Prevalence and Characteristics of Myopia in Adult Rhesus Macaques in Southwest China

Abstract

Purpose: To investigate the prevalence of myopia in a large cohort of adult rhesus macaques at Yunnan Province in southwest China and describe the characteristics of myopic rhesus macaque eyes.

Methods: A total of 219 rhesus macaques 14.07 ± 2.72 years old (range, 8–21) were randomly recruited for this study. We performed fundus photography and measurements of cycloplegic refractive error (RE) and axial length (AL) on macaques.

Results: A total of 429 eyes of 219 macaques were examined. The median RE was −1.25 diopters (D), and the median AL was 18.69 mm. The prevalence of myopia was 62.47%, and one-third of the myopic eyes were highly myopic. The presence of fundus tessellations was higher in myopic eyes than non-myopic eyes (42.54% vs. 6.21%). The cutoff value for the presence of tessellations was −3.52 D for RE and 19.38 mm for AL. In myopic eyes, there were significant differences between grade 1 and grade 3 fundus tessellations on RE (−5.57 ± 2.97 D vs. −8.13 ± 3.51 D) and AL (19.66 ± 0.55 mm vs. 20.60 ± 1.06 mm). Beta-peripapillary atrophy (β-PPA) was found in 48.10% of myopic eyes and 6.83% of non-myopic eyes. The presence of β-PPA is associated with the presence of fundus tessellations, AL, and RE. The presence of β-PPA was higher in grade 3 than grade 1 fundus tessellations (94.4% vs. 76%).

Conclusions: More than half of adult rhesus macaques in southwest China are myopic, and one-third of the myopic ones are highly myopic. Similar to humans, tessellated fundi and β-PPA are the characteristic signs of myopic rhesus macaques. Adult rhesus macaques are optimal animal models for research on the pathogenesis of myopia.

Translational Relevance: This study not only provides a reference for the refractive state and AL in myopic rhesus macaques but also indicates that adult rhesus macaques with spontaneous myopia are optimal animal models for research on the pathogenesis of myopia.

Introduction

Myopia has become a global public health problem since 2000, when it had a prevalence of 23%.¹ By 2020, it had already affected 30% of the world population,² and it is estimated that about 50% of the world population will be myopic by 2050.¹ Furthermore, the outbreak of COVID-19 in 2019 led to house confinement and the online education of students. Intense near work, overuse of electronic equipment, and limited outdoor activities have already caused myopic shifts in students.³ Moreover, it has been demonstrated that children with an earlier onset of myopia and longer duration of myopia progression are prone to developing high myopia later in life.⁴ Not only is high myopia a severe type of myopia, but it also increases the risk of vision-threatened complications such as myopic macular degeneration, cataract, retinal detachment, and glaucoma.⁵ Myopic macular degeneration has become one of the major causes of blindness and irreversible visual impairment in China.⁶ Furthermore, development of high myopia results in increased medical costs and decreased quality of life for individuals, as well as lost productivity and a greater burden for society.² It is obvious that myopia is not simply a refractive state but also a problem affecting society as a whole. 

The pathogenesis of myopia is still obscure. Both genetic and environmental factors play roles in the development of myopia. Research has been done to explore the mechanisms of each of these factors. Animal models of myopia have been developed, including chicken,⁷ mice,⁸ guinea pig,⁹ and tree shrew.¹⁰ But, there are differences between humans and these species in both the size and structure of their eyes. The retina of non-human primates resembles that of humans, including similar macular structures, cell types, and cell circuits.^11,12 Moreover, the DNA sequence of the genus Macaca shares >90% similarity with humans, and protein sequences are highly conserved between these two genera.¹³ The rhesus macaque has been used as a model for human ocular disease.¹³ In the past, myopic animal models were mostly induced by form deprivation or lenses.¹⁴ Compared with spontaneous myopia, these induced myopia models cannot accurately reproduce the natural course of myopia. Therefore, non-human primates with spontaneous myopia would offer great benefits in furthering our understanding of the pathogenesis of myopia. 

