Secondary Logo

Journal Logo

Original Article

Dry Eye

Prevalence and Attributable Risk Factors in a Hospital-Based Population

Sahai, Anshu MS; Malik, Pankaj DNB

Author Information
Indian Journal of Ophthalmology: Apr–Jun 2005 - Volume 53 - Issue 2 - p 87-91
doi: 10.4103/0301-4738.16170
  • Open


Abnormality in preocular tear film causes dry eye. The preocular tear film, classically, is a three-layered structure consisting, from posterior to anterior, of the mucous, the aqueous and the lipid layers. The National Eye Institute/Industry Workshop on Clinical Trials in Dry Eyes1 defined dry eye as “a disorder of the tear film due to tear deficiency or excessive tear evaporation, which causes damage to the inter-palpebral ocular surface and is associated with symptoms of ocular discomfort”. Dry eye is the most frequent disorder in ophthalmology practice.2 According to Whitcher and colleagues,3 the dry eye area most fraught with misinterpretations concerns the patient′s ocular symptoms.

The prevalence of dry eyes varies from 10.8% to 57.1%,45678 thereby showing wide disparity. Much of this disparity stems from the fact that there is no standardisation of the types of patients selected for the study, dry eye questionnaires, objective tests and dry eye diagnostic criteria. Various risk factors for dry eye alluded to in literature include air pollution, cigarette smoking, low humidity, high temperature, sunlight exposure and drugs.49-13

Our aim was to determine dry eye prevalence in the Indian state of Rajasthan and evaluate risk factors associated with dry eye.

Materials and Methods

In this cross-sectional study, 500 patients above 20 years of age presenting with various ophthalmic problems to a tertiary eye care centre were screened for dry eye. The patients were selected randomly and informed about the nature of the study. Patients suffering from acute ocular infections with extensive corneal or conjunctival pathology, contact lens users and those who had undergone extraocular or intraocular surgery within six months of the screening were excluded. Five hundred and seventy-six patients were initially chosen, 53 were excluded and 23 refused to participate in the study. Informed consent was obtained from subjects recruited for the study (86.8%). A single observer who was an ophthalmologist (AS), after eliciting a complete general (including history of systemic disease, especially pertaining to dry eye) and ophthalmic history performed ocular and systemic examination and subsequently administered the dry eye questionnaire. Ocular examination included review of lid surface abnormalities and meibomian gland evaluation. Another ophthalmologist (PM) performed the objective dry eye tests thereafter. The second observer was masked to dry eye information from the questionnaire. The pre-designed dry eye questionnaire was based on models suggested by Hikichi,4 Toda,6 Roche and their colleagues14 and consisted of yes/no responses to 13 symptoms, namely: ocular fatigue, non-sticky eye discharge, foreign body sensation, heavy sensation, dry sensation, discomfort, ocular pain, watering, temporary blurred vision (improved on blinking), itching, photophobia, redness and burning/stinging sensation. A response was defined as positive when the subject reported a symptom to occur sometimes, often or all the time and as negative when reported to occur rarely or never. After ascertaining the responses to each of the questions, the symptom score was calculated. Exposure to sunlight/high temperatures, excessive winds, air pollution, smoking and drugs was inquired for. Objective tests (under room temperature conditions) comprised (in order, each at 10-minute intervals to minimize reflex tearing and ocular surface changes secondary to testing) Lissamine Green staining, Schirmer′s test and tear film breakup time (TBUT). Precut strips for these tests were obtained from a common source, (ContaCare Pvt. Ltd., Baroda), to ensure uniformity. Presence of strands/filaments was also looked for before and after the tests. In those already using tear substitutes, dry eye tests were performed after overnight discontinuation of medication. A symptom score of more than 3, Lissamine green staining score ³ 3 (as per a staining score key proposed by Norn15, Schirmer′s test value £ 5 mm in 5 minutes on Whatman′s filter paper No. 41, TBUT value < 10 seconds and presence of strands and/or filaments in either/both eyes were taken as indicators of dry eye. If three or more of the above 5 tests were positive, the subject was deemed to be suffering from dry eye.

A P-value < 0.05 was considered statistically significant. 95% Confidence intervals were tabulated. Statistical analyses were performed using software SPSS v.10.0 (SPSS Inc., Chicago, Illinois, USA). The likelihood ratio test, which is asymptotically distributed as Chi-square distribution, was used to calculate the P-values. The independent association of environmental risk factors with dry eye was assessed by the multiple logistic regression analysis test. Odds ratio was used to study the strength of the association of environmental risk factors and drugs with dry eye. Analysis of variance test (ANOVA) was used to compare the difference between the mean number of complaints per person (on questionnaire) in the dry eye and the non-dry eye groups.


