Abstract
purpose. To examine associations between myopia and measures of urbanization in a population-based sample of 12-year-old Australian children.
methods. Questionnaire data on sociodemographic and environmental factors including ethnicity, parental education, and time spent in near work and outdoor activities were collected from 2367 children (75.0% response) and their parents. Population density data for the Sydney area were used to construct five urban regions. Myopia was defined as spherical equivalent refraction ≤ −0.50 D.
results. Myopia prevalence was lowest in the outer suburban region (6.9%) and highest in the inner city region (17.8%), with mean refraction tending toward greater myopia by region (outer suburban to inner city), after adjustment for age, sex, ethnicity, near work, outdoor activity, and parental myopia. Multivariate-adjusted analyses confirmed greater odds for myopia in regions of higher population density (P trend = 0.0001). Myopia was significantly more prevalent among children living in apartment residences than other housing types (χ2 < 0.0001), after adjustment for ethnicity, near work, and outdoor activity. Housing density (measured as the number of houses visible from a front door) was not significantly associated with myopia (χ2 = 0.1). For both European Caucasian and East Asian children, myopia was most prevalent in the inner city region (8.1% and 55.1%, for European Caucasian and East Asian, respectively).
conclusions. The higher myopia prevalence in inner city-urban areas compared with outer suburban areas for this large childhood sample suggest that even moderate environmental differences within a predominantly urban setting may be associated with increased odds of myopia. These findings are consistent with previous reports of rural-urban differences in childhood myopia.
It is now widely accepted that the development of myopia is subject to the influences of genetic and environmental factors, although some researchers have suggested that environmental exposures are the major causes of current differences in myopia prevalence.
1 Evidence supporting the importance of hereditary factors has included twin studies, which report higher correlations of refraction in monozygotic than dizygotic twins,
2 3 whereas factors such as higher levels of education
4 5 6 7 may underscore the widespread trends of increasing myopia among the younger cohorts in adult samples.
8 9 10 11
In children, recent studies have reported significant associations between childhood and parental myopia.
12 13 14 Geographic location may also be important for childhood myopia, with higher myopia prevalence reported in urban compared with rural areas.
15 16 17 18 The impact of urbanization, however, has been specifically reported in only one retrospective study.
19 Studies that control for differences in levels of near work and outdoor activity, as well as parental myopia, are likely to provide a better understanding of subtle impacts from differences within an urban environment.
In this study, we sought to evaluate the impact of urbanization by examining the association of childhood myopia with factors such as the area of residence, type of residential housing, and housing density, in a representative sample of Australian school children living in different urban environments but attending the same schools.
Materials and Methods
The sampling and examination methods for the Sydney Myopia Study, conducted from 2003 to 2005, have been described
20 and are briefly outlined here. Secondary schools throughout the Sydney metropolitan region were stratified by socioeconomic status (SES), and 21 secondary schools, including 2 schools from the top SES decile, were randomly selected for participation. The sample included a proportional mix of government (public) and private/religious schools that was representative of the Sydney region. All year seven students in the selected schools were invited to participate.
Approval for the study was obtained from the University of Sydney Human Research Ethics Committee, the New South Wales Department of Education, and the Catholic Education Office. The study adhered to the tenets of the Declaration of Helsinki. Written informed consent was obtained from at least one parent and the verbal assent of all children was obtained before examination.
Participating children completed a 65-item questionnaire that asked about their home environment and the time spent completing near work or in outdoor activity. Specifically, children were asked to specify the type of housing they lived in (house, apartment, terrace house, or other type) and the proximity of their home to other surrounding homes, shops or high-rise buildings (measured as the number of properties visible from the child’s house; categorized as <5, 5–10, or >10). Near-work activities included completing homework, reading for pleasure, playing musical instruments, using a computer, and playing hand-held console games, video games, or board games. Outdoor activities included general activities, such as playing in one’s own backyard, walking, riding a bike or scooter; leisure activities such having a barbeque, or picnic or spending time at the beach or going hiking; and outdoor sports. (Questionnaires defining these activities are available at http://www.cvr.org.au/sms.htm.)
