Independent Influence of Parental Myopia on Childhood Myopia in a Dose-Related Manner in 2,055 Trios: The Hong Kong Children Eye Study


To determine the effects on childhood myopia of parental myopia, parental education, children’s outdoor time, and children’s near work.


Population-based cross-sectional study.


A total of 6,155 subjects in 2,055 family trios (1 child and both parents). Cycloplegic autorefraction was measured for children and noncycloplegic autorefraction for parents. Parental education, children’s outdoor time, and near work were collected by questionnaires. Children were categorized into 10 groups based on parental myopia levels. Associations of the above factors with myopia were evaluated by regression analyses. The areas under the receiver operating characteristic curve (AUROCs) for myopia were evaluated.


Mild parental myopia did not increase childhood myopia’s risk, but the risk was 11.22-folds when both parents were highly myopic. Higher parental education (Father: OR 1.08, P = .046; Mother: OR 1.11, P = .001) and more reading time of children were risk factors (OR 1.21, P = .044). Reduced odds of myopia were associated with more time spent on outdoor activities (OR 0.78, P = .017). Notably, all these factors became insignificant after adjustment, except for parental myopia. Children with more severe parental myopia spent more time on reading, but less on electronic devices. Parental myopic status alone accounted for 11.82% of myopia variation in children. With age and parental myopia, the AUROC for myopia was 0.731.


Among parental and environmental factors, parental myopia confers, in a dose-related manner, the strongest independent effect on childhood myopia. Therefore children with high risk of myopia can be identified for early prevention, based on parental myopia data.


Globally, myopia is the most common ocular disorder, and predominantly so in Asian populations. However, there has been an increasing prevalence over the past decades also in the other populations. It is predicted that nearly half of the world’s population would be myopic by 2050, with as much as 10% being highly myopic. High myopia is associated with excessive eyeball growth leading to sight-threatening complications, including presenile cataract, glaucoma, retinal detachment, choroidal neovascularization, myopic macular degeneration, and macular hemorrhage. It is a major public health concern, posing heavy health and economic burden to the society.

Parental myopia is a known risk factor for childhood myopia development, indicating genetic contribution. , Zadnik and associates demonstrated that history of parental myopia was associated with children’s ocular size. Subsequent studies supported parental myopia as a risk factor for childhood myopia development. However, genetic contribution may not be the only risk, and environmental factors could be linked to parental myopia, which of itself affects children’s vision. , Myopic parents may create a myopigenic environment including habits of intensive near-work and limited time outdoors. Some studies suggested that time outdoor would neutralize the impact of parental myopia on childhood myopia. , Furthermore, the impact of parents’ myopia severity on children’s myopia development has not been established as a result of limited quantitative parental data. Results of several studies, which are based on self-reported history of parental myopia without actual refraction data, indicated a possible relationship between parental and childhood myopia. , , ,

Here we studied 2,055 family trios (1 child and both parents) from the Hong Kong Children Eye Study (HKCES). We investigated known myopic factors: refraction and education level of parents, as well as children’s outdoor time exposure and near-work time. We aimed to establish whether the severity of parental myopia has an influence on childhood myopia, and to evaluate whether this effect is independent of such environmental factors as children’s outdoor time and near work.



The study subjects were recruited from the HKCES, a population-based cohort study of eye conditions in 4,257 children of grade 1 to grade 3 (aged 6–8 years) from primary schools in Hong Kong. In brief, the HKCES was designed to determine the prevalence of children’s ocular disorders, including refractive errors, strabismus, amblyopia, and allergic eye diseases, and to identify the environmental and genetic determinants of these conditions. Sample selection was based on a stratified and clustered randomized sampling frame. In Hong Kong, all primary schools (n = 571) registered in the Education Bureau were stratified into 7 clusters according to population densities. In HKCES, the schools in each cluster region were randomly assigned an invitation priority according to the ranking numbers generated by computer. Invitations to participate in the cohort were sent according to the ranking numbers until the required sample was achieved in each cluster region.

