affect. Disord. 2020, 275, 165–174.
6. Clark, H.; Coll-Seck, A.M.; Banerjee, A.; Peterson, S.; Dalglish, S.L.; Ameratunga, S.; Balabanova, D.; Bhan, M.K.; Bhutta, Z.A.;
Borrazzo, J.; et al. A future for the world′s children? A WHO–UNICEF–Lancet Commission. Lancet 2020, 395, 605–658.
7. Savci, M.; Tekin, A.; Elhai, J.D. Prediction of problematic social media use (PSU) using machine learning approaches. Curr.
Psychol. 2020, doi:10.1007/s12144-020-00794-1.
8. Zendle, D.; Bowden-Jones, H. Is excessive use of social media an addiction? BMJ 2019, 365, doi:10.1136/bmj.l2171.
9. Bányai, F.; Zsila, Á.; Király, O.; Maraz, A.; Elekes, Z.; Griffiths, M.D.; Schou Andreassen, C.; Demetrovics, Z. Problematic social
media use: Results from a large-scale nationally representative adolescent sample. PLoS ONE 2017, 12, e0169839.
10. Van Den Eijnden, R.J.; Lemmens, J.S.; Valkenburg, P.M. The social media disorder scale. Comput. Hum. Behav. 2016, 61, 478–487.
11. Griffiths, M. A ‘components’ model of addiction within a biopsychosocial framework. J. Subst. Use 2005, 10, 191–197.
12. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders: DSM-5; American Psychiatric Associa-
tion: Washington, DC, USA, 2013.
13. Petry, N.M.; Rehbein, F.; Gentile, D.A.; Lemmens, J.S.; Rumpf, H.J.; Mößle, T.; Bischof, G.; Tao, R.; Fung, D.S.; Borges, G.; et al.
An international consensus for assessing internet gaming disorder using the new DSM-5 approach. Addiction 2014, 109, 1399–
1406.
14. Inchley, J.; Currie, D.; Cosma, A.; Samdal, O. Health Behaviour in School-Aged Children (HBSC) Study Protocol: Background, Meth-
odology and Mandatory Items for the 2017/18 Survey; CAHRU: St. Andrews, UK, 2018.
15. Franchina, V.; Vanden Abeele, M.; Van Rooij, A.J.; Lo Coco, G.; De Marez, L. Fear of missing out as a predictor of problematic
social media use and phubbing behavior among Flemish adolescents. Int. J. Environ. Res. Public Health 2018, 15, 2319.
16. Scott, H.; Woods, H.C. Fear of missing out and sleep: Cognitive behavioural factors in adolescents′ night time social media use.
J. Adolesc. 2018, 68, 61–65.
17. Worsley, J.D.; McIntyre, J.C.; Bentall, R.P.; Corcoran, R. Childhood maltreatment and problematic social media use: The role of
attachment and depression. Psychiatry Res. 2018, 267, 88–93.
18. Boer, M.; van den Eijnden, R.J.; Boniel-Nissim, M.; Wong, S.L.; Inchley, J.C.; Badura, P.; Craig, W.M.; Gobina, I.; Kleszczewska,
D.; Klanšček, H.J.; et al. Adolescents′ intense and problematic social media use and their well-Being in 29 countries. J. Adolesc.
Health 2020, 66, S89–S99.
19. Wartberg, L.; Kriston, L.; Thomasius, R. Internet gaming disorder and problematic social media use in a representative sample
of German adolescents: Prevalence estimates, comorbid depressive symptoms and related psychosocial aspects. Comput. Hum.
Behav. 2020, 103, 31–36.
20. Lee, E.W.; Ho, S.S.; Lwin, M.O. Explicating problematic social network sites use: A review of concepts, theoretical frameworks,
and future directions for communication theorizing. New Media Soc. 2017, 19, 308–326.
21. Buda, G.; Lukoseviciute, J.; Salciunaite, L.; Smigelskas, K. Possible effects of social media use on adolescent health behaviors
and perceptions. Psychol. Rep. 2020, doi:10.1177/0033294120922481.
