Adolescents with emotional and behavioural problems are at higher risk to become excessive or addicted Internet users: is this association moderated by gender?

Abstract

Background

The Internet offers many opportunities for adolescents to facilitate their lives. However, its everyday use may lead to excessive behaviour, including addiction. Our aim was to assess the association between emotional and behavioural problems (EBP) and level of internet use, and whether gender moderates this association.

Methods

We used data from a representative sample of 5,433 Slovak adolescents (mean age: 13.51, 48.8% boys) from the Health Behaviour in School-aged Children-study conducted in 2018, collected through online self-report questionnaires. EBP was measured by the Strengths and Difficulties Questionnaire and excessive internet use (EIU) and internet addiction (IA) by the Excessive internet use scale. We analysed using multinomial logistic regression.

Results

More than 25% of the adolescents reported EIU; almost 4% reported being addicted to Internet. EIU was more prevalent in girls, but IA was more prevalent in boys. Adolescents with borderline or increased levels of EBP were more likely to report EIU (odds ratio, OR/95% confidence interval, CI: 1.85/1.60–2.14; 3.16/2.67–3.75, respectively) and IA (OR/95% CI: 2.23/1.57–3.18; 4.89/3.41–7.03, respectively). Adjustment for gender, age or perceived family wealth hardly changed the findings. Moreover, gender did not modify the associations between EBP with EIU.

Conclusion

Adolescents with higher levels of EBP are more likely to become excessive Internet users or Internet addicts. This shows a need of early identification of adolescents with EBP as they seem to be relatively vulnerable to develop EIU or IA.

Introduction

In the last decade Internet became an integral part of the day-to-day existence of adolescents, changing their lives and sometimes causing problems. The everyday presence of Internet changed the traditional ways how adolescents deal with different social issues, such as discrimination, bullying, low level of social competences or problems of psychological nature like depression, anxiety or stress.^1–3 However, the expanding use of Internet to facilitate and improve adolescent lives may lead to addictive behaviour. Consensus lacks on the exact delineation between different levels of internet use (normal, excessive, addictive). Griffith⁴ presents a complex model of Internet addiction with following components: salience, mood modification, tolerance, withdrawal symptoms, interpersonal and intrapersonal conflicts, and relapse. This approach was used also in EU Kids online Survey.⁵ According to Tao et al.,⁶ users need to report preoccupation with Internet, withdrawal symptoms and one more symptom from those mentioned before to be considered as Internet addicts. Durkee et al.⁷ showed that the prevalence of Internet use on the pathological level in 11 European countries is 4.4% and it was higher in boys. However, recent findings of EU Kids Online study reported lower prevalence of excessive Internet use (EIU) that varies between 0% (Italy and Slovakia) and 2.1% (Croatia and Malta).⁸

Emotional and behavioural problems (EBP) occur frequently among adolescents, with prevalence rates between 10–20%,^9,10 and adolescents with EBP have an increased risk of EIU or addiction (IA).¹¹ A reason may be that EBP can cause problems in different life domains, such as academic achievement, social relations, or lead to substance use or violent behaviour.^12,13 In an attempt to deal with these problems, adolescents might start to use Internet excessively.¹⁴ Furthermore, according to the theory of risky and problem behaviours, adolescents who are engaged in some kind of risky or problem behaviours are more likely to engage in other types of problematic behaviours, such as EIU.¹⁵ Research has shown that adolescents with EBP tend to report higher levels of Internet use, to be concrete they tend to spend more time playing online games.^16,17 However, evidence is scarce on the specific levels of Internet use in the adolescents with EBP.