All children and both parents were given complete ophthalmoscopic investigations and assessments of environmental factors by questionnaires. If only 1 parent, either mother or father, completed the assessment, the whole family was not included in this report. Siblings and twins were also not included. The project conformed to the tenets of the Declaration of Helsinki and obtained ethical approval from the Institutional Review Board of the Chinese University of Hong Kong. Informed consent was signed by all participants.

Ocular Examinations

Refractive status was measured for each child before and after cycloplegia using an autorefractor (Nidek ARK-510A, Gamagori, Japan). Two cycles of 1% cyclopentolate (Alcon, Vilvoorde, Belgium) and 1% tropicamide (Santen, Osaka, Japan) are given at 10 minutes apart. Thirty minutes after the last drop, a third cycle of cyclopentolate and tropicamide drops would be administered if pupillary light reflex was still present or the pupil size was less than 6.0 mm. Three or more readings of spherocylindrical autorefraction were taken and averaged at 30 minutes after the last drop of cycloplegic agent. Subjective refraction and best-corrected visual acuity were measured by an optometrist in those children with presenting visual acuity <20/25 in either eye. Noncycloplegic refraction was measured for parents.

Definitions of Myopia

Spherical equivalent refraction (SER) was defined as spherical diopters (D) plus one-half cylindrical diopters. In children, myopia was defined as SER of –0.50 D or less, emmetropia as –0.50 D < SER < +0.50 D, and hyperopia as SER ≥ +0.50 D. Mild myopia was defined as –0.50 D ≥ SER > –3.00 D, moderate myopia as –6 < SER ≤ –3.00 D, and high myopia as SER ≤ –6.00 D. In adults, myopia was defined as SER of –0.75 D or less. Otherwise, the grading of myopia was similar with that of children. Only data from right eye were included in the analysis in view of the high correlation between both eyes.

Questionnaire on Parental Education Level, Children’s Outdoor Activities, and Near Work

Validated questionnaires used in our study derived mainly from the Chinese version of those used in the Sydney Myopia Study (SMS), , so as to facilitate comparison between the 2 studies. The protocol was previously published. First, adjustments for cultural differences and local dialect were implemented by discussing with the representatives of local teachers, parents, and ophthalmologists to make the questionnaires culturally appropriate and linguistically accurate. Second, a pilot study was performed among the parents of 100 children to verify the questionnaires’ reliability and validity. For outdoor activity, it was found that the overall intraclass correlation coefficient between 2 repeated surveys (with an interval of 4 weeks) was 0.72, and that the Cronbach’s alpha coefficient of each item was 0.68. Near-work activities included homework and pleasure reading. Watching television (TV), videos, digital video discs (DVDs), and playing computer games were classified as midrange activities. Diopter-hour was defined as follows: (hours spent studying + hours spent reading for pleasure) × 3 + (hours spent playing video games or working on the computer at home) × 2 + (hours spent watching television) × 1. Total outdoor activities were divided into 2 categories, namely, outdoor for leisure (including walking, riding a bike, playing in park, and picnic) and sport activities. The average number of outdoor activity hours per day was calculated using the formula: [(hours spent on weekday) × 5 + (hours spent on weekend day) × 2]/7. Parental education was categorized according to Hong Kong’s education system: primary school, secondary school, associate degree, bachelor’s degree, and master’s degree or higher.

Questionnaires were administered to parents for their completion with assistance by a trained staff in presence or on the telephone. All data of questionnaires were doubly entered to ensure integrity and precision. And for the missing data in the questionnaires, the parents would be further contacted for completion.

