Ophthalmic & Physiological Optics ISSN 0275-5408
Refractive errors in a Brazilian population: age and sex distribution E. Corrente2, Paula Opromolla1, Carlos Roberto Padovani2 and Fabio H. Ferraz1, Jose 1 Silvana A. Schellini Ophthalmology Departament, Faculdade de Medicina de Botucatu, Universidade estadual Paulista, S~ao Paulo, Brazil, and 2Biostatistics Departament, Instituto de Bioci^ encias de Botucatu, Universidade estadual Paulista, S~ao Paulo, Brazil
1
Citation information: Ferraz FH, Corrente JE, Opromolla P, Padovani CR & Schellini SA. Refractive errors in a Brazilian population: age and sex distribution. Ophthalmic Physiol Opt 2015; 35: 19–27. doi: 10.1111/opo.12164
Keywords: associated factors, Brazil, population study, prevalence, refractive error Correspondence: Silvana A. Schellini E-mail address: [email protected] Received: 5 March 2014; Accepted: 19 September 2014; Published Online: 24 October 2014
Abstract Purpose: To determine the prevalence of refractive errors and their distribution according to age and sex in a Brazilian population. Methods: This population-based cross-sectional study involved 7654 Brazilian inhabitants of nine municipalities of Sao Paulo State, Brazil, between March 2004 and July 2005. Participants aged >1 year were selected using a random, stratified, household cluster sampling technique, excluding individuals with previous refractive or cataract surgery. Myopia was defined as spherical equivalent (SE) ≤0.5D, high myopia as SE ≤3.0D, hyperopia as SE ≥+0.5D, high hyperopia as SE ≥+3D, astigmatism as ≤-0.5DC and anisometropia as ≥1.0D difference between eyes. Age, sex, complaints and a comprehensive eye examination including cycloplegic refraction test were collected and analysed using descriptive analysis, univariate and multivariate methods. Results: The prevalence of astigmatism was 59.7%, hyperopia 33.8% and myopia was 25.3%. Astigmatism had a progressive increase with age. With-the-rule (WTR) axes of astigmatism were more frequently observed in the young participants and the against-the-rule (ATR) axes were more frequent in the older subjects. The onset of myopia occurred more frequently between the 2nd and 3rd decades of life. Anisometropia showed a prevalence of 13.2% (95% CI 12.4–13.9; p < 0.001). There was an association between age and all types of refractive error and hyperopia was also associated with sex. Hyperopia was associated with WTR axes (odds ratio 0.73; 95% CI: 0.6–0.8; p < 0.001) and myopia with ATR axes (odds ratio 0.66; 95% CI: 0.6–0.8; p < 0.001). Conclusions: Astigmatism was the most prevalent refractive error in a Brazilian population. There was a strong relationship between age and all refractive errors and between hyperopia and sex. WTR astigmatism was more frequently associated with hyperopia and ATR astigmatism with myopia. The vast majority of participants had low-grade refractive error, which favours planning aimed at correction of refractive error in the population.
Introduction Refractive Errors (RE) were recently recognised as a public health problem and the leading cause of visual impairment (VI) to approximately 153 million individuals worldwide, of which approximately 8 million are considered blind. It is
estimated that 314 million people have some degree of VI, and RE are a significant cause of VI that can be corrected.1,2 Therefore, RE are considered among the top five priorities of the World Health Organization (WHO) in combating visual dysfunction within the policies of VISION 2020 and regular screening programs with glasses provided at low
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
19
Refractive error in a Brazilian population
F H Ferraz et al.
cost or free of charge for those who have RE was suggested.3 The incidence of RE has increased in recent years having a direct association with some risk factors, such as race and years of schooling.4,5 The need for the correction of RE is significantly greater in urban and literate individuals than in the rural population;3 the lack of RE correction increases with age, mainly linked to low income, low education level and inversely related to access to eye care services and health plans.6,7 Wearing glasses to correct near-sightedness vision in some regions was more frequent among men and uncorrected visual acuity (VA) was predominant above 70 years of age.7 Uncorrected RE are responsible for 16% of blindness and 46% of VI in India and low purchasing power and poor access to health services are the main barriers to the rural population accessing proper correction of myopia and presbyopia.8 Data on the prevalence of RE are scarce in Brazil. VI due to uncorrected RE was estimated at 42% and prevalence of astigmatism being observed more frequently above 70 years of age was 50.4% of the population sample.9,10 The aim of this study was to assess the prevalence of RE and its components in a population of nine municipalities located in the mid-west region of the state of S~ao Paulo, Brazil and to identify the distribution and profile of RE according to specific segments of the sample and their possible associations with age and sex. Methods Sampling procedure A population-based cross-sectional ophthalmic survey of households was conducted in the mid-west region of S~ao Paulo State, Brazil. Census data (Instituto Brasileiro de Geografia e Estatıstica, 2000) was used as the sampling frame. The eligible population consisted of permanent, non-institutionalised residents ≥1–90 years of age, between March 2004 and July 2005. The exclusion criteria included prior cataract or refractive surgery. The study protocol was reviewed and approved by the Institutional Review Board (IRB) (Scientific and Ethics committee) of the Botucatu Medical School, S~ao Paulo, Brazil, and the study was conducted in accordance with the tenets of the Declaration of Helsinki. Permission was obtained from the Health Secretary from each village before starting data collection; informed written consent was obtained in the presence of other family members, and each individual was free to decide whether to participate in the study. The region has 38 municipalities and all were invited to participate. Nine accepted the invitation and were included in the study. The total population from those nine cities in the time of the study was estimated to be 85 000. The 20