Emmetropization of rhesus macaques shares similarities with that of humans.¹⁵ Newborns are hyperopia with a short axial length and steep cornea. Axial elongation and cornea flattening occur during the postnatal process and reach emmetropia at the age of 5 years, which is the age of adolescence for rhesus macaques. The axial length continues to grow up to 12 years of age and begins to shorten after 20 years,¹⁶ when adult rhesus macaques begin to develop myopia. Nevertheless, rhesus macaques with spontaneous high myopia are rarely reported.^16,17 

In this study, we aimed to evaluate the prevalence of myopia in rhesus macaques in southwest China and investigate ocular features at the same time. Fundus photographs were taken to grade myopic fundus. Refractive error (RE) and axial length (AL) were measured. The purpose of this study was to evaluate the prevalence of spontaneous myopia in adult rhesus macaques, to develop a dataset for the refractive state and AL in myopic rhesus macaques, and to describe the characteristics of myopia in rhesus macaques. 

Methods

Animals

We randomly recruited 219 adult rhesus macaques for ophthalmic examination. All animals were housed at an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-accredited facility (Kunming Biomed International, Kunming, Yunnan, China). All of the ophthalmic examinations were performed in accordance with the guidelines of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The project was approved by the animal experimental ethics board of the Ethics Committee (approval no. K001116023-02,01). 

Ophthalmic Examinations

Macaques were fasted and water deprived for 8 hours before examination. Intramuscular injections of 50 mg/mL ketamine and 0.5 mg/mL atropine were administrated. All ophthalmic examinations were completed within 120 minutes under anesthesia. Animals were monitored by an experienced veterinarian during the entire procedure. Animals were positioned prone, with their chin placed on the chin rest of the device. Eyes with obvious lens opacity were excluded. The refractive state was measured by autorefractor keratometry (RC-5000; Tomey Corporation, Aichi, Japan) after cycloplegia. Refractive error was recorded as spherical equivalent in diopters (D), which was calculated as sphere plus half cylinder. Axial length was measured with a Lenstar LS 900 optical biometer (Haag-Streit AG, Köniz, Switzerland). Fundus photographs were captured with a Kowa VX-20 camera (Kowa Company, Nagoya, Japan) with the 45° field centered on the macula. An eye speculum was used during measuring, and sodium hyaluronate eye drops were used to maintain the corneal tear film. Figure 1 provides a flowchart of the examination process. 

Figure 1.

 

Ophthalmic examinations and fundus tessellation grading used in this study.

Grading of Refractive Error

Based on the severity of RE, eyes were grouped into four categories.¹⁸ Eyes without myopia (RE ≥ −0.5 D) were classified as category 0, eyes with low myopia (RE < −0.5 D to > −3.00 D) were classified as category 1, eyes with moderate myopia (RE ≤ –3.00 D to > −6.00 D) were classified as category 2, and eyes with high myopia (RE ≤ –6.00 D) were classified as category 3. 

Assessment of Fundus Tessellations

The fundus tessellation grade was evaluated on the 45° macula-centered fundus photographs. A tessellated fundus appearance was defined as well-defined tessellations that could be clearly observed around the fovea and arcade vessels.¹⁹ With the Early Treatment Diabetic Retinopathy Study grid centered on the fovea, the macula was divided into a foveal region (diameter of 1 mm), a parafoveal region (diameter of 3 mm), and a perifoveal region (diameter of 6 mm). Fundus tessellations were classified into three grades, where grade 1 was defined as tessellations found in the perifoveal region and a region within the upper and lower vascular arcades, grade 2 was defined as tessellations found in the parafoveal region, and grade 3 was defined as involvement of the foveal region (Fig. 1). Two independent retinal specialists (Y.M. and Q.Z.) reviewed the fundus photographs and further performed classification of each grade. Disagreements regarding classification were resolved by a senior retinal specialist (Z.B.J.). 

