Dry Eye Research in 2014: 35 Notable Human Studies

The 2014 picture

What the research shows

The 2014 literature broadened dry eye beyond a single tear-deficiency model. Large surveys clarified who reports symptoms; controlled trials offered several treatment options but often measured short-term outcomes; and diagnostic studies showed why symptoms, tear stability, inflammation, and gland findings must be interpreted together.

Treatments and ocular-surface protection

The strongest treatment evidence comes from randomized comparisons. Smaller uncontrolled series remain useful for questions about feasibility and longer follow-up, but cannot establish that treatment caused an improvement.

1. Lifitegrast improved corneal staining in OPUS-1

   The 84-day, double-masked OPUS-1 trial randomized 588 adults with dry eye to lifitegrast 5% or vehicle drops. Lifitegrast improved the primary inferior corneal-staining measure, but missed the coprimary visual-function symptom measure; some secondary dryness and discomfort ratings favored treatment. The mixed primary results and short follow-up limit claims about broad or lasting benefit.

2. Systemic hydroxychloroquine missed its Sjögren symptom target

   The JOQUER trial randomized 120 people with primary Sjögren syndrome to hydroxychloroquine or placebo for 24 weeks. About 18% in each group achieved the prespecified improvement in at least two of three symptom scales covering dryness, pain, and fatigue. The dryness scale combined mouth and eye symptoms, so this negative systemic trial is informative but not a stand-alone test of ocular dry-eye treatment.

3. Three lid-warming methods gave mixed gland-disease results

   An assessor-masked trial randomized 75 people with meibomian gland dysfunction to a warm towel, EyeGiene, or Blephasteam for three months. Symptom improvement was more common with Blephasteam than a warm towel, but tear breakup, tear production, and gland blockage did not show a clear between-group advantage. Participants knew their treatment, and each arm was small.

4. Thermal pulsation compared with daily lid care

   In an observer-masked randomized trial, 31 people with meibomian gland dysfunction completed three months after one LipiFlow treatment or twice-daily lid warming and massage. Symptoms improved more in the device group, while both groups gained expressible glands and most other objective measures did not differ significantly. The small sample and lack of patient masking limit the symptom comparison.

5. Autologous serum in severe dry eye

   A double-masked randomized crossover trial compared 20% autologous serum drops with preservative-free artificial tears in 20 people with severe, treatment-resistant dry eye. After one month, serum produced a larger symptom reduction and longer tear breakup time. Schirmer wetting and surface staining did not clearly differ, and the small trial cannot settle long-term effectiveness.

6. Loteprednol added to lid hygiene for gland disease

   Patients with moderate to severe meibomian gland dysfunction were randomized to lid scrubs and warm compresses with or without topical loteprednol. Across 70 study eyes, the steroid group had larger early improvements in tear breakup, staining, meibum quality, and several tear cytokines. The two-month study supports short-term anti-inflammatory benefit; it does not answer how often or how long steroid treatment is appropriate.

7. Longer use of 50% autologous serum drops

   A retrospective review followed 63 patients prescribed 50% autologous serum, on average for 12 months. Corneal staining and symptom scores improved at several follow-up points, with no complications reported in the reviewed records. Without a concurrent comparison group, natural fluctuation and other care could explain some of the change.

8. Six-month follow-up after thermal pulsation

   Investigators reassessed 26 people with meibomian gland dysfunction six months after one thermal-pulsation treatment. Symptoms, lipid-layer thickness, and the number of expressible glands improved from baseline; gland atrophy did not reverse. People with less atrophy tended to report more relief. This before-and-after analysis suggests durability but lacks an untreated comparison.

9. Steroid pretreatment when starting cyclosporine

   A double-masked multicenter trial randomized 118 people beginning cyclosporine to two weeks of loteprednol or artificial-tear pretreatment, then continued the assigned add-on for six more weeks. Loteprednol reduced cyclosporine stinging and produced larger improvements in several symptom and surface measures. Follow-up was short, so the trial does not establish a lasting advantage or long-term steroid safety.

10. A heated eye bag for evaporative dry eye

    A single-masked randomized contralateral-eye trial assigned 25 people with gland-related evaporative dry eye to a heated or unheated eye bag for five minutes twice daily over two weeks. The heated side improved on tear-film and gland measures; participants also reported longer-lasting comfort, especially with occasional continued use. Eye-level assignment cannot fully separate whole-person symptoms, and the initial comparison was brief.

11. Cationic emulsion against conventional lubricant drops

    A multicenter open-label trial randomized 79 patients with mild to moderate dry eye to a cationic emulsion or polyvinyl alcohol–povidone drops. At four weeks, tear breakup time and lissamine-green staining improved more with the emulsion, while both groups tolerated treatment. The modest sample, short follow-up, and open-label design leave the size of a patient-important benefit uncertain.