included municipalities and respective population in the year 2003 are presented in Table 1. All examinations were carried out using a mobile eye unit. Data collection The participants were selected using a random, stratified, household cluster sampling technique. The selection of households to be evaluated was done systematically according to local census data: the first house was selected randomly; the next house was the sixth house on the evennumbered side of the street and so on, successively. Only houses on the even numbered side of the street were chosen. The randomly selected household received a letter of invitation to participate in the study. Those who agreed to participate were contacted by telephone to schedule an appointment. All persons of the household were eligible to participate in the study. If there was no answer when the examiners contacted the household or if people refused to participate in the research, the first house to the right was selected. If the next household refused to participate, the first house to the left side of the initial house was selected, and so on, successively. A single survey team conducted the study and all study personnel underwent training. All procedures were standardised prior to commencement. Specific qualitative observations were performed by one or two members of the team in order to minimise inter-observer variability. Trained eye health care workers filled out a detailed questionnaire investigating identification data (sex, ethnicity referred, age), on wearing and availability of glasses, family history of eye diseases, and the presence of any eye abnormalities, with details. Each participant then received
Table 1. Spatial location, the estimated population for the year 2000 and sample size of each participating municipality of the refractive error prevalence study in the middle west region of Sao Paulo State, Brazil Location†
Municipality
South latitude
West latitude
Population‡
Sample (N)
Arandu polis Areio Bofete Conchas Itaı Pereiras Prat^ania Manduri Taguaı
23°080 11″ 22°400 09″ 23°050 53″ 23°000 48″ 23°240 49″ 23°040 24″ 22°480 34″ 23°000 10″ 23°270 07″
49°030 16″ 48°390 47″ 48°150 31″ 48°000 22″ 49°050 34″ 47°580 32″ 48°390 57″ 49°190 28″ 49°240 38″
6065 10 296 7356 14 904 21 039 6226 3950 8271 7468
746 758 692 1013 1020 895 697 1020 813
†
Google Earth, 2012. IBGE (Demographic census in the year 2000).
‡
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
F H Ferraz et al.
a vision and eye examination. The first procedure was evaluation of presenting visual acuity which was measured for the right eye followed by the left with a tumbling E Snellen chart illuminated within a light box at 5 m. The visual acuity was then retested with the patients existing refractive correction. Autorefraction (Topcon KR-7000, Tokyo, Japan) was performed for all subjects over 3 years old independent of the visual acuity. Objective refraction testing of children ≤3 years of age was performed using retinoscopy. Subjective refraction was performed only for those with visual acuity worse than 0.30 logMAR (Snellen 6/12 or 20/40). For individual ≤40 years of age the refractive examination was performed under cycloplegia, obtained using two drops of cyclopentolate 1% in each eye with the refractive examination performed 30 min after instillation of the drops. The optimal visual acuity was recorded using the result of this refraction. If the subject was unable to read the largest letter at 5 m with the objective refraction, testing was repeated at 1 m. If they were unable to read the largest letter at 1 m then the visual acuity was recorded as count fingers, hand movements, light perception or no light perception. To enable statistical analysis, the visual acuity was converted to a logarithm of the minimum angle of resolution (logMAR). Slit lamp biomicroscopy (Shin Nippon SL101, http:// www.shin-nippon.jp) was performed. Fundus examination was performed at the slit lamp utilising 90D Volk lens under mydriasis using tropicamide 1% in those who had not received instillation of cyclopentolate eye drops. Intraocular pressure was evaluated with a non-contact tonometer (Canon TX-F, http://www.canon-europe.com/ Medical/Eye_Care/TX-F/). If intra-ocular pressure was higher than 21 mmHg, the measurement was repeated using a Goldman tonometer. Definitions of refractive error The spherical equivalent (SE) was calculated as the spherical error plus half the cylindrical error. The definitions of refractive error (RE) were adapted from the Baltimore Eye Study.5 Moderate and high myopia were both included as high myopia. Therefore, myopia was defined SE ≤0.5D, moderate to high myopia ≤3.0D, hyperopia SE ≥+0.5D, moderate to high hyperopia ≥+3D and astigmatism ≤0.5DC. Anisometropia was defined as the difference in SE between the right and left eyes ≥1.0D. We categorised the axis of astigmatism into three groups: with-the-rule (WTR; axes of negatively powered correcting cylinder between 0 and 19° and between 160 and 180), against-the-rule (ATR; axes between 70 and 109°) and oblique (axis between 20 and 69° and between 110 and 159°).
Refractive error in a Brazilian population
Statistical analysis There was a high correlation between right and left eye RE data (Spearman r = 0.88), so we decided to use the right eye data for RE analysis. A descriptive analysis was performed using the mean, median and respective measures of dispersion (standard deviation and interquartile range). The proportion and prevalence data are presented in graphs, adopting 95% confidence intervals (CI) and p-values (significant at the p < 0.05 level). The demographic associations of RE with age and sex were assessed using univariate and multivariate analysis (SPSS for Windows, 15 version, SPSS Inc., www.ibm.com/SPSS_Statistics), multiple logistic regression models and adjusted odds ratios (OR) with 95% CI.
