Screen Time and Anxiety: What the Studies Actually Measure

Few topics in mental health generate as much confident commentary as screen time and anxiety. The concerns feel intuitive — of course staring at a phone all day must be bad for mental health. But the research tells a more complicated and less dramatic story than most commentary suggests.

What the correlations actually look like

Cross-sectional studies consistently find a positive association between screen time and anxiety symptoms. More screen time is associated with more anxiety. But the size of this association is important.

A landmark analysis of over 350,000 adolescents found that the association between digital technology use and well-being was negative but tiny — comparable in magnitude to the association between wearing glasses and well-being, or between eating potatoes and well-being [1]. The effect size was approximately r = 0.04 for overall technology use, meaning it explained about 0.16% of the variance in well-being.

A meta-analysis of social network site use and psychological well-being found a pooled correlation of r = −0.06 [4]. That is a real association but an extremely small one.

Other studies find somewhat larger correlations, in the range of r = 0.10–0.15, depending on the population, the measure and the specific type of screen use [2]. But even at the upper end, screen time explains only about 2% of the variation in anxiety. For how to interpret correlation sizes, see correlation vs causation.

The causation problem

Almost all the evidence is correlational. People who are more anxious also spend more time on screens. This could mean:

Screen time causes anxiety. Possible, but hard to demonstrate. The mechanisms proposed — social comparison, sleep disruption, reduced physical activity — are plausible but not established as the primary pathways.

Anxiety causes screen time. Also plausible. Anxious people may use screens for distraction, avoidance, or because social media provides a less threatening form of social interaction than face-to-face contact.

Shared causes. Loneliness, stress, depression and socioeconomic factors could drive both increased screen use and increased anxiety.

Bidirectional effects. Both directions may operate simultaneously, as with sleep and anxiety.

An 8-year longitudinal study of 500 adolescents found no significant long-term association between social media use and mental health outcomes [3]. A large-scale longitudinal analysis found that effects between social media use and life satisfaction were tiny and bidirectional — social media predicted slightly lower well-being, but lower well-being also predicted slightly more social media use [5].

Type of use matters more than total time

One of the more nuanced findings is that what people do on screens matters more than how long they spend on them. An experimental study found that passive Facebook use (scrolling without interacting) reduced well-being, while active use (messaging, commenting) did not [6].

This passive-versus-active distinction makes intuitive sense and has some empirical support, but it’s still based on limited evidence, and the categories are rough. Someone passively scrolling may be relaxing; someone actively posting may be anxiously seeking validation.

Why earlier studies may have overestimated effects

Several factors likely inflated early estimates of the screen time-anxiety link:

Flexible analysis. With many possible measures of screen time, many possible mental health outcomes, and many possible ways to analyze the data, there are many chances to find a significant result. Preregistered studies with locked analysis plans tend to find smaller effects.

Publication bias. Studies finding dramatic effects are more publishable than null results. The publication bias problem applies here as elsewhere.

Recall bias. Most studies measure screen time by asking people to estimate it. People are poor at estimating their phone use, and anxious people may overestimate both their anxiety and their screen time.

Questions to ask about screen time research

1. Is the study cross-sectional or longitudinal?
2. How large is the effect size, and how much variance does it explain?
3. Was screen time measured by self-report or objective tracking?
4. Does the study distinguish between types of screen use?
5. Were the analyses preregistered?

What remains uncertain

Whether specific types of content — news, social comparison, doom-scrolling — have larger effects than generic screen time is understudied with rigorous methods. Whether reducing screen time actually reduces anxiety (as opposed to just correlating with it) has barely been tested experimentally. Whether the relationship is different for people with existing anxiety disorders versus the general population is not well established. The honest summary is that screens probably play some role in anxiety for some people, but a much smaller role than popular narratives suggest.

Thinking about screen time. If you feel that screen use worsens your anxiety, reducing passive scrolling and replacing it with other activities is a reasonable experiment. But it’s unlikely to be a sufficient treatment for an anxiety disorder. If anxiety is affecting your daily life, talk to a GP or mental health professional. If you are in crisis or thinking about harming yourself, contact emergency services or a crisis line now: in the US, call or text 988; in the UK and Ireland, Samaritans on 116 123.