*2.3. Statistical Analysis*

We evaluated the associations of resilience, perceived stress, and depression with IGD using an independent *t*-test. The associations between gender, lower resilience, and IGD were evaluated through Chi-squared analysis. A hierarchical logistic regression model was used to evaluate associations of resilience, perceived stress, and depression with IGD. Finally, we evaluated the associations of resilience, perceived stress, and depression with CGI score among participants with IGD. A logistic regression evaluated the association between items of R14 and IGD with gender, age, and educational level controlled. A value of *p* < 0.05 was considered significant for all analyses, which were performed using the SPSS package (SPSS Inc., Chicago, IL, USA).

A total of 87 participants in the IGD group and 87 in the control group were enrolled in the study after informed consent was obtained. This study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital (KMUH-IRB-990380).

#### **3. Results**

A total of 87 participants from the IGD group and 87 participants from the control group were enrolled in this study. No significant differences were exhibited in gender, age, or education level between the IGD and control groups, as revealed in Table 1. The IGD group scored lower on resilience, scored higher on perceived stress, and received a higher CGI score than the control group. When participants were further classified into the adequate and low resilience groups, participants with low resilience were more likely to be diagnosed with IGD. Furthermore, participants with low resilience exhibited higher perceived stress and depression, as shown in Table 2.


**Table 1.** Associations of age, educational level, gender, perceived stress, depression, Clinical Global Impression score, and resilience with Internet gaming disorder (IGD).

\*\*: <0.01; \*\*\*: <0.001. Resilience: score on 14-Item Resilience Scale; participants scoring 64 or less were classified into a low resilience group. Perceived stress: score on Perceived Stress Scale-10. Depression: score on Center for Epidemiological Studies' Depression Scale. CGI: Clinical Global Impressions scale.


**Table 2.** The difference in perceived stress and depression between individuals with low resilience and those with adequate resilience.

\*: <0.05; \*\*\*: <0.001. Resilience: score on 14-Item Resilience Scale; participants scoring 64 or less were classified into a low resilience group. Perceived stress: score on Perceived Stress Scale-10. Depression: score on Center for Epidemiological Studies' Depression Scale. a: *p* = 0.07.

#### *3.1. Associations Among Resilience, Perceived Stress, Depression, and IGD*

The hierarchical regression analysis in Table 3 demonstrates that low resilience was positively associated with IGD in Model 1 with gender, age, and educational level controlled. Participants with low resilience had a higher odds ratio (OR, 3.46; 95% confidence interval (CI) = 1.72–6.97) for IGD.

Perceived stress was significantly positively associated with IGD in Model 2. As the Wald χ<sup>2</sup> of perceived stress was higher than that of resilience, perceived stress was more associated with IGD than resilience was (Model 2 in Table 3). This result suggests perceived stress was another important factor to IGD. Further, the resilience was significantly associated with IGD with perceived stress controlled for. It suggested that low resilience had an independent association with IGD from perceived stress.

Depression was significantly positively associated with IGD in Model 3 (OR = 1.1, 95% CI = 1.04–1.17). Resilience and perceived stress exhibited no significant associations with IGD when depression was controlled for. This result demonstrated the mediating effect of depression in the associations of resilience and perceived stress with IGD. It also demonstrated that among resilience, stress, and depression, depression was the factor most associated with IGD.


**Table 3.** Hierarchical logistic regression model of Internet gaming disorder for resilience, perceived stress, and depression controlling for gender, age, and education level.

\*: <0.05; \*\*: <0.01. Resilience: score on 14-Item Resilience Scale; participants scoring 64 or less were classified into a low resilience group. Perceived stress: score on Perceived Stress Scale-10. Depression: score on Center for Epidemiological Studies' Depression Scale.

#### *3.2. The Di*ff*erence in Perceived Stress and Depression between Individuals with Low Resilience and Adequate Resilience Among IGD Group*

The *t*-test in Table 2 demonstrates that individuals with low resilience had higher depression among the IGD group (*t* = 0.023; *p* = 0.03). They also had a trend to have higher percieved stress (*t* = 1.85; *p* = 0.07); however, this did not reach significance.

#### *3.3. Association of Resilience Characteristics with IGD*

We firstly evaluated the difference in each item of R14 between IGD and the control group. It demonstrated that the IGD group had a lower score in ability to copy (*t* = 2.15; *p* = 0.03), acceptance (*t* = 2.63; *p* = 0.01), drive (*t* = 2.81; *p* = 0.01), discipline (*t* = 6.94; *p* < 0.001), interest/engagement (*t* = 2.88; *p* = 0.00), self-efficacy (*t* = 2.29; *p* = 0.02), dependable (*t* = 2.59; *p* = 0.01), meaning (*t* = 3.57; *p* < 0.001), and resourcefulness (*t* = 2.55; *p* = 0.01). Then, we regressed IGD on these significant items in R14 with gender, age, and educational level controlled. The logistic regression in Table 4 demonstrates that "I am self-disciplined" was the only resilience characteristic associated with IGD.


**Table 4.** Logistic regression model of Internet gaming disorder for items of 14-Item Resilience Scale with controlling for gender, age, and education level.

\*\*\*: <0.001. Items of 14-Item Resilience Scale: the items were significantly associated with IGD in *t*-test evaluation.

