Dry Eye Research in 2015: 33 Notable Human Studies

The 2015 picture

What the research shows

The 2015 papers advanced drug development and meibomian gland treatment while revealing how different endpoints can tell different stories. Lifitegrast's symptom result contrasted with its staining result; several device and osmolarity studies found useful signals alongside measurement limits. Large surveys and clinical cohorts also showed that discomfort, visual function, and related health conditions deserve attention even when surface findings are modest. Across the collection, patient groups and chosen endpoints shaped how each result could be used.

Treatment studies

The year's interventions included large masked trials, head-to-head comparisons, and small exploratory studies. Results are most useful when both the positive and negative endpoints are read together.

1. Lipid and aqueous lubricants had similar short-term effects

   A multicenter, investigator-masked trial randomized 315 people with dry eye among two lipid-containing and two aqueous artificial tears for 30 days. One lipid formula met the prespecified noninferiority test against an aqueous comparator for symptom-score change; the four products showed no consistent, clinically important separation across other outcomes. Short follow-up and active comparators limit conclusions about lasting benefit.

2. Short-course azithromycin compared with doxycycline for gland disease

   In an open-label randomized study, 110 patients with meibomian gland dysfunction that had not improved with conservative care received five days of oral azithromycin or one month of doxycycline, while continuing lid care and tears. Symptom improvement was similar, but redness, staining, and overall response favored azithromycin at two months. Different treatment durations, unmasked patients, and brief follow-up limit the comparison.

3. Lifitegrast in the OPUS-2 phase 3 trial

   After a placebo run-in, 718 adults were randomized to lifitegrast or placebo drops for 12 weeks in a double-masked trial. Reported eye dryness improved more with lifitegrast, but the other co-primary outcome, inferior corneal staining, did not. Ocular adverse events were more frequent with lifitegrast, and the mixed primary results prevent a claim of broad improvement in both symptoms and surface damage.

4. An early intense pulsed light case series

   A retrospective clinic series followed 91 people with severe dry eye attributed to meibomian gland dysfunction who received repeated intense pulsed light and gland expression. Most of the 78 with available breakup-time data improved, and many reported satisfaction. Because the two procedures were combined and there was no control group, the series could not isolate the light treatment's effect.

5. Intense pulsed light in a paired-eye comparison

   Twenty-eight participants received intense pulsed light around one eye and placebo treatment around the other in a masked prospective study. By day 45, the treated side had a better lipid-layer grade and longer noninvasive tear breakup, while evaporation did not clearly differ. The paired-eye design strengthens the short-term comparison, but the small sample and shared patient-level symptoms limit broader conclusions.

6. Diquafosol versus sodium hyaluronate in Asia

   A randomized study assigned 497 patients in China and Singapore to diquafosol 3% or sodium hyaluronate 0.1% for four weeks. Diquafosol was no worse on fluorescein staining and did better on rose bengal staining, while tear breakup improved similarly. Mild drug reactions were more frequent with diquafosol, and an active-treatment comparison cannot show how either would perform against no treatment.

7. Three lipid-based artificial tear formulations

   In a three-month, double-masked trial, 288 people with dry-eye signs and symptoms received one of three lipid-based lubricants. Both new formulations met the study's noninferiority target against an existing product for symptom relief; one had better tolerability and some secondary measures. With no inactive control, the study chiefly compares formulations rather than the absolute benefit of artificial tears.

8. Omega-3 supplements for contact lens dryness

   A six-month, multicenter, double-masked trial randomized 496 contact lens wearers with dryness to omega-3 capsules or corn-oil placebo. Reported symptoms and lens comfort improved more in the omega-3 group, with some tear-film and cytology measures also favoring supplementation. This was a specific lens-wearing population, and one trial cannot settle whether oral omega-3 helps dry eye more generally.

9. Tacrolimus drops for Sjögren-related dry eye

   Twenty-four patients with Sjögren dry eye were randomized to tacrolimus 0.03% drops or an almond-oil vehicle for 90 days. Staining improved over time in the tacrolimus group, while the vehicle group showed no clear within-group change on the reported tests. The small groups and emphasis on change from baseline make the size of a between-treatment benefit uncertain.

10. Steroid versus cyclosporine after stem-cell transplant

    Seventy-five patients were randomized before hematopoietic stem-cell transplant to loteprednol or cyclosporine drops and followed for a year. Dry eye developed or progressed frequently in both arms, with no statistically clear difference in incidence or progression. Both groups received active treatment, so the trial cannot show whether either prevented dry eye compared with no prophylaxis.

