The association between harmful alcohol use and Internet addiction among college students: Comparison of personality
Cheng-Chung Chen, MD, PhD, Department of Psychiatry, Kaohsiung Medical University Hospital, 100 Tzyou 1st Rd. Kaohsiung City 807, Taiwan. Email:email@example.com
Aims: This study aimed to (i) evaluate the association between Internet addiction and harmful alcohol use, and (ii) evaluate the associated personality characteristics of Internet addiction as well as harmful alcohol use.
Methods: A total of 2453 college students were invited to complete the Chen Internet Addiction Scale, Behavior Inhibition System and Behavior Approach System Scale(BIS/BAS scale), and the Alcohol Use Disorders Identification Test from May 2005 to May 2006.
Results: The results demonstrated Internet addiction was associated with harmful alcohol use among college students. College students with Internet addiction had higher scores on the BIS and BAS fun-seeking subscales. However, college students with harmful alcohol use had higher scores on the BAS drive and fun-seeking subscales, and lower scores on the BIS subscale.
Conclusions: Internet addiction is associated with harmful alcohol use. Furthermore, fun seeking was the shared characteristic of these two problem behaviors and might contribute to the association. However, further studies are necessary to evaluate the underlying mechanisms accounting for the association between Internet addiction and harmful alcohol use.
FOR COLLEGE STUDENTS, Internet use has become one of the most popular recreational and academic activities because of its convenience and availability. Previous reports demonstrate that 8–13% of undergraduates are addicted to the Internet, which has impaired individual psychological well-being, peer and family interactions, and academic performance.1 In universities in Taiwan, free unlimited Internet service is usually provided to freshman college students (who are living apart from their families for the first time). Accordingly, the change of environment makes Internet addiction a very important issue for the mental health of college students.
Until now, there has been no conclusive definition for Internet addiction. Ko et al.2 have developed diagnostic criteria of Internet addiction in adolescents (DC-IA-A) based on systemic diagnostic interview. In that study, Internet addicts and non-addicts were classified according to both the global clinical impression of psychiatrists, and the score on the Chen Internet Addiction Scale (higher than 75th percentile). A total of 13 candidate diagnostic criteria derived from diagnostic criteria of substance use disorder and impulse control disorder in DSM-IV-TR,3 plus diagnostic criteria for Internet addiction in previous studies and clinical experiences were selected to test their diagnostic accuracy in 486 adolescents. Three criteria were excluded because of lower diagnostic accuracy (<75%). The criteria for functional impairment were defined as criteria B. Other criteria, including preoccupation, uncontrolled impulse, use more than intended, tolerance, withdrawal, impairment of control, excessive time and effort devoted to the Internet, and impaired decision-making were defined as criteria A. Then, the best cut-off point with six out of the nine criteria in criteria A was determined based on diagnostic accuracy. It provided a good diagnostic accuracy (95.4%) for adolescent Internet addiction. Next, the diagnostic efficacy of the same criteria was tested in college students with good diagnostic accuracy.4 Thus, the criteria were utilized to define Internet addiction in the presented study.
As the core symptoms of the criteria are nearly identical to the criteria of substance use disorder, Internet addiction has been labeled as ‘behavior addiction’ and it is suggested that it may share the same underlying neurobiological mechanism with substance addiction.5
Furthermore, it has been reported that adolescents with Internet addiction are more likely to have exposure to addictive substances.6 As comorbidity of two disorders may indicate the causal relationship or common cause shared by them,7 evaluation of the comorbidity between Internet addiction and alcohol use disorder (a well-researched addictive disorder) and shared personality factors between them, may shed light on the underlying psychopathology or mechanism of Internet addiction. However, the association between Internet addiction and alcohol use disorder has not yet been evaluated for college students.
Eysenck has suggested that certain personality traits are prominent vulnerability factors for addiction, regardless of the type of addiction.8 According to Gray's Reinforcement Sensitivity Theory, the behavior inhibition system (BIS) responds to punishment and results in behavior withdrawal and arousal, while the behavior activation system responds to reward and results in behavior approach and arousal.9 High BAS-sensitive persons are more prone to engage in approach behavior for stimuli with reward.9 Thus, high level BAS subjects should be more engaged in addictive behavior, including substance and alcohol use.10,11 However, whether high level BAS predicts behavior addiction, like Internet addiction, has not yet been evaluated among college students.
