INFORMATION TO USERS

INFORMATION TO USERS
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and continuing
from left to right in equal sections with small overiaps.
ProQuest Information and Learning
300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA
800-521-0600
VISUAL DEVELOPMENT AND PLASTICITY IN CHILDREN
by
Erin Marie Harvey
A Dissertation Submitted to the Faculty of the
DEPARTMENT OF PSYCHOLOGY
In Partial Fulflllment of the Requirements
For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2002
UMI Number; 3073230
®
UMI
UMI Microform 3073230
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Infomnation and Leaming Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
THE UNIVERSITY OF ARIZONA ®
GRADUATE COLLEGE
As members of the Final Examination Committee, we certify that we have
read the dissertation prepared by
entitled
F.rin Marie Harvey
VISUAL DEVELOPMENT AND PLASTICITY IN CHILDREN
and recommend that it be accepted as fulfilling the dissertation
Doctor of Philosophy
requirement for the Degree of
Feliee L. Bed:
Date
Ph.D.
Date
Williain H. blelson, Ph.D.
^
.
/,
Eliza
^T
t
Date
lisky, Ph.D.
!dni
Date
Rofiemary Ko^r, Ph.D.
i'j
.
2—
Date
Velma Dobson, Ph.D.
'
Final approval and acceptance of this dissertation is contingent upon
the candidate's submission of the final copy of the dissertation to the
Graduate College.
I hereby certify that I have read this dissertation prepared under my
direction and recommend that it be accepted as fulfilling the dissertation
iDissertation Director
Felice L. Bedford, Ph.D.
Date
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an
advanced degree at The University of Arizona and is deposited in the University Library
to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission,
provided that accurate acknowledgment of source is made. Requests for permission for
extended quotation from or reproduction of this manuscript in whole or in part may be
granted by the head of the major department or the Dean of the Graduate College when in
his or her judgment the proposed use of the material is in the interests of scholarship. In
all other instances, however, permission must be obtained from the author.
SIGNED:
4
ACKNOWLEDGEMENTS
I thank Felice Bedford, Ph.D., Velma Dobson, Ph.D., and Joseph Miller, M.D. for
their comments, suggestions, support and guidance throughout the writing of this
dissertation. I also thank my committee members, William Ittelson, Ph.D., Elizabeth
Glisky, Ph.D., and Rosemary Rosser, Ph.D., for their insightful comments and
suggestions, the staff of the Tohono O'Odham Vision Screening Program, h-ene Adams,
Meigan Baldwin, Pat Broyles, Candice Clifford, Jenniffer Funk-Weyant, Natalie Graham,
Frances Lopez, and Kathleen Mohan, for their help with data collection, and the members
of the NIH/NEI Data Monitoring and Oversight Conunittee, Donald Everett, M.A.,
Jonathan Holmes, M.D., Maureen Maguire, Ph.D., Cindy Norris, and Karla Zadnik,
Ph.D., O.D., for their valuable help and support of this research. I extend a special thanks
to the Tohono O'Odham Nation for their support of this project, the San Xavier Mission
School for allowing us to conduct the study in their school, and the children of the San
Xavier Mission School for their participation. This work was supported by a grant from
the National Institutes of Health National Eye histitute (NIH/NEI EY13153).
5
TABLE OF CONTENTS
LIST OF FIGURES
6
LIST OF TABLES
7
ABSTRACT
8
1. INTRODUCTION
10
1.1 The Effects of Uncorrected Astigmatism on Visual Perception
18
1.2 The Effects of Astigmatism-Related Deprivation on Visual Development .27
1.2.1 Recognition (Letter) Acuity
29
1.2.2 Resolution (Grating) Acuity
33
1.2.3 Vemier Acuity
39
1.2.4 Contrast Sensitivity
43
1.2.5 Stereoacuity
48
1.2.6 Summary
53
1.3 The Effects of Cylinder Lenses on Depth Perception
54
1.4 The Effects of Cylinder Lenses on Form Perception
60
1.5 Summary
62
2. A STUDY OF VISUAL DEVELOPMENT AND PLASTICITY
64
2.1 Introduction
64
2.2 Study Design and Methods
65
2.2.1 Subjects
68
2.2.2 Stimuli and Testing Methods
69
2.2.3 Data Analysis and Predictions
77
2.3 Results
81
2.3.1 Baseline Data Analysis: Normal Development in 5- to 14-year-olds...83
2.3.2 Baseline Data Analysis: Effects of Deprivation and Spatial Distortion on
Perception
93
2.3.3 Outcome Measures: Plasticity over 1 Month and 1 Year
110
2.3.4 General Summary and Interpretation of Results
122
3. DISCUSSION
3.1 Normal Visual Development in Grade-School Children
3.2 Effects of Astigmatism-Related Deprivation on Visual Development
3.3 Plasticity Associated with Recovery from Effects of Deprivation
3.4 Case Studies on Plasticity and Recovery from Effects of Deprivation
3.5 Factors Associated with Plasticity: Evidence from Other Paradigms
3.6 Plasticity Associated with Adaptation to Spatial Distortion
3.7 Conclusions
4. REFERENCES
128
128
131
137
139
149
155
159
161
6
LIST OF FIGURES
Figure 1. Demonstration of perception with astigmatic eye
Figure 2. Schematic drawings of visual focus with astigmatic eye
Figure 3. Letter acuity task
Figure 4. Grating acuity task
Figures. Vernier acuity task
Figure 6. Contrast sensitivity task
Figure 7. Stereoacuity task
Figure 8. Meridional size distortion demonstration
Figure 9. Recognition (letter) acuity by age
Figure 10. Resolution (grating) acuity by age
Figure 11. Vernier acuity by age
Figure 12. Contrast sensitivity by age
Figure 13. Mean contrast sensitivity by age and spatial frequency
Figure 14. Stereoacuity by age
Figure 15. Baseline recognition (letter) acuity by group
Figure 16. Baseline resolution (grating) acuity by group
Figure 17. Baseline vernier acuity by group
Figure 18. Baseline contrast sensitivity by group scatter plots
Figure 19. Baseline contrast sensitivity means by group
Figure 20. Baseline stereoacuity by group
Figure 21. Predicted and observed form perception by time plots
Figure 22. Mean change in recognition (letter) acuity by group
Figure 23. Mean change in resolution (grating) acuity by group
Figure 24. Mean change in vernier acuity by group
Figure 25. Mean change in contrast sensitivity by group
Figure 26. Mean change in stereoacuity by group
Figure 27. Case studies: Recognition (letter) acuity by time
Figure 28. Case studies: Resolution (grating) acuity by time
Figure 29. Case studies: Vernier acuity by time
Figure 30. Case studies: Contrast sensitivity for low spatial frequency stimuli by
time
Figure 31. Case studies: Contrast sensitivity for middle spatial frequency stimuli
by time
Figure 32. Case studies: Contrast sensitivity for high spatial frequency stimuli by
time
Figure 33. Case studies: Stereoacuity by time
Figiu-e 34. Mean form perception by trial/stimulus type and group
20
22
30
35
40
44
50
57
84
85
86
87
88
89
96
98
101
104
105
107
109
114
115
116
117
118
141
142
143
144
145
146
147
157
7
LIST OF TABLES
Table 1.
Table 2.
Table 3.
Table 4.
Tables.
Table 6.
Table?.
Correlations Between Age and Measure of Perception
Baseline Data Analysis Sununary
Baseline Grating Acuity Means
Baseline Vernier Acuity Means
Baseline Contrast Sensitivity Means
One Month Data Analysis Summary
One Year Data Analysis Summary
96
100
103
106
109
118
119
8
ABSTRACT
The effects of visual experience on perception were examined using two classic
research paradigms: visual deprivation and perceptual adaptation. The present study
evaluates the extent to which children in the 5- to14-year-old age range have the capacity
for visual plasticity with respect to recovery from the effects of astigmatism-related
visual deprivation and adaptation to spatially distorted visual input.
Visual experience was altered through eyeglass correction of astigmatism, a
condition of the eye that induces degraded (blurred) visual input and causes a form of
visual deprivation. Lenses that correct astigmatism cause two changes in sensory input:
they alleviate the deprivation effects of astigmatism, and cause spatial distortion.
Perception was initially measured when the children first received eyeglass correction,
and change in perception was measured after 1 month of wear, and after 1 year of wear.
Measures included recognition acuity, resolution acuity, vernier acuity, contrast
sensitivity, stereoacuity, and form perception.
Baseline analyses of normal (non-astigmatic) subject data indicated that
recognition acuity, resolution acuity, vernier acuity, and contrast sensitivity continue to
develop within the 5- to 14-year-old age range. Baseline analyses also revealed that
children who experienced astigmatism-related deprivation demonstrated perceptual
deficits, in comparison to non-astigmatic children, on all measures of perception
(although deficits within some measures depended on stimulus orientation (grating acuity
and contrast sensitivity) and spatial frequency of the stimulus (for contrast sensitivity)).
9
and demonstrated measurable distortions in form perception. However, primary outcome
analyses revealed little evidence of plasticity with regard to recovery from the effects of
deprivation and no evidence of plasticity with regard to perceptual adaptation to
distortion.
The results suggest that children in the 5- to 14-year-old age range may be beyond
the sensitive period for recovery from astigmatism-related deprivation through simple
restoration of clear visual input. Discussion focuses on theoretical views on conditions
necessary for plasticity (Bedford, 1993a, 1993b, 1995, Banks, 1988), and their
implications regarding another intervention, discrimination learning, that might be more
effective at inducing plasticity in children and adults who are beyond the sensitive period
for plasticity, and their implications for interpretation of data on adaptation to spatial
distortion observed in the present study.
10
1. INTRODUCTION
Visual plasticity, defined here as the capacity of the human visual perceptual
system to change based on visual perceptual experience, is a key component in the
development and maintenance of a perceptual system that allows for consistent, accurate,
and efficient perception and interaction with our environment (Bedford, 1999, Gibson
and Pick, 2000). The experimental psychology and clinical vision literatures provide a
wealth of examples of how the human visual/perceptual systems adapt based on
information gleaned through visual experience.
The research presented here is based on two classic research paradigms used to
study human visual plasticity. The first paradigm, used primarily in the clinical vision
literature, is the visual deprivation paradigm. The second paradigm, used primarily in the
experimental psychology literature, is the perceptual adaptation paradigm. The study
reported here asks to what extent children in the 5- to 14-year-old age range have the
capacity for visual plasticity with respect to recovery from the effects of visual
deprivation and with respect to adaptation to spatially distorted visual input. As 1 will
explain in the introduction that follows, answers to these questions will begin to fill
significant gaps in both the clinical vision and the experimental psychology literatures,
and will make an important step towards linking research and theory on visual plasticity
across these two literatures that have remained largely separate.
The visual deprivation paradigm has been widely used to evaluate the effects of
experience on visual development. Over the past few decades, this form of research has
11
provided extremely valuable insights into the organization of the visual system,
mechanisms of visual perception, and the variables influencing visual development.
Studies employing the visual deprivation paradigm reflect the results of developmental
plasticity under abnormal circumstances: under conditions of visual deprivation, visual
perceptual systems of infants and young children develop abnormally in comparison to
development of visual perception when clear visual input is present during development.
This pattern of results suggests that typical visual experiences during development are a
necessary component of normal visual/perceptual development, and visual deprivation
studies help us better understand the nature of the visual experiences that are necessary
for normal development of perception.
Because inflicting prolonged deprivation upon human subjects has the potential to
do permanent harm to perceptual systems, particularly if subjects are young children,
most deprivation studies have been conducted with animal subjects. However, the
clinical vision literature has documented some naturally occurring abnormalities and
variations of the human visual system that can induce various forms of visual deprivation,
and the occurrence of these conditions provides researchers and clinicians with unique
opportunities to better understand the effects of deprivation and alleviation of the
deprivation on development and plasticity in the human visual system.
Studies of patients with various naturally occurring conditions that degrade the
quality of visual input to the visual system have shown that individuals who experience
deprivation during early development often demonstrate abnormal visual perceptual
experiences immediately following restoration of clear visual input, indicating that that
effects of the deprivation of clear visual input result in a reduction in our perceptual
abilities (e.g., Maurer and Lewis, 1993). Another example of visual plasticity that also
comes from the clinical vision literature is plasticity associated with recovery from the
effects of visual deprivation. These studies show that while individuals typically
demonstrate reduced perceptual abilities after being deprived of clear visual input, they
often demonstrate gradual improvement in visual perceptual functioning once clear input
is introduced, particularly when normal input in introduced early in development (e.g.,
Maurer and Lewis, 1999).
The present study addresses plasticity associated with recovery from the effects of
visual deprivation caused by an ocular condition called astigmatism. Astigmatism is a
condition of the cornea and/or lens of the eye that, when uncorrected, induces blurred
vision. Fortunately, the deprivation of clear input caused by astigmatism can be
corrected through routine prescription of appropriate eyeglass correction. However,
previous research has shown that when astigmatism remains uncorrected during early
development, individuals experience visual perceptual deficits that persist even after the
cause of the deprivation, astigmatism, is corrected or alleviated with corrective lenses
(Freeman, Mitchell, Millodot, 1972, Mitchell, Freeman, Millodot, and Haegerstrom,
1973). Thus, while the deprivation of clear input induced by uncorrected astigmatism is
optical in origin (i.e., it is caused by abnormalities of the optics of the eye which prohibit
the focus of clear images on the retina), the deficits that persist after the astigmatism is
corrected reflect the capacity of plasticity of the visual perceptual system as a result of
the deprivation. These deficits are believed to be perceptual/neural in origin because they
13
are apparent even when the structural properties of the eye appear to be intact, and even
after appropriate eyeglass correction provides clear focus of images on the retina.
The research presented here focuses on plasticity in children between the ages of
5 and 14 years. This age group was chosen because previous research from the clinical
vision literature has suggested that there is a sensitive period for plasticity associated with
visual deprivation and recovery from the effects of visual deprivation. There are only
limited data regarding the upper age limit of the sensitive period for plasticity associated
with recovery from the effects of astigmatism-related deprivation in humans. One of the
primary goals of this study is to determine the extent to which plasticity still exists in the
visual system of 5- to 14-year-old children in terms of recovery from the effects of
astigmatism-related visual deprivation.
In addition to plasticity related to recovery from the effects of visual deprivation,
the present study also assesses the capacity for another form of visual plasticity,
perceptual adaptation, in the same subjects. Perceptual adaptation has been deflned by
Welch (1978) as "a semi permanent change of perception or perceptual-motor
coordination that serves to reduce or eliminate a registered discrepancy between or within
sensory modalities or the errors in behavior induced by this discrepancy". For over 100
years, perceptual adaptation studies have provided important information regarding the
ability of our perceptual systems to change as a result of detection of perceptual
discrepancies. The ability of our visual system to adapt in this manner allows us to
maintain accurate perception despite changes due to growth (Held, 1965, Bedford, 1999),
14
and allows us to correction for "perceptual drift", i.e., gradual changes in perception that
may occur over time and require realignment (Bedford, 1999).
As I will describe in more detail in the sections that follow, lenses that correct the
eye for astigmatism, cylinder lenses, cause two changes in sensory input. First, as
previously noted, they remove the deprivation effects of astigmatism by allowing the eye
to focus images clearly on the retina. Second, they have spatially distorting effects that
can alter both monocular and binocular spatial perception. In the study that follows, I
will assess the extent to which children adapt to the spatial distortion perceived when
cylinder lenses are first worn. Previous research on adult subjects has indicated that
extended wear of such lenses result in some, but not complete adaptation to the spatial
distortion (Mack and Quartin, 1974, cited in Welch, 1978). However, previous studies
were conducted with adult subjects, leaving open the possibility that complete adaptation
was not achieved because subjects were beyond the sensitive period for this form of
adaptation. Thus, the present study will measure monocular form perception to evaluate
adaptation to the spatial distortion induced by astigmatism lenses in the same sample of
children for whom I will evaluated plasticity for recovery from the effects of visual
deprivation. This data will be an important addition to the experimental psychology
literature on perceptual adaptation, because the literature contains little data on adaptation
in children, does not focus on sensitive periods, as the clinical vision literature on
plasticity does.
Another research paradigm that has documented examples of human visual
plasticity is the discrimination learning paradigm, which has traditionally been reported
15
in the experimental psychology literature. Eleanor Gibson (1963, 1991) describes this
form of plasticity, which she refers to as differentiation, as "any relatively permanent and
consistent change in the perception of a stimulus array, following practice or experience
with this array". In studies of discrimination learning, researchers have demonstrated that
we can improve our abilities to discriminate fine differences between perceptual stimuli
through practice and experience. This facility allows us to tailor our fme perceptual skills
towards performance of perceptual tasks that are relevant to us, allowing for more precise
perception of important stimuli and more efficient interaction with our environment
(Gibson and Pick, 2000). The discrimination learning paradigm was not explicitly
implemented in the present study. However, as I will discuss in more detail in the
sections that follow, I believe that research on discrimination learning fi'om the
experimental psychology literature has important implications for research on other
examples of visual plasticity, particularly plasticity associated with recovery from the
effects of deprivation.
In summary, the primary goal of the present study is to examine plasticity in
children with regard to recovery from the effects of astigmatism-related deprivation and
with regard to perceptual adaptation to the spatially distorting effects of lenses that are
used to correct astigmatism. Recovery from the effects of astigmatism-related visual
deprivation will be assessed for several basic aspects of visual perception: recognition
acuity, resolution acuity, vernier acuity, contrast sensitivity, and stereoacuity. Previous
research indicates that there may be different sensitive periods associated with different
perceptual functions with regard to deprivation and recovery (e.g., Harwerth, Smith,
16
Duncan, Crawford, and von Noorden, 1986, Levi and Carkeet, 1993), and thus there may
be differences in the extent to which each function can recover from the effects of
deprivation. Plasticity as it relates to perceptual adaptation will be assessed for
monocular form perception, bi contrast to the clinical vision literature, the literature on
perceptual adaptation has not experimentally addressed the issue of sensitive periods and
studies of adaptation in children are extremely rare. The present study is unique in that it
evaluates plasticity for adaptation to spatial distortion in children, and allows for
comparisons in the extent to which children show plasticity for recovery from deprivation
and for perceptual adaptation.
To briefly preview the design of the present study, perception will be measured
when the children first receive their eyeglass, after 1 month of eyeglass wear, and after 1
year of eyeglass wear. The experimental manipulation consists of altering visual
experience through eyeglass correction of astigmatism. This manipulation results in two
changes in sensory input: alleviation of the deprivation effects of astigmatism by
allowing the eye to focus clear images on the retina, and spatial distortion. Primary
analyses will focus on evaluation of perceptual change over 1 month and over 1 year for
children with high astigmatism in comparison to a control group of children with little or
no astigmatism. The fact that the same method (same pair of eyeglasses) alters visual
experience for each of these different perceptual fimctions allows for direct comparisons
in plasticity observed based on equivalent amounts of visual perceptual experience across
different measures of perception.
17
While the primary goal of this study is to determine if plasticity occurs as a result
of the eyeglass intervention in children, the baseline measures of perception will also
provide a wealth of information to the literature on developmental changes in perception
in normally sighted children, the effects of astigmatism-related deprivation on different
aspects of visual perception, and on perception of spatially distorted visual input in
children in the S- to 14-year-oId age range. With regard to baseline data on normally
sighted children, the data obtained in the present study are unique in that they include
measures on a large sample of children across a variety of perceptual measures in an age
range that is less frequently examined in developmental studies of vision (most
developmental studies of vision focus on infants and young children, as those periods
encompass the most rapid periods of visual development). While some data on the
effects of astigmatism-related deprivation exist in the literature, the data obtained in
present study are unique in reporting results from a large sample of astigmatic children
for which measures of on a wide variety of perceptual abilities were obtained. These
data will provide a broad description of the effects of astigmatism-related deprivation on
visual development not previously available in the literature. Thus, a secondary goal of
the present study is to closely examine baseline measures, both for children with
astigmatism and for children without astigmatism, in order to gain better insight into
normal visual development in this 5- to 14-year-old age range, and abnormal visual
development associated with astigmatism-related deprivation.
In the sections that follow, I discuss in more detail what is currently known about
the effects of astigmatism and cylinder lenses on visual perception and on visual
18
development. I will begin with a brief discussion of astigmatism, and will describe the
effects of uncorrected astigmatism on perception, and the effects of cylinder lenses on
perception. Finally, I summarize the existing evidence for plasticity associated with
deprivation due to astigmatism, plasticity associated with recovery from the effects of
deprivation due to astigmatism, and plasticity associated with adaptation to spatial
distortion.
1.1 The Effects of Uncorrected Astigmatism on Visual Perception
In this section, I discuss in greater detail the nature of the visual deprivation
induced by uncorrected astigmatism. Regular astigmatism is a condition of the cornea
and/or the lens of the eye in which there is unequal curvature across meridia (i.e.,
orientations) (Rosenfleld, 1988). For example, the surface of a non-astigmatic cornea or
lens is shaped like the surface of a basketball, with equal curvature across directions,
whereas the surface of an astigmatic cornea or lens is shaped somewhat like the surface
of football in which the curvature in one direction is steeper than the curvature in the
orthogonal direction. This unequal curvature affects the refraction of light as it enters the
eye, and influences the quality of images projected on the retina. The effects of
astigmatism on vision are particularly disruptive because, unlike common conditions
such as myopia (nearsightedness), in which distance vision is disrupted but clear vision at
near is often preserved, and high hyperopia (farsightedness) in which near vision is
disrupted but distance vision is often preserved, individuals with astigmatism do not
receive clear visual input at distance or at near.
19
Fortunately, like hyperopia and myopia, ophthalmologists and optometrists can
correct for the optical defocus caused by astigmatism by prescribing lenses. These
lenses, called cylinder lenses, have unequal curvature across different orientations. With
proper prescription, they can correct the unequal refraction of the light that is caused by
the astigmatism and improve the quality of the images the eye receives by allowing the
eye to focus images clearly on the retina.
Before 1 discuss how astigmatism might effect visual development, I will begin
with a demonstration of what astigmatic individuals might experience when astigmatism
is uncorrected, i.e., a demonstration of how the world may look through an astigmatic
eye. These examples provide a sense of the type of visual deprivation experienced by
individuals with astigmatism. Examples are provided in Figure 1. Imagine that the
images of the spoked wheel, the letter acuity chart and the single line on the letter acuity
chart on the left are the stimuli that the observer is viewing. The images on the right are
simulations of what an observer with uncorrected astigmatism might perceive when
viewing these stimuli. These examples illustrate an important feature of astigmatisminduced deprivation; the observer may perceive differences in resolution acuity (the
ability to discriminate fine detail) for lines of different orientations. For example, notice
that the vertical lines are most in focus, and the greatest amount of defocus occurs in the
horizontal lines. As I will now explain in more detail, the orientation of the lines that are
most in focus or out of focus in the presence of uncorrected astigmatism is dependent on
the axis of the astigmatism (direction of most/least curvature), and the type of
astigmatism present.
20
Stimuli
Simulated View through
Astigmatic Eye
Z R K D C
Z R K D C
D N C H V
C D H N R
D N C H V
C D H N R
-R V Z O S—-
R V Z O 8
O S O V z
O 8 D V Z
N O Z C O
N O Z O O
R O N8K
n u N•N
O K • V Z
II M « V #
n • N •• H
i
N O Z C D
N O Z O D
Figure 1. Demonstration of perception with astigmatic eye. Stimuli on the left
represent viewed stimuli, stimuli on the right simulate stimuli as perceived by someone
with uncorrected astigmatism.
21
The graphic in Figure 2 (revised from Gwiazda et al., 1985) provides an example
of how the optical properties of astigmatism influence resolution acuity for lines of
different orientations. The examples provided in this figure are particularly relevant the
study that follows because these examples illustrate a certain type of astigmatism, i.e.,
with-the-rule astigmatism, in which the greatest curvature occurs in the vertical meridian
and the least curvature occurs in the horizontal meridian (i.e., as if you were looking at a
football lying on its side, with pointed ends facing to the right and left). Since all of the
astigmatic subjects in the present study have with-the-rule astigmatism (i.e., the front of
the cornea is most steeply curved in the vertical direction), the examples 1 use here focus
on this form of astigmatism.
Imagine that each eye in Figure 2 is viewing a "+" stimulus (a horizontal and
vertical line). In each example, the diagram of the eye illustrates at what point, relative to
the retina, each line of the
comes into focus. For images that come into focus on the
retina, the result is typically perception of a clear non-blurred image. For images that
come into focus either in fi-ont of or behind the retina, the result is typically perception of
a blurred image.
In a non-astigmatic eye both the horizontal and vertical lines of the "+" will come
in focus in the same location: The + may come in focus on the retina (2a), in front of the
retina (2b, myopic or nearsighted), or behind the retina (2c, hyperopic or farsighted).