11. Additional lubricant drops after cataract surgery

    Across 20 Italian centers, 282 cataract patients were assigned usual postoperative drops with or without a hyaluronate and carboxymethylcellulose lubricant. At five weeks, tear breakup and some symptom and staining measures favored the extra lubricant, while OSDI scores improved in both groups without a clear difference. Open treatment and short follow-up limit the size and durability of the effect.

12. Fluorometholone versus cyclosporine for Sjögren dryness

    Forty patients with Sjögren syndrome were randomized to fluorometholone plus hyaluronate or cyclosporine plus hyaluronate for eight weeks. Both groups improved on several signs and symptoms; the steroid arm improved earlier on selected measures. The small active-treatment comparison offers no untreated benchmark and is too short to assess longer-term steroid safety.

Risk, surgery, and daily experience

Surveys and clinical cohorts connected dry eye with mental health, work, sleep, and postoperative recovery. Most show associations or describe a course over time; they cannot assign a single cause to a symptom.

13. Dry eye and psychological distress in Korean women

    A population-based survey of 6,655 Korean women found that those reporting a dry-eye diagnosis or symptoms also reported more stress and depressive problems after adjustment for measured factors. About 12% reported a diagnosis and 20% symptoms. Because both eye and psychological information were collected at one time, the study cannot establish which condition came first or whether one caused the other.

14. Dry-eye symptoms, pain, and PTSD in a veterans' clinic

    Among 136 patients at a Veterans Affairs eye clinic, dry-eye questionnaire scores tracked nonocular pain and post-traumatic stress more closely than measured tear-film signs. The findings help explain why distress and surface tests can diverge. The mostly older, male clinic sample and cross-sectional design do not make ocular disease irrelevant or establish the source of an individual's symptoms.

15. Recovery after cataract surgery in patients with chronic dry eye

    A retrospective study followed 96 cataract patients who already had chronic dry eye, with records extending as far as two years after surgery. Symptoms, staining, tear breakup, and Schirmer measurements worsened early, then generally returned near preoperative levels by about three months. Without a nonsurgical comparison, the study describes recovery after this operation rather than the precise effect caused by surgery.

16. Chronic dryness after PRK and LASIK

    A prospective, nonrandomized study followed 143 active-duty service members after PRK or LASIK. At one year, chronic dry eye by the study's criteria occurred in 5.0% after PRK and 0.8% after LASIK; lower preoperative Schirmer results and more staining were warning signs in some analyses. The young military sample and nonrandom procedure choice limit direct comparison of the two operations.

17. Occupation and reported dryness in a national survey

    A Korean national health survey analyzed 6,023 adults and found dry-eye symptoms in 16%. Several office and skilled manual occupational groups reported more symptoms than the study's reference group after adjustment. The broad job categories and symptom-based outcome do not identify which workplace exposures matter or prove that a particular job causes dry eye.

18. Short sleep and dry-eye symptoms

    In a survey of 15,878 Korean adults, people reporting five hours or less of sleep also reported dry-eye symptoms more often than those sleeping six to eight hours, even after adjustment for several demographic and lifestyle factors. Sleep and symptoms were self-reported at the same time, so the study cannot say whether short sleep contributes to dryness or discomfort disrupts sleep.

19. Workplace screening in Japanese computer users

    Researchers screened 369 Japanese computer-development workers with symptom questions, an eye-opening task, and ocular tests. Nearly 60% met the study's definite-or-probable dry-eye definition, most in the probable category; simple symptom questions identified many of them. The proposed screening approach was assessed in the same single workplace and needs outside testing before wider use.

20. Dry eye after two cataract-surgery methods

    A prospective, nonrandomized study compared 137 patients receiving femtosecond laser-assisted cataract surgery or conventional phacoemulsification. Both groups had worse dry-eye measures during the first month, and the laser group had more staining and some early symptom differences. The short follow-up and procedure selection by clinical practice limit causal claims about the laser itself.

21. Corneal nerves and dryness in chronic migraine

    Confocal imaging compared corneal nerves in 19 people with chronic migraine and 30 controls; nerve fiber density was lower in the migraine group, and all migraine participants reported symptoms consistent with dry eye. This small case-control study raises a question about shared sensory pathways. It does not determine whether migraine, tear dysfunction, or another factor produced the nerve difference.

How dependable were the tests?

New imaging and automated approaches promised easier measurement, but their performance depended on how dry eye was defined and which patients were sampled. Several studies usefully reported disagreement or variability rather than a simple diagnostic success.

