Dry Eye Research in 2017: 32 Notable Human Studies

The 2017 picture

What the research shows

The 2017 studies made dry eye more concrete as a common, varied condition: millions of U.S. adults reported a diagnosis, but symptoms and clinical signs still diverged in many patients. Large masked drug trials showed distinct results: cyclosporine improved staining without a clear average discomfort gain, while lifitegrast improved an eye-dryness score but not a separate discomfort measure.

Treatments and practical interventions

Trial design varies widely here, from large masked comparisons to small proof-of-concept studies. A change from baseline alone is weaker evidence than a difference against a well-matched control.

1. Cyclosporine emulsion versus vehicle

   The double-masked, multicenter SICCANOVE trial randomized roughly 490 people with moderate to severe dry eye to 0.1% cyclosporine cationic emulsion or vehicle for six months. Corneal staining improved more with cyclosporine, but average ocular discomfort, a co-primary outcome, improved similarly in both groups. A favorable post hoc result in a severe subgroup needs cautious interpretation.

2. One-year follow-up of cyclosporine for severe keratitis

   A report from the SANSIKA trial followed 177 completers through six additional months after its masked comparison, with everyone then receiving 0.1% cyclosporine emulsion. Staining, inflammatory markers, and reported symptoms improved over the year, and no unexpected safety signal appeared. Because the extension had no untreated group, its later gains cannot be attributed solely to continued cyclosporine.

3. Intense pulsed light and tear inflammation

   A double-masked randomized comparison studied three intense-pulsed-light sessions for meibomian gland dysfunction against sham exposure. Tear IL-17A and IL-6 fell more with active treatment, and prostaglandin E2 differed at week 12; several marker changes tracked ocular-surface measures. This supports a biological effect under the tested regimen, while biomarkers alone do not establish lasting symptom relief.

4. Semifluorinated alkane drops in gland-related dry eye

   A prospective multicenter observational study gave perfluorohexyloctane drops to 72 patients with meibomian gland disease for six to eight weeks; 61 completed treatment as directed. Tear breakup, staining, gland expressibility, and symptoms improved from baseline. The absence of a comparison arm leaves placebo effects, concurrent care, and natural variation unresolved.

5. Omega-3 and corneal nerve imaging: a pilot subset

   Investigators imaged the corneal nerves of just 12 participants from a larger randomized omega-3 trial, eight receiving supplement and four placebo. Several nerve, symptom, and osmolarity measures favored omega-3 after 90 days. With only four controls and multiple outcomes, this is exploratory evidence rather than a reliable estimate of treatment benefit.

6. A cooling-receptor agonist for discomfort

   After laboratory testing, researchers tried a topical TRPM8 agonist on 30 people with mild dry eye and then studied two weeks of repeated application in 20. Cooling and tear secretion increased, and discomfort questionnaires improved without reported irritation. These small early human tests support further study; they do not establish routine clinical effectiveness.

7. Chitosan–N-acetylcysteine and tear-film thickness

   A double-masked controlled study first compared a single chitosan–N-acetylcysteine drop with contralateral saline in 21 patients, then explored once- versus twice-daily use for five days in 17. Optical coherence tomography showed a thicker tear film after active dosing. The physiological result and very short follow-up cannot yet establish durable symptom or surface benefit.

8. A desktop humidifier during computer work

   In a masked randomized crossover study, 44 computer users worked for one hour with and without a desktop humidifier on separate days. A modest humidity increase was accompanied by longer noninvasive tear breakup and more reports of comfort, while lipid grade and tear meniscus height did not differ. The trial tests short exposure during screen work, not persistent dry-eye treatment.

9. Manuka honey added to usual gland care

   An open-label randomized trial assigned 114 people with moderate to advanced gland-related evaporative dry eye to usual care alone or usual care plus one of two Manuka honey formulations. All groups improved on several measures; some staining and gland outcomes favored honey. Open labeling, multiple endpoints, and a short eight-week assessment limit certainty about the incremental benefit.

