Internet is one the most important means of communication in the modern society and is deemed as a vital feature in daily lives of most people. Internet usage has increased over the years1 and is required for many daily activities. According to the World Internet Usage Statistics2, the number of internet users worldwide until 30 September 2020 has reached almost 4.9 billion. There are 28.7 million internet users in Malaysia. And 64% of them were between 20 and 30 years of age in 20163. A study conducted in a private Malaysia medical school, showed that the average total daily internet use is 6.56±3.73 h4.
Internet addiction is defined as a psychological addiction that is risky especially for young people, which can lead to health and relationship problems and a lack of time management5. In addition to serving as a necessity in daily life, the excessive use of the internet may also cause many problems. According to Sari and Aydin6, the internet is used for various purposes and is especially widely used by university students. Fast and easy internet access at the university, the need to complete assignments, obtain course information and conduct online research can make university students internet users who were not otherwise using the internet. However, high dependency on the internet and resulting internet addiction can have a negative impact on the well-being of individuals and society, especially among young adults. This situation has the potential to influence the academic performance and social interaction of university students7.
As university students are more likely to have psychosocial and economic problems8, they are at risk of developing internet addiction and eating disorders9. The study of Çelik et al.9 showed a positive correlation between internet addiction and eating disorders. In Malaysia, the prevalence of anorexia nervosa and bulimia nervosa is estimated to be 1% and 3% of the population respectively10. Eating disorders, especially among adolescents and young adults, have become a global concern. Additionally, increased obesity among adolescents and young adults in most countries of the world can be a contributing factor to the development of eating disorders among this age group11. According to the study by Bozkurt et al.12 on young adults in Turkey, young adults with obesity have higher rates of internet addiction. A significant and positive correlation has also been found between internet addiction and body mass index (BMI) among healthy young adults. Individuals with internet addiction often have an irregular diet, snacking between meals and drinking unhealthy drinks, which in turn lead to obesity13.
In general, studies on the relationship between internet addiction and the risk of eating disorders among university students in Malaysia have yet to be conducted. This study was therefore conducted to assess the level of internet addiction and the risk of eating disorders. This study also aimed to identify the relationship between internet addiction, the risk of eating disorders and body mass index (BMI) among UKM students.
MATERIALS AND METHODS
Ethical approval: This study is approved by the Research and Innovation Secretariat Committee, Chancellor Tuanku Muhriz Hospital (reference UKM.PPI.800-1/1/5-JEP- 2019-514).
Subjects and study location: This cross-sectional study was conducted at UKM, Bangi. This study involved undergraduate students (male and female) between 19-27 years of age. The subjects were selected using convenient sampling method from different majors or residential colleges. The sample size was determined using the method described by Daniel14. The selected subjects were healthy without any chronic illness. This study was conducted from September to December 2019. The subjects were briefed on the purpose of the study and the questionnaire prior to the study. The participation in the study was voluntarily.
Sociodemographic and internet usage pattern information: A set of questionnaires was used to collect data regarding the socio-demographic information such as age, year of study and gender.
Anthropometric measurements: All anthropometric measurements were performed in accordance with the standard procedure. Height was measured with the subjects standing upright without wearing shoes using a SECA Body Metre 208 (SECA, Germany) to the closest 0.1 cm. The TANITA Digital Lithium Scale (HD319; TANITA Corporation, Tokyo, Japan) was used to measure weight to the nearest 0.1 kg. The subjects were measured while wearing light clothing and being barefoot. Body weight and height were required to determine the BMI and were calculated based on weight (kg) height1 (m)2. BMI is an indicator of nutritional status that is used in adults; the categories were based on WHO15 criteria and were defined as follows: underweight: BMI <18.5 kg m2; normal: BMI 18.5-24.9 kg m2; pre-obese: BMI 25.0-29.9 kg m2; class one obese: BMI 30.0-34.9 kg m2; class two obese: BMI 35.0-39.9 kg m2 and class three obese: BMI > 40.0 kg m2.
Questionnaires: The questionnaire consisted of two main parts: the Internet Addiction Test (IAT) and Eating Attitude Test (EAT 26).