Previous research has suggested that the level of EBP differs by gender which may also affect the association between EBP and EIU. Magai et al.,¹⁸ Harder et al.¹⁹ and Pathak et al.²⁰ showed that girls tend to report a higher level of EBP, especially in the case of internalizing problems. On the other hand, boys tend to report conduct and prosocial problems.²¹ In addition, previous research²² also suggests gender differences in the way adolescents spend their time online. However, evidence is scarce not only on the association between the level of EBP and EIU, but also on whether gender moderates this association. Moreover, this new survey allowed us to include younger children and to explore the level of adolescents’ internet use in more detail, compared with previous studies.^3,14 Thus, the aim of our study was to examine the association between EBP and the level of Internet use, and whether gender moderates this association. We expect that adolescents with a higher level of EBP are more likely to report EIU or IA.

Methods

Sample and procedure

We used data from the Health Behaviour in School-aged Children-study (HBSC study) that was conducted in 2018 in Slovakia. The study sample was created in accordance with the international HBSC protocol with the purpose to reflect the educational system in Slovakia. Our aim was to obtain data on 11-, 13- and 15-years old children that would be representative for Slovak population. Thus, our study sample was stratified regarding region and type of school (primary school, and early middle school). The HBSC study used a two-step sampling procedure. In the first step, the HBSC Data Management Center (Bergen, Norway) randomly selected 140 schools from a list of all elementary schools and 8-years high schools provided by the Centre for Scientific and Technical Information of the Slovak Republic. All selected schools were contacted and asked to participate in the HBSC study. Thirty-one schools were not willing to participate in the study mostly due to participations in various other studies conducted at their schools. In the second step, at least two classes per one grade in each school were randomly selected. Finally, we used data from 8405 adolescents from 11 to 15 years old (RR = 60%; mean age = 13.43; 50.9% boys) that were collected through online self-reported questionnaires. These questionnaires were filled out in the presence of a research assistant. At the beginning, students received a brief introduction to the study and instructions regarding the questionnaires. They had maximum 45 min to fill out the entire questionnaire.

Our study was approved by the Ethics Committee of the Medical Faculty at the P.J. Safarik University in Kosice (16N/2017). The parents were informed about the study via the school administration and they had the possibility to opt out if they disagreed with their child’s participation. About 10% of parents opted out. Participation in the study was fully voluntary and anonymous with no explicit incentives provided for participation. Due to missing responses on the examined variables, we excluded 2972 respondents, making that analyses were performed on a final sample of 5433 adolescents (mean age = 13.51, 48.8% boys). The flow of respondents leading to the final study sample is shown in more detail in figure 1.

Figure 1.

Flow chart showing the flow of participants leading to the final study sample (5433 Slovak adolescents aged 11–15 years, HBSC Study 2018).

Measures

To measure ‘excessive Internet use’, we used the excessive internet use (EIU) scale consisting of five items, which covers the five dimensions of Internet addiction,⁵ such as salience, tolerance, withdrawal symptoms, conflict and relapse.²³ The responses were (i) never or almost never, (ii) not very often, (iii) fairly often, (iv) very often or symptoms and were dichotomized into categories as fairly often and very often vs. not very often, almost never or never. Then, we used the ‘2 + 1 rule’ proposed by Tao et al.⁶ to distinguish three levels internet use (no excessive internet use, excessive internet and internet addicts). Thus, adolescents reporting the combination of salience and withdrawal symptoms with one more symptom from those mentioned before can be categorized as internet addicts. Those respondents who showed 1–3 symptoms (except the combination of symptoms corresponding to IA) can be categorized as excessive internet users. Cronbach’s alpha in our sample was 0.79.

We measured EBP using the Strengths and Difficulties Questionnaire (SDQ), that covers the most important domains of child psychopathology, such as emotional symptoms, conduct problems, hyperactivity/inattention and peer problems.²⁴ From the SDQ, we used the 20 difficulty items. Responses to these items are: (0) not true, (1) somewhat true, (2) certainly true.²⁵ The total difficulties score (TDS) of the SDQ can range from 0 to 40 and according to original three-band categorization the bandings are: normal (0–15), borderline (16–19) and abnormal (20–40). Cronbach’s alpha in our sample for the SDQ–TDS was 0.64.

