Prevalence and related factors of internet addiction among undergraduate university students in Ethiopia. A community university-based cross-sectional study

Background Globally, more than three billion people use the internet daily with young people being the most common users. Internet addiction among university students in Ethiopia has not been studied. Objectives The main objective of this study was to assess the prevalence and related factors of internet addiction among Wollo university students in Ethiopia. Methods A community university-based cross-sectional study was conducted among Wollo University students from April 10 to May 10, 2019. A total of six hundred three students were participated in the study using a pilot tested and self-administered structured questionnaire. A multistage cluster sampling was adopted for this study. In the rst stage, by the use of lottery method, two colleges (College of medicine and health sciences, and College of natural sciences, and one school (school of law)) were selected. In the second stage, 18 departments were selected from the colleges and school. The Young’s Internet Addiction Test instrument was used to assess the level of internet addiction. Results Generally, the prevalence of internet addiction (IA) among the current internet users was 85% (n=466) with 55.7%(n=305) mild internet addiction, 27.9% (n=153) moderate internet addiction and 1.5% (n= 8) severe internet addiction. In multivariable logistic regression analysis, spending more time on the internet (adjusted odds ratio (AOR) = 10.13, 95%CI: 1.33-77.00)), having mental distress (AOR = 2.69, 95%CI: 1.02-7.06), playing online games (AOR = 2.40, 95%CI: 1.38-4.18), current khat chewing (AOR = 3.34, 95%CI: 1.14-9.83) and current alcohol use (AOR = 2.32, 95%CI: 1.09-4.92) were positively related to internet addiction. Using the internet for more than twelve months (AOR = 0.42, 95%CI: 0.18-0.96) and using the internet by mobile internet (AOR = 0.40, 95%CI: 0.20-0.83) were negatively related to internet addiction. Conclusions The current study documents a high prevalence of internet addiction among Wollo University students.

analysis, spending more time on the internet (adjusted odds ratio (AOR) = 10.13, 95%CI: 1.33-77.00)), having mental distress (AOR = 2.69, 95%CI: 1.02-7.06), playing online games (AOR = 2.40, 95%CI: 1. 38-4.18), current khat chewing (AOR = 3.34, 95%CI: 1.14-9.83) and current alcohol use (AOR = 2.32, 95%CI: 1.09-4.92) were positively related to internet addiction. Using the internet for more than twelve months (AOR = 0.42, 95%CI: 0.18-0.96) and using the internet by mobile internet (AOR = 0.40, 95%CI: 0.20-0.83) were negatively related to internet addiction. Conclusions The current study documents a high prevalence of internet addiction among Wollo University students. Factors independently associated with internet addiction were using the internet greater than ve hours daily, having mental distress, playing online games, current khat chewing and current alcohol use. So, we advise that higher institutions should establish prevention programs focus on educating students about emotional regulation skills and the harmful effects of substance use to minimize internet addiction.

Background
Globally, more than three billion people use the Internet daily with young people being the most common users (1). In the eld of medicine and healthcare, it helps in the practice of evidence-based medicine, research and learning, access to medical and online databases, handling patients in remote areas, and academic and recreational purposes (2,3). Relaxed access and social networking are two of the several aspects of the Internet development of addictive behaviour (4). Internet addiction in puberty can negatively impact personality information and may negatively affect cognitive functioning, lead to poor academic performance and engagement in hazardous activities, and inculcate poor dietary habits (5). Problematic internet use is also related to anxiety and stress (6). It has been found that paranoid ideation, hostility, anxiety, depression, interpersonal sensitivity, and obsessive-compulsive average scores are higher in people with high internet addiction scores than those without internet addiction (7,8).
College students are especially susceptible to developing a dependence on the Internet, more than most other segments of society. This can be quali ed to numerous factors including the following: Availability   of time; ease of use; unlimited access to the Internet; the psychological and developmental characteristics   of young adulthood; limited or no parental supervision; an expectation of Internet/computer use covertly   if not obviously, as some courses are Internet-dependent, from assignments and projects to link with  peers and mentors; the Internet offering a way of escape from exam anxiety, age, making new friendships online, getting into relationships online, being sexual inactive, failure in academic performance, gender, low self-esteem, anxiety, depression, insomnia, attention de cient disorder and hyperactivity symptoms, smoking, visiting pornographic sites, playing online games and potential addictive personal habits of, drinking alcohol or coffee, and taking drugs, duration of internet use and mental distress (9)(10)(11)(12)(13)(14)(15)(16)(17)(18).
Internet addiction is now becoming a serious mental health problem among Chinese adolescents. The researchers identi ed 10.6-13.6% of Chinese college students as Internet addicts (19,20) .A study conducted among Taiwan college students reported that the prevalence of Internet addiction was 15.3% (17).
The prevalence of problematic internet use (PIU) was greater among university students. For instance, the prevalence was 36.9 to 81% in Malaysian medical students (21,22), 16 (11,13,18), 12% IA to 34.7% (PIU) in Greek University students (12), 1.6% IA in Turkey students (26). Internet addiction among university students in Ethiopia has not been studied. The current study aims to explore the prevalence and related factors of internet addiction among undergraduate university students in Ethiopia.

