Study population

A cross-sectional population-based study was carried out from July to September 2013 in order to determine the prevalence of visual impairment and blindness in people aged ≥50 years in the State of Tlaxcala, Mexico. The prevalence of dry eye was also evaluated. A multistage sampling design in keeping with the Rapid Assessment of Avoidable Blindness (RAAB) methodology was used. A total of 39 municipalities were chosen, comprising 4 282 households with 2 574 individuals aged ≥50 years (INEGI, National Institute of Statistics and Geography of Mexico). Of these potentially eligible people, 1 617 accepted to participate in the study, and 1 508 completed the DEQ-5. The evaluation included also several health questionnaires and assessment of visual acuity by trained interviewers at the homes of the participants. People with any visual impairment or general health problems were referred for medical attention with their usual health care provider. This study was performed in accordance with the guidelines of the Declarations of Helsinki. The Institutional Review Boards of Research, Ethics, and Biosecurity of the Conde de Valenciana Institute of Ophthalmology approved the study protocol. Before the interview and blood sampling, all participants signed an informed consent form.

General evaluation and visual acuity

A general questionnaire including sociodemographic information, smoking (age at first cigarette, duration of smoking, and average of cigarettes per day), alcohol consumption (age at first use, duration of alcohol consumption, and average of drinks per day), and chronic comorbidities (such as diabetes, hypertension, and obesity) was administered. Both smoking and alcohol consumption statuses were categorized into never, former, and current use. The intensity of smoking was evaluated by calculating the smoking index:

which quantifies the number of pack-years a person has smoked. Due to the small number of participants with ≥15 pack-years of smoking (4.4%), we categorized the score as follows: never-smoker, <10 pack-years of smoking (low exposure), and ≥10 pack-years of smoking (medium/high exposure). Systolic (SBP) and diastolic (DBP) blood pressure was measured three times using a digital sphygmomanometer, and the average of the last two measurements was used. Hypertension was defined as SBP≥140 mmHg, DBP≥90 mmHg, or intake of antihypertensive drugs regardless of blood pressure values. Diabetes was defined according to the Expert Committee on Diagnosis and Classification of Diabetes Mellitus 2003, as fasting serum glucose >7.0 mmol/l (126 mg/dl) or previous physician’s diagnosis.

Visual acuity and pinhole visual acuity were measured using the Snellen tumbling E chart. Individuals were categorized according to visual acuity in the better eye as normal (better than or equal to 20/60), moderately impaired (worse than 20/60 and equal to 20/200), severely impaired (worse than 20/200 and equal to 20/400), and blind (worse than 20/400). Fasting plasma glucose, total cholesterol, triglycerides, HDL-cholesterol, LDL-cholesterol, and creatinine and albumin levels were determined. All analytical measurements were performed at the Clinical Laboratory at the Conde de Valenciana Institute of Ophthalmology in Mexico City.

Five-item dry eye questionnaire (DEQ-5)

For this study, a previously validated Spanish version of the DEQ-5 questionnaire was used.⁶ The DEQ-5 is a five-item questionnaire that measures a number of symptoms along four dimensions: frequency, intensity in the morning, intensity late in the day, and degree of bother.⁵ The questionnaire score ranges from 0 to 22 points, and the sum of points is categorized as follows: <6, no dry eye symptoms; 6 to 11, mild-moderate dry eye symptoms; and ≥12, severe dry eye symptoms. Persons in the last category must be also examined for the presence of Sjögren&apos;s syndrome (SS), an autoimmune disease that combines dry eyes, dry mouth, and other diseases of the connective tissues.⁵

Statistical analysis

Unadjusted and sex-adjusted prevalence and their respective 95% confidence intervals (95%CI) for dry eye symptoms, both overall and by degree of severity, were estimated by logistic regression analysis. Comparisons between individuals with and without dry eye symptoms were carried out when appropriate, using Pearson chi-square test for categorical variables and analysis of variance (ANOVA) or the Kruskal-Wallis test for continuous variables. The magnitude of the association between dry eye symptoms and various risk factors across severity categories was examined by logistic and multinomial logistic regression models, using the DEQ-5 score, with the participants without dry eye as reference group, and calculating odds ratios (OR) and their 95%CI. Some of the risk factors for dry eye symptoms were age (continuous and dichotomous variable), sex, smoking status (as nominal variable [never, former, and current]), smoking index (never-smokers, <10 pack-years, and ≥10 pack-years), ever alcohol consumption (as dichotomous [yes/no] and nominal variable [never, former, and current]), diabetes (yes/no), antihypertensive treatment (yes/no), visual impairment (none, mild/moderate, and severe), and wearing glasses (yes/no). In a model for individuals with diabetes, duration of diabetes (continuous variable) was added. The fit of the models was tested with the Hosmer-Lemeshow deciles of risk. Also, the outliers and influence statistics were evaluated. All analyses were performed with STATA 14.0 (Stata Corporation, College Station, Tex.).

