Screens strain your eyes. Whether they damage them is a separate, unsettled question.

Prolonged screen use reliably produces eye strain — tired, sore, burning eyes, often with headache and neck pain — and that strain is a symptom, not an injury. A 2026 meta-analysis of 63 studies put the pooled prevalence of asthenopia at 51% across all populations studied, rising to 90% among digital device users [s1]. Whether screens also cause lasting structural change to the eye is a different question with a much weaker answer: a meta-analysis in The Lancet Digital Health found smart device use associated with myopia at an odds ratio of 1.26, with a confidence interval whose lower bound touched 1.00, and reported that not one of its 33 included studies measured screen time reliably [s2].

The strain is real, common and reversible

The asthenopia review searched PubMed to April 2024 for studies from the preceding five years and pooled 63 of them [s1]. Overall prevalence by questionnaire or symptom report was 51% (95% confidence interval 50% to 52%), with subgroup estimates of 90% among digital device users and 77% among computer workers [s1]. During the COVID-19 pandemic, prevalence rose among adults (39% to 45%), university students (36% to 57%) and school-aged children (45% to 64%) [s1].

The symptom picture is broader than most people would list. The commonest ocular complaints were eye tiredness (65%, 95% CI 46% to 84%), eye strain (47%, 95% CI 37% to 58%) and burning or irritation (43%, 95% CI 35% to 51%) [s1]. But neck pain (45%, 95% CI 28% to 62%) and shoulder pain (30%, 95% CI 12% to 48%) were nearly as prevalent, alongside headache (50%, 95% CI 41% to 59%) and difficulty concentrating (44%, 95% CI 32% to 56%) [s1]. A condition in which half of sufferers report neck pain is partly a posture and workstation problem wearing an ophthalmic label.

Among modifiable factors, the review found prolonged screen time associated with asthenopia at an odds ratio of 1.15 (95% CI 1.09 to 1.21), short sleep duration at 1.28 (95% CI 1.04 to 1.57) and prior eye disease at 2.59 (95% CI 1.43 to 4.69) [s1]. Regular breaks were associated with lower odds (OR 0.21, 95% CI 0.09 to 0.51), as were anti-glare filters (OR 0.34, 95% CI 0.19 to 0.64) and knowledge about computer use (OR 0.20, 95% CI 0.13 to 0.30) [s1].

One estimate in that list is worth reading as a caution rather than a finding. Air conditioning use carried an odds ratio of 23.02 — with a 95% confidence interval running from 4.94 to 107.18 [s1]. An interval spanning a twentyfold range is what a very small, very heterogeneous body of data looks like when it is pooled anyway. The direction may well be right; dry moving air and reduced blinking are a plausible pair. The magnitude is not usable.

The myopia question is where the interesting uncertainty is

Eye strain resolves. Myopia does not, and childhood myopia has risen from a pooled global prevalence of 24.32% in 1990 to 35.81% by 2023 across 276 studies of more than five million children [s3]. The timing invites an obvious explanation, and the evidence for it is thinner than the explanation's popularity suggests.

The Lancet Digital Health review searched MEDLINE and Embase to June 2020 for studies of smart device exposure and myopia in people aged three months to 33 years, screening 3,325 articles, including 33 in the systematic review and 11 in the meta-analysis [s2]. Screen time from smart devices alone was associated with myopia at an odds ratio of 1.26 (95% CI 1.00 to 1.60), with heterogeneity of I² = 77% [s2]. Combining smart device use with computer use raised the estimate to 1.77 (95% CI 1.28 to 2.45), with I² = 87% [s2].

Take those two numbers seriously and the picture is not one of a settled hazard. The smart-device estimate's confidence interval reaches exactly 1.00, the boundary of no effect. The heterogeneity figures mean the constituent studies disagreed with one another far more than sampling error explains. And the review's own risk-of-bias assessment is blunt about why: all 33 studies lacked reliable measures of screen time, seven (21%) did not objectively measure myopia, and nine (27%) did not identify or adjust for confounders [s2]. Screen exposure was captured by questionnaire in every single study, with only one also using device-recorded network data [s2]. The authors chose their own verb carefully — smart device exposure "might be associated" with an increased risk of myopia — and called for research using objective measures of both exposure and outcome [s2].

Why this is hard to settle, and what would settle it

The core problem is that time on a screen is confounded with almost everything else that plausibly matters. It is near work; it is time not spent outdoors; it displaces sleep; it clusters with indoor living and urban residence, and urban children already show higher myopia prevalence (28.55%) than the global pooled figure [s3]. Asking a parent to estimate a child's daily screen hours and then correlating that estimate with refraction cannot separate any of those. That is not a criticism of the individual studies so much as a description of what a questionnaire can and cannot buy.

What would resolve it is straightforward to describe and expensive to do: device-logged exposure rather than recall, cycloplegic refraction and axial length rather than self-reported spectacle wear, and enough follow-up to separate onset from progression. Until that exists, the honest summary is that screens are a well-documented cause of temporary eye strain, a plausible but unproven contributor to myopia, and not a demonstrated cause of permanent damage to the eye in their own right.

The strain question, at least, has a practical shape. The factors most strongly associated with fewer symptoms in the pooled data were taking regular breaks and knowing how to set up a workstation — not anything applied to the eye [s1]. Persistent eye pain, change in vision, or symptoms that do not settle with rest are reasons to see an optometrist or ophthalmologist rather than to adjust a monitor.