Methods And Materials
Study area and period The study was done at Wollo University, Dessie campus that is found in South Wollo Zone, Amhara Regional State which is 401 kilometres far from Addis Ababa, Northeastern Ethiopia. It had 5 colleges and 2 schools and a total of 62 departments. The number of regular students in 2018/2019 is 7248; among these 4009 are males and 3239 are females. The study was conducted from April 10 to May 10/ 2019. The sample size was determined using single population proportion formula, taking 50% prevalence of internet addiction with the following assumption: 95% CI, 5% margin of error, 10% nonresponse rate and a design effect of 1.5. So, the nal sample size was 603.
Sampling technique and procedure A multistage cluster sampling technique was adopted for this study. In the rst stage, by the use of lottery method, two colleges (College of medicine and health sciences, and College of natural sciences, and one school (school of law)) were selected. In the second stage, 18 departments (9 from college of medicine and health science, 8 from college of natural science and 1 from school of law) were selected. Students were selected proportionally from the given departments based on the number of students of a particular.

Study design
A community university-based cross-sectional study was carried out to assess the prevalence and related factors of internet addiction among undergraduate students at Wollo University, Amhara Region, Ethiopia. Age of participants ranges from 18 to 30 years.

Study instruments
Self-administered, well-structured and organized English version questionnaire was disseminated to students and data were collected from the individual student. The questionnaire comprised six parts, the rst part consisted of socio-demographic details, the second part consists of Young's Internet Addiction Test (YIAT). The Internet Addiction Test (27) is the most commonly used measure of Internet addiction (28). It includes 20 questions with a scoring of 1-5 for each question and a total maximum score of 100. Based on scoring subjects would be classi ed into normal users (0-30), mild (31-49), moderate (50-79) and severe (80-100) internet addiction groups. Mild addiction, moderate addiction, and severe addiction were considered as having an internet addiction (29)(30)(31). YIAT20 showed that it is more reliable in University students. The Generally Cronbach α in the present study was 0.89, the third part time-related factors, the fourth part reasons for internet use, the fth part psychoactive substance use-related factors and the last part mental health problem-related factors and it was assessed by Kesler10. The Kessler Psychological Distress Scale (K10)(32) is a simple measure of psychological distress. The K10 scale involves 10 questions about emotional states each with a ve-level response scale. The measure can be used as a brief screen to identify levels of distress. Scores will range from 10 to 50. A score under 20 is likely to be well, a score of 20-24 is likely to have a mild mental disorder, a score of 25-29 are likely to have a moderate mental disorder and a score of 30 and over are likely to have a severe mental disorder. Study participants with a score of 20 or more points on the Kesler-10 Likert scale were considered as having mental distress (33).

Data quality control
A structured self-administered questionnaire was developed in English and would be translated to Amharic language and again translated back into English to ensure consistency. Data collectors and supervisors would be trained for two days on the objective of the study, the content of the questionnaire and the data collection procedure. Data would be pilot tested on 5% of the total sample size outside the study area and based on feedback obtained from the pilot test, the necessary modi cation would be done. During the study period, the collected data would be checked continuously daily for completeness by principal investigator and supervisor in the respective departments.

Data processing and analysis
Quantitative data would be cleaned, coded and entered into Epi-data 3.1 and exported to SPSS version 25 for analysis. Descriptive data would be presented by a table, graphs, charts, and means. Multicollinearity was checked by using standard error and there was no correlation between independent variables. The association between independent variables and internet addiction would be made using a binary logistic regression model and all independent variables having p-value ≤ 0.25 would be included in multiple logistic regression models. A P-value less than 0.05 and Adjusted Odds Ratio (AOR) with 95% Con dence Interval (CI) not inclusive of one would be considered as statically signi cant and would be used to determine predictors of internet addiction in the nal model. Hosmer-Lemeshow test was done to check model tness and the model was t.

Results
Socio-demographic characteristics of study participants A total of 603 respondents were participated in this study, of them 548 providing information which makes a response rate of 90.9%. However, fty-ve (n = 55) participants were excluded.
The mean age of the respondents was 21. internet addiction respectively. Nevertheless, the remaining 82 (15%) are free from internet addiction (Fig. 1).
Participants who login permanently had a greater gure of addiction of the internet than those who log in and off occasionally during the day (92.2% versus 83.1%). Those who used the internet for about six months had a greater prevalence of internet addiction than those who used greater than twelve months (91.6% versus 84.1%) ( Table 2).