Table 1 shows the baseline characteristics of the study group. On analysing the presenting complaints, the most common complaint was reduction of vision, found in 278 (55.6%) subjects, followed by watering in 122 (24.4%) and ocular pain in 97(19.4%). Dry eye was present in 92 (18.4%) of the study subjects. The prevalence was significantly higher in patients aged 70 years and above (36.1%) compared to all other age groups (P =0.007; 95% CI 1.34 to 5.67, Table 2). The age group 31-40 years showed a relative peak in dry eye prevalence (20%). Females (22.8%) had significantly higher prevalence than males (14.9%) (p=0.024; 95% CI 1.07-2.66). Prevalence was 22.8% in both non-menopausal (23/101) and in postmenopausal women (28/123). Dry eye appeared to be more common in rural (19.6%) than urban patients (17.5%) (P =0.553; 95% CI 0.55-1.37); however this was not statistically significant.

Table 1:
Baseline characteristics (n=500)
Table 2:
Prevalence of dry eye according to age, sex and place of residence

Farmers/labourers were most afflicted with dry eye (Table 3), followed by the group designated “others with high exposure” consisting of computer operators, drivers, salesmen, field workers, mechanics and cooks. Table 4 shows the strength of association of various environmental exposure factors and drugs with dry eye. All the exposure factors had a propensity for higher risk of dry eye; excessive wind (OR: 2.15), sunlight/high temperature (OR:1.91), air pollution (OR:1.38), smoking (OR:1.42) and drugs (OR:2.04). Subjects exposed to excessive wind (P =0.004), sunlight/high temperature (P =0.014) and drug exposure (P =0.002) were at higher risk of developing dry eye. Commonly used drugs were chlorpheniramine eye drops, antibiotic-corticosteroid eye drops, antiglaucoma drugs, analgesics, bronchodilators, antihypertensives, antihistamines and tranquilizers.

Table 3:
Prevalence of dry eye in various occupational groups
Table 4:
Strength of association of environmental exposure factors and drugs with dry eye

Refractive status was recorded in spherical equivalents for the purpose of this study (Table 5). Dry eye prevalence in emmetropes was 14% (18/129), while in myopes it was 16.8% (30/179) and in hypermetropes 22.9% (44/192). Compared to emmetropes, prevalence was higher in those with corrected and uncorrected refractive errors. Subjects with uncorrected refractive errors had a higher prevalence (42/166=25.3%) of dry eye compared to those with corrected refractive errors (32/205=15.6%). However, taking emmetropes as controls, the difference in dry eye prevalence in corrected (P =0.144) and uncorrected (P =0.083) refractive error groups did not show any statistical significance.

Table 5:
Prevalence of dry eye as per refractive status

The mean number of complaints in the dry eye group were 6 ± 2.4 and in the non-dry eye group 3.7 ± 2.7. The difference was significant (P = 0.001; 95% CI 1.73 - 2.93).


Past studies suggest that dry eye prevalence ranges from 10.8% to 57.1%.45678 The vast disparity in dry eye prevalence stems mainly from the different dry eye diagnostic criteria employed and different cut-off values for objective dry eye tests. The high prevalence in some studies is also because objective dry eye tests have been performed in patients with positive symptom score (thereby introducing a selection bias) or in patients with rheumatoid arthritis and Sjogren′s syndrome, which have proven dry eye components. Our dry eye prevalence of 18.4% falls within this range.

In our study, dry eye prevalence increased progressively with age, which is consistent with findings in other dry eye studies,79 and the age group 31-40 years showed a relative peak. Hikichi4 and colleagues too found this peak but did not explain it. In our opinion, this peak reflects a dry eye state induced by environmental exposure, to which this age group, being the most active occupationally, is exceptionally prone. This phenomenon may be more common in tropical countries where sunlight and wind exposure is immense. More research is required in tropical climates before a final conclusion can be drawn.

Most studies report a higher prevalence of dry eye in females than males.489 Our study was no exception; 22.8% females in the present study had dry eye compared to 14.9% males and the difference was statistically significant (P =0.024). Menopause causes oestrogen deficiency and a consequent change in the local hormonal milieu of the lacrimal gland. It is thought to decrease tear production and occurrence of dry eye in females. In the present study, the prevalence of dry eye in both postmenopausal and non-menopausal females was 22.8%. The increased prevalence in females may also have been due to the higher number of females with dry eye symptoms seeking advice for ocular problems. We noted a higher dry eye prevalence in rural residents than in urban dwellers, (but statistically not significant) contrary to reports from Japan.4 In our opinion, the increased rural prevalence in our study population was a direct consequence of the overwhelming exposure of rural residents, largely farmers and manual labourers, to sunlight, high temperature and excessive wind.

