The study was developed in Villa Maria City, in the province of Córdoba, Argentina from May 2008 to November 2009. A simplified general method for clustersample size surveys of health in developing countries (probability proportion to size, PPS) was used ⁽²⁴⁾. The sample size was calculated from the size of the population for a confidence interval of 5% to obtain a confidence level of 95% ^(24,25). Eleven districts were randomly selected from Villa Maria City. Two stage cluster design were used. First stage sampling was fractions selection. All the fractions available were used (5 fractions). On the second stage two districts inside each fraction were randomly selected. We also added one district from Villa Nueva. These randomized districts were selected by censal-ratio methods, defined according to the concentration of the Villa Maria population given by the 2001 Provincial Census. Household were randomly selected (Kish Grid) from each ratio and all the adults >40 years of age were targeted. A local University helped in promoting the study. 

The tenets of the Declaration of Helsinki were followed and the study was approved by the Institutional Ethics Committee Oulton-Romagosa Joint Committee on Clinical Investigation (C.I.E.I.S. Oulton Romagosa). All concerned volunteers were briefly informed about the scope and purpose of the study. Subjects were ensured strict confidentiality, and then informed consent was taken from each of the participants before procedure. Skilled interviewers approached 40+ year-old members of the targeted households, who explained the characteristics of the study, completed the preliminary data forms and performed a demographic interview at the household. This interview included information about education achievement and family income. Then the subjects were invited to have a complete eye examination at peripherals centers (11 state outpatient clinics). This last exam day was scheduled on Saturdays to avoid losing subjects due to work responsibilities. Only subjects aged 40+, who had good cooperation for the eye examination were included in the analysis.

A team of 13 trained ophthalmologists of the Centro Privado de Ojos Romagosa (Fundación VER) performed the eye examinations at the peripheral center scheduled for that visit. The tasks were distributed according to the field of expertise of each trained doctors (glaucoma specialist, retinal specialist, etc.). Visual acuity was obtained with an ETDRS chart (Sloan Letter Folding Eye Chart 10 ft/3 m, Good Lite Company, USA) for each eye at a distance of 3 m. After tonometry, two drops of a commercially available combination of 5% phenylephrine +0.5% tropicamide (Fotorretin, Poen Laboratories, Argentina) were instilled 5 minutes apart, and after 30 minutes rest, with the objective of obtaining a rapid, persistent and maximum possible dilatation for the fundus examination. Auto refraction was performed in both eyes with a Canon R-10 auto refractor (Canon, USA). The mean of five measurements was registered. The eye examination was then completed with biomicroscopy of the cornea and lens, applanation tonometry and fundus examination. Lens opacity was graded using the Lens Opacities Classification System II (LOCS II) ⁽²⁶⁾. The spherical equivalent was calculated based on the mean auto refraction outcome, adding half of the cylinder value to the recorded sphere. Myopia was defined as a spherical equivalent more negative than −0.50 diopters (D) and hyperopia was defined as a spherical equivalent greater than +0.50 D. Anisometropia was defined as a difference of more than 1 D in spherical equivalent between both eyes. Four age groups were defined with intervals of 10 years. For those with astigmatism greater than 1 D, the axis of the astigmatism was classified to be with-the-rule when the minus cylinder was at the 0º±15º, and to be against-the-rule when the minus cylinder was at the 90º±15º. Other astigmatisms were registered as oblique.

The significance of the differences in percentages of prevalence was obtained with chi square test. One-way ANOVA was used to compare means of spherical equivalent by age groups. In addition to descriptive analyses, odds ratios (OR) with 95% confidence interval (95% CI) were calculated with univariate logistic regression analysis. Univariate logistic regression models were performed with myopic, hyperopic and astigmatic refractive error as dichotomous dependent variables, adding age, sex, nuclear cataract, education, and annual family income as the independent variables. All P values less than 0.05 were considered statistically significant. Data analyses were performed with statistical software (SPSS version 15.0, SPSS Inc., Chicago, IL, USA).

The study population consisted of 646 subjects with complete cycloplegic refraction data. The sample included patients over the age of 40 up to 90, with a mean age of 59.6±10.3 years old. Seventy-one point five percent (71.5%) were females. The level of education of the studied subjects was low, having 28.3% of them incomplete primary school and only 5.2% reaching a university level of more than 12 years of study. Only 28% had a family income that reached more than 377 US dollars per month (Table 1).

