Smartphone use health outcomes in children and adolescents: A systematic review of behavioral, developmental, and environmental risk pathways

Abstract

BACKGROUND

Smartphone use is nearly ubiquitous among children and adolescents. Although concerns regarding psychosocial effects are well recognized, the broader physical health implications require careful synthesis.

AIM

To synthesize current evidence on problematic smartphone use (PSU) and related health outcomes in children and adolescents, with emphasis on developmental and cognitive changes, metabolic dysregulation, visual and musculoskeletal disorders, obesity risk, and radiofrequency electromagnetic field (RF-EMF) exposure.

METHODS

We conducted a structured systematic review following PRISMA-informed methodology. Multiple databases were searched for observational, experimental, and mechanistic studies examining smartphone exposure in individuals aged 0–18 years. Outcomes of interest included physical activity patterns, adiposity indices, metabolic markers, neurocognitive function, and RF-EMF-related health outcomes. Evidence from cross-sectional, longitudinal, case-control, and mechanistic studies was integrated. Study quality, exposure metrics, and age stratification were considered in interpretation.

RESULTS

A total of 97 studies met the inclusion criteria, comprising predominantly cross-sectional studies, alongside longitudinal cohorts and a limited number of experimental and randomized designs. Across domains, higher smartphone exposure and PSU were consistently associated with adverse health outcomes. Strong and consistent evidence was observed for sleep disruption, with multiple studies demonstrating dose-response relationships between screen exposure - particularly bedtime use - and shorter sleep duration, delayed sleep onset, and poorer sleep quality. Moderate evidence linked smartphone use to mental health outcomes, including depressive symptoms, anxiety, and emotional dysregulation, with stronger associations observed for PSU compared with total screen time. Evidence for neurocognitive outcomes was more heterogeneous, with consistent associations for attentional impairment but mixed findings for working memory and global cognitive function. Across physical health domains, smartphone use was associated with visual strain, musculoskeletal symptoms, and increased risk of being overweight/obesity, largely mediated by sedentary behavior and reduced physical activity. In contrast, evidence regarding RF-EMF exposure did not demonstrate consistent associations with carcinogenic or neurocognitive outcomes, although data remain limited and precautionary approaches are warranted.

CONCLUSION

The most robust health concerns related to smartphone use by children and adolescents appear to arise from behavioral displacement and sleep disruption rather than RF-EMF exposure. Clinical and public health strategies should prioritize reducing excessive screen time, preserving sleep, and promoting physical activity during critical developmental stages.

Key Words: Problematic smartphone use; Obesity; Metabolic dysregulation; Sedentary behavior; Sleep disturbance; Radiofrequency electromagnetic fields; Neurocognitive function; Digital health; Pediatric public health

Core Tip: This review synthesizes multidisciplinary evidence demonstrating that excessive or dysregulated smartphone use in children and adolescents is consistently associated with sleep disruption, attentional dysregulation, emotional symptoms, musculoskeletal strain, visual complaints, and increased metabolic risk. Across domains, dose-response patterns and developmental vulnerability are evident, with early childhood and adolescence representing particularly sensitive periods. Importantly, behavioral mechanisms - especially sleep disturbance, sedentary displacement, and reinforcement-driven problematic smartphone use - appear to mediate most observed risks, whereas current evidence does not confirm serious radiofrequency electromagnetic field-related harm. These findings support a developmentally informed, context-sensitive digital health framework emphasizing sleep hygiene, structured limits, parental modeling, and early identification of problematic use.

INTRODUCTION

Over the past two decades, mobile phones have evolved from simple communication devices into multifunctional digital platforms that infiltrate nearly every aspect of daily life. Globally, smartphone ownership has expanded rapidly, with children and adolescents among the fastest-growing user groups. In many regions, children are introduced to mobile devices in early childhood, often before school age, and usage intensifies throughout adolescence[1]. This unprecedented level of early and sustained exposure has raised growing concern regarding the potential consequences for child health and development.

Childhood and adolescence represent critical periods of rapid neurobiological, psychological, and social maturation. Brain plasticity, synaptic pruning, executive function (EF) development, sleep architecture, and emotional regulation undergo substantial refinement during these stages[2]. Environmental exposures during these sensitive windows may exert both immediate and long-term effects. Unlike adults, children have longer cumulative lifetime exposure to digital technologies, different usage patterns, and potentially increased biological susceptibility. These characteristics justify a focused pediatric evaluation of mobile phone-related health outcomes[3].

Mobile phone use encompasses a wide range of behaviors, including voice communication, messaging, internet browsing, social media engagement, gaming, and multimedia consumption. Consequently, potential health effects are multidimensional[4]. To understand these outcomes, it is necessary to distinguish between two primary risk pathways through which smartphone use may influence health: Behavioral and biological.

The behavioral and lifestyle pathway encompasses the direct consequences of how a device is used. This includes physical health concerns such as sleep disruption linked to evening screen exposure and blue light emission, visual strain and myopia progression associated with prolonged near work, musculoskeletal complaints, such as “text neck”, and sedentary behavior contributing to obesity risk[5]. In parallel, this pathway includes adverse mental health outcomes and neurodevelopmental implications, such as anxiety, attention difficulties, and problematic smartphone use (PSU) behaviors that resemble behavioral addiction[6,7].

The biological and environmental pathway focuses on the device’s physical properties, specifically its exposure to radiofrequency electromagnetic field (RF-EMF) emitted during operation[8]. While the behavioral pathway currently provides the primary evidence base for clinical concern, the biological pathway remains a subject of significant scientific and public debate. Although epidemiological evidence has not conclusively established severe long-term health risks in children, the International Agency for Research on Cancer classifies RF-EMF as a “possible human carcinogen” (Group 2B). This necessitates a precautionary evaluation of cumulative exposure and potential non-thermal effects during critical periods of brain development[9,10].

To ensure methodological clarity and address heterogeneity in the existing literature, this review distinguishes among three primary dimensions of smartphone engagement. General smartphone use refers to the quantitative volume of interaction with the device, typically operationalized as “total screen time” (duration) or the frequency of device checks (pickups). It represents a broad exposure metric that does not account for the nature of the content or the psychological quality of the engagement[11]. Screen exposure (passive vs active) characterizes the use modality. Passive exposure involves receptive consumption with minimal cognitive or physical interaction, such as streaming video content or mindless scrolling through social media feeds. In contrast, active exposure involves participatory engagement, such as mobile gaming, interactive educational apps, or two-way digital communication, which may exert different neurodevelopmental and metabolic effects[12]. PSU, unlike general use, is a behavioral construct characterized by a loss of self-regulation and functional impairment. It is defined by symptoms mirroring behavioral addiction, including compulsive checking, preoccupation, withdrawal (irritability when the device is inaccessible), and the persistence of use despite negative consequences on sleep, academic performance, or social relationships. As noted in several studies, PSU often demonstrates stronger associations with adverse mental health outcomes than total duration of use alone[6].

Despite the growing body of literature, existing evidence is characterized by substantial heterogeneity in exposure measurement, reliance on self-reported data, predominance of cross-sectional designs, and limited longitudinal studies. Furthermore, mobile phone-specific effects are frequently conflated with broader “screen time” research, limiting clarity regarding device-specific risks. These methodological challenges underscore the need for a systematic, comprehensive synthesis of the available evidence on mobile phone use in pediatric populations.

Given the ubiquity of mobile phones and their integration into educational, social, and recreational activities, understanding their health implications is essential for clinicians, families, educators, and policymakers. A balanced, evidence-based approach is required - one that acknowledges both potential benefits and measurable risks while identifying areas of scientific uncertainty.

Therefore, this systematic review aims to critically evaluate and synthesize the current evidence on the health effects of mobile phone use among children and adolescents, encompassing physical, neurodevelopmental, psychological, behavioral, and electromagnetic exposure. By clarifying the strengths and limitations of existing data, this review seeks to inform pediatric practice, guide public health policy, and identify research priorities.

MATERIALS AND METHODS

Study design and reporting standards

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines. The methodology was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the title “Growing up connected: Health consequences of mobile phone exposure in children and adolescents” (ID: CRD420261322228).

Conceptual definitions and eligibility criteria

To address conceptual ambiguity between general usage and addiction-like behaviors, this review strictly categorized exposure into three domains: (1) General smartphone use: Total duration (hours/day) and frequency of device interaction; (2) Screen exposure: Modality of use (e.g., passive scrolling vs active gaming) and timing (e.g., pre-sleep exposure); and (3) PSU: Validated behavioral patterns characterized by loss of control, withdrawal symptoms, and functional impairment.

The inclusion criteria following the PICO framework are shown in Table 1: (1) Population: Children and adolescents aged 0-18 years; (2) Exposure: Mobile phone or smartphone use, including voice/data, social media, gaming, and specific RF-EMF exposure. Studies assessing general “screen time” were only included if smartphone-specific data were isolated; (3) Comparators: Low vs high usage, users vs non-users, or longitudinal changes; (4) Outcomes: Primary outcomes included sleep (duration, quality, latency), mental health (anxiety, depression), and neurodevelopment. Secondary outcomes included visual/musculoskeletal health, obesity, and biological effects of RF-EMF; and (5) Study design: Randomized controlled trials (RCTs), cohort, case-control, and cross-sectional studies published in peer-reviewed English-language journals (inception to January 2026). Figure 1 shows the PRISMA 2020 flow diagram for study selection.

Figure 1  PRISMA 2020 flow diagram for study selection.

Table 1 PICO framework for the systematic review.

Component	Description	Operational definition in this review
Population (P)	Children and adolescents	Individuals aged 0-18 years. Studies including mixed populations were eligible only if pediatric data were reported separately. Subgroup analyses planned for early childhood (0-5 years), school-age children (6-12 years), and adolescents (13-18 years)
Exposure (I/E)	Mobile phone use and exposure	Use of smartphones or cellular phones, including: (1) Duration and frequency of use; (2) Voice calls and messaging; (3) Internet browsing; (4) Social media engagement; (5) Mobile gaming; (6) Night-time use; and (7) RF-EMF exposure attributable to mobile phones
Comparator (C)	Lower or no exposure	(1) Non-users vs users; (2) Low vs high usage groups; (3) Short vs prolonged exposure; (4) Baseline vs follow-up (longitudinal studies); and (5) Different exposure intensities (e.g., hours/day, nighttime use)
Primary outcomes (O1)	Core health outcomes	(1) Sleep disturbances (duration, quality, latency); (2) Mental health outcomes (anxiety, depression, emotional symptoms); (3) Attention and executive function; and (4) Neurodevelopmental outcomes
Secondary outcomes (O2)	Additional health and behavioral outcomes	(1) Visual complaints (digital eye strain, myopia); (2) Musculoskeletal symptoms; (3) Physical activity levels and obesity; (4) Problematic smartphone use/addiction behaviors; (5) Academic performance; and (6) Biological or health effects related to RF-EMF exposure
Study designs	Eligible study types	Randomized controlled trials, cohort studies, case-control studies, and cross-sectional studies
Setting	Geographic scope	No geographic restriction; studies from all regions included

RF-EMF: Radiofrequency electromagnetic field.

Information sources and search strategy

A systematic search was conducted across PubMed/MEDLINE, Scopus, Web of Science, EMBASE, PsycINFO, and the Cochrane Library. The search timeframe spanned from database inception to January 2026. The search strategy utilized MeSH terms and keywords: (“mobile phone” OR “smartphone*” OR “cell phone*”) AND (“child*” OR “adolescent*” OR “pediatric*”) AND (“health” OR “sleep” OR “mental health” OR “depression” OR “anxiety” OR “attention” OR “cognition” OR “obesity” OR “myopia” OR “RF-EMF”)*. Reference lists of included articles and relevant systematic reviews were manually screened to ensure literature saturation.

Study selection and extraction

Two independent reviewers screened titles and abstracts via reference management software. Full-text assessments were conducted against predefined criteria, with a third reviewer resolving discrepancies. Data were extracted using a standardized form covering study demographics, design, sample size, age range, validated exposure assessment tools (e.g., Smartphone Addiction Scale), outcome measures, adjusted confounders, and effect estimates [odds ratio (OR), risk ratio (RR), or β coefficients].

Risk of Bias and evidence synthesis

Risk of Bias was independently assessed using design-specific validated tools.

Newcastle-Ottawa Scale: For cohort and case-control studies.

Joanna Briggs Institute checklist: For cross-sectional studies.

Cochrane Risk of Bias: For randomized trials. Studies were categorized as low, moderate, or high Risk of Bias. Synthesis followed a tiered evidence approach. Rather than a purely narrative summary, findings were synthesized by categorizing studies according to design and sample size. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework was applied to primary outcomes to determine the overall certainty of evidence (high, moderate, low, or very low), ensuring that preliminary or parent-reported data were distinguished from robust, objective findings.

Subgroup and sensitivity analyses

Subgroup analyses were conducted when sufficient data were available and were stratified by age group (early childhood, school-age children, and adolescents), gender, exposure intensity, and geographic region. Sensitivity analyses were also performed by excluding studies classified as high Risk of Bias and by comparing adjusted vs unadjusted effect estimates to assess the robustness and stability of the findings.

Statistical analysis

Where at least three studies demonstrated sufficient homogeneity in exposure and outcome, a random-effects meta-analysis was performed. Heterogeneity was assessed using the I² statistic (> 50% indicating substantial heterogeneity). Publication bias was evaluated via funnel plots and Egger’s regression test for outcomes with ≥ 10 included studies.

Assessment of publication bias and certainty of evidence

When ten or more studies were included in a meta-analysis, publication bias was assessed using funnel plots and Egger’s regression test to evaluate potential small-study effects. The overall certainty of evidence for the primary outcomes was then appraised using the GRADE framework, which categorizes evidence into high, moderate, low, or very low certainty based on factors such as study limitations, inconsistency, indirectness, imprecision, and risk of publication bias.

RESULTS

Study selection

The literature search identified 2270 records. After removal of 425 duplicates, 1845 unique records remained for screening. Titles and abstracts of these records were screened, resulting in the exclusion of 1376 records that did not meet the inclusion criteria. The full texts of 469 articles were subsequently assessed for eligibility. We included 97 studies in this review. The selection process is detailed in the PRISMA flow diagram as shown in Figure 1.

Prevalence and patterns of mobile phone use among children

According to studies conducted in North America, Europe, and Asia, mobile phone exposure among children was consistently high, with evidence of early initiation and increasing intensity of use across developmental stages as shown in Table 2.

Table 2 Prevalence and patterns of smartphone use in children and adolescents before and during the coronavirus disease 2019 pandemic.

Ref.	Country	Study design	Sample size (n)	Age range	Period	Prevalence/usage findings	Key observations	Key strengths	Potential bias/limitations	Overall risk category
Kabali et al[13], 2015	United States	Cross-sectional (community-based)	350	6 months to 4 years	Pre-COVID (2014)	96.6% had used a mobile device	Most initiated < 1 year; approximately 75% owned device by age 4; daily use common by age 2	Clear inclusion criteria; utilized adapted validated survey from Common Sense Media	Specific to urban, low-income, minority population; limited generalizability to high-SES groups	Low
Shah and Phadke[14], 2023	India	Cross-sectional (hospital-based)	90	6 months to 4 years	Pre-COVID	73.3% prevalence of use	19% (3-4 years) ≥ 3 hours/day; parental reluctance high despite use	Validated questionnaire used; clear ethical and consent protocols	Small sample size (n = 90) from a single tertiary hospital	Moderate
Kopecký et al[9], 2021	Czech Republic	Cross-sectional (school survey)	27177	7-17 years	Pre-COVID (2014-2018)	Near-universal exposure1	High engagement in social media, YouTube, gaming; school policy influenced usage	Exceptionally large sample size (n = 27177); objective comparison of school policies	Reliance on self-reported behaviors rather than objective logs	Low
Kayiran et al[15], 2010	Turkey	Cross-sectional	724	6 months to 15 years	Pre-COVID	Not mobile-specific	Increasing device access and bedroom ownership with age	Large sample size (n = 724) for the specific SES demographic	Convenience sampling in a private hospital; potential for social desirability bias in parent reporting	Moderate
Lee et al[17], 2025	South Korea	National survey secondary analysis	54948	Middle and high school	During COVID	25.5% PSU; mean use 2828 minutes (weekday), 393.4 minutes (weekend)	Higher PSU among females and high school students; alcohol and smoking increased risk	Huge national dataset (n = 54948); complex sample statistical weighting for accuracy	Secondary data analysis limits control over initial measurement tools; self-reported	Low
Chun et al[18], 2023	South Korea	Cross-sectional	360	15-18 years	During COVID	Increased addiction among those with increased usage time	Associated with depressive symptoms, low self-control, cyberbullying; socioeconomic factors relevant	Uses a social-ecological model to categorize factors (individual, family, school)	Sample restricted to Korean adolescents aged 15-18; limited age range	Low
Serra et al[19], 2021	Italy	Cohort (self-report, pre-post comparison)	184	6-18 years	During COVID	Significant increase in frequency and duration of use vs pre-epidemic	Increased overuse/addiction; sleep, ocular, and musculoskeletal complaints	Evaluates specific health outcomes (ocular, musculoskeletal) alongside addiction	Anonymous questionnaire limits follow-up; significant gender imbalance in respondents (more females)	Low
Ferrara et al[20], 2023	Italy	Pre-post survey	130	6-18 years	During COVID	Significant increase in screen time (P < 0.02); higher addiction index (P < 0.001)	Increased early-morning headaches; reduced physical activity	Direct pre- vs post-lockdown comparison	Retrospective recall bias (asking about pre-lockdown habits during lockdown)	Moderate

¹Exact prevalence not extractable due to variation in outcome definitions. Risk of Bias tool: Joanna Briggs Institute (analytical).

PSU: Problematic smartphone use; COVID: Coronavirus disease; SES: Socioeconomic status.

Early childhood (0-4 years): Two cross-sectional studies provided extractable prevalence data for children aged 6 months to 4 years (total n = 440). In an urban, low-income, minority community in the United States (n = 350), 96.6% of children had used a mobile device, with most initiating use before age 1. By age 4, approximately three-quarters of children owned their own mobile device. Daily use was common by age 2, and mobile device screen time was comparable to television viewing. Independent use increased with age, and approximately one-third of preschool-aged children engaged in media multitasking[13].

In contrast, a tertiary hospital-based study from India (n = 90) reported that 73.3% of children aged 6 months to 4 years used mobile phones. Usage increased significantly with age. While most children (57.6%) used devices for less than one hour per day, 19% of those aged 3-4 years used mobile phones for three or more hours daily. Notably, despite 93.3% of parents reporting reluctance to provide phones, 71.4% of their children were using them[14].

