The current study examined associations between PIU and the presence of mental health disorders, impairment and health behaviors in a large, pediatric and transdiagnostic, self-referred community sample. Across self- and informant-based reporting methods, we found significant PIU associations with depressive disorders, ADHD, combined presentation and ASD. Importantly, PIU independently predicted impairment, even after accounting for the contributions of comorbid DSM-5 disorders. Consistent with prior studies,[36, 50, 53, 76] the presence of PIU was associated with increased sleep disturbances. Surprisingly, when compared with previous work,[72–74] PIU was not consistently negatively correlated with physical fitness, though some results trended towards significance. This may be because some pediatric participants have more structured exercise and play times dictated by caregivers (e.g. Physical Education class, compulsory after-school sports, etc.) and thus the deleterious effects of PIU on physical fitness are prevented. Findings were generally convergent between self and parent reports, though PIU associations with ADHD, combined presentation were subthreshold for parent-report, and those for ASD were not detectable with self report. Overall, findings from the present study suggest that while internet use is not necessarily linked to psychopathology, PIU can be, and may be associated with impairment even after accounting for comorbid psychopathology.
While depressive disorders have been consistently linked to PIU, associations with ADHD have been less commonly found.[39, 45, 51, 53, 87] Caplan has suggested that depressed individuals tend to harbor negative views of their social competency, thus encouraging them to seek social encounters online, as these may be ‘easier’ to achieve since they allow for an increased flexibility in self-presentation through anonymity. Ko et al. have argued that internet use, particularly online gaming, provides necessary immediate reward, rapid response rates, and a plethora of activities to keep boredom at bay, thus making it highly appealing to individuals with ADHD. Additionally, poor self-control and high impulsivity also increase the risk of developing PIU in individuals with ADHD. Finally, in accordance with previous literature [60–62], associations with ASD were found, though solely for parent report, possibly suggesting a limitation in self awareness. The discrepancy in these parent-child reports, as well as the impact of ADHD comorbidity with ASD, are potential avenues for future research to elucidate ASD associations. Our findings have specific implications for the treatment of these disorders and the recognition of PIU, and may provide a pathway to understanding possible mechanisms for the maintenance of these disorders.
We did not confirm findings from earlier studies of an association between PIU with anxiety disorders and symptoms of anxiety.[47, 50, 55, 57] Researchers have theorized that individuals with anxiety, particularly social anxiety, use online relationships to compensate for poor real-life ones in a similar way as posited for individuals suffering from depression.[32, 55] However, many of these previous studies did not account for comorbidity with ADHD, ASD and depression. Therefore, prior associations between PIU and anxiety may have been attributable to comorbid conditions.
In accordance with previous literature,[36, 41, 42, 50, 53, 75, 76] we found an association between PIU and sleep problems, adding to the growing awareness of the impact of inadequate sleep in U.S. adolescents on both mental and physical health.[89, 90] The persistence of this association after adjustment for not only demographic variables, but also ADHD and depression, suggests that internet use may directly influence sleep behavior or vice versa. This suggests that future studies should assess the direct impact of internet use, particularly problematic use, with sleep.
Finally, our findings imply that the effects of PIU on impairment are independent of the effects of comorbid psychopathology on impairment. The potential clinical implications suggest that PIU may not only be problematic, but also may lead to serious impairment, especially for individuals who already suffer from a comorbid mental health disorder. Recent changes to the DSM-5 and ICD-11 have also highlighted the importance of providing more information on negative consequences of specific internet-related activities (i.e. sedentary behavior and sleep deprivation), and how these may contribute to impairment in various domains of life, particularly in samples at risk for psychiatric disorders. Future research should provide more information on the specific negative consequences of different internet-related activities (i.e. sedentary behavior and sleep deprivation), and how these may contribute to impairment in various domains of life, particularly in populations at risk for psychiatric disorders.
There were several limitations to the current work. First, due to the community self-referred nature of the sample, our results may be affected by sampling biases, which can decrease their generalizability to the larger population of youths in the United States. Our sample does have a higher representation of males, likely due to higher frequency of disruptive behaviors, which tend to be among the most common reasons families seek help. Additionally, there is also a greater representation of children from higher SES and Caucasian youth, which we controlled for in our analyses. Importantly, the convergence of our findings with prior studies that have established associations between PIU and psychiatric symptomatology using other sampling designs (i.e. samples not recruited in a clinical setting) [39, 45, 51, 53, 60–62, 87] decreases the likelihood that our results can be substantively explained by the properties of our particular sample. Nonetheless, replication of our findings in more representative samples would ensure their robustness. Second, PIU was only assessed through one questionnaire which, though demonstrated to be reliable, lacks specificity. The questionnaire does not specify what activities “internet use” and “being online” refer to, leaving interpretation to the respondent; it also has no mention of social media or smartphone usage, reflecting the timing of its creation. The Internet Process Addiction Test (IPAT) has been developed to address these issues, though is notably longer and yet to be heavily adopted in the literature. The present work used a general measure of PIU, which is able to capture problematic behavioral patterns associated with internet use, regardless of the specific activities an individual is engaging in. In future work, the acquisition of activity-specific measures would be helpful in discerning differences in patterns of association between mental health disorders and specific forms of internet addiction (e.g., video-gaming, pornography, gambling). Third, while parent-reported PIU was available for children of all ages in the sample (5-21), self-reported questionnaires were only administered to children ages 7 and up. Fourth, the HBN sample is largely composed of individuals affected by one or more mental health or learning disorders, inherently increasing the likelihood of problematic behaviors among participants. As such, the odds ratios calculated for disorders such as ADHD and depression may be an underestimate of what would be obtained in a more traditional, community representative sample. Finally, the sample size of the present work was not large enough to enable stable odds ratio estimates based on typically developing children alone.
Future research should focus on identifying the specific internet-related activities (i.e. social media, gaming, pornography access, gambling, etc.) that are associated with different clinical disorders and comorbidity, as preliminary work has determined that different activities may have different relationships with psychopathology, behaviors, and personality characteristics[56, 57, 87, 92–94]. Most importantly, potential mechanisms through which PIU is associated with impairment and sleep problems after adjustment for psychiatric disorders requires further inquiry. Identification of the directionality of these relationships should be pursued in prospective research in order to generate effective interventions to prevent the negative outcomes associated with PIU.