Urban Living Environment and Myopia in Children | Public Health | JAMA Network Open

Key Points

Myopia has emerged as a significant global health concern over the past 3 decades, particularly in Asia, where it affects a substantial proportion of adolescents and young adults, reaching 80%-90% in urban areas of East and Southeast Asia.1^–4 Notably, the prevalence of myopia is increasing, and the onset of myopia is occurring at younger ages. It is projected that the number of individuals with myopia was approximately 1.4 billion in 2000 and is estimated to reach 4.8 billion by 2050.⁵ While this is typically a benign condition, severe myopia is associated with elevated risk of potentially blinding conditions, such as cataracts, glaucoma, and serious retinal diseases.⁶ The increasing prevalence of myopia is largely attributed to environmental factors primarily related to changes in living environments and lifestyles.^2–8 Researchers have provided evidence that residing in urban areas is associated with a higher likelihood of developing myopia.^9–12

With the projected increase in urbanization, it is estimated that by 2050, around two-thirds of the global population will be residing in cities.13 This rapid urban growth brings about significant environmental changes. Urban areas are characterized by higher population densities, increased housing and building infrastructure, reduced green spaces,¹⁴ and more stressful social conditions.¹⁵ It is important to note that urban residents often have better access to health care compared with those in rural areas.¹⁶ While previous studies have explored the correlation between specific environmental factors associated with urban living, such as green areas, and myopia, there is a lack of understanding regarding the broader effects of the overall urban living environment on myopia.^17–19 It is essential to consider the collective influence of various environmental factors within urbanization.

Satellite remote sensing techniques offer a valuable means of obtaining multidimensional information related to urbanized environments. These maps provide detailed data on various aspects of urbanization, including buildings, roads, nighttime lights, greenery, population density, and numerous other indicators.20 Given their ability to capture multidimensional urban living environment exposure indicators, satellite remote sensing maps present an ideal tool for extracting objective metrics for individual-level studies on urbanization.^21,22 Building on this potential, the present study focused on exploring the association between satellite remote sensing–derived multidimensional urbanization indicators and the incidence and progression of myopia. By delving into this, the aim was to gain a clearer understanding of the association of urban living environments with myopia.

This cohort study was conducted as part of the Investigation on Visual Habits and Eye Health Among Primary and Secondary School Students in Tianjin Project. This study underwent review and received approval from the Tianjin Eye Hospital Ethics Committee. Additionally, it was granted ethical clearance by the same committee, obviating the necessity to procure informed consent, as it was a nonmedical intervention study involving deidentified data. The study was carried out in accordance with the principles outlined in the Declaration of Helsinki.23 It followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Tianjin, 1 of the 4 municipalities directly under the central government in China, spans an area of 11 966.45 km² and comprises 16 districts. These districts can be categorized into central and peripheral areas, exhibiting notable disparities in living environments during the urbanization and development process (eFigure 1 in Supplement 1). Students in grades 1 to 6 from 493 schools across all districts of Tianjin were randomly selected to undergo comprehensive eye examinations in 2021. The examinations conducted for all participants involved visual acuity measurements using the Early Treatment of Diabetic Retinopathy Study (ETDRS) visual acuity chart and noncycloplegic autorefraction (ARK-700A; Nidek Corp) for both eyes. Moreover, data on personal factors, such as age, sex, grade, school socioeconomic status (with schools ranked in the top 100 in Tianjin classified as key schools and others considered normal schools), and refractive correction status (correction or uncorrection) were collected. To ensure regular vision assessments, efforts were made to conduct annual examinations for students spanning from March 1, 2021, to March 31, 2023. Initially selected students in grades 1 to 4 in 2021 who successfully completed 3 vision screenings over a 2-year period were included in longitudinal analysis. Furthermore, a questionnaire was administered to a random sample of students in grades 1 to 6 to provide additional information for the study (eTable 1 in Supplement 1). Sample size and missing value information are presented in eTable 2 in Supplement 1.

An iterative exploratory factor analysis was applied to reconstruct the urban score based on the satellite data. Initially, 12 environmental variables were derived from the data sets (eTable 3 in Supplement 1). We screened candidate environmental variables using population density as a reference, following the most recent public measure for urbanization.^24,25 Subsequently, 4 variables were reserved and formed an urban score²⁶ (eFigure 2 in Supplement 1). Detailed steps are shown in eTable 4 in Supplement 1.

Myopia was defined as a spherical equivalent refraction (SER)≤−0.50 diopters (D) in either eye. The association between the urban living environment and myopia involved steps shown in Figure 1. The prevalence of myopia was assessed among elementary school students in grades 1 to 6, considering different urban scores. For the incidence and progression analysis using longitudinal data, adjustments were performed for age, sex, grade, baseline SER, and school socioeconomic status to assess the association between myopia incidence and the urban environment in the group without myopia at baseline. Additionally, in the group without myopia, adjustments were made for age, sex, grade, baseline SER, refractive correction status, and school socioeconomic status to examine the association between myopia progression, defined by changes in SER, and the urban environment. To further account for the influence of personal factors, information gathered through questionnaires was used. Adjustments were made for age, sex, grade, school socioeconomic status, refractive correction status, parental history of myopia, mean screen time, mean outdoor time, and academic performance to explore the association between the urban living environment and both the prevalence and the severity of myopia.