Thus, the aim of the study was to: (i) evaluate the association between Internet addiction and harmful alcohol use; and (ii) evaluate the associated personality characteristics of Internet addiction as well as harmful alcohol use.
Seven colleges in Taiwan were selected for evaluation from May 2005 to May 2006. The selected colleges included two, two and three schools from urban, suburban, and rural areas, respectively. Eight classes were randomly selected from each college. The research assistant explained the goal and the content of the study to all students in the selected classes, and invited them to participate in this investigation. Those students agreeing to participate in the study were arranged to complete the anonymous questionnaire. A total of 2453 students were recruited, although 27 students refused to participate in the evaluation. A total of 434 participants (161 male, 273 female) were omitted because they failed to complete all rating scales. A total of 1992 participants (581 male, 1411 female) who completed all scales entered the final analysis. The mean of their age was 20.45 ± 2.16 years.
Chen Internet Addiction Scale
The Chen Internet Addiction Scale (CIAS) contains 26 items on a 4-point Likert scale. The internal reliability of the scale and the subscales in the original study range from 0.79 to 0.93.12 According to the diagnostic criteria of Internet Addiction for College,4 the 67/68 cut-off point of CIAS has highest diagnostic accuracy (81.5%) and accepted sensitivity (74.7%) and specificity (86.0%). Accordingly, those with CIAS scores of 67 or more were classified as the Internet addiction group.
Behavior inhibition system and behavior approach system scale (BIS/BAS scale)
The BIS/BAS scales were designed to assess individual differences in the sensitivity of the two motivational systems proposed by Gray13 and were utilized to assess Gray's model of personality in this study. The BIS Scale measures the degree to which respondents expect to feel anxiety when confronted with cues for punishment. The BAS Scale includes subscales of reward responsiveness, drive, and fun-seeking, which measure the degree to which rewards lead to positive emotions, a person's tendency to actively pursue appetitive goals, and the tendency to seek out and impulsively engage in potentially rewarding activities, respectively. Cronbach's alphas of the four subscales were 0.66–0.76.9 Cronbach's alphas and eight-week test-retest reliability of the Chinese version of the BIS/BAS scale have been reported to be 0.75–0.85 and 0.44–0.72, respectively.14 They have also been reported to have good criterion and construct validity.14
Alcohol Use Disorders Identification Test
Alcohol Use Disorders Identification Test (AUDIT) was developed by the World Health Organization to identify hazardous or harmful alcohol use.15 It consists of ten questions pertaining to alcohol consumption, alcohol dependence, and alcohol-related problems during the past 12 months. The defined cut-off point of eight has been reported to diagnose harmful alcohol users with 0.96 sensitivity and 0.85 specificity in the Chinese version of AUDIT.16 In this study, scoring 8 or more on AUDIT was defined as the criterion for inclusion in the harmful alcohol user group. Based on definition of AUDIT, harmful alcohol use refers to alcohol consumption that results in consequences to physical and mental health. Some would also consider social consequences among the harmful effects caused by alcohol.15
Depression was also assessed by the Center for Epidemiological Studies' Depression Scale (CES-D) and a score higher than 28 was classified as having depression.17
The association between Internet addiction and harmful alcohol use was evaluated with χ2 analysis. Furthermore, the association of harmful alcohol use and Internet addiction to Internet behavior, depression and gender were also evaluated. Then, the association between harmful alcohol use and Internet addiction was evaluated by logistic regression with control of gender, age, and depression.
In order to compare the association of BAS/BIS characteristics to Internet addiction and harmful alcohol use, two-factor anova for BAS/BIS characteristics concerning Internet addiction and harmful alcohol use were utilized to evaluate the main effect and interaction effect. The estimated margin of scores on BIS/BAS subscales between four groups classified by Internet addiction or not and harmful alcohol use were provided. A P-value less than 0.05 was considered significant for all tests.
A total of 246 (12.3%) and 131 (6.6%) participants were classified as the Internet addiction group and the hazardous alcohol use group, respectively. The χ2 analysis (Table 1) reveals that college students with harmful alcohol use are more likely to have Internet addiction. Both harmful alcohol use and Internet addiction are associated with depression. Logistic regression analysis revealed that the Internet addiction was still associated with harmful alcohol use (WALD χ2 = 6.75; P = 0.009; odds ratio = 1.84; 95% confidence interval = 1.16–2.90) even with control of gender, age and depression. College students with harmful alcohol use were also more likely to be male, and to use the Internet for more than 30 h/week.
