The Relationship between Growth Spurts and Myopia in Singapore Children

Abstract

Purpose.: We aimed to determine the relationship between puberty and growth spurts with peak spherical equivalent (SE) or axial length (AL) velocity in Singapore schoolchildren.

Methods.: In the Singapore Cohort Study of the Risk Factors for Myopia of 1779 schoolchildren, the longitudinal refractive and pubertal status of 892 boys and 887 girls from ages 6 to 14 years were assessed. The study sample included 1329 Chinese, 316 Malays, 114 Indians, and 20 children of other races. Information regarding puberty parameters, age of peak height velocity, age of menarche, and break of voice (BOV) was obtained. Peak velocity was defined as the greatest change in measurements over a period of 1 year. Tanner stage 1 for pubic hair or breast development, in boys and girls, respectively, at age 12 was categorized as “later puberty,” whereas stages 2 to 5 corresponded to “earlier puberty.” Refractive error was determined by cycloplegic autorefraction using the Canon RK-F5, and AL was measured using the A-scan biometry machine.

Results.: The children were examined annually, and the mean number of visits was 5.7 ± 1.3. Age of peak height velocity occurred earlier in girls than in boys (11.0 ± 1.2 vs. 12.0 ± 1.7 years, P < 0.001). Girls with earlier peak height velocity experienced peak AL velocity and peak SE velocity approximately half a year earlier than those with later puberty (mean age of 10.3 ± 1.6 vs. 10.8 ± 1.7 years, P < 0.001; and 10.0 ± 1.5 vs. 10.6 ± 1.25 years, P < 0.001, respectively). Similarly, boys who had earlier peak height velocity also achieved peak AL and peak SE velocity earlier than those who experienced later peak height velocity (mean age of 10.4 ± 1.6 vs. 11.1 ± 1.8 years, P < 0.001; and 10.1 ± 1.5 vs. 10.6 ± 1.7 years, P = 0.01). Both girls and boys who had early peak height velocity had earlier age of onset of myopia than those with later peak height velocity (9.7 ± 1.4 vs. 10.1 ± 1.5 years for girls, P = 0.04; and 9.9 ± 1.5 vs. 10.4 ± 1.6 years for boys, P = 0.03). Myopia progression, in terms of AL velocity, also occurred earlier in boys and girls with earlier peak height velocity (10.2 ± 1.5 vs. 11.0 ± 1.9 for boys, P < 0.001; and 10.2 ± 1.5 vs. 10.7 ± 1.7 for girls, P = 0.004, respectively). The associations were not significant when Tanner staging, age of menarche, or BOV was used to determine stage of puberty.

Conclusions.: Boys and girls with earlier peak height velocity experienced earlier peak SE and AL velocity, and age of myopia onset. Thus, variations in the onset and peak progression of myopia may be associated with height spurts.

Introduction

There is a high prevalence of myopia, especially in Asian countries. ^1–3 Several environmental factors, such as near work including reading, writing, watching TV, and educational level have been associated with myopia. ^4–6 A history of myopia in parents or siblings also predisposes to myopia. ^7,8 Sex difference in the prevalence and progression of myopia has been reported, although the exact reasons are unknown. ^9–13 In the Beaver Dam Eye Study, the age-adjusted prevalence of myopia was higher in women than men (28.1% vs. 24.0%, P < 0.001). ⁹ In another study in the United States, the rate of myopia progression was fastest from ages 6 to 9 years and girls progressed faster than boys during the ages of 6 to 11 years. ¹³ This timing correlated with the onset of puberty, which was reported to be between 8.9 and 11.2 years. ^14,15 Puberty and peak height velocity occur earlier in girls than in boys. In Singapore, the average sex difference in age of onset of puberty is 1 to 2 years. ^16,17 A recent study in China involving 553 twin pairs demonstrated that the cross-sectional and longitudinal changes of axial length (AL) and height were concomitant in children between the ages of 7 and 15 years. ¹⁸ In this study, the investigators found that girls had greater changes in AL, whereas boys experienced more pronounced changes in stature. This suggested that the growth of the eye and changes in height may be governed by different mechanisms. 

