Global Burden Of Problematic Internet Use: An Umbrella Review and Metanalysis

This umbrella review and meta-analysis of 11 systematic reviews involving over 3 million individuals estimates the global prevalence of problematic internet use behaviors to range from 6% for gaming to 32% for smartphone use, while highlighting substantial methodological heterogeneity and a critical need for higher-quality, geographically diverse research.

The Gist

Imagine the internet as a massive, bustling global city. Most people walk through it, visit shops, chat with friends, and go home. But for some, the city becomes so captivating that they can't leave, get lost in the crowds, or spend all their time in one specific district to the point where it hurts their real life. This paper is like a giant "City Survey" that tried to figure out exactly how many people are getting stuck in this digital city.

Here is what the researchers found, broken down simply:

The Big Picture: A "Review of Reviews"

Instead of going out and asking people on the street themselves, the researchers acted like detectives looking at other detectives' notebooks. They gathered 11 major reports (called "systematic reviews") that had already looked at thousands of smaller studies. They then combined all the data from those smaller studies (involving over 3 million people) to get a clearer, global picture.

Think of it like trying to understand the weather. Instead of checking the sky yourself, you gather every weather report from the last few years, check how reliable the weathermen were, and then calculate the average temperature for the whole planet.

The Findings: How Many People Are "Stuck"?

The researchers looked at four different "districts" of the internet city and found that the number of people getting stuck varies wildly depending on which district you visit:

• The Gaming District (Problematic Gaming): About 6% of people (roughly 1 in 16) are struggling here. This is the most studied area, but it has the lowest "stuck" rate.
• The General Internet District (Problematic Internet Use): About 16% of people (roughly 1 in 6) are having trouble here.
• The Social Media District (Problematic Social Media Use): About 23% of people (roughly 1 in 4) are struggling.
• The Smartphone District (Problematic Smartphone Use): This is the most crowded "stuck" zone, with about 32% of people (roughly 1 in 3) having issues.

The Analogy: Imagine if you asked people if they were addicted to a specific video game, and 6% said yes. But if you asked if they were addicted to their phone in general, suddenly 32% said yes. The "device" itself seems to be the bigger problem than the specific games on it.

The Problem with the Map: Why the Numbers Are Shaky

While these numbers are interesting, the researchers found a huge problem with the maps they were using. They described the quality of the 11 reports they reviewed as "Low" or "Critically Low."

Here is why the map is blurry:

1. Different Rulers: Some studies measured "addiction" with a ruler that was 10 inches long, while others used a ruler that was 20 inches long. One study might say, "If you play 2 hours, you're addicted," while another says, "You need to play 10 hours." Because they used different rules, the numbers don't line up perfectly.
2. Different Crowds: Some studies only looked at university students, others looked at gamers, and others looked at the general public. It's like trying to guess the average height of all humans by only measuring basketball players.
3. Missing Neighborhoods: Most of the data came from Europe and East Asia. There are very few reports from Africa, the Americas, or Southeast Asia. It's like having a weather report for London and Tokyo, but no data for the rest of the world.
4. The "Pandemic" Effect: Many of the studies were done during the COVID-19 pandemic when everyone was stuck at home on their devices. This might have made the numbers look higher than they would be in normal times.

The "Messy" Conclusion

The paper concludes that while a significant chunk of the global population is likely struggling with internet use, we can't trust the exact numbers yet.

The researchers say we need to:

• Standardize the Rulers: Agree on exactly what "problematic use" means so everyone measures it the same way.
• Fill the Gaps: Go look at people in Africa, the Americas, and other under-researched areas.
• Check the Tools: Make sure the questions we ask people are actually measuring the right thing.

In short: The internet city is huge, and a lot of people seem to be getting lost in it, especially on their phones and social media. But because everyone is using different maps and different rules to count them, we need to build a better, more consistent map before we can say exactly how big the problem really is.

Technical Summary

Technical Summary: Global Burden of Problematic Internet Use

Problem Statement
The pervasive influence of the internet has reshaped global interaction, yet it has also introduced a significant public health concern: Problematic Use of the Internet (PUI). PUI encompasses a heterogeneous constellation of behaviors, including problematic gaming, social media use, smartphone use, pornography use, and online shopping. Despite the growing recognition of PUI as a condition causing social, occupational, and mental health impairments, the global burden of disease remains unclear. Existing prevalence estimates vary widely (from 1% to 27%), driven by inconsistencies in operational definitions, measurement instruments, study populations, and geographic representation. Previous evidence syntheses have often focused on specific subtypes (e.g., "internet addiction" or "gaming disorder") without a comprehensive umbrella review to harmonize these diverging conceptualizations or assess the methodological quality of the underlying evidence. Consequently, prioritizing PUI on public policy agendas is hindered by a lack of precise, comparable, and high-quality epidemiological data.

