Social Media and Anxiety: What the Evidence Shows

2 days ago

17 min read

Author: Ukrainian Psychological Hub · Published: September 27, 2026 · Editorial Policy

Social media and anxiety are associated in research, but the size and meaning of that association depend heavily on what researchers measure. Time spent on social media, frequency of use, problematic or dysregulated use, exposure to harmful experiences, social comparison, fear of missing out, and anxiety symptoms are not interchangeable variables. The most current evidence therefore supports a conditional conclusion rather than a universal claim that social media causes anxiety. Tølbøll, 2026

A 2026 European Commission Joint Research Centre systematic review of 55 longitudinal and experimental studies found no single uniform association between social media use and poorer mental health among adolescents and emerging adults. Time- and frequency-based measures produced mixed, often small findings, while problematic or dysregulated use, negative consequences, and harmful online experiences showed more consistent links with poorer outcomes. Cabeza Martínez, Cortesi, & Vlachos, 2026

For anxiety specifically, the evidence is clearer for problematic social media use than for ordinary use. Meta-analyses report moderate positive associations between problematic social networking use and generalized anxiety, social anxiety, attachment anxiety, and fear of missing out. These are associations: they do not by themselves establish that social media use caused the anxiety, that anxiety caused the problematic use, or that every person with either pattern will develop the other. Du et al., 2024

Social Media and Anxiety: The Short Answer

The short answer is that social media can be part of an anxiety-related pattern, but “social media use” is too broad a category to function as a diagnosis or a single causal exposure. Research increasingly separates general use from problematic use and specific experiences. That distinction changes the interpretation of the evidence.

A large 2024 systematic review with meta-analyses found a weak positive association between general social media use and anxiety, while problematic social media use showed stronger and more consistent associations with anxiety and other mental-health outcomes. The review also found substantial heterogeneity, meaning results differed across studies, populations, measures, and designs. Ahmed et al., 2024

The broadest 2026 synthesis reached the same basic conclusion. An umbrella review covering 72 reviews published between 2021 and 2025 found weak and inconsistent associations for general social media use, while problematic social media use was consistently linked with ill-being. Tølbøll, 2026

This is why a useful question is not only “How many hours are you on social media?” It is also “What happens there, why are you using it, how controllable does the use feel, what does it displace, and does it reliably increase distress or impair sleep, school, work, relationships, or daily functioning?” For the broader evidence base beyond anxiety, see Social Media and Mental Health: Risks, Benefits, and What Research Shows.

What Counts as “Social Media Use” in Anxiety Research?

Researchers have used many different exposure measures under the label social media use. Some studies ask for daily time. Others count visits, posts, likes, messages, or checking frequency. Some measure passive viewing, active interaction, problematic use, emotional investment, nighttime use, specific platforms, or exposure to particular content. Still others study cyberbullying, social feedback, appearance comparison, social exclusion, or perceived social-media burden.

These measures answer different questions. A high-frequency user who actively communicates with friends is not necessarily experiencing the same psychological exposure as someone repeatedly checking for reassurance, scrolling through threatening news, monitoring social status, or continuing to use a platform despite significant conflict and distress. The field is moving toward this more component-specific view. Maheux et al., 2025

The same precision is needed on the outcome side. Ordinary nervousness, temporary stress, an elevated score on an anxiety questionnaire, symptoms of generalized anxiety, symptoms of social anxiety, and a clinically diagnosed anxiety disorder are different outcomes. The National Institute of Mental Health notes that occasional anxiety is part of life, while anxiety disorders involve persistent or excessive anxiety that interferes with daily functioning.

What Does the Best Evidence Show in 2026?

General social media use has small or inconsistent links with anxiety

Across reviews, ordinary or general social media use tends to show small, inconsistent, or context-dependent associations with anxiety. This does not mean every observed association is trivial, and it does not mean social media experiences are irrelevant. It means that a single global measure such as daily time is a blunt instrument for detecting who is affected, in what direction, and through which mechanism. Cabeza Martínez, Cortesi, & Vlachos, 2026

A 2025 systematic review focused specifically on adolescents and anxiety included 32 studies. Just over half reported a positive association between social media use and anxiety, but results varied according to how social media was measured, and problematic use was the most common exposure measure. The authors emphasized the need for more accurate measures that do not automatically equate problematic use with time spent. Kerr et al., 2025