Results Of the 8010 subjects enumerated from 3600 residences, 7654 subjects (95.5%) from 3012 residences were examined. The relative contribution of the municipalities was similar, ranging from 9.0 to 13.3%. The reasons for excluding individuals included previous cataract in 137 (1.7%) or refractive surgery in 46 (0.6%) subjects. Of the study participants, 62.7% were women, with a predominance of females in all the age groups. The ages ranged from one to 96 years, with a mean age of 36.9 21.1 years and a median of 40 years (IQ 21–59). Most participants were in the second (15.5%) and fifth (17.1%) decades of life, with a progressive reduction in the higher age groups. The main complaint was asthenopia, which was observed in 59.6% of the participants, most often between 1 and 9 years of age (96.1%) and in the second decade of life (85.9%). The complaint of poor near vision occurred in 39.6% of the participants, mostly after age 40, and complaint of poor distance vision occurred in 23.8% of the participants, more often in young people. The overall prevalence of RE and its average level is shown in Table 2. The distribution of all RE according to sex and age are summarised in Figures 1 and 2. Astigmatism was present in all of the age groups, with a progressive increase with age and with no difference between men and women, reaching 71.1% (95% CI: 65–77.1; p < 0.001) and 74.3% (95% CI: 69.3–79.2; p < 0.001), respectively after 70 years of age. The average cylinder power was 1.16 1.32D, with a slight increase after the eighth decade, reaching 1.41 0.98D. Hyperopia occurred less frequently under 40 years of age, with a significant increase after the fifth decade of life, reaching a frequency of 55.6% for men (95% CI: 50.9–60.3; p < 0.001) and 65.6% for women (95% CI: 61.9–69.2; p < 0.001) in the sixth decade of life. Moderate to high
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
21
Refractive error in a Brazilian population
F H Ferraz et al.
hyperopia affected more women after the age of 60 (15.5– 95.0% CI: 12.1–18.9; p < 0.001). Myopia was more common between the second to third decade of life (43.3–95% CI: 37.5–49.1 for men and Table 2. Distribution of total prevalence, Mean and standard deviation (SD) of the refractive errors according to gender in the inhabitants of middle west region of S~ ao Paulo State, Brazil Refractive error
Prevalence (%)
Mean SD
Myopia Male Female Hyperopia Male Female Astigmatism Male Female
25.3 24.3 26.4 33.8 31.6 34 59.7 60.35 59.32
1.93 1.88 1.97 1.61 1.40 1.73 1.06 1.04 1.07
2.38 2.16 2.49 1.27 1.16 1.31 0.92 0,82 0.97
42.1–95% CI: 38.4–45.7 for women). After this age, there was a progressive decrease in the subsequent age groups, with a further increase in the seventh decade of life (22.3% with 95% CI: 16.7–27.8 for men and 20.6% with 95% CI: 16.0–25.1 for women). Moderate to high myopia was more common in the third decade, affecting 8.8% of males (95% CI: 5.5–12.1; p < 0.001). Anisometropia occurred in 13.2% (95% CI: 12.4–13.9; p < 0.001) of the participants, with a similar distribution for both sexes and was most common after age 60. Axis of astigmatism No differences associated with sex were observed in the axis of astigmatism but there were differences associated with age; WTR astigmatism was more frequently observed in the first and second decades of life, with a frequency of 52.6% (95% CI: 48.7–56.6; p < 0.001) and was lower among the
Figure 1. Distribution of prevalence (%) of refractive errors according to age in males of middle west region of S~ao Paulo State, Brazil.
Figure 2. Distribution of prevalence (%) of refractive errors according to age in females of middle west region of S~ao Paulo State, Brazil.
22
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
F H Ferraz et al.
older age groups, with values over 70 of 11.4% (95% CI: 8.2–14.6; p < 0.001). The opposite was observed with ATR astigmatism; the frequency in the first decade was 15.8% (95% CI: 12.3–19.4; p < 0.001), rising gradually to 55.3% (95% CI: 50.3–60.2; p < 0.001) after 70 years of age. Oblique astigmatism showed no significant variations with age, remaining at approximately 33.6% in almost all of the age groups (Figure 3). Univariate and multivariate analyses (logistic regression) A univariate analysis with RE as dependent variables indicated astigmatism was related to age and had an age-related odds ratio (OR) of 0.48 (95% CI: 0.37–0.63; p < 0.001) and 0.59 (95% CI: 0.46–0.76; p < 0.001) for the first two age groups, indicating that these individuals were half as likely to have astigmatism as those >70 years of age. The OR for the other age groups was relatively lower, but this change was not statistically significant. No significant association was found between astigmatism and sex. There was an association between hyperopia and moderate to high hyperopia with age (p < 0.001), sex (p = 0.02 and p < 0.001) and cylindrical axis (p < 0.001). Hyperopia and moderate to high hyperopia were strongly associated with age, with a lower OR of 0.56 (95% CI: 0.45–0.7; p < 0.001) until the fifth decade of life, showing a low proportion of hyperopic error before this age, with half as great a chance of finding hyperopia before 50 years of age. The OR for the sixth and seventh decades was 1.82 (95% CI: 1.4–2.3; p < 0.001), with nearly twice the prevalence in these age groups.. The OR for sex was 0.8 (95% CI: 0.7–0.9; p < 0.001), such that men had a 0.79-fold lower chance of having hyperopia than women. For moderate to high hyperopia, this difference was even more pronounced, with
Refractive error in a Brazilian population
an OR of 0.4 (95% CI: 0.3–0.6; p < 0.001), revealing a chance to find high hyperopia in men half as frequently as in women. With respect to the axis of astigmatism, WTR astigmatism showed a negative association with hyperopia and with moderate to high hyperopia, compared to ATR astigmatism (OR 0.7; 95% CI: 0.6–0.8; p < 0.001 and OR 0.6; 95% CI: 0.4–0.8; p < 0.001, respectively). A possible relationship between myopia and sex (p = 0.04), age (p < 0.001) and cylindrical axis (p = 0.001) was detected and confirmed by logistic regression. For moderate to high myopia, this association was observed only with age. WTR of astigmatism was more frequently associated with myopia than ATR astigmatism (OR 0.66; 95% CI: 0.6–0.8, p = 0.001). Similarly, moderate to high myopia was twice as likely in the second and third decades of life, compared with after 70 years of age (OR 1.9; 95% CI: 1.1–3.4; p < 0.001, and OR 2.2; 95% CI: 1.3–3.9; p < 0.001), indicating a two-fold greater chance for individuals in these age groups of having moderate to high myopia compared with those who were older. Between 50 and 59 years of age, the OR was 0.46 (95% CI: 0.2–0.9; p < 0.001), indicating that those individuals were half as likely to have moderate to high myopia compared with those over 70 years of age. Also, anisometropia was related to age (p < 0.001). Multiple logistic regression models with RE as dependent variables showed that myopia was more prevalent in the second and third decades, with an OR of 3.16 (95% CI: 2.4–4.1; p < 0.001) and 2.57 (95% CI: 2.0–3.3; p < 0.001), respectively, compared with the eighth decade of life, showing that within these age groups, an individual has a two-to three-fold greater chance of having myopia than older men. However, we found a much lower prevalence between 50 and 69 years of age; the OR for the sixth decade was 0.41
Figure 3. Distribution of frequency (%) of astigmatism axis according to age in the inhabitants of middle west region of S~ao Paulo State, Brazil.
