Internet gaming disorder in adolescents is linked to delay discounting but not probability discounting

Highlights

• IGD adolescents' DD behaviors reflect an inability to delay gratification.

• IGD adolescents' PD behaviors reflect a normal risk-taking tendency.

• Impulsivity might act as a behavioral marker of adolescents with IGD.

Abstract

Internet gaming disorder (IGD) is a behavioral addiction that is gradually becoming a public health problem. We aimed to identify the behavioral characteristics of adolescents with IGD to promote early diagnosis and intervention. As impulsivity and risk taking are the most important traits of addiction, we used several representative methods, including a delay discounting task, a probability discounting task and the Barratt Impulsiveness Scale-11 (BIS-11), to measure the impulsivity and risk-taking tendency of adolescents with IGD and a comparison group. Consistent with previous findings on problematic Internet use (PIU), our results indicated that adolescents with IGD had a larger degree of delay discounting (i.e., more impulsive decision making), regardless of the outcome amount and valence, and higher scores on all three subscales of the BIS-11 than the comparison group. However, there was no intergroup difference in the probability discounting task. In general, the current research posits that impulsive decision making and personality traits, but not risk taking, might co-occur with IGD in adolescents.

Introduction

Internet gaming disorder (IGD) is defined as the persistent and recurrent use of the Internet to engage in games, often with other players. It leads to clinically significant impairments (American Psychiatric Association, 2013) such as insomnia (Lam, 2014), intensified social anxiety (Lo, Wang, & Fang, 2005), elevated levels of depression (Romer, Bagdasarov, & More, 2013), and even suicidal behaviors (Kaess et al., 2014, Zhang et al., 2016). IGD is highly prevalent worldwide, particularly among adolescents (Petry, Rehbein, Ko, & O'Brien, 2015). As IGD is increasingly recognized as a global public health problem, the fifth revision of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) included it as a tentative disorder to foster more research for a better understanding of IGD (American Psychiatric Association, 2013).

Adolescence is a critical period of rapid brain development and behavioral maturation, and adolescents are vulnerable to various types of addictive behaviors (Casey, Jones, & Hare, 2008). Studies have repeatedly shown that most substance users and problem/pathological gamblers begin consuming drugs or gambling during adolescence (Andersen and Teicher, 2009, Giovino, 1999, O'Loughlin et al., 2009, Yip et al., 2011), and the acute and long-term effects of drug use can cause more damage for adolescents than for adults (Meier et al., 2012, Thoma et al., 2011). Moreover, the intensity of alcohol use and gambling in adolescence is correlated with the severity of alcohol and gambling problems in adulthood (Burge et al., 2004, McCambridge et al., 2011). In terms of IGD, adolescents should receive special attention as the prevalence rates of IGD are much higher among adolescents than among adults (Festl et al., 2013, Rehbein et al., 2015). In some studies, the rates of IGD among adolescents exceed 8% (King et al., 2013, Mentzoni et al., 2011). To promote the prevention of and early intervention for IGD, the first step is understanding its etiology by identifying the behavioral traits of adolescents with IGD.

Impulsivity and risk taking are the most important behavioral traits associated with addiction (Everitt and Robbins, 2016, Kreek et al., 2005, Ryan et al., 2013). Impulsivity is defined as “a predisposition toward rapid, unplanned reactions to internal or external stimuli with diminished regard to the negative consequences” (Potenza & de Wit, 2010). It is a multidimensional construct that comprises the impulsive personality trait, impulsive decision making and impulsive motor responses (M. R. Mitchell and Potenza, 2014, Weafer et al., 2014). Impulsivity is regarded as the endophenotype of addiction because it is involved in the initiation, maintenance, and relapse of drug use (Everitt and Robbins, 2016, Jentsch et al., 2014). In parallel, risk taking is the inclination to engage in behaviors that are performed under uncertainty regardless of the possible negative consequences (Kreek et al., 2005). Many studies have shown that risk taking is linked to various addictive behaviors (Fernie et al., 2010, Lejuez et al., 2005, MacPherson et al., 2010, Spurrier and Blaszczynski, 2014). Risk preference is the core feature of pathological gambling. Overall, almost every type of addictive behavior is associated with impulsivity, risk taking or both, and these two traits persistently characterize the behaviors of adolescents (Casey et al., 2008a, Steinberg, 2008). Therefore, for adolescents with IGD, the following question arises: Is excessive online gaming associated with elevated impulsivity or a higher risk-taking tendency?

Some studies have used multiple methods to investigate whether adolescents and adults with IGD exhibit more impulsiveness and a higher risk-taking tendency than those without IGD. First, regarding impulsivity, adolescents with IGD scored higher on the Barratt Impulsiveness Scale (BIS-11), suggesting that they have a heightened impulsive personality trait (Arain et al., 2013; X. Du et al., 2016, Feng et al., 2013, Qi et al., 2015). In contrast, no distinction was found between adolescents with IGD and those without IGD in a Go/No-Go task as the two groups presented similar degrees of motor impulsivity (Arain et al., 2013). Similar to adolescents with IGD, adults with IGD scored higher on both the BIS-11 and the Arnett Inventory of Sensation Seeking (AISS) than adults without IGD (Choi et al., 2014, Ko et al., 2015, Mehroof and Griffiths, 2010). In laboratory tasks, two studies found that adults with IGD presented no deficit in motor response inhibition on the Go/No-Go task (Chen et al., 2015, Ko et al., 2014). It can thus be concluded that compared with those without IGD, adolescents and adults with IGD present elevated degrees of impulsive personality trait and similar degrees of motor impulsivity. However, impulsivity is a multidimensional construct, and the aforementioned study that used the Go/No-Go task measured only the subjects’ motor impulsivity (Arain et al., 2013). It remains unclear whether adolescents with IGD are more likely to make impulsive choices in decision-making tasks.