Parents also completed a 173-item questionnaire that collected detailed sociodemographic data, such as ethnicity, the level of education completed, and the occupation of both parents. To ascertain whether they had myopia, parents were asked whether they needed to use spectacles or contact lenses, the age at first use, and the reason for using spectacles (for distance viewing only, for near work only, or whether spectacles were used for both distance viewing and near work). Spectacle prescriptions were also obtained from parents or their prescribers, where possible. Spectacle prescriptions confirming questionnaire data were available for 23% and 26%, respectively, of the mothers and fathers.
For the purposes of this study, the greater Sydney region was divided into 14 areas, based on spatial units (statistical subdivisions), defined by the Australian Bureau of Statistics (ABS) to delineate relatively homogenous social and economic regions. The 14 areas of greater Sydney were then grouped by population density (ABS data collected in 2004) into five regions: <100 persons/km2 (region 1, outer suburban); 100 to 999 persons/km2 (region 2); 1000 to 1999 persons/km2 (region 3); 2000 to 2999 persons/km2 (region 4); and ≥3000 persons/km2 (region 5, inner city). Participating children were allocated to one of the five regions, according to their residential postcode.
ABS data for 2001, 2003, and 2004 showed that on average, inhabitants of more densely populated regions reported higher levels of educational attainment and greater income, compared with those living in regions with lower population density
(Table 1) .
Examination
After amethocaine was instilled, cycloplegia was induced with cyclopentolate 1% (2 drops). In a few children, tropicamide 1% (1 drop) and phenylephrine 2.5% (1 drop) were also used to obtain adequate mydriasis (minimum pupil diameter 6 millimeters). Autorefraction was performed at approximately 25 to 30 minutes after the last drop (RK-F1 autorefractor; Canon, Tokyo, Japan). Cycloplegic autorefraction was used, as this method of examination has been shown to provide more reliable measurements in young children than cycloplegic retinoscopy or noncycloplegic autorefraction.
21
Definitions
Myopia was defined as spherical equivalent refraction (sphere + ½ cylinder) −0.50 D or less. Parental ethnicity was classified on the basis of self-identification, and in the child if both parents shared that ethnic origin. The ethnic categories (European Caucasian, East Asian, South Asian, Middle Eastern, Pacific Islander, Indigenous Australian, African, and South American) were largely consistent with the Australian Standard Classification of Cultural and Ethnic groups (available at http://www.abs.gov.au, document 1249.0). Children with parents of different ethnicity were classified as having mixed ethnicity. European Caucasian ethnicity comprised 60.0% and East Asian ethnicity 15.0% of the whole sample.
Data Analysis
Data were analyzed using commercial software (SAS ver. 9.1.3; SAS Institute, Cary, NC). Mixed models and generalized estimating equations were used to examine associations and subgroup differences, adjusting for the effects of cluster-sampling. Analyses for myopia and spherical equivalent refraction were performed for the whole sample and separately for the two predominant ethnic subgroups (European Caucasian and East Asian). t-Tests and χ2 tests were used when cluster effects were not significant. All confidence intervals (CIs) are 95%.
Results
Sample Characteristics
Overall, 2367 (75.0%) children attending year 7 in the selected schools had parental permission to participate. Of these, 14 were not examined, because they were absent during the school visit. The mean age of participants was 12.7 years (range, 11.1–14.4 years). Ethnic origins of the participating children were predominantly European Caucasian (64.5%) and East Asian (15.0%). Nonparticipants were similar to participants by sex (51.6% and 50.0% were boys, respectively) and ethnicity (67.8% of nonparticipants were European Caucasian). Overall, 82.8% of nonparticipants and 83.8% of participants attended public schools.
Of the 2353 children examined, only 44 lived outside of the greater Sydney area, boarded at schools, or could not be located by postcode, and were therefore excluded from the analyses. Of the remaining 2309 children with available data, 322 (14.0%) lived in region 1, 703 (30.5%) in region 2, 710 (30.7%) in region 3, 304 (13.2%) in region 4, and 270 (11.7%) in region 5. The majority (94.0%) of children had lived at their current address for 6 months or longer. On average, the samples of children living in more densely populated regions were very slightly but significantly older (
P < 0.0001) and included greater proportions of girls (χ
2 <0.0001) and children of East Asian ethnicity (χ
2 < 0.0001;
Table 2 ). In more densely populated regions, higher levels of parental education and parental myopia were also noted (both χ
2 < 0.0001). The time that children spent in near-work activities was similar across the five regions (
P = 0.1), but significantly less time in outdoor activity was reported by children living in more densely populated areas (
P = 0.0006).