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
23
Refractive error in a Brazilian population
F H Ferraz et al.
(95% CI: 0.3–0.6; p < 0.001), showing a significant drop in prevalence in this range, with half the chance of having myopia compared with the next age group. The association with age was also observed for anisometropia; the OR reached 0.16 (95% CI: 0.1–0.2; p < 0.001) between 40 and 49 years of age, indicating a nearly ten-fold reduction in the chance of finding anisometropia in this age group than in individuals in the eighth decade of life. Table 3 shows the distribution of RE prevalence in the present study compared to reports from the literature. Discussion This study was developed to search for RE in a large number of participants and applying a complex sampling design which provided a representative prevalence estimate for the non-institutionalised Brazilian population. The adherence of the nine municipalities in participating in the research from the 38 that were initially invited might be dependent on the lack of ophthalmic assistance to the population, adding selective bias to our results. Another bias may be the recruitment method, which allowed the inclusion of a majority of female participants probably because the tests were performed at times of the day when men were more likely to be at work. Moderate and high myopes were the main focus of this research, which was directed at the need for and dependence on
spectacles. We believe that myopia worse than -3.00 D requires correction. The highest concentration of SE values remained between 1.62D and 1.85D where almost 80% had visual acuity better than 0.15 logMAR (Snellen 6/9+ or 20/30+). Thus, most individuals in the general population have low RE that does not require corrective lenses, which favour the strategic planning of health programmes and ocular refraction. The most frequent complaint was asthenopia, most likely because this is directly related to astigmatism, especially low-grade. In a study conducted in Colombia with 60 individuals from 6 to 15 years old with low RE, it became clear that the use of corrective lenses was able to reduce symptoms associated with asthenopia, such as prolonged eye hyperaemia, tearing, burning and headache, by almost 30%.11 The other complaints of poor near vision occurred mainly after the 40 years and is probably associated with presbyopia; poor distance vision in the young was associated with myopia. Despite the existence of previous data regarding the association of RE and ethnicity emphasising black populations had lower rates of RE than white, except for hyperopia prevalence, which was comparable in black and white women,5 we decided to focus our analysis on the assessment of RE with age and sex mainly because miscegenation is common in Brazil.
Table 3. Distribution of refractive errors prevalence of the present study (Ferraz et al.) compared to the literature Authors
Place
Participants
Age (years)
Astigmatism%
Hyperopia%
Myopia%
Ferraz et al. (2014)† Adhikari et al. (2014)14§ Lan et al. (2013)15 Wu et al. (2013)16 Mehari & Yimer (2013)3 Casson et al. (2012)12 Kleinstein et al. (2000)13 Wolfram et al. (2014)17 Kim et al. (2013)18 Yoo et al. (2013)19 Pan et al. (2013)20 Schellini et al. (2009)10 Tarczy– Hornoch et al. (2006)21 Bourne et al. (2004)22 Hashemi et al. (2004)23 Wong et al. (2000)24 Dandona et al. (1999)8
Sao Paulo, Brazil Nepal, India Guangzhou, China Shandong, China Ethiopia, Afrique Lao, Asia Beaver Dam, USA Gutenberg, Germany National Survey, Korea Namil, Korea Singapore Botucatu, Brazil La Puente, California Bangladesh, India Tehran, Iran Tanjong Pagar, Singapore Andhra Pradesh, India
7654 2000 2478 6026 4238 2899 2500 15 010 22 562 1532 10 033 2485 5927 12 782 4565 1232
59.7 (+1.25) 31.8 (>+0.5) 24.2 (>+0.5) 41.8 (>+0.5) 31.5 (>+0.5) 51.2 (>+0.5) – 20.6 (>+0.5) 26.0 (>+0.5) 28.4 (>+0.5) 53.4 (>+0.5) 9.8 46.9 (>+0.5) 49.0 (>+0.5)
25.3 (
Refractive errors in a Brazilian population: age and sex distribution E. Corrente2, Paula Opromolla1, Carlos Roberto Padovani2 and Fabio H. Ferraz1, Jose 1 Silvana A. Schellini Ophthalmology Departament, Faculdade de Medicina de Botucatu, Universidade estadual Paulista, S~ao Paulo, Brazil, and 2Biostatistics Departament, Instituto de Bioci^ encias de Botucatu, Universidade estadual Paulista, S~ao Paulo, Brazil
1
Citation information: Ferraz FH, Corrente JE, Opromolla P, Padovani CR & Schellini SA. Refractive errors in a Brazilian population: age and sex distribution. Ophthalmic Physiol Opt 2015; 35: 19–27. doi: 10.1111/opo.12164
Keywords: associated factors, Brazil, population study, prevalence, refractive error Correspondence: Silvana A. Schellini E-mail address: [email protected] Received: 5 March 2014; Accepted: 19 September 2014; Published Online: 24 October 2014