Regardless, in all three examples, both the horizontal and vertical lines come into focus at
the same point, so both are equally in focus (2a) or equally out of focus (2b and 2c). Also
22
A. Normal
-|-
B. Myopic
+
C
C. Hyperopic
+
'
D. Compound Myopic
Astigmatism
+
E. Simple Myopic
Astigmatism
+
F. Mixed Astigmatism
-|-
G. Simple Hyperopic
Astigmatism
+
H. Compound Hyperopic
Astigmatism
+
+
)+
(-1
Figure 2. Schematic drawings of visual focus with astigmatic eye. Illustrations of
the location at which the horizontal and vertical lines come into focus with respect to the
retina for different types of with-the-rule astigmatism.
23
note that the hyperopic eye, with changes in accommodation (i.e., changes in the shape of
the eye's lens), can bring the stimulus into focus when viewed at near or at distance.
Changes in accommodation can not improve the clarity of the stimulus in myopia, but
through changes in viewing distance (bringing the stimulus closer), the myopic eye can
bring the stimulus into focus on the retina. Thus in these three examples, the observers
gain visual experience with clear visual input under certain viewing circumstances, and
their visual experience for stimuli across orientations is equivalent.
1 now turn to examples of astigmatic eyes (Figure 2d-2h). In an uncorrected
astigmatic eye, lines of different orientations come into focus at different locations with
respect to the retina. Individuals with astigmatism can not bring both lines into focus at
the same time through accommodation or through changes in viewing distance.
However, depending on the type of astigmatism and the distance of the object being
viewed, most astigmats are able to bring lines of individual orientations into focus under
the right viewing conditions. This fact, which I will explain in further detail, makes
predictions regarding the effects of astigmatism-related deprivation somewhat
complicated.
In general, there are five types of astigmatism with respect to the focus of stimuli
relative to the retina; simple myopic, compound myopic, mixed, simple hyperopic, and
compound hyperopic. I begin with a discussion of the visual experience of individuals
with myopic astigmatism. In individuals who have myopia without astigmatism (near­
sighted) distant lines of all orientations are focused in front of the retina (see Figure 2b).
Individuals with myopic astigmatism see distant lines of some orientations better than
24
others (see Figures 2d and 2e). Since we relax our accommodation to get our best focus
of distant objects, myopes can not improve focus through accommodation, but can induce
better focus through changes in viewing distance. For example, depending on the amount
of myopia, a myopic astigmat can adjust viewing distance to bring certain lines into
focus, but can not bring all lines into focus at any one viewing distance. In general,
however, the stimulus orientation that is furthest in front of the retina (most myopic,
horizontal lines in our example) is more consistently out of focus for myopic astigmats.
Thus, myopic with-the-rule astigmats experience more consistent deprivation for
horizontal stimuli.
Mixed astigmats have one orientation in focus beyond the retina (hyperopic) and
one focused in front of the retina (myopic) (Figure 2f). As previously noted in the case of
myopic astigmats, at some distances, mixed astigmats will be able to see lines consistent
with the myopic focus clearly, although this may occur relatively infrequently. Mitchell
et al. (1973) assumed that mixed astigmats routinely accommodate to the hyperopic
focus, as they can not accommodate to the myopic focus. This accommodation, while
improving focus for the more hyperopic orientation, would further degrade focus for the
more myopic orientation. Thus, it is likely that mixed with-the-rule astigmats, like
myopic with-the-rule astigmats, experience more consistent deprivation for horizontal
stimuli (more myopic focus) than for vertical stimuli (more hyperopic focus).
Finally, 1 turn to examples of hyperopic astigmatism. For individuals who are
hyperopic and do not have astigmatism (far-sighted), stimuli come into focus behind the
retina (see Figure 2c). Individuals with simple hyperopic astigmatism have one
25
orientation (horizontal lines) in focus on the retina, and the orthogonal orientation
(vertical lines) in focus behind the retina (see Figure 2g). For individuals with compound
hyperopic astigmatism, vertical and horizontal lines are in focus behind the retina, but at
different distances behind the retina (see Figure 2h, horizontal lines are better focus than
vertical lines). Thus, one might assume that both simple and compound hyperopic
astigmats might experience greater deprivation for vertically oriented stimuli, since
stimuli of that orientation are furthest from focus on the retina. However, our ability to
accommodate (adjust the shape of the lens) to compensate for moderate amounts of
hyperopia makes this story more complicated. For example, individuals who are
hyperopic and do not have astigmatism are generally able to bring near objects into focus
by adjusting the shape of the lens in the eye through acconmiodation. Hyperopic
astigmats also have this ability to accommodate, but the accommodation can not bring all
lines into focus at the same time (i.e., they would need to accommodate different amounts
to bring stimuli of different orientations into focus). Since we are interested in the nature
of the visual deprivation that these individuals experience, an important question is how
do these hyperopic astigmats accommodate? Do they routinely accommodate so that one
line, perhaps the one closest to the retina, is in focus, or do they split the difference, and
accommodate between the two extremes? The first scenario would predict greater
deprivation for vertical lines, whereas the second would predict somewhat equivalent
deprivation for both lines. Mitchell et al. (1973) assumed that hyperopic astigmats
routinely acconunodate to the least hyperopic focus (horizontal lines). Later, Freeman
(197S) provided some support for this hypothesis. He measured accommodation in a
26
subject with hyperopic astigmatism and found that the subject accommodated to focus the
stimulus orientation corresponding to the least hyperopic meridian. However, recent
evidence provided by Dobson, Miller, Harvey and Mohan (2002, in press) indicate that
preschoolers with hyperopic astigmatism show equally reduced grating acuity for both
horizontal and vertical stimuli, suggesting that they experience equivalent deprivation for
both stimulus orientations during development. A study, similar in design to that
conducted by Freeman (1975), in which we measure accommodation in a large sample of
uncorrected astigmatic children is currently underway, and should provide evidence for
patterns of accommodation in children with astigmatism, and provide further information
regarding the nature of the deprivation experienced by hyperopic astigmats.
hi summary, while it is clear that there are stimulus orientations that are more
frequently out of focus for individuals with each type of astigmatism, the previous
discussion highlights the fact that most astigmats do not experience constant deprivation
of clear input for stimuli of any single orientation, and are likely to experience some
degree of deprivation of clear input stimuli across orientations. Thus, the exact nature of
the deprivation experience by each child, with regard to frequency and duration of
deprivation for specific stimulus orientations, can not be clearly defined based on the
available literature. For the purpose of the present study, however, I have generated
predictions in accordance with previous theory and research on astigmatism-related
deprivation in adult subjects (Mitchell et al., 1973, Freeman, 1975). More specifically, I
predict that because the stimulus orientation that is furthest from focus on the retina
(horizontal for myopic astigmats and vertical for hyperopic astigmats) or more myopic in
27
the case of mixed astigmatism (horizontal) is the stimulus orientation for which subjects
experienced the greatest and more consistent deprivation of clear visual input, and it is
this stimulus orientation for which subjects should demonstrate reduce perceptual
abilities once the astigmatism is corrected. Because the exact nature of astigmatismrelated deprivation is still in question, the secondary goal of the present study with regard
to baseline measures of the effects of astigmatism-related deprivation on visual
development will make an important contribution to the literature, and may provide
further insight into the exact nature of the deprivation experienced in subjects with
different forms of astigmatism.
1.2 The Effects of Astigmatism-Related Deprivation on Visual Development
Thus far, I have only discussed the effects of uncorrected astigmatism on
perception, i.e., the nature of the deprivation experienced due to the effects of poor optics
induced by the presence of astigmatism. In this section, I turn to issues of plasticity, and
ask what an astigmatic observer perceives when you provide her or him with the
appropriate optical correction (eyeglasses) for correction of astigmatism, so that stimuli
of all orientations are focused on the retina, as shown in Figure 2a. To briefly preview,
some individuals with astigmatism perceive stimuli clearly when wearing the proper
eyeglass correction. However, for individuals who had high astigmatism during early
visual development, often clear vision and equal resolution acuity is not achieved for
stimuli of all orientations even when the appropriate eyeglass correction for astigmatism
is worn. Furthermore, the stimulus orientation for which visual deficits are apparent is
28
consistent with the stimulus orientation for which the uncorrected astigmatism induced
the greatest deprivation during development. In these situations, the observer is said to
have amblyopia, a clinical term for reduced vision that can not be corrected simply
through appropriate eyeglass correction. In addition, some astigmatic individuals show
evidence of a sub-type of amblyopia, meridional amblyopia, defmed by a difference in
acuity across stimuli of different orientations that cannot be corrected with eyeglasses.
In this section 1 summarize the literature on the effects of astigmatism-related
deprivation on visual development, and plasticity as it pertains to recovery from the
effects of astigmatism-related deprivation for five measures of perception included in the
present study: recognition acuity, resolution acuity, vernier acuity, contrast sensitivity,
and stereoacuity. For each visual perceptual function, I will describe typical measures,
normal development, and the available data on the effects of astigmatism-related
deprivation and recovery. Thus, in the literature review that follows, I focus primarily on
the clinical vision literature, as it is this body of research that has focused on deprivation
in humans. However, since recent reports of recovery fi'om the effects of deprivation
have been noted in the literature on discrimination leaming (e.g., Levi, Polat, and Hu,
1997, Polat and Ma-Naim, 2001), I will also include relevant information on
discrimination leaming for each of the visual perceptual functions in the summaries that
follow. In the final discussion, I work towards bringing information fi'om the clinical
vision and experimental psychology literatures together, with the hope that integrating
research and theory across the literatures will lead to a better understanding of human
visual plasticity.
29
1.2.1 Recognition (Letter) Acuity
Recognition acuity is a measure of the perception of fine visual detail that is
typically assessed with black letters or symbols on a white background. Recognition
acuity measurements reflect the smallest stimulus size that the observer can both resolve
and identify. An example of a letter acuity test is shown in Figure 3 (ETDRS Acuity
Chart, Precision Vision, LaSalle, IL). Threshold letter acuity is typically reported in
logMAR units (log value of the minimal angle of resolution), which is often transformed
into more easily recognized and understood Snellen acuity values (e.g., 20/20).
Unlike measures of resolution acuity (e.g., grating acuity), recognition acuity
measures represent the ability to identify the shape of the stimulus in addition to the
ability to discriminate fine detail. Thus, recognition acuity thresholds represent stimulus
sizes that are above detection thresholds, and are the minimum stimulus size at which the
observer can also reliably identify the shape of the stimulus. While recognition acuity is
often used in clinical evaluations of visual acuity, the theoretical interpretation of what
this task measures is not clear (Riggs, 196S). For example, letter charts typically include
several different letters, and it is not clear what aspect of the stimulus subjects use to
make shape identification: for some letters, detection of a landmark feature may aid
detection, and the size of that feature would of course influence identification at near
threshold levels, bi addition, while measures are surely influenced by resolution acuity
thresholds, measures may also be influenced by limitations on shape discrimination.
Despite this ambiguity in the interpretation of recognition acuity measures, these
30
Figure 3. Letter acuity task.
31
measures do offer a practical measure of visual/perceptual performance that has
relevance to everyday perceptual tasks.
As previously noted, different visual perceptual functions appear to have different
developmental time courses, which could potentially make some visual functions more,
or less, susceptible to the effects of deprivation and recovery from these effects during
the 5- to 14-year-old age range. Therefore, before turning to research regarding the
effects of astigmatism on letter acuity, I would like to make a brief note regarding what is
know of the development of recognition acuity in normally developing children. Results
of best-corrected recognition acuity for children in the grade-school age range measured
with logMAR acuity charts have been provided by Dowdeswell, Slater, Broomhall, and
Tripp (1995) who reported average acuity of 20/25 in a sample of 5.5- to 7-year-olds, and
by Myers, Gidlewski, Quinn, Miller, and Dobson (1999) who reported average acuity of
20/20 in a sample of 10 year olds. The results of these studies, suggesting only a 1 line
increase in recognition acuity between age 5-7 and 10, indicates that little development
occurs in the grade-school age range in terms of recognition acuity, and adult levels
appear to be reached at least by age 10.
Due to the orientation-dependent nature of deprivation associated with
astigmatism, most studies of perception of fine visual detail in subjects with astigmatism
have focused on grating acuity measurements for stimuli of different orientations, rather
than recognition acuity measures of fine visual detail. However, three studies that have
focused on members of the Tohono O'Odham Nation, a tribe in Arizona with a
previously documented high prevalence of astigmatism, have reported recognition acuity
32
measures in astigmatic subjects (Kershner and Brick, 1984, Oobson, Tyszko, Miller and
Harvey, 1996, Dobson, Miller, Harvey, and Mohan, 2002). Kershner and Brick (1984)
reported that in their sample of 4"* and S"' grade Tohono O'Odham children, children with
astigmatism (> 1.00 diopter) were more likely to have best-corrected visual acuity worse
than 20/20. Dobson and her colleagues (1996) examined eye exam records for 100
consecutive patients (age 8 or older) seen at the bidian Health Services Hospital
Optometry Clinic located on the Tohono O'Odham Reservation. The results indicated
that in patients with little or no astigmatism (< ID), 13% had below normal bestcorrected letter acuity (i.e., acuity worse than 20/20 while wearing their appropriate
eyeglass correction). This percentage increased in subjects with greater amounts of
astigmatism from 30% in patients with low to moderate astigmatism (1 to < 3 D) to over
60% in patients with high astigmatism (4 D or more). More recently, Dobson and her
colleagues (2002) reported on recognition acuity in 3- to 5-year old members of the
Tohono O'Odham Nation. Rather than letter stimuli, this study used the Lea Symbols
Acuity Chart (Precision Vision, LaSalle, IL), an eye chart that uses shapes (house, square,
circle, heart) so that young children who are not adept at identifying letters can be tested.
The results of the study indicated that for children with little or no astigmatism, average
best-corrected acuity was approximately 20/40, but for children with high astigmatism,
average acuity was approximately 20/50.
Overall, these data indicate that astigmatism is associated with recognition acuity
deficits that persist even when appropriate eyeglass correction is worn, and support the
idea that the presence of astigmatism during development results in deficits that are
33
apparent as early as age 3 to 5 years. These data do not address recovery from
astigmatism-related deprivation, however, and 1 am aware of no existing studies that
address this question prospectively. However, a study recently completed (Miller,
Dobson, Harvey, and Sherrill, 2000) did address recovery from effects of deprivation in
the preschool subjects included in the report by Dobson et al. (2002), and the results of
this study are forthcoming.
Finally, some recent studies have reported that discrimination learning in adult
patients with deprivation-related perceptual deficits can result in improvements in
recognition acuity. Levi, Polat, and Hu (1996) reported that for 2/6 amblyopic observers
trained in a vemier discrimination task, the discrimination learning observed for vemier
discriminations also resulted in improvements in letter acuity: one subject improved from
20/80 to 20/22, the other improved from 20/40 to 20/20. Polat, Ma-Naim, and Belkin
(2001) reported that discrimination learning on contrast detection (across a range of
stimulus orientations and spatial frequencies) resulted in an average improvement in
letter acuity of 2.5 lines on the acuity chart in their sample of 44 subjects with amblyopia.
Thus, while these studies did not include subjects with astigmatism-related amblyopia,
they suggest that plasticity in recognition acuity can occur after discrimination learning
for adult subjects with deprivation-related visual perceptual deficits.
1.2.2 Resolution (Grating) Acuity
Resolution acuity is a measure of the minimum distance between objects that an
observer can detect (Riggs, 1965). Since the proposed study is aimed at evaluating
changes in resolution acuity as a result of orientation dependent visual deprivation,
34
grating acuity measures of resolution acuity for stimuli of different orientations are of
particular interest. Grating acuity is measured as the size of the highest spatial frequency
grating (black and white stripes) that can be resolved or discriminated (when they can not
be resolved, the grating patch is perceived as a uniform gray patch). Spatial frequency is
typically reported as the number of cycles (a cycle equals one black and one white stripe)
per degree of visual angle (a degree of visual angle is roughly equal to the size of a dime
held at arms length). The stimulus in Figure 4 is an example of a grating acuity test. On
each trial, the observer views stimuli mounted behind three circular apertures. One circle
contains a grating stimulus, the others contain gray patches. The stimuli are carefully
constructed so that when the observer cannot resolve the grating, the grating stimulus
appears the same as the gray patches. The observer's task is to indicate which of three
circles contain a grating stimulus. Observers are shown cards with progressively finer
stripes (higher spatial frequencies) until they reach a point at which the they can no
longer perceive the grating, and perceive only a uniform gray shade in each of the circles.
Before discussing the influence of uncorrected astigmatism on development of
grating acuity, I will first address studies on normal development of this visual function.
A study conducted by Mayer and Dobson (1982) used an operant preferential looking
technique to measure grating acuity in children fi-om 5 months to 5 years of age, and
compared development to measures in adult subjects. The results of the study indicated
that by 4 years of age, grating acuity for children did not differ significantly fi'om adult
values. However, Ellemberg, Lewis, Liu, and Maurer (1999) compared grating acuity for
children age 4- to 7- years-old to that of adults, and because the subjects were older and
35
Figure 4. Grating acuity task. The upper photo shows a child being tested and the lower
photo shows a close-up of the test stimuli.
36
were able to respond verbally, a method of limits procedure was used. The results
indicated that grating acuity was adult-like by 6 years of age. The discrepancy in results
between the two studies with regard to the age at which grating acuity reaches adult-like
values (age 4 vs. age 6) is likely to be due to methodological differences between the
studies, but from these studies we can conclude that normal development of resolution
acuity as measured by grating stimuli is essentially complete by the time children reach
4-6 years of age.
What effects does uncorrected astigmatism have on development of grating
acuity? A number of studies have provided evidence that the presence of astigmatism in
infancy or early childhood can lead to reduced resolution acuity for stimuli of certain
orientations even after the astigmatism is corrected (Mitchell, Freeman, Millodot, and
Haegerstrom, 1973, Freeman, Mitchell, and Millodot, 1972, Gwiazda, Bauer, Thorn, and
Held, 1986). Much of the research on the effects of astigmatism-related deprivation
focus on the presence of this pattern of deficits, which is called meridional amblyopia,
i.e., visual acuity deficits that are dependent on stimulus orientation.
Research suggests that, in order for astigmatism-related deficits to develop,
astigmatism must be present during a specific period of visual development, i.e., a
sensitive period. For example, it has been demonstrated that infants less that 1 year old
who have astigmatism do not show evidence of meridional amblyopia (Teller, Allen,
Regal, and Mayer, 1978, Gwiazda, Mohindra, Brill, and Held, 1985). However, several
studies have reported the presence of meridional amblyopia in 3- to 5-year-old astigmatic
children (Mohindra, Jacobson, and Held, 1983, Atkinson, Braddick, Bobier, Anker,
37
Ehrlich, King, Watson, and Moore, 1996, Dobson, Miller, Harvey, and Mohan, 2002).
These data suggest that that the sensitive period for susceptibility to the effects of
astigmatism-related deprivation begins after age 1 year, and prior to age 3 to S years.
The age at which the sensitive period for susceptibility to the deprivation effects
of astigmatism comes to an end is not known, I have been unable to find any specific
reports of the relation between onset of astigmatism after age 7 and development visual
perceptual deficits. Nonetheless, research suggests that other forms of visual deprivation
(e.g., cataract, a clouding of the lens in the eye) that develop in late childhood and
adulthood do not lead to the development of amblyopia (Kuman, Fedorov, and Novikova,
1984, Keech and Kutschke, 1995, Lewis, Maurer, and Brent, 1986).
Evidence regarding the extent of the plasticity associated with recovery from
effects of astigmatism-related deprivation has been reported only in the form of
retrospective studies of adults with astigmatism. In general, these studies indicate that if
astigmatism is corrected with eyeglasses prior to age 7, astigmatic adults do not show
evidence of meridional amblyopia, but if the astigmatism is corrected after age 7, the
astigmatic adults may show evidence of meridional amblyopia. For example, Mitchell et
al. (1973) reported meridional amblyopia in all of his adult astigmats who were corrected
after age 6, and no meridional amblyopia in one astigmat corrected at age 3. Similarly,
Cobb and MacDonald (1978) found that for their 12 astigmatic subjects, there appeared
to be significant increase in meridional amblyopia for the subjects who received glasses
after age 7, in comparison to the subjects who received glasses prior to age 7, who
showed little evidence of meridional amblyopia. Mohindra, Jacobson, and Held (1983)
38
reported that a 3-year old astigmatic child who showed evidence of meridional amblyopia
upon initial testing showed no evidence of meridional amblyopia after 3 months of
eyeglass wear. Thus, the limited evidence in the literature suggests that the sensitive
period for successful reversal of meridional amblyopia through eyeglass treatment is
prior to age 7 years.
In summary, the available research suggests that the sensitive period for
susceptibility to the deprivation effects of astigmatism on resolution acuity begins
towards the end of first year, and is likely to extend at least until the
year of life. The
limited research on recovery from the effects of astigmatism-related deprivation suggests
that the sensitive period for successful treatment of amblyopia is prior to age 7 years.
However, these conclusions are based on retrospective studies of a small number of adult
subjects (with the exception of the prospective report on one child subject reported by
Mohindra, Jacobson, and Held (1983), as discussed above). While it is possible that this
age would also correspond to the end of the sensitive period for the development of
amblyopia due to deprivation, it is not necessarily so, and there is no existing evidence to
indicate that astigmatism-related amblyopia is not likely to develop after age 7.
Are other forms of plasticity relevant to reversing the effects of astigmatismrelated deprivation on grating acuity? Johnson and Leibowitz (1979) and Bennet and
Westheimer (1991) looked at practice effects on grating acuity thresholds over thousand
of trials conducted over a period of several days. The results of the two studies indicated
that learning was not observed for centrally (foveally) presented stimuli, but was
observed for stimuli presented peripherally. These data suggest that discrimination
39
learning techniques may not be effective in improving grating acuity thresholds.
However, since subjects included in these studies all had normal vision, it is possible that
there was a ceiling effect, i.e., that subjects were ah-eady at or near their maximum level
of central grating acuity performance, based on sensory limitations, prior to the start of
the experiment. Subjects with reduced acuity due to deprivation may show
improvements since they are clearly not, by definition, at the maximum acuity level
relative to non-astigmatic subjects.
1.2.3 Vernier Acuity
Vernier acuity represents the fmest amount of positional offset that can be
detected. Figure S shows an example of a vernier acuity test. The observer's task is to
report which line is offset, or "wiggly". Vernier acuity has been called a form of
hyperacuity because the amount of offset that can be detected by a normal observer is
approximately 7 times fmer than their resolution acuity, and corresponds to a size that is
finer than the size of retinal photoreceptors. Because of these features, it has been
suggested that vernier acuity is a measure of positional acuity that is accomplished at the
cortical level of the visual system. In addition, it is a more complex form of pattern
processing than grating acuity and contrast sensitivity.
The developmental course of vernier acuity appears to be somewhat different than
resolution acuity. Zanker and colleagues (1992) studied the development of vernier
acuity fi'om age 2 months to 8 years and reported that vernier acuity becomes adult-like
by age S years. However, their results differ fi'om other studies on several counts,
including differences in vernier thresholds obtained at each age (their thresholds in
40
Figure 5. Vernier acuity task. The upper photo shows a child being tested and the lower
photo shows a close-up of the test stimuli.
41
younger children were higher than previously reported), and on the age at which adult­
like levels are reached, as most studies cite much later ages for full maturity of vernier
acuity. Gwiazda (1987) reported that adult-like vernier acuity is not reached until
approximately age 10, and Carkeet, Levi, and Manny (1997) reported that at age 6 years,
vernier acuity is only half that of adults (they do not report the age at which adult values
are obtained). In summary, these data suggest that vernier acuity development is more
prolonged than grating acuity development, and may not reach adult levels until age 10
years.
A number of studies have provided evidence that the presence of astigmatism in
infancy or early childhood can lead to orientation-dependent vernier acuity deficits that
are present even after the astigmatism is corrected (Mitchell et al., 1973, Gwiazda et al.,
1986). bi addition, like resolution acuity, there is evidence that vernier acuity is
associated with a sensitive period of development. Gwiazda and colleagues reported that
reduced vernier acuity in non-astigmatic children was associated with amount of
astigmatism present when the children were 6-12 months old, but not prior to that age.
These data suggest that the period of susceptibility to vernier acuity deficits due to
astigmatism begins at approximately 6 months of age. The end of the period of
susceptibility for development of vernier acuity deficits is not clear. However, Mitchell
et al. (1973) found orientation dependent vernier acuity deficits in one adult astigmat with
meridional amblyopia who had worn eyeglasses since age 14, suggesting that the
sensitive period ends prior to age 14.
