Dry Eye Research in 2003: 31 Notable Human Studies

The 2003 picture

What the research shows

The 2003 evidence strengthened the view that dry eye is common, clinically burdensome, and varied in cause and presentation. Controlled treatment findings were encouraging for selected groups, while the many proposed laboratory and imaging measures still needed broader validation.

Treatment studies

Several trials compared active treatments with sham procedures or other care. The strongest conclusions are about the specific patients and short periods studied.

1. Oral pilocarpine improved Sjögren-related eye symptoms

   A 12-week randomized trial assigned 85 people with primary Sjögren's syndrome to oral pilocarpine, artificial tears alone, or punctal occlusion plus tears. Pilocarpine improved patient-rated dryness and rose bengal staining more than the comparison treatments, but Schirmer tear production did not differ. Sweating, headache, nausea, and vomiting were reported with pilocarpine, and the single-center study was relatively short.

2. Warm compresses increased the tear-film lipid layer

   In 20 patients with meibomian gland dysfunction, researchers randomly applied a warm, moist compress to one closed eye and a room-temperature compress to the other. Lipid-layer thickness rose by more than 80% in the warm-compress eyes within five minutes and did not meaningfully change in control eyes. This measured an immediate tear-film response, not lasting symptom relief.

3. Punctal occlusion outperformed a sham procedure

   In a masked single-center study, researchers assigned 61 patients with dry eye and conjunctivitis to staged tear-duct plugs or a sham procedure. Symptom scores, staining, and lubricant use improved much more in the occlusion group over eight weeks. The staged treatment and short follow-up do not establish which plug material worked best or how long the benefit lasted.

4. Lower punctal plugs offered most of the short-term benefit

   Sixty-two people with aqueous-deficient dry eye were randomized to temporary collagen plugs in the lower puncta or in both upper and lower puncta. Symptoms, tear volume, and stability improved in both groups, with no clear practical gain from plugging the upper puncta as well. Follow-up ended at 12 days and there was no sham-treated dry-eye group.

5. Fatty-acid supplements improved some inflammatory measures

   A 45-day randomized study assigned 26 patients with aqueous-deficient dry eye to linoleic and gamma-linolenic acid tablets or placebo; both groups used tear substitutes. Symptoms, lissamine staining, and conjunctival HLA-DR expression improved more with the active tablets, while Schirmer and tear-breakup measurements did not separate the groups. The small, brief trial cannot establish a durable treatment effect.

6. Steroid drops produced stronger short-term results than NSAID drops

   Thirty-two patients with keratoconjunctivitis sicca were randomized to artificial tears alone, tears plus a nonsteroidal anti-inflammatory drop, or tears plus a corticosteroid drop. Over 30 days, the steroid group showed greater improvement in symptoms, staining, goblet cells, and an inflammatory cell marker; Schirmer readings did not improve. The trial was too small and short to establish longer-term safety or a preferred steroid regimen.

7. Serum drops helped some patients with graft-versus-host dry eye

   Fourteen patients with severe graft-versus-host dry eye that had not responded to artificial tears used autologous serum drops and were assessed repeatedly. Symptoms and ocular staining improved after four weeks; seven maintained a response for months, while six others later needed punctal plugs. Without a comparison group, the study cannot separate serum's effect from other care or changes over time.

8. A disposable eyelid warmer showed short-term improvements

   A device study first measured the immediate effects of eyelid warming, then followed 17 patients using it daily for two weeks with eight untreated patients as controls. Breakup time, symptoms, lipid-layer appearance, and gland obstruction improved in the treated group. The small, apparently nonrandomized comparison leaves long-term effectiveness uncertain.

Symptoms, daily impact, and clinical course

These studies examined how symptoms relate to examination findings, how much patients value relief, and how dry eye changes during treatment or after surgery.

9. Symptoms and clinical signs were only moderately aligned

   A multicenter study compared questionnaires and ocular tests in 122 people with non-Sjögren dry eye, Sjögren's syndrome, or healthy eyes. Symptoms often intensified by evening, but their correlations with individual tear and staining tests were only moderate. The groups were selected by prior dry-eye status, so the findings do not define a screening test for the general population.

