Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 62
Appendix C
Review of the Progression Literature
There is considerable proliferation and overlap of terminology in studies of myopia. The
phrase myopia progression as it is used here refers to an increase with time in the negative
sphere dioptric power necessary for correction to best visual acuity. This should not be
confused with progressive myopia, which is degenerative or pathological in nature. Myopia
progression can occur either in simple myopia, in which the eve is anatomically normal.
healthy, and capable of normal corrected visual acuity. It can also occur in pathological
,
, ~ ~
~ 1 · 1 ~ 1 1 ~ ~ ~ · ~ . ~ · ~
myopia, In wn1cn one eye opntna~moscop1catly or otherwise snows visible pathological signs,
or has substandard corrected visual acuity, or both (Curtin, 1966, 1979~. The studies
discussed below deal almost exclusively with myopia progression in simple myopia.
REFRACTIVE CHANGES IN THE FIRST She YEARS OF LIFE
Refractive changes in the first six years of life are not usually included under the term
myopia oroeression. nor have they been extensively Flit hilt. t.h~v :'rP hri~flv ;:limm:`ri~A
~_ ~_ ~ A _ ~ ~ ~ ~ ~ ~ ~^ ^ ~ _ ~
~. ~
· · ~· _ · _ -_
~ 1 ~_ ~
here to provide a background for discussions of myopia progression in the school-age years.
I,ongitudinal studies remain to be clone, but refractive changes over this time span can
be implied from refractive error distributions at different ages (Hirsch, 1963~. Such a
comparison is made in Figure C-1. Cook and Glasscock (1951) measured refractive errors of
500 30-hour-old infants (185 white infants, 315 black infants) by retinoscopy under atropine
cycloplegia at the University of Arkansas School of Medicine Hospital. Most frequent
was hyperopia of 1.00 to 2.00 diopters (D.), with a range of -12 to +12 D. of the 1,000
eyes thus measured: 74.9 percent were hyperopic or emmetropic and 25.15 percent were
myopic. Kempf et al. (1928) took retinoscopic measurements of 333 white 6- to 8-year-olds
in Washington, D.C., under homatropine cycloplegia. The distribution of refractive errors
was highly leptokurtic, with a peak at 0 to 1.00 D. of hyperopia and a range of -3.00
D. to +3.00 D. About 97 percent of these children were hyperopic or emmetropic. While
the populations from which these samples were taken were diverse, comparison of the two
distributions would suggest that the variance of refractive errors reduces over the first six
years of life, some children becoming less myopic and some becoming less hyperopic.
Mohindra and Held (1981) reported a cross-sectional study of 400 full-term healthy
infants and children of ages from birth to 5 years, recruited from Boston and Cambridge,
Massachusetts, by mailings. Of these 400 infants and children, 312 were ages 0 to 2 and
88 were between ages 2 and 5. Spherical equivalent refractive errors were determined by
~. . ~
~- - - -- ~-_ _ _ A
62
OCR for page 63
63
40
35
30
-
- o
o-
-
~ 25
of
llJ
a 20
LU
lo
15
10
5
o
,,1
I I ~
,'1
I ~ I ,.-,S., ~ 14
7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1011
._ _ _. Age 6-8
(Kempf et al.)
\
t ~
·
_ Newborn
(Cook & Glasscock)
MYOPIA
HYPEROPIA
FIGURE C-1 Refractive error distributions for newborns based on the data of Cook and Glasscock
(1951) and for 6- to 8-year-olds from the data of Kempf et al. (1928~.
Source: Adated from Hirsch, 1963.
a noncycloplegic retinoscope technique that Mohindra and Held called "near retinoscopy"
(described by Mohindra, 1977; Borghi and Rouse, 1985~. The refractive error distribution
for the neonates (~4 weeks) was very similar to the data from Cook and Glasscock (1951),
with the exception that the distribution from Mohindra and Held (1981) was shifted toward
myopia- a difference that could be explained in part by differences between cycloplegic
~ ~ e ~ ~ r ~ e ~ ~ TO ~ ~ ~ ~ ~ ~ ~ · ~ · ~ ~
and noncyclopleglc results. Mon1n~ra and nela found tnat In sequential groups Irom ages
0-4 weeks to ages 129-256 weeks, the prevalence of myopia and astigmatism declined, the
prevalence of emmetropia (0 ~ 0.99 D.) increased, the prevalence of hyperopia remained
about the same, the mean refractive error shifted toward hyperopia, and the range and
standard deviation of refractive errors decreased.
Studies by Mohindra et al. (1978) using near retinoscopy and by Howland et al. (1978)
using photorefraction suggest that relatively high amounts of astigmatism are quite preva-
lent among infants. Three additional studies (Dobson et al., 1984; Gwiaz~a et al., 1984;
Howland and Sayles, 1984), each using a different refractive technique (retinoscopy under
cycloplegia, near retinoscopy, and photorefraction, respectively) suggest that this is pre-
dominantly against-the-rule astigmatism and that it decreases in amount over the first few
months of life, so that by about age 4, with-the-rule astigmatism is more prevalent.
Premature infants quite often have severe myopia, which is more common as birthweight
decreases over the first few months of life; many infants become emmetropic by age one year
(Fletcher and Brandon, 1955; Scharf et al., 1975; Yamamoto et al., 1979~. Myopia of high
OCR for page 64
64
degree also tends to develop in eyes with retrolental fibroplasia, which is often associated
with premature birth (Yamamoto et al., 1979~. Rabin et al. (1981) proposed that the
opacification of the ocular media in retrolental fibroplasia leads to the severe myopia, and
that this is comparable to the myopia that has been produced experimentally by depriving
animals of clear retinal imagery (Wiese! and Raviola, 1977, 1979; Goss and Criswell, 1981;
Criswell and Goss, 1983~. Reviews on the relationship of refractive error to premature birth
are available in the literature (Banks, 1980; Goss, 1985~.
JUVENILE MYOPIA
Typical Changes During the Sthool-age Ye are
During the middle of this century, thought on refractive changes was dominated by
concepts from cross-sectional and partially longitudinal studies of private practice records
(Brown and Kronfeld, 1929; Jackson, 1932; Brown, 1938, 1942; Slataper, 1950~. Cycloplegic
refraction was negative in sign from age 6 or 7 until the early 30s, the changes being lower
in magnitude after the middle teens. Brown and Kronfeld interpreted the ages 7 to 17 as
the "period of progression." Slataper suggested that there were phases of refractive change
corresponding to different periods of the life-span: (a) the axial myopic change of the young,
involving an average change of -3.33 D. from between ages 8 and 30; (b) the hypermetropic
change of middle age, involving an average change of+~.36 D. between ages 31 and 64; and
(c) senile myopic change, involving an average change of -2.37 D. between ages 65 and 87.
Slataper (1950) also recognized that refractive changes were greater for myopic eyes than
for hyperopic eyes. He reported an average annual change between ages 8 and 22 of -0.29
D. for 1,704 myopic eyes and -0.14 D. for 2,003 hyperopic eyes.
Hirsch (1961, 1964a, 1964b) reported on a twice yearly manifest retinoscopy of school-
children in Ojai, California. He started with 1,200 5- or 6-year-olds and was able to follow
about 500 of them until ages 13 or 14. Hirsch (1964b) found that the refractive error at
ages 5 to 6 did have some predictive value for refractive error at ages 13 to 14. He proposed
that: (a) if a child has any degree of myopia at ages 5 to 6, the myopia will probably
increase; (b) if a 5- to 6-year-old child has a refractive error of +0.50 to +~.25 D., he will
most likely be emmetropic at ages 13 to 14 but may also show myopia or hyperopia; and
(c) with refractions between 0 and +0.50 D. at ages 5 to 6, a child will probably be myopic
by ages 13 to 14. He also noted that children who had 0.25 D. or more aga~nst-the-rule
astigmatism when entering school were more likely to develop myopia than children with
spherical refractions, a relationship that was significant by chi-square analysis. Hirsch's
studies and Langer's, cited below, indicate that the majority of children in a unselected
sample have relatively small changes in refraction, and the myopes are set apart by their
rapid changes by comparison.