42
As previously noted, the developmental course of resolution and vernier acuity
greatly differ, with resolution acuity reaching adult-like values at approximately age five
years, and vernier acuity reaching adult-like values at approximately age ten years.
Therefore, it is possible that the sensitive period for vernier acuity might extend further
than the sensitive period for resolution acuity. Levi and his colleagues have suggested
that plasticity associated with vernier acuity may have a longer sensitive period because
vernier acuity reflects perception of fine differences in spatial position. Changes that
occur during growth (migration of retinal cells, changes in interpupillary distance,
changes in the size of the eyes) would affect perception of spatial position, such as that
which is measured in vernier acuity tasks. Therefore, it would be beneficial for the visual
system to retain some level of plasticity for positional acuity at least until these growth
related changes are complete (Levi and Carkeet, 1993).
While previous research has documented vernier acuity deficits in astigmatismrelated amblyopia, it has not examined improvements in vernier acuity over time after
removal of the deprivation. However, there are some reports regarding improvements in
vernier acuity after removal of the deprivation conditions associated with other forms of
amblyopia. Sinmiers and colleagues (Simmers, Gray, McGraw, and Will, 1999) reported
improvement in vernier acuity in an adult strabismic amblyope and improvements in
vernier acuity in children with various forms of amblyopia as a result of treatment
(eyeglass wear and patching of the "good" eye, to force use of the amblyopic eye).
Is there further evidence in the literature for plasticity relevant to reversing the
effects of visual deprivation on vernier acuity? Quite a bit of discrimination learning
43
research indicates that vernier acuity can be improved with practice, and that
improvement is specific to the trained stimulus orientation. In addition, these effects of
practice have been documented in both normal and amblyopic subjects (Levi, Polat, and
Hu, 1997). However, the amblyopic subjects in these experiments did not have
astigmatism-related amblyopia. Regardless, if individuals with other forms of amblyopia
show improvement, it is possible that individuals with meridional amblyopia will also
show improvements in vernier acuity with practice.
1.2.4 Contrast Sensitivity
Contrast sensitivity represents the smallest difference in brighmess between
stimuli that can be detected, e.g., detection of shades of gray presented on a white
background. Deficits in contrast sensitivity have important implications for visual
perceptual functioning. For example, Skoczenski (2002) explains: " The capacity to see
subtle brightness differences is critical for distinguishing the countless subtle shadings
that define objects, and having poor contrast sensitivity is effectively like seeing one's
surroundings through heavy fog: differences between light and dark are degraded and
sharp edges are blurred".
Tests of contrast sensitivity typically use either letter stimuli or grating stimuli.
Since the present study is aimed at determining the effects of orientation dependent visual
deprivation and recovery, I will focus here on tests of contrast sensitivity that use grating
stimuli. Figure 6 shows an example of a contrast sensitivity test (test revised from
VCTS6500 Test, Vistech Consultants, LaSalle, IL). The observer's task is to report the
orientation of the lines in each stimulus (horizontal, tilted left, or tilted right). Contrast
44
Figure 6. Contrast sensitivity task. The upper photo shows a child being tested and the
lower photo shows a close-up of the test stimuli.
45
sensitivity for each spatial frequency is scored as the lowest contrast level at which the
observer can correctly identify the orientation of the lines. Research indicates that
contrast sensitivity in normal observers is best for mid-range spatial frequency stimuli,
poorer for low spatial frequency stimuli (wide stripes), and worst for high spatial
frequency stimuli (fme stripes). Thus, studies that employ contrast measures typically
include measures over a range of spatial frequencies, and data are presented in the form
of a contrast sensitivity function that relates the size of the stimulus to the threshold
contrast level for detection.
It is important to note that contrast sensitivity and visual acuity, while measuring
different functional aspects of vision, are not functionally independent of each other. For
example, an individual with poor visual acuity is likely to demonstrate poor contrast
sensitivity for high spatial frequency stimuli, particularly if the spatial frequency of the
contrast stimuli is below or near visual acuity threshold. However, this individual may
demonstrate more normal levels of contrast sensitivity when tested on low spatial
frequency stimuli (e.g., wider stripes), particularly if the stimuli are significantly above
visual acuity threshold. In summary, contrast sensitivity measures the smallest difference
in contrast that can be detected, whereas visual acuity measures the smallest amount of
high contrast detail that can be discriminated. While contrast sensitivity may be limited
by visual acuity if high spatial frequency stimuli are used, measurements of contrast
sensitivity for low spatial frequency stimuli may be less affected by visual acuity
limitations.
46
Before discussion of the effects of uncorrected astigmatism on development of
contrast sensitivity, I will first address normal development of contrast sensitivity.
Ellemberg et al. (1999) compared contrast sensitivity for 4- to 7-year-olds to that of adult
subjects. The results indicated that while there was still improvement from age 6 to age
7, at all spatial frequencies measured (0.33,0.5, 1,2, 10, and 20 cy/deg) contrast
sensitivity was adult-like by age 7 years. Bradley and Freeman (1982) reported similar
findings, concluding that contrast sensitivity across spatial fi'equencies reached adult
levels by age 8 years. Three other studies reported different findings, however. Scharre,
Cotter, Block, and Kelly (1990) reported that contrast sensitivity across spatial
frequencies did not reach adult-like values by age 7 years, the oldest age included in their
study. Adams and Courage (2002) reported that the age at which contrast sensitivity
reached adult-like values was dependent on the spatial fi-equency of the stimuli, with
contrast sensitivity for mid-range spatial frequency stimuli (4.8 cy/deg) reaching adult­
like values by age 4 years, and contrast sensitivity for low spatial frequency stimuli (0.40
cy/deg) reaching adult-like values after age 7 years. Also, Gwiazda et al. (1997) reported
that contrast sensitivity did not reach adult like levels by age 8 years, the oldest age group
included in their study. It is likely that differences between studies are the result of
stimulus differences and differences in testing procedures, but overall the data suggest
that contrast sensitivity across spatial frequencies appears to be fully mature by about age
7 or perhaps shortly thereafter. This reflects slightly later attainment of maturity in
comparison to grating acuity, and earlier maturity than vernier acuity.
47
A number of studies of astigmatic adults have provided evidence that the presence
of astigmatism in infancy or early childhood can lead to reduced contrast sensitivity for
stimuli of certain orientations, and that these deficits are present even after the
astigmatism is corrected (Mitchell et al., 1973, Freeman, 1975, Freeman and Thibos,
1975). Freeman and Thibos (1975) found reduced contrast sensitivity across the range
of spatial fi'equencies in meridional amblyopes. However, the age of subjects, and age of
first eyeglass correction was not reported, providing no information relevant to the
determination of a sensitive period for the effects of meridional deprivation on contrast
sensitivity. Mitchell and Wilkinson (1974) reported reduced contrast sensitivity across a
range of spatial frequencies in two adult meridional amblyopes who received optical
correction at the age of 10. These fmdings suggest that the blur associated with
astigmatism affects contrast sensitivity across stimuli of various spatial frequencies, and
suggest that the end of the sensitive period for susceptibility to the deprivation effects of
astigmatism on contrast sensitivity is prior to age 10.
The literature has documented contrast sensitivity deficits in meridional
amblyopes, but has not examined improvements in contrast sensitivity over time after
correction of astigmatism, i.e., recovery from the effects of astigmatism-related
amblyopia. However, as previously noted, Mitchell and Wilkinson (1974) reported
reduced contrast sensitivity in astigmats who received their first correction for
astigmatism at age 10, suggesting that correction of the deprivation effects of astigmatism
must occur prior to age 10 to reverse or avert the development of contrast sensitivity
deficits.
48
Finally, it is important to note that there are some reports in the literature of
improvement in contrast sensitivity with practice in adult subjects. DeValois (1977)
reported improvements in contrast sensitivity for two subjects who were repeatedly tested
in psychophysical contrast sensitivity studies over time, in which their contrast sensitivity
thresholds were determined repeatedly for stimuli of various spatial frequencies.
Recently, Furmanski and Engel (2002) reported discrimination learning for contrast
sensitivity for obliquely oriented grating stimuli, and Sowden, Rose and Davies (2002)
reported spatial frequency specific learning for contrast sensitivity using grating stimuli.
1.2.5 Stereoacuity
Depth perception in humans utilizes both monocular and binocular cues. In
measures of stereoacuity, we are interested in measuring the extent to which we can
utilize binocular cues in perception of depth. More specifically, stereoacuity is a measure
of the extent to which our visual system can utilize the difference in visual images
obtained through our eyes that arises due to the differences in spatial location of the two
eyes (i.e., slightly different viewpoints), and the extent to which this information is
utilized to give rise to the perception of depth. Measures of stereoacuity are threshold
measures of the finest interocular difference between images that we can detect, and
which can result in the perception of depth.
Since monocular cues to depth are very effective, measures of stereoacuity
typically control for monocular cues to depth. The primary technique for eliminating
monocular cues is through the use of random dot stereograms (Julesz, 1971). Random
dot stereograms are generated by creating two identical patches of randomly arranged
49
dots. One of the images is then altered by selecting a section of dots (e.g., a central
square) and shifting them horizontally (i.e., to the right in one image, and to the left in the
other). When the different images are viewed by the two eyes, the difference in location
of the central square across stimuli mimics the interocular difference when viewing realworld objects, and the square appears to float in front of or behind the patch of dots, hi
the literature review and the study that follows, random dot stereograms were used to
assess stereoacuity. Figure 7 illustrates a child being tested with a random dot test of
stereoacuity (Randot Preschool Stereoacuity Test, Stereo Optical Co, Chicago, IL).
hiterpretation of data regarding plasticity in stereoacuity in the present study is
complicated because there are potentially different sources of changes in visual
experience that could reduce best-corrected stereoacuity. First, it is possible that
deprivation due to astigmatism may result in deficits in stereoacuity, as it seems to result
in deficits for other fine perceptual functions. In addition, it is also possible that wearing
cylinder lenses to correct for astigmatism could reduce stereoacuity due to interocular
differences in the spatial distortion caused by the cylinder lenses, as I will discuss in more
detail in the next section. Thus, if astigmats show reduced stereoacuity when glasses are
first worn, it is not clear if the deficits are due to the deprivation effects of the previously
uncorrected astigmatism, or if the deficits are due to spatial distortion induced by the
corrective lenses. Of course, if no deficits in stereoacuity are observed we have evidence
against both of these hypotheses. In what follows, I summarize the available literature on
the relation between astigmatism-related deprivation and stereoacuity, which is very
limited. For this reason, even though stereoacuity data obtained in the present study are
50
Figure 7. Stereoacuity task. The upper photo shows a child being tested and the lower
photo shows a close-up of the test stimuli.
51
confounded, I believe it will add significantly to the currently available data on the
effects of astigmatism-related deprivation, in addition to the literature on adaptation to
spatial distortion, as summarized in the next section.
Estimates of stereoacuity in infants and children have indicated that by age 5,
stereoacuity development is essentially complete. Ciner, Schanel-Klitsch, Herzberg
(1996) measured stereoacuity using random dot stimuli in a sample of children ranging
from age 6 months to S years. The results indicated that by age 2.5 to 5 years mean
stereoacuity was approximately 50 sec arc, approaching adult levels of 20 seconds of arc.
The study also included two children who were between 5 and 5.5 years old, and stereo­
acuity in these children was 20 and 40 sec arc, respectively, suggesting that by age 5,
stereoacuity can reach full maturity. These results are in agreement with those reported
by Fox, Patterson, and Francis (1986), who measured stereoacuity in 3- to 5-year-olds
using the "standard three-rod test" (not a random dot test of stereoacuity), and found that
while there were significant differences in stereoacuity between the adults and children,
median stereoacuity for children approached adult levels at this age.
Does astigmatism-related deprivation result in reduced stereoacuity, even after the
deprivation is alleviated? A recent study found reduced stereoacuity in the presence of
simulated (lens induced) astigmatism (Chen, Kaye, McCloskey, Mfazo, Rubin, and
Harris, 2002). While this study confirms that the uncorrected optical deprivation induced
by astigmatism is sufficient to reduce stereoacuity, it does not address plasticity. That is,
it does not tell us if individuals with high astigmatism show reduced stereoacuity when
the astigmatism is corrected. Only one study that I am aware of has measured the effects
52
of astigmatism-related deprivation on stereoacuity. Mitchell et al. (1973) compared
stereoacuity for horizontal vs. vertical stimuli in one adult astigmat who had meridional
amblyopia, and reported poorer stereoacuity for horizontal stimuli, which was the
stimulus orientation for which the subject had experienced the greatest astigmatismrelated deprivation. While it is likely that this subject had reduce stereoacuity compared
to non-astigmatic subjects, absolute stereoacuity thresholds in comparison to nonastigmatic subjects were not reported in the study (i.e., they reported comparisons
between stereoacuity thresholds for horizontal and vertical stimuli in units of standard
deviations of repeated measures).
While there is very limited research on the relation between stereoacuity and
astigmatism-related deprivation, research on children with strabismus (poor ocular
alignment), anisometropia (difference in quality of visual inputs between eyes), and
cataract (clouding of the lens that restricts visual input) has provided important
information on the susceptibility of depth perception (stereopsis) the effects of visual
deprivation. For example, research has indicated poor spatial correspondence between
eyes can result in a lack of stereopsis in children with strabismus, and that stereo deficits
often persist even after the eyes have been aligned. Studies indicate that from age 1 to 3
years of age, children with strabismus are at risk for loss of stereopsis (Banks, Aslin, and
Letson, 1975). These fmdings suggest that there is a critical period for susceptibility to
the influence of poor ocular alignment on the disruption of stereopsis, and that this
sensitive period extends from age 1 to age 3 years. However, some studies have
demonstrated that older children and even adults with deprivation-related stereo deficits
53
can show improvements in stereoacuity (Brown, Archer, and Del Monte, 1999, MintzHittner and Fernandez, 2000, Morris et al., 1993). However, it is not clear to what extent
these data apply to the effects of astigmatism-related deprivation and recovery from these
possible effects.
Finally, evidence of improvement in stereopsis in normal adults has been reported
in research that has demonstrated improvements in the ability to fuse random dot
stereograms (e.g., Ramachandran and Braddick, 1973, Ramachandran, 1976). However,
some have suggested that improvements may be due to attentional factors, rather than
learning associated with visual processing (Sowden, Davies, and Rose, 1996). Thus, the
literatiu'e does not provide any clear indication that practice can or cannot influence
achievement of stereopsis.
1.2.6 Summary
In summary, the effects of astigmatism-related deprivation have been reported to
some extent for each of the visual perceptual functions to be studied here: recognition
acuity, resolution acuity, vernier acuity, contrast sensitivity, and stereoacuity. However,
prospective studies on plasticity associated with recovery from the effects of
astigmatism-related deprivation through eyeglass wear (restoration of clear images to the
retina) have not been reported prior to the present study. The results of this study will
make a significant contribution to the literature by providing important evidence
regarding plasticity associated with recovery from the effects of astigmatism-related
deprivation in children in the 5- to 14-year-old age range. In addition, analyses of
baseline data will also add significantly to the literature by providing data from a large
54
sample of astigmats on the patterns of perceptual deficits present in children who
experienced astigmatism-related deprivation.
Previous studies of normal development indicate that there are differences in the
age at which visual/perceptual functions reach full maturity: While there is some
disagreement across studies, the data in general indicate that maturity is reached by age 5
years for stereoacuity, by age 4 to 6 years for grating acuity, by age 7 for contrast
sensitivity, at least by age 10 for letter acuity, and approximately age 10 for vernier
acuity. With regard to the age range of subjects in present study, the literature suggests
that development of some functions is essentially complete in our sample of 5- to 14year-olds (grating acuity, stereoacuity), while some development continues within this
age range for others functions (letter acuity, vernier acuity, and contrast sensitivity).
Analysis of control group data from non-astigmatic subjects will add to the literature on
normal development of visual perceptual functions: it will provide a within-subjects
comparison of visual functions across a wide age range of children with normal visual
development.
1.3 The Effects of Cylinder Lenses on Depth Perception
As noted in the previous section, measurements of stereoacuity in the present
study are confounded, i.e., any deficits observed could be attributed to either
astigmatism-related deprivation or to the distortion induced by the cylinder lenses that
correct the astigmatism. In the previous section, I summarized the existing data on the
relation between astigmatism-related deprivation and stereoacuity. hi this section, 1
55
summarize previous work regarding the effects of cylinder lenses on perception of depth,
and provide a summary of the existing evidence for adaptation to such effects.
Viewing the world through a cylinder lens, i.e., a lens that has unequal curvature
across meridia causes two changes in perception. First, there is change orientation
dependent focus or resolution acuity as described in the previous section. Second, there is
a spatial distortion associated with the differential refraction of light across orientations
as it enters the eye. Examples of the types of spatial distortions observed are outlined by
Guyton (1977). In general, the lens distorts the stimulus by disrupting the size ratios
across orientations. Thus, a circle when viewed through a cylinder lens will look oval.
The amount of distortion observed is determined by two factors, the power of the lens
(i.e., the difference in curvature across orientations: the greater the difference in
curvature, the greater the distortion) and the distance of the lens from the pupil (the larger
the distance, the greater the distortion).
The spatial distortion is reduced as the cylinder lens is placed closer to the eye,
and is minimal when the unequal curvature is located in the actual structure of the cornea
or lens of the eye. For example, Guyton (1977) reports only 0.33% distortion per diopter
of astigmatism, whereas cylinder lenses result in approximately 1% distortion per diopter
of cylinder in the lens. Thus, astigmatism in the eye does not cause the same perceived
spatial distortion as cylinder lenses placed in front of the eye: Astigmatic individuals who
are not wearing eyeglass correction do not experience the type of spatial distortion
induced by cylinder lenses.
56
An example of the disruption in spatial correspondence that is likely to occur with
the prescription of cylinder lenses for astigmatism is illustrated in Figure 8. These
disruptions can occur as a resuh of differences in amount of cylinder in the lenses and/or
differences in the axis of the cylinder lenses between eyes. For example, if one eye views
the stimulus shown in Figure 8a through an cylinder lens at an orientation that that results
in the distortion illustrated in 8b and the other eye views through a stronger cylinder lens
that results in the distortion in Figure 8c, you can see that if you overlap these two
stimuli, as demonstrated in 8d, the inter-ocular spatial correspondence between the two
images is disrupted. In this example, there is no difference in the orientation (axis) of
cylinder lenses in front of each eye, but there is a difference in the amount of distortion
between eyes. Differences in the orientation of astigmatism between eyes will also result
in disruption of spatial correspondence between eyes. If one eye views through a lens
oriented such that it results in the distortion illustrated in Figure 8e, and the other eye
views through a lens that results in the distortion illustrated in Figure 8f, we can see as
illustrated in Figiu-e 8g that there will be disruption in spatial correspondence between
eyes. In this example the power of the cylinder lens is the same across eyes, only the
orientation of the cylinder lens differs.
Previous studies of the binocular perceptual effects associated with cylinder
lenses are based on differential spatial distortion between eyes. Much of the research in
this are has been conducted using a special type of lens called a meridional size lens
(MSL). This type of lens has been used, rather than a simple lens with cylinder in it,
because these studies used non-astigmatic observers, and putting cylinder lenses in front
Figure 8. Meridional size distortion demonstration. Figure A represents the standard
stimulus. Figures B and C represent different amounts of meridional distortion, and
figure D represents spatial overlap of distortions illustrated in B and C. Figures E and F
represent different axes of meridional distortion, and G represents spatial overlap of
distortions illustrated in E and F.
58
of a non-astigmatic eye will result in spatial distortion and defocus for stimuli of certain
orientations. MSLs provide meridional size distortion without orientation-dependent
defocus for non-astigmatic observers. Wearing a MSL in front of one eye disrupts the
spatial mapping between the eyes and disrupts interocular spatial correspondence. The
result is the perception of slant in physically flat surfaces, an effect that is most noticeable
in the absence of monocular depth cues (See Ames, 1946, for a detailed illustration and
explanation of these perceptual effects).
Several studies have examined adaptation to the depth distortions induced by
MSLs (e.g., Burrian, 1943, Morrison, 1972, Epstein and colleagues, 1970, 1971,1972).
The results of these studies are quite consistent. When subjects wear MSLs over one eye
(axis 90 degrees, in which the size of the horizontal axis is distorted) with the other eye
unoccluded, they initially experience distortions in depth perception, as described by
Ames (1946). After several days of wear, subjects typically report extensive reduction in
perceived distortion in environments in which monocular depth cues are abundant.
However, when measurements of depth perception are obtained under conditions in
which monocular cues are essentially eliminated, the distortion is still present. These
data suggest that to some degree adaptation to the distortion might consist of suppression
of stereoscopic depth information, and an increased focus on monocular cues. However,
the data from these studies also indicated that the amount of distortion measured under
conditions that eliminate monocular depth cues decreases with the length of wear, and
that a negative aftereffect in depth judgments is observed after the lenses are removed,
although ftill adaptation in the absence of monocular cues has not been reported.
59
Morrison (1972) rq)orts a case of a patient who had worn eyeglasses since
childhood, and then began wearing contact lenses as an adult. As previously noted, the
closer the lens is in relation to the eye, the less distortion occurs. Therefore, based on the
optical principles applied to eyeglasses and contact lens, the eyeglasses should have
provided the patient with a distorted image, whereas the contact lens should not.
However, this patient reported typical symptoms of depth distortion when wearing the
contact lenses. Morrison suggests that the patient had adapted to the distortion produced
by the eyeglasses, such that when the eyeglasses were removed, the patient experienced
an aftereffect. However, the extent of the aftereffect (relative to the distortion of the
eyeglass lenses) was not reported. It is possible that with prolonged wear, perhaps during
a sensitive period of development, ftill adaptation might occur.
Typically, studies have examined binocular effects in individuals wearing
distorting lenses in front of one eye, so the relative distortion between eyes has not been
studied, hi addition, effects have been measured based on gross measures of stereo
vision, rather than fine measures, such as stereoacuity. No data exist to indicate how
much of a difference in power of cylinder lenses between eyes and how much of a
difference in axis of cylinder lenses between eyes is necessary in order to disrupt
stereoacuity. However, it is possible that individuals who are given lenses for the
correction of astigmatism will experience disruption in stereoacuity (minimum amount of
binocular disparity that results in percept of depth), and/or a distorted sense of depth, but
only if they have differences in cylinder power and/or axis between eyes.
60
1.4 The Effects of Cylinder Lenses on Form Perception
bi the previous section, I reported that viewing through an astigmatic lens
produces orientation dependent size distortion and cited evidence that the effects of this
distortion may influence stereoacuity if there are differences in the distortion between
eyes, bi this section, however, I focus specifically on the distortions of form perception,
and literature on plasticity as it relates to adaptation to these distortions in form
perception. Examples of these distortions are provided in Figure 8. Do perceptual
changes occur over time as a result of the introduction of this form of altered visual
experience?
While adaptation to this type of distortion induced by cylinder lenses in clinical
practice has not been directly measured in patients, the distortion induced by such lenses
is common knowledge among clinicians. Guyton (1977) documents several observations
that are apparently well knovm among clinicians. For example, he suggests that the
ability to adapt to cylinder lenses varies across age: "It is a common clinical observation
that children adapt readily to induced distortion from astigmatic spectacle corrections."
More specifically, he notes that".... physiological adaptation is age dependent.... The
ability is well developed in children and decreases rapidly with advancing age". In the
absence of empirical data regarding developmental effects of form adaptation, these
clinical observations represent the only indication available, to my knowledge, that there
are developmental differences in form adaptation, and perhaps a sensitive period for form
adaptation.
61
Studies of adaptation to the types of binocular and monocular distortions that
occur with cylinder lenses have been conducted using meridional size lenses (MSL), and
have been reported in the experimental psychology literatiu'e. Mack and Quartin (1974,
cited in Welch, 1978) conducted a study in which observers wore a MSL over one eye,
with the other eye occluded. The MSLs were worn in an orientation that caused
monocular distortion (elongation) in the horizontal axis (10% and 25% magnification was
used). Subjects wore the lens for 1-8 hours, and adaptation was determined by measuring
the extent of the aftereffect. That is, subjects adjusted a rectangle until it appeared to
them to be a square. Judgments were made first when they initially put on the lens, and
again after they had taken off the lens after 1-8 hours of wear, during which they walked
through corridors. Aftereffects varied from 15% to 49% of the initial distortion,
depending on the degree of magnification and the length of wear. The results of this
study suggest that at least some degree of adaptation in form perception can occur with
short-term wear of MSLs.
hi the present study, I will determine if 5- to 14-year-old children adapt to the
monocular distortions induced by cylinder lenses prescribed for correction of
astigmatism. Studies of meridional size lenses conducted with adult subjects
demonstrated partial adaptation to the size distortion. It is not known if longer term wear
might result in greater, or perhaps full adaptation to the lenses. However, in the study by
Mack and Quartin (1974, cited in Welch, 1978), there was no significant difference
between the amount of adaptation observed after 1 hour and the amount of adaptation
observed after 8 hours of exposure to the distortion, suggesting that it is unlikely that
62
longer wear would produce significant increases in percent adaptation. However, clinical
observations suggest that children may demonstrate greater adaptation to the lenses than
adults (Guyton, 1977). These observations are consistent with the idea put forth by
Pettigrew (1978) (specifically with regard to adaptation to changes in depth disparity
relations that occur through growth) that there is perhaps a sensitive for perceptual
adaptation. Thus, if there is a sensitive period for adaptation, the young subjects in the
present study may show more adaptation than has been previously reported for adult
subjects.