Statistical Analysis

Prevalence and its 95% confidence interval were calculated for myopia in children. Parental influences on childhood myopia was evaluated as follows. First, parental myopia was categorized into 10 groups according to the combinations of paternal and maternal myopia severities . In each of the 10 groups, the prevalence of childhood myopia was determined. Odd ratios (ORs) of risk of childhood myopia in each group was calculated via logistic regression with adjustment for age and gender. Effect of (1) parental myopia; (2) parental education; (3) children’s outdoor time; and (4) children’s near-work time on childhood myopia development was evaluated separately via the logistic regression model with the adjustment of age and gender. Multiple logistic regression models were constructed to evaluate how each of the above factors contributes to childhood myopia. The areas under the receiver operating characteristic (ROC) curve of parental myopia, age, time of outdoors, near work, and parental education level were evaluated with parametric ROC regression. The parametric ROC curve regression model was a probit model; a normal cumulative distribution function with input of a linear polynomial in the corresponding quantile function invoked on a false-positive rate. A P value of less than .05 was considered statistically significant. All analyses were performed using Stata Statistical Software (version 14.0; StataCorp, College Station, Texas, USA).

Table 1

Demographic Comparison of 2,055 Trios With the Hong Kong Children Eye Study Cohort

Characteristic Current Study (N = 2,055) HKCES (N = 4257) P
Male-female ratio 1.08 1.1 .05
Children’s age (years) 7.61 ± 0.95 7.62 ± 0.96 .992
Children’s axial length (mm) 23.15 ± 0.95 23.15 ± 0.95 .876
Children’s SER (Diopter) 0.15 ± 1.59 0.14 ± 1.59 .741
Children myopia rate (%) 24.8% 25.0% .715
Parental SER (Diopter) –2.70 ± 2.88 –2.71 ± 2.95 .339
Low family income rate (%) 25.2% 38.0% <.001

HKCES = Hong Kong Children Eye Study, SER = spherical equivalent refraction.

Unless otherwise noted, values are mean ± SD. Low family income was defined as household income lower than 20,000 HKD.

Table 2

Myopia Prevalence in Children for Different Severities of Parental Myopia

Parental Myopic Status No. of Children Mean (±SD) Age of Children Myopia Prevalence in Children
% (95% CI)
No myopia + No myopia 165 7.66 ± 0.97 12.12 (7.94-18.07)
No myopia + Mild myopia 379 7.56 ± 0.96 12.66 (9.67-16.42)
Mild myopia + Mild myopia 236 7.68 ± 0.93 13.98 (10.11-19.03)
No myopia + Moderate myopia 226 7.66 ± 0.93 22.12 (17.17-28.02)
Mild myopia + Moderate myopia 329 7.61 ± 1.01 27.36 (22.8-32.44)
Moderate myopia + Moderate myopia 157 7.55 ± 0.88 33.76 (26.77-41.53)
High myopia + No myopia 130 7.71 ± 0.94 30.77 (23.41-39.25)
High myopia + Mild myopia 190 7.59 ± 1.04 31.05 (24.86-38)
High myopia + Moderate myopia 183 7.61 ± 0.92 45.36 (38.26-52.64)
High myopia + High myopia 60 7.67 ± 0.93 56.67 (43.86-68.64)
Overall 2,055 24.82 (23.69-26.29)


A total of 6,165 individuals from 2,055 family trios were included in this study. The mean age of the children was 7.61±0.95 years (range 6-8); and for parents, 41.06±5.95 years (range 25-70). The overall myopia prevalence in children aged 6-8 years was 24.8% ( Table 1 ). In parents, the prevalence of no myopia, mild, moderate, and high myopia was respectively 31.5%, 27.8%, 25.6%, and 15.2%. The demographics of this sample group were similar with the HKCES ( Table 1 ), except family income. The proportion of low-income families of the current study was lower than that of the HKCES. The distributions of refraction in father, mother, and children were shown in Figure 1 . The myopia prevalence in children increased with the severity of parental myopia ( Table 2 ).

Figure 1

Distribution of refraction in father, mother, and children.