Langer (1966) reported on a similar vision survey of schoolchildren in Leaside, Ontario,
a suburb of Toronto with a population of generally high socioeconomic status. Children were
seen in grades K, 1, 3, 5, 7, 9, 11, and 13. Refractive measurements were made by manifest
retinoscopy; spherical equivalents were used for analysis. The number of refractions per
child ranged from 3 to 9 (mean 6.9~. For children ages 5-6 through 15-16, the percentage
who were myopic (any amount of negative refractive error) increased, and the percentage
of hyperopes (+1.00 D. refractive error or more) decreased. Both Hirsch (1961) and I.anger
found that the majority of individual plots of refractive error with age were linear. Langer
noted that most of the nonlinear individual plots had slopes that were closer to zero. In 81
percent of the subjects, Langer was able to predict the final refraction within 0.50 D. by
OCR for page 65
65
linear extrapolation from the first three data points, even though 82 percent of the cases
had a time interval between the first and third points of less than 3.5 years.
A study by Hofstetter (1954), although derived from a selected sample, emphasizes
the difference in refractive changes between myopes and nonmyopes. The data used were
manifest subjective refractions from the office of an optometrist in Bloomington, Indiana.
Hofstetter calculated the dioptric change per month, using the mean of the spherical
equivalents of the two eyes. For the 1~ to 2~year-old age group, the majority of myopes
showed negative rates or further increase in myopia. Those subjects who were emmetropic
or hyperopic showed relatively less change (mean and mode rate both near zero), with some
showing positive shifts and some negative shifts.
To summarize the changes typical of childhood, most children show a low negative rate
of change in refractive error, but in most cases this does not result in myopia. Hyperopes
have been reported to have relatively stable refraction. Some of the apparent shifts toward
more hyperopia may represent latent hyperopia becoming manifest. Myopes are readily dis-
tinguishable by their relatively high rates of refractive change (compared with nonmyopes),
which typically stops or slows appreciably during the middle or late teens. Refractive error
distributions remain leptokurtic from age 6 to the teenage years, with the skew shifting
from hyperopia to myopia.
The studies on age of onset of myopia progression, rates of childhood progression, and
age of cessation discussed below are drawn almost exclusively from private practice records
of myopes. While this means, of course, selected samples, it is perhaps less of a detriment in
studying myopia progression than it would be in studying population characteristics. Even
in the absence of regular changes in spectacle correction necessitated by the increases in
myopia, myopes may be likely to return to the practitioner due to breaking or outgrowing
their spectacle frames.
Age of Onset
The incidence of myopia increases throughout the childhood years, beginning at about
age 5 or 6. This is illustrated by the vision surveys done by Hirsch (1952) and Young et
al. (1954b). Hirsch reported on the spherical equivalent refractive errors of the right eye,
determined by manifest retinoscopy, of 9,552 randomly selected elementary schoolchildren
from the Los Angeles area. Included were children between the ages of 5 and 14 at the
nearest birthday. Young et al. (1954a and b) made a survey of several visual characteristics
of 652 schoolchildren in Pullman, Washington, a college town in a wheat farming area
of southeastern Washington. Young et al. stated that the largest occupational groups in
Pullman were college instructors and farmers. Young et al., like Hirsch, used spherical
equivalent refractions for the right eye determined by manifest retinoscopy. The myopia
prevalence data from these studies, as well as those from Langer (1966), are summarized in
Table C-1. Each of these studies shows an increase in prevalence with age. It may be noted
that the largest increase in prevalences for girls occurs between the 7-8 and ~10 age levels
for both the Hirsch (1952) data for all myopes and the Young et al. (1954b) data. For boys
the largest difference occurs between ages 9-10 and 11-12; this would imply that the most
common ages of myopia onset are about age 9-10 for girls and age 11-12 for boys.
Rosenberg and Goldschmidt (1981) selected patient records from the 1974 to 1979 files
of an ophthalmologist in Denmark. Included were patients age 20 years or less, examined
at least twice, with a minimum observation period of 12 months. A total of 280 cases of
myopia (122 males and 158 females) were thus collected. Refraction was by noncycloplegic
subjective refraction or by subjective refraction under 1 percent cyclopentolate in cases of
OCR for page 66
66
TABLE C-1 Myopia Prevalence Among U.S. Schoolchildren in Vanous Populations, 1952-1966
I-Iirsch (1952) Young et al. (1954) Langer (1966)
Myopia Myopia Myopia . Myopia of Myopia
of any amount~ 1 D. > 1 D. any amount > 1 D.
A<,c Girls Boys Girls Boys Girls BoysGirls Boys Girls Boys
5-6 6.15 7.43 0.45 0.67 4.17 0.002.04 0.00 0.00 0.00
7-8 9.71 11.02 0.98 0.90 2.60 5.693.97 3.08 0.00 1.54
9-10 17.18 15.68 2.01 1.82 19.44 9.6812.20 11.68 6.71 5.11
1 1-19 21.60 20.74 5.77 3.08 20.00 27.2729.18 20.48 10.26 5.71
13-~4 25.36 22.53 5.78 5.08 25.71 28.5734.42 34.30 19.48 15.01
suspected unrelaxed accommodation. The age of myopia onset was defined as the age at the
first examination revealing myopia, or as the age at which corrective lenses were prescribed
for those patients already wearing them. The most common onset age for the boys was at
11-12, whereas the distribution for girls peaked at ages 9-10 and 11-12.
Septon (1984) distributed to 500 students in six consecutive second- year optometry
classes at Pacific University in Oregon a survey form with entries for the spherical equivalent
refractive error of the student's more ametropic eye as determined during the previous year,
and for the age at which myopic students first wore corrective lenses. Of the 500 students,
398 were males, and 89 percent were Caucasian. Of the 447 students who responded to the
survey, 332 were myopes. Septon concluded that the ages at which myopes first present for
care tend to cluster in three groups: major ones at ages 8-9 and 12-13 and a minor one at
age 19.
Bucklers (1953) in Germany and Lecaillon-Thibon (1981) in France proposed on the
basis of subjective examination of patient records that the earlier in childhood that myopia is
discovered (and presumably the earlier it begins), the greater is the subsequent progression.
This has been confirmed by several studies (Fletcher, 1964; Francois and Goes, 1975;
Rosenberg and Goldschmidt, 1981; Septon, 1984; Mantyjarvi, 1985a; Goss and Cox, 1985).
Rosenberg and Goldschmidt (1981) calculated the amount of increase of myopia in
terms of the mean of the spherical equivalents of the two eyes and related it to the age of
myopia onset. Higher amounts of increase were more common among those with earlier
onset ages. They also reported that, for 30 girls with onset at age ~10, the average annual
increase in myopia in the right eye was 0.47 D. (SD = 0.28), whereas for 36 girls with onset
at age 11-12, the average annual increase was 0.37 D. (SD = 0.42). Thus, an earlier onset
may be associated with a higher rate of progression.
Mantyjarvi in Finland presented data on 214 myopic children (136 girls and 78 boys),
ages 7 to 15, followed for 1 to 9 years and examined in the Kuopio, Finland, Community
Health Center. Refractive values used were the right eye spherical equivalents as determined
by retinoscopy using 1 percent cyclopentolate. The definition of age of myopia onset as used
in this study was not given. Results on the amount of myopia at age 15-16 as it related to
the onset age are summarized in Table C-2. For both boys and girls, the amount of myopia
was greater for earlier onset ages. M2ntyJarvi was cautious to point out, however, that
OCR for page 67
67
TABLE C-2 Mean Amount of Myopia at Age 15-16 According to the Age of
Myopia Onset
Age of
myopia Final refraction
onset N Mean SD
7-8 Gals 3 -6.17 1.20
Boys 6 -4.42 2.04
9 Gals 8 -4.91 2.52
Boys 9 -4.06 0.95
10 Gals 8 -4.00 1.14
Boys 6 -4.33 0.88
11 Gals 32 -3.27 1.26
Boys 15 -3.00 1.40
12 Gals 34 -2.82 1.06
Boys 15 -2.63 1.14
13 Gals 20 -2.59 0.95
Boys 6 -2.42 - 0.77
14 Gals 16 -2.09 1.18
Boys 12 -2.13 0.94
15 Gals 15 -1.23 0.67
Boys 9 -1.00 0.52
7-10 40 -4.46 1.63
11-13 122 -2.87 1.15
14-15 52 -1.66 1.00
SOURCE: Mansard (1985a).