‘Socioeconomic status was’ measured using perceived family wealth scale that has been used as a subjective measure of SES in various studies before.^26,27 This scale consists of only one question: How well off do you think your family is? Responses were: (1) not at all well off, (2) not so well off, (3) average, (4) quite well off, (5) very well off. A higher score indicated a higher socioeconomic status. Thus, the perceived family wealth and age were used as continuous variables.

Statistical analyses

First, we described the background characteristics of our sample, and the prevalence of EIU and IA for the total sample and stratified by gender and EBP-category. We also explored the differences in EIU regarding the gender and different level of EBP by using Pearson chi-squared test. Next, we assessed the association of gender, age, perceived family wealth and EBP with internet use, crude and mutually adjusted, using multinomial logistic regression. Third, we analysed the degree to which gender modified the association of EBP by adding the interaction of gender with EBP to the regression model. We analysed our data using IBM SPSS statistics 20.0 for Windows.

Results

Description of the sample

Background characteristics of our sample are presented in table 1. Almost 30% of our sample reported EIU and less than 4% of our respondents reported IA (table 1). Moreover, the results of Pearson chi-squared test showed statistically significant differences in EIU regarding the gender and different level of EBP. The prevalence of EIU was higher in girls. In reverse, IA was more prevalent in boys (table 1).

Table 1.

Description of the background characteristics of the sample (5433 Slovak adolescents aged 11–15 years, HBSC Study 2018)

	Total sample	No EIU	EIU	IA	P-value
	N = 5433 (100%)	N = 3762 (69.2%)	N = 1483 (27.3%)	N = 188 (3.5%)
Gender					<0.001
Boys	2653 (48.8%)	1840 (69.4%)	687 (25.9%)	126 (4.7%)
Girls	2780 (51.2%)	1922 (69.1%)	796 (28.6%)	62 (2.2%)
EBP
Normal (0–15)	3382 (62.2%)	2577 (76.2%)	727 (21.5%)	78 (2.3%)	<0.001
Borderline (16–19)	1294 (23.8%)	814 (62.9%)	425 (32.8%)	55 (4.3%)
Abnormal (20–40)	757 (13.9%)	371 (49.0%)	331 (43.7%)	55 (7.3%)
Age (M, SD, range)			13.5 (1.3); (range 11.0–16.0)
Perceived family wealth (M, SD)			3.96 (0.8)

Missing values: EIU 1246 (14.8%), EBP 25 (0.3%), perceived family wealth 1855 (22.1%).

Associations between EBP and the levels of internet use

Adolescents reporting borderline and abnormal levels of EBP were more likely to report EIU and IA than those showing no EBP problems (table 2, crude model). Regarding the results of multinomial regression, boys were more likely to become Internet addicts in comparison to girls. The associations between perceived family wealth and age were associated with statistical significance with internet use regarding both problematic levels (EIU and IA). Next, adding gender, age and perceived family wealth to the model did not change the strength of the association between EBP and level of internet use (table 2, Adjusted model). Finally, gender did not modify the association of EBP with EIU (table 2, bottom row).

Table 2.

The effect of emotional and behavioural problems (EBP), gender, age, perceived family wealth and the interaction of gender × EBP on the level of internet use (5433 Slovak adolescents aged 11–15 years, HBSC Study 2018)

