Why Screen Time Has Been Hard to Study

Screens are near the top of almost every self-reported trigger list. People with migraine describe fluorescent office monitors, late-night phone scrolling, and long video calls as reliable ways to bring on an attack, and clinicians hear it constantly. Despite that, screen exposure has never had a proper evidence synthesis behind it. The literature has been a scattered collection of student surveys, occupational studies, and small clinic samples, each pointing roughly the same direction without anyone pooling them.

A new systematic review and meta-analysis in the Journal of Neurology has done that pooling. A team at the National Institute of Mental Health and Neurosciences in Bengaluru searched five databases without date restrictions, screened everything in duplicate, assessed risk of bias, and rated the certainty of the result using GRADE.

How the Review Worked

The protocol was registered on PROSPERO before the analysis, and the review followed PRISMA 2020 reporting guidance. The team searched Embase, MEDLINE, PubMed, Scopus, and Google Scholar with no date limits, restricted to English-language observational and interventional studies of adolescents or adults with migraine or primary headache that included a quantifiable measure of screen exposure.

Two reviewers independently screened records and extracted data. Risk of bias was assessed with a modified Joanna Briggs Institute and AHRQ checklist for cross-sectional and observational designs. Pooling was done with a random-effects model, and the team ran the standard robustness checks: heterogeneity statistics, Egger's regression for small-study effects, and a leave-one-out sensitivity analysis. Anything that could not be pooled was described narratively.

What They Found

The pooled odds ratio of 1.60 was consistent across age groups. Adults came in at 1.87 (95% CI 1.43 to 2.44) and children and adolescents at 1.46 (95% CI 1.24 to 1.72). Removing any single study one at a time moved the pooled estimate only between 1.52 and 1.67, so no individual study is carrying the result.

The twelve studies that could not be pooled, which reported correlations, regression coefficients, or group comparisons instead of odds ratios, mostly pointed the same way. Two clinic-based comparative studies were exceptions and found no significant difference in headache frequency or severity by smartphone use.

The prediction interval and the Egger's test point in opposite directions.

The Egger's test result points the other way, toward the estimate being inflated. An intercept of 3.71 with p = 0.002 suggests small-study effects, which in practice often means small studies reporting large associations got published while small studies finding nothing did not. When that pattern is present, the pooled figure is likely to be higher than the truth.

Direction of the Association

Migraine causes photophobia. Photophobia is not limited to attacks: many people with migraine have heightened light sensitivity between attacks as well. A bright, flickering, high-contrast screen is exactly the kind of stimulus that gets uncomfortable first.

In a cross-sectional survey, that produces two possible readings of the same data. High screen exposure could be raising headache burden. Or people whose headache burden is already high could be reporting screens as a problem and noticing them more, while people who are relatively unbothered by screens report their use without a second thought. There is also a plausible version in which the arrow runs backwards on the exposure itself, with severely affected people cutting screen time down because of their symptoms.

The review is careful about this. Its own conclusion describes screen exposure as a modifiable and clinically assessable factor while explicitly calling for prospective, objectively measured, and interventional research to establish causality. Objective measurement matters here too, since self-reported screen time correlates only moderately with device-logged usage, and people in pain do not estimate their own behavior accurately.

The Sleep Signal

The most consistent secondary finding across the review was sleep. Screen exposure was associated with sleep disturbance in 18 of the 22 studies, a higher hit rate than visual or ocular discomfort (14 of 22), musculoskeletal symptoms (11 of 22), or psychiatric comorbidity (13 of 22). Reduced quality of life appeared in 20 of 22.

Sleep is one of the few migraine variables with a well-established bidirectional relationship and a decent evidence base for intervention. Insufficient sleep, irregular sleep timing, and poor sleep quality all associate with higher attack frequency, and behavioral sleep interventions have shown benefit in migraine populations. If part of the screen association runs through disrupted sleep, that is the component with the most established treatment options.

Proposed Mechanisms

Across all 22 studies, the mechanistic discussion converged on three pathways. All three are plausible, and none has been demonstrated to drive the association in this population.

The blue light pathway is the one that gets the most attention commercially, and it is worth flagging that blue-blocking lens trials for headache and eye strain have generally been unimpressive. A Cochrane review of blue-light filtering spectacle lenses found no meaningful short-term benefit for eye strain, and the migraine-specific tint evidence points more toward specific narrow-band filters such as FL-41 than toward generic blue blockers.

What to Do With This

Low certainty evidence can still be worth acting on when the change involved is cheap, safe, and reversible, which is the case for screen habits. An odds ratio from cross-sectional surveys is not grounds for overhauling a working life, though a short personal experiment costs very little.

Migraine varies enough week to week that informal impressions are unreliable. Four weeks of baseline data followed by eight to twelve weeks after a change is the minimum needed to see past normal fluctuation, and it is the same approach that makes any trigger experiment interpretable.

For clinicians, the reasonable reading is that screen exposure belongs in the history alongside sleep, caffeine, meals, and stress. It is quick to ask about, it is modifiable, and it opens the door to the sleep conversation, which has better evidence behind it.

Key Takeaways

The association was consistent across 22 studies and both age groups. It also comes from a design that cannot establish direction, in a condition where the outcome plausibly causes the exposure.

Screen habits are cheap to adjust, the changes carry no risk, and the sleep component has independent evidence behind it. The size of any benefit is unknown, and none of this substitutes for a preventive that is working.