10. Patients reported a substantial quality-of-life burden

    Fifty-six patients with varying dry-eye severity completed a time-trade-off survey alongside symptom and vision questionnaires. Average utility scores were 0.78 for moderate and 0.72 for severe dry eye, indicating a meaningful burden from the patients' perspective. This small clinic sample cannot represent everyone with dry eye or be directly equated with another disease's burden.

11. Symptoms improved under care, but complete recovery was uncommon

    Researchers reassessed 97 treated dry-eye patients one to eight years after their original diagnosis, using symptoms, tear tests, staining, and cytology. On average, breakup time and some other measures improved, and patients reported fewer symptoms and less artificial-tear use, but none was completely cured. With no untreated comparison group and varied treatments, the study cannot identify which care caused the changes.

12. LASIK flap orientation was linked to different sensory recovery

    In a randomized fellow-eye study of 52 bilateral LASIK patients, one eye received a superior-hinge flap and the other a nasal-hinge flap. Sensation fell in both eyes after surgery, but the nasal-hinge eyes recovered more by six months and generally had milder dry-eye findings. Different microkeratome models were used for the two flap positions, so hinge direction alone may not explain the difference.

Tear-film tests and ocular surface biology

Researchers developed ways to measure tears and examined proteins, cells, nerves, and immune activity. Many methods were tested only in small or selected samples.

13. Tear proteins differed across forms of dry eye

    A prospective case-control study measured lactoferrin, epidermal growth factor, and aquaporin 5 in tears from 103 dry-eye patients and 16 healthy controls. Lactoferrin and epidermal growth factor were lower across the dry-eye groups, while aquaporin 5 was higher only in Sjögren's syndrome. The small control group and overlapping biology limit use of these proteins as stand-alone diagnostic markers.

14. Obstructive meibomian disease increased measured tear evaporation

    Using a new evaporimetry system, investigators compared 21 patients with obstructive meibomian gland dysfunction and 22 healthy people. Evaporation rates and blink-related changes were higher in the gland-disease group and tracked gland-orifice obstruction scores. The instrument and thresholds were tested in a modest case-control sample rather than independently validated for diagnosis.

15. Lipid-deficient tears spread more slowly after a blink

    Sequential tear-interference images were analyzed in eight patients with lipid tear deficiency and 11 healthy volunteers. Lipid spread took longer in the affected group, and its final layer was thinner and distributed differently. The striking differences came from a very small, selected sample.

16. Meibomian secretions contained a broader mix of polar lipids

    Researchers chemically analyzed meibomian gland secretions from healthy people and patients with chronic blepharitis, identifying several phospholipid and sphingolipid classes. Some fatty-acid patterns differed in samples from patients with meibomianitis, adding detail to how the tear lipid layer may work. This biochemical comparison did not test symptoms or show that a particular lipid change causes dry eye.

17. Conjunctival cell changes tracked clinical severity

    Impression cytology from 143 patients with varying dry-eye severity and 33 controls was compared with clinical tests and tear osmolarity. More severe disease was accompanied by fewer goblet cells and greater squamous changes in surface cells. The proposed grading scheme was developed from this sample and still needed independent validation.

18. Aqueous deficiency also disturbed lipid spread

    Researchers recorded tear-interference images in 17 eyes with aqueous-deficient dry eye and reassessed nine after punctal occlusion. Lipid spread was slow and uneven at baseline and became faster and more even after plugs. The pre-post subgroup was small and uncontrolled, so the change should not be taken as definitive evidence of treatment benefit.

19. Videokeratography offered a noninvasive view of breakup

    A new system recorded sequential corneal topography in 80 eyes from 48 volunteers and compared its breakup measures with slit-lamp fluorescein findings. The measurements were correlated, and the new method identified some symptomatic eyes whose conventional breakup times appeared normal. Diagnostic performance was estimated within this development sample, without an independent reference standard.

20. Sjögren corneas showed changes beyond tear deficiency

    In vivo confocal microscopy compared ten people with primary Sjögren's syndrome with ten matched controls. Surface irregularities, activated-looking stromal cells, and possible nerve sprouting appeared in several affected corneas, although overall nerve density did not differ. The tiny cross-sectional sample cannot show whether those features precede or result from ocular surface disease.