Abstract Purpose: To determine the prevalence of refractive errors and their distribution according to age and sex in a Brazilian population. Methods: This population-based cross-sectional study involved 7654 Brazilian inhabitants of nine municipalities of Sao Paulo State, Brazil, between March 2004 and July 2005. Participants aged >1 year were selected using a random, stratified, household cluster sampling technique, excluding individuals with previous refractive or cataract surgery. Myopia was defined as spherical equivalent (SE) ≤0.5D, high myopia as SE ≤3.0D, hyperopia as SE ≥+0.5D, high hyperopia as SE ≥+3D, astigmatism as ≤-0.5DC and anisometropia as ≥1.0D difference between eyes. Age, sex, complaints and a comprehensive eye examination including cycloplegic refraction test were collected and analysed using descriptive analysis, univariate and multivariate methods. Results: The prevalence of astigmatism was 59.7%, hyperopia 33.8% and myopia was 25.3%. Astigmatism had a progressive increase with age. With-the-rule (WTR) axes of astigmatism were more frequently observed in the young participants and the against-the-rule (ATR) axes were more frequent in the older subjects. The onset of myopia occurred more frequently between the 2nd and 3rd decades of life. Anisometropia showed a prevalence of 13.2% (95% CI 12.4–13.9; p < 0.001). There was an association between age and all types of refractive error and hyperopia was also associated with sex. Hyperopia was associated with WTR axes (odds ratio 0.73; 95% CI: 0.6–0.8; p < 0.001) and myopia with ATR axes (odds ratio 0.66; 95% CI: 0.6–0.8; p < 0.001). Conclusions: Astigmatism was the most prevalent refractive error in a Brazilian population. There was a strong relationship between age and all refractive errors and between hyperopia and sex. WTR astigmatism was more frequently associated with hyperopia and ATR astigmatism with myopia. The vast majority of participants had low-grade refractive error, which favours planning aimed at correction of refractive error in the population.
Introduction Refractive Errors (RE) were recently recognised as a public health problem and the leading cause of visual impairment (VI) to approximately 153 million individuals worldwide, of which approximately 8 million are considered blind. It is
estimated that 314 million people have some degree of VI, and RE are a significant cause of VI that can be corrected.1,2 Therefore, RE are considered among the top five priorities of the World Health Organization (WHO) in combating visual dysfunction within the policies of VISION 2020 and regular screening programs with glasses provided at low
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
19
Refractive error in a Brazilian population
F H Ferraz et al.
cost or free of charge for those who have RE was suggested.3 The incidence of RE has increased in recent years having a direct association with some risk factors, such as race and years of schooling.4,5 The need for the correction of RE is significantly greater in urban and literate individuals than in the rural population;3 the lack of RE correction increases with age, mainly linked to low income, low education level and inversely related to access to eye care services and health plans.6,7 Wearing glasses to correct near-sightedness vision in some regions was more frequent among men and uncorrected visual acuity (VA) was predominant above 70 years of age.7 Uncorrected RE are responsible for 16% of blindness and 46% of VI in India and low purchasing power and poor access to health services are the main barriers to the rural population accessing proper correction of myopia and presbyopia.8 Data on the prevalence of RE are scarce in Brazil. VI due to uncorrected RE was estimated at 42% and prevalence of astigmatism being observed more frequently above 70 years of age was 50.4% of the population sample.9,10 The aim of this study was to assess the prevalence of RE and its components in a population of nine municipalities located in the mid-west region of the state of S~ao Paulo, Brazil and to identify the distribution and profile of RE according to specific segments of the sample and their possible associations with age and sex. Methods Sampling procedure A population-based cross-sectional ophthalmic survey of households was conducted in the mid-west region of S~ao Paulo State, Brazil. Census data (Instituto Brasileiro de Geografia e Estatıstica, 2000) was used as the sampling frame. The eligible population consisted of permanent, non-institutionalised residents ≥1–90 years of age, between March 2004 and July 2005. The exclusion criteria included prior cataract or refractive surgery. The study protocol was reviewed and approved by the Institutional Review Board (IRB) (Scientific and Ethics committee) of the Botucatu Medical School, S~ao Paulo, Brazil, and the study was conducted in accordance with the tenets of the Declaration of Helsinki. Permission was obtained from the Health Secretary from each village before starting data collection; informed written consent was obtained in the presence of other family members, and each individual was free to decide whether to participate in the study. The region has 38 municipalities and all were invited to participate. Nine accepted the invitation and were included in the study. The total population from those nine cities in the time of the study was estimated to be 85 000. The 20