	No EIU	EIU			IA
		Crude model	Adjusted model	Adj., interaction	Crude model	Adjusted model	Adj., interaction
		OR	OR	OR	OR	OR	OR
		(95% CI)	(95% CI)	(95% CI)	(95% CI)	(95% CI)	(95% CI)
Gender
Boys	ref	ref	ref	ref	ref	ref	ref
Girls		1.11 (0.98; 1.25)	0.98 (0.87; 1.11)	0.89 (0.76; 1.06)	0.47 (0.35; 0.64)	0.39 (0.28; 0.53)	0.30 (0.18; 0.52)
Age		1.14 (1.09; 1.19)	1.11 (1.05; 1.16)	1.10 (1.05; 1.60)	1.17 (1.05; 1.31)	1.13 (1.01; 1.27)	1.13 (1.01; 1.27)
Perceived family wealth (1–5)		0.77 (0.72; 0.83)	0.84 (0.78; 1.11)	0.84 (0.78; 0.91)	0.74 (0.62; 0.88)	0.80 (0.67; 0.95)	0.81 (0.67; 0.96)
EBP
Normal	ref	ref	ref	ref	ref	ref	Ref
Borderline		1.85 (1.60; 2.14)	1.79 (1.56; 2.08)	1.64 (1.33; 2.03)	2.23 (1.57; 3.18)	2.34 (1.64; 3.34)	2.12 (1.38; 3.26)
Abnormal		3.16 (2.67; 3.75)	3.02 (2.54; 3.59)	2.57 (1.95; 3.39)	4.89 (3.41; 7.03)	5.46 (3.77; 7.91)	4.39 (2.73; 7.08)
Interaction (EBP × gender)
EBP normal × girls (vs boys)				ref			ref
EBP borderline× girls (vs boys)				1.20 (0.89; 1.59)			1.38 (0.63; 3.04)
EBP abnormal × girls (vs boys)				1.31 (0.92; 1.87)			1.73 (0.79; 3.81)

Ref, reference category; OR, odds ratio; CI, confidence interval; Adjusted, adjusted for the other independent variable.

Discussion

We assessed the associations between EBP and the level of internet use in the Slovak adolescents and the potential moderating effect of gender on this association. We found that adolescents with borderline or abnormal level of EBP were more likely to become excessive Internet users or Internet addicts. Regarding the moderating effect of gender, the tested interaction of the EBP with gender on EIU or IA was not significant.

We found borderline and abnormal levels of EBP to be associated with EIU and IA, confirming previous findings.^16,17,22 Our results can be explained in several ways. First, a study conducted by Paclikova et al.²⁸ showed that adolescents who suffer from a lack of family support or poor family communication tend to report more EBP. Thanks to anonymity that online space offers to the users, adolescents with EBP may consider Internet as a place where they can more easily talk about their problems or find the help and support, they are missing in their everyday life, e.g. help from experts like psychologists or psychiatrists. Nowadays there are several organizations, such as Child line of trust, Blue Line or Teen Line that offer online counselling to the adolescents or adults that suffer from some problems. Regarding distance and costs, this kind of help is more accessible and it allows adolescents to stay in anonymity,²⁹ what can lead to an intensifying of the use of this type of service and consequently to EIU or IA.

A second explanation for the association of borderline and abnormal levels of EBP with EIU and IA is that adolescents with EBP can consider Internet as a way to compensate the failure they face in their real lives. According to the Cognitive-Behavioural Model of Pathological Internet use the development of IA is often motivated by the cognitive biases Internet users have about them in online space, such as: ‘Offline I am nothing, but in the online world I am really worth something’.³⁰ Internet, thanks to its functions and different online activities, can be seen by its users as a suitable tool to reduce their tension or stress,³¹ to facilitate building relationships they are missing in their offline life.³² Moreover, in the online space adolescents have the possibility to escape from their self and low self-control, and manage their self-presentation.³³ In addition, the cyberspace allows adolescents to stay in anonymity and nobody needs to know about their problems. All these benefits of Internet may lead to an intense use that can next convert in IA.