Childhood Myopia and Urbanization
For this cross-sectional sample of 12-year-old children, the prevalence of childhood myopia was lowest (6.9%) in the outer suburban region and highest (17.8%) in the inner city region. The mean childhood spherical equivalent refraction by region (outer suburban to inner city) followed a trend toward greater myopia, after adjustment for age, sex, near work, outdoor activity, and parental myopia
(Table 3) .
Stratifying by ethnicity showed a similar trend of less hyperopic-more myopic refraction from the outer suburban region to the inner city region in both European Caucasian and East Asian ethnic groups
(Table 3) . Across all regions, children of European Caucasian ethnicity were consistently more hyperopic in refraction, with a lower prevalence of myopia, than children of East Asian ethnicity.
Type of Housing and Housing Density
Myopia was significantly more frequent in children who lived in smaller, confined housing types such as terrace houses (21.4%) and apartments (26.3%) than those living in stand-alone or separate houses (11.3%, χ
2 < 0.0001;
Table 4 ). The odds for childhood myopia associated with living in a terrace house was 2.0 (95% CI, 1.4–2.9) and for living in an apartment was 2.2 (95% CI, 1.6–3.1), after adjustment for region. The association between housing type and myopia was significant after separately adjusting for ethnicity (
P = 0.02), near work (
P < 0.0001), and outdoor activity (
P < 0.0001). Housing density (measured as the number of homes seen from the child’s house and categorized as <5, 5–10, or >10) was not significantly associated with childhood myopia in this sample (χ
2 = 0.1).
Multivariate Analyses
In multivariate analyses for childhood myopia, significant variables included region (
P trend = 0.0003), East Asian ethnicity (
P < 0.0001), parental myopia (
P < 0.0001), and outdoor activity (
P = 0.005). In this sample, near work was not significantly associated with childhood myopia (
P = 0.2). The children living in the inner city region had 2.2-fold greater odds for myopia than did those living in the outer suburban region (OR, 2.2; 95% CI, 1.3–3.8,
Table 5 ), after adjustment for age, sex, ethnicity, parental myopia, parental education, and time spent in near work and outdoor activities.
For children of European Caucasian ethnicity, the odds for myopia increased with region (
P trend = 0.0002), such that the inner city region had the greatest odds of having myopia (OR, 7.1; 95% CI, 2.5–19.8, compared with the outer suburban region), after adjustment for age, sex, parental myopia, parental education, and time spent in near work and outdoors
(Table 5) . In multivariate analyses for children of East Asian ethnicity, region was also a significant factor predicting increased odds of myopia (
P trend = 0.0002). There were no statistically significant interactions between ethnicity and our measure of socioeconomic status (interaction term, ethnicity × parental education;
P = 0.07).
Discussion
The findings from this sample of predominantly 12-year-old Sydney school children suggest that the children living in regions with higher population density were significantly more likely to have myopia, after adjustment for factors such as ethnicity, parental myopia, and time spent in near work and outdoor activity. Apartment-style housing was also significantly associated with myopia after accounting for regional differences. Although several studies have reported higher myopia prevalence in rural or remote settings compared with that in urban cities,
15 16 17 18 22 this is the first study to evaluate directly the associations of moderate differences in the urban environment with myopia in samples of children also taught in the same schooling system.
Previous Studies
In comparing the prevalence of childhood myopia from extremes in environmental settings (rural versus urban), distinguishing effects of urbanization from the impact of other factors (e.g., education, schooling, and outdoor activity) can be difficult, so that it is not always entirely clear whether differences in myopia prevalence can be attributed to an urban environment alone. Among 12-year-old school children in China, one prevalence estimate of myopia in an urban sample was 49.7%,
15 but it ranged from 12.0% to 23.0% in semirural samples.
17 In a study of younger children in China attending school in the city or in the countryside, Saw et al.
18 reported that the proportions with myopia were 19.3% and 6.6%, respectively. In these younger children, there were also substantial differences in level of near-work activities, in the proportions with parental myopia and in the level of parental education, so that differences in myopia prevalence could not be attributed to urbanization or geographic region alone. In another study of children in Malaysia and Singapore,
23 ethnicity-specific comparisons consistently showed higher rates of myopia in Singaporean children. Although the authors proposed that environmental differences such as early schooling, population density, and apartment housing may have attributed to the differences in myopia prevalence between the two countries, specific analyses of such factors were not reported. Lithander,
22 in a survey of myopia prevalence in a sample of children from the Sultanate of Oman, found that children in rural remote areas were significantly less myopic than the rest of the sample. This finding was attributed to the high illiteracy rate and the low level of reading or writing undertaken by inhabitants in the rural areas. In a sample of rural school children in China, He et al.