included municipalities and respective population in the year 2003 are presented in Table 1. All examinations were carried out using a mobile eye unit. Data collection The participants were selected using a random, stratified, household cluster sampling technique. The selection of households to be evaluated was done systematically according to local census data: the first house was selected randomly; the next house was the sixth house on the evennumbered side of the street and so on, successively. Only houses on the even numbered side of the street were chosen. The randomly selected household received a letter of invitation to participate in the study. Those who agreed to participate were contacted by telephone to schedule an appointment. All persons of the household were eligible to participate in the study. If there was no answer when the examiners contacted the household or if people refused to participate in the research, the first house to the right was selected. If the next household refused to participate, the first house to the left side of the initial house was selected, and so on, successively. A single survey team conducted the study and all study personnel underwent training. All procedures were standardised prior to commencement. Specific qualitative observations were performed by one or two members of the team in order to minimise inter-observer variability. Trained eye health care workers filled out a detailed questionnaire investigating identification data (sex, ethnicity referred, age), on wearing and availability of glasses, family history of eye diseases, and the presence of any eye abnormalities, with details. Each participant then received
Table 1. Spatial location, the estimated population for the year 2000 and sample size of each participating municipality of the refractive error prevalence study in the middle west region of Sao Paulo State, Brazil Location†
Municipality
South latitude
West latitude
Population‡
Sample (N)
Arandu polis Areio Bofete Conchas Itaı Pereiras Prat^ania Manduri Taguaı
23°080 11″ 22°400 09″ 23°050 53″ 23°000 48″ 23°240 49″ 23°040 24″ 22°480 34″ 23°000 10″ 23°270 07″
49°030 16″ 48°390 47″ 48°150 31″ 48°000 22″ 49°050 34″ 47°580 32″ 48°390 57″ 49°190 28″ 49°240 38″
6065 10 296 7356 14 904 21 039 6226 3950 8271 7468
746 758 692 1013 1020 895 697 1020 813
†
Google Earth, 2012. IBGE (Demographic census in the year 2000).
‡
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
F H Ferraz et al.
a vision and eye examination. The first procedure was evaluation of presenting visual acuity which was measured for the right eye followed by the left with a tumbling E Snellen chart illuminated within a light box at 5 m. The visual acuity was then retested with the patients existing refractive correction. Autorefraction (Topcon KR-7000, Tokyo, Japan) was performed for all subjects over 3 years old independent of the visual acuity. Objective refraction testing of children ≤3 years of age was performed using retinoscopy. Subjective refraction was performed only for those with visual acuity worse than 0.30 logMAR (Snellen 6/12 or 20/40). For individual ≤40 years of age the refractive examination was performed under cycloplegia, obtained using two drops of cyclopentolate 1% in each eye with the refractive examination performed 30 min after instillation of the drops. The optimal visual acuity was recorded using the result of this refraction. If the subject was unable to read the largest letter at 5 m with the objective refraction, testing was repeated at 1 m. If they were unable to read the largest letter at 1 m then the visual acuity was recorded as count fingers, hand movements, light perception or no light perception. To enable statistical analysis, the visual acuity was converted to a logarithm of the minimum angle of resolution (logMAR). Slit lamp biomicroscopy (Shin Nippon SL101, http:// www.shin-nippon.jp) was performed. Fundus examination was performed at the slit lamp utilising 90D Volk lens under mydriasis using tropicamide 1% in those who had not received instillation of cyclopentolate eye drops. Intraocular pressure was evaluated with a non-contact tonometer (Canon TX-F, http://www.canon-europe.com/ Medical/Eye_Care/TX-F/). If intra-ocular pressure was higher than 21 mmHg, the measurement was repeated using a Goldman tonometer. Definitions of refractive error The spherical equivalent (SE) was calculated as the spherical error plus half the cylindrical error. The definitions of refractive error (RE) were adapted from the Baltimore Eye Study.5 Moderate and high myopia were both included as high myopia. Therefore, myopia was defined SE ≤0.5D, moderate to high myopia ≤3.0D, hyperopia SE ≥+0.5D, moderate to high hyperopia ≥+3D and astigmatism ≤0.5DC. Anisometropia was defined as the difference in SE between the right and left eyes ≥1.0D. We categorised the axis of astigmatism into three groups: with-the-rule (WTR; axes of negatively powered correcting cylinder between 0 and 19° and between 160 and 180), against-the-rule (ATR; axes between 70 and 109°) and oblique (axis between 20 and 69° and between 110 and 159°).
Refractive error in a Brazilian population
Statistical analysis There was a high correlation between right and left eye RE data (Spearman r = 0.88), so we decided to use the right eye data for RE analysis. A descriptive analysis was performed using the mean, median and respective measures of dispersion (standard deviation and interquartile range). The proportion and prevalence data are presented in graphs, adopting 95% confidence intervals (CI) and p-values (significant at the p < 0.05 level). The demographic associations of RE with age and sex were assessed using univariate and multivariate analysis (SPSS for Windows, 15 version, SPSS Inc., www.ibm.com/SPSS_Statistics), multiple logistic regression models and adjusted odds ratios (OR) with 95% CI.
