Internet Addiction and Its Associated Factors among Undergraduate Students in Kathmandu, Nepal

Background Internet has penetrated all processes of life and has become an unavoidable part of people's daily life. This widespread use of the Internet has resulted in significant concerns with regard to problematic Internet behaviours and related conditions. The aim of the study was to find out the prevalence of Internet addiction and its associated factors among undergraduate students in Kathmandu. Materials and Methods We included all together 344 undergraduate students from different colleges affiliated to Pokhara University for this cross-sectional study. We used self-administered questionnaire consisting of the Internet Addiction Test scale to assess the Internet addiction. We coded the data, entered it in Epi-Data 3.1 and transferred to IBM SPSS 25 for analysis. We applied bivariate and multivariate logistic regression analysis to identify factors associated with Internet addiction, and p value <0.05 was considered as statistically significantt. Results The prevalence of Internet addiction was found to be 29.90% (95% CI: 25.0–34.9). In the chi-squared test relationship with parents, parental control over the Internet use, perceived feeling of loneliness, and time spent on internet per day were found to be statistically associated (p < 0.05) with Internet addiction. Conclusion This study revealed nearly one-third of the Internet addiction among undergraduate students. Relationship with parents, parental control over the internet use, perceived loneliness feelings, and time spent on internet per day were significantly associated with Internet addiction along other factors. Therefore, it is important to raise awareness among young generation, parents, and teachers towards risk of Internet addiction.


Introduction
Internet is considered to be the most widely used media in the world, and it varies from other types of media. It has penetrated all process of life [1]. Internet invention has changed the various aspects of individual life such as in the way individuals entertain themselves and interact with each other, with the infnite social networking sites [2]. Te number of Internet users has grown exponentially in the world; the number of active internet users in the world far exceeds 4 billion people [3].
Globally, there are 5.32 billion (67% of global population) mobile users, 5.00 billion people (63% of global population) using the internet and there are 4.65 billion (58.7% of global population) social media users around the world in April 2022. Te annual growth rate of internet user's is 4.1% (i.e., 196 million people). Average duration of the internet use is 6 hours and 57 minutes per day on connected activities. Younger people tend to spend more time online than older generations do, with young women spending the greatest amount of time using the internet [4].
Worldwide, the prevalence of internet addiction has been estimated at 6%, considering that only about 39% of the world population has internet access. Tere seems to be a signifcant variation in rates of internet addiction between countries [5]. Te prevalence of severe problematic Internet use (PIU)/Internet addiction ranged from 0 to 47.4%, whereas the prevalence of Internet overuse/possible Internet addiction ranged from 7.4% to 46.4% among students from Southeast Asia [6]. Te extreme use of internet showed the addictive behaviour of the internet use [7]. Physical impairments in the form of insomnia (26.8%), daytime sleepiness (20%), and eye strain (19%) were reported among users [6]. Internet users in Nepal increased by 822 thousand (+7.7 percent) between 2021 and 2022 [8].
Tis widespread use of the Internet has resulted in the signifcant concerns with regards to problematic Internet behaviours and related conditions [9]. Internet addiction (IA) is the lack of ability to control the Internet use and involvement leading to progressive loss of control. With negative social efects, Internet addicts use the web as a social and communication tool, once they experience higher levels of pleasure and satisfaction when online than in real life [10]. It reduces social interaction to family, friends, and community leading to mental health problems such as depression, anxiety, and other psychological problems [11,12]. Internet addiction symptoms express the users' urge to continue being connected despite the desire to stop, experiencing unpleasant emotions when they do not succeed, sleep disturbance, angry or agitated reaction when forced to disconnect, and loosing track of time while online [13]. People having a high level of anxiety use social media more which tentatively increase the negative emotions and leads to mental consequences [14,15].
Internet is being an unavoidable part of day-to-day life because the usage of the internet has been growing explosively worldwide. With the increase in the dependence of internet, people are gradually getting addicted towards it. Tere are only limited studies conducted in Nepal which primarily focus on adolescents and medical students. Tis study focuses on undergraduate level students as this age group mostly uses internet for diferent purposes. Tey are particularly vulnerable to problematic internet use due to several factors such as the psychological and developmental characteristics of late adolescence/young adulthood, ready access to the internet, and an expectation of computer/internet use. Majority of internet users in Nepal are aged 18-24 years, typical age group of undergraduate students [8]. So, this study helps to identify the prevalence of Internet addiction among undergraduate university students and to reveal factors associated with internet addiction in Kathmandu, Nepal.