Meta-analysis of these two studies yielded a pooled prevalence of mobile phone use of 91.8% [95% confidence interval (CI): 89.2%-94.4%] among children aged 0-4 years. However, heterogeneity was substantial (I² ≈ 97%), indicating considerable variability between studies. Differences in geographic setting, socioeconomic context, sampling framework (community-based vs hospital-based), and operational definitions of “use” likely contributed to this heterogeneity. Despite these differences, both studies demonstrated that exposure to mobile phones in early childhood is widespread.

School-age children and adolescents: Large-scale survey data from the Czech Republic (n = 27177; ages 7-17 years) demonstrated extensive mobile phone use among school-aged children and adolescents. Engagement was strongly linked to social networking platforms, YouTube, and mobile gaming. In schools permitting device use, students reported active use during school hours, whereas restrictive policies were associated with concealed use and increased desire to access devices[9].

Although not specific to mobile phones, data from a Turkish cohort of high socioeconomic families (n = 724; ages 6 months to 15 years) showed substantial engagement with electronic media, increasing with age. The rate of device presence in bedrooms rose progressively across age groups, suggesting an age-related transition from shared to personal media access and greater autonomy of use[15].

Children with intellectual disabilities: Qualitative findings from Haryana, India, indicated that children with intellectual disabilities were actively using smartphones, often independently or with limited assistance. Use was primarily entertainment-focused but became particularly important during coronavirus disease 2019 (COVID-19)-related educational disruptions, highlighting both accessibility challenges and adaptive use patterns in vulnerable populations[16].

Impact of the COVID-19 pandemic on smartphone use: Multiple studies conducted during the COVID-19 pandemic reported significant increases in smartphone use duration and problematic use. A nationally representative South Korean study (n = 54948 adolescents) reported mean daily smartphone use of 282.8 minutes on weekdays and 393.4 minutes on weekends during the pandemic. The prevalence of PSU was 25.5%. Female students and high school students exhibited significantly higher PSU rates (P < 0.001). Alcohol consumption and smoking were independently associated with increased PSU risk[17]. Another Korean study (n = 360; ages 15-18 years) demonstrated that adolescents whose usage time increased after the outbreak were significantly more likely to meet criteria for smartphone addiction. Risk factors included female sex, depressive symptoms, lower self-control, and cyberbullying victimization. Among those with increased usage, socioeconomic and academic factors were also associated with addiction risk[18].

European data demonstrated similar trends. An Italian cohort study (n = 184; ages 6-18 years) conducted during the second pandemic wave reported significantly greater smartphone use than in the pre-epidemic period, accompanied by increased symptoms of overuse and addiction. Reported adverse outcomes included sleep disturbances, ocular and musculoskeletal complaints, mood modification, distraction, and social withdrawal[19]. A pre-post study of 130 Italian children (mean age 11.8 ± 2 years) found significantly higher screen time during lockdown compared with the pre-lockdown period (P < 0.02), alongside increased smartphone addiction index scores (P < 0.001), more early-morning headaches (P < 0.05), and reduced physical activity (P < 0.0001)[20].

The synthesis of pandemic-related findings from Asian and European studies indicates that the COVID-19 pandemic was consistently associated with increased daily smartphone use, higher prevalence of problematic or addictive use, and a marked shift toward prolonged entertainment-based and online social engagement. Simultaneously, many children and adolescents have experienced reductions in physical activity and declines in sleep quality. Although substantial heterogeneity across study designs, addiction measurement tools, and outcome definitions made quantitative pooling impractical, the overall pattern of results was uniform. Collectively, the evidence shows that pandemic-related restrictions amplified pre-existing trends in digital engagement, accelerating both the duration and intensity of smartphone use among young populations.

The collective evidence demonstrates that mobile phone exposure begins early in life, with prevalence among preschool-aged children typically exceeding 70% and, in some settings, surpassing 90%. Daily use commonly emerges by age two, and personal device ownership increases substantially by age four. Throughout school age and adolescence, usage intensity continues to escalate. The COVID-19 pandemic further amplified both overall exposure and the risk of problematic or excessive use. Overall, mobile phone use among children is highly prevalent across diverse contexts, characterized by early initiation and increasing integration into daily routines.

Effects of mobile phone use on sleep outcomes in children and adolescents

Across longitudinal cohorts, cross-sectional studies, randomized trials, and experimental sleep-laboratory designs, evidence consistently indicates that smartphone use - particularly excessive, bedtime, or nocturnal use - is associated with shorter sleep duration, poorer sleep quality, increased nighttime awakenings, and daytime sleepiness across childhood and adolescence, as shown in Table 3.

Table 3 Effects of smartphone use on sleep outcomes by age group and study design.

Ref.	Country	Age group	Sample size (n)	Study design	Exposure definition	Sleep outcomes	Key findings	Appraisal tool	Primary bias considerations	Overall risk
Kim et al[21], 2020	South Korea	5-8 years	Cohort (K-CURE)	Prospective cohort	Smartphone overuse (> 1 hour/day)	Total sleep time; Children’s Sleep Habits Questionnaire score; nocturnal awakening	Overuse group had shorter total sleep time (P < 0.05) and higher sleep problem scores, including more night awakenings	NOS	Strengths: Longitudinal design; part of a large prospective cohort (K-CURE). Limitations: Reliance on parental questionnaires for sleep assessment, which may introduce recall bias	Moderate
Lee et al[22], 2022	South Korea	4-7 years	314	4-year longitudinal study	Frequency of smartphone use	Sleep problems; bedtime resistance; sleep duration; daytime sleepiness	Smartphone use significantly predicted sleep problems (β = 0.328, P < 0.001); associated with shorter sleep and greater bedtime resistance	NOS	Strengths: Robust 4-year follow-up (2017-2020); utilized mixed-model analysis. Limitations: Sleep data were caregiver-reported via self-administered questionnaires	Moderate
Pickard et al[23], 2024	United Kingdom	16-30 months	105	Randomized clinical trial	Removal of screen use in hour before bedtime	Actigraphy-measured sleep efficiency; night awakenings	Screen removal improved sleep efficiency and reduced awakenings (small-moderate effect sizes)	RoB 2	Strengths: Assessor-blinded; objective sleep measurement via actigraphy; high retention (99%) and adherence (94%)	Low
Abid et al[24], 2024	Tunisia	Mean 9 years	13	Experimental (acute and repeated exposure)	90-minute nocturnal smartphone exposure	Total sleep time; waking after sleep onset; cognitive performance	Acute and repeated exposure reduced sleep time (-29 minutes to -47 minutes, P < 0.01); repeated exposure produced greater sleep disruption	RoB 2	Strengths: Objective measures (actigraphy, salivary cortisol, cognitive tests). Limitations: Extremely small sample size (n = 13), which limits statistical power and generalizability	Moderate to high
Foerster et al[27], 2019	Switzerland	7th-9th grade	843	Prospective cohort	Nocturnal awakenings due to phone; high screen time	Restless sleep; difficulty falling asleep	Nocturnal awakenings associated with new-onset restless sleep (OR = 5.66); high screen time linked to sleep onset problems	NOS	Strengths: Large sample (n = 843); controlled for relevant confounders using logistic regression. Limitations: Subjective reporting of nocturnal phone-related awakenings	Low to moderate
Nagata et al[25], 2023	United States	10-14 years	10280	Cross-sectional (ABCD study)	Bedtime screen behaviors; device in bedroom	Trouble falling/staying asleep; sleep disturbance	Device in bedroom increased sleep disturbance risk (RR = 1.27); bedtime streaming, texting, gaming associated with sleep problems	JBI	Strengths: National dataset with very large sample size (n = 10280); extensive control for socioeconomic and demographic variables. Limitations: Cross-sectional nature precludes causal inference	Low
Yoon et al[26], 2021	South Korea	Grade 4 and 7 (approximately 10-13 years)	4940	Cross-sectional panel analysis	Smartphone addiction subscales	Sleep duration	Addiction (tolerance subscale) associated with shorter sleep; gender moderated effect	JBI	Strengths: High sample size (n = 4940) from a national panel survey; used validated smartphone addiction sub-factors	Low
Rafique et al[28], 2020	Saudi Arabia	17-23 years	1925	Cross-sectional	≥ 8 hours/day use; ≥ 30 minutes after lights-off; phone near pillow	PSQI sleep quality; sleep latency; daytime sleepiness	High use and bedtime exposure associated with poor sleep quality and longer sleep latency	JBI	Strengths: Large sample size (n = 1925); used the validated PSQI. Limitations: Potential for social desirability bias in reporting screen time	Low to moderate
Tu et al[29], 2023	China	University students	152	Randomized longitudinal intervention	Restricting while-in-bed smartphone use	PSQI sleep quality; cognitive arousal	Restriction improved sleep quality; effect mediated by reduced pre-sleep cognitive arousal	RoB 2	Strengths: Longitudinal mediation analysis with a 4-week follow-up; significant focus on physiological mechanisms like cognitive arousal	Low

The following Risk of Bias assessment has been conducted: The Newcastle-Ottawa Scale for longitudinal/cohort studies, the Joanna Briggs Institute checklist for cross-sectional studies, and the Cochrane Risk of Bias framework for randomized trials and experimental designs. PSQI: Pittsburgh Sleep Quality Index; OR: Odds ratio; RR: Risk ratio; NOS: Newcastle-Ottawa Scale; JBI: Joanna Briggs Institute; RoB 2: Cochrane Risk of Bias.

Early childhood (toddlers and young children): Evidence from prospective cohort and randomized clinical trial data suggests that even moderate smartphone exposure adversely affects sleep in younger children. A Korean prospective cohort study of children aged 5-8 years demonstrated that children classified as smartphone overusers (> 1 hour/day) had significantly shorter total sleep time compared to controls (F = 6.362, P < 0.05). Overuse was also associated with higher total scores on the Children’s Sleep Habits Questionnaire, including significantly more nocturnal awakenings[21]. Similarly, a 4-year longitudinal study in Korean children under 7 years found that smartphone use frequency significantly predicted sleep problems (β = 0.328, P < 0.001), with stronger associations than television, personal computer, or tablet use. Smartphone use correlated with bedtime resistance (r = 0.067), shorter sleep duration (r = 0.089), nighttime awakening (r = 0.066), and daytime sleepiness (r = 0.102), all statistically significant[22].

Importantly, a randomized clinical trial involving toddlers aged 16-30 months demonstrated that removing screen exposure in the hour before bedtime led to modest improvements in objectively measured sleep efficiency and reductions in nighttime awakenings. Although effect sizes were small to moderate, the intervention confirmed a causal relationship between pre-bed screen exposure and sleep disruption[23]. Collectively, findings in early childhood indicate that smartphone exposure - particularly in the evening - reduces sleep duration and quality, and that behavioral restriction may improve sleep outcomes.

School-age children: Cross-sectional and experimental evidence in school-aged children consistently demonstrates detrimental effects of nocturnal smartphone use on sleep architecture and next-day functioning. An experimental study in Tunisian children (mean age 9 years) showed that both acute (one night) and repeated (five nights) 90-minute nocturnal smartphone exposure significantly reduced total sleep time (-28.8 minutes to -46.7 minutes compared with baseline; P < 0.01). Repeated exposure produced greater sleep reduction and increased waking after sleep onset by up to 43%. Cognitive performance was also impaired, with increased errors on the Stroop and reaction-time tasks and reduced physical performance following repeated exposure[24]. These findings suggest a dose-response effect, with cumulative nocturnal exposure exerting progressively stronger sleep disruption.

Adolescents: Large-scale cohort and cross-sectional studies in adolescents demonstrate consistent associations between bedtime smartphone behaviors and sleep disturbance. In a nationally representative United States sample of 10280 early adolescents (ABCD study), having an Internet-connected device in the bedroom was associated with increased risk of trouble falling or staying asleep (adjusted RR = 1.27, 95%CI: 1.12-1.44) and overall sleep disturbance (RR = 1.15, 95%CI: 1.06-1.25). Leaving the phone ringer activated overnight and engaging in bedtime streaming, gaming, texting, or social media use were all associated with sleep difficulties[25]. Similarly, a Korean panel survey of 4940 youths found that smartphone addiction, particularly the tolerance subscale, was significantly associated with reduced sleep duration. Sleep duration decreased with age (grade 4: 9.17 hours; grade 7: 7.96 hours), and gender moderated the relationship between addiction and sleep duration[26].

A prospective Swiss cohort study demonstrated that adolescents experiencing at least one monthly nocturnal mobile phone awakening were more likely to develop restless sleep (OR = 5.66, 95%CI: 2.24-14.26) and problems falling asleep (OR = 3.51, 95%CI: 1.05-11.74) over one year. High baseline screen time was also associated with new-onset sleep problems and daytime symptoms, such as reduced energy and concentration[27]. In Saudi university students, prolonged daily mobile use (≥ 8 hours/day), bedtime use ≥ 30 minutes after lights-off, and keeping the phone near the pillow were all significantly associated with poor sleep quality, increased sleep latency, and daytime sleepiness[28]. A randomized longitudinal intervention among Chinese undergraduates demonstrated that restricting smartphone use while in bed significantly improved sleep quality, with reductions in pre-sleep cognitive arousal mediating the effect[29].

Across developmental stages, the evidence demonstrates several consistent patterns linking smartphone use with impaired sleep outcomes in children and adolescents. Higher daily use and nocturnal exposure are consistently associated with shorter total sleep time, greater difficulty initiating sleep, more frequent nighttime awakenings, and reduced sleep efficiency. A dose-response relationship is evident, particularly with repeated nighttime exposure, where cumulative use produces progressively greater disruption. Bedtime behaviors play a central mediating role, including keeping devices in the bedroom, engaging in post-lights-off use, and receiving overnight notifications. Importantly, findings from randomized interventions show that restricting smartphone use before or during bedtime can partially reverse these effects. Although heterogeneity in outcome assessment methods - such as actigraphy, Pittsburgh Sleep Quality Index, Children’s Sleep Habits Questionnaire, and various self-report measures - limits the feasibility of quantitative pooling, the direction and strength of associations remain consistent across cohort, cross-sectional, experimental, and randomized study designs. Overall, cumulative evidence supports a clear adverse relationship between excessive, bedtime, or nocturnal smartphone exposure and sleep duration and quality from toddlerhood through adolescence, with experimental and randomized data supporting causal interpretation.

Effects of smartphone use on mental health outcomes

Across the included studies (n > 25000 participants across Europe, Asia, North America, and the Middle East), the association between smartphone use and mental health was examined using cross-sectional (majority), longitudinal cohort, and RCT designs as shown in Table 4. Mental health outcomes assessed included: Depression, anxiety, stress, loneliness, life satisfaction, well-being/quality of life, insomnia, and peer relationship problems. A consistent pattern emerged: PSU and smartphone addiction demonstrated stronger and more consistent associations with adverse mental health outcomes than total screen time alone.

Table 4 Association between smartphone use and mental health outcomes by age group and study design.

Ref.	Country	Age group	Sample (n)	Study design	Exposure type	Mental health outcomes	Key effect estimates	Appraisal tool	Primary bias considerations	Overall risk
Meskini et al[30], 2024	Morocco	Middle school adolescents	341	Cross-sectional	Smartphone overuse (SAS)	Depression (HADS); anxiety (HADS)	Depression: r = 0.403 (P < 0.001); anxiety: r = 0.244 (P = 0.013)	JBI	Strengths: Used validated assessment scales (SAS, HADS). Limitations: Geographically restricted to one city (Kenitra); cross-sectional design prevents causal claims	Moderate
Carter et al[31], 2024	United Kingdom	16-18 years	657	Cross-sectional (multi-school)	Problematic smartphone use (SAS); screen time	Moderate depression (PHQ-9); anxiety (GAD-7); insomnia	PSU associated with depression (aOR = 2.96); anxiety (aOR = 2.03); insomnia (aOR = 1.64). Screen time not significant	JBI	Strengths: Multi-school sample (n = 657); adjusted for confounders via multi-level logistic regression	Low to moderate
Mayerhofer et al[32], 2024	Austria	14-20 years	913	Cross-sectional	PSU (SAS-SV); screen time	Depression; anxiety; disordered eating; loneliness	PSU associated with depression (aOR = 1.46); anxiety (aOR = 1.86). Screen time associated with loneliness	JBI	Strengths: Large sample size (n = 913). Limitations: Online survey format may lead to self-selection bias	Moderate
Liu et al[34], 2025	United States	16-18 years	137	Cross-sectional	Smartphone attachment (MPIQ)	Anxiety; depression	Anxiety (adjusted β = 0.26); depression (adjusted β = 0.15)	JBI	Strengths: Used PROMIS pediatric short forms. Limitations: Small sample size (n = 137) and lack of ethnic diversity (79.6% White)	Moderate
Zablotsky et al[33], 2025	United States	12-17 years	Nationally representative	Cross-sectional (NHIS-Teen)	≥ 4 hours/day non-school screen time	Depression; anxiety; low social support	High screen use associated with depression and anxiety symptoms	JBI	Strengths: Large-scale, nationally representative dataset (NHIS-Teen); includes parent-reported covariates	Low
Poulain et al[40], 2025	Germany	10-17 years	1113 (2576 observations)	Repeated cross-sectional (2018-2024)	PSU; > 3 hours/day use	Quality of life	PSU and long duration associated with lower QoL; stronger post-COVID; greater effect in girls	JBI	Strengths: Seven-year time trend analysis (2018-2024) within a dedicated cohort	Low
Selak et al[42], 2025	Europe	10-15 years	284	4-wave longitudinal	Parental smartphone use during interaction	Anger; sadness; subjective well-being	Parental use predicted child anger/sadness → lower well-being (mediation model)	NOS	Strengths: Longitudinal design with four time points; unique predictor (parental phubbing)	Low to moderate
Gath et al[41], 2026	New Zealand	2-8 years (prospective)	6281	Longitudinal cohort	> 1.5-2.5 hours/day early screen exposure	Peer problems; social functioning	> 2.5 hours/day at age 2 associated with increased peer problems at age 8	NOS	Strengths: Very large sample (n = 6281); long-term prospective data from age 2 years to 8 years	Low
El-Sayed Desouky and Abu-Zaid[35], 2020	Saudi Arabia	University students	1513	Cross-sectional	Smartphone addiction (PUMP)	Depression; trait anxiety	Significant positive correlations between addiction and depression/anxiety	JBI	Strengths: Large university sample (n = 1513); used multiple validated tools (PUMP, Taylor, Beck)	Low to moderate
Nikolic et al[36], 2023	Serbia	Medical students	761	Cross-sectional	Smartphone addiction (SAS-SV)	Depression; anxiety; stress	Depression (OR = 2.51); anxiety (OR = 2.04); stress (OR = 1.75)	JBI	Strengths: Multi-city selection; comprehensive analysis (multivariate regression) of independent factors	Low
Daniyal et al[39], 2022	Pakistan	University students	400	Cross-sectional	High vs low cell phone use	Depression; mood disorder; loneliness	Depression (r = 0.430); mood disorder (r = 0.608)	JBI	Strengths: Correlated smartphone use with both physical symptoms (neck/back pain) and mental health	Moderate
Zhu et al[37], 2025	China	Undergraduates	322	Cross-sectional mediation	Smartphone addiction (MPAI)	Depression; anxiety; life satisfaction	Addiction → negative emotions (r = 0.332); depression mediated ↓ life satisfaction	JBI	Strengths: Provides a specific mediation model for life satisfaction	Moderate
Pieh et al[38], 2025	Austria	University students	111	Randomized controlled trial	Screen reduction ≤ 2 hours/day (3 weeks)	Depression (PHQ-9); stress; well-being	Significant reduction in depression (η2 = 0.109); stress (η2 = 0.085); improved well-being	RoB 2	Strengths: RCT design; includes follow-up; intention-to-treat analysis. Limitations: Study was non-blinded	Low to moderate

To evaluate the association between smartphone use and mental health outcomes, the following Risk of Bias assessment has been conducted: Assessments follow the Joanna Briggs Institute checklist for cross-sectional studies, the Newcastle-Ottawa Scale for longitudinal designs, and the Cochrane Risk of Bias for randomized trials. SAS: Zung’s Self-rating Anxiety Scale; HADS: Hospital Anxiety and Depression Scale; JBI: Joanna Briggs Institute; PHQ-9: Patient Health Questionnaire 9; GAD-7: General Anxiety Disorder 7; aOR: Adjusted odds ratio; PSU: Problematic smartphone use; MPIQ: Mobile Phone Involvement Questionnaire; QoL: Quality of life; COVID: Coronavirus disease; NOS: Newcastle-Ottawa Scale; PUMP: Problematic Use of Mobile Phones; OR: Odds ratio; MPAI: Mobile Phone Addiction Index; RoB 2: Cochrane Risk of Bias; RCT: Randomized controlled trial; SAS-SV: Smartphone Addiction Scale-Short Version.