Statistical Analysis

Statistical analysis was performed using the SPSS Statistics 23.0 (IBM, Chicago, IL). Descriptive statistics included mean values (presented as mean ± SD), median, and interquartile range (IQR). Spearman's correlation was used to explore the correlation between RE and AL. The normal distribution of parameters was determined using the Shapiro–Wilk test. The Mann–Whitney U-test was used to compare variables among groups. Receiver operating characteristic (ROC) curves were used to determine the cutoff values of RE and AL in myopic eyes with fundus tessellations. Area under the curve of the receiver operating characteristic (AUROC) was used as an index for accuracy of these two parameters. One-way analysis of variance with Bonferroni correction was used to test the differences among groups. To compare categorical variables, χ² tests were used. Logistic regression was performed to determine the association with β-peripapillary atrophy (β-PPA). P values represent results for two-sided tests, and P < 0.05 was considered statistically significant. 

Results

We randomly screened 429 eyes of 219 rhesus macaques. The mean age was 14.07 ± 2.72 years (range, 8–21). All monkeys were male. The distributions of RE and AL are presented in Figures 2A and 2B. The median RE was −1.25 D (range, −17.25 to 2.96; IQR, −4.88 to −0.04), and the median AL was 18.69 mm (range, 16.98–23.15; IQR, 18.21–19.47). Significant correlation between RE and AL was found (r = −0.686; P < 0.001) (Fig. 2C). The RE and AL data for the left eyes of rhesus macaques in the different age groups are presented in Table 1. 

Figure 2.

 

Distributions of refractive error and axial length of screened eyes. (A, B) RE and AL distribution among the screened eyes of rhesus macaques. (C) RE was significantly correlated with AL (r = −0.686; P < 0.001) and became worse with increased AL. (D) Comparison of ALs in the four categories of RE. Significant differences were found among the four categories (P < 0.001).

Table 1.

 

RE and AL of Rhesus Macaques at Different Ages

 

After grouping eyes into four categories of RE, there were 161 eyes without myopia in category 0, 122 eyes with low myopia in category 1, 55 eyes with moderate myopia in category 2, and 91 eyes with high myopia in category 3. Eyes without myopia showed the shortest AL (18.28 ± 0.51 mm), and eyes with high myopia showed the longest AL (20.53 ± 1.03 mm). These findings suggest that AL increases with RE worsening (P < 0.001) (Fig. 2D, Supplementary Table S1). The prevalence of myopia was 62.47% (268/429) (Fig. 3A). Furthermore, 33.96% (91/268) of these myopic eyes were highly myopic (Fig. 3B). 

Figure 3.

 

Prevalence of myopia in the screened rhesus macaques. (A) The prevalence of myopia was 62.47% (268) in 429 eyes of rhesus macaques. (B) Among these 268 eyes, 45.52% (122) had low myopia, 20.52% (59) had moderate myopia, and 33.96% (91) eyes with high myopia.

Based on the fundus photographs, we evaluated the presence of tessellated fundi in non-myopic eyes and myopic eyes. Fundus tessellations were found in 42.54% (114/268) of the myopic eyes and 6.21% (10/161) of the non-myopic eyes (P < 0.001). We then evaluated the presence of fundus tessellations in each RE category (Supplementary Figs. S1A, S1B). Tessellated fundi were found in 6.21% (10/161) of eyes in category 0, 10.66% (13/121) of eyes in category 1, 56.36% (31/55) of eyes in category 2, and 76.92% (70/91) of eyes in category 3 (Supplementary Table S2). Post hoc tests showed that the presence of tessellated fundi was significantly higher in eyes in categories 2 and 3 (moderate and high myopia, respectively) than eyes in categories 0 and 1 (no myopia and low myopia, respectively) (Supplementary Fig. S1C). These results indicate that fundus tessellations are commonly seen in myopic eyes, especially in eyes with moderate to high myopia. 