16 found that school locality (urban school versus rural school) was significantly associated with myopia and was thought to be the result of various levels of near work and outdoor activity. These factors, however, were not reported.
In this study, we had a unique opportunity to assess myopia risk factors in an age-specific sample of multiethnic children living in a range of residential settings across a large metropolitan area. The significant relationships observed within an urban setting emphasize the relevance of even subtle environmental differences in childhood myopia.
The greater Sydney area in which we conducted our survey, is predominantly suburban, with a broad range of residential settings, including relatively spacious outer suburban regions plus more densely populated inner city areas. Direct comparisons of myopia’s prevalence with previous studies are limited by factors such as ethnicity, levels of near work and study, and the outdoor environment. Nonetheless, the present study shows that the myopia prevalence in the inner city area (17.8%) was substantially lower than that reported in same-age urban childhood samples from China (49.7%)
15 and Malaysia (24.8%).
24 In an urban childhood sample from India, the prevalence of myopia in 12-year-old children was 9.7%.
25 For children of East Asian ethnicity in our sample, the prevalence of myopia in the inner city region (55.1%) was actually quite comparable to the myopia prevalence reported from East Asian counties.
Near Work and Outdoor Activity
By contrast with previous studies that reported urban-rural differences in schooling and near work,
18 we found no statistically significant differences in the time spent engaged in near work among children living in the five regions in our study. Similarly, there were no statistically significant differences in the time spent outdoors, which suggests that, overall, outdoor access is not restricted by locality in Sydney, with public open spaces widely available. Children may also travel to access beach areas or national parks.
A slight but statistically significant protective effect from outdoor activity has been reported in studies of childhood myopia.
13 26 In the present study, the differences in outdoor activity were relatively small (there was a mean difference of 1.4 h/wk between outer suburban and inner city areas). The final multivariate model for myopia also showed statistically significant, though somewhat weaker, effects from outdoor activity (OR, 0.96; 95% CI, 0.95–0.98, for each additional hour per week) by comparison with region (inner city versus outer suburban; OR, 2.5; 95% CI, 1.5–4.1). This suggests that outdoor activity alone is unlikely to be driving the association of myopia with inner city/urban residence.
Parental Myopia and Ethnicity
Other possible contributing factors to the higher childhood myopia prevalence in urban compared to suburban samples could be the greater proportion of parents with myopia, differences in socioeconomic status plus differences in ethnic composition. The higher frequency of parental myopia in the inner urban regions and an uneven distribution of higher income earners in parts of Sydney (for example in inner city areas) could have increased the prevalence of childhood myopia. A relatively greater proportion of East Asian children than European Caucasian children living in the inner city/urban regions may also have contributed to higher myopia prevalence.
In stratified analyses for childhood myopia, there were significant regional associations in both European Caucasian and East Asian children. Multivariate analyses that adjusted for socioeconomic status (parental education), parental myopia, and ethnicity showed that region persisted as a significant variable for childhood myopia, confirming that these findings were independent of both parental myopia and ethnicity.
Housing Density and Housing Type
We defined region using groupings of population density. This definition also matched housing type, so that in more urban areas, apartment or terrace housing was more prevalent (χ2 < 0.0001). Housing type was found to be related to myopia, independent of the effects of region, ethnicity, near work, or outdoor activity. We could not, however, find any associations between housing density and myopia in this sample.
Implication of Findings
The public health impact from the urban environment on eye health and general health in children may become increasingly important in the longer term, particularly as populations continue to rise and residential areas become increasingly built up. With myopia on the rise in urbanized parts of the world such as Hong Kong, Taiwan, and Singapore, it is unclear whether similar trends of increasing myopia will also be evident in countries undergoing rapid urbanization.