22. Kelly, Y.; Zilanawala, A.; Booker, C.; Sacker, A. Social media use and adolescent mental health: Findings from the UK Millen-
nium Cohort Study. EClinicalMedicine 2018, 6, 59–68.
23. Su, W.; Han, X.; Yu, H.; Wu, Y.; Potenza, M. Do men become addicted to internet gaming and women to social media? A meta-
analysis examining gender-related differences in specific internet addiction. Comput. Hum. Behav. 2020, 113, 106480.
24. Valkenburg, P.M.; Koutamanis, M.; Vossen, H.G. The concurrent and longitudinal relationships between adolescents′ use of
social network sites and their social self-esteem. Comput. Hum. Behav. 2017, 76, 35–41.
25. Hussain, Z.; Griffiths, M.D. Problematic social networking site use and comorbid psychiatric disorders: A systematic review of
recent large-scale studies. Front. Psychiatry 2018, 9, 686.
26. Uhls, Y.T.; Ellison, N.B.; Subrahmanyam, K. Benefits and costs of social media in adolescence. Pediatrics 2017, 140, S67–S70.
27. Vannucci, A.; McCauley Ohannessian, C. Social Media Use Subgroups Differentially Predict Psychosocial Well-Being During
Early Adolescence. J. Youth Adolesc. 2019, 48, 1469–1493.
28. Andreassen, C. Online social network site addiction: A comprehensive review. Curr. Addict. Rep. 2015, 2, 175–184.

Int. J. Environ. Res. Public Health 2021, 18, 1885 11 of 11
29. Best, P.; Manktelow, R.; Taylorm, B. Online communication, social media and adolescent wellbeing: A systematic narrative
review. Child Youth Serv. Rev. 2014, 41, 27–36.
30. McCrae, N.; Gettings, S.; Purssell, E. Social media and depressive symptoms in childhood and adolescence: A systematic review.
Adolesc. Res. Rev. 2017, 2, 315–330.
31. Raudsepp, L.; Kais, K. Longitudinal associations between problematic social media use and depressive symptoms in adolescent
girls. Prev. Med. Rep. 2019, 15, 100925.
32. Woods, H.C.; Scott, H. #Sleepyteens: Social media use in adolescence is associated with poor sleep quality, anxiety, depression
and low self-esteem. J. Adolesc. 2016, 51, 41–49.
33. Wilksch, S.; O’Shea, A.; Ho, P.; Byrne, S.; Wade, T. The relationship between social media use and disordered eating in young
adolescents. J. Eat. Disord. 2019, 53, 96–106.
34. Kircaburun, K.; Griffiths, M.D.; Billieux, J. Childhood Emotional Maltreatment and Problematic Social Media Use Among Ad-
olescents: The Mediating Role of Body Image Dissatisfaction. Int. J Ment. Health Addict. 2020, 18, 1536–1547.
35. Shafi, R.M.; Romanowicz, M.; Croarkin, P.E. #SwitchedOn: A call for assessing social media use of adolescents. Lancet Psychiatry
2018, 5, e27.
36. Vannucci, A.; Simpson, E.G.; Gagnon, S.; Ohannessian, C.M. Social media use and risky behaviors in adolescents: A meta-
analysis. J. Adolesc. 2020, 79, 258–274.
37. Uncapher, M.R.; Lin, L.; Rosen, L.D.; Kirkorian, H.L.; Baron, N.S.; Bailey, K.; Cantor, J.; Strayer, D.L.; Parsons, T.D.; Wagner,
A.D. Media Multitasking and Cognitive, Psychological, Neural, and Learning Differences. Pediatrics 2017, 140, S62–S66.
38. Boer, M.; Stevens, G.W.J.M.; Koning, I.; Finkenauer, C.; Van den Eijnden, R.J.J.M. Validation of the Social Media Disorder (SMD)
Scale in Dutch Adolescents: Findings from a Large-Scale Nationally Representative Sample; Working Paper;: 2021.
39. Torsheim, T.; Cavallo, F.; Levin, K.A.; Schnohr, C.; Mazur, J.; Niclasen, B.; Currie, C. FAS Development Study Group. Psycho-
metric validation of the revised family affluence scale: A latent variable approach. Child Indic. Res. 2016, 9, 771–784.