Regarding gender, we found out that while the prevalence of EIU was higher in girls, the boys were more likely to become Internet addicts. This finding supports previous research conducted in this topic^22,34 pointing out the key role of type of preferred online activities among users. Boys tend to spend plenty of time by engaging in online games, often motivated by their need for escapism, need for competition or social contact,³⁵ what can next lead to the development of IA. They can see playing online games as an easy way to build, maintain or expand their social ties as it allows them to be in contact not only with people they know in person, but also with the people they have met online.³⁶ Furthermore, boys are more likely to participate in offline risky behaviours, such as smoking, substance use or aggressive behaviour.³⁷ However, concerning the role of gender in the association between EBP and EIU, our results did not provide evidence that gender moderates these associations. This contradicts previous studies showing that girls with emotional problems, such as depressive symptoms or subjective unhappiness, were at higher risk of problematic level of Internet use than boys with similar difficulties.³⁸ This can be explained as follows. First, the lack of an association with clinical EBP in our study may be due to smaller numbers for clinical EBP levels, which limits the power to detect an interaction. Second, in our study we focused on EBP in general and we did not take into consideration specific psychological problems that could be potential risk factor of EIU or IA, such as emotional problems, hyperactivity or conduct problems [=39]. Therefore, this topic needs a further study.

Strengths and limitations

The major strength of our study is the large and national representative sample of adolescents from 11 to 15 years. Next, to collect our data we used validated measures that were used in various reports before.^5,25,39 Moreover, our study is one of the first in the European context that tried to distinguish adolescents who are excessive or addictive users of Internet.⁶ This study has also some limitations that should be mentioned as well. The first one is its cross-sectional design that does not allow us to explore the causal pathways between used variables. A second limitation of our study can be the use of self-reported data on problem behaviour as adolescents might find it difficult to talk openly about the level of Internet use, however the method of data collection may reduce this effect.

Implications

Our findings on the association of EBP with EIU or IA emphasize the importance of early identification of adolescents with EBP as they seem to be more vulnerable for several adverse outcomes including EIU or IA. Moreover, development of interventions that would reinforce the real-life activities of adolescents may be effective method to reduce the time adolescents spent on the Internet.

Second, in our study we measured IA by using the EIU scale; however, the question we used, needs to replicate to confirm the value of this scale as tool to identify the adolescents who are more likely to become excessive or addictive Internet users. This step can facilitate the development of prevention strategies and lead to earlier initiation of treatment strategies for IA.

Finally, we explored EIU and IA in general; however, a deeper analysis of specific activities that adolescents do online may provide better understanding of the process of development of IA, as previous research showed that activities such as online gaming, use of social networks or online gambling may lead to addictive behaviour.⁴⁰ A longitudinal study is needed to explore the causal pathway between EBP and EIU.

Conclusion

Adolescents reporting borderline or abnormal level of EBP are at higher risk of EIU or IA. Our findings on the association between EBP and EIU imply that adolescents with EBP are at risk of problematic internet use indeed. This requires particular attention and could be accommodated, e.g. in online versions of intervention programmes for adolescents with EBP as they might use internet to gain the support they are lacking in offline life.

Funding

This work was supported by the Slovak Research and Development Agency under the Contract nos APVV-18-0070, APVV-21-0079, APVV-22-0078 and the Scientific Grant Agency of the Ministry of Education, Science, Research and Sport of the Slovak Republic and the Slovak Academy of Sciences, VEGA reg. no. 1/0177/20.

Conflicts of interest: None declared.

Data availability

The raw data supporting the conclusions of this article will be made available by the corresponding author on reasonable request.

Key points.

• Evidence shows various aspects of internet addiction, but consensus still lacks on the exact delineation between Internet addiction and its earlier stages.

• The internet offers a comfortable environment for adolescents experiencing difficulties in their life, but its expanding use to improve their life can lead to addictive behaviour.

• We found that adolescents reporting higher level of emotional and behavioural problems are at higher risk of excessive Internet use or Internet addiction.

• Boys are more likely to become Internet addicts than girls.

• Preventive strategies should focus on increasing the awareness of benefits and risks of the internet among the adolescents experiencing emotional and behavioural problems, as they seem to be the vulnerable group.