In myopic eyes, fundus tessellations tended to present in eyes with worse RE and longer AL (P < 0.001) (Supplementary Table S3). Therefore, ROC analyses were conducted to determine the cutoff values of RE and AL in myopic eyes with tessellated fundi (Supplementary Fig. S2). The curves showed that the cutoff value was −3.52 D for RE (75.97% sensitivity, 85.96% specificity) and 19.38 mm for AL (87.66% sensitivity, 81.58% specificity). AUROC_RE was 0.8718 (95% confidence interval [CI], 0.8303–0.9133) for RE, and AUROC_AL was 0.9199 (95% CI, 0.8870–0.9527) for AL. Both RE and AL showed high accuracy in detecting fundus tessellations. 

Subsequently, we assessed the degree of fundus tessellations in myopic eyes. We identified 25, 18, and 71 eyes in grades 1, 2, and 3, respectively (Supplementary Figs. S3A–S3C). Among these three grades, significant differences were found in RE (F = 4.847, P < 0.01) and AL (F = 8.711, P < 0.001) (Supplementary Table S4). Post hoc tests showed significant difference between grades 1 and 3 in RE (P < 0.01). Significant differences were also found between grades 1 and 2 (P < 0.05) and between grades 1 and 3 in AL (P < 0.001) (Supplementary Figs. S3D, S3E). These findings indicate that the degree of fundus tessellations is correlated with RE and AL in myopic eyes. 

Among the screened eyes, no eye was found with myopic maculopathy more severe than tessellated fundus, whereas β-PPA was present in 140 of 429 eyes: 48.1% (129/268) of myopic eyes and 6.83% (11/161) of non-myopic eyes (Figs. 4A, 4B). The presence of β-PPA differed significantly between myopic and non-myopic eyes (P < 0.001) (Fig. 4C). In the myopic group, eyes with β-PPA showed longer AL and worse RE (Figs. 4D, 4E). Table 2 shows the regression analyses for β-PPA risk factors. In our study, the risk for developing β-PPA increased by 4.82 times for each millimeter increase in AL (P < 0.001) and by 7.04 times for eyes with fundus tessellations (P < 0.001). Increased RE was also associated with the presence of β-PPA (odds ratio, 0.754; 95% CI, 0.862–0.986; P = 0.03). 

Figure 4.

 

Presence of β-PPA in myopic and non-myopic eyes. (A) Normal optic disc in non-myopic eye. (B) Optic disc with β-PPA in myopic eyes. The zoomed-in optic disc is presented on the lower left. (C) The presence of β-PPA was significantly different in myopic and non-myopic eyes (P < 0.001). (D, E) Comparisons of RE and AL between eyes with and without β-PPA (P < 0.001). (F) The presence of β-PPA was significantly different in eyes with grade 1 tessellations and eyes with grade 3 tessellations (P < 0.01).

Table 2.

 

Logistic Regression Analysis for the Determinants for β-PPA

 

Because the presence of β-PPA was closely associated with the presence of tessellated fundus, we further compared the presence of β-PPA in myopic eyes with the various fundus tessellations grades. Significant differences were found among the three grades of tessellated fundi (χ² = 6.106; P = 0.03) (Supplementary Table S5). The presence of β-PPA was found to be 76% (19/25), 88.9% (16/18), and 94.4% (67/71) in grades 1, 2 and 3, respectively. Post hoc tests showed a significant difference between grades 1 and 3 (P < 0.01) but no difference in the presence of β-PPA between grades 1 and 2 or grades 2 and 3 (P > 0.05) (Fig. 4F). These results indicate that the presence of β-PPA is another biomarker of myopia, especially in eyes with obvious fundus tessellations. 

Discussion

In this study, we evaluated the prevalence of myopia in adult rhesus macaques in southwest China. We found that 62.47% of the eyes were myopic, and one-third of these myopic eyes were highly myopic. Because eyes with moderate and high myopia were prone to present fundus tessellations, we conducted ROC analyses. Fundus tessellations can be found in myopic eyes, with cutoff values of −3.52 D for RE and 19.36 mm for AL. Furthermore, after grading myopic eyes based on the manifestation of fundus tessellations, we found that higher grades of tessellation were related with worse RE and longer AL. Finally, we found that the presence of β-PPA was closely associated with AL, RE, and the presence of fundus tessellations; therefore, fundus tessellations and the presence of β-PPA are indicators for myopia. We found that, the higher the grade of myopia, the greater the prevalence of fundus tessellations and β-PPA. 