Internet addiction test (IAT): As shown in Table 1, the problematic internet use are evaluated using IAT16, namely, salience, excessive use, neglect of work, anticipation, self-control and neglect of social life. IAT consists of 20 questions using a 5-point Likert type scale. Based on Table 2, the subjects were asked to rate each of the 20 items on a 5-point Likert scale, from 0 (Not applicable) to 4 (Often). Total scores ranging from 0-30 points were considered to reflect a normal level of internet usage; 31-49 points indicate the presence of a mild level of internet addiction; 50-79 points reflect the presence of a moderate level of internet addiction; and scores of 80-100 indicate a severe addiction of the internet16 (Table 3).
Eating attitude test (EAT-26): The Eating Attitude Test (EAT-26)17 was used to determine if the subject was at risk of eating disorders. The EAT-26 consists of 26 questions related to the subject’s attitudes towards eating. The score is calculated by summing the values from the subject’s responses. The included questions examine the risk of eating disorders, the tendency towards bulimia and the perception of body image. Table 4 shows the scores for items 1-25 and item 26 of the frequency degree ranging from always to never. The final score was then divided into two categories as in Table 5. Subjects with a score of 20 or more were considered to be at risk of eating disorders while scores less than 20 showed a lower risk18.
Data analysis: Collected data were analysed using Statistical Package of Social Sciences (SPSS) version 21.0 software (Inc, Chicago, IL, USA). Descriptive tests including mean, standard deviation and percentage were used for the socio-demographic data, anthropometric measurement, IAT and physical activity levels. The independent t-test was used to determine the differences in the mean scores of anthropometric measurements, IAT and EAT-26 by gender. Associations between male and female subjects with BMI categories, IAT level and risk of eating disorder category were evaluated using the chi-square test. Pearson correlation test was used to determine the relationship between IAT, EAT-26 and BMI. For all data analyses, significant levels were set at p<0.05.
RESULTS AND DISCUSSION
Subject profile: A total of 351 undergraduate students from various majors at UKM, Bangi were involved in this study. The subjects involved consisted of 175 male and 176 female students between 19-27 years of age with a mean age of 21.3±1.5 years for male subjects and 21.1-1.4 years for female subjects. Furthermore, most of the subjects involved in this study were the students of 3rd year (42.2%), followed by 2nd year (22.5%), 1st year (20.2%) and 4th year (15.1%). For marital status, 99.4% of the subjects were single and only 0.6% were married.
Internet usage pattern: Internet usage patterns were assessed to determine the level of internet use, time and frequency that subjects surfed the internet, as shown in Table 6. The prevalence of internet usage for learning by the subjects was 100.0%. This result is indeed expected because as university students, the usage of internet is mainly for learning. Furthermore, the number of hours of internet usage shows that only a small number of subjects (3.4%) use the internet less than two hours in a day. However, more than half of the subjects (51.3%) used the internet for more than five hours a day, with a slightly higher prevalence of male subjects (52.0%) than female subjects (50.6%). Therefore, there was no significant difference in the number of hours of internet use in a day between male and female subjects (p>0.05). The results of this study are supported by Kapahi et al.18 who reported that male subjects (59.0%) used the internet more than five hours a day compared to female subjects (57.0%) and there was no significant difference between the two groups.
The frequency of male subjects used the internet for four to five times a week was 19.4% whereas that of female subjects was 5.7%. Furthermore, most of the subjects of this study used the internet more than five times a week (83.5%). These results show that female subjects have a higher percentage (91.5%) of internet usage frequency more than five times a week compared to male subjects (75.4%) which may be because the female subjects are less socialized than that of the male subjects, therefore the frequency of internet use of female subject is much higher than that of the male subjects. Thus, there was a significant difference in the frequency of internet use in a week between male and female subjects (p<0.001). The results of this study contradict with those of Gordon et al.19 who found that male subjects use the internet more often in a week than female subjects because men often use the internet in their free time but women often use the internet for communication purposes only. Overall, 83.5% of the subjects in this study use the internet more than five times a week followed by 12.5% (four to five times a week), 3.1% (three to four times a week) and 0.9% (less than twice a week).