1.5 Summary
To sum up the previous sections, there is evidence in the literature that
astigmatism-related deprivation results in reduced visual perceptual abilities in terms of
recognition acuity, resolution acuity, vernier acuity, contrast sensitivity, and stereoacuity.
However, the data on extent of the plasticity related to recovery fi-om the effects of this
form of deprivation are very limited. In addition, there is evidence of partial adaptation
to the type of form distortion induced by the lenses that correct for the deprivation effects
of astigmatism, but adaptation to this type of distortion has not been previously reported
in children. In the present study, I will examine plasticity for each of these measures of
visual perception in 5- to 14-year-old children: recognition acuity, resolution acuity,
vernier acuity, contrast sensitivity, stereoacuity, and form perception. Subjects will be
drawn from a sample that spans the upper age limit (age 5 to 14 years) of the
hypothesized end of the sensitivity period (approximately age seven) for development of
63
many of these aspects of visual perception. The primary goal of the present study is to
work towards filling the gap in the clinical vision literature on plasticity associated with
recovery from the effects of astigmatism-related deprivation, and to work towards filling
the gap in the experimental psychology literature on plasticity associated with perceptual
adaptation in children.
Another form of visual plasticity, discrimination learning, is not included in the
experimental design. However, as 1 have noted in the introductory sections, and as I will
further explore in the final discussion, the exchange of ideas and collaboration between
clinical vision research on deprivation and experimental psychology research on
perceptual adaptation and discrimination leaming may provide a particularly valuable
path towards better understanding of human visual plasticity, and may result in the
development of important new treatment options for individuals with visual/perceptual
deficits related to deprivation.
64
2. A STUDY OF VISUAL DEVELOPMENT AND PLASTICITY
2.1 Introduction
The primary goal of the present study is to examine the extent to which 5- to 14year-old children demonstrate plasticity with respect to recovery from the effects of
astigmatism-related deprivation and adaptation to form distortion. Secondary goals of the
present study include evaluation of perceptual development in children between the ages
of S and 14 years and determination of the effects of high astigmatism on visual
perceptual development.
This study is a within-subjects comparison of the limits plasticity associated with
six visual/perceptual functions: recognition acuity, resolution acuity, vernier acuity,
contrast sensitivity, stereoacuity, and monocular form perception. The subjects in the
experimental group are children with high astigmatism, a condition of the eye that
induced a form of visual deprivation. The experimental manipulation will consist of
altering visual experience through eyeglass correction of astigmatism. This manipulation
results in two changes in sensory input: alleviation of the deprivation effects of
astigmatism by allowing the eye to focus clear images on the retina, and meridional size
distortion. Thus, the same method (same pair of eyeglasses) alters visual experience for
each of these different aspects of vision allows for comparisons in plasticity observed
based on equivalent amounts of visual perceptual experience.
65
Analyses will address the following research questions:
1. For S- to 14-year-old children, is there evidence of plasticity associated with
recovery from astigmatism-related visual deficits within a 1 month or 1 year
period of good/clear visual experience, and is there plasticity associated with
adaptation to form distortion within a 1 month period?
2. What do baseline data tell us about normal perceptual development in nonastigmatic children within the 5- to 14-year-old age range, and the effects of
astigmatism-related deprivation on perceptual development?
a. Is there evidence of development of recognition acuity, resolution acuity,
vernier acuity, contrast sensitivity, and stereoacuity in normal (nonastigmatic) children within the 5- to 14-year-oId age range?
b. What patterns of visual perceptual deficits are apparent in children who
experienced astigmatism-related visual deprivation and is there
measurable form distortion when the children are first given their
eyeglasses?
2.2 Study Design and Methods
The present study was conducted on the Tohono O'Odham reservation, located in
southern Arizona. As previously mentioned, this population is of particular interest for
the study of the effects of astigmatism on visual development because previous research
has indicated that Tohono O'Odham children and adults have an unusually high
prevalence of astigmatism. For example, the prevalence of high astigmatism in urban
66
populations of grade-school children has been reported as 3-7% (Hirsch, 1963, Woodruff,
1986, Coleman, 1970), whereas the prevalence of high astigmatism in Tohono O'Odham
children is approximately 33% (Dobson, Miller, and Harvey, 1999, Miller, Dobson,
Harvey, and Sherrill, 2001). Furthermore, previous work with Tohono O'Odham
children has indicated that many are at risk for the development of visual deficits due to
uncorrected astigmatism: few grade school children are currently wearing glasses for
astigmatism, and many report never having worn glasses (Dobson, Miller, Harvey, and
Sherrill, 1999). Thus, this population was chosen because it is likely that there is a high
prevalence of astigmatism-related visual deficits among Tohono O'Odham children, and
pilot data have supported this prediction.
Data collection was performed in five parts; (1) baseline eye exam and
determination of appropriate eyeglass prescription, (2) eyeglass intervention and baseline
vision testing, (3) 1 month follow-up vision testing, (4) 1 year follow-up eye examination
and eyeglass prescription update, and (4) 1 year follow-up vision testing.
At the baseline eye examination, each child received a complete eye examination
including cycloplegic refi-action, conducted by a pediatric ophthalmologist to determine
the if child's eyes were developing normally, and to determine the child's best eyeglass
correction/prescription. Only children who had >/= 2.00 diopters (D) of astigmatism in
either eye, or had uncorrected vision worse than 20/20 and met the following criteria
were prescribed eyeglasses:
Myopia >/= 0.75 D in either meridian
Hyperopia >/= 2.50 D in either meridian
67
Astigmatism >/= 1.00 D in either eye
Anisometropia >/= 1.50 D spherical equivalent
The children were encouraged to wear eyeglasses on an ongoing basis upon dispensing of
the eyeglasses, which occurred at the baseline vision testing session.
Although not all children met the above criteria and were therefore not prescribed
glasses for continual use, all children wore glasses during the vision testing sessions.
This was done so that measurements of all children would reflect their best possible
vision, and also to reduce tester bias, i.e., to mask the testers as to which children had
high astigmatism. For children who had minimal prescriptions and did not meet the
above prescribing criteria, a set of "stock" eyeglasses was used. Thus, if a child did not
require a prescription for long-term-use, the child's prescription was matched to the stock
pair of eyeglasses that were closest to their prescription such that both right and left
lenses were within 0.50 vector dioptric difference from the child's refractive error
(calculation method described by Long, 1976, and modified by Harris, 1990). If a pair of
stock glasses did not meet these criteria, a new pair was ordered for the child to wear
during vision testing.
The baseline vision testing session was conducted on a separate day
approximately 2-3 weeks after the eye exam. A team of several trained testers conducted
vision testing. Vision tests included distance letter acuity, horizontal and vertical grating
acuity, horizontal and vertical vernier acuity, horizontal and vertical contrast sensitivity
for low, middle, and high spatial frequency stimuli, stereoacuity, and monocular form
perception. Details of the test stimuli and procedures are provided in the next section.
68
At the end of the baseline vision testing session, children who were prescribed
eyeglasses and children in the control group with significant myopia (nearsightedness),
hyperopia (farsightedness), or anisometropia (difference in refractive error between eyes)
were given their eyeglasses to wear for a 1 month period. A project staff member visited
the school weekly to encourage the children to wear the eyeglasses, to adjust or repair the
eyeglasses when necessary, and to dispense a new pair if the eyeglasses become lost or
broken. In order to maximize glasses wear, the staff member carried a spare pair of
glasses for each child who required them, so that children were not without glasses for
any length of time if they lost or broke them. A replacement pair of eyeglasses was
immediately ordered once the spare was dispensed, and the number of eyeglass
replacements per child was unlimited.
Follow-up vision testing was conducted at least 1 month after the glasses were
dispensed, and after approximately 1 year. Prior to the 1 year follow-up vision testing
session, each child received another eye examination, and eyeglass prescriptions were
updated. Each child wore her or his updated eyeglass prescription for the 1 year vision
testing session. The follow-up vision testing sessions were identical to the baseline
vision testing session.
2.2.1 Subjects
Subjects were children who attended the San Xavier Mission School, an
elementary school on the Tohono O'Odham Reservation, during the 2000/01 school year
or the 2001/02 school year, and whose parent or guardian provided written consent for
participation. We attempted to recruit all children in grades K through 8 for participation
69
in 2000/01. In addition, in 2001/02 we attempted to recruit all children who were newly
enrolled in the school (including the entire kindergarten class) for participation. The
majority of children were Native American and were members of the Tohono O'Odham
Tribe.
2.2.2 Stimuli and Testing Methods
Due to time constraints and limits on the attention span of some children tested
(particularly the very young children), testing was conducted for the right eye only (i.e.,
the left eye was occluded with an eye patch), with the exception of stereoacuity.
Test order was randomized across subjects for measures of letter acuity, grating
acuity, vernier acuity, contrast sensitivity, and stereoacuity. Each subject was randomly
assigned a test order at the baseline vision testing session (one of the 120 possible test
orders that can be generated using S tests), and each subject was tested in the same order
at baseline and follow-up. This design was implemented to reduce the chance that
differences from baseline to follow-up within a measure of vision might be obtained due
change in test order from baseline to 1 month and 1 year. Subjects participated in the
form perception measurements when they were waiting their turn for other tests.
Recognition (Letter) Acuity: Letter acuity was tested at a distance of 4 meters
using the ETDRS log MAR letter acuity chart (Precision Vision, Inc.). An illustration of
a child being tested is provided in Figure 3. Each line on the chart contains five letters.
Beginning with the top line (20/200), the subject was asked to identify all five letters on
each line of the chart, until he or she could no longer identify any of the five letters on a
line. Visual acuity was scored as the smallest line on which the child could correctly
70
identify at least three of five letters. If children were unsure of their letters, they were
given a lap card that contained all of the letters that appear on the chart, and were asked
to respond by matching the letters on the chart to the letters on the card.
Resolution (Grating) Acuity. Grating acuity stimuli were constructed using
unmounted Teller Acuity Cards (Vistech Consultants, Inc., LaSalle, IL), a commercially
available test of grating acuity that was designed for clinical testing of visual acuity in
infants and individuals who are not able to identify letters using a traditional visual acuity
letter chart. The Teller Acuity Card stimuli could not be used as constructed because
they include only vertical line stimuli.
The Teller Acuity Cards consist of a 12.5 by 12.5 cm patch of grating surrounded
by luminance-matched gray. The cards are constructed so that when an individual is
unable to perceive the stripes, the grating patch appears a uniform gray that matches the
rest of the card. Several sets of cards were purchased, and new stimuli were constructed
from the cards by mounting the stripes and matching gray area from the same card behind
circular apertures to produce 3-altemative-forced-choice (3AFC) task for horizontal,
vertical and oblique grating stimuli. An illustration of a child being tested and the test
stimuli are provided in Figure 4. Subjects' task was to identify which one of the circles
(number 1,2, or 3) contained the stripes. On an individual trial, subjects had a 1/3
chance of correctly guessing the correct location of the stripes. To further reduce the
chance of correct guessing, stimuli for four 3AFC trials were constructed for each grating
spatial fi'equency and line orientation, and subjects were required to correctly identify the
71
location of the stripes on at least three out of four trials. Using this method, the chance of
correctly guessing the location the stripes on three of four trials was 11%.
The stimuli were assembled and organized into a test book. The book includes
stripe widths ranging from 38 cycles/cm to 0.86 cycles/cm (I cycle = one black and one
white stripe). Testing was conducted at a distance of 1.5 meters. At this distance, the
stripe widths range from 104 to 2.3 cycles per degree of visual angle. The book was
organized from widest to finest stripes, interleaving horizontal, vertical, and oblique
stripes together to reduce the chance that differences in measurements across orientations
might be obtained due to subject fatigue (or boredom) if testing of one orientation were
completed prior to testing of the second orientation. Thus, subjects completed the trials
for the widest horizontal, vertical, and oblique stripes for the largest spatial frequency
stripes, then proceeded to the next finer stripe width for each orientation, etc.
In order to further reduce testing time and subject fatigue, the tester started with
the 6.5 cy/cm stripes and asked subjects to complete only the first trial at each stripe
width until the he or she incorrectly identified the location of the stripes on a trial (any
orientation). The tester then went back two stripe widths (wider stripes) for all three
orientations and from then on required the subject to correctly identify the location of the
stripes on at least three of four trials for each orientation/stripe width before continuing
on to the next finer stripe width for that orientation. If the child failed to identify stripe
location on three out of four trials for one orientation of a particular stripe width but
correctly identified stripe location on three out of four trials on one or both of the other
orientations, testing progressed to fine stripe widths only for those remaining orientations
72
on which the child continued to correctly identify the stripes. Grating acuity for each line
orientation was scored as the highest spatial fi'equency stripes (finest stripe width) on
which a subject could correctly locate the stripes on at least three of four trials.
Vernier Acuity. Horizontal, vertical and oblique vernier acuity stimuli were
generated using a computer program and printed on a laser printer with a resolution of
600 dpi (stimuli designed by and program written by Joseph M. Miller, M.D., Miller,
Harvey, and Dobson, 2002). An illustration of the test stimuli and of the test being
conducted are shown in Figure S. The figure of the test stimuli includes all four trials for
a single level vemier offset for vertical stimuli. Ten vemier offset sizes were used. At a
test distance of 1.75 meters, these offsets range from 120 arc sec to 5 arc sec. The stimuli
were printed, mounted, and organized into a test book. Like the grating acuity stimulus
book, horizontal, vertical, and oblique vemier stimuli were interleaved throughout the
book, and it was organized from largest to smallest vemier offset.
The test design was essentially the same as that which was described for grating
acuity, except here the subject's task was to identify which circle (top, middle, or bottom,
or one, two, or three) contains the "wiggly" line. The tester began with the largest offset
size, and ask subjects to complete only the first trial for each offset size/ line orientation
in descending order of offset size until he or she incorrectly identified the location of a
vemier offset on one of the orientations. The tester then went back two offset sizes
(larger offsets) for all three orientations, and from then on for each level of vemier
offset/orientation, required subjects to correctly identify the vemier offset in a 3AFC task
on three out of four trials for each line orientation before continuing on to the next
73
smaller size for that orientation. Vernier acuity for each line orientation was scored as
the smallest vernier offset at which the child could correctly identify the vemier stimulus
on three of four trials.
Contrast Sensitivity. Contrast sensitivity for horizontal and vertical stripes was
determined for three different grating spatial frequencies; 1.5,6, and 18 cycles/degree at a
test distance of 10 feet. Contrast sensitivity stimuli were constructed from a
commercially available clinical test of contrast sensitivity, the VCTS6500 Contrast
Sensitivity Chart (Vistech Consultants, Inc., LaSalle, IL). The task requires subjects to
identify the orientation of grating stimuli, which in the original version of the test are
either vertical, or rotated 15 degrees clockwise or counter-clockwise from vertical. The
original configuration of the test includes only vertical stimuli, and only one trial for each
level of contrast/spatial frequency. Therefore, several charts were purchased, and were
used to construct three test books (one for each of the three spatial frequencies (stripe
widths) that contained four trials at each contrast level/orientation (horizontal/vertical).
An example of four contrast sensitivity trials for a single stripe width (spatial frequency),
contrast level, and stimulus orientation (horizontal) is shown in Figure 6, along with a
photo of the test being conducted. The test design was similar to that which was used to
test vemier acuity and grating acuity: each trial was a 3AFC task (horizontal, rotated
clockwise, or rotated counter-clockwise in the example shown, or vertical, rotated
clockwise, or rotated counter-clockwise (not shown)), and the subject must correctly
identify the stripe orientation on at least three of four trials before continuing on to the
74
next contrast level for that orientation. Horizontal and vertical stimuli were interleaved
within the test book.
Pilot testing indicated that children can most reliably perform this task by holding
up a pen in front of them, and matching the orientation of the pen to the orientation of the
stripes. The tester began with one of the three spatial frequency stimulus books and
presented the first trial at each contrast level/stripe orientation starting with the highest
contrast level and proceeding sequentially towards lower contrast levels, until the child
incorrectly identified the orientation of a stimulus. The tester then went back two
contrast levels for both orientations, and from then on required the child to correctly
identify the orientation of the stripes on at least three of four trials before continuing on to
the next lower contrast level for an orientation. Contrast sensitivity for each stripe
orientation for a given spatial frequency (stripe width) was scored as the lowest contrast
level on which the child was able to correctly identify the orientation of the stripes on at
least three of four trials. Order of testing across spatial frequency (1.5,6, and 18 cycles
per degree) was counterbalanced across subjects, and each child was tested in the same
order at baseline, 1 month, and I year.
Stereoacuity. Stereoacuity was assessed using the Randot Preschool Stereoacuity
Test (Stereo Optical Co., Chicago, IL), a commercially available clinical test of
stereoacuity that utilizes random dot stimuli in order to assess stereoacuity in the absence
of monocular depth cues. Since there is a wide age range included in the present study,
the Randot Preschool Stereoacuity test was chosen so that all subjects would be able to
perform the task. The stereoacuity test included six levels of retinal disparity that range
75
from 800 to 40 seconds of arc, presented in three test books. For testing of stereoacuity,
subjects wore specially constructed polarized glasses over their eyeglasses. It is these
polarized glasses that cause retinal disparity for the random dot display, i.e., they cause
slightly different images to be seen by the right and left eyes. A photo of a child being
tested is shown in Figure 7.
Beginning with the largest disparity level, subjects were required to identify at
least two of three shapes (e.g., star, house, duck) that appear in a random dot display.
The random dot display is presented on the right side of the book, and the possible forms
that can appear are shown in silhouette form on the left side of the book. Therefore,
subjects could respond either verbally, or could point to the form on the left that they see
in the random dot display on the right. Once a subject correctly identified at least two of
three forms at a given disparity level, the tester will then continued on to the next finer
disparity level, until subject was unable to correctly identify at least two shapes.
Stereoacuity was recorded as the smallest disparity at which the subject could correctly
identify two of three shapes in the random dot display.
Monocular Form Perception. Stimuli for evaluation of monocular form
perception were created using PaintShop Pro computer software, and were presented on a
laptop computer. The stimuli were generated from an image of an outline of a circle, and
an outline of a circle filled in with a checkerboard pattern. Each of these two standard
stimuli were distorted along the vertical axis to generate additional test stimuli. In total,
24 stimuli were generated from each standard: 12 that were elongated (made taller) and
12 that were reduced (made flatter) from the standard versions in 1% increments (i.e.
76
from 1% distortion to 12% distortion).
From these stimuli, two Microsoft PowerPoint
slide shows were generated: one for the outline circle and one for the checkerboard circle.
Each page of the slide show contained a single stimulus centered on the screen, and the
pages were ordered in terms of distortion from +12% (circle elongated vertically) to 12% (circle shortened vertically).
Subjects were told that their task was to adjust the stimulus until they found the
one that looked perfectly round, like a ball rather than an egg, and to tell the experimenter
once they had decided which one was the perfect circle. Stimuli were presented on a
laptop computer, and the keyboard was covered with the exception of two keys: the pageup key, and the page-down key. Subjects were instructed that they should press the top
key to make the "circle" taller, and press the bottom key to make the "circle" flatter.
Subjects participated in eight trials: four trials with their glass-on, and four with
their glasses off. For each of the two glasses conditions (glasses on/off), there were two
trials for each stimulus type (outline/checkerboard). For each stimulus type, one trial
started from the most vertically elongated version (+12%), and one trial started from the
most vertically shortened stimulus (-12%). After each trial, the experimenter recorded
which stimulus the subject chose as "the perfect circle". Subjects were not given
feedback regarding how close they were to finding the "correct" circle, but they were
encouraged to take their time and make sure they found the "perfect circle". Order of
testing (glasses on trials first or second) and stimulus order was randomized across
subjects, and subjects conducted the test in the same order at baseline and at 1 month.
77
2.2.3 Data Analysis and Predictions
Subjects with ocular abnormalities other than myopia (nearsightedness) and
hyperopia (farsightedness) and astigmatism were excluded from analyses. In addition,
since predictions for astigmatic children are based on the presence of with-the-rule
astigmatism (plus cylinder axis 90+/- IS degrees), any astigmatic children with
astigmatism that was not with-the-rule were also excluded from analyses.
Subjects were assigned to either the control group or the experimental group
based on the results of the eye examination. Subjects with high astigmatism (>/= 1.00
Diopter in the right eye were assigned to the experimental group, and subjects with low
or no astigmatism (< 1.00 Diopter in the right eye and left eye) were included in the
control group. Subjects in the experimental group were further divided into two
subgroups based on the type of astigmatism present: one group included subjects with
myopic or mixed astigmatism (Figure 2d, 2e, and 20, and the other included subjects
with hyperopic astigmatism (Figure 2g and 2h).
Analyses are divided into three main sections. The first two sections address the
secondary goals of the study, but because they entail analyses of baseline data, they are
presented first in order to maintain the sequential logic of the study design. The first
section addresses normal development across each aspect of vision and includes only
control group subjects. The second section addresses the effects of astigmatism-related
deprivation on visual development and initial measures of perception under conditions of
spatial distortion and includes only baseline data comparing the astigmatism groups to the
control group. The final section addresses the primary goal of the present study,
78
plasticity related to recovery from astigmatism-related deprivation and to adaptation to
form distortion, and compares change in perception from baseline to 1 month and from
baseline to 1 year for the control group and the astigmatism groups.
Before continuing on to the results, I will briefly summarize the predictions
regarding evidence for plasticity for both recovery from deprivation and for form
adaptation. For measures of plasticity associated with recovery from astigmatism-related
deprivation, 1 first predict that perceptual performance at baseline will be poorer for the
astigmatism groups than for the control group. More speciflcally, for grating acuity,
vernier acuity, and contrast sensitivity, measures on which 1 have included stimuli across
different orientations, patterns of deficits observed should be dependent on the type of
astigmatism present. That is, the stimulus orientation that is furthest from focus on the
retina (horizontal for myopic and vertical for hyperopic astigmats) or more myopic in the
case of mixed astigmatism (horizontal) when the astigmatism is uncorrected is the
stimulus orientation for which subjects experienced the greatest and more consistent
deprivation of clear visual input, and is therefore the stimulus orientation for which they
should demonstrate deficits once the astigmatism is corrected. If there is plasticity for
recovery from the effects of astigmatism-related deprivation in children of this age range,
there should be greater improvement from baseline to 1 month and greater improvement
from baseline to 1 year in children in the astigmatism groups than in children in the
control group.
Similarly, for determination of plasticity associated with form adaptation,
analyses focus on change from baseline to 1 month for the astigmatism groups in
79
comparison to the control group. However, analysis of baseline data will first help
determine if the children experience the distortion induced by the cylinder lenses, and if
those distortions our measurable using the procedures implemented in the present study.
Because cylinder lenses that correct for with-the-rule astigmatism have the effect of
shortening perception of stimuli along one axis (flattening, as shown in Figure 8b), true
circles should appear to be ovals when glasses are first worn by subjects in the
experimental group. Thus, it is likely that astigmatic subjects will tend to perceive
stimuli that are vertically elongated as more circle-like, because the elongation along the
vertical axis present in the actual stimulus will essentially balance out the effect of the
horizontal flattening induced by the lens, resulting in a perceived circle. Thus, for
baseline measurements, the prediction is that that children in the control group, both with
and without glasses on (because their glasses have little or no cylinder in the lenses), and
children in the astigmatism group performing the task without their glasses on will be
likely to identify stimuli close to the actual circle as their "perfect circle". Furthermore,
children in the astigmatism group when wearing their glasses (with high cylinder in
them) will be likely to choose an oval that is most likely to result in a percept of circle,
based on the optics of the cylinder lens they are wearing (i.e., the vertically elongated
circle). This pattem of results will confirm that the cylinder lenses that correct for
astigmatism distort form perception, and that the distortion is measurable via this
technique.