21. Dry-eye tests captured related but different features

    Investigators compared Schirmer readings, phenol-red thread results, staining, breakup time, and other examinations in dry-eye patients. Lower tear-volume or production measures were associated with more ocular surface staining in several analyses. Correlation among tests does not establish which combination best diagnoses disease in an unselected population.

22. A symptom inventory distinguished Sjögren's syndrome in its study groups

    Researchers developed a detailed sicca-symptom inventory and compared its scores in women with primary Sjögren's syndrome, other rheumatic diseases, and healthy controls. Oral and ocular symptom scales separated the groups, while symptom severity correlated only moderately with measured tear and saliva output. Initial validation was limited to White women, and responsiveness to treatment had not yet been tested.

23. Conjunctival gene patterns differed in Sjögren's syndrome

    A gene-expression study compared conjunctival cells from 26 people with Sjögren's syndrome and 30 controls across 931 genes. The affected group showed higher expression of several inflammation and keratinization-related genes, consistent with altered surface-cell behavior. This broad exploratory screen did not independently validate the gene set or prove which changes drive symptoms.

24. Graft-versus-host lacrimal glands showed localized T-cell activation

    Biopsies from nine patients with chronic graft-versus-host disease and five with Sjögren's syndrome were examined for immune activation in the lacrimal gland. Activated T cells clustered around ducts and near antigen-presenting cells and fibroblasts in graft-versus-host tissue, suggesting a distinct pathway of gland injury. The small tissue series shows association, not that these interactions alone caused dry eye.

25. Conjunctival folds were a weak discriminator in one small study

    Researchers compared lid-parallel conjunctival folds, tear osmolarity, staining, and redness in 14 people with moderate dry eye and eight controls. The groups differed in osmolarity and several surface findings, but fold counts were similar. This small case-control study cautions against relying on visible folds alone to identify dry eye; it does not settle how folds perform in a broader clinical population.

Prevalence and associated conditions

Population surveys used different symptom and diagnosis rules, so their percentages should be read with each study's definition in mind. Smaller comparisons explored possible risk pathways and contact lens discomfort.

26. A large U.S. survey estimated dry eye in women

    The Women's Health Study surveyed 39,876 U.S. women about a dry-eye diagnosis or severe dryness and irritation. Its age-adjusted estimate was 7.8% among women age 50 and older, with prevalence increasing from younger to older age groups. Because the outcome was self-reported, this is an estimate of reported disease or severe symptoms, not examination-confirmed dry eye.

27. Taiwanese elders often had both symptoms and abnormal tear tests

    The population-based Shihpai Eye Study examined 1,361 Taipei residents age 65 and older with symptom questions and tear, staining, and meibomian gland tests. About one-third reported frequent dry-eye symptoms; among those symptomatic, many had short breakup time or low Schirmer readings. Only two-thirds of selected residents participated, and the estimate depends on the symptom threshold used.

28. An Australian cohort linked symptoms with systemic factors

    In the Blue Mountains Eye Study extension, 1,174 adults age 50 and older completed dry-eye questions and an eye examination. Moderate or severe symptoms were reported by 16.6%, and symptom reports were more common in women and in people with several systemic conditions or medication exposures. These cross-sectional associations cannot establish that the conditions or medicines caused dryness.

29. Androgen insensitivity offered a clue about meibomian function

    Researchers compared people with complete androgen insensitivity syndrome with age-matched female and male controls, examining symptoms, eyelids, tears, and gland secretions. The androgen-insensitive group had more dry-eye complaints and more meibomian abnormalities, consistent with a role for androgen signaling. This rare condition is an unusual comparison group and does not show what hormone treatment would do for ordinary dry eye.

30. Long-term dialysis patients had reduced tear secretion

    A case-control study compared 63 hemodialysis patients with 23 healthy volunteers using Schirmer testing, symptom questions, and other gland assessments. Tear and saliva secretion were lower in the dialysis group, and eye dryness could be present without symptoms. Other health differences between groups prevent attributing the reduction specifically to dialysis treatment.

31. Clinical tear measures partly predicted contact lens intolerance

    Researchers compared 20 successful contact lens wearers with 18 people who had stopped wearing lenses because of discomfort, then assessed comfort after six hours of wear. The intolerant group had lower tear volume and stability; protein tests did not separate groups as well. Predictive formulas had low specificity in validation, so the measures were not reliable enough to rule in intolerance on their own.