Table 1. The association between Internet addiction and harmful alcohol use
|Internet addiction|| || || || || || |
| Yes||32 (13.0)||214||18.90***||–||–||–|
| No||99 (5.7)||1647||–||–||–||–|
|Gender (0)|| || || || || || |
| Male||75 (12.9)||506||53.54***||111 (19.1)||470||34.58***|
| Female||56 (4.0)||1355||–||135 (9.6)||1276||–|
|Depression|| || || || || || |
| Yes||29 (12.0)||212||14.43***||65 (27.0)||176||54.66***|
| No||95 (5.6)||1595||–||173 (10.2)||1517||–|
|Use Internet >30 h/week (6)|| || || || || || |
| Yes||37 (9.6)||350||6.86**||94 (24.3)||293||63.50***|
| No||94 (5.9)||1505||–||151 (9.4)||1448||–|
|Use Internet every day (1)|| || || || || || |
| Yes||102 (7.1)||1331||2.41||215 (15.0)||1218||33.11***|
| No||29 (5.2)||529||–||31 (5.6)||527||–|
Two-factor anova in Table 2 and estimated means in Table 3 revealed that college students with Internet addiction had higher behavior inhibition and that analogues with harmful alcohol use had lower behavior inhibition. Harmful alcohol use was associated with higher BAS drive. Lastly, college students with Internet addiction and harmful alcohol use had higher BAS fun-seeking. The interaction term of Internet addiction and harmful alcohol use was not associated with BAS/BIS characteristics.
Table 2. The association of BIS/BAS characteristics to Internet addiction and harmful alcohol use
|Gender (male)||458.64||61.67*** (1)||50.87||10.11** (1)||34.55||9.04** (1)||1.15||0.33 (1)|
|Age||55.97||7.53** (1)||2.82||0.56 (1)||6.73||1.6 (1)||22.76||6.53* (1)|
|Internet addiction||158.13||21.26*** (1)||15.36||3.05 (1)||2.26||0.59 (1)||45.88||13.16** (1)|
|Harmful alcohol use||30.69||4.13* (1)||0.00||0.00 (1)||14.88||3.89* (1)||19.11||5.48* (1)|
|Interaction†||0.21||0.03 (1)||0.28||0.06 (1)||4.00||1.05 (1)||10.01||0.09 (1)|
|Error||14 770.65||(1986)||9 989.3||(1986)||7 590.65||(1986)||6 922.49||(1986)|
|Total||870 079.00||(1992)||579 868.00||(1992)||248 230.00||(1992)||243 856.00||(1992)|
|R2||0.058|| ||0.009|| ||0.017|| ||0.045|| |
Table 3. The estimated mean margin for the BIS/BAS score among four groups classified by Internet addiction and harmful alcohol use
Previous reports have revealed that Internet addiction is associated with substance use in adults and adolescents.4,18 Our result firstly reports the association between Internet addiction and harmful alcohol use among college students. The association existed with control of gender, age and depression. It indicates that the association could not solely be explained by gender effect or comorbid depression. It also verifies a previously reported hypothesis that addiction-prone phenotypes for substance use would also reflect sensitivity to the alternative behavior reinforcers in an individual's environment.19
Ko et al.6 found that adolescents with Internet addiction or substance exposure are more likely to have higher novelty seeking and lower reward dependence. However, the former had higher harm avoidance and the latter had lower harm avoidance. In line with these results from adolescents, our results revealed that college students with Internet addiction or harmful alcohol use had higher BAS fun-seeking. The former had higher BIS (sensitivity to punishment) and the latter had lower BIS.
In an incentive salience model for addiction,20 substances produced a sensitization of the mesolimbic dopaminergic system, which increased the incentive value of the substance. People with a higher BAS score had a higher sensitivity to rewarding stimuli,21 and a vulnerability to establish the incentive salience to the stimuli. Corresponding to the above theory and previous reports,22 this study demonstrates that higher scores on BAS drive and fun-seeking were associated with harmful alcohol users. The result indicates that responsiveness to reward would make people more sensitive to the rewarding effects of alcohol and increase the risk of harmful alcohol use.
