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

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.