The rate of aging of rhesus macaques is about three times that of humans, and the median life span of rhesus macaques is 27 years.²⁰ Generally, macaques that are <5 years old are juvenile, those that are 5 to 20 years old are adult, those that are 20 to 25 years old are geriatric, and macaques that are >25 years are considered aged.²⁰ In previous studies, ocular biometric measurements have been acquired in infants (3 weeks of age)²¹ and juveniles (up to 5 years of age).^22,23 Seldom have measurements been recorded in geriatric macaques.^17,24 In the present study, the median age of macaques was 14 years (IQR, 12–16), which means that most of the macaques were at midlife and thus we have developed a dataset for refractive state and AL for middle-aged rhesus macaques. 

AL and RE in the present study showed an inclination to myopia. The median RE was –1.25 D and the median AL was 18.69 mm in this cohort. Studies reporting ocular measurements of rhesus macaques are listed in Table 3. There is a trend for AL to increase with age in rhesus macaques. In human beings without myopia, AL has been shown to increase in the first 20 years of life but then decrease 7.3 µm/y between the ages of 20 and 80 years.²⁵ Different from studies in humans, the studies listed in Table 3 reported longer ALs in adult and aged macaques. Therefore, we surmise that the AL of rhesus macaques increases during their entire lifetime and that consequently they develop myopia with age; however, more data on ocular parameters are required to reveal the various stages of ocular development of rhesus macaques. 

Table 3.

 

Summary of Ocular Measurements of Rhesus Macaques

 

In this study, data regarding RE were obtained by autorefraction and data for AL were obtained using optical techniques. RE measurements can be done by both objective and subjective methods. The objective methods, including autorefraction and retinoscopy, are used in the measurement of laboratory animals and offer the benefits of high efficiency, accuracy, and repeatability. Because objective measurements are usually taken with the animal under anesthesia after cycloplegia, the refractive state differs from that when the animal is awake.²⁶ Subjective refraction is a method to measure an animal's functional RE. Investigators use behavioral measurements to obtain the optimal lens power that yields sufficient contrast sensitivity for high spatial frequency grating targets.²⁷ But, this process is time consuming and relies on the cooperation of the animal. In the study by Hung et al.,²⁶ autorefraction and retinoscopy yielded similar results for refractive error, but, compared with subjective measures, the objective measures yielded higher amounts of hyperopia. In most studies of ocular measurements of rhesus macaques, refractive states are obtained by objective methods.^16,21,28 In our study, we also used the convenient and repeatable autorefraction approach to obtain RE in the large-scale screening of rhesus macaques. AL can be obtained with both optical and ultrasound techniques. A-scan ultrasound biometry is the traditional method to obtain ocular dimension data in vivo. The advantage of ultrasound measurement is that it is not affected by optical media opacity. A disadvantage is that it requires placing the ultrasound probe on the corneal apex, which increases the risk of corneal abrasion and infection. Optical methods, based on the principles of low-coherence interferometry, partial coherence interferometry, and swept-source optical coherence tomography, allow ocular parameters to be obtained in a non-contact way.²⁹ In humans, optical measurements have been shown to be more reliable and accurate than ultrasound techniques, because the ultrasound probe compresses the cornea.³⁰ She et al.²³ proposed that inter-instrument discrepancies might be the result of differences in ocular size and component dimensions between animals and humans.²³ In the study by She et al.,²³ the data obtained by optical measurement were consistent with data obtained by ultrasound in young rhesus macaques; therefore, optical measurement was a precise method for measuring AL in this study. 