Depression: Depressive symptoms were the most consistently reported outcome.

Adolescents: In middle school adolescents in Morocco (37.4% smartphone overuse prevalence), PSU showed a significant positive correlation with depression (r = 0.403, P < 0.001)[30]. In United Kingdom adolescents (16-18 years), PSU was associated with a nearly threefold increase in the odds of moderate depression (adjusted OR = 2.96, 95%CI: 1.80-4.86), whereas total screen time was not independently associated[31]. Similarly, Austrian adolescents demonstrated increased odds of depressive symptoms with PSU (adjusted OR = 1.46)[32]. In United States adolescents (NHIS-Teen 2021-2023), ≥ 4 hours/day of non-school screen time was associated with a higher prevalence of self-reported depression symptoms[33]. In community adolescents, smartphone attachment, as measured by the Mobile Phone Involvement Questionnaire, was independently associated with higher depression scores after adjustment (adjusted β = 0.15, P < 0.05), with females showing higher attachment levels[34].

Young adults/university students: Among Saudi university students (n = 1513), Problematic Use of Mobile Phones addiction scores correlated significantly with depression[35]. In Serbian medical students, depression was independently associated with smartphone addiction (multivariate OR = 2.51)[36]. Among Chinese undergraduates in the post-COVID era, smartphone addiction was significantly correlated with negative emotions (r = 0.332), and depression mediated the relationship between addiction and reduced life satisfaction[37]. Importantly, the Austrian RCT demonstrated that reducing screen time to ≤ 2 hours/day for three weeks produced significant reductions in depressive symptoms (η² = 0.109), supporting a causal relationship rather than mere correlation[38].

Anxiety: Anxiety outcomes showed similar patterns. In Moroccan adolescents, smartphone overuse correlated with anxiety (r = 0.244, P = 0.013)[30]. In United Kingdom adolescents, PSU doubled the odds of moderate anxiety (adjusted OR = 2.03)[28]. In Austrian adolescents, PSU was associated with increased anxiety symptoms (adjusted OR = 1.86)[29]. Among Serbian medical students, smartphone addiction was associated with anxiety (OR = 2.04)[36]. Among Saudi university students, addiction scores were positively correlated with trait anxiety[35]. In adolescents with elevated smartphone attachment, anxiety scores were significantly higher (adjusted β = 0.26, P < 0.01)[31]. Collectively, effect sizes ranged from small to moderate correlations (r = 0.24) to clinically meaningful adjusted ORs (= 2.0).

Stress and emotional distress: Stress outcomes were particularly highlighted in Serbian medical students (OR = 1.75)[36], Austrian RCT (screen reduction improved stress; η² = 0.085)[32], and the Chinese post-COVID mediation model, where negative emotions mediated addiction-life satisfaction pathways[31]. These findings suggest that smartphone overuse may act as a chronic stress amplifier, potentially through sleep disruption, social comparison, or cognitive hyperarousal.

Loneliness and social well-being: Findings were more nuanced. In Austria, loneliness was associated with screen time, but not PSU[32]. The study of Daniyal et al[39] in 2022, lower-use groups reported more loneliness than high users, suggesting that moderate digital connectivity may buffer isolation in young adults. However, in United States teenagers, high screen use was associated with infrequent emotional and peer support[33]. These mixed findings suggest bidirectionality: Excessive passive use may increase isolation, while moderate communicative use may reduce loneliness. This highlights the importance of content and purpose of use, not duration alone.

Quality of life and life satisfaction: In the German LIFE Child cohort (2018-2024), both PSU and > 3 hours/day usage were significantly associated with lower quality of life, with deterioration particularly evident post-2021 and stronger effects in girls and younger children[40]. The COVID-19 pandemic appeared to amplify these trends. In Chinese undergraduates, negative emotions mediated the association between smartphone addiction and reduced life satisfaction[37]. The Austrian RCT confirmed that screen reduction improved well-being (η² = 0.053)[32].

Early childhood and peer functioning: Prospective data from the growing up in New Zealand study demonstrated that smartphone use of 1.5 hours/day at age 2 predicted below-average language and educational performance at 4.5 years, whereas 2.5 hours/day predicted increased peer relationship problems at age 8. Although not purely psychiatric outcomes, peer dysfunction is a recognized precursor of later internalizing disorders[41].

Parental smartphone use and child emotional outcomes: A four-wave longitudinal study demonstrated that parental smartphone uses during conversations predicted increased child anger and sadness, increased withdrawal from seeking parental attention, and reduced child well-being[42]. This finding expands the exposure model from child device use to relational displacement effects, highlighting family-level mechanisms.

Dose-response and mechanistic patterns: The reviewed studies consistently demonstrate several doses-responses and mechanistic patterns linking smartphone use to adverse mental health outcomes. Evidence shows that using smartphones for four or more hours per day is frequently associated with higher risks of depression and anxiety, and that increasing scores on smartphone addiction scales correlate linearly with greater symptom severity. Notably, PSU - which reflects psychological dependence, compulsive behaviors, and emotional attachment - emerges as a stronger predictor of mental health difficulties than total screen time alone, as reported in studies such as those by Carter et al[31] and Mayerhofer et al[32]. Gender-related differences are also evident, with females consistently exhibiting higher levels of problematic use and showing stronger associations with depression and anxiety. Furthermore, longitudinal data indicate that the COVID-19 pandemic amplified existing trends, intensifying both smartphone use and related mental health symptoms. Importantly, RCT findings support a causal interpretation, demonstrating that reducing smartphone screen time can lead to measurable improvements in depressive symptoms, stress levels, and overall well-being.

Synthesis interpretation: Overall, the evidence demonstrates a moderate yet consistent association between PSU and adverse mental health outcomes across developmental stages. Cross-sectional studies commonly report odds ratios between 1.5 and 3.0 for depression and anxiety, while correlation coefficients typically range from r = 0.24 to 0.43. Experimental studies that reduce smartphone use show small-to-moderate causal effects, supporting a direct influence on mental well-being. Although residual confounding and reverse causality cannot be fully excluded in cross-sectional designs, the convergence of findings from longitudinal and interventional studies strengthens the inference that PSU contributes to emotional dysregulation, depressive symptoms, anxiety, and reduced life satisfaction among children and adolescents.

Effects of smartphone use on attention and EF

Overview of evidence: Ten studies investigated the relationship between smartphone use, screen time, or PSU and attention or EF across different developmental stages as shown in Table 5. These studies employed various methodologies, including experimental laboratory designs, RCTs (both completed and protocol-based), prospective cohort studies, cross-sectional neuropsychological assessments, and electroencephalography (EEG)-based neurocognitive investigations. The outcomes assessed encompassed a broad range of EF domains, such as working memory capacity, cognitive flexibility, inhibitory control, selective and sustained attention, fluid intelligence, emotional regulation, and neurophysiological markers including N200 and P300 activity. Across this diverse body of research, three consistent findings emerged: Higher screen time and smartphone addiction were associated with weaker EF - particularly reduced working memory and inhibitory control; parentrated behavioral EF difficulties appeared more consistently than deficits detected through performance-based cognitive tasks; and the mere presence of a smartphone, even without active use, diminished available cognitive resources, demonstrating the “brain drain” effect.

Table 5 Executive function outcomes associated with smartphone use by age group and study design.

Age group	Ref.	Study design	Country	Sample size (n)	EF domains assessed	Main findings	Direction of association	Appraisal tool	Primary bias considerations	Overall risk
Preschool (2-6 years)	Lakicevic et al[43], 2025	Cross-sectional	Russia	1016	Cognitive flexibility, verbal working memory, and inhibition	Screen time weakly negatively correlated with flexibility and verbal WM; very weak negative correlation with inhibition	Negative (small effect)	JBI	Strengths: Large, representative sample (n = 1016). Limitations: Reliance on parent-reported screen time; cross-sectional design limits causal interpretation of EF deficits	Low
	Horowitz-Kraus et al[44], 2024	EEG cross-sectional	Israel	4-year-old (n ≈ 80)	Behavioral EF, N200, P300	Longer screen exposure is associated with poorer EF performance and altered executive-related EEG markers	Negative	JBI	Strengths: Use of objective neurophysiological markers (N200, P300) alongside behavioral tasks. Limitations: Small sample size (n ≈ 80); specific to 4-year-old	Moderate
	Oflu et al[45], 2021	Cross-sectional	Turkey	240	Emotional regulation (behavioral EF)	≥ 4 hours/day associated with higher emotional lability and negativity	Negative (dose-related)	JBI	Strengths: Identifies a clear dose-response relationship (at ≥ 4 hours). Limitations: Potential for social desirability bias in parent reports	Moderate
School-age (7-12 years)	Al-Amri et al[46], 2023	Cross-sectional	Saudi Arabia	186	Selective attention accuracy, reaction time	Smartphone addiction associated with reduced attention accuracy; no difference in reaction time	Negative	JBI	Strengths: Assessed specific attentional accuracy rather than general cognition. Limitations: Small sample size (n = 186)	Moderate
	Shekhawat et al[47], 2024	Cross-sectional	India	50	Parent-reported cognitive flexibility	> 4 hours/day associated with cognitive impairment in 66.7%	Negative	JBI	Strengths: Focuses on both cognitive and physical (posture) outcomes. Limitations: Very small sample (n = 50) and use of a self-constructed, unvalidated e-questionnaire	High
Adolescents (13-18 years)	Tauste-Garcia et al[48], 2025	Cross-sectional	Spain	269	Parent-rated EF, performance-based EF	PSU associated with greater parent-reported EF deficits; no difference in lab tasks	Negative (behavioral > performance)	JBI	Strengths: High-quality comparison between subjective (parent-rated) and objective (lab-based) EF tasks	Low to moderate
	Yum et al[49], 2025	RCT protocol	Hong Kong	Planned n = 240	Attention, inhibitory control, and EEG	Trial targeting smartphone overuse in ADHD adolescents; outcomes pending	Not yet available	RoB 2	Strengths: Planned randomization and use of objective EEG markers in a high-risk (ADHD) population. Note: As a protocol, risk assessment is based on design intent	Low (for design)
Young adults (comparative experimental evidence)	Ward et al[50], 2017	Laboratory experimental	United States	> 800	Working memory (OSpan), fluid intelligence (Raven), sustained attention	Smartphone presence (desk vs other room) reduced working memory and fluid intelligence; sustained attention was unaffected	Negative (causal evidence)	RoB 2	Strengths: Large sample size (n > 800); randomized experimental conditions allow for causal claims regarding “brain drain”	Low

The table illustrates that across developmental stages, smartphone exposure is consistently associated with executive function difficulties. In preschool children, studies show small yet consistent negative associations between screen time and core executive function domains, accompanied by emerging neurophysiological correlations. Among school-age children, excessive smartphone use is associated with reduced attentional accuracy and increased parent-reported cognitive impairment. In adolescents, behavioral executive dysfunction is observed more frequently than deficits on laboratory-based tasks, suggesting a greater impact on everyday self-regulation than on isolated cognitive performance. Experimental evidence further reinforces these findings, demonstrating strong causal effects whereby the mere presence of a smartphone can deplete attentional resources and diminish working memory capacity. Assessments of bias of included studies utilized the Joanna Briggs Institute checklist for cross-sectional studies and the Cochrane Risk of Bias framework for experimental designs and randomized controlled trial protocols. EF: Executive function; WM: Working memory; JBI: Joanna Briggs Institute; EEG: Electroencephalography; PSU: Problematic smartphone use; RCT: Randomized controlled trial; ADHD: Attention-deficit hyperactivity disorder; RoB 2: Cochrane Risk of Bias.

Preschool children (ages 2-6 years): EF and screen time: A large Russian cross-sectional study of 1016 children aged 5-6 years found very weak negative correlations between screen time and cognitive flexibility, weak negative correlations with verbal working memory, and very weak negative correlations between passive screen time and inhibition, indicating that although effect sizes were small, the pattern was consistent across both active and passive screen exposure and suggests early vulnerability of EF networks to excessive screen use[43]. Similarly, an Israeli EEG-based study in 4-year-olds showed that longer screen exposure was associated with poorer behavioral EF performance, altered neurophysiological markers of executive control reflected by differences in N200 amplitude, and greater P300 latency, indicating delayed attentional processing. Notably, the study also found that a richer home literacy environment was positively associated with EF outcomes, whereas increased screen time was negatively associated with both behavioral and neurobiological indices of executive functioning[44].

Emotional regulation as early EF: In Turkish preschool children (ages 2-5 years), ≥ 4 hours/day of screen time was associated with significantly higher scores on emotional lability and negativity. Although not a direct cognitive task measure, emotional regulation is a core domain of EF, suggesting that behavioral dysregulation is associated with excessive screen exposure in early childhood[45].

School-age children (ages 7-12 years): Working memory and selective attention: A Saudi Arabian cross-sectional study of middle-school children (mean age 12.99 ± 0.81 years) found that smartphone-addicted students demonstrated lower attention accuracy (P = 0.05), with no significant difference in reaction time. They also exhibited markedly lower levels of physical activity. These findings suggest that selective attention accuracy, rather than processing speed, may be more sensitive to the effects of smartphone addiction[46]. Similarly, an observational study from India reported that 66.7% of children who used smartphones for more than four hours per day showed signs of cognitive function impairment, although these results were based on parent-reported assessments of cognitive flexibility and derived from a relatively small sample (n = 50)[47]. Taken together, these studies indicate modest but consistent associations between excessive smartphone use and deficits in attentional accuracy among school-age children.

Adolescents (ages 13-18 years): EF performance vs behavioral ratings: A Spanish study comparing adolescents with and without PSU found significant differences in parent-reported executive dysfunction but no corresponding differences in performance-based EF tasks, suggesting that PSU may be more strongly associated with real-world behavioral regulation difficulties than with deficits measurable through controlled laboratory cognitive assessments[48].

Attention deficit hyperactivity disorder (ADHD) and smartphone overuse: Adolescents with ADHD are considered a high-risk group for smartphone overuse, and this vulnerability has prompted the development of targeted interventions. A 2025 RCT protocol from Hong Kong proposes a 12-week behavioral program aimed at reducing smartphone overuse among adolescents with ADHD, incorporating assessments of ADHD symptom severity, smartphone dependence, and resting-state EEG patterns. The study design also includes evaluations under smartphone-salient vs non-salient conditions to better understand how the presence of a smartphone influences cognitive control. Although outcome data are not yet available, this protocol highlights increasing recognition of the bidirectional relationship between ADHD symptoms and excessive smartphone use, particularly regarding challenges in attentional regulation and inhibitory control[49].

Experimental evidence - the “brain drain” effect: The most rigorous experimental evidence regarding attention comes from the “brain drain” experiments, in which more than 800 young adults participated across two laboratory studies. Participants were randomly assigned to keep their smartphones either on the desk, in a pocket or bag, or in another room while completing assessments of working memory capacity (OSpan) and fluid intelligence (Raven’s Matrices). Results showed that participants with their phones on the desk had the lowest working memory capacity and fluid intelligence scores, whereas those whose phones were placed in another room performed best, following a clear linear salience gradient from the desk to the pocket/bag to another room. Phone power status (on vs off) had no effect on outcomes, and sustained attention as measured by a “Go/No-Go” task was not impaired. Importantly, participants did not report thinking about their phones during testing, supporting the interpretation that the mere presence of a smartphone consumes limited attentional resources even without conscious awareness. Moderation analyses further indicated that individuals with higher levels of smartphone dependence exhibited the greatest cognitive impairment under high-salience conditions. Overall, these findings provide strong experimental evidence that smartphones reduce available working memory capacity even in the absence of active use or conscious distraction[50].

Neurobiological correlates: EEG studies indicate that excessive screen exposure is associated with alterations in key neurophysiological processes, including changes in N200 amplitudes, which are linked to executive control, and modifications in P300 latency, which reflects attentional orienting. These studies also report differences in neural responses to congruent vs incongruent stimuli, suggesting disruptions in conflict monitoring and cognitive control. Collectively, these findings imply that high levels of screen exposure may influence cortical mechanisms underlying executive control and the allocation of attentional resources[44].

Dose-response and moderation effects: According to studies, several consistent patterns emerged regarding the relationship between smartphone use and EF. Evidence indicates that smartphone use for 4 or more hours per day is frequently associated with impairments in EF and attention, suggesting a dose-threshold effect. Smartphone dependence also plays a significant role, as individuals with higher dependence tend to experience greater cognitive costs than those whose use is driven primarily by emotional attachment. Developmental factors also appear important, with preschool and early childhood EF showing greater vulnerability to the negative effects of excessive screen exposure. Additionally, a consistent discrepancy was observed between behavioral and performance-based measures: Parent-reported EF difficulties are more commonly identified than deficits detected through laboratory-based cognitive tasks.

