The diagnostic status of problematic internet and social media use in formal classification systems is, stated directly, minimal. Neither the DSM-5-TR nor ICD-11 recognizes a disorder of internet use broadly construed. Internet Gaming Disorder appears in DSM-5 Section III as a condition requiring further study — not as a recognized diagnosis, and explicitly limited to gaming rather than general internet use. ICD-11 formally recognized Gaming Disorder as a condition within the chapter on addictive disorders, effective 2022 — again, gaming-specific and not applicable to social media or broader internet use. For clinicians working with presentations centered on social media, compulsive news consumption, online pornography (absent primary CSBD features), or general internet use dysregulation, there is no recognized diagnostic home in either major classification system. The practical coding solution in most jurisdictions is a residual impulse-control or behavioral disorder code accompanied by clinical narrative, analogous to the situation described for CSBD prior to ICD-11 recognition.
The research literature attempting to operationalize “internet addiction” or “problematic internet use” as clinical constructs is substantial in volume but limited in coherence. A 2014 systematic review identified over twenty different assessment instruments in use across the published literature, with minimal standardization of construct operationalization, threshold criteria, or reference populations. Prevalence estimates vary between 0.8% and 26.7% across studies — a range so wide as to indicate that the studies are not measuring the same phenomenon. This measurement heterogeneity is not merely a technical problem; it reflects genuine conceptual disagreement about what the construct is, who it affects, and what severity threshold distinguishes disorder from heavy use.
A specific and underappreciated diagnostic challenge is what might be termed the generality problem: the internet is not an activity but a medium. Pathological engagement with specific internet-mediated activities — gaming, pornography use, gambling — is clinically distinguishable from pathological engagement with the internet as a delivery mechanism for those activities. The analogy would be attempting to diagnose “television disorder” rather than recognizing that the television can be the delivery mechanism for problematic engagement with sports betting, pornography, or other specific behavioral patterns. The clinical utility of the category depends on whether “problematic internet use” identifies a coherent population with shared mechanisms and treatment needs, or whether it aggregates heterogeneous presentations that happen to share a delivery technology.
Several researchers — notably Starcevic and Billieux — have argued persuasively that the most productive clinical approach treats internet-mediated behavioral problems as specific-activity disorders (gaming disorder, compulsive pornography use, compulsive social media use) rather than as expressions of a unitary internet use disorder. This specificity argument is clinically compelling: it generates more targeted formulations, maps more cleanly onto specific mechanisms, and aligns with the direction in which formal diagnostic systems have moved. The practical consequence is that the clinician’s first task is to identify what, specifically, the patient is doing online that is causing problems, rather than treating internet use as a monolithic behavioral target.
A substantial proportion of the available research on problematic internet and social media use has been conducted in adolescent populations, partly because adolescents are among the heaviest users and partly because concerns about social media’s effects on adolescent wellbeing have generated significant research funding. This concentration of research in adolescent populations creates a generalizability problem for adult clinical practice: developmental considerations that may be central to adolescent presentations — identity formation, peer orientation, the normative salience of social comparison — are not necessarily operative in adult presentations, and the mechanisms documented in adolescent research should not be assumed to apply uniformly across the lifespan. The large-scale studies examining social media use and adolescent mental health — including work by Twenge and colleagues suggesting population-level increases in adolescent depression associated with smartphone and social media adoption — are population-level associations whose application to individual clinical assessment is not straightforward.
In the absence of recognized diagnostic criteria, clinicians should document problematic internet and social media use presentations using the most specific available code combined with clinical narrative. In ICD-10 contexts, F63.89 (Other Specified Impulse-Control Disorder) or F63.9 (Impulse-Control Disorder, Unspecified) are the most commonly used residual codes. Clinical narrative should specify the specific behavioral pattern, functional impairment domains, duration, and evidence of failed self-regulatory attempts. The specificity argument — that treating problematic social media use as a discrete clinical target produces more useful formulations than aggregating it under “internet use disorder” — has practical implications for assessment. Clinicians should map the full behavioral topography of internet use before formulating: which platforms, which activities, which functions, which time patterns, and which consequences. This mapping informs both formulation and intervention specificity. The wide prevalence range in the published literature (0.8%–26.7%) should be treated as a signal of construct invalidity rather than true population variability. When screening instruments are used in clinical practice, their results should be treated as prompts for fuller clinical assessment, not as diagnostic thresholds. A high score on a problematic internet use scale indicates that further assessment is warranted; it does not establish a diagnosis.