Currently, the intent of treatment for myopia is vision correction, with options including the use of spectacles, contact lens, and refractive surgery. Other treatment options for myopia currently under evaluation in experimental studies include orthokeratology and the use of pharmacologic therapies such as atropine and pirenzepine. In addition to pursuing treatment in myopia, there is a growing need to identify potentially modifiable factors for implementation in primary interventional studies. The current findings could have substantial implications for intervention at the level of the local community in areas such as town planning and residential housing, particularly in areas of low myopia prevalence.
Strengths and Limitations
Several factors contribute to the strengths of this study on myopia and the urban environment. Data were collected from a large representative sample of children living in a predominantly urban setting, studying in the same school year, and educated within the same school system. By sampling within these environmental conditions, we were able to minimize possible confounding effects from factors such as education and study. Adjustment for other factors such as outdoor activity and parental myopia was also included in our analyses as a rigorous test of the significance of regional differences. Standardized examination and test procedures used at all schools and the objective measures of refraction (cycloplegia with cyclopentolate) have also strengthened the findings of this study.
In the existing literature on refraction, myopia has been defined using various measurement methods (retinoscopy versus autorefraction), using a variety of cutoffs, defined either by spherical equivalent refraction or based on horizontal and vertical meridians. Although myopia can be considered a continuum along the spectrum of refractive error rather than a distinct condition, the Refractive Error Study in Children (RESC)
27 surveys conducted in China, India, Malaysia, South Africa, and Chile have all used a protocol of cycloplegic autorefraction, with spherical equivalent refraction cutoffs defining myopia at −0.50 and/or −1.00 D.
In our study, myopia was defined in a similar manner, by using spherical equivalent refraction, and tested by autorefraction in subjects under cycloplegia. This definition, which enables direct comparison with previous studies, could also be more meaningful than a higher cutoff for the purpose of evaluating the early processes affecting eye geometry in myopia, since myopic refractive error under cycloplegia, even at −0.50 D, could suggest that the growth of the eye has already progressed to a stage beyond emmetropia. As Mutti and Zadnik
28 have shown, refractive error at a young school age in children is a strong predictor of later myopia. Using −0.50 D as a cutoff for myopia in the present study could also provide some indication of those children more likely to be identified as myopic in later life, regardless of the defining cutoff.
One of the limitations of this study was the inability to collect refraction data on nonparticipants, because children with existing refractive errors may have been more likely to participate, which could have resulted in greater statistical significance for some of the associations reported. The relatively high participation rate and the similarity of demography between participants and nonparticipants, however, provide some assurance that this was a fairly representative sample and that participation bias was only minimal. Although housing density was indirectly assessed in the present study by proximity to other houses, this may have been a fairly unreliable indicator of urbanization, particularly for children residing in apartment housing. Assessment of other factors such as diet, and green space availability and its utilization, could be examined in future studies.
In conclusion, findings from this sample of 12-year-old Australian school children confirm previous reports of higher myopia prevalence for children living in city areas with higher population density. This association was found to be independent of conventional myopia risk factors such as ethnicity, parental myopia, the levels of near work, and outdoor activity. These data provide evidence supporting the hypothesis that certain features of the urban environment may influence the development of myopia in childhood. Further work to characterize which specific aspects of urban living contribute to myopia would be useful.
The Sydney Myopia Study was supported by the Australian National Health and Medical Research Council (Grant 253732), the Westmead Millennium Institute, University of Sydney, and the Vision Co-operative Research Centre of the University of New South Wales. IGM was supported by the ARC Centre of Excellence in Vision Science (COE561903).
Submitted for publication November 12, 2007; revised April 15, 2008; accepted July 9, 2008.
Disclosure:
J.M. Ip, None;
K.A. Rose, None;
I.G. Morgan, None;
G. Burlutsky, None;
P. Mitchell, None
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “
advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Corresponding author: Paul Mitchell, Department of Ophthalmology, Centre for Vision Research, University of Sydney, Westmead Hospital, Hawkesbury Road, Westmead, NSW, Australia, 2145;
paul_mitchell@wmi.usyd.edu.au.