Study Design, Period, Setting, and Population.
Tis crosssectional study was conducted among the undergraduate students of Kathmandu district of Nepal and data were collected during 30 June to 15 July of 2022.

Inclusion and Exclusion
Criteria. Undergraduate students above the age of 18 years currently enrolled at any course of Pokhara University inside Kathmandu district were eligible to be included in this study and no exclusion were made in terms of any attributes.

Sample Size Determinations, Sampling Techniques, and
Procedures. Te past study conducted among higher secondary students in Kathmandu had noted the prevalence of internet addiction at 34.35% [16]. Considering this prevalence rate and margin of error at 5%, using Cochran's formula, the initial sample size was estimated at 312 which was optimized to 343 adjusting the 10% nonresponse rate.
Simple random sampling technique was used to choose six colleges from the Kathmandu district using a lottery method where a total of 344 students were enrolled at the time of the study. Te required number of undergraduates from each selected colleges were estimated based on the number of students enrolled in each college and its programs to ensure proper representation of the undergraduate students. Finally, all the estimated number of students from each college and program were approached by enumerating all the students present in the randomly selected class at the time of data collection. When the number of students in class was more than the estimated sample size, the surplus data were also taken. Te ethical approval for research was taken from IRC-CiST prior to the data collection and the approval was taken from concerned colleges to conduct the research. Informed consent was taken from the participants with explanation of aim and nature of the study prior to data collection.

Data Collection Tools and Procedures. Internet Addiction
Test (IAT) is a validated instrument to measure Internet addiction [16]. Te Internet Addiction Test was primarily developed by Dr. Kimberly Young, which is a 20-item 5-point Likert scale ranging from 0 to 5 (0 = less extreme behaviour to 5 = most extreme behaviour) that measures the severity of self-reported compulsive use of the internet [17]. Te sum of the ratings was calculated for the 20-item responses for the total IAT score. Te maximum IAT score is 100 points. IAT scores were categorized as internet users who scored <50 were considered average user and who scored ≥50 were considered internet addicted. Te questionnaire consisted of four sections where the frst section consisted of sociodemographic information, second section consist of behavioural factors related to the internet use, third section was about the perceived psychological status and interpersonal relationship, and the fnal section consisted of Young's 20-item Internet Addiction Test (IAT). Te questionnaire was pretested among 10% of the sample population prior to data collection. Data were collected by the help of a self-administered method. During the process of data collection, the nature of the study was explained in detail to the participants; the details regarding the duration of the study, informed consent, and confdentiality concerns. Questionnaires were distributed to all the students possessing inclusion criteria.

Data Quality Control
Issues. One of the researchers was self-involved for data collection and supervised by the other one. Each day at the end of data collection, every questionnaire was checked, reviewed by the supervisor, and organized for completeness and consistency. Pretesting of the questionnaire was carried out in one of the management colleges afliated to Tribhuvan University in Kathmandu. Pretesting of the instrument was carried out in 10% of the sample. Cronbach's alpha of IAT tools was found to be 0.799 in this study.

Data Processing and Analysis.
Data were entered in Epi-Data (version-3.1) and exported to Statistical Package on Social Sciences (IBM SPSS) version 25. Te Pearson chisquared test and binary logistic regression were applied to access the association between diferent independent variables and dependent variables (internet addiction) at 95% confdence interval and 5% level of signifcance i.e., p value<0.05. Te unadjusted odds ratio (UOR) has also been reported along with the adjusted odds ratio (AOR) for those variables which were signifcant in bivariate analysis. For the multivariate analysis, the variance infation factor (VIF) test was performed to check multicollinearity among the independent variables. Te Hosmer-Lemeshow test (the HL test) for goodness-of-ft and Nagelkerke R square test were also performed for the model.

Results
A total of 400 questionnaires were distributed among the undergraduate students of the selected colleges, of which 344 questionnaires were received covering complete response to all the provided questions. Tus, the response rate of 86% was achieved.

Prevalence of Internet Addiction.
Out of total 344 students who provided their complete response, 164 (47.7%) were found to had a mild level of Internet addiction under the IAT score of (31-49) whereas 96 (27.9%) were found to had a moderate addiction level (score 50-79) and 7(2.0%) were found to be severe addict (score 80-100). Te overall prevalence of possible addict/internet addict based on IAT with a score of ≥50 was at 29.9% (95% CI: 25.0-34.9)

Sociodemographic Characteristics of Respondents.
Te mean age of the students who participated in the study was 20.68 ± 1.863 years while the minimum and maximum age of the participants ranged between 18 and 30 years. Tere was nearly equal participation of the female and male undergraduate students as 53.8% female and 46.2% were male. Te majority were unmarried and reported to have a nuclear family (Table 1).