12. Scleral lenses in complex ocular-surface disease

    A tertiary-center chart review examined 212 people evaluated for scleral lenses across several ocular-surface conditions. Among 115 who completed fitting, therapeutic goals were met in all but two, and average visual acuity improved. The selected referral population and lack of a comparator mean the results describe successful use after fitting, not the chance that any dry-eye patient will benefit.

Population patterns, risk, and daily impact

These studies drew from different countries, ages, and disease definitions, and one examined reading in older adults. Their percentages describe their own samples and should not be combined into one universal prevalence estimate.

13. Older adults with dry eye reported more reading difficulty

    The Salisbury Eye Evaluation examined 1,981 older adults with dry-eye symptom questions, reading habits, and a timed spoken-reading task. Participants with symptoms were more likely to report reading difficulty or avoid newspaper reading, yet adjusted spoken-reading speed was not clearly slower. Its cross-sectional measures cannot establish that dry eye caused the change in reading behavior.

14. Dry eye in the Beaver Dam Offspring Study

    A cohort analysis of 3,275 adults found dry-eye symptoms in 14.5%, with higher prevalence among women. Contact lens wear, allergies, arthritis, thyroid disease, and some medications were associated with dry eye, which also tracked with lower general and vision-related quality-of-life scores. The questionnaire-based definition and observational design cannot establish causes.

15. Prevalence in a British female twin cohort

    A cross-sectional survey of 3,824 women in TwinsUK found that 9.6% reported a dry-eye diagnosis with artificial-tear use, while 20.8% reported recent symptoms. Age, allergy, arthritis, and several chronic pain conditions were associated with disease measures. This larger survey reuses the TwinsUK cohort studied for pain sensitivity in 2013, rather than providing an independent population; self-reported measures also limit generalization.

16. A national dry-eye estimate from South Korea

    An analysis of 11,666 adults in Korea's national health survey estimated 8.0% with a reported clinical diagnosis and 14.4% with symptoms. Female sex, age, previous eye surgery, stress, and thyroid disease were associated with both outcomes. These survey associations identify groups worth studying, but they do not show that any one factor causes dry eye.

17. Where dry eye clustered across South Korea

    A separate spatial analysis of 16,431 Korean national-survey participants found higher reported dry eye in urban than rural areas, with associations involving humidity and sunshine duration. It broadens the environmental picture, but regional exposure measures and a cross-sectional survey cannot isolate an individual's cause. This paper partly overlaps the population behind the national prevalence report above.

18. Screen use and dry eye in schoolchildren

    Investigators compared 28 children classified with dry eye and 260 controls. Longer total screen exposure and smartphone use were associated with dry eye, whereas computer time alone was not clearly associated. The observational study raises a question about device habits but cannot separate screen exposure from other differences between the children.

19. Sleep loss as an experimental tear-film stressor

    Twenty healthy men were allocated to a sleep-deprivation condition or control and had repeated tear measurements. By the next morning, the sleep-deprived group had higher tear osmolarity, shorter tear breakup time, and less Schirmer wetting. The controlled short-term experiment supports an effect of acute sleep loss on tear physiology, but it did not test established dry-eye patients or repeated sleep restriction.

20. Gland findings in long-hour screen workers

    A case-control study compared 53 people using visual displays more than four hours daily with 40 shorter-duration users. Longer-use workers had worse symptom, staining, tear-breakup, and meibomian-gland measures, even when tear volume was normal. Exposure was not assigned, so work patterns and other factors may account for part of the difference.

21. Diet, vitamin D, and dry-eye measures

    A cross-sectional study measured diet, serum vitamin D, symptoms, and tear signs in 247 older male veterans. Greater adherence to a Mediterranean diet was unexpectedly associated with more dry-eye classification, while vitamin D levels were not associated with disease presence and showed only a small symptom association. This is hypothesis-generating observational evidence, not a reason to change diet or prescribe supplements for dry eye.

22. How much dry-eye variation runs in families?

    A twin analysis surveyed 3,930 British women and examined dry-eye signs in 606. Identical twins resembled one another more than nonidentical twins for several outcomes; modeled heritability was 29% for symptoms and 41% for diagnosis with tear use. Estimates varied widely by sign, and shared sampling and measurement choices shape these population-level estimates.

23. Tear-film function in type 1 diabetes

    Researchers compared 53 people with type 1 diabetes with 40 age-matched controls and assessed neuropathy in the diabetes group. Tear stability, quantity, lipid quality, and corneal sensitivity were lower with diabetes, while symptom scores were similar. Worse neuropathy tracked with poorer tear stability, but the cross-sectional comparison cannot determine whether neuropathy produced the surface changes.