In the present study, I measured both "reduction of effect" and "negative
aftereffects" for form adaptation. In measurements of reduction of effect, baseline and
80
post-exposure measurements are made under conditions of altered sensory experience:
perception is measured when the altered sensory experience is first induced (i.e., when
glasses are first dispensed), and then after prolonged exposure (i.e., after 1 month of
glasses wear). In measurements of negative aftereffects, baseline and post-exposure
measurements are made under typical conditions: perception is measured prior to the
introduction of the altered sensory experience, and then after the altered sensory
conditions are removed. Measurements at baseline and follow-up were obtained under
conditions of altered sensory experience (while subjects are wearing eyeglasses), and
under unaltered sensory experience (while the subjects were not wearing eyeglasses).
While measurements of negative aftereffects are the preferred measure of perceptual
adaptation, the presence of sensory deprivation (blur induced by astigmatism) under
unaltered sensory conditions in the present study makes measurements of "aftereffects"
difficult. However, since subjects are likely to be able to perform the monocular form
adaptation task even without their glasses on, I measured both "reduction of effect" and
"negative aftereffects" for monocular form adaptation: measures at baseline and 1 month
were conducted both with and without eyeglasses.
At the 1 month follow-up session, control group subjects will be expected to
perform as they did at baseline: they will identify circles close to the actual circle as their
"perfect circle" both with and without their glasses on. However, if plasticity exists for
monocular form perception as a result of 1 month of cylinder lens wear, subjects in the
astigmatism group will become more likely to choose stimuli that are less vertically
elongated than those that they chose at baseline, both with and without their glasses on.
81
This result would demonstrate adaptation both in tertns of "reduction of effect", i.e., with
their glasses on, they will demonstrate a reduced perceptual distortion from baseline to
follow-up when tested with their glasses on, and a "negative aftereffect, i.e., without their
glasses on, they will tend to choose an oval that is distorted in the opposite direction as
the distortion induced by the glasses.
2.3 Results
A total of 157 children were enrolled in the study. Of these children, 12 were
excluded from analyses for the following reasons: child refused eye drops (n=l),
exotropia (n=2), active conjunctivitis (n=l), blepharophimosis (n=l), iris coloboma
(n=l), ptosis (n=l), astigmatism present, but axis was not with-the-rule (n=l), did not
meet criteria for either the control or astigmatism group (RE astigmatism < 1, LE
astigmatism >/= 1, n=4), prescription found to require adjustment after first eye exam
(n=l). Thus, the final sample consisted of 144 children in grades K-8. The children
ranged in age from 5.4 to 14.4 years of age (mean = 9.4 years, SD = 2.7), and 53% of the
children were female.
Based on the results of the eye examination, children were assigned to groups. Of
the 144 children, 96 were assigned to the control group (right eye and left eye
astigmatism < 1.00 D), 29 were assigned to the myopic/mixed astigmatism group (right
eye astigmatism >/= 1.00 D, and plus cylinder sphere < 0), and 19 were assigned to the
hyperopic astigmatism group (right eye astigmatism >/= 1.00 D, and plus cylinder sphere
>/= 0). The prevalence of high astigmatism found in this sample (33%, 48/144) is
82
remarkably consistent with previous research, which also found the prevalence of high
astigmatism in Tohono O'Odham preschool children to be approximately one-third
(Dobson, Miller, and Harvey, 1999, Miller, Dobson, Harvey, and Sherrill, 2001).
There were significant differences among groups with regard to age
(F(2,141)=6.19, p=0.003). Post-hoc comparisons indicated that the hyperopic group
(mean=7.85, SD 2.07) was significantly younger than the myopic/mixed group (10.52,
SD 2.41) and the control group (9.42, SD 2.70), although the myopic/mixed group did
not differ significantly from the control group. This pattern is consistent with previous
research on changes in refi'active error with age which indicates that hyperopia is more
prevalent in infancy and early childhood, and tends to decrease through mid to late
childhood (for summary, see Zadnik and Mutti, 1998). Since age differences existed
across groups, analyses were conducted with age entered as a covariate to reduce the
possibility that significant differences among groups might be due to differences in age
(differences in level of development), rather than to differences in visual experience (i.e.,
type of astigmatism present). Amount of astigmatism did not significantly differ between
the myopic/mixed (mean = 2.84 diopters, SD=1.60, range = 1.00 to 8.25) and hyperopic
groups (mean = 2.33 diopters, SD =1.27, range = 1.00 to 5.25) (p=0.24).
Sample sizes at the two follow-up points were reduced relative to baseline. Of
144 children in the final baseline sample, 141 (98%) participated in the 1 month followup. At the 1 year follow-up, the sample size was reduced to 84/144 (58%) for several
reasons: (1) attempts were not made to follow the 8"* grade class which had graduated
and gone on to other schools, nor were attempts made to follow several children who
83
were no longer enrolled in the school when we returned the second year for testing; (2)
parents of one child refused participation in year 2; and (3) some children were newly
enrolled in the school during year two of the study, and therefore baseline and 1 month
data were collected in year two rather than 1 year follow-up data (this had the effect of
increasing the sample size for baseline and 1 month data, but reducing the % follow-up
for 1 year follow-up).
2.3.1 Baseline Data Analysis: Normal Development in 5- to 14-year-olds
Since, as a secondary research goal, I was interested in evaluating normal
developmental trends for each of the measures of vision in the developmental analysis,
these analyses were conducted only on children in the control group, i.e., children who
had structurally normal eyes with no potentially amblyogenic conditions. This section
contains detailed results of statistical analyses. A general summary and interpretation of
results is provided at the end of the Results Section.
Figures 9-14 plot individual data and means on each measure by age. Table 1
ranks the results of correlation between age and each measure from lowest to highest.
While some correlations appeared stronger than others, statistical analyses indicated that
there was a significant relation between age and performance on each measure, with the
exception of stereoacuity (p < 0.08).
In order to provide a more fine grained analysis of when significant development
in each of these perceptual functions is occurring within this 5- to 14-year-old age range,
I conducted ANOVAs on each measure to compare performance across several age
categories: 5- and 6-year-olds, 7- and 8-year-olds, 9- and 10-year-olds, and
84
a
•
(Q
2
8*
o o
A
Io
o.o-
a
A o
•
o
o o o o
A • <ji
(ji i{)
A o
AA
(|)
i|) i|)
-.2-
-.4
4.0
Rsq = 0.3411
6.0
8.0
10.0
12.0
14.0
16.0
Age (Years)
22
te
7 to < 9
9 to < 11
Age (Years)
Figure 9. Recognition (letter) acuity by age. Individual data are plotted in A, and means
+/- 1 standard error are plotted in B. In A, each line on the symbols represents data from
one subject.
85
Age (Yean)
B
S
O
^ 0 ^ ^ a
0 0 DO
^
A ji
S
V
Iw
s
S 10
5
^ 0 0 -4- 0
o0
^
00
AAA
o
a
flO
80
100
120
Age (Years)
140
teo
Age (Years)
Age (Years)
Figure 10. Resolution (grating) acuity by age. Individual data are plotted in A, B, and C
for vertical, horizontal and oblique stimuli, respectively. Each line on the symbols
represents data from one subject. Means +/-1 standard error are plotted in D.
86
am
i
8
I
>
>
40
60
14 0
120
10 0
80
160
Age (Years)
0
o
0
a
ooO
o^
B
o
0
o
o
a
o
o
A. ^ o o A ^
^ ir^
0
0^o
0
0^
100
120
Age (Years)
Aoe (Years)
m
*8
i
I
>A
>
I
s
40
60
80
100
120
14 0
18 0
Age (Years)
Figure 11. Vernier acuity by age. Individual data are plotted in A, B, and C for vertical,
horizontal and oblique stimuli, respectively. Each line on the symbols represents data
from one subject. Means +/- 1 standard error are plotted in D.
87
0a
<
o
g
10
B
f«>
0
• 00
0
«
o
o
o
3
>
0Q
O
• -f ° o
£
o
3
Z
f
m
I "
&
o
A^ 0
»*>*-<
0^
^-
o
o
o
o 0 ^ A A # 0 AJ""""""^
o
80 80 too 120 1*0 too
a
o
a
a
a
AiQ'01489
40
60
SO
Age (Years)
fOO
120
t40
{00
Age (Years)
5
i
6I 10
3
Z
g .10
3
>
I& ^ ^
jO^ OQ
0 4-^
^oaoa0.Aw-4-^Ai
a GO
eo
ao
too
t2 0
i4 0
0 •• ^
4 ^ M. o M. ^ ^-k^rj
a a o
^o
i«a
40
Aoe (Years)
60
ao
too
t2o
140
tao
Age (Years)
o^o-^oaooo
3
Z
o o ^ ^ ^ ^ Ak ^ ^ ^
•^OO ^ 4 . 0 ^ 0 - ^ ^
ooa
60
ao
too
t20
140
ao^^^oAo^
• o ^ ^ ^ ^ oAoA-#leo
Age (Years)
Figure 12. Contrast sensitivity by age. Individual data are plotted in A and B for low
spatial frequency stimuli, C and D for middle spatial fi'equency stimuli, and E and F for
High spatial frequency stimuli. Data for vertical stimuli are plotted in A, C, and E, data
for horizontal stimuli are plotted in B, D, and F.
88
Vertical Contrast Sensitivity (Control Group Only)
-2.2
-2.0
-1.8
to
2 -1.6
c
o
O
g) -1.4
5 and 6 year olds
7 and 8 year olds
9 and 10 year olds
11 to 14 year olds
-1.2
-1.0
-0.8
1
,
1
1.5
6.0
18.0
Spatial Frequency
Horizontal Contrast Sensitivity (Control Group Only)
-2.2
-2.0
-1.8
to
S -1-6
c
o
O
g»
—I
5 and 6 year olds
7 and 8 year olds
9 and 10 year olds
11 to 14 year olds
-1.2
-1.0
-0.8
T
1.5
6.0
18.0
Spatial Frequency
Figure 13. Mean Contrast sensitivity by age and spatial frequency. Means +/-1 standard
error for vertical stimuli (A) and horizontal stimuli (B).
89
2.4
2.32.2-
CD
*5
(0
g,
2.0-
3
i
s
1.8-
CO
(ji
4.0
6.0
8.0
10.0
12.0
14.0
16.0
Age (Years)
Figure 14. Stereoacuity by age. Each line on the symbols represents data from one
subject.
90
Table 1. Correlations between age and measures of perception.
Measure
Correlation
Stereoacuity
0.15
Grating Acuity - Oblique
0.18*
18 cy/deg Contrast Sensitivity - Vertical
0.27*
6 cy/deg Contrast Sensitivity - Horizontal
0.30*
18 cy/deg Contrast Sensitivity - Horizontal
0.32*
Vernier Acuity - Horizontal
0.33*
6 cy/deg Contrast Sensitivity - Vertical
0.34*
1.5 cy/deg Contrast Sensitivity - Vertical
0.38*
Grating Acuity - Horizontal
0.39*
Grating Acuity - Vertical
0.41 *
Vernier Acuity - Oblique
0.42*
1.5 cy/deg Contrast Sensitivity - Horizontal
0.43*
Vernier Acuity - Vertical
0.46*
Letter Acuity
0.58*
• P < 0.01
91
11- to 14-year-olds. Plots of means by age are provided for each age in Figures 9-14.
ANOVAs on each of the measures indicated a significant main effect of age group for all
measures with the exception of stereoacuity (which approached significance, p=0.06) and
grating acuity for oblique stimuli (p=0.13).
In order to narrow down the age range during which most development was
occurring, post-hoc comparisons focused first on determining if there were significant
improvements between adjacent age groups (5/6 vs. 7/8, 7/8 vs. 9/10, 9/10 vs. 11+). For
letter acuity, the ANOVA showed a significant main effect of age group (F(3,93)=13.87,
p < 0.001), and post-hoc comparisons indicated that there was significant improvement
between the 7/8 and 9/10 age groups (p = 0.024), but not between other adjacent groups.
This finding is apparent in the plot of means, which shows a greater slope between the
means for these two age groups.
For vertical and horizontal grating acuity, there were significant main effects of
age group (F(3,93)=6.58, p< 0.001 for vertical and F(3,93)=5.45, p= 0.002 for
horizontal). There was no significant difference between adjacent age groups, but there
was significant change from 5/6 to 11+ (p=0.005 for horizontal and p=0.038 for vertical),
and from 7/8 to 11+ (ps=0.002). Thus, as can be seen in the plot of means for vertical
and horizontal grating acuity, much of the development occurs between the 7/8 and 11+
age groups, suggesting gradual improvement between ages 7 and 11+. As previously
noted, no main effect of age group was observed for grating acuity for oblique stimuli,
although the general shape of the developmental function is similar to that obtained for
horizontal and vertical stimuli.
92
For vernier acuity, there was a significant main effect for age group
(F(3,94)=10.09,p< 0.001, F(3,92)=3.48, p=0.019, and F(3,93)=6.73, p <0.001 for
vertical, horizontal and oblique stimuli, respectively) but no significant improvement was
observed between adjacent age groups. However, there was significant improvement
between the 5/6 to 9/10, and the 7/8 to 11+ age groups for vertical (p=0.001 and p=0.039,
respectively) and oblique stimuli (p=0.029 and p=0.04, respectively), and significant
improvement from the 5/6 to the 11+ age groups for horizontal stimuli (p=0.024). Thus,
even more so than for grating acuity, improvements in vernier acuity appear to be gradual
across this age range.
Analyses revealed significant main effects of age group for contrast sensitivity for
vertical and horizontal low spatial frequency stimuli (F(3,91)=7.49, p < 0.001 for vertical
and F(3,91)=8.14, p<0.001 for horizontal), mid range spatial frequency stimuli
(F(3,91)=4.83, p=0.004 for vertical and F(3,91)=3.70, p=0.015 for horizontal), and high
spatial frequency stimuli (F(3,92)=3.52, p=0.018 for vertical and F(3,92)=4.64, p=0.005
for horizontal). Post-hoc analyses on contrast sensitivity for low spatial frequency
vertical and horizontal stimuli indicated that there was significant improvement between
the adjacent 7/8 and 9/10 age groups (p=0.001 for vertical and p=0.002 for horizontal).
For mid range spatial frequency stimuli, improvement between the 7/8 and 9/10 year age
groups for vertical and horizontal stimuli approached significance for both stimulus
orientations (ps=0.09 and 0.05), and improvement from 7/8 to 11+ age groups was
significant for vertical stimuli only (p=0.006). Finally, for the high spatial frequency
stimuli, there was significant improvement for vertical stimuli between the 7/8 and 11+
93
age groups (p=0.034), with the 7/8 vs. 9/10 comparison approaching signiflcance
(p=0.n), and there was significant improvement for horizontal stimuli between the 7/8
and 9/10 year age groups (p=0.03S). This pattern of results is clear in plots of mean
contrast sensitivity in Figure 13, which show a relatively large jump in improvement
between the 7/8 and 11+ age groups.
As previously noted, the signiflcant main effect of age for stereoacuity
approached statistical significance (p=0.063). Therefore, post-hoc comparisons for this
measure were also conducted, and indicated a marginally significant improvement in
stereoacuity between the 5/6 and 9/10 year age groups (p=0.055), suggesting small
gradual improvement within the 5 to 10 year age range.
2.3.2 Baseline Data Analysis; Effects of Deprivation and Spatial Distortion on Perception
This section contains baseline comparisons between the control group and the
astigmatism groups on baseline measures of perception, that is, perceptual performance
when the children first put on their eyeglasses. For recognition acuity, resolution acuity,
vernier acuity, contrast sensitivity, and stereo acuity, differences between the control and
astigmatism groups reflect the influence of astigmatism on visual perceptual
development. For measures of form perception, differences between the control and
astigmatism group reflect the spatially distorting effects of the cylinder lenses used to
correct for astigmatism. Separate analyses were conducted for each measure. This
section contains detailed results of statistical analyses. For comparison across measures,
a summary of baseline analyses is provided in Table 2, and a general sunmiary and brief
interpretation of results is provided at the end of the Results Section.
94
Table 2. Baseline Data Analysis Summary. All analyses conducted with age as a
covariate to control for age (i.e., developmental differences) across groups. Sample sizes
for the control, myopic/mixed, and hyperopic groups, respectively, provided in
parenthesis.
Post-Hoc Tests
Control
vs.
Myopic/
3 Groups ANOVAs
Control vs.
Mixed vs.
Myopic/
Hyperopic
Hyperopic
Mixed
»(<0.001)
• (<0.001)
Group * (<0.001)
NS (0.93)
Letter Acuity (94,28,19)
•(=0.018)
•(=0.001)
NS (0.98)
V
• (=0.028)
•(<0.001)
NS (0.13)
Grating Acuity
H
Orientation x Group *
(=0.017)
•
(<0.001)
(94.28,18)
•(<0.001)
NS (0.84)
0
• (=0.029)
NS (0.73)
NS(0.16)
V-H
•
(=0.003)
•
(=0.003)
NS (0.31)
V
Group * (0<0.001)
•(=0.001)
• (=0.002)
NS (0.78)
Vernier Acuity
H
Orientation x Group (NS)
•(=0.001)
•(<0.001)
(93,27,19)
NS (0.20)
0
NS (0.97)
NS (0.99)
NS (0.26)
V-H
NS (0.20) NS (=0.08)
NS (0.67)
V
Contrast
Orientation x
• (=0.012)
Group
*
• (0.004)
NS
(0.90)
Sensitivity
H
(0.047)
• (0.028)
1.5 cy/deg
NS (0.41)
NS (0.26)
V-H
•
(=0.007)
•(<0.001)
NS(0.15)
V
Contrast
Orientation x
•(=0.001)
• (=0.002)
NS (0.30)
Sensitivity
H
Group (NS)
6.0 cy/deg
NS (0.58)
NS (0.63)
NS (0.44)
V-H
•(<0.001)
•(<0.001)
NS (0.27)
V
Contrast
Orientation x
• (<0.001)
Group *
•(<0.001)
NS (0.99)
Sensitivity
H
(=0.004)
•(=0.001)
NS
(0.93)
18.0 cy/deg
NS (0.23)
V-H
»(<0.001)
• (=0.037)
• (<0.001) NS (0.062)
Stereoacuity (95,28,18)
* Statistically significant before Bonferroni correction applied (p values represent
uncorrected significance level).
V = Vertical Stimuli, H = Horizontal Stimuli, O = Oblique Stimuli, V-H =Vertical
- Horizontal.
Orientation x Spatial
Frequency x Group
(=0.016)
Measure
(sample sizes)
95
Recognition (Letter) Acuity: Sample sizes for baseline letter acuity measurements
were 94,28, and 19 for the control, myopic/mixed, and hyperopic groups, respectively.
A scatter plot of individual subject data and a plot of group means is provided in Figiu'e
15. For letter acuity, higher numbers represent poorer acuity, i.e., thresholds of higher
logMAR values (minimum angle of resolution) represent poorer acuity. Mean acuity was
0.03 logMAR (SD 0.14) for the control group, 0.20 logMAR (SD 0.16) for the
myopic/mixed group, and 0.25 logMAR (SD 0.22) for the hyperopic group. These mean
logMAR values correspond to Snellen values of 20/21 for the control group, 20/32 for the
myopic/mixed group, and 20/36 for the hyperopic group. A one-way ANOVA yielded
significant main effect of group (F(2,141) = 26.76, p < 0.001). Post-hoc analyses
indicated that mean acuity was significantly reduced for the myopic/mixed and hyperopic
groups in comparison to the control group (ps < 0.001), but the two astigmatism groups
did not significantly differ from each other.
Resolution (Grating) Acuity: Sample sizes for baseline grating acuity
measurements were 94, 28, and 18 for the control, myopic/mixed, and hyperopic groups,
respectively. Means for each stimulus orientation by group are provided in Table 3, and
scatter plots of individual subject data and a plot of group means is provided in Figure 16.
For measures of grating acuity, higher numbers represent better grating acuity, i.e.,
thresholds of more cycles (1 cycle = one black and one white stripe) per degree of visual
angle represent finer acuity. A repeated measures group by orientation ANOVA yielded
a significant interaction between orientation and group (F(4, 272) = 3.08, p < 0.02). Post
hoc analyses indicated that horizontal, vertical and oblique grating acuity for the
96
•
A
a
+
0
o
a
*
*
A
#
#
*
-f
*
*
*
*
Myopic
Hyperopic
A
#
*
a
Control
19
Control
Myopc/Mixed
Hyperopic
Figure IS. Baseline recognition (letter) acuity by group. Individual data are plotted in A,
and means +/- 1 standard error are plotted in B. In A, each line on the symbols represents
data from one subject.
Table 3. Baseline grating acuity means by group.
Stimulus
Mean
Standard
N
GROUP
(log cy/deg)
Deviation
Control
1.5569
0.15765
94
Myopic/Mixed
1.4794
0.12943
28
Hyperopic
1.4214
0.20543
18
Total
1.5240
0.16595
140
Control
1.5500
0.15015
94
Myopic/Mixed
1.4014
0.15975
28
Hyperopic
1.4400
0.13565
18
Total
1.5061
0.16239
140
Control
1.5337
0.13965
94
Myopic/Mixed
1.3883
0.15321
28
Hyperopic
1.3437
0.11067
18
Total
1.4802
0.15860
140
Orientation
Vertical
Horizontal
Oblique
COfW
Myope
HyptfOpc
D
-OLJSZ
I
A vwut
I
> N«t0U
I
o omm*
CO(«Ol
Myope
Hypwopc
Figure 16. Baseline resolution (grating) acuity by group. Individual data are
plotted in A, B, and C for vertical, horizontal and oblique stimuli, respectively. Means
+/-1 standard error are plotted in D.
99
myopic/mixed group was significantly poorer than for the control group. The difference
between vertical and horizontal grating acuity was greater than this difference in the
control group, although this comparison reached statistical significance before, but not
after correction for multiple comparisons. Oblique grating acuity for the hyperopic group
was significantly poorer than the control group, and while vertical and horizontal grating
acuity were poorer in the hyperopic group than in the control group, these effects reached
statistical significance before, but not after correction for multiple comparisons. The
difference between vertical and horizontal grating acuity for the hyperopic group did not
significantly differ from the control group. Myopic/mixed and hyperopic groups did not
significantly differ on vertical, horizontal or oblique grating acuity, nor did they differ in
the difference between vertical and horizontal grating acuity.
Vernier acuity: Sample sizes for baseline vernier acuity measurements were 93,
27, and 19 for the control, myopic/mixed, and hyperopic groups, respectively. Means for
each stimulus orientation by group are provided in Table 4, and scatter plots of individual
subject data and plots of means by group are provided in Figure 17. For vernier acuity,
higher scores represent poorer acuity, i.e., thresholds of greater size (more seconds of arc)
represent poorer vernier acuity. A group by orientation ANOVA yielded a significant
main effect of group (F(2,135) = 11.83, p < 0.001), but the main effect of orientation and
group by orientation interaction was not significant. Post hoc analyses indicated that
mean horizontal, vertical and oblique vernier acuity for the myopic/mixed group and for
the hyperopic group were significantly poorer than for the control group, although the
100
Table 4. Baseline vernier acuity means by group.
Stimulus
Mean
Standard
GROUP
Orientation
N
(log sec of arc)
Deviation
Control
1.0860
0.28459
93
Myopic/Mixed
1.2258
0.25720
27
Hyperopia
1.3914
0.32326
19
Total
1.1549
0.30328
139
Control
1.0897
0.25670
93
Myopic/Mixed
1.2384
0.32278
27
Hyperopic
1.3749
0.32374
19
Total
1.1576
0.29684
139
Control
1.1811
0.24281
93
Myopic/Mixed
1.3262
0.24680
27
Hyperopic
1.4745
0.28634
19
Total
1.2494
0.26973
139
Vertical
Horizontal
Oblique
101
MyopK
A CVAVLOOl
> CVAtCOOl
o CVAOLOQI
MyocK
HvparopK;
Figure 17. Baseline vernier acuity by group. Individual data are plotted in A, B, and C
for vertical, horizontal and oblique stimuli, respectively. Means +/- 1 standard error are
plotted in D.
102
astigmatism groups did not differ from each other. The difference between vertical and
horizontal vernier acuity did not significantly differ from the control group for either the
myopic/mixed or hyperopic groups, nor did the astigmatism groups differ from each
other on this measure.
Contrast sensitivity. A summary of contrast sensitivity means is provided in
Table 5, scatter plots of individual subject data are shown in Figure 18, and plots of mean
data by group are shown in Figure 19. Higher contrast sensitivity values represent poorer
contrast sensitivity, i.e., thresholds with greater percentage contrast values represent
poorer contrast sensitivity. Separate analyses were conducted for each spatial frequency.
