High Myopia and Its Associated Factors in JPHC-NEXT Eye Study: A Cross-Sectional Observational Study

The increasing prevalence of high myopia has been noted. We investigated the epidemiological characteristics and the related factors of high myopia in a Japanese adult population. Japan Public Health Center-Based Prospective Study for the Next Generation (JPHC-NEXT) Eye Study was performed in Chikusei-city, a rural area in mid-east Japan, between 2013 and 2015. A cross-sectional observational analysis was conducted to investigate prevalence and related factors of high myopia. A total of 6101 participants aged ≥40 years without a history of ocular surgeries was included. High myopia was defined as a spherical equivalent refraction of ≤−6.00 diopters according to the American Academy of Ophthalmology. Potential high myopia-related factors included intraocular pressure (IOP), corneal structure, corneal endothelial cell density, age, height, body mass index, heart rate, blood pressure, biochemical profile, and current history of systemic and ocular disorders. The odds ratios of high myopia were estimated using the logistic regression models adjusted for the associated factors. The prevalence of high myopia was 3.8% in males and 5.9% in females with a significant difference. Age was inversely associated, IOP was positively associated, and none of other factors were associated with high myopia in both sexes. In conclusion, only age and IOP were associated with high myopia in this community-based sample.

positive association between IOP and high myopia. Other factors were not associated with high myopia for either men or women.

Conclusions
In this community-based sample, age was inversely, and IOP was positively associated with high myopia in both the men and the women.

Background
Myopia is one of the most prevalent conditions of the eye. It causes visual impairment in both children and adults that is usually correctable by optical aids such as glasses and contact lenses. High myopia is generally determined as -6.00 diopters (D) or less in refraction and axial lengths of 26.5 mm and more by the American Academy of Ophthalmology. [1] High myopia is associated with progressive and excessive elongation of the eyeball, which results in various funduscopic changes in the posterior fundus, and increases the risk of pathologic myopia, which may cause irreversible vision loss such as glaucoma, retinal detachment, and macular degeneration. [2] High myopia is a major cause of blindness in many countries [3], and the prevalence of myopia and high myopia is expected to increase globally from 2000 to 2050.
[4] Thus, it is important to manage myopia progression and to prevent myopia-related ocular complications and vision loss the approximately 1 billion people with high myopia. [4] Over the past few decades, some studies have provided information on the prevalence and risk factors for myopia, including genetic predisposition and environmental factors such as extended near work, less exercise, and luck of outdoor activities. [ Human lifestyles are rapidly changing. The factors that may affect the progression of myopia are essential to understand to find countermeasures for myopia. Communitybased, population-based research is a reliable way to elucidate the associated factors. In order to elicit the high myopia associated factors, it was thought to be important to eliminate biases and prejudices as much as possible. We established this study to screen all the possible factors which were available in the collected data and personal information without manipulation. In addition, this study targeted adult population as subjects. Pathological myopia, which is usually preceded by high myopia and seen mostly in adults, is theoretically thought to be originated from myopia in children. To know possible associated factors for high myopia in adults is considered to be a crucial key to render solutions to diminish the number of future high or pathological myopia in children.
We conducted the JPHC-NEXT Eye Study, an ancillary study of the Japan Public Health Center-Based Prospective Study for the Next Generation (JPHC-NEXT), to examine the prevalence of refractive status of the participants and factors associated with high myopia in Chikusei, a rural city in mid-east Japan. To our knowledge, this is the first large community-based study to determine the factors associated with high myopia.

Methods
Association's Declaration of Helsinki. The study protocol was approved by the Institutional Review Boards of Keio University, Osaka University, the University of Tsukuba, and the National Cancer Center. Written informed consent was obtained from all of the participants.

Information on past medical history
Information on histories of hypertension, diabetes and dyslipidemia was collected through face-to-face interview at the baseline survey. Likewise, inquiry about smoking history and alcohol history as well as histories of ocular disease and its surgery were also performed.