In an animal experiment, there was a significant difference in the AL of the eyes between rats injected with growth hormone and rats injected with saline (0.6245 ± 0.0079 cm vs. 0.5932 ± 0.0149 cm) (Solomon AS and Hagin D. IOVS 2004;45:ARVO E-Abstract 1242). We postulate that the onset of puberty may contribute toward higher rates of myopia progression reported in girls. In this study, we aimed to describe and correlate various measures of puberty (height spurts, Tanner staging, break of voice, and menarche) with the onset and progression of myopia denoted by peak AL and spherical equivalent (SE) velocity in a longitudinal study of children aged 6 to 14 years in Singapore. 

Methods

A total of 1979 children in grades 1 to 3 (aged 7–10 years) from three schools located in the northern (n = 705), western (n = 961), and eastern (n = 313) parts of Singapore were enrolled in the Singapore Cohort Study of the Risk Factors for Myopia (SCORM) in 1999 and 2001. There were 1329 Chinese, 316 Malays, 114 Indians, and 20 children of other races. Children with serious medical problems, such as heart disease or leukemia, and those with eye disorders that impair vision, like syndrome-related myopia or congenital cataracts, were excluded from the study. The study was approved by the Ethics Committee, Singapore Eye Research Institute, and followed the tenets of the Declaration of Helsinki. Written informed consent was obtained from all parents after the nature of the study was explained. Analyses were performed using data collected yearly when the child was enrolled in school (aged 6–14 years; i.e., from 1999–2006 for two schools and from 2001–2006 for the third school). Children (n = 145) who attended fewer than three examinations between grades 1 and 8 were excluded from the study. A further 3% of children (n = 55) with unstable refractive error pattern, defined as both myopic and hyperopic but at different school visits, were also excluded. In this study, 1779 (89.8%) of 1979 children were followed until they were in grade 8. The methodology and initial cross-sectional results have been described. ^6,19  

Refractive Error Measurements

The children were examined yearly at their respective schools by a team of ophthalmologists, optometrists, and research assistants. The distance-corrected and uncorrected log minimum angle of resolution (Log MAR) visual acuity in each eye was measured according to standard protocol. ²⁰ Cycloplegia was induced using 3 drops of 1% cyclopentolate solution at 5-minute intervals after instillation of an anesthetic drop (0.5% proparacaine). Thirty minutes after the last drop of cyclopentolate solution was instilled, one of two autokeratorefractometers (model RK-5; Canon, Inc., Ltd., Tochigiken, Japan) was used to obtain the average of five consecutive measurements in the steeper and flatter meridian. The average of the keratometry readings was obtained. 

An A-scan biometry ultrasound unit (probe frequency of 10 mHz; Echoscan US-800; Nidek Co. Ltd., Tokyo, Japan) was used to measure AL. The average of six values was taken if the SD of the six measurements was less than 0.12 mm. One drop of 0.5% proparacaine was instilled into each eye before ultrasound biometry measurements were made. 

The age of onset of myopia referred to the age at which nonmyopic children developed SE of more than or equal to −0.5 D of SE. Peak AL velocity was defined as the greatest change in AL measurements over a period of 1 year. Peak SE velocity was defined as the largest increase in dioptric power over a period of 1 year. The corresponding age at which peak AL and SE velocity occurred was termed “age of peak AL or SE velocity.” 

Pubertal Assessment and Height Measurements

Height was measured annually for 8 years in adherence to a standard protocol. Each child's height was measured standing and without shoes. ^21,22 The age of peak height velocity was defined as the age of the largest difference of height measured between 2 consecutive years during the period of the study. The age of peak height velocity may be an indicator of puberty. 

