Four in Ten Nepali Teens Show Problematic Internet Use, Study Finds

Nearly half of older teenagers attending higher secondary schools in Nepal’s capital are showing signs of problematic internet use, according to a new cross-sectional study published in PLOS Global Public Health. The research, conducted among 406 students in grades 11 and 12 across Kathmandu Metropolitan City, found that 42.6 percent of participants met the criteria for problematic internet use, a pattern of online behavior so excessive or compulsive that it begins to interfere with daily functioning, academic performance, and social relationships. The finding adds Nepal to a growing list of low- and middle-income countries where rapid expansion of mobile connectivity is outpacing the public health infrastructure needed to manage its consequences.

The study, led by Manish Rajbanshi and colleagues, was designed to fill a conspicuous evidence gap. While problematic internet use, often abbreviated as PIU, has been extensively studied in East Asia, Europe, and North America, researchers have had comparatively little data from South Asia, and almost none focused specifically on Nepali adolescents in the critical final years of secondary school. The authors note that access to electronic devices and internet connectivity has expanded rapidly in Nepal, making adolescents in developing countries an increasingly vulnerable population. Excessive use has been linked in prior literature to academic difficulties, inattention, and social isolation, outcomes with obvious relevance for students preparing for the high-stakes examinations that shape university admission in the country.

Methodologically, the team employed a cross-sectional design with multistage random sampling, a technique that first selects schools and then randomly samples students within them, reducing the risk that the sample reflects only a handful of particularly wired institutions. Data collection relied on a self-administered questionnaire built around the Internet Addiction Test, or IAT, one of the most widely used screening instruments in the field. The IAT asks respondents to rate the frequency of behaviors such as failing to control time online, concealing use from family members, and feeling restless when disconnected, producing a score that researchers can threshold to classify respondents as having problematic use. A total of 406 students participated, with a mean age of 17.22 years and a standard deviation of 1.03, placing the cohort squarely in late adolescence.

The statistical approach followed a conventional two-stage logic. Bivariate analyses using the Chi-square test and bivariate binary logistic regression identified variables associated with PIU at a liberal significance threshold of p less than 0.1. Those candidate variables were then entered into a multivariate logistic regression model, which estimates the independent effect of each factor while holding the others constant. Results were reported as adjusted odds ratios, or AORs, with 95 percent confidence intervals. An AOR above 1 indicates elevated odds of problematic use compared with the reference group, and a confidence interval that excludes 1 signals a statistically reliable association. This layered approach matters because crude associations in bivariate analysis can dissolve once confounders such as age, family income, and usage patterns are accounted for.

What emerged from the multivariate model was a coherent portrait of who is most at risk. Students aged 19 and above had roughly double the odds of PIU compared with younger classmates, with an adjusted odds ratio of 2.16 and a confidence interval spanning 1.81 to 5.24. Family economic context also mattered: adolescents from households with non-agricultural income sources showed higher odds, at an AOR of 1.27 (95 percent CI: 1.07 to 2.25), a finding the authors interpret in light of greater device ownership and disposable income in urban, non-farming households. Platform preference proved telling as well. Students who favored YouTube had an AOR of 1.70 (95 percent CI: 1.07 to 2.72), and those who gravitated toward Facebook had an AOR of 1.65 (95 percent CI: 1.40 to 2.06), suggesting that algorithmically curated and socially driven feeds may be particularly effective at capturing and holding adolescent attention.

Timing and duration of use were among the strongest predictors in the model. Students who primarily used the internet in the evening had an AOR of 2.01 (95 percent CI: 1.09 to 3.69), while those whose use extended into the night had an AOR of 2.03 (95 percent CI: 1.21 to 3.38). The most striking figure belonged to heavy users: adolescents spending more than eight hours per day online had an AOR of 2.69 (95 percent CI: 1.48 to 4.91), meaning their odds of problematic use were nearly triple those of lighter users. The evening and night findings align with a broader body of research linking late-night screen exposure to sleep displacement, circadian disruption, and next-day fatigue, all of which can feed back into the cycle of compulsive use by degrading the self-regulation needed to log off.

The prevalence figure itself deserves scrutiny. At 42.6 percent, with a 95 percent confidence interval of 37.0 to 47.0 percent, the estimate suggests that problematic internet use is not a marginal phenomenon among Kathmandu’s older students but a mainstream behavioral health concern. The confidence interval is wide enough to encompass a range from just over a third to nearly half of all students, reflecting the sample size of 406, yet even the lower bound would represent a substantial burden. Cross-sectional studies of this kind cannot establish causation; they capture a snapshot in which, for example, depression-like withdrawal and heavy internet use may reinforce each other. Still, the consistency of the associations with duration, timing, and platform choice gives the findings practical traction for intervention design.

The authors argue that their results point toward targeted rather than blanket interventions. Awareness programs for students and parents, school-based preventive strategies embedded in the academic calendar, and regulation of the time adolescents spend online are the three pillars they highlight for promoting healthier internet behavior. The platform-specific associations offer a concrete lever: if YouTube and Facebook users carry elevated risk, digital literacy curricula could focus on how recommendation algorithms and social feedback loops are engineered to maximize engagement, teaching students to recognize the mechanics of the feeds they scroll. Similarly, the evening and night associations suggest that simple structural changes, such as device-free hours before bedtime or school counseling that addresses sleep hygiene alongside screen time, could yield outsized benefits.

The study also carries implications for how Nepal’s education and health authorities monitor the digital transition. Because risk rose with age within the school cohort, and because non-agricultural household income marked greater vulnerability, screening efforts might be concentrated in urban schools serving older students from economically diversified households. The researchers emphasize that the evidence base for Nepal remains thin, and this study of one metropolitan area cannot be generalized to rural districts where connectivity, device access, and family livelihoods differ substantially. Repeated surveys using the same IAT instrument would allow trend tracking as internet penetration continues to climb, turning a one-time snapshot into a surveillance tool.

For a country where smartphone ownership has become nearly universal among urban youth in less than a decade, the Kathmandu findings read as an early warning from the front lines of the global digital shift. The study’s core message is measured but urgent: problematic internet use among Nepali adolescents is common, patterned, and identifiable with standard screening tools, which means it is also addressable. The authors call for coordinated action across schools, families, and policymakers to promote healthy internet behavior before the cohort currently in grade 11 and 12, already showing a 42.6 percent prevalence, ages into adulthood with habits that are harder to change. As low- and middle-income countries absorb the same technologies that transformed wealthier nations, studies like this one provide the local evidence needed to craft responses grounded in data rather than alarm.

Subject of Research: Prevalence and risk factors of problematic internet use among higher secondary school students in Kathmandu, Nepal

Article Title: Problematic internet use and its associated factors among the students of higher secondary schools of Kathmandu Metropolitan City, Nepal

Article References: Rajbanshi, M., Thapa, S., Aryal, R., Joshi, J., Bishwokarma, B., Neupane, D. R., Shrestha, S. M., & Pahari, D. P. (2026). Problematic internet use and its associated factors among the students of higher secondary schools of Kathmandu Metropolitan City, Nepal. PLOS Global Public Health, 6(10), e0006511. https://doi.org/10.1371/journal.pgph.0006511

Image Credits: AI Generated

DOI: 10.1371/journal.pgph.0006511

Keywords: problematic internet use, internet addiction, adolescents, Nepal, Kathmandu, Internet Addiction Test, cross-sectional study, screen time, public health, school students, digital behavior, logistic regression