10. Royal jelly supplementation in a small placebo trial

    Forty-three adults with dry-eye symptoms were randomized, double-masked, to oral royal jelly or placebo for eight weeks. Tear secretion increased with royal jelly, especially among those with lower starting Schirmer values; the publication also reported separate animal experiments. The small clinical sample and short course do not justify treating the supplement as established dry-eye care.

11. Intranasal stimulation and conjunctival goblet cells

    A double-masked randomized crossover experiment compared intranasal tear neurostimulation with sham stimulation in 15 people, ten with dry eye and five controls. Impression cytology showed more conjunctival goblet-cell degranulation after active stimulation. The mechanistic response is promising, but the study was too small and brief to show sustained clinical relief.

12. Lifitegrast for eye dryness in the OPUS-3 trial

    OPUS-3 randomized 711 adults to twice-daily lifitegrast 5% or placebo for 12 weeks under double masking. The eye-dryness rating improved more with lifitegrast by day 14 and at the primary day-84 assessment, while a separate ocular-discomfort score did not differ significantly. The result supports a specific symptom benefit, not improvement on every symptom measure.

Who was affected and how dry eye was experienced

Large surveys offer scale, while smaller clinical samples reveal effects on work, pain, and other conditions. Their populations and disease definitions differ substantially.

13. Diagnosed dry eye among U.S. adults

    Using 75,000 responses to a U.S. health survey, researchers projected that 6.8% of adults, about 16.4 million people, had diagnosed dry eye. Estimates rose with age and were higher in women than men. Because the survey asked about prior diagnosis, it does not count everyone with symptoms or confirm each diagnosis by examination.

14. Dry-eye symptoms in a Saudi Arabian sample

    A multistage survey of 1,858 adults attending primary-care centers in Al-Ahsa estimated frequent dry-eye symptoms in 32.1% after age adjustment. Female sex, older age, smoking, and diabetes were associated with symptoms. This symptom-based regional estimate is not directly comparable with U.S. diagnosed-disease prevalence and cannot identify causes.

15. Eye symptoms during the office workday

    Two online surveys in three offices received 505 and 213 responses. Dry-eye-related complaints were more frequent at work than at home, and many respondents reported some interference with daily work. Self-selected responses and no clinical examination limit estimates of disease prevalence or the effect of any particular office condition.

16. Serum metabolites and dry-eye symptoms

    A metabolomics analysis in 2,819 TwinsUK participants tested 222 known serum metabolites against questionnaire-defined dry eye. Lower androsterone sulfate and epiandrosterone sulfate showed strong corrected associations, especially with symptoms. The observational finding is biologically interesting but does not show that changing androgen levels would prevent or treat dry eye.

17. How gland-disease definitions affect prevalence

    A Japanese clinic study examined 510 adults aged 50 or older who were scheduled for cataract surgery. With a definition requiring both lid-margin abnormalities and blocked gland openings, 47.5% met criteria for meibomian gland dysfunction and 18.0% had symptomatic disease. This selected surgical population and the chosen definition shape the estimate.

18. Dry eye and satisfaction after cataract surgery

    Researchers compared 27 satisfied and 27 dissatisfied patients two months after uncomplicated cataract surgery. Visual acuity and most dry-eye signs were similar, but dissatisfied patients reported worse dry-eye symptoms, visual function, and health anxiety. The small cross-sectional comparison cannot determine which factor caused dissatisfaction.

19. Pain severity does not track all surface signs

    Among 84 clinic patients with dry eye, ocular pain was common and correlated with symptom questionnaire scores. Pain severity did not clearly correlate with tear breakup or corneal staining, and antidepressant use was associated with more pain in adjusted analysis. Medication use may mark other conditions; this study cannot establish its causal role.

20. Type 2 diabetes and ocular-surface changes

    An Indian case-control study compared 53 people with type 2 diabetes with 30 healthy controls. The diabetes group had less tear production and stability, more staining, and cytology evidence of goblet-cell loss. The clinic-sized observational comparison supports closer study of the diabetic ocular surface but cannot separate diabetes from all related health differences.

21. Which patients still had severe symptoms a year later?

    A prospective study followed 120 U.S. veterans with dry-eye symptoms for one year. About three-quarters of those with severe symptoms at baseline remained in that category, and baseline pain, sleep, and mental-health measures were associated with later symptom severity. The mostly male veterans sample and observational design limit generalization and cannot show that those factors caused persistent dry eye.