OCR for page 68
68
there was considerable individual variation, as can be noted from the standard deviations
in Table C-2. Of children with onset before age 10, 12.5 percent remained below 3.00 D. of
myopia, 70 percent developed myopia of 3.00 to 5.75 D., and 1-7.5 percent had myopia of
6.00 D. or more. Of those children with onset at ages 11-15, 66.7 percent had less than 3.00
D. of myopia at age 15-16, 32.2 percent ended up with 3.00 to 5.75 D. of myopia, and 1.1
percent reached 6.00 D. or more myopia.
Goss and Cox (1985) presented data from private optometry practice records obtained
from five locations in the central United States. The refractive data used were the refractive
errors in the principal meridian nearest horizontal in the right eye as derived from the man-
ifest subjective refraction recordings. A linear regression equation of diopters of refractive
error versus age in years was calculated for subjects with four or more refractions between
the ages of 6 and 15. An index of age of myopia onset was determined by extrapolation
to zero refractive error. An index of the final amount of myopia after childhood myopia
progression was derived from the mean of the amount of myopia found at examinations
after age 17. The coefficients of correlation of onset age and final amount of myopia were
0.42 for 1-pales In = 49), and 0.61 for females In = 31~. Both of these correlations were
statistically significant at the 0.01 level.
These studies indicate that myopia continues to increase in prevalence throughout the
school-age years, that girls have earlier onset ages, and that the earlier myopia appears, the
more likely it will progress to a high degree.
Rates of Progression
For Hirsch's (1964a) nonvisually selected sample, the average rate of refractive error
change was -0.07 D./yr. (SD = 0.09~. He stated that rate of change for myopes was higher
than for the rest of the sample. Most of the reports in which rates of childhood myopia
progression are given are studies of various treatment modalities tested for their ability to
alter the course of myopia development (Goss, 1982~. Control group data can be used to get
an idea of common rates. In the United States, mean rates of childhood myopia progression
were in the neighborhood of -0.30 to -0.60 D./yr. (Nolan, 1964; Roberts and Banford,
1967; Baldwin et al., 1969; Oakley and Young, 1975; Goss, 1984), whereas in Japan mean
rates were in the area of -0.50 to -0.80 D./yr. (Tokoro and Kabe, 1964, 1965; Matsuo,
1965; Otsuka, 1967~.
Mantyjarvi (1985b) provided a comparison of rates of refractive change for 46 hyperopes
(32 girls and 14 boys) and 133 myopes (75 girls and 58 boys), who were followed for
observation times varying between 5 and 8 years, up to age 15. Refractions were spherical
equivalents in the right eye under 1 percent cyclopentolate cycloplegia. The children were
seen annually or biennially in most cases. They were examined at the Kuopio, Finland,
Community Health Center following a screening by school doctors and nurses. The mean
(+ SD) annual rate of change for the hyperopes was -0.12 D./yr. (+ 0.14), with a range
of +0.11 to -0.45 D./yr. For the myopes, the mean (+ SD) was -0.55 D./yr. (+ 0.27),
with a range of 0.00 to -1.63 D./yr. Mantyjarvi (1985b) also reported on 30 schoolchildren
who were initially hyperopic but became myopic during the observation time. The mean
(+ SD) rate while they were hyperopic was -.021 D./yr. (+ 0.21), with a range of +0.25
to -0.75 D./yr. The mean (+ SD) rate while they were myopic was -0.60 D./yr. (+ 0.45),
with a range of-0.08 to-1.63 D./yr. It would appear that the rate of r~fra~.tiv~ John
accelerates about the time a child becomes myopic.
· O
Goss and Cox (1985) reported on longitudinal records of 559 myopes from five optometry
practices collected on the basis of the following criteria: (1) at least four examinations
OCR for page 69
69
TABLE C-3 Frequency Distribution of Rates of Childhood Myopia Progression
for Patients from Five Optometry Practices in the Central United States
Range of Myopia (D.) Males (N) Females (N)
+ 0.20 to 0.00 ~0 4
-0.01 to 0.20 37 20
-0.21 to -0.40 51 48
-0.41 to -0.60 41 44
-0.61 to -0.80 15 23
-0.81 to -1.00 12 7
-1.01 to -1.20 2 1
-1.21 to -1.40 0 0
-1.41 to -1.60 0 1
SOURCE: Goss and Cox (1985~.
between ages 6 and 24, (2) myopia of at least 0.50 D. sometime during the course of the
clinical record, (3) astigmatism never manifested in excess of 2.50 D., (4) no strabismus
or amblyopia, (5) no contact lens wear prior to the last refractive data recorded for use
In this Study iu~, no ocular ~atholozY~ and (71 no systemic pathology that might affect
. . . · . . , ~ .
1 ~ 1- 1 · ·1 1- 1 ~ ~ 1 ~] · ~1 ~ ~ · ~ ~ 1
ocular Endings, such as Juvenlle diabetes or other conditions. the retractive data used were
refractive errors in the principal meridian nearest horizontal in the right eye as derived
from the manifest subjective refraction recordings. Rates of childhood myopia progression
in diopters per year were calculated by linear regression, using points at or before age 15
in those cases in which four or more refractions were recorded during that age span. Rates
were thus determined for 158 mates and 145 females. The mean rates of progression were
-0.40 D./yr. (SD = 0.24; range = -0.01 to -1.09 D./yr.) for males, and -0.43 D./yr.
(SD = 0.25; range = +0.12 to -1.52 D./yr.) for females. The difference in these means
was not statistically significant at the 0.05 level. A frequency distribution of the rates is
shown in Table C-3. Most rates fell within the -0.21 to -0.40 D./yr. and -0.41 to -0.60
D./yr. ranges. While there was considerable individual variation, positive rates (indicating
a decrease in myopia) and rates more negative than -1.00 D./yr. were uncommon.
Cessation Age
Ophthalmic clinicians have long recognized that childhood myopia progression stops
or slows down in the middle to late teens. Bucklers (1953), for instance, noted this from
OCR for page 70
70
examination of patient records from his practice. The average annual refractive changes in
the data of Brown (1938, 1942) and Slataper (1950), which also included emmetropes and
hyperopes, were less negative after about age 16.
Goss and Winkler (1983) investigated this on a quantitative basis. They selected 299
longitudinal records of myopic patients from three optometry practices. These were the first
299 records collected in the larger sample of Goss and Cox (1985~. The refractive data were
manifest subjective refractions. Examples of plots of myopia progression in individuals from
this sample are shown in Figure C-2. The mode! used by Goss and Winkler was one of two
straight lines, one through the points during the time when childhood myopia progression
was occurring and another through points at later ages, with the cessation age being the
point at which the two lines met. On the bash of this model, they determined cessation
age by four different graphical and statistical methods. The mean cessation age for females
was earlier than that for males, the difference generally being about one year. While the
most common cessation ages were in the middle teens (about age 15 for females and age 16
for males), considerable individual variation we nOt.P~ with ~t.:`nA~rr] A-`riati~r~c! Of -
two years.
Goss and Cox did not find a statistically significant correlation between onset age, as
derived by the method described above, and cessation age, as derived by Goss and Winkler's
method 2. The coefficients of correlation were 0.07 for males (n = 49), and 0.31 for females
(n = 31~. This would imply that the duration of childhood myopia progression is not
constant from one individual to another.
The cessation age could be viewed as a point of transition from the generally rapid
childhood myopia progression to a period of relatively smaller or negligible increases in
myopia. The increases in myopia that do occur after the cessation age, especially after ages
1~20, are referred to in this appendix as young adult myopia progression.
~ ~ ~ ·, ~ v^^ ~ ~ In a ~ ~ ~ AW HAVE- Vat ~ V ~
YOUNG ADULT MYOPIA
Variables Affectmg Progression
Numerous factors undoubtedly affect the progression of myopia. These may include
growth, health, nutrition, personality, race, ethnic heritage, hereditary factors, near work,
and education (Baldwin, 1964, 1981; Goldschmidt, 1968; Borish, 1970; Curtin, 1970; Sorsby,
1979; Angle and Wissmann, 1980a). Studies are needed to effectively and systematically
investigate these factors. One variable that is known to affect progression is gender (WeaTe,
1983~. As shown above, females have earlier onset age and earlier cessation age than males.