Table 1. Select Socioeconomic Factors in Sydney Regions (Statistical Range) Stratified by Population Density
Table 1. Select Socioeconomic Factors in Sydney Regions (Statistical Range) Stratified by Population Density
| Region 1* (Outer Suburban) | Region 2 | Region 3 | Region 4 | Region 5 (Inner City) |
| Population density (persons/km2) | 49.1–78.5 | 446.6–921.0 | 1,098.3–1,160.7 | 2,282.0–2,812.4 | 3,033.5–4,162.9 |
| Socioeconomic factors | | | | | |
| Individual annual taxable income (AUD) | 39,026–40,138 | 36,091–57,098 | 37,867–44,509 | 36,727–38,507 | 47,323–65,120 |
| Completed university degree (%) | 5.9–7.7 | 6.0–18.6 | 7.8–10.8 | 7.7–10.3 | 16.1–22.5 |
| Unemployed (%) | 4.6–8.0 | 2.6–6.2 | 3.5–7.2 | 3.9–7.6 | 3.7–4.8 |
| Ancestry (%), † | | | | | |
| European | 78.0–87.4 | 48.6–82.6 | 62.9–73.7 | 47.3–49.1 | 59.9–69.1 |
| East Asian | 1.3–3.1 | 2.5–17.3 | 8.5–9.1 | 12.5–13.9 | 6.1–13.4 |
| Middle Eastern or North African | 0.7–1.2 | 0.3–3.9 | 1.9–2.5 | 10.5–12.5 | 0.8–3.2 |
| Speaks language other than English at home (%) | 9.3–14.7 | 11.4–55.3 | 26.1–28.5 | 43.9–52.7 | 22.7–41.8 |
Table 2. Characteristics of Participating 12-Year-Old Children, Stratified by the Population Density of the Area of Residence
Table 2. Characteristics of Participating 12-Year-Old Children, Stratified by the Population Density of the Area of Residence
| Region 1 (Outer Suburban) n = 322 | Region 2 n = 703 | Region 3 n = 710 | Region 4 n = 304 | Region 5 (Inner City) n = 270 |
| Mean age (SD)* | 12.7 (0.4) | 12.6 (0.4) | 12.6 (0.4) | 12.8 (0.4) | 13.0 (0.4) |
| Sex (n, %) | | | | | |
| Girls, † | 152 (47.2) | 327 (46.5) | 326 (45.9) | 184 (60.5) | 152 (56.3) |
| Boys | 170 (52.8) | 376 (53.5) | 384 (54.1) | 120 (39.5) | 118 (43.7) |
| Ethnicity (n, %), † | | | | | |
| European Caucasian | 210 (65.2) | 446 (63.4) | 437 (61.6) | 91 (29.9) | 186 (68.9) |
| East Asian | 24 (7.5) | 121 (17.2) | 85 (12.0) | 70 (23.0) | 49 (18.2) |
| Highest parental education (n, %), † | | | | | |
| High school | 86 (31.4) | 104 (16.6) | 180 (28.6) | 70 (28.3) | 15 (6.0) |
| TAFE | 116 (42.3) | 249 (39.8) | 259 (41.1) | 79 (32.0) | 47 (18.8) |
| University | 72 (26.3) | 273 (43.6) | 191 (30.3) | 98 (39.7) | 188 (75.2) |
| Parental myopia (n, %), † | | | | | |
| No parents with myopia | 166 (67.5) | 320 (58.1) | 378 (68.5) | 126 (63.6) | 116 (52.3) |
| Two parents with myopia | 9 (3.7) | 49 (8.9) | 19 (3.4) | 18 (9.1) | 31 (14.0) |
| Mean time spent doing near work (SD) | 26.8 (14.6) | 26.9 (13.1) | 27.6 (15.1) | 29.4 (14.2) | 27.5 (11.0) |
| Mean time spent in outdoor activity (SD)* | 13.3 (8.6) | 11.6 (7.1) | 12.8 (7.6) | 11.2 (7.6) | 11.9 (7.0) |
Table 3. Refraction and Myopia Prevalence in 12-Year-Old Children Stratified by Region
Table 3. Refraction and Myopia Prevalence in 12-Year-Old Children Stratified by Region
| n | Mean Spherical Equivalent Refraction (SEM)* | Proportion with Myopia (95% CI) |
| Whole sample | | | |
| Region 1 (outer suburban) | 322 | 0.28 (0.10) | 6.9 (4.1–9.6) |
| Region 2 | 703 | 0.24 (0.07) | 11.7 (9.3–14.0) |