Definitive Examination
The spherical equivalent refraction (SEq) was calculated from the refraction using the following formula: the full spherical power plus half of the cylindrical power. Initially, the mean values of the SEq of the right eyes and the left eyes were compared in the entire population of the current study. There was no significant difference in the median SEq between the right eyes and the left eyes (-0.13 vs -0.00 (D), p=0.10). Thus, we used the SEq of the right eyes to evaluate the refractive status in this study. A high myopia was defined as SEq of -6.00 D or less based on the American Academy of Ophthalmology criteria. [1] Statistical Analysis Statistical analyses were performed using IBM SPSS statistics version 23 (IBM, Armonk, NY, USA). Differences across the groups were tested using an analysis of (co)variance or the chi-square test. Student's t-test was used for independent variables showing normal distribution, and nonparametric tests such as Mann-Whitney's U test were applied to TG, GLU, SEq, IOP, corneal radius, central corneal thickness, and corneal endothelial cell number. To analyze the relationships of alcohol intake and smoking history with high myopia, Fisher's exact probability test and logistic analysis were performed. Age and multivariable adjusted OR and 95% CI of the high myopia according to the quartiles of each risk factor were calculated sex-specifically using logistic regression models The multivariate models included age, height, body mass index, systolic blood pressure (SBP), triglycerides, HDL cholesterol, LDL cholesterol, HbA1c, IOP, and central corneal thickness.
The linear trend of ORs across the quartiles were tested by using the continuous variables of the explanatory variables. A p value of less than 0.05 was considered statistically significant.

Participants
The prevalence of high myopia in the current study was 5.0%, 3.8% in men and 5.9% in women. High myopia was more prevalent among women than men (p<0.001). According to the age distribution, the prevalence of high myopia in the 40s, 50s, 60s, 70s, 80s, and 90s was 10.6%, 8.8%, 3.5%, 2.0%, 1.2%, and 0.0%, respectively (Fig 2). The median SEq of the right eyes in total was -0.13 D ranging from -23.13 D to +10.25 D, with 1st quartile -1.38 and 3rd quartile +0.88 D. In the men, the median SEq was -0.13 D ranging from -18.88 D to +10.25 D, with 1st quartile -1.25 and 3rd quartile +0.88 D. In the women, the median SEq was -0.13 D ranging from -23.13 D to +6.88 D, with 1st quartile -1.63 and 3rd quartile +0.88 D. The median SEq of the right eye was not significantly different between genders (p=0.52).

Associations between systemic factors and high myopia
We evaluated the associations between systemic factors and high myopia (Table1). The mean age for the high myopia group was significantly lower than non-high myopia group (55.7±10.5 vs 62.8±10.3 years, p<0.001). The peak of the high myopia prevalence by age was from 40 to 49 years (6.2% among the men and 13.0% among the women). The mean height for the high myopia group was significantly higher than others (158.8±8.0 vs 157.5±8.7 cm, p=0.009). A similar difference was also found in the age and heights of men and women, respectively (Table 1). No significant difference between the high myopia group and non-high myopia group was observed in weight, waist, and diastolic blood pressure. Among the laboratory data, T cholesterol, triglycerides, and LDL cholesterol were significantly different between the two groups in men. Glucose and HbA1c were significantly different between the two groups in the women. Regarding the relationship between alcohol intake and high myopia, ORs for high myopia and alcohol intake history were 1.166 (95%CI 0.371-3.662) in the men and 0.900 (95%CI 0.467-1.732) in the women. ORs for high myopia and tobacco use history were 0.598 (95%CI 0.340-1.051) in the men and 0.967 (95%CI 0.593-1.577) in the women (S1 Table).