Clinical course, measurements, and mechanisms

Objective measurements can explain part of dry eye, yet symptoms and signs often diverge. The studies below show what particular tests captured in selected patients and where their conclusions remain narrow.

24. Changes in tear osmolarity did not track symptom relief

    A retrospective study compared two visits for 186 dry-eye patients receiving routine care. Changes in tear osmolarity did not meaningfully track changes in symptoms or corneal staining, while staining and symptoms had a modest relationship. The treatments were not standardized, and a two-visit clinic review cannot rule out value for osmolarity in other settings.

25. Dry-eye and gland changes after cataract surgery

    A prospective series followed 48 patients before cataract surgery and at one and three months afterward. Symptoms, lid-margin findings, and tear breakup worsened, and meibum expression declined by month three; measured gland structure and several tear-volume measures did not clearly change. Without a nonsurgical comparison, the study cannot fully separate surgery from time and postoperative care.

26. Low humidity triggers measurable surface changes

    Nineteen dry-eye patients and 20 symptom-free controls spent two hours in a chamber at 5% relative humidity. Both groups showed more corneal staining, shorter tear breakup, and higher tear MMP-9 afterward. The controlled exposure demonstrates an acute response to very dry air, although it does not measure longer-term disease under typical indoor conditions.

27. Mapping corneal epithelial thickness

    Optical coherence tomography mapped the corneal epithelium in 100 symptomatic dry-eye patients and 35 controls. The superior epithelium was thinner in the dry-eye group, and greater disease severity was associated with more irregular thickness patterns; central thickness did not distinguish groups well. This cross-sectional study proposes an additional imaging measure rather than a stand-alone diagnostic rule.

28. The best test depends on the underlying condition

    A comparative study measured symptoms, osmolarity, breakup time, staining, and tear production in 125 patients across six defined ocular or systemic conditions and 24 controls. Test patterns differed by disease; no measure captured every group equally. The very high accuracy reported for one combined test set was derived in this selected sample and needs external validation.

29. Inflammatory cytokines in Sjögren and other dry eye

    Tear samples from 20 people with Sjögren dry eye, 20 with non-Sjögren dry eye, and 20 controls showed higher IL-17 and IL-22 in both disease groups, especially Sjögren disease. Cytokine levels correlated with some symptom and surface measures. The small cross-sectional comparison supports an inflammatory signal but does not prove the cytokines caused the condition.

30. Dry eye after SMILE and LASIK

    A prospective nonrandomized study followed 176 patients after SMILE or two LASIK techniques. Tear-film breakup and symptoms worsened early after all procedures, but several six-month measures favored SMILE over LASIK. Choice of surgery was not randomized, so baseline or treatment differences may contribute to the apparent advantage.

31. A point-of-care inflammatory marker test

    At four sites, investigators compared a rapid tear MMP-9 test with a clinical dry-eye classification in 237 patients. Depending on whether symptoms were required in the reference definition, positive agreement was 81%–86% and negative agreement 97%–98%. Agreement with a constructed clinical definition is useful, but it does not make MMP-9 an independent gold-standard diagnosis.

32. Gland loss versus expressed meibum

    In 26 people with obstructive meibomian gland dysfunction, lower lids showed more gland loss and worse expressed meibum than upper lids. Meibum quality tracked tear breakup and staining, while the amount of gland loss did not. This small cross-sectional study suggests that gland function and appearance provide different clinical information.

33. Early ocular-surface changes after refractive surgery

    A prospective comparison followed 15 SMILE and 32 femtosecond-LASIK patients through three months, measuring tear signs and inflammatory mediators. SMILE had shorter-lived increases in some tear markers and better early tear breakup and corneal sensitivity; other mediators did not differ. The small, nonrandomized groups limit causal claims and long-term predictions.

34. Light scattering and visual quality in dry eye

    A comparison of 35 dry-eye patients with 20 controls measured forward light scatter and corneal backscatter. Both dry-eye subgroups had more forward scatter, while greater corneal backscatter was most evident when central punctate staining was present. The findings connect surface disruption with optical quality, but this small study did not test a treatment or functional visual task.

35. When dry-eye symptoms and tear tests disagree

    A population survey and examination of 2,262 rural Chinese residents compared reported symptoms with tear breakup and Schirmer results. About 42% were placed in a group where symptoms and measured signs disagreed. The substantial mismatch cautions against relying on one measure alone, though the exact percentage depends on the study's definitions and two selected tests.