Problematic or dysregulated use shows a stronger association

The relationship becomes more consistent when studies examine problematic social media use: a research construct centered on impaired control, preoccupation, conflict, negative consequences, or functional disruption rather than time alone. A 2024 meta-analysis of 209 studies involving 252,337 participants found moderate positive associations between problematic social networking use and several anxiety-related outcomes, including generalized anxiety and social anxiety. Du et al., 2024

Another 2024 systematic review and meta-analysis focused on social anxiety and problematic social media use and included 53 studies with 59,928 participants. It also found a strong positive correlation, with the size of the association varying by measurement instrument, sex, publication year, and platform type. Wu, Huang, & Yang, 2024

Problematic social media use remains a research construct rather than an automatic clinical diagnosis. A high score on a screening scale can identify a pattern worth examining, but it does not by itself establish a DSM or ICD diagnosis. The English Hub treats functional impairment, loss of control, distress, and context as distinct from mere frequency. See Problematic Social Media Use: Signs, Risk Factors, and Functional Impairment.

Longitudinal evidence does not support a simple one-way story

Longitudinal studies are more informative about temporal order than one-time surveys, but they still do not automatically prove causation. Their results are also mixed. An eight-year study of 500 adolescents found that within-person increases in time spent on social media were not associated with later increases in depression or anxiety. Coyne et al., 2020

A four-wave Norwegian cohort study followed children from ages 10 to 16 and used psychiatric interviews for anxiety and depression symptoms. Within-person changes in posting, liking, commenting, and related social media behaviors did not predict later changes in anxiety or depression symptoms, and symptoms did not predict later changes in those behaviors. Steinsbekk, Nesi, & Wichstrøm, 2023

A 2026 Swedish cohort study followed participants from age 15 or 16 into young adulthood. Time spent on social media initially appeared related to later anxiety and depression, but the association disappeared after adjustment for confounding factors. The authors concluded that time spent alone may have limited long-term explanatory value and called for research on specific social-media activities. Zetterqvist et al., 2026

At the same time, some longitudinal work on problematic use finds prospective relationships with emotional symptoms, and some studies support bidirectional pathways. A 2026 three-wave study of 1,562 Italian adolescents used a random-intercept cross-lagged model specifically to separate stable between-person differences from within-person change and reported longitudinal coupling between problematic social media use and emotional symptoms. Remondi et al., 2026

Taken together, these findings make directionality a central part of the answer. Anxiety can shape how a person uses social media, social-media experiences can shape anxiety in some contexts, and shared factors can influence both.

Does Social Media Cause Anxiety?

The current evidence does not support a universal claim that social media causes anxiety disorders. It supports a more specific claim: certain patterns of use and certain online experiences are associated with anxiety symptoms, and some longitudinal or experimental findings are consistent with effects in particular contexts. Causal effects are likely heterogeneous rather than identical across people.

The 2026 JRC review was designed specifically around longitudinal and experimental evidence. It concluded that time- and frequency-based measures were mixed and often small, reverse or bidirectional pathways remained plausible, and experimental evidence was more informative for short-term mechanism-linked outcomes than for broad, durable symptom change. Cabeza Martínez, Cortesi, & Vlachos, 2026

This is a stronger basis for interpretation than cross-sectional correlations. If people with more anxiety also report more social-media use at one point in time, the observation alone cannot tell us whether use increased anxiety, anxiety changed the use, another factor influenced both, or several of these processes occurred together.

Why Might Social Media Be Linked With Anxiety?

Research proposes several pathways. None is a single universal mechanism, and several can operate at the same time. The most useful model treats anxiety as emerging from interactions among platform features, content, social environment, developmental stage, individual differences, goals, and existing vulnerabilities. Orben, Meier, Dalgleish, & Blakemore, 2024

Social comparison and self-evaluation

Social comparison on social media becomes especially salient because platforms make other people's achievements, appearance, relationships, travel, popularity, and lifestyles unusually visible. Upward comparison can sometimes motivate, but it can also intensify self-evaluation, uncertainty, or perceived inadequacy. The effect depends on the content, the person's interpretation, and the relevance of the comparison domain.