Factors related to internet addiction
On univariable analysis, internet addiction was associated with using the internet for more than 12 months (OR = 0.49; CI: 0.24-0.96). Those who were using the internet for more than 5 hours per day were more likely to develop internet addiction than those who were using less than 5 hours (OR = 8.87; CI: 1.21-65.25). Mode of internet access was related to internet addiction i.e. those who used mobile internet were 45% lower risks of having internet addiction than those who used data cards (OR = 0.55; 95% CI: 0.28-1.07). Those who were permanently online were most likely to have internet addiction than those who were not (OR = 2.39; 95% CI: 1. 16-4.93 In the nal model, spending more time on the internet, having mental distress, playing online games, current khat chewing and current alcohol use were positively related to internet addiction. Using the internet for more than twelve months and using the internet by mobile internet were negatively related with internet addiction ( Table 2).

Discussions
The current study documents a high prevalence of internet addiction among Wollo University students. In the nal model, factors independently associated with internet addiction were spending more time on the internet, having mental distress, playing online games, current khat chewing and current alcohol use.
The prevalence of internet addiction in the present study was higher than the prevalence of internet  (18). The discrepancy might be due to the cut-off point of YIAT-20, instrument difference, mental health policy, cultural difference like time utilization, the difference in study participants such as in our study the participants were from two colleges and one school, and all participants were internet users, sample size and the time difference between the studies. The study in Malaysian University was conducted among medical students only and focusing on mild internet addiction and moderate internet addiction and not on severe internet addiction.
In our study spending more time on the internet were 10 times more likely to develop internet addiction than those who are spending less time. The nding of this study in line with similar studies done on college students in Taiwan and three medical schools across three countries (17,31). The possible explanation for the association between Internet usage time and Internet addiction is that it might be as much a symptom as it is a cause. However, this study design was cross-sectional and no causal relationship can be clari ed, further studies ought to examine whether Internet usage time is an essential factor for determining Internet addiction. Likewise, students who had mental distress were 2.7 times more likely to develop internet addiction as compared to their counterparts. Study ndings in these areas showed that students who had mental distress were related to higher levels of internet addiction than students who hadn't mental distress (10, 13-15, 18, 26). This could be due to the Khantzian's(37) selfmedication hypothesis, indicating that mentally distressed university students might come to rely on the Internet as a method for coping with their mental distress. Hence, they will devote more and more time on the Internet and headway toward addiction if their mental distress symptoms are not cured (38).
Students who had playing online games were 2.4 times higher to have internet addiction than their counterparts. A similar nding was also reported in Greek University (12,30). Furthermore, students who chewed khat currently were three times most likely to develop internet addiction than students who reported no current khat chewing which is in line with the study nding in Greek University students (12). In this study, students who drank alcohol currently were 2.3 times most likely to have internet addiction as compared with students who didn't drink alcohol.
Other studies reported a similar nding (12). Probable reasons involve either common forerunner factors that encourage liability to the co-occurrence between drug use or the existence of other personal habits/dependencies and internet addiction, and that they represent different conditions along a spectrum of related disorders of addiction. Nevertheless, susceptibility factors relating to internet addiction and substance use disorders persist obscure. This coexistence of internet addiction with other personal habits/dependencies is a similar pattern to that observed in the coexistence of pathological gambling and substance abuse. Though this comparison might be unequal, it is valuable to pull upon it as an example, to pursue aetiological factors (39).
Students who used the internet by mobile internet were 60% of lower risks of having internet addiction as compared to those students who used data cards. This might be due to inadequate nance to use the internet on mobile internet. So, the students may refrain from using the internet through mobile internet.
Students who used the internet for more than 12 months were 52% less likely to have internet addiction than their counterparts. The current nding is not supported by other studies in the world. The alpha in ation from multiple testing. Another limitation is that the analysis did not account for the complex sampling strategy in adjusting the standard errors.

Conclusions
The current study documents a high prevalence of internet addiction among Wollo University students. Factors independently associated with internet addiction were using the internet greater than ve hours daily, having mental distress, playing online games, current khat chewing and current alcohol use. So, we advise that higher institutions should establish prevention programs focus on educating students about emotional regulation skills and the harmful effects of substance use to minimize internet addiction.

Declarations
Ethics approval and consent to participate The study was conducted after getting ethical clearance from Wollo University College of medicine and health science institutional review board. A formal letter of permission was obtained from the student service directorate of Wollo University. The respondents were informed about the aim of the study. Con dentiality and privacy of the respondents were maintained. Written consent was obtained from each participant before administering the questionnaire.

Consent for publication
Not applicable Availability of data and materials The dataset supporting the conclusions of this article is available with the corresponding author and will be made available on reasonable request.

Competing interests
The authors report no con icts of interest in this work Funding Not applicable   Figure 1 Internet addiction by severity among undergraduate university students in Ethiopia, 2019 (n =548) Figure 2 Reasons for internet use among undergraduate university students in Ethiopia, 2019 (n=548)