In addition to assessing impulsivity, different laboratory tasks have been implemented to investigate risk taking among adolescents and adults with IGD. Adults with IGD were more willing than those without IGD to take risks in both the Cups Task and the Game of Dice Task (Pawlikowski and Brand, 2011, Yao et al., 2015). Regarding risk taking among adolescents with IGD, a study using the Balloon Analog Risk Task (BART) reached a somewhat ambiguous conclusion (Qi et al., 2015). In that study, although adolescents with IGD made more total pumps than controls, there was no significant intergroup difference in the two indexes used as the primary measures of risk-taking tendency (i.e., the average adjusted pumps and the number of total pops) (Lejuez et al., 2002). Consequently, additional suitable paradigms should be utilized to clarify the associations between IGD and impulsiveness and risk taking among adolescents.

Delay discounting is the most commonly used measure of impulsive decision making (de Wit, 2009). It refers to the devaluation of a future reward (gain) as a function of the delay before its receipt (Ainslie, 1975, Rachlin et al., 1991). In a typical delay discounting task, subjects are instructed to make a series of binary choices between smaller, faster gains and larger, slower gains to calculate their unique delay discounting rate, which represents their preference for the smaller, faster gains and is akin to the ability to delay gratification (MacKillop et al., 2011). The higher a person's impulsivity level is, the larger his delay discounting rate will be. Delay discounting is considered a behavioral marker for addiction (Bickel, Koffarnus, Moody, & Wilson, 2014). Research has repeatedly demonstrated that addictive behaviors are associated with higher delay discounting rates in adults and adolescents (MacKillop et al., 2011). Several longitudinal studies have found that the delay discounting rate can predict both the initial foray into drug use and future abstinence (Audrain-McGovern et al., 2009, Mueller et al., 2009). In parallel, probability discounting is a phenomenon whereby the subjective value of an uncertain reward declines as the odds against receiving it increase (Green and Myerson, 2004, Rachlin et al., 1991). Though similar to delay discounting in terms of definition and experimental procedure, probability discounting is utilized to measure risk taking. In probability discounting tasks, subjects are presented with choices between smaller, certain gains and larger, probabilistic gains. Four studies using a probability discounting task consistently reported that gamblers had lower probability discounting rates than controls, suggesting that gamblers have an elevated risk-taking propensity and revealing the robust validity of the probability discounting task for detecting risk preference (Holt et al., 2003, Ligneul et al., 2013, Madden et al., 2009, Miedl et al., 2012).

Although delay discounting and probability discounting of gains have been widely examined in addiction research, most studies have focused only on the gain condition and have neglected decision-making processes that involve losses. The valence of the outcome has significant impacts on choices. People discount losses to a smaller degree than they do gains in both types of discounting tasks; this phenomenon is referred to as the “sign effect” (or valence effect) (Estle, Green, Myerson, & Holt, 2006). The delay discounting and probability discounting of losses play an important role in daily addictive behaviors because the decision to use drugs or to play games has future, possibly negative, consequences. A few experimental studies have found that cigarette smokers present larger delay discounting rates than nonsmokers for both money and health losses, suggesting their ignorance of the future in a loss condition (Baker et al., 2003, Johnson et al., 2007, Odum et al., 2002). Only one study revealed that smokers did not present a higher risk tendency than non-smokers in a probability discounting task of losses (Ohmura, Takahashi, & Kitamura, 2005). The amount of the outcome also affects decision making in both delay discounting and probability discounting tasks. People are more impulsive and more willing to take risks for small outcomes than for large outcomes, a phenomenon called the “magnitude effect” (or amount effect) (Estle et al., 2006). Heightened impulsivity and risk-taking tendency have been observed in people with addictive behaviors in discounting tasks using wide range of outcome amounts (Holt et al., 2003, MacKillop et al., 2011).

To our knowledge, no research has systematically investigated the impulsiveness and risk preferences of adolescents with IGD for both gains and losses. The current study compared adolescents with IGD with a comparison group using the monetary delay discounting task and the probability discounting task in both gain and loss conditions and with both large and small amounts. Previous studies revealed that adults with problematic Internet use (PIU) had higher delay discounting rates than adults without PIU and that there was no group difference for the probability discounting task (Q. Li et al., 2016, Saville et al., 2010). Considering that adolescents with IGD might share multiple traits with adults with PIU, two hypotheses were tested in the current study. First, we hypothesized that adolescents with IGD may be less sensitive to delayed outcomes and may show greater delay discounting than adolescents in a comparison group. Second, we hypothesized that adolescents with IGD may have an intact tendency to take risks and may demonstrate probability discounting similar to that of the comparison group.