A group by orientation ANOVA on contrast sensitivity for low spatial frequency
stimuli (1.5 cy/deg) yielded a significant orientation by group interaction (F(2,134) =
3.14, p<0.05). The myopic/mixed group did not differ from the control group on
horizontal, vertical, or on the difference between horizontal and vertical contrast
sensitivity. The hyperopic group had significantly poorer horizontal contrast sensitivity
than the control group, but contrast sensitivity for vertical stimuli did not significantly
differ from the control group, nor did the hyperopic and control groups differ on the
difference between vertical and horizontal contrast sensitivity after correction for
multiple comparisons was applied. Contrast sensitivity for vertical stimuli did not differ
between myopic/mixed and hyperopic groups, and horizontal contrast sensitivity was
poorer for the hyperopic group than the myopic/mixed group, although these effect did
not reach statistical significance after correction was applied. The two astigmatism
groups did not differ on difference between vertical and horizontal contrast sensitivity.
103
Table 5. Baseline contrast sensitivity means by group.
Stimulus
Mean
GROUP
0.330356
92
Myopic/Mixed
-1.66076
0.219898
29
Hyperopic
-1.41164
0.452425
17
Total
-1.58505
0.333570
138
Control
-1.61797
0.321131
92
Myopic/Mixed
-1.65216
0.182413
29
Hyperopic
-1.28246
0.500514
17
Total
-1.58382
0.342926
138
Control
-1.96401
0.402125
92
Myopic/Mixed
-1.79139
0.350602
29
Hyperopic
-1.48632
0.535774
17
Total
-1.86889
0.437682
138
Control
-1.84564
0.395783
92
Myopic/Mixed
-1.64247
0.285723
29
Hyperopic
-1.41251
0.507702
17
Total
-1.74959
0.416605
138
Control
-1.18950
0.450867
92
Myopic/Mixed
-.83865
0.404703
29
Hyperopic
-.59876
0.304847
17
Total
-1.04300
0.476542
138
Control
-1.19484
0.439436
92
Myopic/Mixed
-.63934
0.319904
29
Hyperopic
-.60721
0.290817
17
Total
-1.00572
0.480830
138
6.0 cy/deg
Vertical
6.0 cy/deg
Horizontal
18 cy/deg
Vertical
18 cy/deg
Horizontal
N
-1.59322
1.5 cy/deg
Horizontal
Std. Deviation
Control
1.5 cy/deg
Vertical
(LogCS)
104
Control
Myope
Hyperopc
Control
Myooc
Hyperopic
Control
Myope
Hyperopc
?•
Control
Myope
Hyperopc
Myope
Hyperopc
Myope
Hyperopc
Figure 18. Baseline contrast sensitivity by group scatter plots. Individual data are plotted
in A and B for low spatial frequency stimuli, C and D for middle spatial frequency
stimuli, and E and F for High spatial frequency stimuli. Data for vertical stimuli are
plotted in A, C, and E, data for horizontal stimuli are plotted in B, D, and F.
105
A vtrtcai
UyOpC/MMd
Myopc(Mwn
HvVtrapC
Hvpvepc
Figure 19. Baseline contrast sensitivity means by group. Means +/- 1 standard error for
vertical and horizontal stimuli are plotted in A for low spatial frequency stimuli, B for
middle spatial frequency stimuli, and C for high spatial frequency stimuli.
106
A group by orientation ANOVA on contrast sensitivity for middle spatial
frequency stimuli (6.0 cy/deg) yielded a significant main effect of group (F(2,134) =
10.80, p<0.001), but no significant main effect of orientation or orientation by group
interaction. Both the myopic/mixed and hyperopia groups differed from the control
group on horizontal and vertical contrast sensitivity, but neither astigmatism group
differed from the control group on the difference between vertical and horizontal contrast
sensitivity. The myopic/mixed and hyperopic groups did not differ on any measure.
A group by orientation ANOVA on contrast sensitivity for high spatial frequency
stimuli (18.0 cy/deg) yielded a significant orientation by group interaction (F(2,135) =
5.82, p < 0.005).
Both the myopic/mixed and hyperopic groups differed from the
control group on horizontal and vertical contrast sensitivity, but only the myopic/mixed
group differed from the control group on V-H contrast sensitivity. The myopic/mixed
and hyperopic groups did not differ on any measure.
Stereoacuity: Sample sizes for baseline stereoacuity measurements were 95, 28,
and 18 for the control, myopic/mixed, and hyperopic groups, respectively. The mean
stereoacuity for was 1.66 log sec (SD 0.13) for the control group, 1.72 log sec (SD 0.22)
for the myopic/mixed group, and 1.91 log sec (SD 0.39) for the hyperopic group. Figure
20 plots individual stereoacuity scores and mean stereoacuity scores for each group. For
stereoacuity measures, higher scores represent poorer stereoacuity, i.e., thresholds with
larger difference (more seconds of arc) in images between eyes necessary to give rise to
the perception of depth indicate poorer stereoacuity. A one-way ANOVA yielded a
107
3.0
a
A
2.8
?
o
2 2.6
O
0]
T3
8
Si
o
a
g» 2.2
!<
o
S
o
CO
1.6
*
m
m
A
*
*
«
*
1.4
Control
m
Myopic
Hyperopic
2.8-
Control
Myopic/Mixed
Hyperopic
Figure 20. Baseline stereoacuity by group. Individual data are plotted in A, and means
+/- 1 standard error are plotted in B. In A, each line on the symbols represents data from
one subject.
108
significant main effect of group (F(2, 137) = 11.38, p < 0.001). Post-hoc analyses
indicated that stereoacuity for the hyperopic group was significantly poorer than the
control group. Other comparisons did not reach statistical significance: stereoacuity for
the myopic/mixed group was not significantly reduced in comparison to the control group
after correction for multiple comparisons, and stereoacuity for the hyperopic group was
not significantly reduced in comparison to the myopic/mixed group.
Form Adaptation. For measurements of form perception, raw data ranged from 12 to +12 in whole numbers, with negative numbers representing vertically compressed
circles, 0 representing a true circle, and positive numbers representing vertically
elongated circles. Negative and positive numbers also represent the percentage that the
stimulus is compressed or elongated, relative to a true circle, e.g., -6 represents a 6%
vertically compressed circle.
Mean scores for each wear (glasses on/glasses off) condition at baseline were
determined by averaging responses on four trials: two trials each for the outline circle and
checkered circle stimuli, one on which the child started the task on the -12 stimulus, and
one on which the child started the task on the +12 stimulus.
A group by wear repeated measures ANOVA was conducted on baseline data.
Figure 21 illustrates the predicted pattern of results, along with the observed pattem of
results. At baseline, I predicted that there would be no difference in mean response for
the control group with glasses-on vs. glasses-off (since there is little or no cylinder in the
lenses), and that the astigmatism group would not differ fi'om the control group in the
glasses-off condition at baseline. However, the astigmatism group, on average, should
109
0
(0
c
Reduction of Effect
2
(0
« 0
a " "O"
c
s>
i -2
Negative After Effect
o
Baseline
1 Month
f-
Baseline
1 Month
Control Group, Glasses On
Control Group. Glasses Off
Astigmatism Group, Glasses On
Astigmatism Group, Glasses Off
Figure 21. Predicted and observed form perception by time plots. Predicted patterns are
shown in A, observed means +/- 1 standard error are shown in B.
110
tend to choose more vertically elongated stimuli as "the perfect circle", since cylinder
lenses that correct for with-the-rule astigmatism vertically compress images.
The results, shown in the lower portion of Figure 21, indicated that there was a
significant interaction between group and wear (F(l,65)=9.66, p=0.003). Post-hoc
comparisons indicated mean responses for the control group did not differ between the
glasses-on and glasses-off condition. In contrast, the astigmatism group did significantly
differ between glasses-on and glasses-off conditions, such that the children tended to
choose more vertically elongated stimuli in the glasses-on condition as predicted.
2.3.3 Outcome Measures: Plasticity over I Month and 1 Year
Analyses in this section focus on evaluating change in vision from baseline to I
month and fi'om baseline to 1 year for the astigmatism groups in comparison to the
control group. This section provides details of the results of statistical analyses. A
general summary and interpretation is provided at the end of the Results Section.
Group by time repeated measures ANOVAs were conducted, and the reports that
follow focus on the significance of the interaction between group and time. Main effects
of group are not of primary interest, as differences among groups were reported at
baseline. In addition, main effects of time are of limited interest for the purpose of this
report because main effects of time, without significant interactions between time and
group, would simply indicate that there were improvements across groups due to visual
development or, more likely in the case of 1 month data, improvements due to practice
effects. Developmental effects were evaluated in previous analyses.
111
Summaries of analyses are provided in Table 6 for 1 month data and Table 7 for 1
year data. In addition. Figures 22-26 plot mean change for each measure from baseline to
1 month, and from baseline to 1 year.
Recognition (Letter) Acuity. Both the myopic/mixed and hyperopic groups had
significantly poorer mean acuity than the control group at baseline. To determine if there
was improvement with 1 month of glasses wear, a group (control, myopic/mixed,
hyperopic) by time (baseline vs. 1 month) repeated measures ANOVA on letter acuity
data was conducted. The results indicated that there was no significant interaction
between group and time: the amount of change that occurred from baseline to 1 month
did not differ across groups.
A group (control, myopic/mixed, hyperopic) by time (baseline vs. I year)
repeated measures ANOVA on letter acuity data was conducted to determine if there
were improvements with I year of glasses wear. The results indicated that there was no
signiflcant interaction between group and time: the amount of change that occurred from
baseline to 1 year did not differ across groups.
Resolution (Grating) Acuity. Baseline analyses indicated that the myopic/mixed
group had significantly poorer grating acuity for vertical, horizontal and oblique lines, in
comparison to the control group, and the hyperopic group had significantly poorer grating
acuity for oblique lines, in comparison to the control group. Vertical and horizontal
grating acuity for the hyperopic group in comparison to the control group approached but
did not reach significance. To determine if there was improvement with 1 month of
glasses wear, a group (control, myopic/mixed, hyperopic) by time (baseline vs. 1 month)
112
Table 6. One Month Data Analysis Summary Table. All analyses conducted with age as
a covariate to control for age (i.e., developmental differences) across groups. Sample
sizes for the control, myopic/mixed, and hyperopia groups, respectively, provided in
parenthesis.
3 Groups ANOVAs
Group X Time
Interaction
Measure
(Sample Sizes)
Letter Acuity (91,28,19)
V
Grating Acuity
H
(91,28,17)
0
V-H
V
Vemier Acuity
H
(89,26,17)
O
V-H
V
Contrast
H
Sensitivity
1.5 cy/deg
V-H
Contrast
Sensitivity
6.0 cy/deg
Contrast
Sensitivity
18.0 cy/deg
V
H
V-H
V
H
V-H
Stereoacuity (92,28,18)
NS (0.27)
Group X
VNS
Orient, x
HNS
time
ONS
NS
Group X
VNS
Orient, x
HNS
Time
ONS
NS
Group X
Orient, x
Time
NS
Group X
Orient, x
Time
NS
Group X
VNS
Orient, x
H • (0.05)
Time
NS
•(=0.015)
Post-Hoc Tests:
Grou p X Time Interactions
Control
Myopic/
vs.
Control vs.
Mixed vs.
Myopic/
Hyperopic
Hyperopic
Mixed
*
(0.048)
NS
(0.073)
NS (0.61) • (=0.007)
NS
(0.64)
NS
(=0.07)
* Statistically significant before Bonferroni correction applied (p values represent
uncorrected significance level). V = Vertical Stimuli, H = Horizontal Stimuli, O =
Oblique Stimuli, V-H = Vertical - Horizontal, NS = not statistically significant.
113
Table 7. One Year Data Analysis Summary Table. All analyses conducted with age as a
covariate to control for age (i.e., developmental differences) across groups. Sample sizes
for the control, myopic/mixed, and hyperopic groups, respectively, provided in
parenthesis.
Measure
(Sample Sizes)
3 Groups ANOVAs
Group x Time
Interaction
Post-Hoc Tests:
Group X Time Interactions
Control
Myopic/
vs.
Control vs.
Mixed vs.
Myopic/
Hyperopic
Hyperopic
Mixed
NS (=0.52)
Letter Acuity (56,17,9)
V
VNS
Group X
Grating Acuity
H
HNS
Orient, x
(55,17,8)
0
Time NS
ONS
V-H
V
Group X
VNS
Vernier Acuity
H
Orient, x
HNS
(54,16,9)
0
Time NS
ONS
V-H
V
Contrast
Group X
VNS
Sensitivity
Orient, x
* (0.006)
• (0.028)
H
NS (0.97)
H • (0.013)
1.5 cy/deg
Time NS
V-H
V
Contrast
Group X
VNS
Sensitivity
Orient,
x
• (0.026)
H
NS (0.77)
* (0.005)
H • (0.012)
6.0 cy/deg
Time NS
V-H
V
Contrast
Group X
VNS
Sensitivity
Orient, x
H
HNS
18.0 cy/deg
Time NS
V-H
•(p=0.017)
Stereoacuity (57,17,9)
* (0.005)
NS (0.14)
NS (0.85)
* Statistically significant before Bonferroni correction applied (p values represent
uncorrected significance level). V = Vertical Stimuli, H = Horizontal Stimuli, O =
Oblique Stimuli, V-H = Vertical - Horizontal, NS = not statistically significant.
114
c
9
to
S
I9
S9
oC
S
.05
Control
MyopioMued
Hyperopc
Control
Myopic/Maed
Myp^fOpC
Figure 22. Mean change in recognition (letter) acuity by group. Mean +/- 1 standard
error of differences between baseline and 1 month follow-up measures are shown in A,
and mean +/- 1 standard error of differences between baseline and 1 year follow-up
measures are shown in B.
115
» 00
a Veftical
> Honzoniai
O Ottique
Controi
Myopic/Maed
Hyperopc
A Verticai
> Honzontai
O Oblique
Control
Myopic/Mutd
Hyperopic
Figure 23. Mean change in resolution (grating) acuity by group. Mean +/- I standard
error of differences between baseline and I month follow-up measures are shown in A,
and mean +/-1 standard error of differences between baseline and 1 year follow-up
measures are shown in B.
116
A Vertical
> Honzoraal
0 OMque
Control
Myope/Mixed
Hyperoptc
I
^ Vertical
o
CO
> Honzontal
O Oblique
Control
Hyperoptc
Figure 24. Mean change in vernier acuity by group. Mean +/-1 standard error of
differences between baseline and 1 month follow-up measures are shown in A, and mean
+/-1 standard error of differences between baseline and 1 year follow-up measures are
shown in B.
117
tSCy/Otg
..
9
k
i
f•I
I
3
c
5K
•
'
,
.
'
B
I'
..
^
3
ISCy/0«
XX rr
I
I
& imisMM
S'
. a
C
1>
aocy/txg
-
I '
3 .
H
5a ^
«
2
IT
SOCy/Dao
I
I
3t '
i
I
16 0 CyoMO
lBOCy/D«g
Figure 25. Mean change in contrast sensitivity by group. Mean+/-1 standard error of
differences between baseline and follow-up measures are shown in A and B for low
spatial frequency stimuli, C and D for middle spatial frequency stimuli, and E and F for
high spatial fi-equency stimuli. Change fi'om baseline to 1 month are plotted in A, C, and
E, and change from baseline to 1 year are shown in B, D, and F.
118
Conlrol
Myopic/Mixed
Hyperopic
Conlrol
Myopic/Mixed
Hyperopic
Figure 26. Mean change in stereoacuity by group. Mean +/-1 standard error of
differences between baseline and 1 month follow-up measures are shown in A, and mean
+/-1 standard error of differences between baseline and 1 year follow-up measures are
shown in B.
119
rq)eated measures ANOVA on grating acuity data for each stimulus orientation was
conducted. The results of the analyses yielded no significant interactions between group
and time, indicating that there were no differences in the amount of change observed
from baseline to 1 month across groups.
To determine if there was improvement with 1 year of glasses wear, a group
(control, myopic/mixed, hyperopic) by time (baseline vs. 1 year) repeated measures
ANOVA on grating acuity data for each stimulus orientation was conducted. The results
of the analyses yielded no significant interactions between group and time, indicating that
there were no differences in the amount of change observed from baseline to 1 year
across groups.
Vernier acuity. Baseline measures indicated that mean horizontal, vertical and
oblique vernier acuity for the myopic/mixed group and for the hyperopic group were
significantly poorer than the control group. To determine if there was improvement with
1 month of glasses wear, a group (control, myopic/mixed, hyperopic) by time (baseline
vs. 1 month) repeated measures ANOVA on vernier acuity data for each stimulus
orientation was conducted. The results of the analyses yielded no significant interactions
between group and time, indicating that there were no differences in the amount of
change observed from baseline to 1 month across groups.
To determine if there was improvement with 1 year of glasses wear, a group
(control, myopic/mixed, hyperopic) by time (baseline vs. 1 year) repeated measures
ANOVA on vernier acuity data for each stimulus orientation was conducted. The results
of the analyses yielded no significant interactions between group and time, indicating that
120
there were no differences in the amount of change observed from baseline to 1 year
across groups.
Contrast Sensitivity. Group by time ANOVAs were conducted on 1.5,6.0, and
18,0 cy/deg contrast sensitivity for horizontal and vertical stimuli to determine if there
were improvements from baseline to 1 month. The analyses yielded no significant
interactions (group x time) for 1.5 and 6.0 cy/deg contrast sensitivity for horizontal or
vertical stimuli, and no significant interaction for 18 cy/deg contrast sensitivity for
vertical stimuli. However, the interaction was marginally significant for 18 cy/deg
contrast sensitivity for horizontal stimuli (F(2,131)=3.06, p=0.05). Post-hoc analyses
indicated that there was greater improvement for the myopic/mixed group in comparison
to the control group, although this effect did not reach significance after correction for
multiple comparisons. The difference between the hyperopic group and the control group
did not reach significance.
Group by time ANOVAs were also conducted on 1.5,6.0, and 18.0 cy/deg
contrast sensitivity for horizontal and vertical stimuli to determine if there were any
significant improvements from baseline to 1 year. The analyses yielded no significant
interactions (group x time) for vertical stimuli at any of the three spatial frequencies.
However, the interaction was significant for horizontal stimuli at 1.5 (F(2,76)=4.63, p <
0.02) and 6 cy/deg (F(2,76)=4.73, p = 0.02), but not at 18 cy/deg. Post-hoc analyses
indicated that for both 1.5 and 6.0 cy/deg horizontal stimuli, there was significantly
greater improvement for the hyperopic group, but not the myopic/mixed group, in
121
comparison to the control group. Comparisons between the hyperopic and myopic/mixed
group were significant before, but not after correction for multiple comparison.
Stereoacuity. At baseline, the hyperopic group had significantly poorer mean
stereoacuity than the control and myopic/mixed groups. To determine if there was
improvement with 1 month of glasses wear, a group (control, myopic/mixed, hyperopic)
by time (baseline vs. 1 month) repeated measures ANOVA on stereoacuity data was
conducted, and the analysis yielded a significant interaction (F(2,133)=4.37, p < 0.02).
Post-hoc analyses indicated that there was significantly greater improvement from
baseline to 1 month for the hyperopic group in comparison to the control group. The
amount of change in the myopic/mixed group did not differ from that observed in the
control group or the hyperopic group.
To determine if there was improvement with 1 year of glasses wear, a group
(control, myopic/mixed, hyperopic) by time (baseline vs. 1 year repeated measures
ANOVA on stereoacuity data was conducted, and the analysis yielded a significant
interaction (F(2,80) = 4.3, p< 0.02). Post-hoc analyses indicated that there was
significantly greater improvement from baseline to 1 year for the hyperopic group in
comparison to the control group. The amount of change in the myopic/mixed group did
not differ from that observed in the control group or the hyperopic groups.
Form Adaptation. Mean scores for each wear (glasses on/glasses off) x time
(baseline/1 month) condition were determined by averaging responses on four trials: two
trials each for the outline circle and checkered circle stimuli, one on which the child
started the task on the -12 stimulus, and one on which the child started the task on the
122
+12 stimulus. As previously noted, raw data ranged from -12 to +12 in whole numbers,
with negative numbers representing vertically compressed circles, 0 representing a true
circle, and positive numbers representing vertically elongated circles
Since analysis of baseline data did suggest that the eyeglasses do induce distorted
perception for the astigmatism group, analyses were aimed at determining if subjects in
the astigmatism group adapted to the distortion after 1 month of wear. Figure 21
illustrates the predicted pattern of results and the observed pattern of results. Adaptation
was determined both in terms of reduction of effect, based on changes in perception when
glasses were on, and in terms of negative aftereffects, based on changes in perception
when glasses were off. Separate repeated measures ANOVAs were conducted on data
obtained when glasses were on, and data obtained when glasses were off The analyses
yielded no main effect of time, and no significant interactions between group (control vs.
astigmatism) and time (baseline vs. 1 month), indicating that there was no difference
between groups in the pattern of form perception measures at baseline vs. follow-up.
2.3.4 General Summary and Interpretation of Results
Baseline Data Analyses: Normal Development in 5- to 14-year-olds. Results
suggest that, with the exception of stereoacuity, significant development in basic
perceptual functions occurs in the S- to 14-year-old age range. Recognition acuity,
resolution acuity, vernier acuity, and contrast sensitivity were significantly correlated
with age, and the correlation between age and stereoacuity approached significance.
Closer examination of change across age groups revealed differences in patterns of
development across perceptual measures. Developmental data revealed that recognition
123
acuity in this age range develops significantly between 7/8 year-old and 9/10 year-old age
groups. Grating acuity develops gradually between age 7/8 year-old and 11- to 14-yearold age groups, and vernier acuity develops gradually between 5/6 year-old and 11- to
14-year-old age groups. Contrast sensitivity, like resolution acuity, appears to develop
notably between the 7/8 year-old and 9/10 year-old age groups, as difference between
these groups reached or approached significance for low, middle, and high spatial
fi'equency stimuli.
Baseline Data Analyses: Ejfects of Deprivation and Spatial Distortion on
Perception. Results of analyses suggested that astigmatism-related deprivation had
significant effects on development of all measures of perception studied here, and that
wear of cylinder lenses resulted in measurable distortion in form perception.
Recognition (Letter) Acuity. Both myopic/mixed and hyperopic
astigmatism groups had significant deficits for resolution acuity (in comparison to
the control group). These results were as predicted, and suggest that astigmatismrelated deprivation results in deficits in recognition acuity.
Resolution (Grating) Acuity. The myopic/mixed group had significant
deficits for perception of horizontal, vertical and oblique grating acuity stimuli,
and showed greater deficits for horizontal than for vertical stimuli (this effect
approached but did not reach significance). The hyperopic group showed
significant deficits for perception of oblique grating acuity stimuli, showed
deficits for perception of horizontal and vertical stimuli that neared significance,
and showed no evidence of orientation dependent (horizontal vs. vertical)
124
differences in resolution acuity deficits. These data were as predicted patterns for
the myopic/mixed group: there appeared to be greater perceptual deficits for the
stimulus orientation for which subjects experienced greatest deprivation, i.e.,
horizontal stimuli. Data from the hyperopic group were not quite as predicted,
and show essentially equivalent deficits across stimulus orientation. Predictions
were that there would be greater deficits for vertical than for horizontal stimuli,
since these subjects presumably experienced greater deprivation for vertical
stimuli.
Vernier Acuity. Both myopic/mixed and hyperopic astigmatism groups
had significant deficits for perception of horizontal, vertical and oblique vernier
acuity stimuli, although no orientation differences (horizontal vs. vertical) in
perceptual deficits were observed in either astigmatism group. Patterns of deficits
across stimulus orientation were not as predicted for either the myopic/mixed
astigmatism group, in which it was predicted that there would be greater deficits
for perception of horizontal stimuli, nor for the hyperopic astigmatism group, in
which it was predicted that there would be greater deficits for perception of
vertical stimuli.
Contrast Sensitivity. For low spatial frequency contrast sensitivity, the
myopic/mixed group demonstrated no significant deficits, while the hyperopic
group showed significant deficits for only horizontal stimuli, and a marginally
significant difference in perception of horizontal and vertical stimuli. For the
hyperopic group, these results were the opposite of the predicted pattern of
125
results: it was expected that perception of vertical stimuli would be poorer than
for horizontal stimuli. These results also did not support the prediction that
myopic/mixed astigmats would show poorer perception for horizontal than
vertical stimuli.
For middle spatial frequency contrast sensitivity, both the myopic/mixed
and hyperopic astigmatism groups showed significant deficits for perception of
both horizontal and vertical stimuli, and no difference between perception across
stimulus orientation. These results suggest that astigmatism-related deprivation
influences perception of middle range spatial fi-equency stimuli, but they do not
support the prediction that myopic/mixed astigmats would show poorer
perception for horizontal stimuli, and hyperopic astigmats would show poorer
perception for vertical stimuli.