Pubertal status was assessed by doctors in the school premises when the children were in grade 6 (age 12 years) according to a local standard criteria. ¹⁶ Breast development in girls and pubic hair development in boys were used to determine pubertal stage based on Tanner's standards. ^23,24 The age of menarche and break of voice (BOV) were based on subject recall. Boys or girls who experienced BOV or menarche before or after the median age of the subjects were said to have early or late puberty, respectively. 

Early puberty in boys and girls was defined as age of peak height velocity, menarche, or BOV that was lower than the 50th percentile of the study cohort. Tanner stage 2 to 5 of pubic hair development in boys or breast development in girls at 12 years of age corresponded to earlier puberty. “Later puberty” was denoted as age of peak height velocity, menarche, or BOV greater than the 50th percentile. Tanner stage 1 of pubic hair or breast development in boys and girls respectively at age 12 years was defined as “later puberty.” 

Data Analysis

The 2-sample t-test was used to compare the mean ages of peak AL velocity and peak SE velocity and myopia onset between boys and girls, and children with earlier and later puberty defined using different terms. An ANOVA test was performed to evaluate the association between age of onset of myopia, peak AL, and SE velocity (dependent variables) and early and late puberty (independent variables), after adjusting for confounding factors, such as sex, father's education, ethnicity, reading books per week, and parental myopia. The analyses were performed for all children, and repeated for children who were nonmyopic (age of onset of myopia, peak AL and peak SE velocities) and for children who were myopic at baseline (peak AL and SE velocities). The rate of myopia onset was defined as the number of new cases of myopia detected in a specific year divided by the number of nonmyopic children during the baseline visit. 

All probabilities quoted were two-sided and were considered statistically significant when less than 0.05. Data analysis was conducted using SPSS (PASW Statistics 18; SPSS, Inc., Chicago, IL). 

Results

There were 621 children (326 boys and 295 girls) who were myopic at baseline and 622 children (291 boys and 331 girls) who developed myopia subsequently. The total number of children who were myopic at the end of the study was 1243 (49.6% boys and 50.4% girls). The age of peak height velocity was notably earlier in girls with a mean of 11 (SD 1.2, range 7.7–14.8) vs. 12 years (SD 1.7, range 7.4–15.0) in boys (P < 0.001) (Table 1). The age of peak AL velocity occurred marginally earlier in girls (10.6 ± 1.7 years) than in boys (10.7 ± 1.7 years) (P = 0.04). Sex differences had no impact on age of peak SE velocity (P = 0.65) (Table 1). Figures 1 and 2 show the distribution of age of peak SE and AL velocity in children with early and late puberty. 

Figure 1. 

 

Distribution of age of SE peak velocity in children in early and late puberty.

Figure 2. 

 

Distribution of age of AL peak velocity in children in early and late puberty.

Table 1.

 

Comparison of the Age of PHV, AL Velocity, and SE Velocity for All Children, Boys, and Girls (n = 1779)

		All, n = 1779	Boys, n = 892	Girls, n = 887	P Value*
Age of PHV, y	Mean (SD)	11.47 (1.53)	11.97 (1.65)	10.98 (1.22)	<0.001
	Range	(7.38–14.99)	(7.38–14.99)	(7.69–14.82)
Age of peak AL velocity, y	Mean (SD)	10.64 (1.7)	10.72 (1.73)	10.55 (1.66)	0.04
	Range	(7.37–14.99)	(7.37–14.99)	(7.37–14.98)
Age of peak SE velocity, y	Mean (SD)	10.31 (1.58)	10.33 (1.64)	10.29 (1.52)	0.65
	Range	(7.39–14.98)	(7.39–14.98)	(7.40–14.90)

 PHV, peak height velocity.

*    P value for comparing the mean age differences between boys and girls, based on t-test.

