Dry Eye Research in 2026: 26 Studies on Treatments, Symptoms, and Testing

The 2026 picture so far

The treatment studies below tell different stories depending on what was measured. Some drops improved both symptoms and signs of eye-surface damage. Other interventions changed a tear-film measurement without improving symptoms more than the comparison treatment.

In these summaries, corneal staining means dye uptake that makes surface damage visible during an eye exam. Tear breakup time measures how quickly the tear film becomes unstable between blinks. Each linked heading opens a PubMed record, publisher article, or DOI page for the study.

What the research shows so far

The strongest treatment trials reported benefits on specific signs, symptoms, or both, over the periods they studied. The symptom studies show why those outcomes deserve separate attention. Diagnostic research is testing ways to measure dry eye more efficiently, while the risk-factor studies provide leads for further investigation.

Treatment and care

These randomized trials tested specific formulations and patient groups over defined follow-up periods, with results that sometimes differed between symptom scores and eye exams.

1. Acoltremon in two pivotal phase 3 trials

   Two masked trials assigned 465 and 466 adults to acoltremon or inactive drops for 90 days. At day 14, a tear-production test improved by at least 10 mm in 42.6% versus 8.2% in COMET-2 and 53.2% versus 14.4% in COMET-3. Symptoms favored acoltremon in both trials, but the prespecified symptom result reached statistical significance in only one. The trials enrolled patients with specific signs and symptoms, and follow-up lasted three months.

2. Lifitegrast drops in a phase 3 trial

   In a multicenter trial, 615 adults with moderate to severe dry eye used lifitegrast or comparison drops without the drug for 84 days. Lifitegrast improved the prespecified corneal-staining measure and several symptom scores more than the comparison drops. Follow-up lasted 12 weeks, and sponsor employees were among the authors.

3. OT202 drops in a phase 2 trial

   A multicenter masked trial compared two strengths of OT202 with vehicle over eight weeks. The 1% drops improved the primary corneal-staining score 0.82 points more than vehicle and also favored a symptom questionnaire. The 0.5% drops did not show a clear benefit. This short phase 2 result needs confirmation in a larger trial.

4. SJP-0132 dose finding: a missed primary endpoint

   In a separate Japanese trial, 344 patients were assigned to one of three SJP-0132 strengths or placebo for four weeks. None of the doses significantly improved the prespecified corneal-staining measure. Some other signs and symptoms favored treatment, but those secondary signals cannot overturn the primary null result. This trial used a different patient group from the phase 3 study below.

5. Motugivatrep drops in a phase 3 trial

   Researchers assigned 535 patients in Japan to motugivatrep or placebo for eight weeks. At week four, the drug improved a dry-eye quality-of-life score 2.4 points more than placebo, meeting the trial's primary statistical endpoint (p=0.043). The clinical importance of that difference is unclear; the benefit was no longer statistically significant at week eight, and tear-breakup and corneal-staining results did not clearly differ.

6. Reproxalap: longer-term safety

   A randomized safety study treated 757 people across six-week and 12-month cohorts; 111 reproxalap and 72 comparison-drop recipients completed the yearlong arms. It reported no treatment-related serious eye events, but more participants stopped reproxalap because of adverse events (11.1% versus 2.8%). The study cannot establish whether the drug improves symptoms.

7. SHJ002, an investigational anti-microRNA drop

   In an 85-person, double-masked phase 2 trial, 12 weeks of SHJ002 improved total corneal staining and one symptom scale more than comparison drops. A separate eye-dryness score did not reach statistical significance. The early result needs confirmation in a larger trial.

8. Two formulations of hyaluronic acid artificial tears

   A 292-person randomized trial compared thinner and more viscous 0.3% hyaluronic acid drops for 12 weeks. The thinner drops performed at least as well on corneal staining within the study's preset margin and caused less temporary blur after use. Other efficacy measures did not show better dry-eye control.

9. Cyclosporine over three years

   Researchers first treated 336 patients with severe keratitis and dry eye with cyclosporine for a year, then randomized 245 who had improved staining to continue treatment or switch to comparison drops. Ocular surface complications were uncommon in both groups through year three. The uncontrolled first year and the selection of responders limit what this study can say about long-term benefit for all patients.

10. Low-level light added to intense pulsed light

    In a 58-person sham-controlled trial for meibomian gland dysfunction, or eyelid oil-gland problems, both groups received intense pulsed light; one also received active low-level light. The added treatment improved the oily layer of the tear film at 24 weeks, but symptoms, tear stability, and staining did not improve more than with intense pulsed light alone.

11. Meibomian gland expression plus lid hygiene

    Sixty-four patients were randomized to daily lid hygiene alone or hygiene plus gland expression every two weeks; 62 were analyzed. Imaging suggested faster improvement in visible gland area at four weeks with expression, but the groups did not differ on that primary measure at weeks eight to 12. This short study cannot establish lasting gland regeneration.

12. Punctal plugs for persistent dry eye

    A 24-week trial assigned 66 patients with persistent dry eye to plugs in the upper and lower tear-drainage openings of both eyes or a sham procedure. Symptoms and corneal staining improved more with plugs, but about 39% of treated patients had a plug come loose and 33% reported skin irritation. The small, patient-masked study tested an intensive approach in a selected group.