Internet Use-Related Characteristics of Respondents.
With regards to the internet use pattern, most (208 i.e., 60.5%) of the participants reported have started using internet between 11 and 15 years of age. Likewise, more than half 201(58.4%) noted that they spent more than 5 hours per day over the internet. More than two-third (72.1%) of the students reported that their internet use is mostly for educational, recreational, and entertainment purposes, i.e., 256 (74.4%). Similarly, half of the participants (50.9%) reported that their social media as one of the major platforms for their internet use while 39 (11.3%) reported to be engaged in online jobs (Table 2).

Perceived Psychological Status and Interpersonal
Relationship of Respondents. It was noted that nearly a quarter of the participants perceived themselves to be stressed and had a lower self-esteem. Almost two-third 226 (65.7%) reported having a wonderful relation with their parents while only few reported to have wonderful relation with their teachers and peers. More than half of the students i.e., 193 (56.1%) experienced loneliness sometimes in past one week (Table 3).

Factors Associated with Internet Addiction.
In bivariate analysis, no statistically signifcant relationship was observed between student's sociodemographic characteristics such as age, gender, marital status, parental education, and economic status of the family and their internet addiction status (Table 4). In regards to the internet use pattern, a statistically signifcant relationship was observed between time spent by the students on daily basis over internet and their internet addiction status. Similarly, use of internet for social networking and media was also found to be associated with internet addiction at p < 0.05 (Table 5).
In context of perceived psychological factors and interpersonal relationship of the students, nature of students, relationship with their parents, parental control over their internet use, and the perceived level of loneliness were found to have a statistically signifcant relationship with their internet addiction status at p < 0.05 (Table 6).
For multivariate analysis, the variance infation factor (VIF) test among the independent variables was performed where the highest reported VIF was 1.894 suggesting no issue of multicollinearity. Students who spent fve hours or more on the internet were found to be almost twice more at odds (AOR: 1.780, 95% CI: 1.052-3.012) of experience internet addiction as compared to those who used internet for less than 5 hours a day. Similarly, higher odds of internet addiction were observed among the students who had good (AOR: 1.957, 95% CI: 1.022-3.745) and satisfactory (AOR: 2.832, 95% CI: 1.354-5.614) relation with their parents as compared to those who had wonderful relation. Likewise, students who reported their parents have higher control over their internet use were found to be thrice more at odds (AOR: 3.643, 95% CI: 1.687-7.863) of internet addiction as compared to students whose parents do not control their internet use. Students experiencing loneliness most of the time were also found to be have three-folds increase in their odds (AOR: 3.105, 95% CI: 1.264-7.629) of internet addiction in comparison to those who reported not experiencing loneliness (Table 7).

Discussion
It was seen that the prevalence of internet addiction among this study group was 29.90% which was supported by a metaanalysis conducted in prevalence of Internet addiction in medical students in diferent countries, the prevalence of IA was 30.1%, [18]. Furthermore, a study conducted among undergraduate university students in Ethiopia was found to be 29.4% moderate to severe IA [19]. Another study conducted among young adults in Bangladesh revealed prevalence of Internet addiction was 27.1%, [20], and online survey of problematic Internet use (PIU) and its correlates among undergraduate medical students of Nepal found the prevalence of PIU to be 31.9% [21].
A study conducted among medical students in Nepal found low prevalence of Internet addiction than this study where out of 100 students, 21 students were found to be slightly addicted to using the Internet [22]. In a study, among adolescents in a peri-urban setting in Nepal, 21.5% of the participants were identifed with borderline Internet addiction and 13.3% with possible internet addiction [23]. A study among adolescent Turkish students prevalence of IA was 17.7%.
A study conducted among undergraduate students in Nepal revealed 35.4% prevalence of Internet addiction [24] and a study among higher secondary level students in Kathmandu district revealed possible addicts/Internet addicts to be 34.35% [16] which is slightly higher than this study.
Te possible reason for variations in prevalence of Internet addiction across diferent country or even in the study conducted in same country might be because of sample size, sampling procedure, diference in social context and background of the participants, purpose of the Internet use, knowledge about the problems of Internet addiction, or aware about the proper use of Internet.