For high spatial frequency stimuli, both the myopic/mixed and hyperopic
astigmatism groups showed significant deficits for perception of horizontal and
vertical stimuli, and the myopic/mixed group showed greater deficits for the
horizontal stimuli than for vertical stimuli. These results suggest that astigmatismrelated deprivation influences perception of high spatial frequency contrast
sensitivity, and support the prediction that myopic/mixed astigmats would show
poorer perception for horizontal than for vertical stimuli. However, the data did
not lend support to the prediction that hyperopic astigmats would show
demonstrated poorer perception for vertical than for horizontal stimuli.
126
Stereoacuity. Deficits for stereoacuity were significant for the hyperopic
group, and approached but did not reach significance for the myopic/mixed group.
These results suggest that astigmatism-related deprivation influences stereoacuity
in hyperopic astigmats, but not myopic/mixed astigmats.
Form Perception. Comparisons in baseline form perception measures
indicated mean responses for the control group did not differ between the glasseson and glasses-off condition, hi contrast, the astigmatism group did significantly
differ between glasses-on and glasses-off conditions, such that the children tended
to choose more vertically elongated stimuli in the glasses-on condition. In
general, this pattern is consistent with what was predicted, and suggests that this
task is in fact measuring the distortion experienced when the cylinder lenses are
worn by the children in the astigmatism group. However, one aspect of the
pattern of the results is not consistent with the original prediction: I predicted that
the children in the astigmatism group would not differ from the control group
when they were not wearing their glasses, but they would differ from the control
group when they were wearing their glasses. However, the results yielded the
opposite pattern of results: the astigmatism group did not differ from the control
group when they were wearing their glasses, but differed from the control group
when they were not wearing their glasses. These results suggests that the children
in the astigmatism group perceive the circles differently than the control group
when they are not wearing their glasses, i.e., the perceive vertically reduced ovals
as more circular, in comparison to the control group who reported that stimuli
127
close to real circles were most circle-like. However, as previously noted, data
fix)m the astigmatism group did show the glasses do resuh in changes in form
perception in the direction predicted by the distortion induced by the cylinder
lenses.
Outcome Analyses: Plasticity over I month and I year. The data revealed little
evidence for plasticity with regard to recovery from the effects of astigmatism-related
deprivation. After 1 month, there was evidence for significant improvement only for
stereoacuity in the hyperopic astigmatism group. After 1 year, there was evidence of
significant improvement in perception of low and middle range spatial frequency contrast
sensitivity for horizontal stimuli in the hyperopic astigmatism group, and significant
improvement in stereoacuity for the hyperopic group. Improvement in contrast
sensitivity for horizontal stimuli is an unexpected result, as ware baseline findings that
indicated that the hyperopic group demonstrated significant deficits for perception of
horizontal contrast sensitivity stimuli: predictions suggested that there would be greater
deficits for vertical than for horizontal stimuli in this group.
The results indicated that there was no difference among groups in the pattern of
form perception at baseline vs. 1 month follow-up across the astigmatism and control
groups. Thus, these data do not provide any indication that the subjects in the
astigmatism group adapted to the distortion induced by the cylinder lenses.
128
3. DISCUSSION
The study presented here offers several contributions to the literature on visual
development, deprivation, recovery from deprivation, and perceptual adaptation.
Baseline analyses provide a comparison of recognition acuity, resolution acuity, vernier
acuity, contrast sensitivity, and stereoacuity development in between the ages of 5 to 14
years, and a large-sample description of the effects of astigmatism-related deprivation on
visual development in children. Primary outcome measures provided a prospective
analysis of plasticity with regard to recovery from the effects of astigmatism-related
visual deficits in children and adaptation to spatial distortion in children. In what
follows, I discuss the findings related to each of these aspects of the present study in
greater detail.
3.1 Normal Visual Development in Grade-School Children
Analysis of baseline data in for non-astigmatic children provided evidence for
development in several basic perceptual functions between ages S and 14 years.
Developmental analyses suggested that there was a significant relation between age and
visual performance on all measures, with the exception of stereoacuity, which
approached but did not reach significance. These results indicate that letter acuity,
grating acuity, vernier acuity, and contrast sensitivity continue to develop in this age
range.
129
Analyses aimed at narrowing down the grade-school age range during which most
development occurs for each perceptual function yielded consistent results for letter
acuity and contrast sensitivity: most development occurred between ages 7 and 10.
Grating acuity also appeared to have a period of greater development starting around age
7, and continuing to age 11-14. In contrast, vernier acuity appeared to develop more
gradually across the 5- to 14- year age range, and stereoacuity appeared to develop
slowly between S and 10 years. Unfortunately, these data can not tell us if subjects
reached adult levels, as adult subjects were not included in the present study.
It is interesting that for several measures, i.e., letter acuity, grating acuity, and
contrast sensitivity, the age range from 7 to 10 seems to be associated with marked
improvements. These data can not provide clear answers as to why development
increases after a relatively slow period between age 5 and 7. However, one possibility is
that the type of visual experience children are exposed to at different ages may influence
these developmental patterns. For example, as children make their way through grade
school, it is likely that they spend increasing amounts of time doing near work that
requires good vision for detail, e.g., letters and numbers. Perhaps this experience
provides a type of discrimination learning, where vision improves as a result of "practice"
in discriminating fine perceptual stimuli. The early slow period of grade-school visual
development, age S to 7 years, includes the kindergarten and first grade years, during
which children are first learning to read, but tend to do so with large letter stimuli. The
rapid period of development, age 7 to 9 years, include second and third grade, when
children are becoming more adept at reading, and are spending more time reading with
130
relatively smaller letter stimuli. Thus, the general pattern of results obtained here are at
least consistent with an experiential effect on pattems of visual development.
Freeman (1978) reported some interesting findings regarding configurational
differences in visual acuity that may be related to the developmental pattems observed
here, and may provide an example of how typical visual experience beyond early
childhood can lead to discrimination leaming. Freeman reported differences in letter
visual acuity based on the configuration of the test display; visual acuity was significantly
better for letters presented in rows than for letters presented in columns. In addition, he
provided experimental evidence to support the hypothesis that these acuity differences
are experience dependent. First, he reported that orientation differences were not
obtained for non-letter acuity stimuli (Landolt C). Second, native Chinese readers
demonstrated no such acuity difference for Chinese characters read in rows vs. columns,
a finding that is particularly significant to the experience-dependent hypothesis because
Chinese is printed and read in either rows or columns. Finally, young children who could
identify letters but not yet read also failed to demonstrate acuity differences for letters
presented in rows vs. columns. Freeman suggested that these data provide evidence that
specific visual experiences, particularly reading, influence visual resolution. I believe
these data provide further support for the hypothesis that pattems of development in
grade-school reflect the influence of visual experience during this period of development.
While the goal of these developmental analyses was to better understand the
development of visual perception, it is important to note that another possible
interpretation of any age effects is that older children tend to perform better on the tasks
131
due to attentional or motivational factors, or may have more liberal criteria for
responding in these detection tasks once they approach threshold. Significant attempts
were made to reduce the cognitive demands of the tasks used for assessment and to
engage the children in the task, and thus the likelihood that this alternative explanation
can account for the developmental effects observed here is reduced, but can not be
completely ruled out.
Overall, these data indicate that some development of these visual functions still
occurs in this age range. Furthermore, the patterns of development indicate that visual
experiences, such as reading, may influence development during this age range,
suggesting the use of discrimination leaming strategies during this age range may have
notable effects on development, and perhaps may be useful in the treatment of
developmental visual disorders, such as recovery from the effects of early visual
deprivation. This possibility will be discusses further in the following sections.
3.2 Effects of Astigmatism-Related Deprivation on Visual Development
Baseline comparisons between astigmatism and control groups contributed
important data on the extent to which various visual perceptual functions are influenced
by the presence of high astigmatism during development. Vision was measured under
conditions in which each child wore his or her best optical correction. Thus, any deficits
detected could not be attributed to optical effects, and therefore must originate fi'om
higher levels of the visual system.
132
Baseline analyses indicated that the presence of astigmatism during development
was associated with reduced perceptual capabilities on all measures of visual function
reported here: recognition acuity, resolution acuity, vernier acuity, contrast sensitivity,
and stereoacuity. In what follows, I discuss the implications of these findings in the
context of previous research, and potential for further research.
Letter acuity was significantly reduced for both the myopic/mixed and hyperopic
astigmatism groups. In the control group, mean acuity was approximately 20/20,
whereas mean acuity in the astigmatism groups was approximately 20/32. Because letter
acuity requires fme resolution acuity in addition to the ability to identify fme letter
stimuli, it is not clear if reduced letter acuity is associated only with deficits in resolution
acuity, or if it may also be associated with deficits in higher level form perception.
Previous research has indicated that other forms of amblyopia are associated with higher
level perceptual deficits including form perception even when stimuli are presented
above resolution acuity limits (Hess, Campbell, and Greenhalgh, 1978, Hess, Wang,
Demanins, Wilkinson, and Wilson, 1999, Lewis, Ellemberg, Maurer, Wilkinson, Wilson,
Dirks, and Brent, 2002). Research is currently is progress to determine if astigmatismrelated deprivation results in deficits in perception of global form.
Grating acuity results indicated that the myopic/mixed group showed greater
deficits for horizontal than for vertical stimuli. These results are consistent with data
firom adults indicating orientation-dependent deprivation can result in reduced
visual/perceptual capabilities for stimuli of the deprived orientation (Freeman, Mitchell,
and Millodot, 1972, Mitchell, Freeman, Millodot, and Haegerstrom, 1973). Uncorrected
133
myopic/mixed with-the-rule astigmats experience greater defocus for horizontal lines
than for vertical lines when viewed at distance (see Figure 2d-0. and these data indicate
that reduced vision persists even after the optical cause for reduced vision, astigmatism,
is corrected. Dobson et al. (2002) reported the same pattern of results for preschoolers
with myopic/mixed astigmatism.
Grating acuity results for the hyperopic astigmatism group indicated that these
children have equally reduced acuity for stimuli across orientations. I should note,
however, that deflcits for horizontal and vertical stimuli did not reach statistical
significance. The smaller sample size and greater variability (possibly due to younger
mean age) in this group resulted in lower statistical power, and is likely to have been an
important reason that statistical significance was not reached. However, this pattern of
results is consistent with the results of Dobson et al. (2002), who reported equally
reduced acuity for horizontal and vertical stimuli in preschoolers with hyperopic withthe-rule astigmatism. Mitchell et al., (1973), however, predicted and reported reduced
acuity for stimuli corresponding to the more hyperopic focus (vertical, in the case of our
sample) in hyperopic astigmats, reasoning that uncorrected hyperopic astigmats would be
likely to acconmiodate to focus the nearer focal point (horizontal, in the case of our
subjects), and thus would experience greater deprivation for vertical stimuli. Freeman
(1975) provided support for this prediction when he reported that measurements of
accommodation in a subject with hyperopic astigmatism indicated that he accommodated
to focus stimuli at the least hyperopic meridian. However, the results presented here, and
those reported by Dobson et al. (2002) suggest that hyperopic astigmatic children do not
134
accommodate in the manner that Mitchell et al. predicted for hyperopic astigmats. The
observed reduced acuity across orientations suggests that these children, when
uncorrected, may accommodate someplace between the anterior and posterior focal
points, or may fluctuate accommodation, thus resulting in visual experience that does not
provide them with consistently clear input for stimuli of any orientation (Dobson et al.,
2002). We are currently conducting a study to determine where hyperopic astigmats
focus for distant and near targets when astigmatism is uncorrected. This research should
provide us with better understanding of the nature of the visual experience of uncorrected
hyperopic astigmats, and the resulting visual/perceptual deficits reported here.
Vernier acuity was significantly reduced for stimuli of all three orientations for
both the myopic/mixed and hyperopic astigmatism groups. These data suggest that
astigmatic defocus induces deficits in perception of flne spatial relationships. It is not
clear why deficits for vernier acuity were not dependent upon stimulus orientation.
Perhaps resolution of vertical and horizontal vernier acuity stimuli require clear
perception of both horizontal and vertical stimulus information because while the overall
stimulus is one orientation (e.g., horizontal), the offset to be detected is the orthogonal
orientation (vertical). However, previous studies have reported orientation dependent
deficits in subjects with high astigmatism (Mitchell et al., 1973, Gwiazda et al., 1986).
Contrast sensitivity deficits for astigmatic subjects varied across spatial
frequencies. For low spatial frequency stimuli (1.5 cy/deg), deficits were apparent only
for the hyperopic astigmatism group, and only for horizontal stimuli. For middle and
high spatial frequency stimuli, both astigmatism groups had significant deficits for
135
percq>tion of both horizontal and vertical stimuli, and for high spatial frequency stimuli,
the myopic/mixed group showed significantly greater deficits for horizontal than for
vertical stimuli. Thus, the results for high spatial frequency contrast sensitivity mirrored
the grating acuity results for the myopic/mixed group, providing further evidence that the
deprivation was greater for the more myopic stimulus orientation (horizontal). It is not
clear why there were deficits for low spatial frequency horizontal stimuli for the
hyperopic group but high and middle range contrast sensitivities were poor for both
horizontal and vertical stimuli, further suggesting that hyperopic astigmats may not
habitually bring the more anterior focal line (horizontal stimuli) into focus through
accommodation.
In general, the contrast sensitivity results provide us with a greater understanding
of the how deficits associated with astigmatism-related deprivation may influence
perception in real-world circumstances. The fact that deficits were observed for high
spatial frequency stimuli is not surprising given the reduced recognition and resolution
acuity in astigmatic subjects - the astigmatic children clearly have difficulty perceiving
fine visual detail even with the appropriate optical correction. However, deficits
observed for mid range spatial frequency stimuli suggests that visual perceptual deficits
may extend beyond fine perceptual tasks, and may influence perception of their general
environment to a greater extent than we might have expected based solely on recognition
and resolution acuity results. In particular, deficits for both high and mid range spatial
frequency information under low contrast conditions are likely to result in difficulty in
136
differentiating much of our visual environment, which includes a significant amount of
low contrast visual information.
Stereoacuity was significantly reduced only for the hyperopic group. It is not
clear why this group in particular showed such reduced stereoacuity. Children with
strabismus are at risk for poor stereoacuity. While strabismus was not detected in any of
the children in the hyperopic astigmatism group, children with high hyperopia are at risk
for a form of strabismus called accommodative esotropia. Thus, it is possible that some
of the children with hyperopic astigmatism may have had mild accommodative esotropia
that was not detected at the eye exam, but may have resulted in the small reductions in
stereoacuity observed here.
It is not clear if stereo deficits are associated with the spatial distortions induced
by the cylinder lenses that correct for the optical effects of astigmatism, or if these are
higher level deficits in stereoacuity. Since use of contact lenses to correct astigmatism
minimizes spatial distortions induced by the lens curvature, it is possible to evaluate
whether stereo deficits are present in the absence of spatial distortion in astigmatic
subjects. Unfortunately, due to time and financial constraints, and the fact that putting
contact lenses on children would be both invasive and would be significantly complicated
in terms of creating a sterile environment within the school setting, this was not evaluated
in the present study. However, the present findings of reduced stereoacuity in astigmats
provide support for continuing to pursue answers to such questions in further studies.
137
3.3 Plasticity Associated with Recovery from Effects of Deprivation
The primary question addressed in this study was the extent to which 5- to 14year-old children demonstrate plasticity with respect to recovery from the effects of
astigmatism-related deprivation and with respect to adaptation to spatial distortion.
Results regarding recovery from deprivation were generally negative. On average,
glasses wear over a I month period and over a 1 year period did not result in significant
improvements in letter acuity, grating acuity, or vernier acuity. Significant improvements
were seen at 1 month and 1 year for the hyperopic group in terms of stereoacuity,
although it is not clear if those improvements were due to adaptation to the spatial
distortion induced by the lenses, or if they were due to recovery from the effects of visual
deprivation on stereoacuity. Some marginally significant improvements were seen for
low and mid range spatial frequency contrast sensitivity for horizontal stimuli in the
hyperopic astigmatism group. It is important to note that although they were not
statistically significant, there were trends towards greater in improvement in the
astigmatism groups in comparison to the control group on measures of recognition acuity,
resolution acuity, and vernier acuity, suggesting that evidence of plasticity on these
measures may have been obtained with greater statistical power (increased sample size,
reduced variability).
It is important to note that the conclusion that eyeglass intervention during the 5
to 14 year age range is ineffective in treating this form of amblyopia is only one possible
interpretation of the data presented here. In studies aimed at evaluating treatment
effectiveness, the issue of subject compliance is always a primary concern. In the study
138
reported here, many efforts were made to encourage compliance in glasses wearing.
However, there were some children who were not compliant with treatment, and it is
possible that the inclusion of these children weakened treatment effects. Of 23 astigmatic
children on whom repeated compliance reports were obtained over the year of eyeglass
intervention, only 61% had "very good" or "excellent" compliance. An alternative
interpretation of the null treatment effects is that perhaps many of the children had
benefited from use of eyeglasses prior to the start of the study. Of the 48 astigmatic
children, 35% were wearing eyeglasses when they arrived at the first eye exam. Thus,
treatment effects that might have occurred due to eyeglass wear may have already
occurred for some subjects, and again may have weakened treatment effects. Finally, it is
possible that younger children responded to treatment and older children did not, again
weakening treatment effects since data were pooled over a large age range. It would be
helpful to compare results for older vs. younger children in this age range to see if there
is evidence of plasticity in younger children, but the sample size of astigmatic children in
the present study does not allow for meaningful analysis of age effects in treatment
effectiveness. However, a larger scale study is currently underway in which recovery
from astigmatism-related deprivation in K - 2"'' grade children will be compared to
recovery from astigmatism-related deprivation in 4"^- 6"* grade children.
In summary, three factors (compliance, previous treatment, and age effects) may
have contributed to the failure to fmd significant improvement in the children with
astigmatism, when viewed as a group. Thus, in is important to interpret the null
treatment effects cautiously. In the next section, I look at individual subject data, to
139
determine if there is evidence of treatment effects under conditions in which,
theoretically, we would expect them to be strongest.
3.4 Case Studies on Plasticity and Recovery from Effects of Deprivation
The limited sample size in the present study, although greater than previous
studies of the effects of plasticity related to recovery from the effects of astigmatismrelated deprivation, is not sufficient to conduct sub-analyses based on factors such as
compliance, previous eyeglass wear, and treatment age. However, given the potential
theoretical and clinical importance of the present study with regard to visual plasticity
associated with astigmatism-related deprivation, it seems worthwhile to look at some
individual cases to determine if there is any evidence of plasticity related to recovery
from astigmatism-related deprivation in this age group.
Since previous treatment and treatment compliance are important variables that
could not be statistically controlled for in the present study due to limited sample size, 1
reviewed the data to determine which astigmatic children did not have a history of
previous treatment (glasses wear for full correction of astigmatism) and which children
were highly compliant in wearing glasses throughout the I year study. 1 determined that
there were just three children who met both of these criteria.
Subject AA\ This child was 11 years old when his glasses were first dispensed as
part of the present study. He had 8.25 D of mixed astigmatism in the right eye (-5.75
+8.25 X 078) and 7.50 D of mixed astigmatism in the left eye (-5.75 +7.50 x 096). These
are unusually high amounts of astigmatism even for a child in this highly astigmatic
140
population. This child did have a history of previous glasses wear, but his eyeglasses
only partially corrected his astigmatism (right eye wear was -1.25 +2.62 x 092 at the first
eye exam), leaving him with over 5.00 D of uncorrected astigmatism. AA was
consistently compliant in wearing his eyeglasses throughout the year.
Subject CJ: This child was 6.6 years old when her eyeglasses were first dispensed
as part of the present study. She had 2.75 D of simple myopic astigmatism in the right
eye (-2.75 +2.75 x 077) and 3.00 D of mixed astigmatism in the left eye (-2.75 +3.00 x
092). CJ had no history of eyeglass wear, and was compliant with wearing eyeglasses
throughout the year, with the exception of a few short periods (days) during which her
eyeglasses were broken.
Subject CR: This child was 8.9 years old when her eyeglasses were first
dispensed as part of the present study. She had 2.50 D of mixed astigmatism in the right
eye (-2.25 +2.50 x 092) and 3.00 D of mixed astigmatism in the left eye (-2.75 +3.00 x
081). CR had no history of previous eyeglass wear, and was compliant with wearing her
eyeglasses throughout the study.
Data for each of these three subjects were compared to control group subjects of a
similar age, i.e., each subject was compared with mean data from control group subjects
who were +/-1 year of the subject's age in order to generate age-appropriate
(developmentally comparable) comparison group for each case study. Data are illustrated
in Figure 27 (letter acuity), 28 (grating acuity), 29 (vernier acuity), 30-32 (contrast
sensitivity), and 33 (stereoacuity). Review of data overall suggest that at baseline,
subject AA demonstrated poor perception in comparison to the control group, but showed
141
Subject AA
I
E
o
•0 2
•
o
AA
Control Group Mean +/-1 sd (n=16)
Baseline
One-Year
08 •
Subject CJ
q:
<
s
02
3
00 •
«
•
o
CJ
Control Group Mean *1- sd (n=13)
Baseline
One-Year
Subject OR
tr
02•
§
I
fl!
00 •
J,,.
•
o
CR
Control Group Mean •••/-sd (n=13)
Baseline
One-Year
Figure 27. Case studies: Recognition (letter) acuity by time. Individual subject data for
subjects AA, CJ, and CR plotted with age-matchedi control group mean + 1 standard
deviation and mean - 1 standard deviation.
142
Subject AA
•—
I
—o—
AA, Vertical
Control Group+/-sd(n=16),VerUcal
AA, Horizontal
Control Group+/-sd(n=16). Horizontal
Baseline
One-Year
Subject CJ
S*
I
Ica 12-^
I
O
CJ, Vertical
—^ — Control Group +/- sd (n=12). Vertical
—•— CJ, Horizontal
—o — Control Group •/- sd (n=12). Horizontal
08
ZO
Baseline
One-Year
Subject CR
5
S"
I
3
12
S
a
to
08
•—
I
—o —
CJ, Vertical
Control Group •/- sd (n=12). Vertical
CJ, Horizontal
Control Group •/- sd (n=12). Horizontal
Baseline
One-Year
Figure 28. Case studies: Resolution (grating) acuity by time. Individual subject data for
subjects AA, CJ, and CR plotted with age-matched control group mean + 1 standard
deviation and mean - 1 standard deviation.
143
Subject AA
©
0.5 -
c
w
—A —
—•
— -o —
Q)
>
AA. Vertical
Control Group •/- sd (n-l6). Vertical
AA. Horizontal
Control Group +/- sd (n=16), Horizontal
Baseline
One-Year
20-
Subject CJ
2
(0
0
1
03
<
0)w
1
•—
I
—o—
CJ, Vertical
Control Group +/- sd (n=11), Vertical
CJ. Horizontal
Control Group+/-sd (n=13), Horizontal
Baseline
One-Year
20 -
Subject CR
2m
"O 15-
®
05 •
E
o
>
CJ. Vertical
Control Group *1- sd (n=13). Vertical
CJ. Horizontal
Control Group *1- sd (n=13). Horizontal
00 -
Baseline
One-Year
Figure 29. Case studies: Vernier acuity by time. Individual subject data for subjects AA,
CJ, and CR plotted with age-matched control group mean + 1 standard deviation and
mean - I standard deviation.
144
•1 0
Subject AA
>
C(A
$
(A
15
co
o
•J?
-12
•1
4•
•1 6 -
-18 -2 0 -
s= = = =
=6
1
-2.2 -
u
-2
>»
4-
•—
I
--a—
AA, Vertical
Control Group *1- sd (n=16). Vertical
AA, Horizontal
Control Group+/-sd(n=16), Horizontal
-2 6 -
Baseline
One-Year
^ -'0-
Subject CJ
>
I -'2-14 (O
5 -1 6 5 •'®? -2 0 -
o -2 4
in
-2.6
'"(05
c
(0
2
c
O
•JS"
Q>«
o
(A
•1 0
-12
-1 4
-1 6
-1 8
-20
-22
-24
•2.6
—
—
—•
—o —
CJ. Vertical
Control Group •/- sd (n=13), Vertical
CJ. Horizontal
Control Group •/- sd (n=13), Horizontal
Baseline
One-Year
Subject CR
'— CR. Vertical
• — Control Group •/- sd (n=13). Vertical
I— CR. Horizontal
—o — Control Group *1- sd (n=13), Horizontal
Baseline
One-Year
Figure 30. Case studies: Contrast Sensitivity for low spatial frequency stimuli by time.