The mechanism is especially important when feeds repeatedly present curated or idealized content that touches existing concerns about appearance, success, belonging, or social status. Experimental evidence is stronger for some short-term self-evaluative outcomes than for broad claims that comparison inevitably produces an anxiety disorder. Cabeza Martínez, Cortesi, & Vlachos, 2026

Social evaluation, feedback, and uncertainty

Likes, comments, views, reply timing, follower counts, public metrics, and ambiguous silence can become social-evaluative cues. For a person already sensitive to rejection or negative evaluation, these cues may increase monitoring and worry. The psychological issue is not the existence of a number on a screen by itself; it is what that cue means to the person and how much attention, reassurance seeking, or avoidance becomes organized around it.

Fear of missing out and repeated checking

Fear of missing out (FOMO), can connect uncertainty about social inclusion with repeated checking. A person may check because they feel anxious about missing information or social activity, receive temporary reassurance, and then return when uncertainty rises again. That pattern is consistent with reinforcement-based accounts of checking, although FOMO is a research construct rather than a diagnosis.

Importantly, FOMO can be both an antecedent and a correlate of digital behavior. Its relationship with anxiety is not reducible to a one-way sequence in which social media simply creates FOMO and FOMO creates anxiety.

Harmful interactions, harassment, and cyberbullying

Online harassment, humiliation, exclusion, discrimination, threats, rumor spreading, and cyberbullying are qualitatively different exposures from ordinary social-media use. They can plausibly contribute to distress and anxiety, and they deserve direct attention instead of being averaged into a generic screen-time number.

The U.S. Surgeon General's advisory emphasizes that social media can provide meaningful benefits while also exposing young people to harmful content and interactions. The advisory recommends focusing on safety, developmental vulnerability, content, and context rather than treating all use as equivalent. U.S. Surgeon General, 2023

Sleep disruption and late-night use

Social media can intersect with anxiety through sleep when use delays bedtime, interrupts sleep, or keeps a person engaged during periods when they intended to disengage. Sleep is a separate health domain, so the relevant point here is narrower: anxiety-related outcomes can be partly shaped by what digital behavior displaces or disrupts, not only by the content viewed.

A 2024 systematic review with meta-analyses found that problematic social media use was associated with sleep problems as well as anxiety and depression, while results for general social media use were less consistent. Ahmed et al., 2024

Information load, vigilance, and negative content

Some feeds repeatedly expose users to conflict, crisis, threat, alarming news, or emotionally charged content. For people who are already anxious, this can interact with vigilance and threat monitoring. The relevant construct is the person's exposure pattern and response, not a claim that a platform has “rewired” the brain or caused a universal loss of attention.

Doomscrolling is a related behavioral pattern, but it has its own canonical ownership in the Digital Life & Attention cluster. This article therefore treats negative-content scrolling as one possible context within the broader social-media/anxiety relationship rather than absorbing doomscrolling-specific evidence.

Social Anxiety and Social Media

Social anxiety deserves separate attention because social platforms can simultaneously reduce and increase social-evaluative pressure. Text-based or asynchronous communication may feel easier than face-to-face interaction for some people. It can offer more time to compose a response, control self-presentation, or connect with others from a lower-pressure environment. For others, public feedback, comparison, monitoring, and uncertainty about how they are perceived can intensify social-evaluative concerns.

The 2024 meta-analysis by Wu and colleagues found a strong positive correlation between social anxiety and problematic social media use across 53 studies. That result is compatible with several directions: social anxiety may predispose some people toward dysregulated use, problematic use may maintain or amplify social-evaluative concerns, or both may be influenced by additional variables. Wu, Huang, & Yang, 2024

This is also why “social media anxiety” should not be treated as a formal diagnosis. A person may experience anxiety in relation to posting, messaging, public feedback, comparison, or online conflict, but diagnosis depends on the broader pattern, duration, severity, and impairment.

Are Teenagers More Vulnerable?

Adolescence is a period of heightened importance for peer relationships, identity development, social evaluation, and belonging. Social media can extend these developmental processes into persistent digital environments. That makes adolescence scientifically important without implying that every teenager is harmed by the same platform in the same way.

The 2026 umbrella review of adolescent research found weak and inconsistent associations for general use and more consistent links between problematic use and ill-being. Tølbøll, 2026 The 2025 annual review likewise argued that adolescent social-media use should not be treated as a monolith because content, features, functions, and individual differences matter. Maheux et al., 2025

Age also changes what social media means. A platform can function as a place for peer belonging, creative expression, identity exploration, school coordination, activism, entertainment, relationship maintenance, or exposure to social pressure. Those experiences can coexist.