Anthropometric data: Table 7 shows the anthropometric characteristics of the subject by gender. The overall mean weight of the subjects was 62.1±14.7 kg. Comparison of the subjects by gender showed that the mean weight, height and BMI of male subjects was significantly higher than that of female subjects (p<0.001). Table 8 shows the classification of the body weight status by gender of the subjects based on BMI classification according to WHO15. The gender comparison showed that the majority of male and female subjects were classified as having normal weight with a prevalence of 55.4% and 64.2%, respectively. Furthermore, the prevalence of female subjects classified as underweight was much higher (17.0%) than that of the male subjects (6.3%).
However, for the prevalence of overweight or obesity, male subjects was muchhigher (38.3%) compared with female subjects (18.2%). Thus, there was a significant difference between male and female subjects for BMI classification (p<0.001). The findings of this study are consistent with the study by Gopalakrishnan et al.20 on medical students in Malaysia which showed a much higher prevalence of being underweight in female subjects (17.0%) than that of the male subjects (12.2%), while the prevalence of overweight or obese was higher among male subjects (41.2%) compared to female subjects (31.4%). The study by Pengpid and Peltzer21 also showed that the prevalence of overweight among male students was higher (28.8%) compared to female students (22.4%).
Overall, the majority of subjects involved in this study were of normal weight (59.8%) followed by overweight (20.5%) and underweight (11.7%). Overall, the lowest prevalence was in the obese category which is only 8.0%. The study by Yusoff et al.22 on university students in Terengganu showed that 14.7% of the subjects were classified as underweight, 57.8% as normal weight, 16.8% as overweight and 10.6% of the subjects were classified as obese. Meanwhile, compared to the study by Yusof et al.22, the prevalence of overweight subjects in this study was higher by 3.7%, although, the percentage of obese subjects in this study was lower by 2.6%, comparatively. These findings are consistent with the previous study conducted by Gan et al.23 who showed that there was a significant difference in weight status between male and female university students in Malaysia (p<0.05).
Internet addiction test (IAT)
Mean score of internet addiction test (IAT): Internet addiction includes some symptoms of chemical addiction, however, there is no consumption of harmful as in drug and alcohol addiction18. Internet addiction involve changes in daily life, social life, life performance, feelings, emotions and others.
Table 9 shows the total mean scores of internet addiction by gender. The total mean scores of internet addiction for male subjects are 39.6±10.0 and 38.3±10.2 for female subjects. The total mean score of male subjects were slightly higher but non significantly different from female subjects, demonstrating that male and female subjects had the same level of internet addiction (p>0.05).
This finding is supported by Teong and Ang24 who conducted the study on students at four universities in Sabah and Sarawak and showed that the mean scores of male and female subjects did differ significantly because the use of the internet by university students is now increasing among both genders. Due to the advancement of technology, easy access to the internet and the daily use of notification tools through the internet in various forms, such as for education, entertainment, communication and shopping, the internet has become a very important tool for both male and female students25.
Furthermore, these results are also consistent with a previous study conducted by Kittinger et al.26 on university students in the United States which reported that the mean score of internet addiction for male subjects (37.76±9.92) was slightly lower than female subjects (38.60±10.36), although there was no significant difference (p>0.05). However, these findings are inconsistent with Bui et al.27 who conducted a study on young adults in France and found that the mean score of internet addiction for male subjects was higher (35.51±11.83) than female subjects (32.29±11.32) and there were significant differences (p<0.05).
In a study involving university students in Turkey, Chou et al.28 found that male are at risk for internet addiction due to their stereotypical use of sexual content while women may not have such use. This finding is supported by Morahan-Martin and Schumacher29 who reported that the difference between male and female is due to online activities such as gaming, pornography and gambling, which are more frequently associated with male than female.
Classification of subjects by level of internet addiction: The level of internet addiction is categorized into four parts namely normal, mild, moderate and severe16. Table 10 shows the classification of subjects based on the level of internet addiction by gender. Overall, the prevalence of internet addiction in this study was 80.1%, in which 66.7% (male: 66.3%, female: 67.0%) of the subjects involved in this study had a mild level of internet addiction. Meanwhile, 19.9% (male: 18.9%, male: 21.0%) are categorized as having normal internet addiction and as many as 13.4% had a moderate level of internet addiction (male: 14.9%, female: 11.9%). However, none of the subject were categorized as having a severe level of internet addiction.