Astigmatism may be a factor in myopia progression (Baldwin, 1957; Hirsch, 1964b; Fulton
et al., 1982~. It has also been noted that, among strabismics and amblyopes, refractive
changes are greater in the fixating eye than in the deviating eye (Lepard, 1975; Leffertstra,
1977; Bielik et al., 1978; Nastri et al., 1984~.
Numerous attempts made to control the progression of myopia have been reviewed by
Baldwin (1967), Borish (1970), Grosvenor (1980,1982), Goss (1982), and Grosvenor et al.
(1987~. No method has been shown to be consistently and universally effective. Apparent
success in reducing rates of myopia progression in some reports could possibly be attributed
to investigator bias, inadequate experimental design, poor matching of experimental and
control groups, and numerous uncontrolled variables. The most common treatment modal-
ities have been rigid contact lenses, cycloplegic drugs, and bifocal lenses. The success of
rigid contact lenses in controlling myopia progression can be attributed to flattening of the
cornea, which causes reduction in the refractive power of the eye. In the only study in which
OCR for page 71
71
-8.00
-7.00
-6.00
-5.00
o
-3.00
cr
I
-4.00
-2.00
-1 .00
0.00
1.00
2.00
-8.00
-7.00
-6.00
-5.00
o
- -4.00
LIJ
cr
T
-3.00
-2.00
-1 .00
0.00
1.00
2.00
A
F
,~
1
-
1 1 1 1
1 1
4 6
r
B
_ ~=
of
8 10
12 14 16 18 20 22 24
AGE (yrs)
~1
4 6 8 10
12 14 16 18 20 22 24
AGE (yrs)
FIGURE C-2 Examples of myopia progression in five males (A) and five females (B) from north central
United States private practice records: plotted on the abscissa is age in years, and on the ordinate is
refractive error in the principal meridian nearest horizontal in the right eye. Each set of common symbols
represents one individual.
Source: Adapted from Goss and Winkler, 1983.
OCR for page 72
72
corneal curvature changes were controlled (Baldwin et al., 1969), refractive error changes
and axial length increases in rigid contact lens wearers and spectacle lens wearers were sim-
ilar. Studies with cycloplegic drugs have not considered the effect of long-term reduction
in ciliary muscle tonus on the results. That is, apparent reductions in rate may represent
continued reduction in baseline ciliary tonus (Sato, 1957) rather than prevention of axial
elongation. The disadvantages of long-term application of cycloplegics include compliance
and their substantial visual and potentially systemic side effects. The results of studies on
bifocals have varied widely. Potential explanations for this, in addition to those mentioned
above, include the manner in which the bifocals were employed and the types of cases in
which they were used. It is possible, for instance, that bifocals may be effective in the types
of cases in which they reduce asthenopia, i.e., high lag of accommodation (accommoda-
tive response lags behind the accommodative stimulus) (Roberts and Banford, 1967; Goss,
1986a) and near-point esophoria.
Typical Reilacti~re Changes in Young A(lults
The majority of pertinent studies in this area deal with selected samples. Brown
(1942), in his study compiled from cycloplegic refractions in a private practice, found mean
annual refractive error changes that were negative in sign from ages 6 to 34. In STataper's
(1950) series, they were negative from ages 7 to 31. The magnitude of the average changes
were much less after about age 18 or 20 than they were in childhood (see Table C-4.
In Hofstetter's (1954) sample of private practice patients in Bloomington, Indiana, most
persons in the 21- to 34-year-old age range showed little or no change in refractive error.
Most of the myopes who did change became more myopic, and most of the hyperopes who
changed became more hyperopic. While stable refraction was the rule for emmetropes,
changes in both directions did occur. Since the data in Hofstetter's study were based on
manifest refractions, some of the shifts toward hv~eronia were nrnhahlv H~ to. mz..nif~c~.~.imn
of previously latent hyperopia.
~ ~rat 1 fitly_ ~ n~-rt ~ ^~~
,, ~7 ,¢ ~< ~^~ ~ _ ~ ~ ~ ~ ^~
wrosvenor t'Y``a, I, ~Y-~iCJ conducted a mall survey of male optometrists' own
spectacle corrections from ages 20 to 40. He asked only for the spectacle correction, not
the refractive error. Analysis was based on the most positive meridian of the right eye. At
age 20, 59 (53.2 percent) of the subjects were in the emmetropic group (piano to +0.87
spectacle correction); at age 40, 44 (39.6 percent) of the subjects were in this group. The
hyperopic group (+1.00 and over) had a net gain of 10 subjects, and the myopic group
(any amount of negative power spectacle correction) had a net gain of 5 subjects during the
20-year period. For the 59 subjects originally in the emmetropic group, 24 had no change
in spectacle correction over the 20 years, 22 had a change in the positive direction (toward
hyperopia), and 13 had a negative change, the maximum being -1.50 D. For the 41 subjects
originally in the myopic group, 12 had no change in spectacle correction, 4 had a positive
change, and 25 had a negative change, the maximum being -2.00 D. For the 11 subjects
originally in the hyperopic group, 4 had no change, 6 had a positive change, and 1 had a
negative change. The mean spectacle correction was -0.08 D. (SD = 1.47) at age 20 and
-0.18 D. (SD = 1.92) at age 40.
Morgan (1958, 1960) presented a longitudinal study in California of individuals seen
20 years apart, first at about age 13 in 1934, and later at about age 33 in 1954. In
all, 51 females and 44 males consented to return 20 years after the study done on about
150 students entering seventh grade in two junior high schools. Vertical and horizontal
meridians from the subjective refraction were both used for analysis. The mean change in
refractive error was negative in sign. The number of emmetropes decreased over the 20-year
OCR for page 78
78
rate of about -.024 D./yr. The maximum change recorded was -2.12 D. and the minimum
was -0.25 D. There were 3 eyes that changed in the hyperopic direction. There were 363
eyes in the category of having spectacle lenses both at entry and at the beginning of the
fourth year; of these, 348 were myopic. There were 270 of the 363 eyes that had a change in
refractive error, the average change being -0.55 D. (SD = 0.69), which would represent a
rate of about -0.18 D./yr. The maximum myopic shift was -3.25 D., while the minimum
was -0.12 D. There were 26 eyes with a change toward hyperopia. The average amount of
myopic shift was greater for those who did not wear spectacles at entry (-0.72 D.) than
for those who did (-0.55 D.~. Brown also pointed out that a -5.50 D. eligibility limitation
for entry to West Point appears to be reasonable, because it would be unlikely for a 5.50
D. myope to develop 8 D. of myopia (the limit for the commissioning of an officer) by the
beginning of the fourth year.
In an unpublished study of the U.S. Air Force Academy class of 1980, Goodson (1983)
compared the spherical equivalent refractive error taken at the beginning of the senior
year with that indicated on the students' entering medical records. Data were spherical
equivalents of the worst eye. Of 914 records, 225 reported refractive error, in addition to or
instead of acuity. Of these, 196 or 76.9 percent showed a myopic shift of at least -0.25 D.
Over the three-year period, with a mean change of -0.83 D. in those who did have a myopic
shift. Of the 86 cadets whose entrance refractive error fell within the acceptance region
for Flying Class ~ (i.e., not exceeding +1.75 D. or -0.25 D.) 40 (46.5 percent) remained
pilot-qualified on this standard.
A more recent unpublished study by O 'Neal et al. (1986) was based on the Air Force
Academy medical records of the class of 1985, of whom 89.3 percent were men. The
data were spherical equivalent refractions of 497 records reported in terms of numbers of
eyes. Omitted from the total possible number of class records (i.e., 944) were those absent
on the day of data collection, those indicating use of near glasses or contact lenses, and
incomplete entries. Entry-leve! refractions were performed by the individual's private vision
specialist, and many or most probably were not taken under cycloplegia, but all those taken
approximately two years later at the beginning of the third year of study were done by
military optometrists and were taken after instillation of two drops of cyclopentolate. At
entrance, 105 eyes (10.6 percent) were emmetropic, 409 (41.1 percent) were hyperopic, and
480 (48.3 percent) were myopic. Emmetropia was defined as a spherical equivalent of zero.