Journal of Addiction
Te association between time spent on Internet per day and IA was found to be statistically signifcant (p value 0.010). Tis association is supported by various studies such as a study conducted in Nepal [16], Ethiopia [19], Bangladesh [20], and Bengaluru, India [25]. As the dependence on Internet increases, individuals spent more time on Internet. Relationship with parents and IA have signifcant association (p < 0.001) which is also indicated by several studies [16,[26][27][28][29][30]. Internet addiction was found to be higher among those who had satisfactory relationships within family. Self-control was negatively linked to Internet addiction as prevalence of Internet addiction was found to be higher in individuals with poor self-control than individuals with good self-control [30][31][32]. Te loneliness of the participants was found to be signifcantly associated with Internet addiction (p < 0.001) similar with the various studies [6,33,34]. Lonely individuals prefer to increase their communication through social networks to meet their emotional needs that was included in the meta-analysis conducted in Iran [35]. Perceived stress did not show any signifcant association with Internet addiction which is found to be consistent with the study conducted in East Malaysia [36] and the study conducted among Chinese adolescents [37]. However some previous studies reported signifcant association between perceived self-esteem and internet addiction, [38,39] but we did not fnd the association i.e., low self-esteem increased susceptibility to IA. Terefore, more studies are needed to understand the signifcant association of self-esteem and IA.
Majority of the participants used internet for entertainment and refreshment purpose followed by education or to get new information. Use of Internet for social networking purpose showed signifcant association with IA (p value 0.024). Tis fnding is lined with the study conducted in Northern Tanzania [40] and Saudi Arabia [41]. Most commonly used app/website was found to be social media among the participants supported by the study among young adults in Bangladesh [20].
Te restrictive parenting approach was found to be signifcantly associated with IA (p < 0.001) similar to the fnding of the study conducted among adolescents in Hong Kong, the greater the number of rules and the stricter the enforcement of rules concerning the Internet use, the more likely it is that adolescents will become addictive users [29].

Strength and Limitations.
Te target population of this study was undergraduate university students of Nepal, which group is highly vulnerable to Internet addiction and majority of Internet users are also of this group. Tis study included a considerably high number of socio-demographic variables and also the Internet use and behavioural related and family and college related variables.
Te fnding of the study is based on the primary information collected using the standard tool by the active involvement of the researchers. Since this is a cross-sectional study conducted with minimal required sample size, it could    not cover all disciplines and universities hence insufcient to conclude that Internet addiction is high among undergraduate students in Nepal. So, further studies using larger sample size is necessary for fnding the causal associations on Internet addiction among undergraduate students in Nepal.

Implications of the Study.
Problematic use of Internet/ Internet addiction can result in the various health problems and many factors are associated with it. Tis study identifed the purpose of the Internet use, level of Internet addiction, and factors associated with Internet addiction. So, by identifying the factors associated with Internet addiction and strength of association between them, possible intervention action can be taken and will ultimately help to reduce the various health and educational problems, social life, and family relationship related problems arising due to excessive use of Internet.

Conclusion. Prevalence of possible addiction of In-
ternet use was about one-third among undergraduate students. Tere are numbers of factors which play a signifcant role to point out the alarming situation such as time spent on Internet per day, relationship with parents, parental control over Internet use, and loneliness feelings were found to be statistically associated with Internet addiction. Major purpose of Internet use was found to be entertainment and refreshment, and use of internet for social networking showed signifcant association with IA. Most commonly used app/website was found to be social media among the participants. As family relationship and a restrictive parenting approach were found to be signifcantly associated, familybased prevention strategies need to be developed and applied to achieve healthy family interactions through improving parents-child communication and strengthening family functionality rather than directly restricting the Internet use. Further studies should be conducted in the consequences of problematic internet use in mental health.

Policy Recommendations.
Family-based prevention strategies need to be developed and practice to achieve healthy family interactions through improving parents-child communication and strengthening family functionality rather than directly restricting the Internet use. Education programs need to be carried out on regular basis regarding the addictive behaviour and coping strategies also involving education and sensitization of the students as well as teachers.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Ethical Approval
Ethical approval for research was taken from IRC-CiST (REF NO IRC 190/078/079) prior to the data collection, and the approval was also taken from concerned colleges to conduct the research.

Consent
Informed consent was taken from the participants with explanation of aim and nature of the study. Participants were not forced to participate in the research and were allowed to withdraw from the study.

Conflicts of Interest
Te authors declare that they have no conficts of interest regarding the publication of this article.

Authors' Contributions
SA generated concept, developed proposal, collected data, and prepared preliminary manuscript. LA contributed to data entry, cleaning, and result preparation and supported manuscript preparation. SK supported for proposal fnalization, supervision on data collection and quality control, and manuscript writing. SP coordinated with college prior to data collection, analysed data, and wrote the manuscript. MK fnalized the proposal, analysed data, and prepared and reviewed manuscript. All authors have read and fnalized the manuscript.