Results Of the 8010 subjects enumerated from 3600 residences, 7654 subjects (95.5%) from 3012 residences were examined. The relative contribution of the municipalities was similar, ranging from 9.0 to 13.3%. The reasons for excluding individuals included previous cataract in 137 (1.7%) or refractive surgery in 46 (0.6%) subjects. Of the study participants, 62.7% were women, with a predominance of females in all the age groups. The ages ranged from one to 96 years, with a mean age of 36.9 21.1 years and a median of 40 years (IQ 21–59). Most participants were in the second (15.5%) and fifth (17.1%) decades of life, with a progressive reduction in the higher age groups. The main complaint was asthenopia, which was observed in 59.6% of the participants, most often between 1 and 9 years of age (96.1%) and in the second decade of life (85.9%). The complaint of poor near vision occurred in 39.6% of the participants, mostly after age 40, and complaint of poor distance vision occurred in 23.8% of the participants, more often in young people. The overall prevalence of RE and its average level is shown in Table 2. The distribution of all RE according to sex and age are summarised in Figures 1 and 2. Astigmatism was present in all of the age groups, with a progressive increase with age and with no difference between men and women, reaching 71.1% (95% CI: 65–77.1; p < 0.001) and 74.3% (95% CI: 69.3–79.2; p < 0.001), respectively after 70 years of age. The average cylinder power was 1.16 1.32D, with a slight increase after the eighth decade, reaching 1.41 0.98D. Hyperopia occurred less frequently under 40 years of age, with a significant increase after the fifth decade of life, reaching a frequency of 55.6% for men (95% CI: 50.9–60.3; p < 0.001) and 65.6% for women (95% CI: 61.9–69.2; p < 0.001) in the sixth decade of life. Moderate to high
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
21
Refractive error in a Brazilian population
F H Ferraz et al.
hyperopia affected more women after the age of 60 (15.5– 95.0% CI: 12.1–18.9; p < 0.001). Myopia was more common between the second to third decade of life (43.3–95% CI: 37.5–49.1 for men and Table 2. Distribution of total prevalence, Mean and standard deviation (SD) of the refractive errors according to gender in the inhabitants of middle west region of S~ ao Paulo State, Brazil Refractive error
Prevalence (%)
Mean SD
Myopia Male Female Hyperopia Male Female Astigmatism Male Female
25.3 24.3 26.4 33.8 31.6 34 59.7 60.35 59.32
1.93 1.88 1.97 1.61 1.40 1.73 1.06 1.04 1.07
2.38 2.16 2.49 1.27 1.16 1.31 0.92 0,82 0.97
42.1–95% CI: 38.4–45.7 for women). After this age, there was a progressive decrease in the subsequent age groups, with a further increase in the seventh decade of life (22.3% with 95% CI: 16.7–27.8 for men and 20.6% with 95% CI: 16.0–25.1 for women). Moderate to high myopia was more common in the third decade, affecting 8.8% of males (95% CI: 5.5–12.1; p < 0.001). Anisometropia occurred in 13.2% (95% CI: 12.4–13.9; p < 0.001) of the participants, with a similar distribution for both sexes and was most common after age 60. Axis of astigmatism No differences associated with sex were observed in the axis of astigmatism but there were differences associated with age; WTR astigmatism was more frequently observed in the first and second decades of life, with a frequency of 52.6% (95% CI: 48.7–56.6; p < 0.001) and was lower among the
Figure 1. Distribution of prevalence (%) of refractive errors according to age in males of middle west region of S~ao Paulo State, Brazil.
Figure 2. Distribution of prevalence (%) of refractive errors according to age in females of middle west region of S~ao Paulo State, Brazil.
22
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
F H Ferraz et al.
older age groups, with values over 70 of 11.4% (95% CI: 8.2–14.6; p < 0.001). The opposite was observed with ATR astigmatism; the frequency in the first decade was 15.8% (95% CI: 12.3–19.4; p < 0.001), rising gradually to 55.3% (95% CI: 50.3–60.2; p < 0.001) after 70 years of age. Oblique astigmatism showed no significant variations with age, remaining at approximately 33.6% in almost all of the age groups (Figure 3). Univariate and multivariate analyses (logistic regression) A univariate analysis with RE as dependent variables indicated astigmatism was related to age and had an age-related odds ratio (OR) of 0.48 (95% CI: 0.37–0.63; p < 0.001) and 0.59 (95% CI: 0.46–0.76; p < 0.001) for the first two age groups, indicating that these individuals were half as likely to have astigmatism as those >70 years of age. The OR for the other age groups was relatively lower, but this change was not statistically significant. No significant association was found between astigmatism and sex. There was an association between hyperopia and moderate to high hyperopia with age (p < 0.001), sex (p = 0.02 and p < 0.001) and cylindrical axis (p < 0.001). Hyperopia and moderate to high hyperopia were strongly associated with age, with a lower OR of 0.56 (95% CI: 0.45–0.7; p < 0.001) until the fifth decade of life, showing a low proportion of hyperopic error before this age, with half as great a chance of finding hyperopia before 50 years of age. The OR for the sixth and seventh decades was 1.82 (95% CI: 1.4–2.3; p < 0.001), with nearly twice the prevalence in these age groups.. The OR for sex was 0.8 (95% CI: 0.7–0.9; p < 0.001), such that men had a 0.79-fold lower chance of having hyperopia than women. For moderate to high hyperopia, this difference was even more pronounced, with
Refractive error in a Brazilian population
an OR of 0.4 (95% CI: 0.3–0.6; p < 0.001), revealing a chance to find high hyperopia in men half as frequently as in women. With respect to the axis of astigmatism, WTR astigmatism showed a negative association with hyperopia and with moderate to high hyperopia, compared to ATR astigmatism (OR 0.7; 95% CI: 0.6–0.8; p < 0.001 and OR 0.6; 95% CI: 0.4–0.8; p < 0.001, respectively). A possible relationship between myopia and sex (p = 0.04), age (p < 0.001) and cylindrical axis (p = 0.001) was detected and confirmed by logistic regression. For moderate to high myopia, this association was observed only with age. WTR of astigmatism was more frequently associated with myopia than ATR astigmatism (OR 0.66; 95% CI: 0.6–0.8, p = 0.001). Similarly, moderate to high myopia was twice as likely in the second and third decades of life, compared with after 70 years of age (OR 1.9; 95% CI: 1.1–3.4; p < 0.001, and OR 2.2; 95% CI: 1.3–3.9; p < 0.001), indicating a two-fold greater chance for individuals in these age groups of having moderate to high myopia compared with those who were older. Between 50 and 59 years of age, the OR was 0.46 (95% CI: 0.2–0.9; p < 0.001), indicating that those individuals were half as likely to have moderate to high myopia compared with those over 70 years of age. Also, anisometropia was related to age (p < 0.001). Multiple logistic regression models with RE as dependent variables showed that myopia was more prevalent in the second and third decades, with an OR of 3.16 (95% CI: 2.4–4.1; p < 0.001) and 2.57 (95% CI: 2.0–3.3; p < 0.001), respectively, compared with the eighth decade of life, showing that within these age groups, an individual has a two-to three-fold greater chance of having myopia than older men. However, we found a much lower prevalence between 50 and 69 years of age; the OR for the sixth decade was 0.41
Figure 3. Distribution of frequency (%) of astigmatism axis according to age in the inhabitants of middle west region of S~ao Paulo State, Brazil.