At entry, 563 eyes (56.6 percent) fell in the range of +.050 to -0.50 D. Of the total number
of eyes (994), 66.7 percent shifted in the myopic direction of -0.60 D. on average. Of the 514
emmetropic and hyperopic eyes on the entering record, 282 (54.9 percent) showed a shift in
the myopic direction of at least -0.12 D. Of the 480 myopic eyes at entry, 381 (79.4 percent)
progressed in the myopic direction of at least -0.12 D. On average over the two-year period,
the change in refraction equalled -0.18 (+ 0.42) D. for the emmetropic eyes, -0.18 (+ 0.37)
D. for the hyperopic eyes, -0.54 (+ 0.62) D. for the myopic eyes, and -0.34 (+ 0.53) D. for
the total. At the time of the second exam, 32.8 percent of eyes were hyperopic, 10.9 percent
were emmetropic, and 56.3 percent were myopic. For persons having a shift toward myopia,
the average shift was -0.60 D. Of the 612 eyes whose entering refractions fell within the
Flying Class ~ standard, 506 (82.7 percent) remained pilot-qualified.
Provines et al. (1983) reported a survey of 1,105 navigators and 1,295 pilots stratified
by major air commands. The study provided a random sample of the number of pilots and
navigators required to wear corrective lenses and of initial and current spherical equivalent
refractive errors. The study included persons who had entered undergraduate training at
ages 20 to 25 and who had server} on active duty for 20 years or less. Myopia was defined
as a spherical equivalent refractive error of more than -0.25 D. In analyzing the number of
OCR for page 79
79
TABLE C-6 Pilots and Navigators Who Developed More Than 0.25 D. of Myopia by
Various Initial Refractive Error Levels
Did Did Become
Initial spherical Not Become Myopic (0.25D.
equivalent Myopic (0.25D. or more)
refractive error N or more)
Pilots:
-0.25 to piano 248 205 43 (17.3~o)
+0.12to+0.37 251 231 20(8.0)
+O.SOto+0.75 193 191 2(1.0)
> to +.87 100 100 0 ( 0.0)
Total 792 727 65 (8.2)
Navigators:
-0.25 to piano 153 109 44 (28.8)
+0.12to +0.37 142 132 10(7.0)
+ 0.50 to + 0.75 119 112 7 ( 5.9)
+0.87to+1.12 31 29 2(6.5)
> +1.25 16 16 0(0.0)
Total 461 398 63 (13~7)
SOURCE: Provines et al. (1983).
persons initially hyperopic who became myopic, three trends are apparent: (1) navigators
were more likely to become myopic: 65 (8.2 percent) of 792 pilots became myopic, compared
with 63 (13.7 percent) of 461 navigators. (2) The incidence was more likely in persons who
had greater time in service. For example, among persons with 1 to 5 years of service, 2.4
percent of pilots and 7.8 percent of navigators had become myopic. This can be compared
with 17.6 percent of pilots and 24.7 percent of navigators who became myopic during 16 to
20 years of service. (3) As Table C-6 indicates, both pilots and navigators were more likely
to develop myopia if the initial refraction was near piano. The highest incidences were for
persons with initial refractions of -0.25 D. to piano.
Shotwell (1981, 1984) has studied the effects of reading lens prescription on refractive
error changes in young adults. In a 1981 report, he randomly assigned 232 students in the
OCR for page 80
80
Naval Academy Preparatory School (ages 18 to 20) to one of three reading prescription
groups: (1) distance correction with no. 1 pink tint (placebo group); (2) +1.25 D. added to
the distance correction with two prism diopters base-in over each eye; (3) bifocals with a
+1.50 D. added. There was a very high rate of attrition in the study: 100 subjects did not
wear their spectacles when reading and were not included in the analysis, and 44 subjects
disenrolled or were not available for retest. Only persons with refractive errors between
+1.00 D. and -1.00 D. initially were used in the statistical analysis. This left 21 subjects in
the placebo group, 20 in the plus and prism group, and 19 in the bifocal group. The mean
changes in each group were not significantly different from zero.
Shotwell followed 61 of these students for an additional four years spent at the U.S.
Naval Academy. The primary cause of the attrition was disenroliment at the Academy (116
subjects); other causes were failure to wear the lenses full time for reading (40 subjects) and
voluntary withdrawal (18 subjects). These 61 subjects had initial refractive errors between
-0.37 D. and +0.75 D. (mean = +0.22; SD = 0.43~. The subjects were predorn~nantly
white with a mean age at the start of the study of 18.7 years (SD = 1.3; range = 17 to 21
years). The three types of lens prescriptions were the same as in the 1981 study. Pre- and
post-test refractions were done using 1 drop of 1 percent tropicamide, with the examiner
blind to the subject's group assignment. The mean refractive changes in four years were:
(a) placebo group, -0.23 D. in = 21; SD = 0.28~; (b) plus with prism group, -0.27 D. in
= 27; SD = 0.27~; (c) bifocal group, -0.15 D. in = 13; SD = 0.37~. These means were
not significantly different. The control group had a mean rate of change that was about
-0.06 D./yr. Shotwell also divided the 21 subjects in the control or placebo group into 10
in "high-risk" groups and 10 in "Iow-risk" groups according to six theoretical risk factors
for changes toward myopia: number of family members wearing glasses, near-point cover
test phoria, cup-to-disk ratio, reading distance, reading style (deliberate versus fast), and
time spent on near work versus time spent in outdoor activities. Of the six risk factors, only
the time spent on near work versus time spent outdoors showed a statistically significant
difference in mean refractive change between the high-risk and low-risk groups. While this
study was well designed and executed, the high attrition rate is a considerable limitation.
Diamond (1957) reported a study of 67 airline pilots whom he followed for employment
periods of time ranging from 5 to 18 years, all of whom had 20/20 unaided visual acuity at the
time of employment. For the 16 pilots who became myopic enough to reduce unaided visual
acuity or worse, the ranges of ages and refractive errors at the beginning of employment
were 21 to 31 years and -0.25 to +0.25 D. The spherical equivalent refractive error shifts
ranged from -0.25 to -1.25 D. (mean, -0.65 D.) over periods of time that varied from 5 to
17 years (mean, 12.3 years). Diamond stated that the development of the myopia was slow
and gradual over the years of observation. The average annual change was -0.05 D. for the
16 who became myopic. Individual myopic shifts as a function of initial age and length of
employment are depicted in Figure C-3. Although the number of persons in each initial age
category was not over four, the mean rates of change were similar in each category. The 51
pilots who did not move into myopia were 20 to 30 years old initially, had refractive errors
of -0.25 to +2.25 at the beginning of their employment, and were followed for 9 to 18 years.
Dunphy et al. (1968) discussed refractive error changes in 100 graduate students ages
20 to 30 (200 eyes) enrolled in the Harvard business and law schools. At annual visits,
cycloplegic refractions were done 15 minutes after the instillation of 2 drops of 1 percent
tropicamide five minutes apart. Refractive errors wars ~.n~.lv~.~1 in t.-rme ~fer~h~r;~]
_ ~ _ _ ~_ ~1 . ~
~ ^wv^, _ art ~1~ abut ~ 111 b~1-111~ U1 ~ll~[lCa1
equivalents. In most cases' data was expressed in terms of number of eyes rather than
Suer of persons. - ~ -ne study group of law students had a greater proportion of students
with uncorrected acuity of 20/20 or better than their entire class (more than 50 percent
OCR for page 81
81
+0.25
+0.12
,O
LL-0.12
~-0.25
a)
Q-0.37
co
~ -0.50
._
cn -0.62
._
Q
o
>a
0
a)
a)
a)
-1 .25
Age Group
-0.049
Overall Av.
_ .
-0.75 .
-0.87
.-1 .00
-1 .12 .
*14
~ .