Clinical tests, nerves, and inflammation

These studies investigate specific pieces of the dry-eye picture. Their measurements add useful context, but none replaces an assessment of symptoms, lids, tear film, and the ocular surface together.

22. Patterns of fluorescein tear breakup

    A cross-sectional study classified tear-breakup patterns in 106 dry-eye patients and compared symptoms, tear meniscus, lipid appearance, staining, and Schirmer results. Several measures differed across the five pattern groups, suggesting that breakup appearance may reflect different tear-film problems. The proposed classification needs validation between examiners and in independent patients.

23. Corneal nerves and immune cells under the microscope

    Confocal imaging compared eyes with Sjögren dry eye, non-Sjögren dry eye, and healthy controls. Corneal nerve density was lowest and dendritic-cell density highest in the Sjögren group, alongside greater nerve tortuosity in disease groups. The relatively small comparison shows structural associations, not whether nerve or immune changes came first.

24. Persistent pain after surface anesthesia

    A cross-sectional study of 224 veterans with dry-eye symptoms used topical anesthesia and pain-sensitivity testing at the forehead and forearm. People whose ocular pain persisted after anesthesia reported more neuropathic-like symptoms and greater sensitivity outside the eye, while most tear signs did not differ. The predominantly male clinic sample limits generalization.

25. Fluorescein volume changes tear-breakup readings

    In a prospective crossover measurement study of 41 adults, conventional-volume fluorescein produced shorter tear-breakup times than either a tiny fluorescein dose or noninvasive measurement. Minimal-dose and noninvasive results were similar. The finding shows that the testing method itself can affect a dry-eye sign and argues for consistent technique when tracking patients.

26. Tear cytokines in patients with depression

    Investigators compared tears from 32 patients treated for depression with 34 controls. IL-6, IL-17, and TNF-alpha were higher in the depression group; IL-17 and TNF-alpha tracked dry-eye severity measures. Depression, its treatment, and other differences could each contribute, so the study cannot identify a single mechanism.

27. Why central corneal staining matters for vision

    A visual-function comparison involved 52 dry-eye patients and 20 controls, all with standard corrected acuity better than 20/20. Dry-eye patients with central punctate staining had lower contrast sensitivity than those without central staining, even though ordinary acuity remained good. The cross-sectional result helps explain visual complaints but does not show that treating staining restores contrast.

28. HLA-DR as a surface inflammation marker

    An analysis pooled baseline data from three studies involving 311 patients with moderate to severe dry eye. Conjunctival HLA-DR correlated modestly with corneal staining and more weakly with symptoms and tear tests; it fell with cyclosporine versus vehicle in two studies. These correlations support research use of the marker, not a stand-alone clinical diagnosis.

29. Computer-assisted grading of corneal staining

    Clinicians graded photographs of corneal staining with a conventional scale or a computer-assisted index. Interobserver agreement was higher with the index, and the systems placed many images into different severity ranges. Because the comparison used photographs of epitheliopathy rather than patient outcomes, better numerical agreement still needs clinical validation.

30. Distinct conjunctival T-cell patterns in dry eye

    Using impression cytology and flow cytometry, investigators identified resident and circulating T-cell subsets in the human conjunctiva. Patients with dry eye clustered into two immune patterns associated with redness or shorter tear breakup. The observational analysis suggests biological diversity but does not yet define a validated test or subtype-specific treatment.

31. Bacteria and meibomian gland dysfunction

    Cultures from 201 patients with gland dysfunction and 84 matched controls found more positive aerobic and anaerobic samples in the disease group, both at gland openings and on the conjunctiva. The finding is an association: ordinary ocular bacteria were present in controls too, and culture results alone do not establish infection or a need for antibiotics.

32. Corneal nerves and tear substance P in diabetes

    A small cross-sectional comparison of nine people with diabetes and 17 controls found lower corneal nerve density and less tear substance P in diabetes, with the two measures correlated across participants. This suggests a possible nerve-related tear marker, but the 26-person sample is far too small to establish a useful diagnostic threshold.