The study by Rosliza et al.7 on the students of Universiti Putra Malaysia showed that the prevalence of a mild, moderate and severe level of internet addiction was 35.71%, 56.52 and 7.76%, respectively demonstrating that moderate and severe levels of internet addiction in that study were much higher than in the present study. This finding is probably because the subjects of their study had sufficient financial resources, as Mazhari30 reported that subjects with better financial resources, either through educational loan or family contributions; had better opportunities to access the internet, thus the risk of excessive internet usage was higher. The study by Eren et al.31 on students of Erzurum Atatürk University in Turkey showed that the prevalence of internet addiction was 78.5% (mild: 52.6%; moderate: 24.2%; severe 1.7%). The difference in this prevalence may be related to several factors such as the heterogeneity of the sample, the sample size, socio-cultural differences and access to the internet32.
Eating attitude test-26 (EAT-26)
Min skor EAT-26: Assessment of eating disorder risk in this study was determined using the EAT-26 questionnaire. Table 11 shows the total mean scores of the EAT-26 by gender. Comparison between the genders showed that the total mean score for male subjects was higher (12.5±9.3) compared to female subjects (11.3±9.4) but there was no significant difference between them (p>0.05).
The results of this study are consistent with a previous study conducted by Jie and Sivapathy33 on students of UCSI University, Malaysia, who reported that there was no significant difference in the EAT-26 scores between male and female students. This finding is due to the fact that advertisements promote lean bodies, as reported by Swami et al.34, thus thin body is liked by females and muscular body frame is preferred by males.
Subject classification based on eating disorders risk: The classification of subjects based on the risk of eating disorders by gender is shown in Fig. 1. These results indicate that the risk of eating disorders is higher in men (18.9%) than women (14.2%), however, there was no significant difference between the genders (p>0.05). Musaiger et al.35 conducted a study on university students in Kuwait also found that the prevalence of eating disorders is higher (33.6%) in male students than female students (31.8%). This is because male students are highly involved in physical activities and researchers reported that high levels of physical activity are associated with eating disorders36. However, in contrast Ahmed and Uddin37 reported that female university students were more likely to have eating disorders than males with 96 (80.7%) females compared to 23 (19.3%) male experiencing eating disorders. Overall, a total of 16.5% of subjects in this study were at risk of eating disorders. The study by Musaiger et al.35 reported a higher risk of eating disorders (32.7%) compared to the present study.
Higher level of stress in males increased the psychological problems, which in turn leads to eating disorders23. Additionally, male students desire larger muscle mass while female students want to be leaner38. This study is also supported by Ebrahim et al.39, who found that among male students, the desire to increase muscle mass and reduce body fat was associated with the risk of eating disorders. Furthermore, muscular bodies are more attractive among males as their participation in body building has increased and this practice can lead to negative perception related to body image40. Muscle dysmorphia can force male to use steroids which leads to substance abuse and thus leading to an increased risk of eating disorders41,42.
Relationship between IAT and risk of eating disorders: Excessive or inappropriate use of the internet is referred to as internet addiction. As university students are more likely to have psychosocial and economic problems8, they are at risk of internet addiction and eating disorders9. Correlation tests were conducted to determine the relationship between IAT and the risk of eating disorders among the subjects. The strength of the relationship between the variables can be determined based on three levels: namely, weak (r = 0.10-0.29), moderate (r = 0.30-0.49) and strong (r = 0.50-1.00)43.
Table 12 shows the relationship between IAT and EAT-26 scores by gender. The results of this study showed no significant correlation between IAT v and EAT-26 scores for male subjects (p>0.05). However, there was a weakly significant and positive correlation between IAT and eating disorder risk for female subjects (r = 0.187, p<0.05). These findings are consistent with the study by Bui et al.27 on young adults in France who reported that there was no correlation between internet addiction and eating disorder risk for male subjects but there was a correlation for female subjects (r = 0.33, p<0.001).
These findings are supported by Rodgers et al.44 who reported that the gender differences may be due to several factors such as differences in body image perception and risk of eating disorders. Consistent with previous findings32, female students with severe internet addiction obtained higher scores in the EAT-40 subscale compared to those without internet addiction. This is because internet addiction such as spending time on social media sites (Facebook, MSN) and body image perception is found to be an important predictor of eating disorders among women