Abstract: Objective To report the prevalence of myopic maculopathy in people over 40 years old in rural area of China, and to explore the influencing factors of myopic maculopathy based on Synthetic Minority Over-Sampling Technique（SMOTE） algorithm. Design Cross-sectional study. Participants 1177 myopia patients in the Handan Eye Disease Study from 2012 to 2013, aged 40 years and above. Methods Questionnaire was done to obtain subjects’ basic demographic data and behavioral data such as smoking and drinking history; Subjects’ height, weight and blood pressure were measured through physical examination; blood samples of each subject was collected for laboratory examinations; Early Treatment Diabetic Retinopathy (EDTRS) eye chart was used to measure the subject’s presenting visual acuity and best-corrected visual acuity (BCVA); Kowa applanation tonometer was used to measure the introcular pressure (IOP); Topcon automatic refractor (model KR8800) was used to measure diopter; A-ultrasound (10 MHz) was used to measure the eye axis and other biological parameters; use Canon fundus camera (model CR-2) to take two fundus photos that centered on the optic disc and the macula; the diagnosis method of maculopathy: based on the myopic population (spherical equivalent refraction less than -0.5 D), any of the following symptoms in the fundus was judged to be maculopathy: diffuse chorioretinal atrophy, patchy chorioretinal atrophy, macular atrophy, lacquer cracks, and myopic choroidal neovascularization appeared in the posterior pole. Fuchs spots were classified as myopic choroidal neovascularization. Using the sixth census data of China, the prevalence of maculopathy was graded and gender-standardized. The SMOTE algorithm was used to balance data. Logistic regression was used to explore the influencing factors of myopic maculopathy. the ROC curve was used to estimate the predictive power of the influencing factors. Main Outcome Measures Crude prevalence and age-sex standardized prevalence of maculopathy. Results Of the 5055 people over the age of 40 with clear-readable fundus photographs, 1177 had myopia, of which 35 had myopic maculopathy. In the myopic population, the prevalence of myopic maculopathy was 2.97%, and the age-gender standardized prevalence was 2.65%. Univariate analysis showed: for age (t=10.910, P<0.001), education (χ2=21.511, P<0.001), marital status (χ2=62.488, P<0.001), BMI (t=-6.400, P<0.001), blood glucose (t=-4.199, P=0.014), diopter value (Z=-15.766, P<0.001), intraocular pressure (Z=-7.384, P<0.001) and BCVA (t=-24.384, P<0.001), there was statistically significant difference between the myopic maculopathy group and the control group. Logistic regression results showed that age and diopter were risk factors for myopic maculopathy. IOP was a protective factor for myopic maculopathy. The odds ratio (OR) values were 1.056 (95% confidence interval, CI: 1.025, 1.087), 0.667 (95% CI: 0.603, 0.738) and 0.841 (95% CI: 0.762, 0.928) respectively; ROC curve analysis showed that using age, diopter and IOP to predict myopic maculopathy, the area under the curve can reach 0.936 (95% CI: 0.918, 0.953). Conclusion The age-gender standardized myopic maculopathy prevalence rate in myopic population over 40 years old in rural area of China is 2.65%. Age, spherical refractive error, and IOP are influence factors of myopic maculopathy. Older age, lower IOP, and larger refractive error will increase disease risk. (Ophthalmol CHN, 2021, 30: 177-183)

Key words: myopic maculopathy, prevalence, influence factor, SMOTE algorithm