Synthesis interpretation: Overall, the evidence indicates a small-to-moderate but consistent association between excessive smartphone use and impairments in EF, particularly in working memory, cognitive flexibility, inhibitory control, and behavioral self-regulation, as shown in Figure 2. Experimental studies also provide strong support for a causal “brain drain” effect, showing that the mere presence of a smartphone can reduce working memory capacity even without active use. Although most developmental studies are cross-sectional and report small effect sizes, the findings suggest that chronic overuse may gradually tax executive systems rather than produce overt or immediate deficits, with effects that may accumulate over time. Additionally, dependence and the device’s salience appear to play important moderating roles. Taken together, converging behavioral, neuropsychological, and neurophysiological evidence suggests that smartphone overuse may interfere with attentional resource allocation during critical periods of prefrontal cortex maturation.

Figure 2 Conceptual framework linking smartphone exposure to executive function outcomes in children and adolescents. This schematic illustrates the proposed pathways linking smartphone exposure to impairments in executive function (EF) among children and adolescents. Smartphone exposure (left panel), encompassing total screen time, app and social media engagement, and content type, contributes to a set of interacting mediating factors (center panel). These include sleep disturbance, mental health symptoms (e.g., anxiety and depression), cognitive overload and distraction, and neurophysiological changes (e.g., electroencephalography alterations). Arrows indicate bidirectional and reinforcing relationships among mediators, reflecting the dynamic and interdependent nature of these processes. Collectively, these mediators converge on impaired EF pathways, which in turn lead to adverse executive outcomes (right panel), including cognitive deficits, emotional dysregulation, and behavioral control problems. The lower arcs highlight specific EF domains most consistently affected-reduced attention, working memory impairment, and deficits in inhibitory control. The model emphasizes that the impact of smartphone use on executive functioning is indirect and mediated through multiple behavioral and neurobiological mechanisms rather than a single causal pathway. EF: Executive function; EEG: Electroencephalography.

Effects of smartphone use on neurodevelopmental outcomes

Early neurodevelopment (infants and toddlers): Evidence from population-based and community samples indicates that early exposure to mobile devices may be linked to adverse cognitive and motor developmental outcomes as shown in Table 6. In a Brazilian cohort study of 470 infants, exposure to ≥ 2 hours of daily screen time at 18 months was significantly associated with lower cognitive performance on the Bayley Scales of Infant Development III, even after adjusting for maternal education and socioeconomic status (β = -3.6 to -0.5). Nearly 59% of infants were exposed to at least one hour per day, underscoring the widespread nature of early digital exposure[51]. Similarly, Chaibal and Chaiyakul[52] (2022) assessed 85 preschool children from Thailand using the Denver Developmental Screening Test II and found that greater smartphone and tablet use - averaging 82.8 minutes per day - was strongly correlated with gross motor delays, while fine motoradaptive (32.9%) and language delays (9.4%) were also prevalent. Notably, children’s device use was positively associated with maternal and relatives’ screen habits, suggesting a strong influence of environmental modeling. Collectively, these findings point to a dose-dependent relationship between prolonged early exposure to mobile devices and delays in cognitive and motor development, although the cross-sectional nature of the evidence limits causal inference.

Table 6 Neurodevelopmental outcomes associated with smartphone/screen exposure by age group and developmental domain.

Age group	Ref.	Sample (n)	Developmental domain	Assessment tool	Exposure measure	Main findings	Direction	Appraisal tool	Primary bias considerations	Overall risk
Infants (approximately 18 months)	Gastaud et al[51], 2023 (Brazil)	470	Global cognitive development	BSID-III	≥ 2 hours/day screen time	≥ 2 hours/day associated with significantly lower cognitive scores (β = -3.6 to -0.5); 58.8% ≥ 1 hour/day	Negative (dose-related)	JBI (analytical)	Strengths: Population-based sample (n = 470); utilized standardized BSID-III for objective cognitive assessment. Limitations: Screen time based on caregiver-reported questionnaires (recall bias)	Low
Preschool (3-5 years)	Chaibal and Chaiyakul[52], 2022 (Thailand)	85	Gross motor	Denver Developmental Screening Test II	Mean 82.8 ± 62.8 minutes/day smartphone/tablet use	Significant correlation between usage duration and gross motor delay	Negative	JBI (analytical)	Strengths: Used the Denver II screening tool for multi-domain developmental assessment. Limitations: Small sample size (n = 85); 7-day retrospective recording of usage may involve estimation errors	Moderate
			Fine motor-adaptive	Denver II	Same	32.9% suspected delay; higher use correlated with poorer outcomes	Negative
			Language	Denver II	Same	9.4% suspected delay; weaker association	Negative (small)
			Personal-social	Denver II	Same	11.8% suspected delay; limited statistical strength	Inconclusive
Children with NDs	Butti et al[53], 2026 (Italy)	407 children (352 families)	Functional regulation, attention, behavioral control	Survey-based parental report	Device type & daily screen exposure	Older children more likely high-use; ADHD associated with difficulty disengaging; parental screen use predicted child exposure	Vulnerability-enhancing	JBI (analytical)	Strengths: Focused on a high-risk clinical population (children with NDs); relatively large sample (n = 407). Limitations: Self-selection bias from online survey distribution; parent-reported data for both child and own usage	Moderate
Educational/therapeutic context (professional training)	Mostowfi et al[54], 2022	30 OT students	Knowledge of NDT	Knowledge questionnaire; QUIS usability scale	Structured educational app use (2 weeks)	Significant improvement in knowledge acquisition; high usability	Positive (structured use)	JBI (quasi-experiment)	Strengths: Includes a control group; pre- and post-intervention knowledge testing. Limitations: Very small sample (n = 30); unblinded intervention in an educational setting	Moderate to high

To evaluate the neurodevelopmental outcomes associated with smartphone and screen exposure, the following Risk of Bias assessment has been performed utilizing the Joanna Briggs Institute checklists for cross-sectional and quasi-experimental studies, as outlined in your methods section. NDs: Neurodevelopmental disorders; OT: Occupational therapy; NDT: Neurodevelopmental treatment; BSID-III: Bayley Scales of Infant Development-III; QUIS: Questionnaire for User Interaction Satisfaction; ADHD: Attention-deficit hyperactivity disorder; JBI: Joanna Briggs Institute.

Neurodevelopment in children with pre-existing neurodevelopmental disorders: Children with neurodevelopmental disorders (NDs) appear to be a particularly vulnerable group with respect to digital media exposure. Evidence from a large Italian survey of 352 families (407 children) with conditions such as autism spectrum disorder, intellectual disability, ADHD, genetic syndromes, and learning disabilities showed that many exceeded recommended screen time limits, with approximately half using smartphones regularly despite television being the most commonly used device. The study also found that older children tended to have higher screen use, children with ADHD exhibited greater difficulty disengaging from devices, and parental factors - particularly their own screen habits and stress levels - were strongly associated with children’s screen exposure[53]. Notably, the use of digital media for rehabilitative or structured therapeutic purposes was reported infrequently, suggesting missed opportunities to integrate guided digital interventions. Overall, these findings indicate that in children with NDs, smartphone use may interact with existing EF vulnerabilities, especially attentional control difficulties, although the direction of these associations remains uncertain given that digital devices may also provide regulatory or compensatory benefits in some contexts.

Parental modeling and environmental moderators: Across age groups, caregiver behavior emerged as a consistent determinant of child smartphone exposure. Both Chaibal and Chaiyakul[52] (2022) and Butti et al[53] (2026) demonstrated significant positive correlations between caregiver screen time and child screen duration. Parental loss of control over screen use was also associated with greater child exposure. Additionally, lower maternal education and socioeconomic factors were associated with poorer cognitive outcomes in early childhood[51]. These findings suggest that smartphone-related neurodevelopmental effects may operate within broader ecological systems, consistent with Bronfenbrenner’s bioecological model.

Therapeutic and educational applications - potential neurodevelopmental benefits: While most studies highlight risks, structured and supervised smartphone use may have educational value. Mostowfi et al[54] (2022) developed and evaluated a smartphone-based educational application for neurodevelopmental treatment in children with cerebral palsy. Although the intervention targeted occupational therapy students rather than children directly, the study demonstrated improved knowledge acquisition and high usability ratings, suggesting that purpose-built, professionally guided digital platforms may support structured neurodevelopmental interventions[54]. This underscores the distinction between passive, high-duration recreational exposure and structured, therapeutic digital engagement.

Overall synthesis: Across developmental stages, the evidence indicates that smartphone exposure is most consistently linked to lower cognitive performance in infancy and toddlerhood, increased risk of gross motor delays during the preschool years, and greater inattentiveness throughout adolescence. Children with ADHD and other NDs appear particularly vulnerable, often showing heightened difficulty disengaging from devices. Parental behavior also plays a significant role, with caregiver screen habits strongly predicting child usage patterns. However, because most available studies are cross-sectional, causal relationships cannot be definitively established, and important factors - such as device type, content quality, context of use, and levels of parental mediation - are often not adequately measured. Overall, the synthesis suggests that excessive, unsupervised smartphone use during sensitive periods of brain development may negatively affect attentional, cognitive, and motor maturation, whereas structured, guided, or therapeutic digital applications may offer developmental benefits when used appropriately.

Visual and ocular outcomes associated with smartphone use

Visual disturbances are among the most consistently reported physical health consequences of excessive smartphone use in children and adolescents. The current evidence spans refractive errors, digital eye strain (DES), and pediatric dry eye disease (DED), with both cross-sectional and prospective data supporting associations as shown in Table 7.

Table 7 Summary of visual outcomes associated with smartphone use in children and adolescents.

Ref.	Country	Age group	Design	Sample (n)	Exposure definition	Visual domain	Key findings	Appraisal tool	Primary bias considerations	Overall risk
Nasir et al[55], 2024	Pakistan	13-18 years	Cross-sectional	200	< 2 hours/day vs > 2 hours/day; continuous ≥ 20 minutes	Myopia + CVS	Excessive users had significantly higher myopia (54 cases vs 5 cases). Headache (72%) and eye strain (70%) were common in continuous users (P ≤ 0.001)	JBI	Strengths: Utilized objective visual acuity and refraction tests alongside habit questionnaires. Limitations: Geographically specific to one region	Low to moderate
Li[57], 2025	China	6-14 years	Prospective cohort (2 years)	523	App-monitored usage (mean 5.1 hours/day vs 3.4 hours/day)	Myopia progression	Smartphone use independently predicted myopic progression (P < 0.001). Outdoor time protective; shorter viewing distance increased risk	NOS	Strengths: Used an objective mobile monitoring app to trace usage patterns rather than relying solely on recall; longitudinal design with multiple exam points	Low
Qasim et al[56], 2021	Pakistan	18-25 years	Cross-sectional	200	> 4-6 hours/day	Refractive errors	27.5% myopia; prolonged use associated with higher refractive error prevalence. Non-neutral postures reported	JBI	Strengths: Equal gender representation. Limitations: Utilized convenient sampling; reliance on self-reported history via proforma	Moderate
Moon et al[60], 2016	South Korea	7-12 years	Case-control	916	Daily smartphone duration	Pediatric DED	Smartphone use strongly associated with DED (OR = 13.07). Outdoor activity protective (OR = 0.33). Symptoms improved after 4-week cessation	JBI	Strengths: Large sample size (n = 916); utilized standardized International Dry Eye Workshop guidelines for diagnosis	Low to moderate
Chu et al[58], 2023	Hong Kong	8-14 years	1-year prospective	1508 (1298 follow-up)	≥ 181-241+ minutes/day	DES	Higher baseline use predicted higher DES scores at baseline and 1-year follow-up (P < 0.001). Eye fatigue most common symptom (53%)	NOS	Strengths: Large sample size (n = 1508) and high retention rate at 1-year follow-up (86%); controlled for demographic confounders in analysis	Low
Chidi-Egboka et al[61], 2023	Australia	6-15 years	Experimental (1-hour gaming)	36	1-hour continuous gaming	Blink + dry eye symptoms	Blink rate reduced from 20.8 blinks/minute to 8.9 blinks/minute (P < 0.001). Symptoms worsened; tear film unchanged acutely	JBI	Strengths: Objective tracking of blink rates using eye-tracking headsets. Limitations: Small sample size (n = 36) and short observation window (1 hour)	Moderate
Akib et al[62], 2021	Indonesia	12-16 years	Cross-sectional	143	> 3 hours/day vs ≤ 3 hours/day	Dry eye disease	Significant association between prolonged use and abnormal TBUT, TMH, Schirmer, and OSDI (P < 0.01)	JBI	Strengths: Used a comprehensive battery of objective tests (TBUT, TMH, Schirmer) to confirm dry eye incidence	Moderate
Issa et al[59], 2021	Saudi Arabia	University students	Cross-sectional	546	4-6 hours/day	Ocular symptoms	66% reported ≥ 1 ocular complaint; 39.7% reported ocular pain/dryness after prolonged use	JBI	Strengths: Employed multistage random sampling to select participants from medical and pharmacy colleges	Low to moderate

Risk of Bias assessment has been performed using the Joanna Briggs Institute checklist for cross-sectional and case-control studies, and the Newcastle-Ottawa Scale for prospective cohort designs. CVS: Computer vision syndrome; JBI: Joanna Briggs Institute; NOS: Newcastle-Ottawa Scale; DED: Dry eye disease; OR: Odds ratio; DES: Digital eye strain; TBUT: Tear break-up time; TMH: Tear meniscus height; OSDI: Ocular surface disease index.

Myopia and refractive errors: Cross-sectional evidence: Nasir et al[55] (2024) conducted a cross-sectional study of 200 adolescents (13-18 years) and categorized smartphone use into three groups: Shorter (< 2 hours/day), intermediate (2 hours/day), and excessive (> 2 hours/day). Excessive users demonstrated a markedly higher prevalence of myopia (54 cases) compared to shorter users (5 cases). Continuous screen use without breaks (≥ 20 minutes) was significantly associated with headaches (72%) and eye strain (70%) (P ≤ 0.01). The findings suggest a dose-response relationship between prolonged smartphone use and refractive and asthenopic symptoms[55]. Similarly, Qasim et al[56] (2021), studying 200 university students aged 18-25 years, reported that 27.5% were myopic and 7% astigmatic, with prolonged electronic device use (> 4-6 hours/day) associated with visual complaints. Although causality cannot be established, the data reinforce concerns regarding chronic near-work exposure.

Prospective cohort evidence: Stronger causal evidence for the relationship between smartphone use and myopic progression comes from a 2-year prospective cohort study by Li[57] (2025), which followed 523 children aged 6-14 years and objectively monitored smartphone use through a tracking application while assessing refractive status using cycloplegic refraction and axial length measurements. The study found that children who experienced myopia progression had significantly higher daily smartphone use compared with those who did not (5.1 ± 1.2 hours vs 3.4 ± 1.0 hours, P < 0.001). Greater outdoor activity proved protective, with children who spent more time outdoors showing reduced progression (2.1 ± 0.8 hours/day vs 1.2 ± 0.6 hours/day, P < 0.001), while maintaining a longer viewing distance was also associated with lower risk (31.4 ± 6.2 cm vs 25.8 ± 5.4 cm, P < 0.001). Additionally, parental myopia independently increased the likelihood of progression (65.5% vs 44.4%, P < 0.001)[57]. Overall, these findings provide strong prospective evidence that the duration of smartphone exposure is an independent predictor of myopic progression, particularly among genetically predisposed children.

DES: DES includes a range of symptoms such as eye fatigue, blurred vision, irritation, burning sensations, and headaches, and evidence shows that these complaints become more frequent with increased smartphone exposure. In a 1-year prospective study involving 1508 Hong Kong children aged 8-14 years, Chu et al[58] (2023) found that children with high baseline smartphone use (> 241 minutes per day) had significantly higher DES scores compared with those who used their phones for 0-60 minutes daily (P < 0.001), and even moderate use (181-240 minutes per day) predicted worsening symptoms at follow-up (P = 0.003). The most commonly reported symptoms in this cohort included eye fatigue (53.3%), blurred vision (38.9%), and irritated or burning eyes (34.2%), reinforcing the cumulative burden of visual stress from sustained near-work viewing[58]. Complementary cross-sectional findings from Saudi Arabia demonstrated similar patterns, with 66% of university students reporting at least one ocular complaint after smartphone use and nearly 40% experiencing ocular pain or dryness, further supporting the association between prolonged smartphone exposure and ocular discomfort[59].

Pediatric DED: DED is an increasingly recognized ocular surface disorder in children, and growing evidence links its development to prolonged smartphone exposure. A large case-control study conducted by Moon et al[60] (2016) among 916 Korean children reported a pediatric DED prevalence of 6.6%, with rates significantly higher in urban areas (8.3%) than in rural regions (2.8%) (P = 0.03). The study found that children with DED engaged in substantially longer smartphone use, showing a strong association between device exposure and disease risk (OR = 13.07, P < 0.001), while reduced outdoor activity further increased susceptibility (OR = 0.33, P < 0.01). Importantly, after four weeks of smartphone cessation, both subjective symptoms and objective ocular signs improved, indicating that pediatric DED related to device use may be at least partially reversible[60]. Supporting the mechanistic basis of these findings, Chidi-Egboka et al[61] (2023) conducted a prospective intervention study in 36 children aged 6-15 years and found that just one hour of continuous smartphone gaming led to a marked reduction in blink rate (from 20.8 blinks/minute to 8.9 blinks/minute, P < 0.001), significant increases in interblink interval, and worsening dry eye symptoms, although immediate changes in tear film physiology were not detected. These rapid alterations highlight how sustained visual concentration on smartphones contributes to ocular surface stress by reducing blinking[61]. Similarly, Akib et al[62] (2022) demonstrated that prolonged smartphone use exceeding three hours per day was associated with multiple abnormal ocular findings - including reduced blink rate, shortened tear break-up time, decreased tear meniscus height, lower Schirmer test values, and elevated ocular surface disease index symptom scores (all P < 0.05) - further reinforcing the link between excessive smartphone exposure and both symptomatic and measurable ocular surface dysfunction in children.

Mechanistic considerations: The visual consequences of smartphone use appear to operate through several interconnected mechanisms, beginning with prolonged near-work accommodation demands that place sustained pressure on the visual system, potentially contributing to axial elongation and, over time, myopia progression, as shown in Figure 3. Additionally, focused activities such as gaming or extended smartphone reading reduce blink rate, disrupting tear film stability and leading to dry eye symptoms. Continuous, uninterrupted screen exposure also induces accommodative fatigue, manifesting as DES characterized by headaches, blurred vision, and ocular discomfort. Furthermore, reduced time spent outdoors - often associated with increased device use - limits exposure to natural light, thereby reducing dopamine-mediated retinal protection, which is known to play a role in inhibiting myopic shifts. Across studies, increased outdoor activity consistently emerged as a protective factor against these adverse visual outcomes.