| Region 3 | 710 | 0.18 (0.07) | 9.7 (7.5–11.9) |
| Region 4 | 304 | −0.02 (0.10) | 18.0 (13.6–22.4) |
| Region 5 (inner city) | 270 | 0.12 (0.10) | 17.8 (13.2–22.4) |
| European Caucasian | | | |
| Region 1 (outer suburban) | 209 | 0.93 (0.08) | 1.9 (0.04–3.8) |
| Region 2 | 446 | 0.84 (0.05) | 3.8 (2.0–5.6) |
| Region 3 | 437 | 0.79 (0.05) | 5.3 (3.2–7.4) |
| Region 4 | 91 | 0.72 (0.12) | 3.3 (0.4–7.1) |
| Region 5 (inner city) | 186 | 0.75 (0.08) | 8.1 (4.1–12.0) |
| East Asian | | | |
| Region 1 (outer suburban) | 24 | −0.43 (0.47) | 29.2 (9.6–48.8) |
| Region 2 | 121 | −0.68 (0.20) | 41.3 (32.4–50.2) |
| Region 3 | 85 | −0.66 (0.24) | 26.5 (16.8–36.2) |
| Region 4 | 70 | −1.03 (0.26) | 44.9 (32.9–57.0) |
| Region 5 (inner city) | 49 | −0.91 (0.31) | 55.1 (40.7–69.5) |
Table 4. Proportion with Myopia by Housing Type
Table 4. Proportion with Myopia by Housing Type
| Proportion with Myopia n (%) |
| Separate house (n = 1714) | 189 (11.3) |
| Terrace house (n = 126) | 126 (21.4) |
| Apartment (n = 137) | 36 (26.3) |
| χ2 | <0.0001 |
Table 5. Multivariate Adjusted Odds Ratio (CI) for Myopia in Children
Table 5. Multivariate Adjusted Odds Ratio (CI) for Myopia in Children
| Region 1 (Outer Suburban) | Region 2 | Region 3 | Region 4 | Region 5 (Inner City) | P |
| Whole sample (n = 2309)* | 1.0 | 1.2 (0.7–2.1) | 1.8 (1.2–2.9) | 2.0 (1.3–3.1) | 2.2 (1.3–3.8) | 0.0003 |
| European Caucasian (n = 1028), † | 1.0 | 3.1 (1.1–8.8) | 5.9 (2.0–17.0) | 4.1 (1.4–12.5) | 7.1 (2.5–19.8) | 0.0002 |
| East Asian (n = 245), † | 1.0 | 1.3 (0.6–3.3) | 1.0 (0.5–2.3) | 1.6 (0.7–3.9) | 2.0 (0.9–4.6) | 0.0002 |
MorganI, RoseK. How genetic is school myopia?.
Prog Retin Eye Res. 2005;24:1–38.
[CrossRef] [PubMed]LyhneN, SjolieAK, KyvikKO, GreenA. The importance of genes and environment for ocular refraction and its determiners: a population based study among 20–45 year old twins.
Br J Ophthalmol. 2001;85:1470–1476.
[CrossRef] [PubMed]DiraniM, ChamberlainM, ShekarSN, et al. Heritability of refractive error and ocular biometrics: the Genes in Myopia (GEM) Twin Study.
Invest Ophthalmol Vis Sci. 2006;47:4756–4761.
[CrossRef] [PubMed]Au EongKG, TayTh, LimMK. Education and myopia in 110,236 young Singaporean males.
Singapore Med J. 1993;34:489–492.
[PubMed]WuHM, SeetB, YapEP, SawSM, LimTH, CHiaKS. Does education explain ethnic differences in myopia prevalence? A population-based study of young adult males in Singapore.
Optom Vis Sci. 2001;78:234–239.
[PubMed]XuL, LiJ, CuiT, et al. Refractive error in urban and rural adult Chinese in Beijing.
Ophthalmology. 2005;112:1676–1683.
[CrossRef] [PubMed]WensorM, McCartyCA, TaylorHR. Prevalence and risk factors of myopia in Victoria, Australia.
Arch Ophthalmol. 1999;117:658–663.
[CrossRef] [PubMed]KempenJH, MitchellP, LeeKE, et al. The prevalence of refractive errors among adults in the United States, Western Europe, and Australia.
Arch Ophthalmol. 2004;122:495–505.
[CrossRef] [PubMed]AtteboK, IversRQ, MitchellP. Refractive errors in an older population: the Blue Mountains Eye Study.
Ophthalmology. 1999;106:1066–1072.