On the other hand, the study reveals that the Internet addiction group had higher BAS fun-seeking scores. It demonstrates that college students with higher BAS fun-seeking scores had higher sensitivity not only to alcohol but also the Internet. Pleasure of control and fluidity of identity online have been reported as seductive properties of the Internet.23 Also, the group interactive design of gaming and online chatting and the predictable achievement in them, might allow college students to satisfy their (Maslow's24 hierarchy of) basic needs, including love and belonging, esteem and self-actualization in the virtual world. These immediate satisfactions might contribute to the reinforcement potential of Internet use. Thus, college students with higher BAS fun-seeking, which contributes to sensitivity to reward, would be prone to approach the Internet. Also, this sensitivity to reward may play a key role in rapid neural sensitization and to the development of strong associations with conditioned online behavior as per Robinson and Berridge's incentive salience hypothesis for substance dependence.20 This effect will contribute to persisting, habitual and uncontrolled Internet behavior. However, this hypothesis for association between BAS fun-seeking and Internet addiction should be examined by further research.
As per previous reports,9 our study reveals that college students with harmful alcohol use had lower scores of BIS. As BIS illustrates the anxiety when confronted with cues for punishment and reaction to conditioned aversive stimuli, college students with lower BIS are prone to imbibe large quantities of alcohol without considering the negative psychological or physiological consequences.
However, college students with Internet addiction had higher scores of BIS. College students with higher sensitivity to reward stimuli and higher BIS might select a perceived lower hazard behavior, for example, the Internet, to reward themselves. Moreover, higher BIS may play two conflicting roles on addiction to Internet. Faceless interaction, anonymity, lack of direct physical harm and being able to enter and leave without limitation in Internet activities allow the user to feel relief from immediate anxiety for their behavior. Even though troubles may occur sometimes, conflict in a chat room or failure in a game for instance, they could escape from chatting immediately or play again to compensate for the loss. For college students with higher anxiety for immediate consequences, the Internet virtual world provides an easier environment than their real-world activities. Alternatively, the anxiety for punishment stimuli ought to make them worry about the long-term negative consequences of Internet addiction. Why did the higher BIS not protect college students from addiction to the Internet?
Firstly, college students with higher BIS are more likely to establish aversive conditioning to immediate negative consequence and reflect their anxiety to such results. However, the negative consequences of Internet addiction are usually delayed for a long time and might not be directly caused by Internet use. Thus, it might be more difficult for the aversive condition to segue into immediate discomfort to alarm the BIS system. Secondly, it has been suggested that sensitivity to punishment plays a minor role compared to BAS fun-seeking in alcohol use behavior among college students.22 When the reinforcing potential of the Internet has established the incentive salience on college students because of their higher BAS fun-seeking, addiction to the Internet could not be stopped by sensitivity to delayed aversive outcome. As a result, Internet addiction becomes established and enhances loss of control even though the user may worry about the long-term consequences. However, this hypothesized mechanism to explain the association between higher BIS and Internet addiction should be examined by further studies.
There are several limitations in this study that should be considered when interpreting its findings. First, a higher proportion of female college students are included than in the general population. Thus, the prevalence rate of Internet addiction would be under-estimated. Second, our investigation relied on self-reported data from adolescents. Third, social restrictions on alcohol use may discourage college students from admitting alcohol use even in anonymous questionnaires. Finally, the cross-sectional research design of the present study could not confirm causal relationships between personality, Internet addiction and harmful alcohol use.
The study demonstrates the association between Internet addiction and harmful alcohol use. It reveals the necessity to screen harmful alcohol use for college students with Internet addiction, and Internet addiction among college students with harmful alcohol use. The results found that college students with both Internet addiction and harmful alcohol use had higher scores on the BAS fun-seeking measure. However, college students with the former had higher BIS and those with the latter had lower BIS scores. The underlying mechanism accounting for the association between Internet addiction and harmful alcohol use should be further evaluated to develop preventive media policies for Internet use of college students.
This study was supported by a grant from Kaohsiung Medical University Hospital (97-KMUH-6R-01) and a grant from Center of Excellence for Environmental Medicine (KMU-EM-97-2.3a).