Methodology
This study employed an umbrella review methodology, adhering to JBI guidelines and PRISMA 2020 reporting standards. The research team conducted a comprehensive search across six databases (MEDLINE, Embase, Scopus, Web of Science, CINAHL, and the Cochrane Review Library) from inception to February 2026.

• Eligibility Criteria: The review included systematic reviews and meta-analyses (and one scoping review) reporting prevalence estimates for PUI behaviors (gaming, general internet use, social media, smartphones, sexting) across any geographic region, time period, or demographic. Only English-language studies were included. Narrative and critical reviews were excluded to facilitate quantitative synthesis.
• Data Extraction and Quality Assessment: Data were extracted from 11 included reviews, representing 1,109 primary studies and over 3.1 million individuals. The methodological quality of the included reviews was critically appraised using the AMSTAR-2 tool, categorizing confidence levels as High, Moderate, Low, or Critically Low.
• Meta-Analysis: To generate aggregated global prevalence estimates, the authors extracted non-overlapping primary studies from the included reviews. Overlap was assessed using the Corrected Coverage Area (CCA). A random-effects meta-analysis with double arcsine transformation was performed to pool prevalence estimates for four main subtypes: problematic gaming, problematic internet use, problematic smartphone use, and problematic social media use. Heterogeneity was measured using $I^2$, and small-study effects were examined via funnel plots.

Key Contributions

• First Comprehensive Umbrella Review: This is the first study to synthesize existing evidence syntheses on the epidemiology of PUI across five broad categories (sexting, general internet use, gaming, social media, and smartphones) in a global context.
• Methodological Appraisal: The study provides a critical assessment of the quality of existing PUI literature, revealing that 9 out of 10 systematic reviews were rated as having "Low" or "Critically Low" confidence according to AMSTAR-2 standards.
• Harmonized Prevalence Estimates: By re-analyzing non-overlapping primary data, the study provides pooled global prevalence estimates that account for the heterogeneity of previous syntheses.
• Identification of Gaps: The review highlights significant disparities in research focus (heavily skewed toward gaming), geographic representation (concentrated in Europe and East Asia), and the lack of standardized diagnostic criteria and measurement tools.

Results

• Prevalence Estimates: The pooled global prevalence estimates for the four analyzed subtypes were:
  • Problematic Gaming: 6% (95% CI, 5%–7%)
  • Problematic Internet Use: 16% (95% CI, 15%–17%)
  • Problematic Smartphone Use: 32% (95% CI, 28%–35%)
  • Problematic Social Media Use: 23% (95% CI, 19%–28%)
• Heterogeneity: Substantial between-study heterogeneity was observed across all subtypes ($I^2 > 99\%$), reflecting variations in study methodologies, sampled populations, and definitions of PUI.
• Quality of Evidence: The critical appraisal indicated that most included reviews suffered from methodological flaws, particularly the failure to include a list of excluded studies with justification (70% of reviews) and a lack of a priori protocol establishment (40% of reviews).
• Regional and Demographic Trends: Research is heavily concentrated in the European and Western Pacific regions, with scarce data from Africa, the Americas, and Southeast Asia. Risk factors identified across subtypes include older age, male gender, and psychological distress.
• Funnel Plot Asymmetry: Evidence of funnel plot asymmetry suggests the presence of small-study effects, indicating that smaller studies may report higher prevalence rates than larger ones.

Significance and Claims
The paper claims that PUI behaviors affect a substantial proportion of the global population, with prevalence rates varying significantly by subtype (ranging from 6% for gaming to 32% for smartphone use). However, the authors emphasize that these findings must be interpreted with caution due to the "Low" or "Critically Low" methodological quality of the underlying evidence syntheses.

The study asserts that the current state of PUI research is fragmented, with a lack of harmonized conceptualizations and standardized measurement tools. The authors conclude that while PUI is a pressing digital public health issue, the current evidence base is insufficient to support precise policy development. They call for:

1. High-Quality Epidemiological Research: Specifically in underrepresented regions (lower- and lower-middle-income countries) and for under-researched subtypes (social media, smartphones, pornography).
2. Standardization: The development of clear definitions and standardized criteria for PUI behaviors to reduce heterogeneity in future studies.
3. Representative Sampling: A shift away from convenience or snowball sampling toward robust, representative population samples to generate more accurate prevalence estimates.

The paper positions itself not as a definitive solution to the measurement of PUI, but as a critical step in identifying the gaps and methodological limitations that must be addressed to establish a reliable foundation for public health interventions.