40. Inchley, J.; Currie, D.; Young, T.; Samdal, O.; Torsheim, T.; Augustson, L.; Mathison, F.; Aleman-Diaz, A.; Molcho, M.; Weber,
M.; et al. Growing Up Unequal: Health Behaviour in School-Aged Children (HBSC) Study: International Report from the 2013/2014
Survey, Health Policy for Children and Adolescents, No. 7; WHO Regional Office for Europe: Copenhagen, Denmark, 2016.
41. Brown, B.; Mounts, N.; Lamborn, S.; Steinberg, L. Parenting practices and peer group affiliation in adolescence. Child Dev. 1993,
64, 467–482.
42. Paakkari, O.; Torppa, M.; Kannas, L.; Paakkari, L. Subjective health literacy: Development of a brief instrument for school-aged
children. Scand. J. Public Health 2016, 44, 751–757.
43. Paakkari, O.; Torppa, M.; Boberova, Z.; Välimaa, R.; Maier, G.; Mazur, J.; Kannas, L.; Paakkari, L. The cross-national measure-
ment invariance of the health literacy for school-aged children (HLSAC) instrument. Eur. J. Public Health 2019, 29, 432–436.
44. Paakkari, O.; Torppa, M.; Villberg, J.; Kanna, L.; Paakkari, L. Subjective health literacy among school-aged children. Health Educ.
2018, 118, 182–196.
45. Ravens-Sieberer, U.; Erhart, M.; Torsheim, T.; Hetland, J.; Freeman, J.; Danielson, M.; Thomas, C. HBSC Positive Health Group.
An international scoring system for self-reported health complaints in adolescents. Eur. J. Public Health 2008, 18, 294–299.
46. Prochaska, J.J.; Sallis, J.F.; Long, B. A physical activity screening measure for use with adolescents in primary care. Arch. Pediatr.
Adolesc. Med. 2001, 155, 554–559.
47. Kaplan, G.A.; Camacho, T. Perceived health and mortality: A nine-year follow-up of the human population laboratory cohort.
Am. J. Epidemiol. 1983, 117, 292–304.
48. European Commission. Digital Economy and Society Index (DESI) 2020 Finland. Available online: https://ec.europa.eu/digital-
single-market/en/scoreboard/finland (accessed on 12 November 2020).
49. Boursier, V.; Gioia, F.; Griffiths, M.D. Do selfie-expectancies and social appearance anxiety predict adolescents’ problematic
social media use? Comput. Hum. Behav. 2020, 110, 106395.
50. Haidt, J.; Allen, N. Scrutinizing the effects of digital technology on mental health. Nature 2020, 578, 226–227.
51. Yang, C.C. Instagram use, loneliness, and social comparison orientation: Interact and browse on social media, but don′t com-
pare. Cyberpsychol. Behav. Soc. Netw. 2016, 19, 703–708.
52. Viner, R.M.; Gireesh, A.; Stiglic, N.; Hudson, L.D.; Goddings, A.L.; Ward, J.L.; Nicholls, D.E. Roles of cyberbullying, sleep, and
physical activity in mediating the effects of social media use on mental health and wellbeing among young people in England:
A secondary analysis of longitudinal data. Lancet Child Adolesc. Health 2019, 3, 685–696.
53. Hunt, M.G.; Marx, R.; Lipson, C.; Young, J. No more FOMO: Limiting social media decreases loneliness and depression. J. Soc.
Clin. Psychol. 2018, 37, 751–776.
54. Heffer, T.; Good, M.; Daly, O.; MacDonell, E.; Willoughby, T. The longitudinal association between social-media use and de-
pressive symptoms among adolescents and young adults: An empirical reply to Twenge et al. (2018). Clin. Psychol. Sci. 2019, 7,
462–470.
55. Kosola, S. Lasten ja nuorten netti-ja peliriippuvuus: Pitääkö olla huolissaan? Suom Lääkäril 2020, 75, 324–329.
56. Hormes, J.M.; Kearns, B.; Timko, C.A. Craving Facebook? Behavioral addiction to online social networking and its association
with emotion regulation deficits. Addiction 2014, 109, 2079–2088.