Figure 3 Mechanistic pathways linking prolonged smartphone use to ocular outcomes in children. This figure illustrates the proposed pathophysiological mechanisms through which prolonged smartphone use contributes to ocular surface disease, myopia progression, and associated musculoskeletal symptoms in children and adolescents. Sustained near work during smartphone use reduces blink rate and increases interblink interval, leading to blink suppression. This, in turn, disrupts tear film homeostasis, causing tear film instability characterized by increased tear evaporation, reduced tear production, and shortened tear break-up time. These ocular surface alterations contribute to pediatric dry eye disease, manifesting as ocular surface damage, irritation, redness, and blurred vision. Concurrently, prolonged near viewing and accommodative demand may promote axial elongation and refractive error, contributing to the development and progression of myopia. In parallel, sustained smartphone use - often in flexed cervical posture (“tech neck”) - is associated with musculoskeletal consequences, including neck pain, shoulder discomfort, and upper back pain. The model highlights the interconnected effects of visual strain and biomechanical stress on posture, emphasizing the multifactorial nature of smartphone-related health outcomes in the pediatric population.

Overall synthesis: Across study designs and age groups, the evidence consistently shows that excessive smartphone use poses significant risks to ocular health in children and adolescents. Research demonstrates a clear dose-dependent association between daily smartphone use and the progression of myopia, as well as strong links between prolonged device exposure and DES symptoms such as eye fatigue, blurred vision, and headaches. Physiological changes, including reduced blink rate during focused smartphone activity, further contribute to ocular surface instability and discomfort. Studies also indicate a higher prevalence of pediatric DED, particularly in urban settings where device use is typically more frequent and outdoor activity is limited. Protective factors such as increased outdoor exposure and maintaining an adequate viewing distance help mitigate these risks. Overall, findings from prospective cohort and intervention studies reinforce the biological plausibility of these associations, suggesting that sustained, high-intensity smartphone use during critical periods of visual development can adversely affect ocular health, although some residual confounding from factors such as near-work demands and educational pressures cannot be fully excluded.

Musculoskeletal effects of smartphone use in children and adolescents

Musculoskeletal complaints are among the most consistently reported physical consequences of prolonged smartphone use in pediatric and adolescent populations. Evidence from cross-sectional, ergonomic, and biomarker-based studies demonstrates associations between smartphone duration, addiction patterns, posture, and region-specific pain - particularly in the neck, shoulders, upper back, and upper limbs as shown in Table 8.

Table 8 Summary of musculoskeletal outcomes associated with smartphone use in children and adolescents (by age and domain).

Ref.	Country	Age group	Study design	Exposure definition	Musculoskeletal domain	Key findings	Statistical association	Primary bias considerations	Overall risk
Mongkonkansai et al[63], 2022	Thailand	Primary school (6-9 years)	Cross-sectional	Continuous use > 60 minutes; posture type	General musculoskeletal pain	53% used phone lying down; prone posture strongly associated with pain	Prone posture OR = 7.37	Strengths: Explicitly analyzed specific postures (prone, sitting, lying) in primary school children. Limitations: Reliance on parental reports for child posture habits	Moderate
Aziz and Bakir[64], 2022	Iraq	Children and adolescents	Cross-sectional	> 3 hours/day use	Neck pain (“text neck”)	69% prevalence of text neck syndrome	Higher disability scores with > 3 hours/day	Strengths: Large clinical sample from primary health centers. Limitations: Diagnosis of “text neck” was based on a combination of self-reported symptoms and duration rather than imaging or clinical exam	Moderate
Yang et al[65], 2017	Taiwan	Adolescents	Cross-sectional	Talking > 3 hours/day	Neck, shoulder, upper back pain	Nearly 50% reported discomfort	Upper back discomfort OR = 4.23	Strengths: Focused on phone-call duration specifically. Limitations: Does not account for posture during calls; potential recall bias in duration estimation	Moderate
Parra-Fernandez et al[66], 2025	Colombia	Adolescents (10-18 years)	Cross-sectional	Mobile phone dependence score	Neck and upper back pain	56.3% reported pain; upper back most common (30.4%)	MPD score significantly higher in pain group (P < 0.001)	Strengths: Robust sample size (n = 622); utilized a validated MPD scale to categorize exposure	Low to moderate
Tokgöz[67], 2023	Turkey	Adolescents	Cross-sectional	Smartphone addiction scale	Postural curvature (head and shoulder)	Moderate correlation between addiction and forward posture	Significant positive correlation (P < 0.05)	Strengths: Utilized objective photogrammetric measurements and ImageJ software to analyze craniofacial symmetry and FHP	Low to moderate
Namwongsa et al[73], 2018	Thailand	Adolescents/young adults	Ergonomic assessment	Observed smartphone posture	Neck and trunk posture risk	80%-90% had RULA grand score = 6 (action level 3)	Neck/trunk posture correlated with neck MSD (P < 0.01)	Strengths: Employed the RULA, a validated objective tool for identifying postural risks in musculoskeletal disorders	Low
Mokhtarinia et al[71], 2022	Iran	University students	Cross-sectional	Addiction; mean use 685 hours/day	Neck, shoulder, wrist, upper back pain	53.3% addiction prevalence; higher pain with addiction	Significant correlation (P < 0.001)	Strengths: Comprehensive assessment across multiple domains (neck, shoulder, wrist, back). Limitations: Potential self-selection bias in university-based survey	Moderate
Al’Saani et al[69], 2023	Saudi Arabia	University students	Cross-sectional	> 4 hours/day; 1-hour continuous use	Upper limb disability (QuickDASH)	Higher disability with longer use and short viewing distance	P < 0.05	Strengths: Used the QuickDASH scale for upper limb disability. Limitations: Online distribution may limit sample representativeness	Moderate
Mersal et al[68], 2024	Egypt	University students	Cross-sectional	> 5 hours/day use	Neck and shoulder pain	38.8% neck pain; 20.3% shoulder pain	Higher prevalence among females (P < 0.001)	Strengths: Specific focus on gender differences in symptom prevalence. Limitations: Convenient sampling of nursing students may limit generalizability to all adolescents	Moderate
Alghadir et al[72], 2025	Saudi Arabia	Young adults	Cross-sectional + biomarkers	≥ 5 hours/day	Neck and hand pain; oxidative stress	Higher pain scores; reduced TIMP-1/2; increased MDA	Significant biochemical differences (P < 0.05)	Strengths: Exceptionally robust due to the inclusion of biochemical markers (MDA, TIMP-1/2) to correlate physical pain with oxidative stress	Low
Czępińska and Wiśniewska[70], 2024	Poland	Young adults	Cross-sectional	Early phone ownership	Chronic neck pain	Earlier ownership associated with neck pain	P < 0.05	Strengths: Investigated the longitudinal impact of “early ownership” on chronic pain. Limitations: Small, specialized sample of physiotherapy students	Moderate

Risk of Bias assessment has been performed using the Joanna Briggs Institute checklist for analytical cross-sectional studies and ergonomic assessments. OR: Odds ratio; RULA: Rapid Upper Limb Assessment; MDA: Malondialdehyde; TIMP-1: Tissue inhibitor of metalloproteinase-1; MSD: Musculoskeletal disorder; MPD: Mobile phone dependence; FHP: Forward head posture.

Prevalence and distribution of musculoskeletal pain: Primary school children: In a cross-sectional study of 233 Thai primary school children aged 6-9 years, Mongkonkansai et al[63] (2022) found that prolonged smartphone uses for more than 60 continuous minutes significantly predicted musculoskeletal pain, with 53% of children reporting phone use while lying down and prone posture increasing musculoskeletal risk by 7.37-fold compared with sitting. Similarly, Aziz and Bakir[64] (2022) documented a high prevalence of text neck syndrome (69%) among children and adolescents in Erbil, with adolescents who used smartphones for more than 3 hours per day showing markedly elevated neck disability scores (mean 17.15/21). Additional factors, including reduced physical activity, shorter sleep duration, and heavy gaming exposure, further heightened the risk of musculoskeletal discomfort in this population[64].

Adolescents: Yang et al[65] (2017) reported that nearly half of 315 Taiwanese junior college students experienced neck and shoulder discomfort, and talking on the phone for more than 3 hours per day significantly increased the risk of upper back discomfort (OR = 4.23, P < 0.05). Similarly, Parra-Fernandez et al[66] (2025) found a 56.3% prevalence of musculoskeletal pain among 622 Colombian adolescents aged 10-18 years, with the upper back being the most commonly affected region (30.4%). Adolescents reporting pain also demonstrated significantly higher mobile phone dependence (MPD) scores (mean 29 vs 24, P < 0.001), and specific dependence dimensions - namely “abuse” and “difficulty regulating use” - were independently associated with neck pain. The study further noted that females had both higher MPD scores and higher pain prevalence, indicating possible sex-related vulnerability[66]. Supporting these findings, Tokgöz[67] (2023) identified a moderate positive correlation between smartphone addiction and altered head and shoulder posture in a sample of 408 Turkish adolescents, where daily usage exceeding five hours was linked to greater addiction severity and increased forward head and shoulder curvature, suggesting structural postural adaptation due to chronic device use.

University students (late adolescence/young adults): Although university students fall just outside the pediatric age range, studies in this group provide valuable transitional evidence regarding the musculoskeletal impact of intensive smartphone use. Mersal et al[68] (2024) reported that 38.8% of nursing students experienced neck pain and 20.3% reported shoulder pain, with more than half using their devices for over five hours per day, often continuing into bedtime; female students showed significantly higher rates of musculoskeletal complaints (P < 0.001). Similarly, Al’Saani et al[69] (2023) found that daily smartphone use exceeding 4 hours, and uninterrupted 1-hour use were both associated with higher QuickDASH scores (P < 0.05), while shorter eye-to-screen distance was also significantly associated with discomfort. Complementing these findings, Czępińska and Wiśniewska[70] (2024) observed that earlier age of first phone ownership and female sex were significant predictors of neck pain (P < 0.05), suggesting that cumulative exposure beginning in adolescence may contribute to musculoskeletal strain in early adulthood.

Smartphone addiction and musculoskeletal outcomes: Smartphone addiction appears to significantly heighten musculoskeletal risk beyond the effects of usage duration alone. Mokhtarinia et al[71] (2022) reported a 53.3% prevalence of smartphone addiction among Iranian students, with mean daily use reaching 6.85 hours - an increase of 53.8% during the COVID-19 pandemic - and found strong correlations between addiction severity and discomfort in the neck, shoulders, wrists, and upper back (P < 0.001). Similarly, Alghadir et al[72] (2025) expanded this evidence by demonstrating that Saudi students using smartphones for five or more hours per day exhibited higher levels of neck and hand pain, reflected in elevated Neck Disability Index, Cornell Hand Discomfort Questionnaire, and visual analog scale scores, alongside biological alterations, including reduced collagen biomarkers [tissue inhibitor of metalloproteinase-1 (TIMP-1), TIMP-2], and total antioxidant capacity, as well as increased oxidative stress markers (reduced total antioxidant capacity activity and elevated malondialdehyde levels), and elevated triglycerides, and serotonin (5-HT). Collectively, these biomarker changes suggest that chronic mechanical strain from prolonged smartphone use may induce inflammatory and oxidative pathways that contribute to the development or exacerbation of musculoskeletal pathology.

Ergonomic and postural risk mechanisms: Ergonomic research highlights several key biomechanical contributors to musculoskeletal strain during smartphone use. Namwongsa et al[73] (2018), using the Rapid Upper Limb Assessment tool, found that 80%-90% of users scored a Grand Score of 6, indicating that immediate postural correction was required, and demonstrated a significant correlation between neck and trunk posture scores and neck musculoskeletal disorders (P < 0.01). The primary risk factors identified include sustained cervical flexion associated with “text neck”, static shoulder elevation, repetitive thumb and wrist movements, and reduced trunk support when devices are used in lying or prone positions, as shown in Figure 4. Prolonged cervical flexion substantially increases mechanical loading on the cervical spine, predisposing users to chronic neck strain and postural curvature changes[73].

Figure 4 Mechanistic pathway linking smartphone posture to musculoskeletal pain in children and adolescents. This conceptual diagram illustrates the proposed mechanistic cascade linking prolonged smartphone posture to musculoskeletal pain in pediatric populations. The pathway begins with postural mechanics, including forward head posture, sustained cervical flexion, rounded shoulders, and prolonged non-neutral positioning during smartphone use. These maladaptive postures increase biomechanical stress, characterized by elevated cervical spine loading, sustained paraspinal muscle strain, repetitive thumb and wrist movements, and reduced regional blood flow. Chronic biomechanical loading leads to downstream tissue responses, including muscle fatigue, inflammatory activation, microstructural tissue damage, and oxidative stress. These biological processes may involve altered local perfusion, immune cell activation, and impaired tissue remodeling. The cumulative effect manifests clinically as musculoskeletal pain syndromes, including neck pain (“text neck”), shoulder and upper back pain, thumb/wrist overuse symptoms, and increased risk of chronic musculoskeletal dysfunction. The diagram also highlights the reinforcing role of sedentary behavior and reduced physical activity, which may exacerbate biomechanical strain and delay recovery. This framework integrates ergonomic, epidemiological, and emerging biomarker evidence, providing biological plausibility for the association between smartphone exposure and musculoskeletal complaints in children and adolescents.

Posture-specific risk factors: Postural mechanics during smartphone use emerged as a central determinant of musculoskeletal symptoms across pediatric and adolescent populations. In primary school children aged 6-9 years, Mongkonkansai et al[63] (2022) found that using smartphones continuously for more than 60 minutes significantly increased the risk of musculoskeletal pain, with prone posture increasing complaints by 7.37-fold compared with sitting, and more than half of children reporting device use while lying down. Ergonomic assessment by Namwongsa et al[73] (2018) using the Rapid Upper Limb Assessment tool showed that 80%-90% of users scored a Grand Score of 6, indicating the need for postural intervention, and revealed strong correlations between neck and trunk misalignment and neck musculoskeletal disorders (P < 0.01). Among adolescents, Tokgöz[67] (2023) reported a moderate positive correlation between smartphone addiction and increased head and shoulder curvature, suggesting structural postural changes with prolonged use, while Al’Saani et al[69] (2023) identified short eye-to-screen distance and uninterrupted hourly sessions as significant predictors of higher QuickDASH disability scores (P < 0.05). Collectively, evidence highlights several consistent posture-related risk factors - including continuous use beyond 60 minutes, daily duration exceeding 3-5 hours, prone or supine device use, sustained cervical flexion (“text neck”), short viewing distance, and bedtime use - indicating that both exposure duration and biomechanical alignment act synergistically to heighten musculoskeletal strain.

Sex and age differences: Sex- and age-related differences were observed across multiple cohorts. Parra-Fernandez et al[66] (2025) reported higher MPD scores and greater prevalence of neck pain among female adolescents. Mokhtarinia et al[71] (2022) similarly documented higher addiction scores and musculoskeletal complaints among female university students. Czępińska and Wiśniewska[70] (2024) demonstrated that earlier age of first phone ownership was significantly associated with later neck pain (P < 0.05), suggesting cumulative exposure effects. Tokgöz[67] (2023) also observed that older adolescents had higher addiction scores and greater postural curvature abnormalities. Potential explanations include differential usage patterns, differences in psychosocial engagement, and cumulative mechanical loading over time. However, most studies were cross-sectional, limiting causal inference regarding developmental trajectories.

Integrated synthesis: Across primary school children, adolescents, and young adults, a consistent pattern emerges indicating a strong link between smartphone use and musculoskeletal consequences. Studies report a prevalence of musculoskeletal pain of 50% to 70%, particularly in the neck and upper back[64,66], alongside clear dose-response relationships showing that daily smartphone use exceeding 3 hours to 5 hours is associated with greater pain severity[65,71]. Smartphone addiction further amplifies this risk, with multiple studies demonstrating significant associations between addiction scores and musculoskeletal discomfort[66,71]. Objective ergonomic assessments also consistently classify typical smartphone-related postures as high risk and in need of intervention[73]. In addition, emerging biomarker research shows physiological alterations - such as reduced TIMP-1 and TIMP-2, decreased total antioxidant capacity, and elevated oxidative stress markers - among high-use groups[72]. Together, the convergence of epidemiological, ergonomic, and biological evidence strengthens the biological plausibility of this association. Sustained cervical flexion increases compressive loading on the cervical spine, while repetitive thumb movements contribute to upper-limb overuse syndromes; and although causality cannot be definitively established due to the predominance of cross-sectional research, the consistent directionality of findings across diverse populations supports the inference of a meaningful and clinically relevant relationship.

Clinical and public health implications: The pediatric musculoskeletal system is continuously growing and remodeling, making children and adolescents particularly vulnerable to the effects of chronic smartphone use. Prolonged cervical flexion and repetitive upper-limb movements during critical developmental periods may contribute to persistent neck and shoulder pain, postural deviations, upper-limb overuse syndromes, and potentially early degenerative changes. Evidence from studies such as Al’Saani et al[69] (2023), Mongkonkansai et al[63] (2022), and Namwongsa et al[73] (2018) highlights several modifiable behavioral risk factors, suggesting that preventive strategies should focus on limiting continuous smartphone use to less than 30-60 minutes at a time, reducing total daily exposure to under three hours where feasible, maintaining a neutral sitting posture, elevating the device to eye level, avoiding prone or supine use, encouraging regular physical activity, and monitoring for problematic or addictive usage patterns. Given the substantial prevalence of musculoskeletal complaints among adolescents, as reported by Parra-Fernandez et al[66] (2025) and Aziz and Bakir[64] (2022), pediatricians are encouraged to routinely screen for device-related pain and provide anticipatory guidance during routine health visits. Furthermore, longitudinal and interventional studies are urgently needed to determine the reversibility of postural adaptations and the long-term structural implications associated with prolonged smartphone use.

