Our results showed that age was significantly associated with axial length (AL), axial length-to-corneal radius ratio (AL/CR), and interocular difference in axial length (ALD) during childhood and puberty. Sex and parental myopic state were also significantly associated with AL and AL/CR.
Previous studies showed that the annual ocular axial elongation of children in lower grades (age 6 to 9 years) was between 0.21 mm12 and 0.70 mm13. In children aged 6 to 12 years, the annual axial elongation was 0.36 mm14. In our study, the difference in AL between either two consecutive ages is 0.12 mm to 0.38 mm (from age 3 to 15 years), which is consistent with previous studies. Although data on cycloplegic refraction was not available in the current study, we analyzed the ratio of axial length to the corneal radius (AL/CR) as a substitute for refraction, which is also a reliable parameter in presenting refractive errors15. The sex difference in AL distribution was similar to that described in the previous studies16–18, while a larger corneal radius (CR) was also observed in boys. A higher mean AL/CR in the left eyes of boys was found, implying that there might be some biological differences that influence refractive development between sexes. We found that parental myopic state was significantly correlated with AL and Likewise, the myopic state of parents19–22, parental age10,23, and education level18 are correlated with the refractive status of the children. Parental socioeconomic background18 and living environment24 also affect myopic patterns.
Physical growth synchronizes with ocular growth, in which children with taller body height23 or greater height spurts25 are more likely to experience increasingly faster myopic progression. Even though the children’s height or bodyweight data were not acquired in the current study, we discovered that the AL of teenagers was not significantly longer than that of younger children after the age of 15 years. Our findings suggested that ocular growth adapts to physical development, and the average AL progression reaches a steady-state as physical growth stabilizes.
It has been recognized that outdoor durations, indoor studying, near work26, and sleep also have roles in ocular development, and these light-oriented activities are indispensable for normal ocular growth. There have been many studies
One of the vital chemicals in the light-related pathway, or circadian rhythm, is melatonin, also known as the most potent Zeitgeber27. Lieberman et al28 first hypothesized that sufficient natural outdoor illumination and artificial indoor lighting might suppress melatonin secretion. The high intensity of light exposure stimulates the intrinsically photosensitive retinal ganglion cells (ipRGCs), and triggers dopaminergic pathways29. Furthermore, Opn4−/− mice with impaired melanopsin-expressing retinal ganglion cells (mRGCs) were susceptible to myopic shift and reduced retinal dopamine and 3,4-dihydroxyphenylacetic acid (DOPAC) in form-deprivation30. Higher morning melatonin was observed in myopes than in emmetropes and in those who spent more time outdoors or were exposed to light31. Likewise, in myopic young adults32, the concentration of melatonin was higher than that in nonmyopic patients, but myopic ones were discovered to have poorer subjective sleep than nonmyopic patients31. Children with high myopia were also estimated to have severer sleep problems33. Short-wavelength light like blue light, which can be perceived by melanopsin, was shown to increase ipRGC- induced melatonin suppression in mice’s pineal gland34. Moreover, it was the middle-wavelength light instead of the short-wavelength light that induced a more myopic shift. It could be implicated that the effect of melanopsin on sleep is not dose-dependent. In our study, weekends sleep duration is positively associated with AL/CR in primary school students. In contrast, weekdays sleep duration is negatively associated with AL/CR in junior high school students. The difference between the two age periods might be contributed to other factors, including the difference in the workload.
Excessive “screen time” was significantly correlated with sleep deprivation in preschoolers, school-aged children35, adolescents36, and young adults37. Liu et al24 further suggested that it was the late bedtime that had a better predictive value in myopia progression. Some studies have ascribed myopia to electronic device usage or TV watching 18,38. Using electronic devices or watching TV is associated with a potential risk of eye overuse at a near or moderate distance, accommodative spasm, or acute acquired comitant esotropia39 in some cases. Inadequate sleep, much exposure to electronic devices or TV, or a heavy workload were raising concerns of myopic progression in children and teenagers worldwide, especially in eastern Asia. Our study found that neither outdoor durations nor total media exposure was significantly associated with AL or AL/CR in all ages except in individual age groups, which could not be conclusive due to the cross-sectional design. Since online courses have gained popularity in recent years, we should be aware that more “visual display terminal related syndrome” is taking place, and the progressively overwhelming study pressure may deprive children and teenagers of their sleep, aggravating myopic development.
The occurrence of anisometropia, or asymmetrical ocular development, is mainly attributed to monocular light suppression or abnormal visual stimuli. Therefore, activities that affect both eyes are less likely to be noticed. Our study found that age was significantly correlated with ALD and interocular difference in AL/CR (△AL/CR), while outdoor duration and total media exposure were not. The ALD of teenagers was not significantly larger than that of younger children after the age of 12 years. These findings highlight the variability in anisometropia during physical growth.