22. Tear osmolarity and cataract lens calculations

    A three-practice observational study repeated keratometry in 50 cataract candidates with high tear osmolarity and 25 with normal readings. The high-osmolarity group had more variable corneal measurements and predicted lens powers, even when patients did not necessarily report dryness. The selected groups and observational design show a planning concern, not proof that treating osmolarity improves surgical outcomes.

23. A short symptom scale compared with the OSDI

    In 114 clinic patients, researchers compared the two-item SANDE visual scale with the 12-item OSDI at baseline and follow-up. Scores were moderately correlated and their average levels were similar after normalization, but changes over time agreed less closely. The shorter scale may save time, while individual results should not be treated as interchangeable with OSDI scores.

24. Repeated tear-osmolarity readings varied widely

    A tertiary-center study took repeated osmolarity readings from 37 people with Sjögren syndrome, blepharitis, or neither condition. Average readings did not clearly separate the groups, and within-session measurement error was larger in the two patient groups than in controls. The small, often treated samples limit generalization, but the variability warns against relying on one reading alone.

25. Automated tear imaging did not match older tests

    A study of 296 eyes compared automated Keratograph tear breakup and meniscus measures with fluorescein-based examination in people with ocular-surface disease and controls. Automated breakup did not clearly distinguish the groups, and its meniscus values moved in the opposite direction from the traditional method. Meibography remained useful for describing gland structure, but the devices' numbers could not simply be substituted for one another.

26. Automated tear-surface breakup as a screening marker

    A prospective comparison of 28 dry-eye patients and 17 controls tested an automated, noninvasive breakup measure against tear hyperosmolarity. The shortest breakup reading separated the selected groups well and was more repeatable than manual noninvasive assessment. The cutoff and accuracy were determined in only 45 people, so they require validation in a broader clinic population.

27. Agreement between noninvasive and fluorescein breakup

    Sixty participants had repeated noninvasive Keratograph and fluorescein tear-breakup measurements at two visits. Within- and between-visit agreement was only moderate, with wide ranges of possible differences for an individual. The study shows why a change in test method can change the result even when both methods address tear stability.

Meibomian glands and ocular-surface biology

Human imaging and clinical studies connected dry eye with gland structure, corneal nerves, and inflammation. These observational patterns and small treatment subgroups are leads to investigate, not validated tests or proof of cause.

28. Baseline corneal nerves may predict short-term response

    Researchers imaged corneal nerves in 60 people with meibomian gland dysfunction before randomizing them to artificial tears, loteprednol, or a steroid-antibiotic combination for four weeks. Symptom and staining improvement was seen mainly in participants with near-normal baseline nerve length in the artificial-tear and loteprednol groups. These small subgroups do not establish a validated cutoff or prove that nerve imaging should guide treatment.

29. Tear production alongside meibomian gland loss

    A multicenter comparison studied women with meibomian gland dysfunction, aqueous-deficient dry eye, both conditions, or neither. Despite similar symptom scores in two patient groups, gland findings and tear measurements differed; within the gland-dysfunction group, greater gland loss related to higher Schirmer values. Researchers proposed compensatory tearing, but a cross-sectional correlation cannot demonstrate compensation.

30. Corneal immune cells across dry-eye subtypes

    Retrospective confocal images from 37 dry-eye patients and 20 controls showed denser, more branched corneal dendritic cells in dry eye, especially aqueous-deficient and immune-related cases. This suggests differing levels of surface immune activity. The pilot sample was small and counted some eyes from the same patients, so the proposed pattern needs independent confirmation.

31. NLRP3 inflammatory signals in patient tears and cells

    Tear and conjunctival samples from 54 patients with Sjögren dry eye, 50 with other dry eye, and 46 controls were analyzed for NLRP3 and related inflammatory products. Several signals were higher in both patient groups, often most clearly in Sjögren disease. The comparison supports an inflammatory lead but does not show that this pathway initiated disease or that blocking it would help patients.

32. Meibomian gland loss around stem-cell transplantation

    A retrospective imaging study compared eyelid glands in 86 patients with ocular graft-versus-host disease, transplant patients without it, patients before transplant, and healthy controls. Gland loss was greater in the affected group, but it was also evident before transplantation in some patients. That pattern argues for multiple possible contributors and does not validate gland imaging as a predictor of later disease.

33. Galectin-3 changes in dry-eye tears

    Tear washes from 16 dry-eye patients and 11 controls showed more galectin-3 in patient tears, with a partially broken-down form in about half of the patient samples. Conjunctival and laboratory experiments explored how the protein might be released or cleaved. The small discovery set cannot establish galectin-3 as a diagnostic test or prove its role in surface-barrier damage.