Screen time and social media use were positively associated with anxiety while sleep duration showed a protective effect, with a high-screen and low-sleep behavioral profile comprising 34.3% of the sample exhibiting the highest mean anxiety levels.
Key Findings
Results
Screen time was positively associated with anxiety in a global adult sample across 2020–2024.
Each additional hour of screen time corresponded to a regression coefficient of b = 0.30 (p < 0.001).
The study used hierarchical regression models applied to a repeated cross-sectional global dataset (N = 3,000) spanning 2019–2024.
Statistical significance was evaluated at p < 0.05.
Results
Social media use was positively associated with anxiety, independently of general screen time.
Each additional hour of social media use was associated with a regression coefficient of b = 0.25 (p < 0.001).
This association was identified using hierarchical regression models in the same global adult sample.
The effect size was slightly smaller than that observed for total screen time (b = 0.30).
Results
Sleep duration showed a protective effect against anxiety.
Each additional hour of sleep was linked to a regression coefficient of b = -0.36 (p < 0.001).
The magnitude of the protective effect of sleep (-0.36) was larger than the harmful effect of screen time (0.30) or social media use (0.25).
Sleep adequacy was also tested as a moderator in the analysis.
Results
The association between screen time and anxiety was stronger among adults younger than 40 years than among those aged 40 years and older.
The regression coefficient for screen time and anxiety was b = 0.323 for adults under 40 years.
The regression coefficient for screen time and anxiety was b = 0.265 for adults aged 40 years and older.
The age-based interaction was statistically significant (p = 0.039).
These findings support age-stratified interventions emphasizing sleep hygiene and nighttime device boundaries for younger adults.
Results
A high-screen and low-sleep behavioral profile was identified that comprised 34.3% of the sample and exhibited the highest mean anxiety level.
The high-screen and low-sleep cluster had a mean anxiety level of M = 6.27.
Behavioral profiles were identified using k-means clustering (person-centered clustering techniques).
This profile represented the largest single high-risk subgroup at 34.3% of the total sample (N = 3,000).
The authors recommend screening for this behavioral profile in clinical and educational settings.
Methods
The study used a repeated cross-sectional design applied to a publicly available global dataset on adult mental health and lifestyle.
The dataset covered the period 2019–2024 with a total sample size of N = 3,000.
Both hierarchical regression models and k-means clustering were applied.
Moderation by age and sleep adequacy was assessed.
The authors note that future longitudinal research using objective measures is needed to clarify causality.
What This Means
This research suggests that spending more time on screens and social media is linked to higher anxiety levels in adults, while getting more sleep is linked to lower anxiety levels. The study analyzed data from 3,000 adults globally between 2020 and 2024, finding that each extra hour of screen time corresponded to a meaningful increase in anxiety scores, each extra hour of social media use was also associated with higher anxiety, and each extra hour of sleep was associated with lower anxiety — with sleep appearing to have the strongest relationship of the three factors studied.
The research also found that younger adults (under 40) appeared more vulnerable to the anxiety-related effects of screen time than older adults, suggesting that age matters when thinking about who might be most at risk. Additionally, by grouping participants into behavioral profiles, the researchers identified that about one in three people in the sample fell into a 'high-screen, low-sleep' pattern, and this group had the highest average anxiety levels of any subgroup identified.
This research suggests that public health and clinical efforts to address anxiety might benefit from targeting both screen use habits and sleep quality together, particularly for younger adults. However, because this was a cross-sectional study — meaning it captured snapshots of people's behavior rather than following individuals over time — it cannot definitively prove that screen time causes anxiety or that better sleep directly reduces it. The authors call for future studies using objective measurements and longer follow-up periods to better understand these relationships and inform targeted interventions.
Liu W. (2026). Digital behavior and anxiety in the post-pandemic era: a five-year analysis of screen time, sleep, and behavioral risk profiles.. Frontiers in public health. https://doi.org/10.3389/fpubh.2026.1766808