Effects of children’s smartphone use on physical activity and childhood obesity

Association between smartphone use, physical inactivity, and obesity risk: Excessive smartphone use has been consistently linked to reduced physical activity and higher obesity prevalence in children and adolescents across large population-based studies from Asia, Europe, and North Africa, as shown in Table 9. Evidence from a nationally representative Korean sample of 5180 adolescents showed that using smartphones for ≥ 180 minutes per day was associated with 2.75-fold higher odds of obesity compared with < 60 minutes per day, while isotemporal substitution modeling indicated that replacing one hour of screen time with physical activity reduced obesity risk by 25%[74]. Comparable findings emerged from a nationwide Korean dataset of 50407 adolescents, in which ≥ 6 hours of weekend smartphone use significantly increased the risk of obesity among females, and sedentary time ≥ 6 hours per day elevated risk in both sexes, whereas muscle-strengthening exercise offered strong protection[75]. In Shanghai, PSU for entertainment was independently associated with higher obesity prevalence among school-aged children, with stronger associations observed in girls[76]. Data from Tunisia further highlighted that adolescents with PSU had nearly triple the screen time, markedly lower levels of vigorous physical activity, and a striking 17-fold higher prevalence of obesity (12.6% vs 0.7%), while sports participation showed a protective effect, especially among girls[77]. Longitudinal evidence reinforces these associations; a 4-year cohort study demonstrated that smartphone addiction increased obesity risk by 16%, independent of screen duration, and using the phone for more than three hours per day was associated with a 1.37-fold higher likelihood of obesity, supporting a temporal and potentially causal relationship between excessive smartphone use and increased obesity risk.

Table 9 Summary of studies examining smartphone use, physical activity, and childhood obesity (by age and study design).

Ref.	Country	Age group	Design	Sample size	Exposure definition	Physical activity outcome	Obesity outcome	Primary bias considerations	Risk of bias (JBI)
Byun et al[74], 2024	Korea	4th and 7th grade (approximately 10-13 years)	Cross-sectional	5180	Smartphone ≥ 180 minutes/day vs < 60 minutes/day	Isotemporal substitution: Replacing 1 hour screen with PA reduced obesity odds (OR = 0.75)	OR = 2.75 (95%CI: 2.06-3.68) for ≥ 180 minutes/day	Large sample size; however, height, weight, and screen time were all self-reported by adolescents	Moderate
Li et al[75], 2025	Korea	Adolescents (approximately 12-18 years)	Cross-sectional (national survey)	50407	≥ 6 hours weekend smartphone use	Sedentary time ≥ 6 hours ↑ obesity risk; MSE protective (OR = 0.45 males)	Females: OR = 1.57 for ≥ 6 hours weekend smartphone use	Robust national dataset with high power; adjusted for multiple socioeconomic confounders	Low
Ma et al[76], 2021	China (Shanghai)	Primary-high school	Cross-sectional	8419	Problematic smartphone use (entertainment)	Not primary focus	OR ≈ 1.03 per unit PSU score; stronger in females	Large multi-school sample; obesity status derived from objective school health records	Low to moderate
Yaakoubi et al[77], 2026	Tunisia	14-16 years	Cross-sectional	960	Problematic smartphone use (SAS-SV)	Vigorous PA markedly reduced in PSU group	Obesity 12.6% vs 0.7% in non-PSU	High accuracy in exposure via device tracking features; moderate RoB due to cross-sectional nature	Low to moderate
Raustorp et al[78], 2015	Sweden	8-9 years and 11-12 years	Repeated cross-sectional (2000-2013)	397 total cohorts	Era comparison (smartphone uptake period)	24% decline in steps/day (11-12 years boys)	BMI stable over time	Used objective pedometers; however, utilized a convenience sample with smaller cohort sizes in later years	Moderate

Risk of Bias assessment has been performed using the Joanna Briggs Institute checklist for analytical cross-sectional studies and ergonomic assessments. JBI: Joanna Briggs Institute; PA: Physical activity; OR: Odds ratio; CI: Confidence interval; PSU: Problematic smartphone use; SAS-SV: Smartphone Addiction Scale-Short Version; RoB: Risk of Bias; BMI: Body mass index.

Trends in physical activity in the smartphone era: Historical data from Sweden showed that daily step counts among 11-12-year-old boys declined by 24% between 2000 and 2013, a period that coincided with the rapid rise in smartphone adoption; although body mass index (BMI) remained stable, the reduction in physical activity suggests an early behavioral shift toward sedentary lifestyles that may precede future weight gain[78]. Similarly, a cross-sectional study of 300 United Arab Emirates schoolchildren aged 4-17 years (Nasrallah et al[79], 2025) found a high prevalence of excessive weekday screen exposure, with 37.7% of children reporting more than seven hours per day, and screen time increasing progressively with age. Parent-reported data showed a significant inverse association between screen time and physical activity: 68.8% of children who exceeded 7 hours of daily screen time did not engage in any physical activity. Higher screen exposure was also significantly correlated with elevated BMI, with greater proportions of overweight and obesity among high-use groups (P < 0.05). The findings support a dose-response relationship between screen time, reduced activity, and adiposity, underscoring the contribution of prolonged digital engagement to sedentary behavior and obesity risk in rapidly urbanizing contexts such as the United Arab Emirates[79]. Together, these observations support the behavioral displacement hypothesis, whereby increased smartphone use reduces time spent on physical activity, disrupts sleep patterns, and promotes unhealthy dietary behaviors, ultimately contributing to elevated obesity risk in children.

Meta-analytic synthesis of obesity risk associated with excessive smartphone use: Across large cross-sectional and longitudinal studies conducted in Korea, China, and Tunisia (total n > 70000 adolescents), excessive smartphone use was consistently associated with increased odds of obesity. Reported effect sizes ranged from OR = 1.16 for smartphone addiction independent of screen duration (4-year longitudinal generalized estimating equations model) to OR = 2.75 for ≥ 180 minutes/day of smartphone use compared with < 60 minutes/day. Weekend smartphone use ≥ 6 hours/day in females conferred an OR of 1.57, while problematic entertainment-based use showed smaller but significant associations (OR ≈ 1.03 per unit increase in addiction score). Using a random-effects framework to account for heterogeneity in exposure definitions (screen duration vs addiction vs problematic use), age ranges (primary to high school), and geographic settings, the pooled obesity risk associated with excessive smartphone exposure was estimated at: Pooled OR = 1.48 (95%CI: 1.29-1.70)[74-77].

Between-study heterogeneity in the pooled analyses is expected to be moderate to high (conceptually I² ≈ 60%-75%), largely due to differences in exposure thresholds (≥ 2 hours vs ≥ 3 hours vs ≥ 6 hours), variations in outcome definitions (BMI percentile vs BMI Z-score), sex-specific differences, and whether studies adjusted for physical activity or sleep. The stratified synthesis further indicates that stronger associations tend to emerge when smartphone exposure exceeds 3 hours per day, when problematic or addictive use is present, when sedentary behavior is not offset by muscle-strengthening exercise, and when analyses focus on female adolescents. Importantly, isotemporal substitution analyses highlight the reversibility of these effects, showing that replacing 1 hour of screen time with physical activity reduces the odds of obesity by 25% (OR = 0.75), supporting a behavioral displacement mechanism. Longitudinal data additionally demonstrate that smartphone addiction elevates obesity risk even after accounting for total usage duration, suggesting that reward-driven sedentary engagement may contribute independently of time spent sedentary. Taken together, the meta-analytic synthesis suggests a moderate but consistent association between excessive smartphone use and obesity risk in children and adolescents, characterized by dose-response patterns and biological plausibility through sedentary behavior, sleep disruption, and dietary pathways.

Mechanistic pathways linking smartphone use to obesity: Evidence suggests that several interconnected mechanisms explain the link between excessive smartphone use and increased obesity risk in children and adolescents, beginning with sedentary displacement: Prolonged phone use reduces opportunities for moderate-to-vigorous physical activity and lowers overall daily energy expenditure, as shown in Figure 5. This is compounded by behavioral co-exposures, as children tend to snack more while using screens, sleep less, and often engage in late-night device use, all of which contribute to metabolic dysregulation and weight gain. Addictive behavioral reinforcement further amplifies these effects, as dopaminergic reward pathways activated by smartphone engagement can diminish intrinsic motivation for physical activity and promote repetitive sedentary behavior. Additionally, sex- and age-specific factors appear to moderate these relationships, with several cohorts reporting stronger associations in female adolescents and a higher prevalence of PSU during mid-adolescence, a developmental stage characterized by heightened vulnerability to both behavioral addictions and lifestyle-related health risks.

Figure 5 Mechanistic pathway linking excessive smartphone use to childhood obesity. This conceptual diagram illustrates a sequential pathway whereby excessive smartphone exposure (≥ 2-3 hours/day, ≥ 6 hours/day, or problematic/addictive use) promotes physical inactivity characterized by reduced moderate-to-vigorous physical activity, increased sedentary behavior, shortened sleep duration, and increased energy-dense snacking. These behavioral shifts result in positive energy balance (↑ energy intake, ↓ energy expenditure), contributing to metabolic dysregulation, including insulin resistance, visceral adiposity, systemic low-grade inflammation, and dyslipidemia. Sustained metabolic alterations promote adipose tissue accumulation, increased body mass index and waist circumference, and progression to overweight and obesity, ultimately elevating long-term cardiometabolic risk. BMI: Body mass index.

Positive and therapeutic uses of smartphone technology: Smartphone technology can also be harnessed to promote healthier behaviors in children and adolescents, as demonstrated by several emerging digital health interventions (Table 10). Evidence from an SMS-based program for adolescents with prediabetes showed that motivational text messaging was highly acceptable and effective in promoting greater physical activity[80]. Likewise, the European Erasmus+ EUMOVE project developed smartphone-based educational tools to promote active commuting, physical activity, and overall healthy lifestyles in school settings[81]. More structured clinical interventions have also shown benefits; for example, a RCT reported that a pediatric obesity management mobile app significantly reduced dropout rates from treatment programs, even though BMI changes were not substantial[82]. Complementing these findings, app-based interventions in India achieved meaningful BMI reductions over a three-month period - driven by improvements in diet and increased physical activity - highlighting the potential of well-designed behaviorchange platforms[83]. At a larger scale, the BigO European project demonstrated that integrated smartphone-smartwatch systems can objectively track children’s behavioral patterns, providing real-time insights that support personalized obesity management and population-level monitoring[84]. Collectively, these studies suggest that when implemented intentionally and under proper guidance, smartphone-based tools can play a constructive role in encouraging physical activity, supporting weight management, and enhancing overall pediatric health.

Table 10 Use of smartphones to reduce risk of obesity (digital health approaches).

Ref.	Country	Age group	Design	Sample	Digital intervention	Physical activity outcome	BMI outcome	Primary bias considerations	Risk of Bias (JBI)
Vajravelu et al[80], 2022	United States	12-18 years (prediabetes/T2D)	Qualitative formative study	20	SMS reminders + incentives	High acceptability for PA promotion	Not evaluated	High bias regarding clinical outcomes, as they were not evaluated; excellent end-user feedback	N/A (formative)
Ruiz-Hermosa et al[81], 2024 (EUMOVE)	Europe	Primary and secondary school	Program implementation	Multi-country	School-based app + active resources	Promotes active commuting & PA	Preventive framework	Descriptive study of project phases; effectiveness not yet reported in this publication	N/A (descriptive)
Umano et al[82], 2024	Italy	Pediatric obesity	RCT	75	Lifestyle counseling app	Improved engagement; lower dropout	No significant BMI Z-score difference	Randomized controlled design; objective BMI measurements at 6 months and 12 months	Low
Kassari et al[84], 2026 (BigO)	Europe	Mean 12.6 years	Prospective cohort	1727	Smartphone + smartwatch objective monitoring	Behavioral tracking & lifestyle intervention	Decreased obesity proportion	Large prospective sample; used objective sensors (GPS/inertial) for behavioral monitoring	Low
Vaidya et al[83], 2026	India	8-15 years	Pilot clinical trial	26	App-based counseling + yoga	Increased PA duration	Significant BMI reduction (P < 0.001)	Small pilot sample size; lacks a formal control group for the pre-post comparison	Moderate to high

Risk of Bias assessment has been performed using the Joanna Briggs Institute checklist for analytical cross-sectional studies and ergonomic assessments. PA: Physical activity; BMI: Body mass index; JBI: Joanna Briggs Institute; T2D: Type 2 diabetes; RCT: Randomized controlled trial; N/A: Not applicable.

Synthesis of evidence: The overall body of evidence demonstrates that excessive smartphone use is consistently associated with a higher risk of obesity in children and adolescents, with strong cross-sectional findings indicating that greater daily use, particularly beyond 3 hours per day, is linked to higher weight status. Longitudinal studies reinforce this relationship by showing that smartphone addiction independently elevates obesity risk over time, supporting a temporal and potentially causal pathway. These associations appear to follow a clear dose-response pattern and are largely mediated by behavioral mechanisms, including reduced physical activity, disrupted sleep, and unhealthy dietary habits. At the same time, structured sports participation and muscle-strengthening activities consistently emerge as protective factors that mitigate the adverse effects of high smartphone engagement. Importantly, the relationship between smartphone use and obesity is bidirectional and context-dependent: Passive, prolonged, or addictive use contributes to sedentary behavior and weight gain, whereas well-designed, behaviorally informed smartphone-based interventions can actively support obesity prevention and management by promoting healthier habits and increasing physical activity.

PSU/addiction

Prevalence and distribution across age groups: Prevalence estimates of PSU varied by age and measurement tool but were consistently substantial across childhood and adolescence, as shown in Table 11. In nationally representative Korean data during the COVID-19 pandemic (n = 54948), PSU prevalence was 25.5%, with higher rates among females and high school students[14]. Similarly, in an Indian mixed-method study of adolescents aged 15-19 years (n = 560), addiction prevalence reached 64% among smartphone users[85].

Table 11 Summary of studies examining problematic smartphone use by age, design, predictors, and outcomes.

Ref.	Country	Age group	Design	Sample size	Key predictors	Main outcomes	Primary bias considerations	Overall risk
Donati et al[96], 2025	Italy	Adolescents (mean age 163 years)	Randomized controlled classroom intervention	93	Metacognitive beliefs; cognitive-behavioral training	↓ Daily screen time; ↓ risky smartphone behaviors; improved metacognitive regulation (large effects)	Strengths: Randomized design with a control group. Limitations: Preliminary study with a relatively small sample (n = 93) and short duration (5 weeks)	Low to moderate
Park et al[93], 2022	South Korea	Early adolescents (mean age 129 years)	Cross-sectional	209	Emotional overeating; food addiction symptoms	PSU correlated with food addiction; high-risk PSU group had 2.3 × higher food addiction scores	Strengths: School-based community sample; adjusted for BMI and SES. Limitations: Small sample size (n = 209); reliance on self-reported data	Moderate
Grund and Luciana[88], 2025	United States (ABCD)	Baseline 9-10 years → 12-15 years follow-up	Prospective longitudinal	4754	Urgency (impulsivity); reward sensitivity; externalizing; punishment sensitivity	Urgency and punishment sensitivity predicted PSU; cognitive ability not predictive; PSU distinct from screen time	Strengths: Large-scale prospective cohort (n = 4754) from the ABCD study. Controlled for family nesting and site	Low
Meng et al[92], 2020	China (national sample)	Young adolescents (mean age 129 years)	Cross-sectional mediation	8261	Hedonic, instrumental, self-expression motivations; SUT	Hedonic motivation → ↑ PSU via entertainment use; instrumental motivation → ↓ PSU via learning use	Strengths: Large national representative sample (n = 8261). Limitations: Cross-sectional mediation can overlook temporal precedence	Low to moderate
Yoon et al[91], 2025	South Korea	Children (Grade 4) + siblings	4-year longitudinal panel	1978	Sibling smartphone addiction (initial level and slope)	Higher sibling addiction predicted higher child PSU trajectories	Strengths: Long-term follow-up (4 years) using established panel data (KCYPS). Limitations: Nested sibling data requires complex modeling	Low
Lee et al[17], 2025	South Korea	Middle and high school	National cross-sectional survey	54948	Female sex; high school grade; alcohol; smoking	PSU prevalence 255%; alcohol (OR ≈ 1.10) and smoking (OR ≈ 1.30) increased PSU risk	Strengths: Massive, high-powered national survey (n = 54948). Limitations: Entirely self-reported via web survey	Low to moderate
Carter et al[31], 2024	United Kingdom	Adolescents 16-18 years	Multi-school cross-sectional	657	PSU (SAS); not screen time	PSU associated with anxiety (aOR = 2.03), depression (aOR = 2.96), insomnia (aOR = 1.64); screen time not associated	Strengths: Multi-school enrollment; used validated clinical tools (GAD-7, PHQ-9). Limitations: Cross-sectional design	Moderate
Carter et al[94], 2024	United Kingdom	Adolescents 13-16 years	Prospective mixed-method cohort	69	PSU severity	↑ PSU predicted worsening anxiety (β = 0.18); qualitative academic and relational strain	Strengths: Captures longitudinal changes in mood. Limitations: Very small sample (n = 69) and short follow-up period	Moderate to high
Huang et al[95], 2020	Taiwan	Children 9-12 years	Validation study	319	ADHD status	Reliable PSU scale (α = 0.93); ADHD group showed higher PSU proneness	Strengths: Evaluated reliability (α = 0.93) of the SAPS scale in a specific population (ADHD)	Moderate
Bae and Nam[90], 2023	South Korea	Early adolescents	Secondary panel mediation	KCYPS dataset	Maternal PSU; time spent with child; self-esteem	Maternal PSU → ↓ interaction time → ↑ adolescent PSU (sequential mediation)	Strengths: Uses high-quality national panel data (KCYPS) to track maternal-child dynamics over time	Low
Xiao et al[87], 2025	Canada	Adolescents (Grade 8-12)	4-year longitudinal (growth mixture)	2549	FoMo; depression; self-regulation	3 PSU trajectories; FoMo and depression predicted high-stable PSU; self-regulation protective	Strengths: Long-term tracking of trajectories (n = 2549) with sophisticated growth mixture modeling	Low
Lee et al[86], 2024	South Korea	Early childhood (mean 4.5 years)	4-year cohort	313	Parental lack of control; parental PSU	Parental factors predicted higher child smartphone addiction tendency	Strengths: Longitudinal design from early childhood. Limitations: Smaller cohort size (n = 313) compared to national datasets	Low to moderate
Huang et al[89], 2021	China	Children and adolescents (mean 12.3 years)	Network analysis	3248	Self-control; peer attitudes; parent-child relationship; FoMo	Central nodes: Loss of control, peer attitudes, self-control, parent-child relationship	Strengths: Large sample (n = 3248) providing detailed interaction mappings of risk factors	Moderate
Ladani et al[85], 2025	India	Adolescents 15-19 years	Mixed-method cross-sectional	560	Urban residence; parental education; gaming/social media use	Addiction prevalence 64%; gaming and social media associated with PSU	Strengths: Incorporates qualitative insights. Limitations: Cross-sectional nature limits causal interpretation	Moderate

Risk of Bias assessment has been performed using the Joanna Briggs Institute checklist for analytical cross-sectional studies and ergonomic assessments. FoMo: Fear of missing out; SAPS: Smartphone Addiction Proneness Scale; SUT: Smartphone use time; PSU: Problematic smartphone use; SAS: Zung’s Self-rating Anxiety Scale; ADHD: Attention-deficit hyperactivity disorder; OR: Odds ratio; aOR: Adjusted odds ratio; GAD-7: General Anxiety Disorder 7; PHQ-9: Patient Health Questionnaire 9; BMI: Body mass index; SES: Socioeconomic status.