14
~ N::
- *Numbers refit r to lengt ~ of duty (years)
21 ~221 23 1 24 ~ 25 26 27 28 2c
-0.036 1 1 -0.048 1 -0.024 1 -o. 111 -0.044 -0.048 -0.048
Averane Annual Shift
~9
17
13
12
27
.
-0.048
17
17
28
.
-0.048
12
. 30
-0.041
15 10
31
-0.045
FIGURE C-3 Refractive error changes in individual pilots who developed myopia according to age at
beginning of employment and length of employment. The brackets indicate the two eyes of one individual.
Beginning and ending spherical equivalent refractive errors are represented by dots.
Source: Adapted from Diamond, 1957.
compared with 39 percent). The changes in one year and the corresponding numbers of eyes
were: +0.50 to +0.25 D., 29 eyes; +0.125 to -0.125 D., 89 eyes; -0.25 to-0.50 D., 73 eyes;
and -0.625 to -0.875 D., 9 eyes. All 9 eyes in this greater negative change category were
myopic (amount unspecified) at the beginning of the study. Of 40 law students followed
for a second year, 40 of the 80 eyes changed -0.25 to -1.50 D. in two years. Of these 40
students, the 24 students with initial refractions of -0.125 to +1.50 D. had a mean change
in two years of -0.12 D., 11 students with -0.75 to -0.25 D. refractions averaged -0.47 D.
change in two years, and 5 students with -5.25 to -2.00 D. refractions averaged -0.01 D.
change.
Kent (1963) presented six case reports of myopia onset at age 18 or later. Some of
the recorded refractions were done under cycloplegia and some were not. The means of
the spherical equivalents of the two eyes were used for analysis. Excluding one case due to
lack of complete data, the rates of refractive error change of the remaining five varied from
-0.04 to -0.16 D./yr. over time spans varying from 9 to 26 years. The myopia developed
was all low in degree and generally developed over a period of several years. In another
case, keratometer measurements were taken on several occasions. There were increases in
spherical equivalent keratometer powers of 0.58 D. in the right eye and 0.65 D. in the left
eye, corresponding to spherical equivalent refractive error changes of -0.94 D. in the right
eye and -0.75 D. in the left eye. It would appear that this subject's myopia developed
largely as a result of an increase in corneal power. Kent stated that three subjects first
manifested myopia while they were students, but that most of their myopia developed after
completing school, and that three other subjects "were not engaged in activities requiring
OCR for page 82
82
much close work when their myopia developed." He also noted that two subjects in the first
group and one in the other each observed transient distance blur occurring after prolonged
nearwork just prior to their myopia onset. This study is notable in that it is apparently the
first to report changes in an ocular component in adult myopia progression.
Goss et al. (1985) studied young adult myopia progression the same optometry practice
population of 559 subjects used by Goss and Cox (1985~. The refractive error data consisted
of the refractive error in the principal meridian nearest horizontal in the right eye. Goss
et al. reported three analyses: (1) subjective categorization of patterns of refractive error
change typical of young adulthood in myopes; (2) rates of refractive error change; and
(3) changes in keratometer readings. There was a sufficient number of data points for
subjective categorization of young adult patterns for 63 males and 53 females. The majority
of subjects, 68.3 percent of males and 86.8 percent of females, fit the pattern they termed
adult stabilization, characterized by rapid increase in myopia in childhood and adolescence,
followed by stabilization in adulthood. Adult continuation was found in 25.4 percent of
the males and 13.2 percent of the females in the sample. It is characterized by myopia
progression in adulthood, but at a slower rate than that during childhood. The least
common pattern (6.3 percent of males and none of the females) was adult acceleration, in
which myopia progression accelerates in adulthood. In this category they included cases
both of acceleration of existing myopia and the adult onset of myopia. It may be noted
that patterns with refractive changes toward myopia were more common among males
(see Figure Cab. They calculated rates of young adult myopia progression, using linear
regression slopes (D./yr.) for the 57 cases with 3 or more data points at age 20 and later,
and separately for the 108 cases with 3 or more data points at age 18 and later. Most of
the records ended in the middle 20s, but some continued to the middle 30s. The mean
rates for age 20 and over were -0.07 D./yr. (SD = 0.09) for males and -0.03 D./yr. (SD ~
0.09) for females, while for age 18 and over they were -0.08 D./yr. (SD = 0.~) for males
and -0.02 D./yr. (SD = 0.06) for females. The difference in mean rates for males and
females was statistically significant for both age criteria (p ~ 0.05 for age {S and over and
p ~ 0.01 for age 20 and over). The rates for age 18 and older ranged from +0.19 to -0.25
D./yr. for females and from +0.09 to -0.36 D./yr. for males. Low negative rates appeared
to be most common. For 11 subjects in the adult stabilization and adult acceleration
categories and 20 subjects in the adult continuation category, Goss et al. (1985) had data
on three or more refractions and three or more keratometer measurements recorded at age
18 and beyond. For the keratometer readings, like the refractions, the principal meridian
nearest horizontal in the right eye was used for analysis, and rates were calculated by linear
regression analysis. For all 11 of the adult continuation and adult acceleration subjects with
three or more keratometer readings, there was a reduction in the corneal radius (range of
rates, -0.003 to -0.042 mm/yr.), which would cause a shift toward myopia. The coefficients
of correlation of corneal radius with refractive error for these 11 subjects varied from +0.12
to +0.98, based on an average of only 3.9 points per subject. For the 20 adult stabilization
subjects, there were 11 increases in corneal radius and 8 decreases, most of them small
in amount and with rates ranging from +0.014 to -0.036 mm./yr. Data on other ocular
dioptric components for this sample were not available.
Goss and Erickson (1987) reported a further analysis of the relation of refractive error
and corneal power changes in their samDIe to ask the contrih,,~.i~n of rr~rn~z`1 char tm
~,# ~ ___ Vie ~ V ~;~O MU
· _
myopia progression. for analysis they selected the records with three or more refractions
and three or more keratometer readings at age 18 or later: 22 males and 15 females
met this criterion. Right eyes were used for analysis. Most records stopped at about
age 25. Rates of refractive error change and rates of keratometer power change were
OCR for page 83
83
-7.00 _
-6.00
-5.00
O -4.00
-
G
N
G
o
I
-3.00
-2.00
-1 .00
0.00
1.00
2.00
-
9 12
-
1 1 1 ~ 1 1 ' 1 1 1 1
15 1 8 21 24 27 30 33 36
AGE IN YEARS
3 6
FIGURE C-4 Examples of young adult myopia progression patterns based on the classification system
of Goss et al., 1985. Each set of common symbols represents refractive data for one subject: A-adult
continuation; B adult stabilization; C adult acceleration.
Source: Adapted from Goes and Cox, 1985.
calculated by linear regression. Statistically significant correlations were found between
rate of refractive change and rate of keratometer power change for young adult myopia
progression. (Significant correlations were not found for childhood myopia progression.)
The regression slopes of refractive error change (y) on keratometer power change (x) were
around 0.7. This slope is close to the 0.68 slope found by Erickson and Thorn (1977) for
corneal changes in orthokeratology, and it indicates that across the sample, about two-thirds
of a diopter change in keratometer power results in one diopter shift in refraction. Non-zero
y-intercepts in the regression equations may be due to the effects of other components or to
corneal changes not measured by the keratometer (Erikson, 1978~. These findings indicate
that the cornea can play some role in the progression of myopia in young adults, but the
contribution of other ocular components cannot be ruled out.
Adams (1987) reported a case in which adult myopia progression was apparently not
primarily due to corneal steepening. From ages 24 to 42 the manifest refraction showed
changes of -3.75 D. in each eye in the principal meridians nearest horizontal and -3.25
D. in each eye in the principal meridians nearest vertical. The corresponding changes in
keratometer power were: horizontal, +0.75 D. in the right eye and +1.00 D. in the left eye;
vertical, +0.25 D. in the right eye and +0.37 D. in the left eye. In unpublished theses, Schell
et al. (1986) reported that most of the adult myopic shift found in a group of optometry
students could be accounted for by axial length changes, and McBrien (1986) found that
axial elongation accounted for myopic changes in his young adult sample.
