Spatial Technology Assessment of Green Space Exposure and Myopia

Abstract

The prevalence of myopia has increased substantially within 1 generation in East Asia, especially China, which is thought to be attributable largely to changing environmental factors.1Morgan I.G. Ohno-Matsui K. Saw S.-M. Myopia. Lancet. 2012; 379: 1739-1748Abstract Full Text Full Text PDF PubMed Scopus (997) Google Scholar Green space is a potential environmental factor influencing population health2Chiesura A. The role of urban parks for the sustainable city.Landscape and Urban Planning. 2004; 68: 129-138Crossref Scopus (1347) Google Scholar and has been suggested to have changed over the past decades because of rapid city development and the mass migration of populations from rural to urban areas in China. Children from urban areas have 2.6 times (95% confidence interval [CI], 1.8–3.9) higher odds of having myopia than those from rural areas that are characterized by more green environments.3Maas J. Verheij R.A. Groenewegen P.P. et al.Green space, urbanity, and health: how strong is the relation?.J Epidemiol Community Health. 2006; 60: 587-592Crossref PubMed Scopus (1150) Google Scholar A potential association between green space and myopia may exist, but it has not been investigated thoroughly. With the development of spatial technology, green space can now be monitored, computed, and analyzed based on satellite data using remote sensing technology for health research.4Beck L.R. Lobitz B.M. Wood B.L. Remote sensing and human health: new sensors and new opportunities.Emerg Infect Dis. 2000; 6: 217Crossref PubMed Scopus (260) Google Scholar,5Dadvand P. Sunyer J. Alvarez-Pedrerol M. et al.Green spaces and spectacles use in schoolchildren in Barcelona.Environ Res. 2017; 152: 256-262Crossref PubMed Scopus (32) Google Scholar We evaluated the association between green space in schools and the prevalence and incidence of myopia. A large-scale cohort design used satellite imaging to quantify green space and clinical examinations and questionnaires to document myopia. We conducted this study in Shenzhen, the largest immigrant city in China, which is undergoing rapid urbanization. This study was based on data collected by the Environmental Health and Myopia Prevention and Control Project, which focused on evaluating the impact of environmental factors on the development of myopia (National Clinical Trial identifier, 04161326). A total of 113 schools in Shenzhen participated in the study. We included schoolchildren from grade 1 (6 years of age) to grade 4 (9 years of age), representing a period of rapid refractive error changes. This study was approved by the ethics committees of the Zhongshan Ophthalmic Center. Written informed consent obtained from all guardians was in accordance with the tenets of the Declaration of Helsinki. Eye refraction data were collected annually from 2016 through 2019. Students were excluded from this study if they reported a history of wearing rigid contact lenses or if they had a history of medical treatment or a disease that might affect their vision or vision development, such as amblyopia or diabetes. Myopia was defined as spherical equivalent refraction (SER, sphere plus one-half cylinder) of at least –0.5 diopters (D). The increase in myopia prevalence was defined as the cumulative increase in myopia prevalence in each school from baseline (2016–2017) to the last data collection period (2018–2019), which is a continuous variable. Data from the right eyes were used for the analyses. The normalized difference vegetation index is used as the measure of green space exposure (Fig S1, available at www.aaojournal.org). The normalized difference vegetation index was defined as the ratio of the difference between near-infrared reflectance and red visible reflectance to their sum, which ranges between –1 and 1, with higher numbers indicating more greenness. A total of 159 895 students (grades 1 through 4) were eligible for the study (Fig S2, available at www.aaojournal.org). Students who had unreliable data or missing data were excluded further from the analyses. Of the 142 865 students (89.3%), boys constituted 56.4% of the total sample, with the proportion of children at each age ranging from 26.6% for 6-year-old children to 22.4% for 9-year-old children. A total of 17 646 students (12.4%) had myopia with an average SER of –1.636 diopters (D) at baseline. The total of 113 schools had an average myopia prevalence of 11.9% (standard deviation, 3.6%). School socioeconomic ranking information (key schools, moderate schools, and normal schools) was used to reflect the school socioeconomic status. The study included 24 key schools (21.2%), 52 moderate schools (46.0%), and 37 normal schools (32.7%) at baseline. The mean normalized difference vegetation index value for all schools was 0.202 (standard deviation, 0.057). At baseline, no association with green space exposure was observed for any of the other covariates that were assessed at school level (male percentage: r = 0.007, P = 0.427; average age: r = 0.124, P = 0.189; average SER: r = –0.180, P = 0.057; myopia prevalence: r = 0.117, P = 0.217; school socioeconomic status: r = –0.177, P = 0.061), for myopia (odds ratio, 1.063; 95% CI, 0.944–1.196; P = 0.313), and for SER of