[CrossRef] [PubMed]WongTY, FosterPJ, HeeJ, et al. Prevalence and risk factors for refractive errors in adult Chinese in Singapore.
Invest Ophthalmol Vis Sci. 2000;41:2486–2494.
[PubMed]DayanYB, LevinA, MoradY, et al. The changing prevalence of myopia in young adults: a 13-year series of population-based prevalence surveys.
Invest Ophthalmol Vis Sci. 2005;46:2760–2765.
[CrossRef] [PubMed]SawSM, ShankarA, TanSB, et al. A cohort study of incident myopia in Singaporean children.
Invest Ophthalmol Vis Sci. 2006;47:1839–1844.
[CrossRef] [PubMed]MuttiDO, MitchellGL, MoeschbergerML, JonesLA, ZadnikK. Parental myopia, near work, school achievement, and children’s refractive error.
Invest Ophthalmol Vis Sci. 2002;43:3633–3640.
[PubMed]IpJM, HuynhSC, RobaeiD, et al. Ethnic differences in the impact of parental myopia: findings from a population-based study of 12-year-old Australian children.
Invest Ophthalmol Vis Sci. 2007;48(6)2520–2528.
[CrossRef] [PubMed]HeM, ZengJ, LiuY, XuJ, PokharelGP, EllweinLB. Refractive error and visual impairment in urban children in southern china.
Invest Ophthalmol Vis Sci. 2004;45:793–799.
[CrossRef] [PubMed]HeM, HuangW, ZhengY, HuangL, EllweinLB. Refractive error and visual impairment in school children in rural southern China.
Ophthalmology. 2007;114:374–382.
[CrossRef] [PubMed]ZhaoJ, PanX, SuiR, MunozSR, SperdutoRD, EllweinLB. Refractive Error Study in Children: results from Shunyi District, China.
Am J Ophthalmol. 2000;129:427–435.
[CrossRef] [PubMed]SawSM, HongRZ, ZhangMZ, et al. Near-work activity and myopia in rural and urban schoolchildren in China.
J Pediatr Ophthalmol Strabismus. 2001;38:149–155.
[PubMed]ParitsisN, SarafidouE, KoliopoulosJ, TrichopoulosD. Epidemiologic research on the role of studying and urban environment in the development of myopia during school-age years.
Ann Ophthalmol. 1983;15:1061–1065.
[PubMed]OjaimiE, RoseKA, SmithW, MorganIG, MartinFJ, MitchellP. Methods for a population-based study of myopia and other eye conditions in school children: the Sydney Myopia Study.
Ophthalmic Epidemiol. 2005;12:59–69.
[CrossRef] [PubMed]HarveyEM, MillerJM, DobsonV, TyszkoR, DavisAL. Measurement of refractive error in Native American preschoolers: validity and reproducibility of autorefraction.
Optom Vis Sci. 2000;77:140–149.
[PubMed]LithanderJ. Prevalence of amblyopia with anisometropia or strabismus among schoolchildren in the Sultanate of Oman.
Acta Ophthalmol Scand. 1998;76:658–662.
[CrossRef] [PubMed]SawSM, GohPP, ChengA, ShankarA, TanDT, EllweinL. Ethnicity-specific prevalence rates of refractive errors vary in Asian children in neighboring Malaysia and Singapore.
Br J Ophthalmol. 2006;90(10)1230–1235.
[CrossRef] [PubMed]GohPP, AbqariyahY, PokharelGP, EllweinLB. Refractive error and visual impairment in school-age children in Gombak district, Malaysia.
Ophthalmology. 2005;112:678–685.
[CrossRef] [PubMed]MurthyGVS, GuptaSK, EllweinLB, et al. Refractive error in children in an urban population in New Delhi.
Invest Ophthalmol Vis Sci. 2002;43:623–631.
[PubMed]RoseKA, MorganIG, IpJ, et al. Outdoor activity reduces the prevalence of myopia in children. Ophthalmology. .Published online February 20, 2008
NegrelAD, MaulE, PokharelGP, ZhaoJ, EllweinLB. Refractive Error Study in Children: sampling and measurement methods for a multi-country survey.
Am J Ophthalmol. 2000;129:421–426.
[CrossRef] [PubMed]MuttiDO, ZadnikK. The utility of three predictors of childhood myopia: a Bayesian analysis.
Vision Res. 1995;35:1345–1352.
[CrossRef] [PubMed]