Previous studies have shown that in children aged 3 to 5 years, a higher level of hyperopia was the leading risk factor for amblyopia and strabismus40,41; in children aged 7 to 8 years who had anisometropic amblyopia (interocular difference of spherical equivalent refraction > 3D), there was an average ALD of 1.57 mm (average 0.32 to 3.16 mm)4; and in those age 11–12 years with unilateral amblyopia (BCVA < 80 ETDRS letters (0.8 Snellen) and a ≥ 2-line difference between the eyes), the average AL was 0.6 mm shorter than that of the other eyes42. Anisometropic amblyopia often features a significantly shorter axial length, and the eyeball would undergo extension during emmetropization. However, the hyperopic anisometropia produced by strabismus is often not compensable. Smith et al1 found that relative myopic defocuses in the deviating eye suppressed axial elongation, while the improved accommodative function in the fixating eye reduced opticalDATA error and initiated the emmetropization process increasing the interocular difference in AL or AL/CR. We also found that preschoolers' esotropic eyes had shorter AL, less AL/CR, and ALD. To avoid strabismus-related axial anisometropia in the future, early detection, and intervention of strabismic amblyopia with anisometropia are crucial for normal ocular development.
Studies that have focused on axial anisometropia are listed in Table 5.
Table 5
Reports on refractive/axial anisometropia during childhood and adolescence
No.
|
Author, year
|
Age
|
Number
|
ALD (mm)
|
Highlights
|
1
|
Abrahamsson et al43
|
1 year until 4 years
|
310
|
N/A
|
Anisometropia was variable during emmetropization or decreased from infancy.
|
2
|
Tong et al44
|
7–9 years
|
1979
|
0.70(0.65)
|
Anisometropia was correlated with ALD and was more prominent in myopic anisometropia
|
3
|
Chia et al45
|
9 years
|
543
|
0.05
|
1. Right eyes were longer(0.05 mm)
2.Dominant eyes were less astigmatic (0.20D).
|
4
|
Deng et al46
|
6-month
5 years, 12-15years
|
1827
|
N/A
|
1. The prevalence of anisometropia increases in children aged 12–15 years
2.Anisometropia was more prominent in nonemmetropes.
|
5
|
Donoghue et al6
|
6–7 years
12–13 years
|
1050
|
0.40(anisometropia ≥ 1D) and 0.10(anisometropia < 1D)
|
1. Anisometropia was more common in children aged 12 to 13 years with hyperopia ≥ + 2DS
2. Anisometropic eyes had greater ALD.
|
6
|
Deng et al47
|
9.29 ± 1.30
(at baseline)
|
358(at baseline)
|
< 0.025 mm/y
(93.3%)
|
1. Children who had more axial elongation would have greater ALD.
2. The amount of anisometropia at commencement did not affect myopia progression.
|
7
|
Hu10
|
10.0 ± 3.3 years
(4–18years)
|
6025
|
N/A
|
1. Refractive anisometropia was associated with longer AL and larger ALD.
2. Myopic anisometropia was correlated with paternal education and more time indoors, while hyperopic anisometropia did not connect with eye care habits.
|
8
|
Palamar et al48
|
11.09 ± 5.27
(4 to 33 months)
|
42
|
-0.95 ± 0.50
|
AL and mean keratometry were the leading causes of hyperopic anisometropia.
|
9
|
Bach et al49
|
30.62 ± 18.04 months
(3 months to 7 years)
|
165
|
N/A
|
1. The steepest increase in AL was present at ten months of age
2. ALD was not significantly different.
|
N/A: not available |
Our study has several limitations. First is the lack of cycloplegic refraction data. It prohibited us from comparing our findings with other myopic or anisometropic epidemiological studies. However, we provided a complete analysis of ocular development in students aged 3 to 17 years. This could be a reference for other studies in the absence of cycloplegia. Second is the cross-sectional nature of this study; we could not conclude whether children with longer ALs were more likely to have axial anisometropia, and a long-term prospective study is warranted. Despite a large population being involved, some statistical fluctuations might make the results less representative. The third is the extensive age range of the study population, rendering the comparisons of age-related daily activities difficult. To reasonably eliminate the age effect, we divided the study populations into four based on their study stage, and age still played a vital role in AL or ALD in the four age groups.
In conclusion, age was the main factor associated with axial length, and the interocular difference in axial length and sex difference between AL and ALD was also recognized in our study. Outdoor duration and total media exposure did not strongly correlate with AL or ALD in our study. Instead, sleep duration on weekends or weekdays might associate with the AL or ALD. Young age esotropia was correlated with shorter AL, AL/CR, or ALD. The measurement of axial length is a direct way to estimate the prevalence of anisometropia at young ages, and the mechanism between sleep durations and ocular development needs more biological investigations.