Associated between ocular features and high myopia
Associations between ocular features and high myopia were also evaluated (Table1). The median IOP was significantly higher for the high myopia group than the others (14.30 vs 13.30 mmHg, p<0.001). A similar difference was found in the median IOP of the men or women, respectively. A significant difference was also observed in the median corneal radius between the high myopia group and the others (7.62 vs 7.64 mm, p=0.024).
However, no significant difference was found in the median corneal radius of the men and women, respectively. There was no significant difference in central corneal thickness and corneal endothelial cell number between the two groups.
IOP and age associated with high myopia A multivariable logistic regression analysis was performed to identify the factors associated with high myopia ( Table 2). The ORs adjusted for multivariable factors were significant for age and IOP in the men and the women. In men, the values were as follows: the ORs adjusted with other factors for high myopia in each age group were 40-58 years

Discussion
This community-based study showed the prevalence of high myopia and associated factors including physical, ocular and demographic factors among adult Japanese for the first time. The factors associated with myopia were widely investigated as we analyzed biochemistry tests, blood pressure, height, body weight, habit of smoking, alcohol intake, past medical history, and present diseases. In this current study, we found that high myopia is more prevalent in women, younger age, and it has higher IOP.
High myopia affects approximately 1-4% of adults aged ≥40 years, and its prevalence was higher in some studies of East Asian adults and adolescents. [19][20][21][22][23][24][25][26] Our findings showed that the prevalence of high myopia was 5.0%, which was no less than the generally affected rates, although Chikusei-city is in a rural area where its prevalence has usually been lower compared to urban areas.
[23] In our study, the high myopia rate in the older population was relatively low and the younger generation had a higher prevalence of high myopia, which may be reflect cohort effects. Although the reason for this cohort effects is unknown, the prevalence of high myopia could be expected to increase in future.
Some previous studies reported the relationship of height[27-29] and BMI[15, 30-32] with myopia; however, we did not find such association. A possible reason for this discrepancy is confounding; the previous reports did not adjust for age and any other confounding factors. In fact, BMI and height were shown to be associated with high myopia in unadjusted models, and adjustment for age resulted in the elimination of this significance in our study, presumably because age is a strong predisposing factor for BMI and height reflecting cohort effects.
As for the other laboratory factors, there was no significant difference. A few studies suggested that hyperglycemia and hyperlipidemia led to myopic shift, whereas other studies revealed that the refractive shift was more likely hyperopic with hypoglycemia.
[33-36] Further analysis is needed to elucidate the influence of metabolic shift.
In terms of alcohol intake and smoking history, both factors did not show any associations with high myopia according to our results. Also, liver functions, represented by GOT, GPT, and GGTP, did not show any relationship to high myopia. Previous reports also found that there were no significant trends observed between smoking and refractive errors. [37] The percentage of high myopia in men and women was 3.8% and 5.9%, respectively, indicating significant gender difference. It has been reported that female sex had a predisposition of high myopia.
[38] Likewise, female sex was proven to be high risk for The mean IOP linearly increased parallel to the myopic progression. It has been reported that IOP was associated with central corneal thickness, age, and blood pressure. [41,42] Even after adjustment for these factors, we still found that high IOP was significantly associated with high myopia while there are conflicting evidences regarding relationship of high myopia and IOP. [43][44][45] This study has several limitations. First, since this was a cross-sectional study, the causal relationships cannot be determined. Second, some known risk factors of myopia such as natural guardians, near work, outdoor activities, and academic backgrounds were unavailable in this study. Moreover, neither measuring of the axial length nor lens examinations has been performed. With respect to the lenses, factors of cataracts remain to be considered. We also should consider environmental differences across generations which might have affected the result of associations of age with high myopia.
As discussed, the factors associated with high myopia were undetermined. This study did not suggest that height, BMI, blood glucose, hypocholesteremia, liver dysfunction, kidney dysfunction, smoking, and alcohol intake were associated with high myopia, whereas women, young age and high IOP were found to be related to high myopia. Meanwhile, high IOP and young age were found to be risk factors for high myopia, which may indicate the path for future studies concerning myopia control.

Conclusions
In conclusion, this epidemiological study performed in a Japanese rural area revealed

Consent for publication
Consent for publication was obtained from each person partaking in this study in written consent forms.

Availability of data and materials
All data generated or analysed during this study are included in this published article and its supplementary information.

Competing Interests statement
The authors declare that they have no competing interests.