Factors that influence the incidence of myopia

Paper title: Environmental risk factors can reduce axial length elongation and myopia incidence in 6-to-9 years old children

Authors: J. Willem. M. Tideman, MD, MSc, Jan Roelof Polling, BoH, Vincent W. V. Jaddoe, Md, PhD, Johannes R. Vingerling. MD, PhD, Caroline C.W. Klaver, MD, PhD

Date:  August 2018

Reference: Ophthalmol 2019;126:127-36 [Link to open access paper]

The authors evaluated the Generation R dataset to establish that change to axial eye length (AL) and incidence of myopia from age 6 to 9 yrs in children born in Rotterdam, The Netherlands, is associated with the following risk factors:

• Parental myopia
• 1 or more books read per week
• Time spent reading
• No participation in sports
• Non-European ethnicity
• Less time spent outdoors
• Baseline Axial Length (AL): Corneal Radius (CR) ratio

The authors found that the risk factors they investigated, when combined, had a prediction power of 0.78 for onset of myopia between age 6-9rs.

The full list of predictor factors that were assessed is presented in Table 3 of the open access paper

Of these factors parental myopia was the most influential fixed factor with increased risk of becoming myopic with one parent who is myopic, and greater risk if both parents are myopic.

From the modifiable factors that were assessed, increased time spent on reading and number of books read per week had the greatest prediction score.

• Clearly there is a balance to strike between education demands and risk for developing myopia, but nonetheless the reported relationship between close work and myopia onset from 6-9yrs age needs to be understood by eye care practitioners for them to be able to provide reliable advice to children and their parents.
• Reading should not be vilified but eye care practitioners are well placed to advise the benefits of reducing time spent on closework where this is not detrimental to a child's education and wellbeing.

Other than change to axial eye length alone, which has an obvious relationship to myopia onset, the most useful measurable biometric factor reported in this study was axial length to corneal radius (AL-CR) ratio.

• AL-CR ratio is calculated by dividing AL by CR. E.g. for a child with AL 23.0mm and CR of 7.8mm would have a AL-CR ratio of 23.0/7.8 = 2.95.
• AL-CR ratio below 2.80 had no predictive value of myopia onset between 6 and 9yrs age
• Above this, risk increased with greatest risk of myopia onset in this age range when AL-CR ratio exceeded 3.00.
• In their analysis the authors truncated participants into quartiles. In the highest quartile AL-CR ratio was greater than 2.916 in which 24% of the children became myopic between 6 and 9yrs age.
• The authors also reported that children with higher AL-CR were more likely to benefit from interaction to reduce environmental risk factors.

If you have access to axial length measurement, then this analysis suggests that it is worth taking the extra step to measure CR and calculate and record the AL-CR ratio.

• Improving your familiarisation of this ratio across the patients you see may benefit you in understanding which young patients to monitor more closely for onset of myopia and to advise on modifying environmental factors towards trying to slow or prevent onset of myopia.  

The analysis benefits from a large sample size that is representative of the population from which it is drawn, however this naturally means that care needs to be taken when extrapolating study results across different populations. 

It needs to be kept in mind that the large population US based CLEERE study[1] did not find environmental factors to have any predictive value for onset of myopia. The predictive effect reported in this current study indicates:

• the need for more research on this topic.
• potential for variation across different populations in how predictor factors can be used to identify risk for children to develop myopia.

Success of cycloplegia may have influenced the calculated predictor score for ethnicity - the study being based in The Netherlands means that there are likely to be more children with dark irises that are less responsive to cyclopentolate in the non-European cohort.

The authors found strong associations between ocular biometry measured at 6yrs and myopia onset between 6 and 9yrs:

• These factors are the result of genetic variation and eye growth.
• however, they also reflect risk behaviour before measurement at 6yrs age, i.e. not risk behaviour between age 6 and 9yrs.
  • The authors suggest that this may result in underestimation of the profit that can be gained by behavioural change in children identified as higher risk from ocular biometry at age 6yrs, i.e. higher AL-CR ratio. 

The prediction scores in this study are less reliable in children younger than 6yrs or older than 10yrs.

• Research is needed in this younger and older populations to understand differences in value of these myopia onset predictors

Axial eye length (AL) and incidence of myopia were captured from 2136 children included in the Generation R study,[2-3] Rotterdam, The Netherlands at age 9rs who were found to be not myopic at age 6yrs. Daily life activities and demographic characteristics had been captured by questionnaire and were used to determine associations between individual lifestyle/demographic factors and incidence of myopia and axial eye elongation.    

The original dataset included 6690 children with ocular biometry measurements at age 6yrs and 5862 at age 9yrs with 4734 having measurements at both visits and included in the risk factors for axial eye growth analysis. Cycloplegic refraction was only approved 1.5yrs into the age 9yrs measurements resulting in 3362 children being eligible for cycloplegic refraction from which 898 refused cycloplegia. Of the remaining 2464 children a further 39 were excluded having been identified as myopic at age 6yrs. From these 2136 also had axial length elongation analysis measurements making them eligible for inclusion in the risk factors for myopia incidence analysis. Univariate and multivariate analyses were performed to determine the influence of individual and combined factors towards incidence of myopia and axial elongation. 

Axial eye length (AL) and corneal curvature (CR) were both measured in mm using the Zeiss IOLMaster 500 and used to calculate the AL-CR ratio, a measurement previously shown to be highly correlated with SE,[4] by dividing AL by CR.  

Automated cycloplegic spherical equivalent (SE) refraction was performed at age 6yrs when visual acuity was >0.1 logarithm of the minimum angle of resolution using LEA symbol visual acuity charts at 3m. Children with VA ≤0.1 logarithm of the minimum angle of resolution or no glasses or ophthalmic history were classified as non myopic. Children with cycloplegic refraction measures at age 9yrs with SE -0.50D or more were considered myopic, with incident myopia being the proportion of myopic children at age 9rs who were not myopic at age 6yrs.