Table 1 Monthly income in Villa Maria

The present study compares well with the Latino Eye Study ⁽¹⁹⁾, which reported the prevalence of myopia in a Mexican Hispanic population. For the age range 40–50, with a similar cut-off point of −0.50 D, that study reports 19.6% myopia for men and 22.3% for women. In that population, 36% (901/2,472) of males had more than 12 years of education, so their academic achievement was somehow higher than that of Villa Maria subjects and this could be the cause of higher myopia prevalence in US Latinos.

The prevalence in both studies is low compared to that of East Asian urban environments ⁽³⁰⁾, in which subjects have very high academic achievement and low outdoor exposure. The prevalence of high myopia (cut-off point of −5 D) was very low in both Hispanic studies (2.4% for Latino Eye Study and 2.17% for the present one). This prevalence of high myopia in low academic environments with high outdoor exposure probably represents the cases with genetic background for myopia. The Latino Eye Study has not presented data of spherical equivalent or hyperopia prevalence so studies cannot be compared in that sense. It is noteworthy that according to recent genetic analysis in Argentineans, 78.9% of the mean ancestry components come from European descent, 15.8% from Amerindian and only 4.15% from African descent ⁽³⁴⁾. A recent publication from the European Eye Epidemiology Consortium with a similar population, when compared to our study, showed that the prevalence of myopia with a cut point of −0.75 D was 40.2% in subjects aged 40–44. This higher prevalence of myopia may be possibly due to the different academic achievement and outdoor exposure of people living in Argentina compared to Europe ⁽³⁵⁾. Another study on refractive errors in our country ⁽³⁶⁾, based on noncycloplegic refractions in an unselected sample of officeworkers coming for a general health checkup aged 25–65, with high academic degrees (mean of 6 years of University study), showed a prevalence of myopia of 32.9% for subjects aged 40–50 (with the same cut point), much higher than the results obtained in our report (13.5%). This greater prevalence of myopia in the study by Cortinez et al. ⁽³⁶⁾ could be related to greater academic achievement of that sample, as it was shown in other studies ^(30,31). In the present study, the low prevalence of myopia goes along with the lack of academic achievement (90% less than 12 years of studies), and very low income (66.7% less than 377 US dollars). The low prevalence of myopia is in accordance with recent data from Iran in Shahroud ⁽³⁷⁾, where the prevalence has been low in similarly aged subjects. Considering nuclear opacity, only 3.6% of the sample had advanced nuclear opacity, which has been related to the development of myopic shifts in refraction ^(38,39), however, the involved numbers were very low to reach significance. As the number of hyperopic subjects was greater, we found that clear lenses were significantly related to hyperopic refractive error.

As regard astigmatism, we found a prevalence of 26.0%, which is comparable with studies that used a similar methodology as our study showing a range between 18.5–29.6% ^(12,15,40). On the other hand, the prevalence is much higher in studies that use a >0.50 D criterion to define astigmatism ^(40-42). In spite of a majority of withthe-rule astigmatism found in our sample, there was a tendency to increase to against-the-rule astigmatism with age (Table 4). This tendency has already been described in other studies ^(14-20,41). A possible cause of this finding could be changes in lens opacity or more probably due to corneal modifications as other studies have suggested ^(42-45).

In spite of the randomly selected sites according to the local population census, some variables could be biased due to different factors. In this report there was a higher participation of women. This result was probably linked to the voluntary nature of the examination that may reflect factors like time availability and major possibility of awareness about the campaign, cultural factors that make women more predisposed to assist. Many similar studies also showed a higher enrolment of women ^(7,15,19,46). Other concern is the low participation rate in this study, probably related to the fact that the demographic interview was scheduled first and the eye examination in a different visit (26.7% failed to return). Future studies in our country could improve this participation rate if scheduled as only 1-day visit for the whole procedure.

In summary, our work shows the prevalence of refractive error in an adult population of Hispanic residents in Argentina. The prevalence of myopia was very low when compared with the hyperopic refractive defect that was found in the studied participants. Refractive errors are an important cause of vision impairment in Córdoba, Argentina. We believe that this study shows that there is a great need of eye-care services and spectacle distribution to help improving visual-specific quality of life. Since our study data were collected several years ago, further studies are needed to report the trends on refractive error or current situation in the province, to evaluate if new policies or interventions are required currently.