Active vs Passive Use: Is Scrolling the Problem?

Popular advice often divides social-media behavior into active use, such as posting or messaging, and passive use, such as scrolling. Research does not support a simple rule that active use is healthy and passive use is harmful.

A 2024 meta-analysis of 141 studies and roughly 145,000 participants found that most associations between active or passive use and mental-health or well-being outcomes were negligible. Some associations varied by age, social-media context, and the type of active use. Active use was also associated with greater perceived online support. Godard & Holtzman, 2024

A more useful approach asks what the person is doing, with whom, around what content, for what purpose, and with what emotional consequence. Passive viewing of supportive community content is psychologically different from passive comparison with idealized appearance content. Active messaging with a trusted friend is different from repeatedly posting for reassurance.

How Much Social Media Is Too Much for Anxiety?

There is no universal adult medical cutoff at which social media becomes pathologically excessive or anxiety-producing. Research frequently measures hours per day, but time is only one exposure variable. A threshold observed in one adolescent study should not be converted into a universal diagnostic line.

The U.S. Surgeon General has highlighted observational findings in which more than three hours per day among young people was associated with greater risk of mental-health problems, including anxiety and depression symptoms. That figure is a population-level association from specific studies, not a diagnostic threshold and not proof that crossing three hours causes an anxiety disorder. U.S. Surgeon General, 2023

Recent longitudinal work reinforces this caution. In the 2026 Swedish cohort, self-reported time at age 15 or 16 did not predict anxiety and depression five years later after adjustment for confounders. Zetterqvist et al., 2026

For an individual, functional questions are often more useful than a single hour count: Is use repeatedly displacing sleep or obligations? Is it creating conflict? Does the person feel unable to stop despite wanting to? Does checking become organized around reassurance? Do specific feeds or interactions predictably trigger distress? Is anxiety interfering with daily functioning?

Anxiety Symptoms Are Not the Same as an Anxiety Disorder

A study can report an association with anxiety symptoms without showing an increase in clinically diagnosed anxiety disorders. Many studies use self-report questionnaires, and elevated scores can reflect distress without establishing a diagnosis.

Clinical anxiety disorders involve patterns of anxiety, fear, or worry that are persistent, excessive, and impairing. Diagnosis requires evaluation of symptoms, duration, context, functional impact, and other possible explanations. NIMH, Anxiety Disorders

This distinction matters for search terms such as “social media anxiety.” The phrase can describe a real experience of anxiety connected with online life, but it is not itself a standard standalone DSM or ICD diagnosis.

Can Deleting Social Media or Taking a Break Reduce Anxiety?

Temporary abstinence or reduction can help some people, but it should not be presented as a universal treatment. Trials differ in whether participants completely abstain, reduce time, stop selected platforms, or change only certain behaviors. Outcomes also differ by duration, population, adherence, and what replaces the time.

A 2025 meta-analysis of 20 randomized controlled trials found a small overall improvement in well-being from social-media detox interventions, with substantial heterogeneity across studies and moderation by cultural background and intervention duration. Liu et al., 2025

That evidence supports experimentation with boundaries when use feels unhelpful; it does not show that complete deletion is medically necessary, that abstinence works for everyone, or that stopping social media automatically treats an anxiety disorder. A useful intervention targets the person's actual problem: comparison, checking, harassment, nighttime use, conflict, information overload, or loss of control.

Practical Ways to Reduce Anxiety-Related Social Media Friction

Identify the specific trigger instead of blaming the entire medium

For a week, notice when anxiety rises in relation to social media. The most informative details are often the platform, content, interaction, time of day, motive for opening the app, and what happened afterward. A pattern may emerge around appearance content, work updates, political conflict, relationship uncertainty, read receipts, public metrics, comparison, news, or nighttime checking.

Change the feed before changing your whole life

Unfollow, mute, block, or reduce exposure to accounts and topics that reliably intensify distress without providing meaningful value. Increase direct access to people and communities that provide support, useful information, or genuine enjoyment. This targets content and context rather than assuming all social-media minutes are equivalent.

Reduce cues that trigger automatic checking

If repeated checking is the problem, notification settings, badge counts, lock-screen previews, app placement, and scheduled check windows can change how often a cue interrupts another activity. The goal is not to prove that notifications are inherently harmful; it is to make use more intentional when cue-driven checking is reinforcing anxiety.

