Purchase this article with an account.
Zhongqiang Zhou, Xiaochen Ma, Hongmei Yi, Xiaopeng Pang, Yaojiang Shi, Mirjam Meltzer, Mingguang He, Scott Rozelle, Ian George Morgan, Nathan G Congdon; Factors underlying large differences in myopia prevalence among primary school children in adjoining provinces of western China. Invest. Ophthalmol. Vis. Sci. 2014;55(13):3627.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To study reasons for differences in myopia prevalence between middle-income Shaanxi (ranked #14 of 31 Chinese provinces on per capita income) and poor Gansu (ranked #30). These neighboring Chinese provinces both have populations > 90% ethnically Han
Primary school children with uncorrected visual acuity (VA) <= 6/12 in either eye underwent cycloplegic automated refraction. Myopia was defined as spherical equivalent refractive error (SE) <= -0.5D in both eyes and uncorrected VA <= 6/12 in at least one eye. Socioeconomic, demographic and behavioral factors were assessed by questionnaire. School performance was assessed using a 90-minute mathematics test . Population density was calculated at the township level (each of 253 participating schools was located in a different township).
Myopia prevalence among 9667 children in Shaanxi (mean age 10.4 (1.0 ) years, 53.6% male) was 23.1%, nearly twice that among 10,308 children (mean age 10.7 ( 1.2 ) years, 50.6% boys) in Gansu at 13.4% (P < 0.0001). Spectacle ownership was low among children with refractive error in both Shaanxi (464/2362 = 19.6%) and Gansu (250/1472 = 17.0%). In multiple regression modeling, predictors of myopia included older age (Relative risk [RR] = 1.08, P <0.001), female gender (RR = 1.25, P < 0.001), family wealth (RR = 1.13 for middle versus lowest tercile P = 0.04; RR =1.24 for highest versus lowest tercile, P < 0.001), spectacle wear by parents (RR = 1.62, P < 0.001), math scores at the beginning of this study (RR = 1.21, P < 0.001) and residence in Shaanxi (RR = 1.18, P < 0.001), but not near work time, middle distance work time, outdoor activity, parents’ highest education, or parents having out-migrated for work. Lower population density in Shaanxi (RR = 0.79, P = 0.03) and higher population density in Gansu (RR = 1.27, P = 0.04) were associated with myopia in separate province-specific models.
The predominant non-demographic predictors of myopia in this study were socioeconomic (family wealth), academic (math scores) and familial (parental spectacle wear), but these do not fully explain the very low prevalence of myopia in Gansu versus Shaanxi. The impact of population density on myopia is complex in this setting. It seems likely that there may be important determinants of myopia prevalence in China which are still not well understood.
This PDF is available to Subscribers Only