Dry Eye Research in 2018: 33 Notable Human Studies

The 2018 picture

What the research shows

The strongest 2018 result was DREAM's year-long finding that high-dose omega-3 did not outperform placebo. Smaller trials offered more targeted signals, often on tear or gland measurements rather than patient-level symptoms. New diagnostic approaches and exposure studies added useful context, but their results were sensitive to how dry eye was defined and whom researchers enrolled.

Treatments and clinical trials

The largest trial tested an oral supplement for a full year. Most other interventions were shorter, smaller, or focused on a specific cause of ocular dryness.

1. DREAM found no added dry eye benefit from high-dose omega-3

   The multicenter DREAM trial randomized 535 people with moderate to severe dry eye to 3,000 mg of fish-derived omega-3 fatty acids daily or an olive-oil placebo for one year. Symptom scores improved in both groups, but the between-group difference was not statistically significant, and staining, tear breakup, and Schirmer tests also showed no clear advantage. The findings apply to this dose and comparator in the enrolled population, rather than every possible diet or supplement regimen.

2. Eyelid intense pulsed light improved gland output, with mixed symptom results

   In a sham-controlled paired-eye trial involving 44 people with meibomian gland dysfunction, one eye received three direct-eyelid intense pulsed light treatments and the other sham treatment; both eyes also received gland expression and artificial tears. Treated eyes had greater gains in meibomian secretion and tear breakup time, but symptom and corneal-staining changes did not differ clearly between eyes. The paired-eye design and shared symptom ratings limit conclusions about patient-level relief.

3. A lifestyle program eased reported symptoms in a small office trial

   Forty-one Japanese office workers with definite or probable dry eye were randomized to a two-month program involving diet, exercise, and positive-thinking advice or a control group; 36 completed follow-up. Symptoms improved within the intervention group, but staining, tear breakup, and Schirmer results did not differ significantly between groups. The small study cannot identify which part of the combined program helped or whether any effect lasts.

4. Diquafosol and artificial tears both helped short breakup-time dry eye

   A five-week randomized study assigned 27 people with short tear breakup-time dry eye to diquafosol or preservative-free artificial tears. Both groups reported less discomfort and pain sensitivity; diquafosol increased tear-meniscus height as well as breakup time, while artificial tears improved breakup time. The tiny, subtype-specific trial does not establish a broad symptom advantage for diquafosol.

5. Scheduled lubricant dosing missed its primary staining goal

   A multicenter randomized trial compared a polyethylene-glycol/propylene-glycol artificial tear used four times daily with use as needed in 97 people with dry eye. At day 28, regular dosing was not superior on the prespecified total surface-staining endpoint, although a secondary symptom-bother score favored it. The result concerns one product and a short treatment period, and the symptom finding should be read as secondary.

6. Three lubricants improved symptoms; biomarker differences were exploratory

   Eighteen people with dry eye used three lubricant formulas in randomized crossover order, with tear physiology and inflammatory proteins measured after four weeks of each treatment. All three improved symptoms and noninvasive tear stability; one carboxymethylcellulose, glycerin, and castor-oil formula showed the largest reduction in several measured inflammatory markers. The small crossover trial did not show a corresponding difference in evaporation or tear osmolarity, and its biomarker pattern needs confirmation.

7. Oral pilocarpine raised tear flow in Sjögren syndrome

   A double-blind randomized trial assigned 72 people with Sjögren syndrome to oral pilocarpine or an oral artificial-saliva comparator for 12 weeks. Pilocarpine produced greater lacrimal and salivary flow and better overall dryness ratings, with nausea and excess saliva among the reported adverse effects. Because the comparator targeted mouth dryness and the trial was brief, it does not establish longer-term ocular benefit against other eye treatments.

8. Punctal plugs improved ocular-surface signs in glaucoma-drop users

   In a randomized paired-eye trial, 48 patients with symptoms while using prostaglandin glaucoma drops received a lower punctal plug in one eye and no plug in the other. At six weeks, plugged eyes had longer tear breakup time, less staining, and lower tear osmolarity; some plugs extruded and some eyes developed increased tearing. The short study concerns treatment-related ocular-surface disease, and whole-person symptom ratings cannot be attributed neatly to one eye.

9. A gland-expression device improved MGD measures over one month

   Seventy patients with meibomian gland dysfunction were randomized to a mechanical gland squeezer or a control group, and 65 completed follow-up. At one month, symptoms, tear breakup time, staining, and meibum measures improved more in the treated group; mild pain or discomfort was reported during the procedure. The trial was short and did not establish how long the gains persist.

10. Cyclosporine before transplant showed a benefit in a dry-eye subgroup

    A randomized study began topical cyclosporine before allogeneic stem-cell transplantation and continued it afterward, comparing outcomes with no pretreatment in 58 patients who completed three months. Among patients who already had low tear production and short breakup time before transplant, the treated group had better corresponding tear tests afterward. This subgroup finding does not prove prevention of ocular graft-versus-host disease in all transplant recipients.