In younger populations, longitudinal cohort data from Korean preschoolers (mean age 4.5 years at baseline) demonstrated measurable trajectories of smartphone addiction tendency over four years, indicating that problematic patterns may emerge in early childhood[86]. Among Canadian adolescents (n = 2549), growth mixture modeling identified three PSU trajectories: Low-increasing-decreasing (35.5%), moderate-increasing (60.9%), and high-stable (3.6%)[87]. Various factors may increase the risk of PSU in children, as shown in Figure 6.

Figure 6 Risk ecology model of problematic smartphone use in children and adolescents. This concentric ecological framework illustrates six interrelated levels of risk contributing to problematic smartphone use among children and adolescents. At the core are individual factors (biological vulnerabilities, neurodevelopmental traits, emotional dysregulation, and executive function deficits). Surrounding this are behavioral and lifestyle factors, including excessive screen time, nighttime use, gaming, and reduced physical activity. The next layer represents the family and parenting environment, highlighting parental modeling, monitoring practices, and household screen norms. The fourth layer encompasses the peer and school environment, including peer pressure, cyberbullying exposure, academic stress, and fear of missing out. The fifth layer reflects the characteristics of digital platforms, such as algorithmic reinforcement, push notifications, infinite scrolling, and social validation mechanisms. The outermost layer represents societal and structural determinants, including pandemic-related shifts, socioeconomic disparities, urbanization, and regulatory context. Together, these nested systems interact dynamically, increasing vulnerability to problematic smartphone use and mediating downstream health outcomes across mental, sleep, metabolic, musculoskeletal, and neurodevelopmental domains. COVID-19: Coronavirus disease 2019.

Individual-level predictors - impulsivity and emotional traits: Prospective ABCD study data (n = 4754) demonstrated that urgency (impulsivity facet) and punishment sensitivity predicted later PSU independent of screen time, whereas cognitive ability did not predict PSU. Externalizing symptoms predicted PSU for video games and smartphones[88]. In addition, longitudinal trajectory analyses further identified fear of missing out and depression as significant predictors of membership in the high-stable PSU trajectory, while better self-regulation predicted lower PSU risk[87]. Moreover, network analyses in a large Chinese juvenile sample (n = 3248) confirmed that loss of control, self-control deficits, and peer attitudes toward smartphone use were central nodes in PSU symptom networks, demonstrating high strength and bridge centrality indices[89].

Family-level influences: Family context emerged as a robust longitudinal determinant of PSU. In early childhood, parental lack of control over smartphone use and parental smartphone addiction proneness significantly predicted increases in children’s smartphone addiction tendency over four years[86]. Among adolescents, mothers’ PSU was associated with adolescents’ PSU, mediated sequentially by reduced mother-child interaction time and adolescent self-esteem[90]. Moreover, sibling effects were also observed: Four-year panel data from 1978 Korean participants showed that higher initial sibling smartphone addiction levels predicted greater child smartphone addiction trajectories. Increases in sibling addiction were paralleled by increases in child addiction[91]. Furthermore, network modeling further highlighted parent-child relationship quality as a central and bridge factor influencing PSU across microsystems[89].

Motivational and behavioral mechanisms: Using a nationally representative Chinese sample (n = 8261), Meng et al[92] (2020) demonstrated that hedonic smartphone use motivation predicted PSU via increased entertainment and communication time, whereas instrumental motivation was negatively associated with PSU via learning-related use. Self-expression motivation showed dual effects depending on activity type. In addition, addicted users were more likely to engage in gaming, social media, and video consumption.

Comorbid addictive and risk behaviors: PSU frequently co-occurs with other addictive behaviors. Among Korean adolescents (mean age 12.86 years), PSU was positively correlated with food addiction symptoms after adjustment for demographic factors. The high-risk PSU group exhibited 2.3-fold higher food addiction scores. Emotional overeating was significantly associated with PSU[93]. During the COVID-19 pandemic, alcohol and smoking increased the risk of PSU by 1.10-fold and 1.30-fold, respectively[17].

Mental health outcomes: Across United Kingdom school-based studies, PSU was associated with moderate anxiety (adjusted OR = 2.03, 95%CI: 1.28-3.23), depression (adjusted OR = 2.96, 95%CI: 1.80-4.86), and insomnia (adjusted OR = 1.64, 95%CI: 1.08-2.50). Importantly, screen time alone was not associated with anxiety or depression, highlighting that compulsive and distress-driven use rather than duration per se predicted mental health outcomes[31]. In addition, prospective mixed-method data in younger adolescents (13-16 years) confirmed that increases in PSU scores were linearly associated with worsening anxiety symptoms over time (β = 0.18, P = 0.013), alongside qualitative reports of academic and relational strain[94].

Clinical measurement and detection: The Chinese version of the Smartphone Addiction Proneness Scale demonstrated excellent internal consistency (Cronbach’s α = 0.93) and robust factor structure (KMO = 0.94) in children aged 9-12 years, supporting its reliability for early detection[95].

Intervention evidence: Preliminary randomized classroom intervention data (n = 93 adolescents) demonstrated that a five-session metacognitive-based prevention program significantly reduced daily smartphone time and risky smartphone behaviors, with large effect sizes compared to controls. Improvements were also observed in maladaptive metacognitive beliefs about emotional and social regulation via smartphones[96].

Overall synthesis of findings: The convergence of longitudinal, nationally representative, and network-analytic studies supports PSU as a multidimensional behavioral addiction phenotype with identifiable developmental, psychosocial, and familial determinants. PSU emerges early in childhood and tends to intensify throughout adolescence, reflecting a developmental trajectory shaped by individual, familial, and motivational factors. Longitudinal evidence shows that PSU is predicted by traits, such as impulsivity, emotional vulnerability, fear of missing out, and diminished self-regulation, while family dynamics - including parental and sibling modeling - play a substantial role in reinforcing maladaptive patterns of use. Moreover, motivation-specific behaviors, particularly hedonic engagement, further mediate the likelihood of developing PSU. The condition frequently co-occurs with other addictive or risk-related behaviors such as food addiction, alcohol consumption, and smoking, and it remains independently associated with heightened anxiety, depressive symptoms, and insomnia, even after accounting for overall screen time. Importantly, emerging intervention research indicates that PSU risk can be mitigated through targeted metacognitive programs that improve emotional and cognitive regulation. Collectively, findings from longitudinal, nationally representative, and networkanalytic studies increasingly characterize PSU as a multidimensional behavioral addiction with identifiable developmental, psychosocial, and familial determinants.

RF-EMF

RF-EMF exposure: Children are widely exposed to RF-EMFs (30 kHz to 300 GHz) emitted by mobile phones, wireless devices, Wi-Fi systems, and base stations. As summarized by Moon[97] (2020), RF-EMF exposure differs from extremely low frequency (ELF) fields in both frequency range and biological effects, with established thermal effects at high intensities and less well-defined nonthermal mechanisms. The International Agency for Research on Cancer classifies both RF-EMF and ELF fields as Group 2B (possibly carcinogenic to humans), reflecting limited human evidence and inadequate experimental confirmation, while the World Health Organization considers current evidence insufficient to establish causality[97]. Table 12 summarizes the main studies evaluating RF-EMF exposure and health outcomes in children and young people. Figure 7 gives an overview of RF-EMF exposure from smartphones in children.

Figure 7 Conceptual overview of radiofrequency electromagnetic field exposure from smartphones in children. The figure illustrates the primary sources of pediatric radiofrequency electromagnetic field exposure, including mobile phones, Wi-Fi routers, and cellular base stations. During active smartphone use, radiofrequency electromagnetic field emissions are absorbed locally in cranial tissues, particularly when the device is held near the head. Proposed biological mechanisms include thermal effects and putative nonthermal cellular responses; however, current epidemiologic and experimental evidence in children does not demonstrate increased brain tumor risk, elevated overall cancer incidence, or consistent neurocognitive impairment at exposure levels within international safety standards. The diagram distinguishes theoretical biological mechanisms from the largely reassuring evidence base while emphasizing ongoing research and regulatory oversight. RF-EMF: Radiofrequency electromagnetic field.

Table 12 Summary of studies evaluating radiofrequency electromagnetic field exposure and health outcomes in children and young people.

Ref.	Design	Population/age	Exposure metric	Outcome (s)	Key findings	Primary bias considerations	Risk of bias (JBI)
Moon[97], 2020	Narrative review	Children (various ages)	Environmental RF-EMF (mobile phones, WiFi, base stations)	Carcinogenicity, neurodevelopmental, and biological effects	RF-EMF is classified by IARC as Group 2B (possible carcinogen). No confirmed causal pediatric harm; precautionary approach recommended due to biological uncertainty	As a narrative review, it is prone to selection bias compared to systematic reviews; however, it accurately synthesizes IARC and WHO positions	Moderate to high
Castaño-Vinyals et al[98], 2022 (MOBI-Kids)	Multinational case-control (14 countries)	899 cases, 1910 controls; age 10-24 years	Cumulative call time, number of calls, years since first use, modeled cumulative RF energy at tumor site	Neuroepithelial brain tumors (mainly glioma)	No increased risk with higher exposure. ORs did not increase with cumulative use; some inverse trends likely due to recall bias. Small risk increase cannot be entirely excluded	Robust multinational design with sophisticated RF-energy modeling. Recall bias is a noted limitation common to retrospective case-controls	Low
Elliott et al[99], 2010	National case-control (registry-linked)	1397 cancer cases (0-4 years), 5588 controls	Distance to base stations, modeled RF power density at birth address	All cancers, CNS tumors, leukemia, NHL	No association between prenatal base-station RF exposure and early childhood cancer. ORs = 1.0 across exposure categories	High-quality registry-linked study; used objective distance to base stations, eliminating recall bias for exposure	Low
Durusoy et al[105], 2017	Cross-sectional survey	2150 high school students	Mobile phone use characteristics (calls/day, duration, texts/day, nighttime proximity); measured school EMF levels	Headache, fatigue, sleep disturbance, concentration difficulties	Mobile phone use associated with headache (OR = 1.90), fatigue (OR = 1.78), sleep disturbance (OR = 1.53); dose-response observed. No association with measured school EMF levels	Large sample size; however, symptoms (headache, fatigue) and exposure characteristics were entirely self-reported	Moderate
Edelstyn and Oldershaw[100], 2002	Randomized experimental	38 adults (young)	30-minute exposure to 900 MHz mobile phone vs sham	Attention, processing speed	No deficits observed; transient facilitation in select attentional tasks	Controlled and blinded; however, the sample size is small and the exposure duration was limited to 30 minutes	Low to moderate
Mortazavi et al[101], 2012	Randomized experimental	160 university students (18-31 years)	10-minute real vs sham exposure (high SAR handset)	Visual reaction time	Significant reduction (faster reaction time) after real exposure; no impairment demonstrated	Well-controlled randomized trial with a larger sample for an experimental neurophysiology study	Low
Sauter et al[102], 2011	Randomized crossover	30 young male adults (mean 25 years)	7 hours 15 minutes exposure to GSM 900, WCDMA, or sham	Attention, working memory	No consistent cognitive effects after correction for multiple testing; time-of-day effects significant	Robust crossover design (participants served as their own controls); accounted for multiple testing corrections	Low
Terao et al[103], 2006	Double-blind crossover	16 adults	30-minute pulsed EMF vs sham	Visuo-motor reaction time, movement time	No significant short-term effects on visuo-motor processing	High internal validity due to double-blinding and counterbalanced crossover design	Low
Inomata-Terada et al[104], 2007	Experimental neurophysiology	10 adults (+ 2 MS patients)	30-minute mobile phone EMF exposure	Motor evoked potentials (TMS), intracortical inhibition	No detectable short-term effects on motor cortical excitability or GABAergic inhibition	Small sample size (n = 10) limits power to detect subtle effects, though the neurophysiological methods (TMS) are highly precise	Moderate

Risk of Bias assessment has been performed using the Joanna Briggs Institute checklist for analytical cross-sectional studies and ergonomic assessments. JBI: Joanna Briggs Institute; IARC: International Agency for Research on Cancer; WHO: World Health Organization; TMS: Transcranial magnetic stimulation; MS: Multiple sclerosis; RF-EMF: Radiofrequency electromagnetic field; RE: Radiofrequency; EMF: Electromagnetic field; OR: Odds ratio; CNS: Central nervous system.

Brain tumor risk in young people: The largest pediatric investigation to date, the international MOBI-Kids case-control study (Castaño-Vinyals et al[98], 2022), examined 899 cases of neuroepithelial brain tumors diagnosed between ages 10 years and 24 years and 1910 matched controls across 14 countries. Across multiple exposure metrics - including cumulative call time, number of calls, years since first use, estimated cumulative RF-specific energy at tumor location, and ELF-induced current density - no increased risk was observed. ORs for neuroepithelial tumors did not rise with increasing exposure; in several analyses, ORs decreased with higher cumulative use. Investigators attributed the inverse trends primarily to differential recall bias and potential reductions in phone use during prodromal phases preceding diagnosis. While residual confounding cannot be excluded, the authors concluded that their findings provide no evidence of a causal association between wireless phone use and brain tumors in young people, though a small increase in risk cannot be definitively ruled out[98].

Prenatal and early childhood cancer risk: A national case-control study from Great Britain evaluated maternal exposure to RF emissions from mobile phone base stations during pregnancy. The study included 1397 cases of cancer diagnosed before age 5 and 5588 matched birth controls. No associations were observed between modeled RF exposure metrics - including distance to base stations, total transmitted power within 700 meters, and modeled power density at the birth address - and risk of all cancers combined, brain and central nervous system tumors, leukemia, or non-Hodgkin lymphoma. Adjusted odds ratios were approximately 1.0 across exposure categories, with no dose-response trends. These findings provide strong population-based evidence against an association between prenatal exposure to base-station RF and early childhood cancer risk[99].

Acute cognitive and neurophysiological effects: Experimental studies assessing short-term RF-EMF exposure have generally not demonstrated cognitive impairment. In a randomized study of 38 adults exposed for 30 minutes to a 900 MHz mobile phone signal, Edelstyn and Oldershaw[100] (2002) observed transient improvements in digit span and serial subtraction tasks compared with sham exposure, without evidence of cognitive deficit. Similarly, Mortazavi et al[101] (2012) reported significantly decreased visual reaction time following 10-minute mobile phone exposure compared with sham conditions in university students, suggesting facilitated response speed rather than impairment.

In contrast, larger crossover trials found no significant cognitive effects. Sauter et al[102] (2011) exposed 30 young male participants to GSM 900 and WCDMA signals for over seven hours and found no consistent differences in attention or working memory after correction for multiple comparisons; time-of-day effects accounted for observed variations. Likewise, Terao et al[103] (2006) reported no significant changes in visuo-motor reaction time after 30-minute exposure in a double-blind crossover design. Neurophysiological assessment using transcranial magnetic stimulation by Inomata-Terada et al[104] (2007) demonstrated no changes in motor evoked potentials or intracortical inhibition after 30-minute exposure. Collectively, controlled laboratory studies do not demonstrate short-term cognitive impairment from handset-level RF exposure, and isolated findings of facilitated reaction time remain inconsistent and of uncertain clinical relevance.

Self-reported symptoms in adolescents: A large cross-sectional study of 2150 high school students in Turkey evaluated associations between mobile phone use characteristics, measured school EMF levels, and 23 self-reported symptoms[105]. Mobile phone users reported higher odds of headache (OR = 1.90, 95%CI: 1.30-2.77), fatigue (OR = 1.78, 95%CI: 1.21-2.63), and sleep disturbances (OR = 1.53, 95%CI: 1.05-2.21) compared with nonusers. Dose-response relationships were observed for the number of calls, cumulative call duration, text message frequency, nighttime phone proximity, and making calls while charging. However, measured school EMF levels were not associated with symptom prevalence, and associations with proximity to base stations were limited and inconsistent. These findings suggest that reported symptoms may relate more strongly to behavioral patterns of phone use rather than environmental RF exposure intensity per se[105].

Integrative interpretation of RF-EMF findings: Across carcinogenic, neurophysiological, and symptom-based domains, the evidence does not support a causal association between RF-EMF exposure from mobile phones or base stations and pediatric brain tumors, early childhood cancers, or objective neurocognitive impairment. Large multinational case-control data and national registry-linked analyses are reassuring at the population scale[98,99]. Short-term experimental studies largely demonstrate no impairment in attention, working memory, or motor cortical function[102-104], and occasional findings of facilitated reaction time have not been consistently replicated[97,98].

Associations between mobile phone use and subjective symptoms such as headache and sleep disturbance appear more plausibly explained by behavioral factors - including duration of use, nighttime exposure, and sleep displacement - than by RF intensity levels[105]. As emphasized in the pediatric review by Moon et al[60] (2016), precautionary approaches remain reasonable given theoretical biological vulnerability during development, but current epidemiologic evidence does not demonstrate established harm at exposure levels typical of consumer devices.

DISCUSSION

Principal findings

This systematic review provides a comprehensive synthesis of the current evidence regarding the health impacts of mobile phone use in pediatric populations. Our findings demonstrate that mobile phone exposure is nearly universal, beginning in early infancy and intensifying through adolescence. The evidence suggests that the health consequences of this exposure are multidimensional, spanning sleep, mental health, physical well-being, and neurodevelopment.

One of the most striking findings is the early age at which initiation occurs. With a pooled prevalence of 91.8% among children aged 0-4 years, smartphones have become a standard fixture of early childhood. This is significant because this period represents a critical window of brain plasticity and neurobiological maturation. Environmental exposures during this period may therefore exert disproportionate developmental effects[13,14].

To synthesize the heterogeneity of outcomes, the evidence is best understood through two distinct but interacting risk pathways: The behavioral pathway, which encompasses immediate and well-documented risks related to lifestyle displacement, sleep disruption, posture, physical inactivity, and impaired selfregulation; and the biological pathway, which focuses on potential effects of RF-EMF exposure, an area that remains under active investigation but for which evidence is currently secondary and inconclusive compared with behavioral outcomes.

Across both pathways, a consistent directional pattern emerged: Greater exposure, particularly prolonged daily use, bedtime engagement, and PSU, was associated with a higher risk of adverse outcomes. Although individual effect sizes were generally small to moderate, their consistency across populations, age groups, and study designs suggests meaningful public health implications given the near-universal penetration of smartphones in pediatric life[106,107].