© 2014 The Authors Ophthalmic & Physiological Optics © 2014 The College of Optometrists Ophthalmic & Physiological Optics 35 (2015) 19–27
23
Refractive error in a Brazilian population
F H Ferraz et al.
(95% CI: 0.3–0.6; p < 0.001), showing a significant drop in prevalence in this range, with half the chance of having myopia compared with the next age group. The association with age was also observed for anisometropia; the OR reached 0.16 (95% CI: 0.1–0.2; p < 0.001) between 40 and 49 years of age, indicating a nearly ten-fold reduction in the chance of finding anisometropia in this age group than in individuals in the eighth decade of life. Table 3 shows the distribution of RE prevalence in the present study compared to reports from the literature. Discussion This study was developed to search for RE in a large number of participants and applying a complex sampling design which provided a representative prevalence estimate for the non-institutionalised Brazilian population. The adherence of the nine municipalities in participating in the research from the 38 that were initially invited might be dependent on the lack of ophthalmic assistance to the population, adding selective bias to our results. Another bias may be the recruitment method, which allowed the inclusion of a majority of female participants probably because the tests were performed at times of the day when men were more likely to be at work. Moderate and high myopes were the main focus of this research, which was directed at the need for and dependence on
spectacles. We believe that myopia worse than -3.00 D requires correction. The highest concentration of SE values remained between 1.62D and 1.85D where almost 80% had visual acuity better than 0.15 logMAR (Snellen 6/9+ or 20/30+). Thus, most individuals in the general population have low RE that does not require corrective lenses, which favour the strategic planning of health programmes and ocular refraction. The most frequent complaint was asthenopia, most likely because this is directly related to astigmatism, especially low-grade. In a study conducted in Colombia with 60 individuals from 6 to 15 years old with low RE, it became clear that the use of corrective lenses was able to reduce symptoms associated with asthenopia, such as prolonged eye hyperaemia, tearing, burning and headache, by almost 30%.11 The other complaints of poor near vision occurred mainly after the 40 years and is probably associated with presbyopia; poor distance vision in the young was associated with myopia. Despite the existence of previous data regarding the association of RE and ethnicity emphasising black populations had lower rates of RE than white, except for hyperopia prevalence, which was comparable in black and white women,5 we decided to focus our analysis on the assessment of RE with age and sex mainly because miscegenation is common in Brazil.
Table 3. Distribution of refractive errors prevalence of the present study (Ferraz et al.) compared to the literature Authors
Place
Participants
Age (years)
Astigmatism%
Hyperopia%
Myopia%
Ferraz et al. (2014)† Adhikari et al. (2014)14§ Lan et al. (2013)15 Wu et al. (2013)16 Mehari & Yimer (2013)3 Casson et al. (2012)12 Kleinstein et al. (2000)13 Wolfram et al. (2014)17 Kim et al. (2013)18 Yoo et al. (2013)19 Pan et al. (2013)20 Schellini et al. (2009)10 Tarczy– Hornoch et al. (2006)21 Bourne et al. (2004)22 Hashemi et al. (2004)23 Wong et al. (2000)24 Dandona et al. (1999)8
Sao Paulo, Brazil Nepal, India Guangzhou, China Shandong, China Ethiopia, Afrique Lao, Asia Beaver Dam, USA Gutenberg, Germany National Survey, Korea Namil, Korea Singapore Botucatu, Brazil La Puente, California Bangladesh, India Tehran, Iran Tanjong Pagar, Singapore Andhra Pradesh, India
7654 2000 2478 6026 4238 2899 2500 15 010 22 562 1532 10 033 2485 5927 12 782 4565 1232
59.7 (+1.25) 31.8 (>+0.5) 24.2 (>+0.5) 41.8 (>+0.5) 31.5 (>+0.5) 51.2 (>+0.5) – 20.6 (>+0.5) 26.0 (>+0.5) 28.4 (>+0.5) 53.4 (>+0.5) 9.8 46.9 (>+0.5) 49.0 (>+0.5)
25.3 (