On the basis of observations that the rates of increase of myopia typical of youth
are different from those in adulthood and that the mechanisms associated with each may
be different, Goss has proposed that different terminology should be applied to the two
phenomena. He and his colleagues have referred to the increases in myopia preceding the
middle to late teens as childhood myopia progression and any increases subsequent to
that as young adulthood myopia progression. After the cessation age of childhood myopia
OCR for page 84
84
progression (Goss and Winkler, 1983), refraction may either be stable or may show young
adult myopia progression. Some emmetropic young adults also have myopia appear with
subsequent young adult myopia progression. As discussed in the next section, the rates of
adult myopia progression following adult onset are different from those of childhood myopia
progression.
CHANGES IN DIOPTRIC COMPONENTS O1? THE EYE
Sorsby et al. (1961) presented both cros~sectional and longitudinal data. In the cross-
sectional study were 1,432 children (671 boys and 761 giris) ages 3 to 15 drawn from
London County Council schools and day nurseries in South London. About 30 to 40 percent
of those contacted volunteered for the study. Refractive error was determined by retinoscopy
under scopolamine cycloplegia, and corneal power by keratometer. Anterior chamber depth
and lens thickness were calculated from slit lamp cross-sectional photographs. Radii of
curvature of the crystalline lens were derived from comparison phakometry. Axial length
was calculated using the above information. Over the age span studied (ages 3 to 14 in boys
and ages 3 to 15 in giris), the mean vertical meridian refractive error shifted toward less
hyperopia by 1.40 D. in boys and 2.30 D. in gird. Between ages 3 and 13, mean axial length
increased about 1 mm for both boys and girls, and mean crystalline lens power decreased a
little less than 2 D. Changes in mean corneal power were slight.
The longitudinal study in the Sorsby et al. monograph consisted of 440 subjects,
who were reexamined 2 to 6 years after the initial test; 54 cases were excluded due to
inconsistencies in findings, leaving 386 cases (183 boys and 203 girls). In 135 cases (34.9
percent) axial length increased 0.0 or 0.1 per year. In these cases, refractive errors and
cornea and {ens also tended to be close to stationary. In 195 cases (50.5 percent) there was
a yearly axial length increase of 0.2 or 0.3 mm. Sorsby et al. noted that, if uncompensated
by other components, this would cause a shift toward myopia of 0.6 to 0.8 D. Only 13 of
these 195 (6.7 percent) had refractive changes of this order. Most of the children had full
or partial compensation for this axial elongation by reductions in corneal or lens power.
The remaining 56 cases (14.5 percent) showed an annual axial elongation of 0.4 mm or
more. There was considerable variability in the corresponding changes in refractive error:
12 of the 56 (21.4 percent) showed adequate compensation (refractive error changes of 0.0
to 0.2 D./yr.~; 24 of the 56 (42.8 percent) had annual refractive changes of 0.3 to 0.5
D; and 20 of the 56 (35.7 percent] shifted toward mvoDia bv n 6 Or more nor ~r~:`r lifer
~ ~ - ~ ~ ~ -A-Rae- ~ ~ ~- ~^ ^~^ ~ e-~ ~ ~ . a- V1
those children first examined at younger than 10 years of age, the numbers with the higher
and lower levels of axial elongation were approximately equal for all beginning refraction
categories (myopia, emmetropia, and mild and severe hyperopia), but higher amounts of
axial elongation predominated for children who became myopic. For subjects first tested at
10 years of age or older, there were proportionately more cases of higher axial elongation
among those who were myopic or became myopic. That is, among the older children, 10
of 83 (12 percent) of the hyperopes showed marked axial elongation, compared with 15 of
43 (34.9 percent) of the emmetropes, 6 of 14 (42.9 percent) of the myopes, and 7 of the 17
(41.2 percent) who became myopic.
In a later monograph, Sorsby and Leary (1970) discussed follow-up observations of 129
children in the 1961 study, using the same examination methods: 68 of these children had
the follow-up examination at or before age 14. They divided these 68 cases into 49 with
slight shift toward myopia (< 1.31 D. in an average of 8.25 years) and 19 with greater shifts
(> 1.51 D.~. The group with higher refractive shifts showed greater axial elongation but
similar amounts of corneal and lens power decrease. In general, the greater the axial
OCR for page 85
85
elongation, the lower the corneal and lens power. However, adequate compensation does
not occur in some subjects, so that the mean refractive shift toward myopia increases as the
amount of axial elongation increases. There was great individual variability in refractive
error and component changes. The work by Sorsby and his colleagues suggests that myopia
appears and childhood myopia progression occurs most often when axial elongation is greater
than usual or when average amounts of axial elongation are not adequately compensated
for by corneal and lens power reductions.
Tokoro and Suzuki (1968, 1969) measured the refractive components of 56 eyes over the
time span of 1960 to 1967: 33 eyes in 18 subjects were examined four times, and 23 eyes in
13 subjects were examined two or three times. The subjects ranged in age from 7 to 21 in
1960. Refraction was determined 24 hours after the instillation of two drops of 1 percent
atropine. The radius of curvature of the anterior surface of the cornea was measured by
ophthalmometry. Anterior chamber depth and crystalline lens thickness were determined
by Jaeger's apparatus (Duke-Elder and Abrams, 1970~. Radii of curvature of the crystalline
lens were found by photographic phakometry. Axial length was determined by calculation
and by ultrasonography, but it was unclear which was used in the analysis. Refractive index
and power of the crystalline lens were derived by calculation.
Tokoro and Suzuki found significant correlations of refraction, lens power, and axial
length at the initial examination with the annual change-in refraction. They also calculated
average changes in refractive components according to age (see Figure C-5. There was
little change in corneal power at any age. There was some reduction in lens power up to age
11 or 12, which would partially compensate for the axial elongation that occurs at this time.
Most obvious ~ an increase in axial length. This axial elongation is greatly diminished after
age 17 or 18. Plots by Tokoro and Suzuki of individual change in refractive error with age
and plots of axial length versus age show remarkably similar patterns of increases until the
middle or late teens, followed by a plateau or period of led change. Such a comparison might
lead one to suggest that the cessation of childhood myopia progression could correspond to
the cessation of axial elongation of the globe.
In 1971, Larsen reported a major cross-sectional study of the ocular components in
children conducted in Norway (Larsen 1971a, 1971b,1971c): 80 full-term newborns (43 boys,
37 girls) were examined with 0.2 percent oxibuprocaine or tetracaine without cycloplegia.
For children of ages 6 months to 7 years, a general anaesthetic was used with cycloplegia
(1 percent cyclopentolate HCL). In the age groups 8 to 13 years, surface analgesia was
used with c~cloole~ia. There were 465 bays and 381 girls. excluding newborns. Larsen
_ ~_
_ ~ . . ., , . ., ~ ~ . ~ . . ~ ~ ~ . ~ ~ .
t1971a' points to three phases In the development ot anterior chamber depth dimensions:
(1) rapid postnatal growth phase from birth to one and a half years, with anterior chamber
depth increases of approximately 0.9 to 1.0 mm in both sexes; (2) a slower infantile growth
phase from 1 to 7 years, with increases of approximately 0.3-0.4 mm; and (3) a slow
juvenile growth phase from 8 to 13 years, with an increase of 0.1 mm. No difference was
found between the values for the anterior chamber depth at age 13 and the values for
20 emmetropic adults in the age group 20 to 40 years. Negative correlations of anterior
chamber depth and refractive error (deeper anterior chamber in myopes) were found from
the second year of life onward. For 76 eyes of 12-year-old girls, this correlation was r =
-0.88 (Larsen 1971a).
In 80 newborns, Larsen (1971b) found that the mean value of the lens thickness taken
with cycloplegia was 3.93 mm in 43 boys and 3.99 mm in 37 girls. Lens thickness decreased
to 3.36 mm in 12 boys and 3.45 mm in 24 girls in the 13-year-old age group. The drop
was consistent through the entire age range. For both boys and girls, since the drop was
obtained with the application of cycloplegia in the age range 1 to 13, it was not an artifact
OCR for page 86
86
D
0.4
03 08
0.2 0.6
0 1 04
0.2
O O
-0.2
-0.4
1.41
1.40
Annual change of
R: refraction
1
R
I : axial length
Dc: refractive power of the cornea
DL: refractive power of the lens
n2: refractive index of the lens
10 15 20 25 Age
FIGURE C-5 Average annual changes in refractive components with age.