students at the individual level (β coefficient, –0.025; 95% CI, –0.083 to 0.034; P = 0.412). Of the students included in the baseline analyses, 118 369 (82.9%) participated in the final data collection. We analyzed the association between green space exposure and school- and individual-level outcomes. At the school level, a 0.1 increase in green space exposure was associated with a 3.6% lower (95% CI, 1.8%–5.5%; P = 0.0001; Fig 1; Table S1, available at www.aaojournal.org) increase in the school myopia prevalence over 2 years, adjusting for 4 covariates at baseline: average age, male percentage (percent of boys in a school), myopia prevalence at baseline, and school socioeconomic status. According to individual-specific refraction at baseline, the students were divided into nonmyopic and myopic groups. In the nonmyopic group, the model showed that a 0.1 increase in green space exposure was associated with a 19.8% (adjusted odds ratio, 0.802; 95% CI, 0.724–0.889; P < 0.0001; Table S1) reduction in the myopia risk. In the myopic group, the model showed that green space exposure was not associated with myopia progression (P = 0.058; Table S1). The student-level analyses were adjusted for 4 covariates: age, SER at baseline, sex, and school socioeconomic status. To estimate further potential effects caused by other possible confounding factors, a random sample of 8400 students was selected and a total of 7996 students’ parents (95.2%) completed the questionnaires. Based on the subset, student-level analyses were adjusted for 5 additional covariates: paternal myopia, maternal myopia, average screen time per day, average reading time per day, and average outdoor activity time per day after school. Adding the covariates addressed in the questionnaire in terms of myopia incidence, a 0.1 increase in green space exposure was associated with a 27.9% (adjusted odds ratio, 0.821; 95% CI, 0.737–0.805; P < 0.0001; Table S1) reduction in the myopia risk in nonmyopic students. Regarding myopia progression, green space exposure was not associated significantly with myopia progression over 2 years in myopic students (P = 0.133; Table S1). Our study showed a negative association between green space exposure and myopia, providing a potential basis for the development of prevention strategies targeting the onset of myopia. No interactions or correlations were found between green space exposure and other known risk factors. Therefore, assuming a causal association between more green space in school and reduced myopia prevalence and that all the students respond in the same way to risk factors for myopia, if the green spaces of all schools were increased to 0.3 or more, the overall myopia prevalence would be reduced by approximately 4 percentage points per year. This would be sufficient to achieve the ideal myopia prevalence reduction suggested by the Ministry of Education in China (Fig 1).6Ministry of Education of the People's Republic of ChinaImplementation Plan for Comprehensive Prevention and Control of Myopia in Children and Adolescents.http://www.moe.gov.cn/srcsite/A17/moe_943/s3285/201808/t20180830_346672.htmlDate: 2018Date accessed: August 30, 2018Google Scholar Additionally, our study demonstrated that satellite imaging can mitigate the challenge of quantifying the environmental exposures at a macroscopic level by elucidating medical patterns using remotely sensed data.4Beck L.R. Lobitz B.M. Wood B.L. Remote sensing and human health: new sensors and new opportunities.Emerg Infect Dis. 2000; 6: 217Crossref PubMed Scopus (260) Google Scholar,7Obermeyer Z. Emanuel E.J. Predicting the future—big data, machine learning, and clinical medicine.N Engl J Med. 2016; 375: 1216Crossref PubMed Scopus (1223) Google Scholar Such methods can be extended further to study the relationship between the global environment and human vision health. Our study has several limitations. First, we did not consider the students’ green space exposure outside their schools; therefore, we cannot exclude the impact of exposure to green space during the students’ after-school time on myopia development. Second, the students in our study underwent refractive error measurement by noncycloplegic autorefraction instead of cycloplegic refractometry, which can lead to the overestimation of myopia. Third, despite the design of our study, our risk estimates indicate only correlations. Randomized controlled trials are still needed to investigate the effect of increasing green space as a community intervention for myopia. In summary, our study proved a strong association between green space and myopia in schoolchildren based on satellite imaging techniques. The implications are relevant to policymakers because greater access to green space potentially provides eye health benefits, which may help to contain the exploding epidemic of myopia in an increasing number of countries worldwide. The authors thank all the students, their families, the schools involved, the Education Bureau, and the Rescue Eyes Service Centre (http://www.star-eyes.org) for supporting this study and Bryan Spencer for editing the manuscript. Download .pdf (1.01 MB) Help with pdf files Fig S1 Download .pdf (.51 MB) Help with pdf files Fig S2 Download .pdf (.08 MB) Help with pdf files Table S1

Topics & Concepts