Separate Effect on Childhood Myopia of Parental Myopia, Parental Education Level, Children’s Outdoor Activities, and Children’s Near Work After Adjustment of Age and Gender

Parental myopia has a dose-related effect on childhood myopia. Mild parental myopia conferred no effect (OR 1.15, 95% CI 0.65-2.02, P = .168; Table 3 ). The risk was 2.2-fold higher when one parent was moderately myopic (OR 2.20, 95% CI 1.24-3.91, P = .002; Table 3 ), and 11.22-fold higher when both parents were highly myopic (OR 11.22, 95% CI 5.49-22.93, P < .001; Table 3 ). The risk effect of myopia in children with a highly myopic parent may be reduced if the other parent was nonmyopic (OR 3.43, 95% CI 1.86-6.33, P < .001; Table 3 ) or mildly myopic (OR 3.74, 95% CI 2.11-6.63, P < .001; Table 3 ). Of note, the risk of childhood myopia was the same from paternal refraction (Beta coefficient: 0.12, P < .001, R 2 =12.5%; Figure 2 ) and from maternal refraction (Beta coefficient: 0.13, P < .001, R 2 =12.5%; Figure 2 ).

Table 3

The Influence of Parental Myopia and the Other Risk Factors on Children’s Myopia

Factors OR a (95% CI) P a Value OR b (95% CI) P b Value OR c (95% CI) P c Value
Parental myopic status
No myopia + No myopia Reference Reference Reference
No myopia + Mild myopia 1.15 (0.65-2.02) .168 1.09 (0.59-1.99) .788 1.09 (0.6-1.99) .78
Mild myopia + Mild myopia 1.27 (0.69-2.33) .414 1.24 (0.65-2.35) .514 1.23 (0.65-2.34) .52
No myopia + Moderate myopia 2.20 (1.24-3.91) .002 2.01 (1.08-3.74) .027 2.02 (1.08-3.75) .027
Mild myopia + Moderate myopia 3.09 (1.81-5.3) <.001 3.08 (1.73-5.47) <.001 3.10 (1.74-5.50) <.001
Moderate myopia + Moderate myopia 4.31 (2.39-7.75) <.001 4.23 (2.24-7.99) <.001 4.25 (2.25-8.01) <.001
High myopia + No myopia 3.43 (1.86-6.33) <.001 3.54 (1.84-6.83) <.001 3.55 (1.84-6.85) <.001
High myopia + Mild myopia 3.74 (2.11-6.63) <.001 3.84 (2.08-7.1) <.001 3.84 (2.08-7.10) <.001
High myopia + Moderate myopia 7.03 (3.99-12.37) <.001 7.78 (4.19-14.46) <.001 7.79 (4.20-14.47) <.001
High myopia + High myopia 11.22 (5.49-22.93) <.001 11.58 (5.35-25.06) <.001 11.65 (5.38-25.23) <.001
Paternal educational level 1.08 (1.00-1.17) .046 0.90 (0.80-1.00) .053 0.90 (0.81-1.01) .068
Maternal educational level 1.11 (1.04-1.18) .001 1.02 (0.92-1.15) .699 1.02 (0.91-1.15) .68
Time of outdoor 0.90 (0.8-1.02) .092 NA NA 1.00 (0.83-1.21) .978
Outdoor for sports 0.94 (0.77-1.13) .499 NA NA NA NA
Outdoor for leisure 0.79 (0.65-0.96) .017 NA NA NA NA
Diopter∗hour 0.99 (0.96-1.01) .326 0.99 (0.95-1.03) .721 1.00 (0.96-1.05) .935
Total near-work time (hours per day) 1.06 (0.96-1.17) .244 NA NA NA NA
Watching TV 0.88 (0.76-1.01) .074 NA NA NA NA
Doing home work 1.02 (0.84-1.24) .842 NA NA NA NA
Reading 1.21 (1.01-1.48) .044 NA NA NA NA
Using computer 1.02 (0.86-1.21) .804 NA NA NA NA
Using electronic devices 0.80 (0.69-0.93) .005 NA NA NA NA
Total near-work time 1.06 (0.96-1.17) .244 NA NA NA NA

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Aug 17, 2020 | Posted by in OPHTHALMOLOGY | Comments Off on Independent Influence of Parental Myopia on Childhood Myopia in a Dose-Related Manner in 2,055 Trios: The Hong Kong Children Eye Study

Full access? Get Clinical Tree

Get Clinical Tree app for offline access