The increase in AL and SE peaked at an earlier age (P < 0.001) in all girls and in boys with earlier peak height velocity (Table 2). In addition, the age of SE peak velocity was earlier than the age of AL peak velocity in both boys and girls. Nonmyopic children who experienced peak height velocity earlier developed myopia earlier (girls: 9.7 ± 1.4 years, P = 0.01; boys: 9.9 ± 1.5 years, P = 0.03) (Table 3). The progression of myopia, defined as peak SE or AL velocity in children who were already myopic, was also significantly faster in those with earlier peak height velocity and slower in those with later peak height velocity (10.2 ± 1.5 vs. 10.9 ± 1.8 for age of peak AL velocity, P < 0.001; 9.8 ± 1.4 vs. 10.2 ± 1.5 for age of peak SE velocity, P = 0.001) (Table 4). This was true for both boys and girls; however, the difference in peak SE velocity was not significantly different between boys with earlier peak height velocity and those with later peak height velocity (P = 0.13) (Table 4). 

Table 2.

 

Age of Peak AL and SE Velocity in All Children for Earlier and Later Puberty Onset as Defined by Age of PHV

		Early Puberty		Late Puberty		P Value*
		n	Mean (SD)	n	Mean (SD)
Age of peak AL velocity, y	All	888	10.34 (1.58)	891	10.94 (1.76)	<0.001
	Boys	446	10.36 (1.57)	445	11.09 (1.82)	<0.001
	Girls	442	10.31 (1.59)	446	10.79 (1.69)	<0.001
Age of peak SE velocity, y	All	888	10.05 (1.49)	891	10.56 (1.63)	<0.001
	Boys	446	10.09 (1.54)	445	10.57 (1.71)	<0.001
	Girls	442	10.02 (1.45)	446	10.56 (1.54)	<0.001

 Data presented are unadjusted mean (SD). Earlier puberty in girls is defined as age of PHV ≤ 10.9 (median) and in boys as age of PHV ≤ 12.2 (median). Later puberty in girls is defined as age of PHV > 10.9 (median) and in boys as age of PHV > 12.2 (median).

*    P value for early and late puberty in the multivariate linear models with age of peak AL and SE velocity as dependent variable, adjusting for sex, father's education, ethnicity, reading books per week, and parental myopia.

Table 3.

 

The Age of Onset of Myopia and Age of Peak AL and SE Velocity in Nonmyopic Children Who May Develop Myopia for Earlier and Later Puberty Onset as Defined by Age of PHV

		Early Puberty		Late Puberty		P Value*
		n	Mean (SD)	n	Mean (SD)
Age of onset of myopia, y	All	314	9.79 (1.46)	311	10.24 (1.51)	0.001
	Boys	145	9.86 (1.51)	146	10.41 (1.57)	0.03
	Girls	169	9.72 (1.43)	165	10.10 (1.48)	0.01
Age of peak AL velocity, y	All	314	10.20 (1.58)	311	10.84 (1.64)	<0.001
	Boys	145	10.16 (1.52)	146	10.92 (1.70)	<0.001
	Girls	169	10.24 (1.64)	165	10.78 (1.58)	0.001
Age of peak SE velocity, y	All	314	10.03 (1.41)	311	10.57 (1.57)	<0.001
	Boys	145	9.94 (1.46)	146	10.58 (1.60)	0.004
	Girls	169	10.11 (1.37)	165	10.57 (1.55)	0.004

 Data presented are unadjusted mean (SD). Earlier puberty in girls is defined as age of PHV ≤ 10.9 (median) and in boys as age of PHV ≤ 12.2 (median). Later puberty in girls is defined as age of PHV > 10.9 (median) and in boys as age of PHV > 12.2 (median).

*    P value for early and late puberty in the multivariate linear models with age of onset of myopia, age of peak AL and SE velocity as dependent variable, adjusting for sex, father's education, ethnicity, reading books per week, and parental myopia.

Table 4.