Source: Adapted from Tokoro and Suzuki, 1969.
of the absence of cycloplegia in the newborns. No correlation was found between refractive
status and lens thickness or between lens thickness and anterior chamber depth. Larsen
(1971c) also reported that vitreous depth averaged 10.48 mrn and 10.22 mm in the male
and female newborns, respectively. Comparable values at one year were 13.62 and 13.24,
showing major growth during this period. For years 10 to 13, mates averaged 15.62 mm,
females 15.31 mm. These levels closely approximated those of 10 emmetropic men (15.44
mm) and 10 emmetropic women (15.18 mm). The relation between the ametropias and
vitreous length was reported for the 1- to 3-year-old age group, separately for boys and girls.
For the 20 male myopic, 86 emmetropic, and 158 hyperopic eyes, the vitreous measured
14.13 mm, 14.04 mm, and 13.72 mm, respectively. The corresponding values for the 8 female
myopic, 106 emmetropic, and 102 hyperopic eyes were 14.24 mm, 13.86 mm, and 13.67 mm.
For 12-year-old girls (76 eyes), the coefficient of correlation of vitreous depth and refractive
error was -0.88. Considering the total axial length of the eye (i.e., the sum of the anterior
chamber depth, axial diameter of the lens, and the length of the vitreous), Larsen (1971a)
suggested three growth periods: (1) a rapid postnatal phase with an increase in length of
3.7 to 3.8 rum in the first year and a half; (2) a slower infantile phase from the second to
the fifth year of life, with an increase in length of 1.1-1.2 mm; and (3) a slow juvenile phase
continuing to age 13, with an increase of 1.3 to 1.4 mm. Most of the increase in axial length
is an increase in vitreous depth. Larsen (19716) also noted that the greater the myopia, the
greater the axial length. For 12-year-old girls, the coefficient of correlation of axial length
and refractive error was -0.82.
FIedelius (1980, 1981a, 1981b, 1981c, 1982a, 1982b) reported on the follow-up at age 18
of 137 of 539 persons who were originally examined at about age 10 to assess the ophthairnic
risks associated with premature birth (as defined by low birthweight, < 20QO g). The
follow-up sample consisted of 70 persons who were born with low birthweight (I,BW group)
and 67 who were born full term (FT group), and it deliberately overrepresented myopes
In = 58) due to the selection criteria used. Of the LBW group, 52.9 percent were myopes
OCR for page 87
87
TABLE C-7 Refraction and Axial Length and Their Changes from Ages 10 to 18 in a Sample of Low Birthweight and
Full-Term Subjects
Av. ref. Change in refractive
error error from aide 10 to 18
at age
N 18 (D.) Mean SD
Average Change in axial
actual length from age Mean age at
length lOtol8 exam (vrs.)
at age
Range 18 (mm.) Mean Range
-
Initial Follow-up
LBW males 36 -2.2 -1.75 1.66 0 to 24.23 0.90 ~.17 to 10.0 18.3
-6.3 + 2.67
FT males 36 ~.2 -1.26 1.23 + 0.2 to 24.19 0.73 ~.22 to 10.3 18.4
=.3 +2.19
LBW females 34 -0.9 -1.29 1.03 Oto 23.38 0.64 -0.29 to 9.8 18.2
4.0 + 1.92
FT females 31 ~.6 -1.07 0.91 +0.5 to 23.73 0.48 ~.11 to 10.7 18.6
-3.8 + 1.99
Note: LBW = low-birthweight group; FT = full-term group
SOURCE: Fledelius (1980,1981a, 1981b).
and, of the FT group, 31.3 percent were myopes. It was estimated that, if the entire initial
group of 539 had been followed up at age 18, 17.6 percent of the LBW group and 13.!
percent of the FT group would have been myopes (FIedelius, 1980~. Corneal curvature was
measured by keratometry; anterior chamber depth, lens thickness, vitreous depth, and axial
length were determined by ultrasonography. Refraction values were spherical equivalents
from retinoscopy with 1 percent cyclopentolate and 1 percent cyclopentolate and 1 percent
tropicamide. Right eyes of the 137 subjects were used for analysis, along with 20 left eyes
of anisometropes in the sample.
FIedelius (1980, 1981b) noted that of the 137 subjects, 127 had a change in refractive
error toward myopia, with the range being from +0.50 to -6.25 D. change between ages
10 and 18. The refractive error changes were generally greater in myopes (median change,
-1.70 D.) than in emmetropes and hyperopes (median change, -0.70 D.~. Among both
males and females, LBW subjects had a more myopic mean refractive error, a more negative
mean change in refractive error, and a greater mean axial elongation (see Table Cab. The
differences may in part be due to selection factors, since there were proportionately more
myopes in the LBW group. The most noticeable ocular dioptric component alteration was
an axial elongation (see Table Cub. Axial elongation was especially prorn~nent in the eyes
with greater changes in refractive error.
FIedelius (1981a, 1981b) noted that for comparable refractive error levels, LBW subjects
had lesser corneal radii of curvature and shorter axial lengths than FT subjects. FIedelius
(1981c) also distinguished between a "myopia of prematurity" and a "juvenile myopia"
in the LBW group on the basis of onset age. He characterized myopia of prematurity
as being diagnosed in infancy or the preschool years, of high amount, and rather static,
compared with juvenile myopia as appearing around ages ~ to 15 and progressing thereafter
to generally low or medium amounts. Both LBW categories show a small corneal radius of
curvature ("steep" cornea). Premature birth is a risk factor for myopia development early
in life, but it is unclear whether it is a risk factor for myopia development in later periods.
OCR for page 88
88
TABLE C-8 Mean Refractive Error Change and Ocular Dioptric Components at Age 18
in a Sample of Low-B~rthweight and Full-Term Subjects
Corneal Anterior
Refractive radius of chamber Lens Axial
error curvature depth thickness length
N (D.) (mm.) (mm.) (mm.) (mm.)
LBW males 36 -1.75 -0.004 + 0.08 + 0.03 + 0.90
FT males 36 -1.26 +0.02 +0.17 -0.02 +0.73
LBW females 34 -1.29 -0.005 + 0.08 -0.01 + 0.64
FT females 31 -1.07 +0.003 +0.09 -0.02 +0.48
Note: LBW = low-birthweight group; FT = full-term group.
SOURCE: Pledelius (1981b, 1982a)
COMPARISON OF RATES OF MYOPIA PROGRESSION 1lOR
ADULT- ONSET AND CHILDHOOD MYOPIA
The rates of young adult myopia progression after onset in adulthood are lower than
the rates of childhood myopia progression. In the Derby study (1985), of the 29 emmetropes
at college entrance who became myopic by graduation, 20 (69 percent) developed 1.00 or
less myopia, and only 1 developed as much as 2.00 D. of myopia. Thus, the majority were
progressing at a rate of -0.25 D./yr. or less minus. In the Diamond study (1957), the
16 airline pilots who became myopic had an average annual change of -0.05 D. For five
adult-onset myopes followed by Kent (1963), the rate of refractive change varied from -0.04
to -0.16 D./yr. In the Goss et al. report (1985), three of the four patients classified as adult
acceleration types were adult-onset myopes, and all were males. These three had rates of
young adult myopia progression of -0.16, -0.13, and -0.04 D./yr. Based on tabular data
in the O'Neal report (1986), there were 51 emmetropes (spherical equivalent equal to zero
diopters) who became myopic in their first two years at the Air Force Academy. Of these
cadets, none developed more than 1.50 D. of myopia; the mode was 0.25 D. of myopia. The
average change was -0.49 D. for a rate of about -0.25 D./yr. For comparison purposes, one
can look at the rates of childhood myopia progression from Mantyjarvi (1985b) and from
Goss and Cox (1985~. Mantyjarvi found a mean rate of childhood myopia progression of
-0.55 D./yr. in Finland. Goss and Cox found a mean rate of -0.40 D./yr. for boys in the
central United States.
Representative terms from entire chapter:
myopia progression