Exposure to excessive wind, sunlight/high temperature and drugs were significantly related to dry eye causation (Odds ratio: 2.15, 1.91 and 2.04 respectively). The odds were 1.42 for smoking and 1.38 for air pollution. Smoking, air pollution and drugs have been suggested as risk factors in various studies.456789-13 Smoking predisposes the eye to tear film instability by its direct irritant action on the eyes and represents a modifiable risk factor in dry eye causation. A drug too may disrupt one or more components of the tear film causing it to become unstable.

Our finding of increased dry eye prevalence in those with refractive error (corrected and uncorrected) compared to emmetropes is consistent with the observation by Moss and colleagues.9 It is postulated that persons with refractive errors have an increased tendency to rub their eyes and apart from the introduction of infective material, sebum and sweat, could cause the lodgement of particulate foreign substances into the eye that predispose to tear film instability. This study also demonstrated that dry eye was more prevalent in hypermetropes (22.9%) than in myopes (16.8%) and was least in emmetropes (14%). We have no definite explantation for this observed phenomenon despite the suggestions by Shimmura and colleagues.16 The authors contend that it would be worthwhile to incorporate refractive error studies in all dry eye study designs in the future. A limitation of the present study was the exclusion of patients with severe corneal/conjunctival pathology, leading to an underestimation of dry eye prevalence. In most of these cases, the tests would have been difficult to perform and the results impossible to interpret. It is also essential to realise that the dry eye prevalence is usually expected to be higher in contact lens users, who were excluded in this study.

A discussion of the sensitivity, specificity, predictive value of the various dry eye questionnaire responses and the objective tests employed is beyond the scope of this article.


Mahendra Kumar Bijarnia M.Sc., Ph.D (Statistics), Research Officer at the Institute of Development Studies, Jaipur assisted in statistical analyses and the interpretation of the data presented here. The authors thankfully acknowledge Late Dr. Ram Mohan Sahai, under whose guidance this study design was formulated and the study was initiated

1. Lemp MA. Report of the National Eye Institute/Industry Workshop on clinical trials in dry eyes CLAO J. 1995;21:221–32
2. Murube J, Wilson S, Ramos-Esteban J. New horizons in the relief and control of dry eye Highlights of Ophthalmology. 2001;29:55–64
3. Whitcher JP Jr, Gritz DC, Daniels TE. The dry eye: A diagnostic dilemma Int Ophthalmol Clin. 1998;38:23–37
4. Hikichi T, Yoshida A, Fukui Y, Hamano T, Ri M, Araki K, et al Prevalence of dry eye in Japanese eye centers Graefes Arch Clin Exp Ophthalmol. 1995;233:555–8
5. Farrell J, Grierson DJ, Patel S, Sturrock RD. A classification for dry eyes following comparison of tear thinning time with Schirmer tear test Acta Ophthalmol (Copenh). 1992;70:357–60
6. Toda I, Fujishima H, Tsubota K. Ocular fatigue is the major symptom of dry eye Acta Ophthalmol (Copenh). 1993;71:347–52
7. Albietz JM. Prevalence of dry eye subtypes in clinical optometry practice Optometry Vis Sci. 2000;77:357–63
8. Versura P, Cellini M, Torreggiani A, Profazio V, Bernabini B, Caramazza R. Dryness symptoms, diagnostic protocol and therapeutic management: A report on 1,200 patients Ophthalmol Res. 2001;33:221–7
9. Moss SE, Klein R, Klein BE. Prevalence of and risk factors for dry eye syndrome Arch Ophthalmol. 2000;118:1264–8
10. Lee AJ, Lee J, Saw SM, Gazzard G, Koh D, Widjaja D, et al Prevalence and risk factors associated with dry eye symptoms: A population based study in Indonesia Br J Ophthalmol. 2002;86:1347–51
11. Chopra SK, Saramma G, Daniel R. Tear film break-up time in non-contact lens wearers and contact lens wearers in normal Indian population Indian J Ophthalmol. 1985;33:213–16
12. Crandall DC, Leopold IH. The influence of systemic drugs on tear constituents Ophthalmology. 1979;86:115–25
13. Gupta SK, Gupta V, Joshi S, Tandon R. Subclinically dry eyes in urban Delhi: An impact of air pollution? Ophthalmologica. 2002;216:368–71
14. Bandeen-Roche K, Munoz B, Tielsch M, West SK, Schein OD. Self-reported assessment of dry eye in a population-based setting Invest Ophthalmol Vis Sci. 1997;38:2469–75
15. Norn MS. Lissamine green: Vital staining of cornea and conjunctiva Acta Ophthalmol (Copenh). 1973;51:483–91
16. Shimmura S, Shimazaki J, Tsubota K. Results of a population-based questionnaire on the symptoms and lifestyles associated with dry eye Cornea. 1999;18:408–11

Dry Eye; Lissamine Green; Tear film Break-Up Time; Schirmer′s Test; Dry Eye Questionnaire

© 2005 Indian Journal of Ophthalmology | Published by Wolters Kluwer – Medknow