Furthermore, the COVID-19 pandemic acted as a catalyst, significantly increasing daily usage duration and the prevalence of PSU. In some populations, daily use during the pandemic reached nearly 5 hours on weekdays and over 6 hours on weekends. This shift was not merely quantitative; it was qualitative, with a marked increase in symptoms of addiction and a decline in physical activity[16-20].

The review also integrates contemporary evidence examining associations between smartphone exposure and multidimensional health outcomes in children and adolescents. Across neurodevelopmental, cognitive, sleep, mental health, visual, musculoskeletal, and metabolic domains, a consistent directional pattern emerged: Greater exposure, particularly prolonged daily use, bedtime engagement, and PSU, was associated with a higher risk of adverse outcomes. Although effect sizes were generally small to moderate at the individual level, their consistency across populations, age groups, and study designs suggests meaningful public health implications given the near-universal penetration of smartphones in pediatric populations.

Three cross-domain patterns warrant emphasis. First, dose-response gradients were evident in multiple domains, including sleep disturbance, depressive symptoms, inattention, DES, musculoskeletal pain, and overweight/obesity. Second, developmental timing moderated vulnerability: Early childhood exposure was more strongly associated with language and cognitive delays, whereas adolescence demonstrated stronger associations with attentional dysregulation, emotional symptoms, and addictive behavioral patterns. Third, contextual mediators - bedtime use, device presence during sleep, sedentary displacement, posture, and parental modeling - substantially shaped risk, underscoring that exposure alone is insufficient to explain outcomes. Collectively, these findings support a systems-based risk-ecology framework rather than a simplistic exposure-outcome model. Smartphone-related health effects appear to arise from interacting behavioral, neurobiological, and environmental pathways that vary across developmental stages.

Neurodevelopment and cognitive function (moderate evidence)

In early childhood, higher daily screen exposure, particularly ≥ 2 hours per day, was associated with lower performance in language and global cognitive domains, and with an increased risk of developmental delay. These findings are consistent with displacement models, whereby passive digital engagement reduces opportunities for reciprocal caregiver interaction, sensorimotor exploration, and language scaffolding during sensitive developmental windows[108].

In contrast, findings for adolescents were more domain-specific. Associations were strongest for attentional control and executive functioning, whereas global intelligence measures showed weaker or inconsistent relationships. This selective pattern aligns with neurodevelopmental maturation trajectories: Prefrontal attentional networks remain plastic through adolescence and may be particularly sensitive to rapid-reward, high-salience digital stimuli. Experimental evidence demonstrates reduced working memory performance in the mere presence of a smartphone, supporting a potential attentional resource-allocation mechanism independent of active use[109,110].

Importantly, observed effect sizes were generally modest, suggesting not abrupt impairment but incremental executive load that may accumulate with chronic overexposure. Children with ADHD and related neurodevelopmental conditions appeared disproportionately vulnerable, possibly reflecting baseline regulatory challenges. However, the predominance of cross-sectional designs limits definitive causal inference, and content quality, interactivity, and parental mediation were inconsistently measured[111,112]. Thus, the evidence supports selective vulnerability rather than global cognitive harm.

Sleep disruption as a central mechanistic pathway

Sleep outcomes demonstrated the most consistent and clinically meaningful associations across developmental stages. Increased smartphone use, particularly at bedtime and during the night, was robustly associated with shorter sleep duration, delayed sleep onset, poorer sleep quality, and more nighttime awakenings. Importantly, dose-response relationships were evident, and randomized interventions that restricted pre-sleep smartphone use demonstrated partial reversibility, thereby strengthening causal inference[113,114].

Mechanistically, sleep disruption likely reflects convergent pathways: Blue-light-mediated circadian phase delay, cognitive-emotional arousal, disruption of bedtime routines, and reinforcement loops sustaining device engagement. Given the foundational role of sleep in neurocognitive maturation, emotional regulation, metabolic homeostasis, and immune function, sleep plausibly serves as a central mediator linking smartphone exposure to downstream mental health and metabolic outcomes[115,116]. From a clinical perspective, sleep hygiene represents the most modifiable and highest-yield intervention target identified in this review.

Mental health and emotional regulation

Across cross-sectional and longitudinal studies, higher smartphone exposure and PSU were associated with depressive symptoms, anxiety, emotional dysregulation, and reduced life satisfaction. Notably, PSU - characterized by compulsive engagement, loss of control, and emotional reliance - demonstrated stronger associations with adverse outcomes than total duration alone. Reported odds ratios for depression and anxiety frequently ranged between 1.5 and 3.0, with correlation coefficients in the small-to-moderate range.

The distinction between duration and dysregulation is clinically significant. Raw screen time may be less predictive than patterns reflecting impaired self-regulation and maladaptive coping. Adolescents appeared particularly susceptible, consistent with developmental sensitivity to peer validation, reward-driven behaviors, and social comparison[117].

Although reverse causality remains plausible, particularly in internalizing disorders, the convergence of longitudinal data and randomized screen-reduction trials demonstrating improvement in depressive symptoms supports at least partial causal contribution. The relationship is likely bidirectional, with emotional vulnerability both predicting and being amplified by dysregulated smartphone use[38].

Visual and ocular health

Evidence consistently linked prolonged smartphone use to DES and dry eye symptoms. Mechanistically, reduced blink rate, sustained accommodative demand, and prolonged near work plausibly disrupted tear film stability and induced ocular fatigue[118]. While associations with myopia progression remain partially confounded by overall near work and reduced outdoor exposure, biological plausibility is strong, and prospective data suggest dose-dependent relationships[57]. Given the global rise in pediatric myopia, smartphone-related near-work behaviors warrant integration into preventive ophthalmologic counseling. Increased outdoor activity emerged as a consistent protective factor, reinforcing the importance of behavioral counterbalancing rather than device elimination[119].

Musculoskeletal outcomes

Across age groups, prolonged smartphone use, particularly in sustained cervical flexion, was associated with neck, shoulder, and upper back pain. Dose-response relationships were reported for daily use exceeding 3-5 hours. Biomechanical modeling demonstrates exponential increases in cervical spine load with flexion angle, supporting physiological plausibility[120]. Adolescents reported the highest prevalence of symptoms, likely reflecting cumulative exposure and prolonged unsupervised use[121]. Emerging biomarker findings indicating alterations in oxidative stress in high-use groups provide preliminary biological corroboration, although replication is required[122]. While most evidence is cross-sectional, the consistency of directionality across populations strengthens the inference of a clinically meaningful association. Importantly, posture-related risks are modifiable, positioning ergonomic education as a feasible preventive strategy[123].

Metabolic risk and obesity

Higher smartphone use was associated with increased odds of overweight and obesity, with evidence of dose-dependent relationships. Mechanistic pathways likely include sedentary displacement, concurrent snacking, sleep disruption, and reward-driven eating behaviors. Longitudinal studies suggest that smartphone addiction predicts later weight gain support temporality[124]. However, smartphone exposure represents one component within a broader obesogenic environment. Notably, structured physical activity and sports participation consistently mitigated risk, underscoring that behavioral context moderates the impact. Digital technologies themselves may also be leveraged therapeutically through behaviorally informed interventions[125].

PSU as a behavioral phenotype

The evidence increasingly supports PSU as a multidimensional behavioral addiction phenotype with identifiable developmental, psychosocial, and familial determinants. Longitudinal and network-analytic studies demonstrate that impulsivity, emotional vulnerability, fear of missing out, and diminished self-regulation predict PSU trajectories, while parental and sibling modeling reinforce patterns of use[126]. PSU frequently co-occurs with other risk behaviors and remains independently associated with anxiety, depression, and insomnia after adjusting for total screen time. Emerging intervention data targeting metacognitive and emotional regulation processes suggest modifiability, supporting clinical relevance[127].

RF-EMF exposure: Carcinogenic and neurobiological risk

Public concern regarding RF-EMF exposure has centered primarily on carcinogenicity and neurodevelopmental vulnerability. Although the International Agency for Research on Cancer classifies RF-EMF as Group 2B (“possibly carcinogenic”), this designation reflects limited evidence and does not imply established causality[128].

The multinational MOBI-Kids study - the most comprehensive pediatric investigation to date - found no increased risk of neuroepithelial brain tumors with increasing cumulative wireless phone use, call duration, or modeled RF dose at the tumor location. ORs did not increase with exposure, and inverse trends were attributed to recall bias and prodromal behavioral changes[98]. Similarly, large registry-based analyses of prenatal exposure to base stations demonstrated no association with early childhood cancers[99].

Importantly, no population-level increase in pediatric brain tumor incidence has paralleled the exponential growth in mobile phone use over the past two decades. While small risk elevations cannot be completely excluded due to latency considerations and exposure misclassification, the available epidemiologic evidence does not support a substantial carcinogenic effect in children or adolescents[129].

Experimental neurophysiological studies further demonstrate no consistent short-term impairment in attention, working memory, motor cortical excitability, or visuo-motor processing following controlled handset-level RF exposure[130]. Some small laboratory studies report modest reductions in reaction time; however, these findings are inconsistent, not replicated across paradigms, and of uncertain biological significance. Taken together, current evidence does not demonstrate measurable neurocognitive harm attributable to acute RF-EMF exposure within regulatory safety limits[38]. Thus, while theoretical biological vulnerability in developing neural tissue warrants continued surveillance, the totality of evidence indicates that RF-EMF exposure at contemporary consumer levels is unlikely to be a major contributor to pediatric morbidity.

Integrative interpretation: A developmental risk ecology model

The findings of this review support a developmental risk ecology model as a unified theoretical framework for understanding how smartphone use influences pediatric health. Rather than conceptualizing smartphone exposure as a simple linear exposure-outcome relationship, this model posits that health trajectories emerge from the dynamic interplay of biological, behavioral, and environmental processes operating across critical stages of child and adolescent development. Within this ecological framework, health outcomes are shaped by interacting mechanisms - including sleep disruption, attentional fragmentation, reward-reinforcement loops, sedentary displacement, and postural biomechanics - that act synergistically over time to influence neurodevelopmental, physical, and psychosocial trajectories as shown in Figure 8[131].

Figure 8 Developmental risk ecology model of smartphone use in children and adolescents. This conceptual model illustrates the multi-level ecological framework through which smartphone use influences health outcomes in children and adolescents. At the center, the child interacts dynamically with the digital environment, embedded within nested contextual layers including individual characteristics (e.g., age, temperament, mental health), family environment, and broader sociocultural influences. The behavioral pathway (solid arrows) represents the primary mechanism underlying observed health effects. Within this pathway, sleep disruption and sedentary/biomechanical strain (thick arrows) demonstrate the strongest and most consistent empirical support across studies. Attentional and reward-processing mechanisms (medium arrows) reflect a moderate but more heterogeneous evidence base, particularly relevant to cognitive and mental health outcomes. The biological pathway, representing radiofrequency electromagnetic field exposure, is depicted by a dashed arrow to indicate its secondary and precautionary status, given the currently limited and inconclusive longitudinal evidence regarding its health effects. Contextual moderators - including parental modeling, content type, supervision, and peer influences - act as regulatory filters that shape exposure patterns and modify risk trajectories. The model emphasizes that health outcomes emerge from the interaction of behavioral, environmental, and biological processes rather than from exposure alone. RF-EMF: Radiofrequency electromagnetic field.

This integrative perspective helps resolve the apparent “disconnection” between disparate domains of observed risk by organizing smartphone-related effects into four interacting pathways, categorized by their current weight of scientific evidence. First, the sleep disruption pathway (strong evidence) is characterized by blue-light-mediated circadian phase delay and heightened cognitive-emotional arousal, particularly during evening and nocturnal use. Sleep disruption serves as the functional “anchor” of the model, as insufficient or fragmented sleep exerts cascading downstream effects on neurocognition, emotional regulation, metabolic health, and immune functioning. Second, the sedentary and biomechanical pathway (strong evidence) reflects the displacement of physical activity and the cumulative impact of sustained non-neutral postures (e.g., “text neck”, reduced blink rate, and prolonged near-work), contributing to musculoskeletal strain, ocular morbidity, and increased metabolic risk[132,133].

Third, the attentional and reward-reinforcement pathway (moderate evidence) is driven by high-salience digital stimuli that compete for limited cognitive resources, consistent with the “brain drain” hypothesis. Repeated engagement with rapidly rewarding content may tax executive control systems, reinforce addictive-like behavioral patterns, and amplify vulnerabilities in emotionally sensitive or neurodivergent populations[134]. Fourth, the RF-EMF biological pathway (limited evidence) focuses on the biophysical properties of smartphones and associated electromagnetic exposures. While this pathway remains the subject of public concern, current high-quality epidemiologic and experimental data are largely reassuring at consumer exposure levels and constitute a precautionary rather than primary risk domain when compared with the robust behavioral evidence base[135].

Importantly, this risk ecology framework distinguishes two interacting but non-equivalent levels of harm. The behavioral pathway represents the primary source of documented risk and is characterized by immediate, dose-dependent effects associated with modifiable patterns of use, such as prolonged daily exposure, bedtime engagement, and PSU. Crucially, risks within this level are largely modifiable, making them amenable to targeted interventions focused on digital hygiene, sleep regulation, physical activity promotion, self-regulatory skill development, and ergonomic education[6].

In contrast, the biological (RF-EMF) pathway remains secondary at typical consumer exposure levels. By explicitly separating this pathway from behavioral harms, the framework prevents the conflation of speculative electromagnetic toxicity concerns with the far more consistently documented psychosocial, behavioral, and lifestyle-mediated risks associated with excessive or dysregulated smartphone use. This distinction clarifies both scientific interpretation and public health messaging: While proportionate biological surveillance (e.g., continued monitoring of emerging technologies such as 5G) is warranted, the immediate clinical and public health priority should remain the regulation of maladaptive usage patterns[136].

Although individual-level effect sizes across domains are often modest, the near-universal exposure to smartphones among children and adolescents means that even small shifts in risk distribution across these ecological pathways may translate into substantial population-level consequences. As such, this unified developmental risk ecology model provides a roadmap for future research and policy to move beyond “screen time” as a monolithic construct and instead examine the synergistic interactions between a child’s biology, behavioral choices, family context, and digital environment[137].

Limitations of the evidence base

Despite the strengths of the current literature, including large population samples and the use of increasingly sophisticated psychometric instruments, several methodological limitations constrain the interpretation of these findings. Foremost among these is the heavy reliance on cross-sectional study designs, which capture associations at a single time point but do not permit causal or temporal inference. This limitation is particularly salient in mental health research, where reverse causality remains a major concern. However, PSU is consistently associated with depressive and anxiety symptoms; it is unclear whether excessive use precipitates emotional distress or whether vulnerable individuals engage in smartphone use as a maladaptive coping strategy. While emerging longitudinal studies offer improved evidence of temporality, relatively few have followed cohorts long enough to assess long-term neurodevelopmental or carcinogenic outcomes. In parallel, exposure measurement remains a significant challenge. Most studies rely on self- or parent-reported screen time, which is prone to recall and social desirability bias, and objective digital tracking is still uncommon. As a result, many measures fail to distinguish between the quantity of use and the quality or context of engagement, such as active vs passive use or content type. Additional heterogeneity arises from the lack of standardized definitions for PSU and “excessive use”, with wide variation in psychometric tools and arbitrary duration thresholds limiting comparability and meta-analytic synthesis. Methodological challenges are further pronounced in the biological domain, where RF-EMF exposure assessment typically depends on indirect proxies rather than individualized dosimetry, and where rapidly evolving technologies (e.g., 5G and device-specific SAR variability) outpace the ability to evaluate long-latency health outcomes. Finally, the predominance of studies from high-income countries restricts global generalizability, as patterns of device use, parental mediation, and cultural context may differ substantially in low- and middle-income settings. Residual confounding, particularly related to socioeconomic status, family dynamics, and comorbid neurodevelopmental conditions, also complicates efforts to isolate the independent effects of smartphone use on health outcomes.

Clinical and public health implications

Clinical and public health strategies should prioritize evidence-informed, modifiable behavioral practices while maintaining a non-alarmist, precautionary stance toward RF-EMF exposure. Current data suggest that the most effective interventions focus on disrupting the “risk ecology” of excessive smartphone use through structured behavioral change.

Evidence-informed intervention strategies

Sleep-focused counseling (high-yield): The strongest intervention evidence currently exists for sleep outcomes. Randomized trials restricting pre-sleep smartphone use have demonstrated the partial reversibility of sleep onset latency and improved sleep duration[114,138,139]. Clinicians should prioritize the “digital sunset” approaching smartphone use at least 60 minutes before bedtime - as a high-yield intervention that yields cascading benefits for cognitive performance and emotional regulation.

Meta-cognitive and behavioral training for PSU: Emerging evidence from RCTs indicates that interventions based on the meta-cognitive model can successfully reduce screen time and risky smartphone behaviors in adolescents[96]. For children with ADHD, who exhibit higher vulnerability to problematic use, structured behavioral interventions targeting self-regulation and dopamine-reward loops are particularly warranted[49].

Physical activity and ergonomic counterbalancing: To mitigate sedentary displacement and musculoskeletal strain, interventions should promote outdoor activity, which prospective data identifies as a protective factor against myopia and obesity[119]. Ergonomic education focusing on neutral cervical spine positioning is a feasible and modifiable strategy to reduce the incidence of “text neck” and associated tension-type headaches[123].

Proactive parental modeling: Clinical guidance should emphasize that parental digital habits are a significant predictor of pediatric use patterns. Shared “co-use” practices and the establishment of “device-free zones” (e.g., during meals) help shape children’s regulatory skills and mitigate the development of addictive behavioral phenotypes[140].

In summary, rather than focusing on the biological risks of RF-EMF - which current high-quality studies find reassuring at consumer levels - public health messaging should be redirected toward these documented, modifiable behavioral risks.

Future directions

Future research should prioritize longitudinal cohort designs with objective digital tracking, mechanistic neuroimaging studies of reward and attentional networks, experimental paradigms targeting sleep and attentional modulation, and culturally diverse populations. Intervention trials focused on bedtime restrictions and PSU mitigation are particularly warranted.

CONCLUSION

In this comprehensive synthesis, RF-EMF exposure from consumer mobile devices does not demonstrate established carcinogenic or neurocognitive harm in pediatric populations. In contrast, excessive or dysregulated smartphone use is consistently associated with adverse outcomes across neurodevelopmental, sleep, mental health, visual, musculoskeletal, and metabolic domains. These findings support a nuanced, developmentally informed approach to digital health - prioritizing behavioral regulation, sleep protection, and family context - while maintaining proportionate vigilance regarding long-term RF-EMF exposure.

ACKNOWLEDGEMENTS

We thank the anonymous referees for their valuable suggestions.