Tessellated fundi and β-PPA are the most common myopic fundus changes in humans.³¹ Fundus tessellations result from transparency of the retinal pigment epithelium and choriocapillaris and increased pigmentation of choroidal stroma.³² Tessellated fundi can be seen in myopia,³³ primary open-angle glaucoma,³² and aging.³⁴ In studies of humans, the grade of fundus tessellations can represent the severity of early-stage myopic maculopathy.^35,36 In the present study, we classified myopic eyes into three grades based on the involvement of tessellations. There was a trend that eyes with longer AL and worse RE showed more severe fundus tessellations. The outcomes have been the same with studies conducted in humans.³⁵ 

β-PPA is characterized by an absence of photoreceptors, atrophy of the retinal pigment epithelium and choriocapillaris, and visible large choroidal vessels and sclera.³⁷ The pathogenesis of β-PPA is still controversial. Kim et al.³⁸ suggested that it is a result of scleral stretching; however, in the study by Lee et al.,³⁹ the authors observed that the β-PPA enlarged in the same direction as the vascular trunk dragging during ocular elongation. Because the attachment between the retinal pigment epithelium and Bruch's membrane is tighter than that between the inner retinal structure and outer walls (lamina cribrosa and sclera), the lamina cribrosa would shift due to overwhelming growth of sclera during axial elongation. The presence of β-PPA was found to be around 40% to 55% in population-based studies,^40–42 compared with 75% to 80% in myopes.^43,44 In high myopes with AL exceeding 27.5 mm, the prevalence can even reach 98%.⁴⁵ The presence of β-PPA is associated with age, myopic refractive error, long AL, and visual acuity,^45,46 and the area of β-PPA is positively associated with AL.⁴⁷ Yan et al.³² reported that fundus tessellations were associated with β-PPA. Also, AL has been found to be an independent factor associated with β-PPA in high myopia.³¹ We found the same pattern in macaque monkeys. In our study, the presence of β-PPA was significantly higher in myopic eyes (48.10%), especially in eyes with longer ALs and higher grades of tessellations. This result further confirms β-PPA as a biomarker for the myopic rhesus macaque. 

In this study, we classified myopic eyes into three grades based on the involvement of tessellations. Significant differences in RE, AL, and the presence of β-PPA were found among the three grades, although no difference was found when grade 2 was compared with the other two groups. One possible reason is that grade 2 is an intermediate state between grade 1 (mild) and grade 3 (severe). Another explanation is that the small sample size for grade 2 (n = 18) made the result statistically insignificant. 

There are some limitations of our study. First, it was a cross-sectional study, and a longitudinal study design would perhaps be preferred to observe the progression of myopia in rhesus macaques. Second, retinal function was not tested in this study, in which we mainly focused on the prevalence of myopia and morphological changes. Studies examining the visual function of myopic macaques would further verify the phenotype–function relationship. Finally, the macaques in this study were all male. Because female macaques are a scarce resource for research, only ocular biometry measures of male macaques were acquired. It is unknown whether the development of myopia differs between sexes. Further studies with data acquired from both male and female macaques would address this limitation. 

In conclusion, we found a high prevalence of myopia of adult rhesus macaques in southwest China. We propose that the cutoff values of –3.52 D for RE and 19.38 mm for AL would serve as indicators of the presence of fundus tessellations in myopic rhesus macaques. We found that tessellation grades and the presence of β-PPA in adult rhesus macaques shared the same patterns with humans; that is, longer ALs and a higher grade of myopia corresponded to a higher prevalence of tessellations and β-PPA. This study provides support for the notion that adult rhesus macaques, especially aged ones, can serve as animal models for the spontaneous development of myopia in research on the pathogenesis of myopia. 

Acknowledgments

The authors thank all members of the 502 team for their discussions; Zhou Dong for development of the algorithm; Shan-Shan Jin, PhD, for statistical analyses; and Lichuan Yang, DVM, and the veterinary staff with Kunming Biomed International for their assistance. 

Supported in part by grants from the National Natural Science Foundation of China (82125007) and Beijing Natural Science Foundation (Z200014) and by the Youth Beijing Scholar program. 

Disclosure: Y. Ma, None; Q. Lin, None; Q. Zhao, None; Z.-B. Jin, None