A total of 177 894 elementary school students from grades 1 to 6 participated in this study; 48.3% were female, and 51.7% were male. The mean (SD) age of the students was 10.27 (1.75) years (eTable 5 in Supplement 1). The questionnaire was administered to a random sample of 3000 students in grades 1 to 6, resulting in a response rate of 87.3% (n = 2620). We stratified the students into 2 distinct groups based on their urban scores, which ranged from 0 to 1, with a mean (SD) of 0.42 (0.23). These groups were delineated into high urbanization (urban score ≥0.5) and low urbanization (urban score <0.5) groups. The prevalence of myopia among students living in high urbanization areas was 67.2%, while in low urbanization areas, the prevalence of myopia was slightly lower at 64.2% (Figure 2A). In the longitudinal cohort (137 087 participants from grades 1-4; 47.7% female; 52.3% male; mean [SD] age, 8.97 [1.21] years), the prevalence of myopia was consistently higher in the high urbanization group compared with the low urbanization group in 2021, 2022, and 2023. Over the 2-year period from 2021 to 2023, the prevalence of myopia increased by 32.1% in the high urbanization group, which was slightly higher than the 31.1% increase observed in the low urbanization group (Figure 2B). In the population with myopia in 2021, the mean (SD) SER was −1.878 (1.361) D in the high urbanization group and −1.879 (1.377) D in the low urbanization group, with no significant difference between the 2 groups (P = .48). However, after 2 years, the severity of myopia in the low urbanization group became greater than that in the high urbanization group (mean [SD] SER: −2.418 [1.674] D vs −2.362 [1.630] D; P < .001) (Figure 2C).

To analyze the factors associated with the incidence of myopia, individuals who did not have myopia at baseline were selected. First, a comparison was made between those who developed new-onset myopia and those who did not develop myopia during the 2-year period. The statistical results indicated that age, sex, grade, SER at baseline, school socioeconomic status, and urban score were significantly different between the groups (eTable 6 in Supplement 1). To further analyze the association between urbanization level and myopia incidence, mixed-effects logistic regression models were used while adjusting for age, sex, grade, SER at baseline, and school socioeconomic status. The results revealed a significant association between elevated urban scores and an augmented risk of myopia. For each 1-unit increment in the urban score, there was an increased risk of myopia over a 1-year period (odds ratio [OR], 1.09; 95% CI, 1.01-1.15; P = .02) and a 2-year period (OR, 1.53; 95% CI, 1.50-1.57; P < .001). Furthermore, an examination of specific indicators pertaining to urban living environments consistently bolstered these associations. Over a 2-year time frame, a 1-unit increase in the population density index was associated with a higher risk of myopia (OR, 1.56; 95% CI, 1.52-1.59; P < .001). Similarly, a 1-unit increase in the night light index was associated with an increased risk of myopia (OR, 1.41; 95% CI, 1.38-1.45; P < .001). Conversely, a 1-unit increase in the enhanced vegetation index was associated with a reduced risk of myopia (OR, 0.46; 95% CI, 0.43-0.50; P < .001), and a 1-unit increase in the walking time to the nearest hospital index was associated with a diminished risk of myopia (OR, 0.79; 95% CI, 0.67-0.87; P < .001). The consistent trend observed in these results showed an association between urban living conditions and myopia incidence (Table 1).

Individuals who had myopia at baseline were selected to examine the association with myopia progression. Significant differences in myopia progression over the 2-year period were observed based on age, sex, grade, school socioeconomic status, and refractive correction status. Correlations between myopia progression and factors including age (r = −0.024; P < .001), baseline SER (r = 0.069; P < .001), and urban score (r = −0.068; P < .001) were observed. To account for potential confounding factors, generalized mixed linear models were adjusted for age, sex, grade, baseline SER, school socioeconomic status, and refractive correction status. The adjusted results demonstrated that each 1-unit increase in the urban score was associated with a significant decrease in myopia progression at 1 year (OR, 0.84; 95% CI, 0.82-0.86; P < .001) and 2 years (OR, 0.73; 95% CI, 0.70-0.75; P < .001). Furthermore, analyzing the specific factors related to urban living environments yielded consistent results. Living in areas with higher population density was associated with a reduction in myopia progression over 2 years per 1-unit increase in the population density index (OR, 0.85; 95% CI, 0.82-0.88; P < .001). Similarly, over a 2-year period of living in an urbanized living environment, a 1-unit elevation in the night light index was associated with a decrease in myopia progression (OR, 0.76; 95% CI, 0.74-0.79; P < .001), while an increase in the enhanced vegetation index by 1 unit was associated with an increase in myopia progression (OR, 1.41; 95% CI, 1.35-1.48; P < .001). Moreover, a 1-unit increase in the walking time to the nearest hospital index was associated with an increase in myopia progression (OR, 1.25; 95% CI, 1.21-1.30; P < .001). In summary, the findings indicated a negative association between the urban living environment and myopia progression (Table 1).