Symptoms and disease course

How dry eye feels does not always track what an eye exam shows. These studies explore that gap and how findings can change over time.

13. DREAM: when symptoms and clinical signs disagree

    A secondary analysis followed 535 people in a dry-eye trial. Under one definition, 77% had a mismatch between symptom severity and exam findings at baseline; 46% had discordant symptom and sign patterns between months three and 12. Those percentages may differ in people outside this selected moderate to severe dry-eye group.

14. SICCA: which complaints track measured signs?

    In 3,514 adults from a registry focused on Sjögren disease, frequent artificial-tear use and blurred vision tracked measured eye-surface abnormalities more closely than pain descriptions did. This comparison does not make any single complaint a diagnostic test, and the registry population may not represent everyone with dry eye.

15. SICCA: two to three years of follow-up

    Among 427 people with established ocular surface dryness, those with Sjögren disease had greater worsening of staining, tear breakup, and tear production than those without it. Reported symptom changes did not differ significantly. This and the preceding SICCA study draw on the same registry, so they do not independently confirm one another.

16. Does time of day affect dry-eye measurements?

    In a Norwegian clinic cohort of 1,044 patients, symptoms were often worse in the morning or evening, while measured tear-film signs showed no clinically meaningful daily pattern. The cross-sectional study relied on recalled symptoms and could not account for every change in patients' daily surroundings.

Detection and measurement

Screening studies can reveal signs in people who do not report severe symptoms. New assays and questionnaires may add useful information, but performance in a development study is only a first step.

17. Eye-surface findings before cataract surgery

    At 12 centers, 1,482 cataract candidates without a previous dry-eye diagnosis received eye-surface exams. Short tear-breakup time was found in 67.6% and meibomian gland dysfunction in 66.9%, while about one-fifth reported severe symptoms. An abnormal test alone does not mean every person had dry-eye disease.

18. TeaRx, a five-biomarker tear test

    A single-center study tested the assay in 593 people, including 495 with dry eye. Its chosen model correctly flagged about 72% of those with dry eye and correctly cleared about 63% of healthy controls. The model was selected and assessed in the same study population; several authors worked for the test's developer, and independent validation is needed.

19. Tear MMP-9 as a diagnostic marker

    Researchers compared diagnostic models in 580 people with and without dry eye across three centers. Adding the inflammatory marker MMP-9 to standard symptom and exam measures improved classification. A shorter model using symptoms, Schirmer's tear-production test, and MMP-9 also performed well when each center was held out in turn. The underlying trial was funded by the assay manufacturer, and the models need testing in routine care.

20. The DEAL daily-life questionnaire

    Using answers from 312 patients, researchers refined a dry-eye questionnaire to 15 items covering daily activities, mood and sleep, eye-drop relief, and ease of use. It may capture experiences missed by existing questionnaires. Validation is preliminary, and the study did not establish how well the tool detects change after treatment.

Possible risk factors

These observational studies examine community patterns and selected patient groups, including screen use, vitamin D status, activity, and diabetes. They cannot establish that changing an exposure would prevent or treat dry eye.

21. Screen use in a one-year Beijing cohort

    Researchers examined 1,040 adults twice, a year apart, and grouped them by screen use. Higher-use groups had more dry eye at both visits. The study reported a 9.27% new-case rate over one year using eyes as the unit of analysis. Screen exposure was not randomized, so it cannot show that reducing screen use prevents disease.

22. Vitamin D deficiency in health records

    A large matched U.S. electronic-record analysis found more newly coded dry-eye diagnoses among people labeled vitamin D deficient than among matched controls (3.3% versus 2.7%). Diagnosis codes and differences the researchers could not measure limit the finding. The study does not show that vitamin D supplements prevent or treat dry eye.

23. Physical activity measured by accelerometer

    After one week of activity monitoring, researchers followed 84,065 UK Biobank participants; 726 met the study's dry-eye outcome definition. People who reached at least 150 minutes of moderate to vigorous activity a week had lower observed risk than less active people, whether that activity was spread out or concentrated in one or two days. The subgroup result was not statistically significant in men, and the overall association does not prove that exercise prevents dry eye.

24. Sedentary time and dry-eye incidence

    Another UK Biobank analysis followed 374,398 people and recorded 452 new self-reported dry-eye cases. An apparent link with sedentary time disappeared after adjustment for other factors. The self-reported outcome and relatively few cases limit how decisively to read this result. This study and the preceding activity study draw on the same research resource, so they should not be treated as independent confirmations.

25. Dry eye in a Beijing community cohort

    Researchers sampled adults in 15 Beijing communities and analyzed 1,091 baseline assessments. Using a definition that required symptoms and short tear-breakup time, age- and sex-standardized dry-eye prevalence was 27.0%. This is the baseline of the ADEC cohort used in the screen-use follow-up above, not a second independent population. The estimate also depends on the chosen definition and urban sample.

26. Corneal nerves in children with type 1 diabetes

    A hospital study compared corneal nerve images in 50 children and adolescents with type 1 diabetes, grouped by dry-eye status. Those with dry eye had lower nerve fiber density and length and fewer nerve branches. The selected sample was small, and this cross-sectional comparison cannot show whether nerve changes preceded dry eye or whether the findings apply to children without diabetes.