Description of the study population

The study population included 1 508 individuals (900 women [59.6%] and 608 men [40.3%]; the mean age was 64.7 years [s.d. 10.6]). According to the classification of the degree of severity of dry eye symptoms, 887 (58.8%) participants did not have the condition, 455 (30.2%) had mild/moderate symptoms, and 166 (11.0%) had severe dry eye symptoms. Comparison of participants across DEQ-5 categories showed significant differences by sex, occupation, wearing of glasses, visual acuity in the better eye, and visual function (table I). Although no significant differences between dry eye symptoms groups by smoking status or smoking index were observed, the smoking index of ≥10 pack-years was more prevalent in current than in former smokers with mild/moderate (36.9 vs. 13.4%, respectively) or severe dry eye symptoms (36.8 vs. 11.8%, respectively).

Prevalence of mild/moderate and severe dry eye symptoms

The overall prevalence of any degree of severity of dry eye symptoms was 41.1% (95%CI 38.6-43.6). The prevalence of any severity of dry eye symptoms was significantly higher in women than in men (45.3% [95%CI 42.1-48.6] vs. 35.0% [95%CI 31.3-38.9], respectively), as well as in individuals with ≥10 pack-years of smoking (58.4% [95%CI 48.5-67.7]) and in those with<10 pack years of smoking (46.3% [95%CI 41.4-51.2]), compared to never-smokers (36.1% [95%CI 32.6-39.8]). The prevalence of any severity of dry eye symptoms was also higher in current (42.6% [95%CI 38.7-46.5]) and former (45.0% [95%CI 40.3-49.8]) alcohol consumers than in never-consumers (35.1% [95%CI 30.6-39.8]), as well as in individuals with moderate visual impairment (59.4% [95%CI 46.8-70.9]) and severe visual impairment/blindness (68.6% [95%CI 56.4-78.6]), compared to those with normal visual acuity (38.9% [95%CI 36.4-41.6]). A similar trend was observed in the prevalence of mild/moderate and severe dry eye symptoms (table II). Additionally, the prevalence of any degree of dry eye symptoms increased significantly with worsened visual acuity in those who did not wear glasses (normal vision, 34.6%; moderate visual impairment, 54.3%; and severe visual impairment/blindness, 75.8%; p<0.001). No significant trend due to visual impairment was observed in participants who wore glasses.

Risk factors associated with the prevalence of dry eye symptoms

In a multiple logistic regression analysis, the odds for any severity of dry eye symptoms were higher in women (OR=2.26, 95%CI 1.70-3.00), in individuals with a smoking index of <10 pack-years (OR=1.40, 95%CI 1.05-1.87) and ≥10 pack-years (OR=1.29, 95%CI 1.44-3.63), subjects with a history of ever alcohol consumption (OR=1.31, 95%CI 1.02-1.70), and those receiving antihypertensive treatment (OR=1.29, 95%CI 1.00-1.65) (table III). In addition, when the association between mild/moderate and severe dry eye symptoms and a smoking index of ≥10 pack-years was evaluated by multinomial logistic regression, the odds remained largely unchanged (OR=2.29, 95%CI 1.39-3.78 and OR=2.27, 95%CI 1.09-4.76, respectively).

In a multiple logistic regression analysis in individuals with diabetes, the likelihood of dry eye symptoms was higher in women (OR=2.28, 95%CI 1.30-3.97), in individuals with a smoking index of ≥10 pack-years (OR=2.75, 95%CI 1.12-6.73), and in those with longer diabetes duration (OR per 5 yrs. =1.14, 95%CI 1.03-1.25). No difference was observed in those taking antihypertensive medication (table IV). We were not able to estimate the association between smoking index categories and moderate and severe dry eye symptoms, because of the small number of participants with a high smoking index (≥10 pack-years).