11. Thermal pulsation for contact lens wearers with gland disease

    An open-label randomized study assigned 55 soft contact lens wearers with meibomian gland dysfunction to one thermal-pulsation treatment or no immediate treatment. Gland secretion, symptom scores, and comfortable lens-wearing time improved more in the treated group over three months. The absence of a sham procedure and the short follow-up make the size and durability of the benefit uncertain.

12. Two lubricant formulas compared over six weeks

    A double-masked trial randomized 99 patients to a hydroxypropyl guar–hyaluronic acid lubricant or sodium hyaluronate for 42 days. The combination met the prespecified noninferiority test for total ocular staining but did not clearly improve tear breakup over the comparison drop. Noninferiority on one sign does not establish that the formulas are equally helpful for every patient or over longer use.

13. Added lubricant ingredients after LASIK

    At 14 centers, 148 people undergoing LASIK were randomized under double masking to a multi-ingredient carboxymethylcellulose–hyaluronic acid drop or carboxymethylcellulose alone for 90 days. Dry-eye symptom scores improved in both groups, with no significant between-group difference at day 90; tear and staining measures were also similar. Participants had no more than mild preoperative dry eye, so the trial does not answer whether the added ingredients help people with established severe disease.

Surgery, blinking, and patient experience

Prospective surgical cohorts and patient studies described symptom persistence, blink behavior, and the outcomes people with dry eye most want researchers to measure.

14. Early tear and gland findings tracked persistent symptoms after cataract surgery

    A single-center prospective cohort followed 116 cataract patients, with 96 completing dry eye assessments through three months. Thirty-one of those followed still had elevated symptom scores at three months; worse baseline symptoms and selected one-month tear and gland findings were associated with persistence. These associations came from one surgical cohort and do not establish an independently validated prediction rule.

15. Incomplete blinking accompanied more dry eye and gland loss

    Researchers compared 77 people with visible incomplete blinks with 77 matched people without them during a single eye examination. The incomplete-blink group more often met dry eye criteria, reported more symptoms, and had more meibomian gland dropout; blink frequency alone was not linked to the measured surface findings. The cross-sectional comparison cannot show whether incomplete blinking caused the gland and tear changes.

16. Symptom scores differed by sex even at similar levels of clinical signs

    A cross-sectional analysis of 755 patients in a Dutch sicca cohort compared women and men within strata of dry eye signs. Women reported more symptoms in the mild and moderate sign groups, while the symptom difference narrowed in the severe group; symptoms and signs also correlated less closely in women. Referral and reporting differences may contribute, so the study does not isolate a biological cause.

17. Sjögren dry eye also involved meibomian gland dysfunction

    Investigators compared women with primary or secondary Sjögren syndrome, a separate meibomian gland dysfunction group, and controls using eye examinations and dry eye questionnaires. Sjögren groups had more blocked gland openings and poorer meibum quality, consistent with an evaporative component alongside aqueous tear deficiency. The selected cross-sectional groups cannot determine how often this combination occurs in all Sjögren patients.

18. Abnormal tear tests were common before cataract evaluation

    Two U.S. practices prospectively evaluated 120 patients presenting for cataract surgery using tear osmolarity, MMP-9, surface examination, and symptom questionnaires. Eighty percent had at least one abnormal tear test, including many who did not report symptoms. The broad test-based definition and selected surgery candidates mean this is not a population prevalence estimate or proof that every positive test needs treatment.

19. Patients prioritized pain and daily impact in dry eye research

    A two-round online survey asked 420 people with dry eye from 15 countries to rank research questions and outcomes. Burning, discomfort, and ocular pain ranked highest, while several patient-prioritized outcomes rarely appeared in the trials the researchers reviewed. The volunteer sample was largely older, female, and White, so its priorities may not represent every patient group.

Diagnostic methods and inflammatory markers

Studies compared tear-film tests, evaluated blinking as a screen, revised diagnostic categories, and examined tear or conjunctival markers. The most promising measures still had limits as stand-alone tests.

20. Maximum blink interval offered only moderate screening accuracy

    A Japanese clinic study measured how long 292 people could keep their eyes open and compared that interval with tear breakup time and dry eye classification. Blink interval was shorter in dry eye, but a proposed 12.4-second cutoff had 82.5% sensitivity and only 51.0% specificity. The single-center, selected sample and low specificity make it a possible screen, not a diagnosis by itself.

21. A noninvasive breakup test distinguished symptoms better than fluorescein time

    In a randomized crossover test-order study, 74 young adults with and without dry eye symptoms received automated noninvasive tear breakup testing and conventional fluorescein breakup testing. The automated result separated the selected groups better, but its diagnostic area under the curve was only 0.68, with 68% sensitivity and 70% specificity at the chosen cutoff. The symptom-based group definition and narrow age range limit generalization.

