The Aravind Comprehensive Eye Survey (ACES) is a population-based survey of adults 40 years of age or older to assess the burden of ocular morbidity and vision impairment in rural south India. The study was approved and is annually reapproved by the Committee on Human Research at the Johns Hopkins Bloomberg School of Public Health and by the Ethical Review Committee of the Aravind Eye and Children’s Hospitals and adhered to the tenets of the Declaration of Helsinki. The details of the methodology of the study and sample selection have been published.
2 To summarize, a two-stage, random cluster sampling technique was used to identify 50 study sectors from within 50 rural villages of three southern districts of the state of Tamil Nadu in south India. This sample can be considered to be representative of rural areas in southern India but not necessarily of urban areas in southern India or of rural or urban areas in other parts of India.
We measured presenting and best-corrected visual acuity using illiterate E ETDRS charts at 4-m distance. Participants failing to read the largest letters at 4 m were retested at 2- and 1-m distances. Participants were deemed to have sufficient visual acuity to read a particular line if a minimum of four of five letters in a line were identified correctly. Visual acuity was then transformed to log of minimum angle of resolution (logMAR) units. We assigned a visual acuity of 1.7 logMAR units for participants who were unable to read any of the letters, even at 1 m. We used presenting visual acuity for further analysis relating to this manuscript, as visual function in daily life is a function of presenting rather than best-corrected vision.
We considered a person to have uncorrected refractive error if the difference between presenting and best-corrected vision was more than 2 lines and optical correction was responsible for this difference. We used the Lens Opacities Classification System (LOCS) III to grade the lens at the slit lamp under standard testing conditions.
14 We determined individuals to have current, definite age-related cataract if they had LOCS III nuclear opalescence ≥3.0 and/or cortical cataract ≥3.0 and/or posterior subcapsular cataract (PSC) ≥2.0 in either eye.
4 We performed gonioscopy to examine the angles of the anterior chamber and classified the angle based on the system proposed by Shaffer.
15 We measured intraocular pressure with Goldmann applanation tonometry and performed a central 24-2 visual field examination by automated perimetry (Humphrey Visual Field Analyzer; Carl Zeiss Meditec, Dublin, CA) in all study participants. The presence of glaucoma was defined independent of intraocular pressure. The methods of diagnosing and classifying glaucoma have been published previously.
3 Posterior segment examinations were performed with indirect ophthalmoscopy and 78-D lens examinations at the slit lamp after the pupils were dilated. We classified age-related macular degeneration based on the classification system proposed by the International ARM Epidemiologic Study Group.
16
We measured blood pressure in study participants by using standard procedures, and the median of three readings was used in the analysis. We defined systemic hypertension as measured systolic blood pressure ≥ 140 mm Hg and/or diastolic blood pressure ≥ 90 mm Hg or current use of systemic antihypertensive medications.
17 18 We used a glucometer and strips to test for blood sugar levels. We obtained capillary blood for examination after a finger prick with a sterile lancet 2 hours after the subject had eaten breakfast. We defined diabetes as measured postprandial blood sugar of ≥180 mg/dL or current use of blood sugar–lowering medications. Facilities for glycosylated hemoglobin estimations were not available in the study districts during the period of study and hence were not performed.
We used an instrument previously validated for use in this population to determine the vision-specific quality-of-life and visual function scores.
11 This quality-of-life questionnaire included questions concerning the activities of daily living, mobility, social activities, and mental state. For each of these questions, if the subject responded that he or she had any difficulty with the activity, the subject was asked whether the difficulty was related to vision. The quality-of-life scores presented herein are those specific to difficulty with activities with which vision was associated. The visual function questionnaire included questions on general vision, visual perception, sensory adaptation, and depth perception. Social workers previously trained in the administration of these questionnaires conducted interviews at the hospital before visual acuity measurements and ocular examination. Each item in the questionnaire was answered with a four-point scale ranging from “not at all” to “a lot.” We calculated a total score for each of the questions and subscales and expressed this score as a percentage of the total possible obtainable score ranging from 0 to 100, with higher scores indicating better results.
The purpose of these analyses was to examine the associations between quality-of-life scores and demographic, systemic, and ocular morbidity and vision-specific variables. Similarly, we assessed the association between these variables and visual function scores. Demographic variables included age (categorized by decade), sex, education (none, primary or middle, and secondary or higher), and occupation (none, farmer or laborer, professional/business, government). Morbidity included hypertension, diabetes, cataract, glaucoma, age-related macular degeneration, and refractive error. Presenting visual acuity in the better eye was categorized as 20/60 or better, <20/60 to 20/200, <20/200 to 20/400, or worse than 20/400. We also ran models using visual acuity as a continuous variable using the logMAR scale. Quality-of-life scores were skewed to higher values, but visual function scores were approximately normally distributed. However, because of the large sample size, we fit linear regression models for both outcomes and used t-based 95% confidence intervals for the regression coefficients. After selecting the demographic and morbidity covariates to be included in the models, we ran the models, with and without visual acuity. Both sets of models are presented to demonstrate the changes in coefficients associated with quality-of-life and visual function scores, with and without visual acuity as an explanatory variable. We also examined the subscales of quality of life and vision function scores, by using the same linear regression models for the overall scores.