Children's day-to-day activities were assessed by questionnaire to determine: 

• Age 6yrs
  • Hours per day  spent on computers and watching television
  • Hours per day spent on sports activities
  • Hours per day spent outdoors
  • Parent's myopic status
  • Maternal education
  • Family income
  • Parents country of birth (European vs non-European)
  • Height, weight and body mass index
  • Vitamin D
• Age 9yrs
  • Number of books read per week
  • Time spent reading
  • Interval duration of reading

Height and weight was measured at 6yrs age and used to calculate body mass index (BMI). 25-Hydroxy vitamin D levels were also measured. Birth parameters and gestational age were obtained using medical records.

The results from the study showed that from 6yrs to 9yrs old, an average axial elongation was approx. 0.21mm/year with no significant difference between boys and girls. Those children classed as myopic at 9rs age showed greater elongation over time 0.34mm per year vs 0.19mm per year for non-myopes. 

When each predictive factor was evaluated individually statistical analysis revealed that greater axial elongation was associated with several factors:

• Being from non-European descent
• Having at least one myopic parent
• Lower family income
• Low vitamin D levels from spending less time outdoors
• More time spent indoors on computers and reading
• A higher A -CR ratio at 6yrs old.  

When the predictive factors were combined in multivariate analysis, the associations were very similar and the same conclusions could be drawn as to the individual predictive factors.

223 children from the analysed 2136 cohort became myopic between age 6 to 9yrs indicating 10.4% incidence of myopia between the two visits. Myopic children who became myopic by 9yrs age were more often from low socioeconomic families, of non-European descent, more often had parents who were myopic, spent less time outdoors, read more books and had higher AL and AL-CR ratios when they were 6yrs old. 

From the multivariate analysis the authors calculated a myopia prediction score for each of the predictor variables with each assigned prediction score point being equivalent to an odds ratio of 1.43:

• Myopic parents: 1 = 1.2; 2 = 1.9
• Time spent outdoors: 0.5
• >5 hrs spent reading per week: 1.2

AL-CR ratio was also scored, with the reference of 2.80 being zero. As the ratio increased, so did the predictor score with a ratio of >3.00 having a predictor score of 12.8.  

The full list of predictor scores is presented in Table 3 of the open access paper

When combined, the risk factors assessed in this paper were reported to have an area under the curve (AUC) predictive value for myopia onset between 6-9yrs age of 0.78. Increase to AL was found to have the greatest predictive value for myopia onset in this age range of 0.85, which is not a surprise given that increasing eye length if not balanced by change in refractive power will push ocular focus in a myopic direction.

To test these predictor scores further, the authors applied the scores to individuals to find that of the 571 children with a predictor score of 3.5 or less, only 8 (1.4%) became myopic from age 6 to 9yrs. At the other extreme, 50 children from the cohort had a predictor score of 11 or more with 27 (54%) of these becoming myopic over the same 6yrs to 9yrs age range.   

AL-CR ratio at baseline showed statistically significant interactions with:

• Parental myopia
• Number of books read per week
• Reading distance
• Ethnicity 
• Environmental risk score 

AL-CR ratio >2.916 represented the highest quartile of this ratio in which 24% of the children became myopic between 6 and 9yrs age and all predictors except for sports participation showed the highest association with AL elongation in comparison to the other AL-CR ratio quartiles.  

The authors reached the conclusion that there are ocular and environmental factors involved in myopia development and that the environmental factors can play a large part especially for those children in the top 25% for having the highest AL-CR ratio.  They felt these children would benefit most from intervention strategies that if implemented should help limit myopia progression.

Paper title: Environmental risk factors can reduce axial length elongation and myopia incidence in 6-to-9 years old children

Purpose: To identify risk factors for axial length (AL) elongation and incident school myopia.

Design: Population-based prospective birth-cohort study.

Participants: Four thousand seven hundred thirty-four children examined at 6 and 9 years of age from the Generation R Study in Rotterdam, The Netherlands.

Methods: Axial length and corneal radius (CR) were measured with an IOLMaster 500 and daily life activities and demographic characteristics were obtained by questionnaire. Three thousand three hundred sixty-two children (71%) were eligible for cycloplegic refractive error measurements. Linear regression models on AL elongation were used to create a risk score based on the regression coefficients resulting from environmental and ocular factors. The predictive value of the prediction score for myopia (≤-0.5 diopter) was estimated using receiver operating characteristic curves. To test if regression coefficients differed for baseline AL-to-CR ratio, interaction terms were calculated with baseline AL-to-CR ratio and environmental factors.

Main outcome measures: Axial length elongation and incident myopia.

Results: From 6 to 9 years of age, average AL elongation was 0.21±0.009 mm/year and myopia developed in 223 of 2136 children (10.4%), leading to a myopia prevalence at 9 years of age of 12.0%. Seven parameters were associated independently (P < 0.05) with faster AL elongation: parental myopia, 1 or more books read per week, time spent reading, no participation in sports, non-European ethnicity, less time spent outdoors, and baseline AL-to-CR ratio. The discriminative accuracy for incident myopia based on these risk factors was 0.78. Axial length-to-CR ratio at baseline showed statistically significant interaction with number of books read per week (P < 0.01) and parental myopia (P < 0.01). Almost all predictors showed the highest association with AL elongation in the highest quartile of AL-to-CR ratio; incidental myopia in this group was 24% (124/513).

Conclusions: Determination of a risk score can help to identify school children at high risk of myopia. Our results suggest that behavioral changes can offer protection particularly in these children.

[Link to open access paper]

Myopia Profile - why binocular vision matters in myopia management [Link]