22. Automated MGD tests traded sensitivity for specificity

    Researchers compared automated tear and gland measurements in 149 patients with meibomian gland dysfunction and 27 controls. Counting an abnormal breakup time or gland-loss result as positive yielded 86% sensitivity but only 39% specificity; requiring both reversed the tradeoff to 39% sensitivity and 86% specificity. The selected case-control sample and weak single-test discrimination support using these measures alongside a clinical gland exam.

23. New diagnostic rules reclassified many probable dry eye cases

    A Japanese study applied both 2006 and 2016 dry eye criteria to 250 clinic patients. Under the newer symptom-plus-short-breakup rule, 71 of 89 patients previously labeled probable dry eye became definite cases. The rise reflects a changed definition in a selected clinic sample, not evidence that dry eye itself suddenly became more common.

24. Different short questionnaires suited different lens-wearing groups

    A psychometric study evaluated the SPEED questionnaire in 284 people, including 150 contact lens wearers, and the CLDEQ-8 in the lens-wearing subgroup. Shortened forms tracked self-reported dryness differently: SPEED better identified dry eye status in nonwearers, while the shorter CLDEQ form performed better in wearers. Neither questionnaire tracked the measured tear-meniscus height or phenol-red thread test, limiting stand-alone diagnostic claims.

25. Tear cytokines varied across dry eye subtypes

    A laboratory comparison measured 11 inflammatory molecules in tears from Sjögren and non-Sjögren aqueous-deficient dry eye, meibomian gland dysfunction, and control groups. Several cytokines were higher in dry eye, and IL-8 and TNF-alpha differed among patient subtypes. Small selected groups and the absence of validated clinical cutoffs keep these markers exploratory.

26. Conjunctival immune-cell percentages tracked Sjögren eye severity

    Investigators sampled conjunctival cells in Sjögren-associated keratoconjunctivitis sicca and compared antigen-presenting cell percentages with clinical severity and goblet-cell density. More HLA-DR-positive immune cells correlated with worse ocular findings and fewer goblet cells. This observational tissue-marker result does not establish a diagnostic threshold or prove that the cells cause the damage.

Prevalence, exposures, and associated conditions

Large surveys and clinical comparisons examined dry eye in different populations and settings. These associations depend on case definitions and generally cannot establish cause and effect.

27. A Brazilian survey found 12.8% self-reported dry eye

    A cross-sectional questionnaire study of 3,107 participants across Brazil's five regions found 12.8% met its self-reported dry eye measure. Older age, female sex, contact lens wear, prolonged computer use, and several other factors were associated with higher odds. Symptoms and previous diagnosis were self-reported without eye examinations, and sampling and regional differences limit a precise national disease estimate.

28. Dry eye was linked to poorer glucose control in a diabetes cohort

    A community-based Chinese study examined 1,360 people with type 2 diabetes using symptoms, tear tests, staining, and eye examinations. Seventeen and a half percent met the study's dry eye definition, and higher blood glucose and HbA1c were associated with dry eye. The single-time-point design cannot determine whether glucose control caused the ocular findings.

29. Traffic-related pollution and weather tracked dry eye diagnoses

    A Taiwan case-crossover study linked 25,818 first dry eye diagnoses from health-insurance records with nearby air-monitoring data on diagnosis days and comparison days. Carbon monoxide, nitrogen dioxide, and temperature were positively associated with diagnoses, while humidity was inversely associated. Exposure was estimated from monitors near clinics rather than each person's actual surroundings, so the links do not prove causation.

30. Four hours of smartphone viewing worsened short-term tear measures

    Eighty healthy volunteers were assessed before and after smartphone viewing or computer-screen use for up to four hours. The smartphone group reported more dryness and fatigue and had shorter tear breakup times, with higher measured oxidative-stress markers after prolonged use. A short laboratory task in healthy adults cannot establish the risk of lasting dry eye from ordinary phone use.

31. Dry eye reports correlated with depression and suicidal thoughts

    A Korean national survey analysis of 16,408 adults compared reported dry eye diagnosis or symptoms with questionnaire reports of depression and suicidal ideation. Both dry eye measures were associated with higher adjusted odds of these mental-health reports. Because all conditions were assessed through survey responses at one time, the study cannot show that dry eye caused depression or suicidal thoughts.

32. Meibomian gland loss tracked symptoms and age in one sample

    A German study used lower-eyelid gland imaging and dry eye assessments in 112 adults. Gland loss was associated with age and dry eye status across the full sample, while symptom score was a stronger correlate within the dry eye group and sex was not a clear predictor. The modest cross-sectional sample cannot establish which changes came first.

33. Gland dysfunction and eyelid friction accompanied shorter tear breakup

    In a cross-sectional study of 449 dry eye patients, researchers recorded meibomian gland dysfunction, friction-related ocular-surface findings, and tear tests. Breakup time was shorter among patients with gland dysfunction, and friction-related findings also accompanied shorter breakup time in aqueous-deficient dry eye. The associations do not show whether treating either finding will improve symptoms.