In the cross-sectional data analysis, additional factors including individual genetics and behaviors (eg, screen time, outdoor time, and academic performance) were incorporated. After adjusting for age, sex, grade, school socioeconomic status, parental history of myopia, mean screen time, mean outdoor time, and academic performance, the association between the urban living environment (as represented by urban score) and myopia prevalence was assessed. The analysis revealed that each 1-unit increase in urban score was associated with a higher myopia prevalence (OR, 1.62; 95% CI, 1.08-2.42; P = .02). Thus, urban living remained positively associated with myopia prevalence. Furthermore, when adjusting for age, sex, grade, school socioeconomic status, refractive correction status, parental history of myopia, mean screen time, mean outdoor time, and academic performance, it was found that individuals with myopia residing in areas with higher levels of urbanization had a lower severity of myopia as indicated by a larger SER (OR, 1.46; 95% CI, 1.07-1.99; P = .02). In other words, the urban living environment was negatively associated with myopia severity. A detailed overview of the results obtained from the analysis of the 4 specific factors related to the urban living environment is given in (Table 2).

Previous studies have indirectly suggested that urban living contributes to an increased risk of myopia. In addition to studies that generalized to simple urban-rural divisions,9^,27–29 Zhang et al¹⁹ found that average population density in administrative divisions was associated with myopia risk among Chinese children independent of factors such as academic activities, outdoor time, family educational level, and economic development. Similarly, studies conducted in Africa projected an increase in childhood myopia prevalence in urban settings.³⁰ Additionally, the presence of green spaces within school campuses is associated with a lower prevalence of myopia at the school level and potentially with reduced risk of myopia development in individuals.^17,18 In our study, we systematically examined various environmental variables with population density as a reference point. We included 4 specific indicators: population density, night light index, enhanced vegetation index, and walking time to nearest hospital. When analyzing these indicators individually, we observed a higher level of agreement with the results presented by the urban score. Our inclination was to place greater confidence in the notion that areas with higher population density, more intense nighttime lights, fewer green spaces, and improved medical infrastructure would be associated with a higher risk of myopia. Thus, it can be inferred that the overall urbanized living environment was associated with higher risk of myopia. However, additional research is required to ascertain whether these indicators are independently associated with myopia or whether other factors in the urbanized living environment contribute to the condition. Furthermore, our findings indicated a relatively gradual progression of myopia in highly urbanized living environments. This association may, in part, be attributed to the greater prevalence of effective myopia correction methods in urban centers. Conversely, students residing in less urbanized areas, potentially due to limited medical access, inadequate education, lack of awareness about health care, and socioeconomic challenges, may face barriers in accessing timely and appropriate myopia control measures.^31,32 While we accounted for the impact of refractive correction status in our analysis, we only used a basic categorization of whether or not refractive correction was performed without considering the presence of undercorrection. It is important to note that undercorrection of myopia is a well-established factor that contributes to the advancement of myopia.^33,34 Therefore, our results underscore the need for vigilant management of populations with myopia in low-urbanized areas despite the inherent risks associated with highly urbanized living environments.

There are certain limitations to our study that should be acknowledged. First, the optometry results used for diagnosing myopia were conducted under nonciliary muscle paralysis, which may have led to an overestimation of the prevalence of myopia among students. Second, our study relied on myopia data collected from 2021 to 2023, during which the COVID-19 pandemic impacted regular urban living environments, potentially introducing variations in the analysis of the role of urbanization compared with previous years.35^,36 Choi et al³⁷ found a correlation between the size of one’s residence and both axial length and refractive error. Nonetheless, this association could be further accentuated during specific periods of prolonged indoor activities. Third, our study merely suggests an association between urbanized living environments and the onset and progression of myopia. However, further investigation is necessary to identify the specific factors that contribute to this association.

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2023 Li X et al. JAMA Network Open.

Corresponding Author: Wei Zhang, MD, PhD, Tianjin Eye Hospital, No. 4, Gansu Rd, Heping District, Tianjin 300020, China (zhangwei3066@126.com).

Author Contributions: Drs X. Li and Zhang had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs X. Li, L. Li, and Qin were co–first authors.

Concept and design: X. Li, Qin, Z. Li, Zhang.

Acquisition, analysis, or interpretation of data: X. Li, L. Li, Qin, Cao, Mu, Liu.

Drafting of the manuscript: X. Li, L. Li, Cao, Mu, Liu, Z. Li.

Critical review of the manuscript for important intellectual content: X. Li, Qin, Zhang.

Statistical analysis: X. Li, Qin, Cao, Liu, Z. Li.

Obtained funding: Zhang.

Administrative, technical, or material support: L. Li, Mu.

Supervision: Qin, Zhang.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by grant 22JCZDJC00160 from the Tianjin Municipal Science and Technology Commission, Tianjin Research Program of Application Foundation (W.Z.) and NKYKD202202 from the Nankai University Eye Institute (W.Z.).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 2.