 

Age of Peak AL and SE Velocity in Myopic Children for Earlier and Later Puberty Onset as Defined by PHV

		Early Puberty		Late Puberty		P Value*
		n	Mean (SD)	n	Mean (SD)
Age of peak AL velocity, y	All	305	10.19 (1.48)	316	10.86 (1.80)	<0.001
	Boys	163	10.17 (1.45)	163	11.00 (1.88)	<0.001
	Girls	142	10.21 (1.53)	153	10.70 (1.70)	0.004
Age of peak SE velocity, y	All	305	9.80 (1.44)	316	10.18 (1.49)	0.001
	Boys	163	9.89 (1.47)	163	10.17 (1.61)	0.13
	Girls	142	9.69 (1.41)	153	10.19 (1.34)	<0.001

 Data presented are unadjusted mean (SD). Earlier puberty in girls is defined as age of PHV ≤ 10.9 (median) and in boys as age of PHV ≤ 12.2 (median). Later puberty in girls is defined as age of PHV > 10.9 (median) and in boys as age of PHV > 12.2 (median).

*    P value for early and late puberty in the multivariate linear models with age of peak AL and SE velocity as dependent variable, adjusting for sex, father's education, ethnicity, reading books per week, and parental myopia.

When puberty was determined by Tanner staging of pubic hair in boys and breast development in girls, there was no significant correlation between timing of puberty and age of peak AL velocity (boys, P = 0.48; girls, P = 0.69) or SE velocity (boys, P = 0.88; girls, P = 0.82). This was true for all children (myopic or nonmyopic at baseline).There was also no significant correlation between timing of puberty, as defined by Tanner staging and the onset of myopia (boys, P = 0.08; girls, P = 0.77). Likewise, when age of menarche or BOV was used as a determinant of pubertal status, puberty had no significant association with the onset of myopia (boys, P = 0.53; girls, P = 0.77). Age of peak AL (boys, P = 0.37; girls, P = 0.38) and SE velocities (boys, P = 0.06; girls, P = 0.71) also did not correlate significantly with the onset of BOV or menarche. 

For children who were myopic at the baseline visit, the average SE change was 0.54 ± 0.29 D per year (boys: −0.49 ± 0.28 D per year; girls: −0.60 ± 0.29 D per year). Change in both SE and AL decreased with increasing age. The annual changes in SE and AL were 0.60 ± 0.55 D and 0.33 ± 0.33 mm, respectively, at 6 to 7 years of age, and were 0 ± 0.29 D and 0.15 ± 0.30 mm, respectively, at 13 to 14 years of age. Myopia onset rates were 3.3% at 6 years old, 14.4% at 7 years old, 20.9% at 8 years old, 15.6% at 9 years old, 7.4% at 10 years old, 5.1% at 11 years old, 2.4% at 12 years old, 0.4% at 13 years old, and 0.1% at 14 years old, respectively. 

Discussion

In our prospective study, Singaporean children aged 6 to 14 years with earlier growth spurts as defined by age of peak height velocity had myopia at an earlier age and were found to have earlier peak AL and SE velocity for all children and myopic children. However, we found that other measures of puberty, including Tanner staging, menarche, and BOV, were not associated with the onset and progression of myopia in both boys and girls. Thus, the onset and peak SE or AL velocity may be determined by height spurts alone and not puberty. 

A key question is the sex difference observed in the progression of myopia between teenage girls and boys. Some studies have found that myopia progression rates were higher in girls. ^25–28 In one study, the prevalence of myopia was significantly higher in girls than in boys (14.1% vs. 9.7%) between ages 11.1 and 14.4 years. ²⁵ In the study by Pärssinen et al., ²⁸ progression of myopia was significantly faster among the girls than boys in the third (P < 0.001) and fourth grades (P = 0.038). The differences in myopia progression rates were attributed partly to the fact that girls spent a greater amount of time on reading and close work, and relatively less time outdoors. ²⁸ The difference between the rates of progression of myopia in boys and girls of grades 3 and 4 could be attributed to the peak height velocity during this age. Our study shows that between ages 6 and 14, girls had earlier peak AL and SE velocity as compared with boys, although this was not statistically significant (Table 1). 