#### **4. Discussion**

#### *4.1. Perceived Stress of Individuals with IGD*

This study demonstrated that participants with IGD experienced higher perceived stress. This result corresponds with those of other studies [14,23]. Stress is conceptualized as a risk factor contributing to progressive long-term changes in the brain and a subsequent drug-prone state characterized by craving and increased risk of relapse [34]. Accordingly, our results might support the possible role of perceived stress in developing IGD addiction. IGD may also have negative consequences. Ko et al. [35] reported that 46.5%, 25.4%,19.7%, and 4.2% of participants with IGD experienced impaired social interaction, professional failures, impaired family relationships, and failures on examinations related to career opportunities, respectively. Such negative consequences may stress adults with IGD and explain the increased perceived stress levels in our study. Further prospective studies to investigate the addiction process may illuminate the causal relationship between stress and IGD.

#### *4.2. Associations among Resilience, Perceived Stress, Depression, and IGD*

As in the studies of Wu et al. [36] and Canale et al. [23], the interview in the present study demonstrated an association between resilience and IGD. Furthermore, we demonstrated that those with low resilience have a 3.46 times odds ratio of being diagnosed with IGD. This suggests that low resilience plays a risk role in online gaming addiction. In the R14, resilience is represented as a stable personal resource and a positive personality trait, such as discipline, that can contribute to personal competence, self-acceptance, and life satisfaction [37]. These characteristics may facilitate stress coping and attenuate stress-induced depression [21]. The lower perceived stress and depression among the adequate resilience group demonstrated in our results might support such a claim. Therefore, the characteristics of resilience can be promoted to prevent IGD and to attenuate perceived stress and depression among the general population.

Two studies have evaluated the buffering effect of resilience on the association between stress and IGD. Wu et al. [36] did not observe significant protective effects of resilience against the effects of stress on IGD. However, Canale et al. [23] identified a moderating role of resilience and demonstrated that higher perceived stress and lower resilience were associated with increased gaming time. The difference in stress measurement methods might have contributed to the inconsistent results. However, the assessment of IGD was performed using questionnaires in the two aforementioned studies. Our study, which used similar measurements to Canale's [23], demonstrated that individuals with IGD had lower resilience and higher perceived stress. Our regression analysis also demonstrated

that perceived stress had a higher association with IGD than resilience. Although resilience plays a role in the development of IGD, perceived stress may be another influential factor and should be a target of intervention to prevent IGD.

Individuals with IGD could experience the psychosocial distress results from excessive gaming [35]. However, individuals with IGD usually cope with stress through gaming [15], exhibiting escapist gaming. This might aggravate, not resolve, their psychosocial problems. In the current study, the within-group analysis demonstrated IGD individuals with low reliance had higher, but not significantly, perceived stress than those with adequate resilience. This suggests a limited buffering effect of resilience on stress among individual with IGD relative to that in the general population. This might suggest further intervention aside of promoting resilience should be provided to copy the stress associated with IGD. Therefore, alternative stress coping strategies, such as problem-focus coping, exercise, or mindfulness [38] should be provided to replace escapist gaming for IGD individuals under stress.

Depression is one of the most reported factors associated with IGD [39,40]. With depression controlled, perceived stress and resilience were not significantly associated with IGD. This suggests that depression is the most proximal factor contributing to IGD. According to Baron and Kenny [41], depression may play a mediating role in the association between resilience, perceived stress and IGD. Depression could be the outcome of IGD [26], as IGD leads to psychosocial stress. Without adequate resilience to buffer dysphoric mood and psychosocial stress, individuals of IGD can use gaming as an escape from depression. However, excessive gaming without limitation might lead to further psychosocial problems and result in a vicious cycle. Further, these results also suggest that depression as a mood status has a stronger association with IGD than perceived stress and resilience. This study demonstrated that individuals with IGD had lower resilience and higher depression. Thus, depression could also be the outcome of low resilience under high stress which then contributes to IGD. The causalrelationship or bi-direction interaction among resilience, stress, depression, and IGD cannot be confirmed in this cross-sectional study. Nevertheless, comorbid depression should be well assessed and effective intervention—such as antidepressants, cognitive behavior therapy [42], exercise, or promoting resilience—for depression should be provided for individuals with IGD, particularly those with lower resilience or higher perceived stress. Although the depression could be the outcome of IGD [26], gaming could be used to temporarily escape from dysphoric moods for those with limited alternative coping strategies. Thus, depression should be well intervened before aggressive intervention for gaming behavior of individuals with IGD.

#### *4.3. Resilience Characteristic Most Associated with IGD*

Discipline was the only resilience characteristic associated with IGD when other items were controlled. Individuals with IGD had lower discipline than controls did. Self-regulation was reported to mediate the association between impulsivity and pathological video gaming among youth [43]. Individuals with lower discipline might have had difficulty in controlling their excessive gaming, leading to a risk of IGD. Conversely, a loss of control in gaming with negative consequences might represent a lifestyle with lower discipline. The causal relationship between discipline and IGD cannot be confirmed with a cross-sectional study. However, our study suggests it is important to promote discipline to prevent risk of IGD among the general population and to prevent the deterioration of daily life functioning among individuals with IGD. Future prospective studies regarding the predictive ability of discipline for IGD or the treatment effect of promoting discipline in IGD may clarify this causal relationship.

#### **5. Limitations**

This study had limitations that should be considered when interpreting the findings. First, IGD was evaluated only through a diagnostic interview with participants. Additional information from family members or partners could have increased the validity of the diagnosis. Second, the cross-sectional

research design could not confirm causal relationships between resilience, perceived stress or escapist online gaming behaviors and IGD. Lastly, we excluded those take psychotropic medication from this study to prevent its effect on depression and perceived stress. However, it could limit the generation of this study's results to populations with psychotropic medication.