Protect activities that anxiety and scrolling are displacing

Sleep, face-to-face relationships, exercise, school, work, meals, and restorative downtime can be crowded out when social-media use expands into every available gap. Protecting those activities can matter even when total daily screen time changes only modestly.

Experiment with reduction rather than treating detox as a cure

A short reduction experiment can help answer a practical question: Does this particular pattern contribute to my anxiety? Compare mood, worry, sleep, concentration, and urge to check before and during the experiment. If anxiety remains high, that is useful information too; the underlying problem may need broader attention.

Use clinical treatment for clinical anxiety

If anxiety is persistent, difficult to control, or impairing school, work, relationships, sleep, or daily functioning, a mental-health professional can assess the broader pattern. Evidence-based anxiety treatment is aimed at the anxiety disorder and its maintaining processes; social-media habits can be incorporated when they are relevant.

When Social Media May Be Helping Rather Than Hurting

Social media can provide social support, access to communities, identity exploration, information, creative expression, and connection with people who share uncommon experiences. These benefits matter because reducing social media can also remove valued social resources.

The strongest contemporary reviews describe both risk and benefit pathways and emphasize individual differences. Orben et al., 2024 A person whose anxiety decreases when they can communicate asynchronously with supportive peers may have a very different digital profile from someone whose use revolves around comparison, harassment, and reassurance checking.

The practical target is therefore a healthier relationship with digital social environments, not a moral judgment about being online.

What the Evidence Does Not Show

Current research does not justify treating all social-media use as harmful, diagnosing an anxiety disorder from screen time, assuming that frequent use equals addiction, or claiming that a particular platform has “rewired” every user's brain. It also does not justify turning correlational findings into one-way causal statements.

Research also does not support the idea that every anxious social-media user should quit. The same person can encounter support and stress on the same platform, and the same feature can have different effects depending on context.

The most defensible conclusion in 2026 is narrower and more useful: social media is one environment in which anxiety-relevant processes can occur. Problematic or dysregulated use and harmful experiences show more consistent associations with anxiety than raw time alone, while causal direction varies across studies and individuals.

Frequently Asked Questions

Can social media cause anxiety?

Certain social-media experiences can plausibly contribute to anxiety for some people, and some longitudinal or experimental evidence supports effects in specific contexts. The overall evidence does not establish a universal causal rule that social media causes anxiety disorders. General use shows small or mixed associations, while problematic use and harmful experiences show stronger and more consistent links.

Why does social media make me anxious?

Common possibilities include social comparison, fear of missing out, public evaluation, uncertainty about replies, repeated checking, harassment, negative content, information overload, sleep-displacing use, and existing social anxiety. The useful question is which mechanism is operating for you.

Is social media anxiety a real diagnosis?

“Social media anxiety” is a descriptive phrase, not a standalone standard clinical diagnosis. Anxiety related to social-media experiences can be psychologically real. A clinician would evaluate whether symptoms fit an established anxiety disorder or another condition and whether they cause significant distress or impairment.

Does social media cause social anxiety?

Research shows a strong correlation between social anxiety and problematic social media use, but directionality is not settled. Socially anxious people may use platforms differently, some digital experiences may reinforce social-evaluative concerns, and both processes may occur together.

How many hours of social media are safe?

There is no universal adult medical cutoff. For adolescents, public-health guidance sometimes highlights observational thresholds associated with greater risk, but those are not diagnostic lines. Content, context, problematic use, developmental stage, and functional impact matter alongside time.

Will deleting social media cure anxiety?

No universal treatment effect has been established. Randomized trials of social-media reduction or abstinence show small average improvements in well-being, with substantial variation among studies and people. If someone has an anxiety disorder, social-media changes may be one part of a broader evidence-based treatment plan rather than a substitute for assessment and care.

Is passive scrolling worse than active use?

Not consistently. A large meta-analysis found that most associations for active and passive use were negligible and context-dependent. What is being viewed or shared, why the person is using the platform, and what psychological process is engaged can matter more than the active/passive label alone. Godard & Holtzman, 2024

Should parents focus on screen time?

Time can be useful as one signal, especially when social media displaces sleep, school, exercise, relationships, or other important activities. It should be combined with questions about content, online experiences, problematic use, developmental needs, and how the young person is actually feeling and functioning.