A novel finding of our study is the association between peak height velocity and peak SE or AL velocity in Singapore children. Children with earlier onset and peak progression of myopia were found to achieve peak height velocity at an earlier age than their counterparts who developed myopia later. This could be a result of the surge in growth hormone during the time when children achieve peak height velocity. One recent study showed that the administration of growth hormones in rats resulted in significant axial elongation and progression of myopia (Solomon AS and Hagin D. IOVS 2004;45:ARVO E-Abstract 1242). 

Interestingly, the study by Wang et al. ¹⁸ showed that the age of peak AL change (at 8 years) occurred 2 years before the age of peak height change (at 10 years) in the study group of both males and females. This was consistent with the findings in our study, in which peak AL velocity occurred earlier at 10.6 years and peak height velocity at 11.5 years (Table 1). The authors postulated that eye growth and stature may be influenced by different hormones, as greater AL elongation was noted in girls, but height change was greatest in boys and most significant in younger children. This hypothesis was also substantiated in a study of 5150 adult women aged between 40 and 85 years (median age of 50), from Madurai, Tirunelveli, and Tuticorin districts of south India. Higher rates of myopia were found in women (257 subjects, 38.8%) who had age of menarche earlier than 14 years. ²⁹  

Our study demonstrated a positive association between height velocity and myopia. We found significant association between myopia and peak AL and SE velocities, and peak height velocity, but not with other features of puberty, such as Tanner staging and menarche. We postulate that growth spurts and possibly growth hormone surges (Solomon AS and Hagin D. IOVS 2004;45:ARVO E-Abstract 1242), rather than pubertal changes, ³⁰ may be the significant mediators of the progression of myopia and changes in AL and SE velocities. One possible reason is that the measurement of height is more objective than the assessment of puberty by Tanner staging. Another explanation is that growth spurts rather than puberty per se may influence peak SE or AL velocity. Although peak height velocity is an integral feature of puberty, the timing of peak height velocity during the pubertal period may vary between individuals and the tempo of pubertal development. To the best of our knowledge, this is one of the first longitudinal studies to examine peak AL and SE velocity with growth spurts or puberty. 

Another important finding of our study is that the age of SE peak velocity was observed to be earlier than the age of AL peak velocity. The mechanism for this finding is unclear. We hypothesize that the emmetropia status in children is not only maintained by AL but also by lens and or corneal power. SE is a composite of AL, corneal power, and lens thickness. Thus, the impact of peak SE velocity on peak AL velocity may have occurred after some time, as there may have been more immediate changes in corneal power and lens thickness. Therefore, the age of SE peak velocity was noted to be earlier than the age of AL peak velocity. 

Our study has considerable implications for control or treatment of myopia. Daily instillation of a long-acting cycloplegic agent, such as low-dose atropine or multizone contact lenses, has been assessed in clinical trials. Our study suggests that atropine eye drops may be continually used throughout the period of major growth spurts in order to retard progression of myopia. In addition, monitoring the changes in height during treatment is useful, as this predicts spurts of progression of myopia. Hence, interventions such as atropine eye drops should not be discontinued until 2 to 3 years after the age of peak height velocity. The age of cessation of treatment should be at least 14 to 15 years for boys and 13 to 14 years for girls. Growth spurts are often known only after the process has started, which could be late for slowing eye growth. Thus, height measurements should be conducted more frequently than routine visits to the child's doctor. 

In conclusion, growth spurts may influence the age of peak SE and AL velocity. Both boys and girls with earlier peak height velocity had correspondingly earlier onset and progression of myopia as denoted by peak SE and AL velocities. There were no associations of onset and progression of myopia with other pubertal parameters. 

Footnotes

 Supported by National Medical Research Council Grant NMRC/0975/2005.

Footnotes

 Disclosure: V.C.-H. Yip, None; C.-W. Pan, None; X.-Y. Lin, None; Y.-S. Lee, None; G. Gazzard, None; T.-Y. Wong, None; S.-M. Saw, None