Individual subject data for subjects AA, CJ, and CR plotted with age-matched control
group mean + I standard deviation and mean - 1 standard deviation.
145
Subject AA
.2 .t«-
g
<s
-16-
2
c
5 .2 2S" -2 4 •
-2 6 •
Q
O -2'
CO
—•
• -o —
AA, Vertical
Control Group •/-sd (n=16).Verticat-o
AA. Horizontal
Control Group •/- sd (n=l6). Horizontal
Baseline
One-Year
Subject CJ
>
oS
2
3
= == =g
8"
• —
I
—a —
Q
a
<o
CJ. Vertical
Control Group •/- sd (n=13). Vertical
CJ, Horizontal
Control Group •/- sd (n=13). Horizontal
•30
Baseline
One-Year
•t 0
Subject CR
>
12
3 •"
>>
o
(O
Z Z ~ = r : r := ==Q
•—
I
D—
CR. Vertical
Control Group */-sd(n=13), Vertical
CR. Horizontal
Control Group+/-sd (n=13), Horizontal
Baseline
One-Year
Figure 31. Case studies: Contrast Sensitivity for middle spatial frequency stimuli by
time. Individual subject data for subjects AA, CJ, and CR plotted with age-matched
control group mean + 1 standard deviation and mean - 1 standard deviation.
146
2'S
C -0 5 -
Subject AA
Si
0)
g -1 0 •
c
8g,
-.5-
>•
u
® -2 5
cfc:=: = = = = = r:i:ii —:z
'—
•—
I—
—-o—
:
AA, Vertical
Control Group+/-sd(n=16). Vertical
AA. Horizontal
C o n t r o l G r o u p + / - s d ( n = l6 ) , H o r i z o n t a l
Baseline
One-Year
Subject CJ
'>
C -05 Si
en
12 -10c
5
g, -.5>»
a
®
-2 5 -
'—
•—
I—
—° —
CJ, Vertical
Control Group •/- sd (n=13). Vertical
CJ. Horizontal
Control Group •/- sd (n=13). Horizontal
Baseline
One-Year
Subject CR
>
*5)
C -05
(O
Q
E -10
C
a
>>
u
® -25
CR. Vertical
' — Control Group +/- sd (n=13). Vertical
I— CR. Horizontal
—o — Control Group •/- sd (n=13), Hotizontal
Baseline
One-Year
Figure 32. Case studies: Contrast Sensitivity for high spatial frequency stimuli by time.
Individual subject data for subjects AA, CJ, and CR plotted with age-matched control
group mean + 1 standard deviation and mean - 1 standard deviation.
147
2.8
a
(D
O
in
Subject AA I
26
24 -
22-
20
18
<
o
£
o
(0
o—
o— •
t4
•
°
^2
AA
Control Group Mean *1-1 sd (n-16)
Baseline
^8
ta
0
o
16
One-Year
Subject CJ
^6
2.4 -1
22
1
3 18
2.0-1
§
Q)
CO
O
J6
14 -(
Control Group Mean •*•/- sd (n=l4)
12 •
Baseline
Subject CR
y 26•
CO
"5
s
24
•
w
20-
3
'a-l
§ '6H
.9
(0
12•
One-Year
o— •
•
0
CR
Control Group Mean •*•/- sd (n=13)
Baseline
One-Year
Figure 33. Case studies; Stereoacuity by time. Individual subject data for subjects AA,
CJ, and CR plotted with age-matched control group mean + 1 standard deviation and
mean - 1 standard deviation.
148
improvement over the course of the year. Subjects CJ and CR, however, demonstrated
perception that was within normal limits for most measures at baseline, and appeared to
show improvements similar to that demonstrated by the control group over the course of
the year. The failure to find astigmatism-related deficits in these two subjects may be due
to the fact they had moderate, rather than severe astigmatism, or perhaps their
astigmatism had developed fairly recently, and thus did not result in the development of
amblyopia.
Overall, these analyses indicate that while there is little evidence of astigmatismrelated deficits and thus no evidence of recovery from deficits in subjects CJ and CR, the
data from AA indicated notable deficits at baseline, and improvement over time once
clear sensory input was restored through use of eyeglasses. This finding is particularly
notable because subject AA was 11-years-old when he received in glasses to correct his
astigmatism, several years older than age 7, which previous research has indicated as
being the upper age limit for plasticity associated with recovery from astigmatism-related
deficits. However, it should also be noted that even after 1 year of treatment, AA still
demonstrated evidence of perceptual deficits: e.g., letter acuity was 20/63 at baseline, and
20/32 at 1 month and 1 year. These data suggest that although plasticity can occur in a
child of this age, there may be limits on the amount of change that can occur with the
introduction of clear sensory input, as full recovery was not observed.
149
3.5 Factors Associated with Plasticity: Evidence from Other Paradigms
The data presented here suggest that, for 5- to 14-year-old age children with
astigmatism-related visual deficits, restoration of clear sensory input through prescription
of eyeglasses did not result in significant improvements in vision for the astigmatism
group as a whole, although there was some evidence of improvement in at least one child
who had relatively severe deficits at the start of the study (i.e., subject AA).
Research on plasticity associated with visual deprivation in humans has primarily
been reported in the clinical vision literatures in ophthalmology and optometry, with little
crossover into the experimental psychology literature. I believe that research and theory
from the experimental psychology literature can provide important insights and new ways
of thinking about plasticity associated with visual deprivation and recovery, and that this
contribution could be particularly meaningful in the case of recovery from visual
deprivation, and in the search for effective treatments for deprivation-related
visual/perceptual deficits, hi this section, I discuss some theoretical perspectives on
plasticity and perceptual change from the experimental psychology literature, and explore
ways in which these perspectives could be applied to research and theory on the visual
deprivation and recovery. Specifically, I will focus on theory regarding factors necessary
to induce plasticity.
Bedford (1993a, 1993b, 1995) addressed the question of what conditions are
necessary for plasticity to occur, hi her general "Perceptual Learning Theory", she
argues that all changes in perception that result from experience must result from the
150
detection of an internal perceptual error that needs to be corrected. She applied the
theory to plasticity associated with perceptual adaptation and suggested that plasticity
results when two conditions are met. The first condition, following the work of others
(Wallach, 1968, Welch, 1978), is that a sensory discrepancy must be detected. However,
she reasoned that the presence of a sensory discrepancy alone does not provide sufficient
evidence to the perceptual system that plasticity is necessary, i.e., it does not allow us to
distinguish between an internal perceptual error that needs to be corrected vs. the
detection of something new about our environment. Thus, she reasoned further that a
second condition that must be met is the detection of a violation of constraint internalized
by perceptual systems, which often involves knowledge about objects. It is the violation
of internal constraints that provides the cue that there is an internal perceptual error that
needs to be corrected.
I believe that applying Bedford's theory to deprivation-related plasticity can
provide important insights regarding factors that might lead to greater plasticity in
recovery from deprivation. To apply the Perceptual Learning Theory framework to
recovery from deprivation, we should start by asking two questions regarding the
conditions that lead to plasticity: (1) what is the sensory discrepancy that is detected, and
(2) what is the internal constraint that is violated by the discrepancy. With regard to
recovery from deprivation, I believe that the sensory discrepancy introduced is the
improved optics and resolution provided at the level of retinal inputs through use of
eyeglass correction for astigmatism, thus resulting in new sensory patterns of stimulation.
However, as Bedford (1993) reasoned, detection of this new sensory information may
151
simply indicate that there is something new about the environment; how do our
perceptual systems know that the discrepancy is reflecting an internal perceptual error
that requires correction or plasticity? What is the internal constraint that is violated by
this discrepancy? I believe that the internal constraint that is violated relates to object
constancy, hi general, our knowledge of the world and objects in the world tells us that
they do not change suddenly, i.e., the same objects should not result in different patterns
sensory stimulation over time. If they do, this would signal the violation of an internal
constraint, and signal the perceptual system that there is an internal error that requires
correction, as predicted by Perceptual Learning Theory.
How is the Perceptual Learning Theory framework helpful in pointing towards
ways in which we might induce more plasticity in recovery from deprivation? If
plasticity results when a discrepancy is detected, and when that discrepancy violates an
internal constraint, perhaps we could use methods to enhance detection of these
conditions. I believe that one method of accomplishing this is through use of
discrimination learning paradigms.
In the present study, the initial hypothesis was that with restoration of normal
visual experience (eyeglass wear to correct for optical effects of astigmatism),
improvement in visual perceptual functioning would occur. The basis for this assumption
was that normal visual experience in early childhood is suflicient to support normal
development of perceptual capabilities, and that perhaps restoration of normal input
would also be suflicient for the development of these perceptual capabilities in previously
visually deprived children. However, if subjects were beyond the sensitive period for this
152
type of change, perhaps more intensive visual stimulation e.g., through use of
discrimination learning trials, would be required in order for perceptual systems to detect
the sensory change, or error. As noted in the introduction, the literature on discrimination
learning has provided ample evidence that visual perceptual abilities in adults can be
improved through intensive training paradigms.
Reinke (1999) recently applied Perceptual Learning Theory (Bedford, 1993a,
1993b, 1995) to plasticity associated with discrimination learning. She suggested that the
perceptual error detected in these studies was that there was "not enough sensitivity".
The information regarding violation of internal constraints in these paradigms may come
from top-down processes, such as attention, feedback, and task-related demands and
information. Bedford (personal communication, 2002) has suggested that perhaps the
internal constraint is that two different objects should not look exactly the same. For
example, if task related information indicates that the two objects are different, but
sensory inputs provide information that they look the same, this condition would signal
that there is an internal error (not enough sensitivity, as suggested by Reinke, 1999) that
needs to be corrected. Perhaps this type of top-down information regarding detection of
sensory discrepancy and violations of internal constraints would provide the additional
motivation for perceptual change in subjects who are towards the end or beyond the
sensitive period for recovery from astigmatism-related deprivation.
The literature on perceptual adaptation may also provide further clues for
development of effective treatments for deprivation-related perceptual deficits. For
example. Banks (1988) noted that adult studies of perceptual adaptation report that
153
complete adaptation occurs in some instances but not others, i.e., perceptual aftereffects
are less than 100% of the distortion experienced. Banks suggests that complete
adaptation can occur under the following conditions: (1) when experimentally induced
distortions mimic the effects of normal developmental processes, and (2) when there is
adequate visual and visuomotor experience to determine what sort of distortion has
occurred and how to compensate for it. Perhaps these principles might also be relevant
for recovery from deprivation. The data presented here indicate that full recovery of
perceptual abilities did not occur in children with astigmatism-related perceptual deficits
treated with eyeglass wear. Applying Banks' hypothesis to the interpretation of these
data suggests that perhaps the failure to observe improvement in these subjects is due in
part to the fact that either the restoration of clear input through use of eyeglasses does not
adequately mimic normal development or that subjects did not receive adequate visual
experience in order for them to determine what type of sensory change had occurred.
How might we create a situation for these subjects that would more accurately
mimic normal development of perceptual abilities, and heighten detection of changes in
their visual experiences, i.e., increase the likelihood that they will detect a perceptual
error? In experimental paradigms, this is often achieved through feedback, whether it is
through trial and error motor interactions with their environment in prism adaptation
experiments, or through experimenter-provided error feedback in discrimination learning
paradigms. In the case of recovery from deprivation, the use of direct feedback
regarding detection of visual stimuli seems most appropriate, and therefore
discrimination learning paradigms might be a useful means of providing subjects with
154
feedback regarding the accuracy of their perceptions, and cues to the presence of
perceptual errors.
In summary, research and theory on plasticity from the experimental psychology
literature suggests several factors that may increase plasticity in subjects recovering from
the effects of visual deprivation. These ideas focus on the conditions necessary for
plasticity to occur, such as detection of an sensory discrepancy (Banks, 1988, Bedford,
1993a, 1993b, 1995) and violation of internal constraints that would signal that there was
an internal error that required correction, or plasticity (Bedford, 1993a, 1993b, 1995).
The most efficient way to increase the likelihood that these conditions would be detected
by perceptual systems might be through use of discrimination learning paradigm. With
use of this type of paradigm, feedback and task demands could be used to stimulate topdown processes to increase the chance of detection of internal perceptual errors, and
gradual increases in discrimination difficulty over time could be used to simulate normal
developmental changes and to ensure continuous detection of perceptual error over time.
In addition, visual perceptual experience could be directed at targeting specific perceptual
deficits.
Recently, researchers have begim to evaluate the effects of this type of treatment
strategy on patients with deprivation-related perceptual deficits (e.g., Polat et al., 2001).
The data presented in the present study, specifically baseline data, have essentially
provided a blueprint of the nature of perceptual deficits that occurs as a result if
astigmatism-related deprivation, and provide important information for development of
discrimination learning type treatment for these deficits.
155
3.6 Plasticity Associated with Adaptation to Spatial Distortion
This research also addressed the extent to which children adapt to the shape
distortion induced by cylinder lenses, i.e., lenses that correct for the optical defocus
caused by astigmatism. I predicted that there would be no difference in mean response
for the control group with glasses on vs. glasses off (since there is little or no cylinder in
the lenses), and that the astigmatism group would not differ from the control group in the
glasses-off condition at baseline because minimal distortion is produced by an astigmatic
cornea or lens. However, I predicted that, if the task were sensitive enough to measure
the distortion induced by the cylinder lenses, the astigmatism group, on average, would
tend to choose more vertically elongated stimuli as "the perfect circle", since cylinder
lenses that correct for with-the-rule astigmatism vertically compress images.
Results for the control group were as predicted: mean responses were close to 0
(i.e., true circle), and did not differ between the glasses-on and glasses-off conditions, nor
did they differ from baseline to 1 month. The astigmatism group tended to choose more
vertically elongated stimuli in the glasses-on condition at baseline, as predicted,
indicating that the task was sensitive enough to measure the distorting effects of the
cylinder lenses. However, the data suggest that children in the astigmatism group
perceive the circles differently than the control group when they are not wearing their
glasses: on average, the astigmatic children perceived vertically compressed stimuli as
more circle-like, whereas the children in the control group chose stimuli close to the true
circle as circle-like. This result is puzzling, because the original prediction was that
children in both groups would choose a stimulus close to the true circle when glasses
156
were not worn, i.e., when no distortion was induced. As previously discussed,
astigmatism of the eye does not cause distortion like cylinder lenses because the amount
of distortion perceived is dependent in part on the distance between the unequal curvature
and the location where light enters the eye, i.e. the pupil, and the greater the distance
from the pupil, the greater the distortion. Thus, when unequal curvative is located on the
cornea, distortion is minimal (Guyton, 1977), whereas when the unequal curvature is
located further from the pupil, e.g., with eyeglass lenses, distortion is greater. However,
even if astigmatism at the cornea did distort form perception in a manner similar to the
distortion induced by cylinder lenses, one would assume that the children would have
adapted to this distortion.
Several additional post hoc analyses were conducted to verify this unusual
finding. For example, in Figure 34,1 plot mean scores for the control group and the
astigmatism group on each type of trial (baseline/follow-up x stimulus type (outline
circle/checkered circle) x start stimulus (-12/+12)). As you can see from the comparison
of the two plots, in the glasses on condition, mean measurement by trial type is similar
for the astigmatism vs. the control group: lines connecting measurements for each trial
type for control and astigmatism groups are generally flat. In contrast, in the glasses off
condition, regardless of trial type, the astigmatism group tended to choose more vertically
compressed stimuli as the "perfect circle". These patterns are consistent across trial type,
despite what appear to be strong main effects of start stimulus (filled vs. open symbols),
such that subjects tend to choose more elongated stimuli when they start the task from the
+12 stimulus (filled symbols), and they tend to choose more compressed stimuli when
157
A: Glasses on
--IT
"5
Control
Astigmatism
B: Glasses off
aH
Control
Aibgmalisni
B = Checkered Circle Stimuli, Subjects started task with +12 Stimulus
• = Checkered Circle Stimuli, Subjects started task with -12 Stimulus
^ = Outline Circle Stimuli. Subjects started task with -»-12 Stimulus
^ = Outline Circle Stimuli, Subjects started task with -12 Stimulus
= Baseline Measurements
= 1 Month Measurements
Figure 34. Mean form perception by trial/stimulus type and group. Mean +/-1 standard
error of form perception responses for each trial/stimulus type under glasses-on (A) and
glasses-off (B) conditions.
158
they start the task from the -12 stimulus (open symbols). Thus, this pattern seems quite
robust, occurring regardless of stimulus type, start stimulus, or time of measurement.
One possibility is that the effects of orientation-dependent blur might influence
shape perception in the astigmatic subjects. For example, in the case of the
myopic/mixed astigmats, increased blur for horizontal lines may make true circles appear
taller (top and bottom of the circle would appeared blurred than the sides), thus causing
them to choose flatter stimuli to compensate for the effects of the blur. Also, since these
subjects have poorer acuity for horizontal even with glasses on, the effect might have
been observed for both the glasses-on and glasses-off conditions. Hyperopic astigmats,
who appear to focus between horizontal and vertical (based on presence of equivalent
deflcits for both orientations) may not have such a bias related to blur. However, it is
possible that this type of effect is driving the data for the astigmatism group towards
selection of flatter stimuli, and thus explaining the observed effects.
Regardless of this unexpected flnding, baseline data indicate that the children in
the astigmatism group perceive shape differently when the glasses are on, in comparison
to when they are off, and that this distortion is in the direction as predicted by the type of
distortion the lenses induce. In addition, the data are also clear with regard to measures
of perceptual adaptation to these lenses - there was no indication in the data that the
children were adapting to the distortion over a I month period. However, Perceptual
Learning Theory (Bedford, 1993a, 1993b, 1995) would suggest that, since perception in
the astigmatism group with glasses-on appears to be accurate, no perceptual error would
be detected, and therefore, the plasticity would not be expected to occur.
159
Finally, from a clinical perspective, these data suggest that the form distortion
induced by the cylinder lenses provides minimal disruption to perception of form. From
a theoretical perspective, the absence of conditions that would lead to plasticity did not
allow us to determine the extent of plasticity related to perceptual adaptation in children
in the present study. 1 hope that future studies directly comparing perceptual adaptation
for adults vs. children will help determine if there is a sensitive period for this form of
visual plasticity, as there is for other forms of visual plasticity, such as plasticity related
to visual deprivation and to recovery from the effects of deprivation.
3.7 Conclusions
Since the publication of key research studies by Mitchell, Freeman and their
colleagues in the early 1970s (Freeman et al, 1972, Mitchell et al, 1973), only a handful
of studies have addressed questions regarding the effects of astigmatism-related
deprivation on visual/perceptual development and function in humans (e.g., Gwiazda et
al., 1986, Dobson et al., 2002). Presumably, this is due to the limited subject pool for
such studies due to the low prevalence of astigmatism in the general population. The
presence of such a high prevalence of astigmatism among some Native American Tribes
has allowed us to conduct studies that have led to a better understanding the influence of
astigmatism on visual perceptual development, and have allowed for prospective studies
on plasticity for recovery from the effects of astigmatism-related deprivation. Ultimately,
the goal of this line of research is to gain better insight into the development and
organization of the human visual system and the conditions that lead to visual plasticity.
160
and from a clinical perspective, to develop effective treatments for astigmatism-related
visual deficits, and other forms of deprivation-related visual/perceptual deficits.
The results of the present study suggest that recovery from the effects of
astigmatism-related deprivation in 5- to 14-year-old children are minimal at best with
simple re-introduction of clear vision (i.e., eyeglass treatment). However, I hope that the
discussion presented here has highlighted areas of research and theoretical ideas that may
lead to treatment alternatives for children and adults who are beyond the sensitive period
for recovery from these types of visual perceptual deficits. Many of these ideas are
discussed primarily in the experimental psychology literature, which has focused little on
deprivation-related plasticity. Greater interaction of research and theory on visual
plasticity between clinical vision and experimental psychology could lead to significant
advances for both research traditions.
161
4. REFERENCES
Abraham, J.E., and Volovick, J.B. (1972). Preliminary Navajo optometirc study. Journal of
the American Optometric Association. 43.1257-1260.
Adams, R.J., and Courage, M.L. (2002). Using a single test to measure human contrast
sensitivity from early childhood to maturity. Vision Research. 42. 1205-1210.
Ames, A. (1946). Binocular vision as affected by relations between uniocular stimuluspatterns in commonplace environments. The American Journal of Psychology. 59.333357.
Atkinson, J., Braddick, O., Bobier, B., Anker, S., Ehrlich, D., King, J., Watson, P., Moore, A.
(1996). Two infant vision screening programmes: Prediction and prevention of
strabismus and amblyopia from photo- and videorefractive screening. Eye, 10, 189-198.
Banks, M.S. (1988). Visual recalibration and the development of contrast and optical flow
perception. In A. Yonas (Ed.), Perceptual development in infancy, (pp. 145-196).
Hillsdale, NJ: Lawrence Erlbaum Associates.
Banks, M.S., Aslin, R.N., Letson, R.D. (1975). Sensitive period for the development of
human binocular vision. Science. 190.675-677.
Bedell, H.E., Flom, M.C., Barbeito, R. (1985). Spatial aberrations and acuity in strabismus
and amblyopia. Investigative Ophthalmology and Visual Science. 26, 909-916.
Bedford, F.L. (1993a). Perceptual and cognitive spatial learning. Journal of Experimental
Psychology: Human Perception and Performance. 19. 517-530.
Bedford, F.L. (1993b). Perceptual Leaming. In D. Medin (Ed.), The Psychology of Learning
and Motivation. (Vol.30, pp. 1-60), San Diego, CA: Academic Press.
Bedford, F.L. (1995). Constraints on perceptual leaming: Objects and dimensions. Cognition.
M, 253-297.
Bedford, F.L. (1999). Keeping perception accurate. Trends in Cognitive Sciences. 3,4-11.
Bennett, R.G., Westheimer, G. (1991). The effect of training on visual alignment
discrimination and grating resolution. Perception and Psvchophvsics. 49. 541-546.
Boniuk, V. (1973). Refractive problems in native peoples (the Sioux Lookout Project).
Canadian Journal of Ophthalmology. 8,229-233.
162
Bradley, A., Freeman, R.D. (1982). Contrast sensitivity in children. Vision Research. 22.953959.
Brown, S.M., Archer, S., Del Monte, M.A. (1999). Stereopsis and binocular vision after
surgery for unilateral infantile cataract. Journal of AAPQS. 3,109-113.
Burian, H.M. (1943). Influence of prolonged wearing of meridional size lenses on spatial
localization. Archives of Opthalmoloev. 30.645-666.
Carkeet, A., Levi, D.M., Manny, R.E. (1997). Development of vernier acuity in childhood.
Optometry and Vision Science. 74. 741-750.
Chen, S.I., Kaye, S.B., McCloskey, L., Mfazo, H., Rubin, A., Harris, W.F. (2002). The effect
of induced astigmatism on stereoacuity thresholds. 2002 Annual Meeting Abstract and
Program Planner [on CD-ROM]. Association for Research in Vision and Ophthalmology.
Abstract 4679.
Ciner, E.B., Schantel-Klitsch, E., Herzberg, C. (1996). Stereoacuity development: 6 months to
5 years. A new tool for testing and screening. Optometry and Vision Science. 73.43-48.
Cobb, S.R., MacDonald, C.F. (1978). Resolution acuity in astigmats: Evidence for a critical
period in the human visual system. British Journal of Physiological Optics. 32.38-49.
Coleman, H.M. (1970). An analysis of the visual status of an entire school population. Joumal
of the American Optometric Association. 41.341-347.
DeValois, K.K. (1977). Spatial frequency adaptation can enhance contrast sensitivity. Vision
Research. 17. 1057-1065.
Dobson, v.. Miller, J.M., Harvey, E.M. (1999). Corneal and refractive astigmatism in a
sample of 3- to 5-year-old children with a high prevalence of astigmatism. Optometry
and Vision Science, 76, 855-860.
Dobson, v.. Miller, J.M., Harvey, E.M., Mohan, K.M. (in press). Amblyopia in astigmatic
preschool children. Vision Research.
Dobson, v.. Miller, J.M., Harvey, E.M., Sherrill, D.L. (1999). Prevalence of astigmatism,
astigmatic anisometropia, and glasses wearing among preschool- and school-age Native
American children. In Vision Science and Its Applications, Vol. 1, OSA Technical
Digest Series (pp. 177-80). Washington, DC: Optical Society of America.
163
Dobson, v., Tyszko, R.M., Miller, J.M., Harvey, E.M. (1996). Astigmatism, amblyopia, and
visual disability among a Native American population. In Vision Science and its
Applications, 139-142.
Dowdeswell, H.J., Slater, A.M., Broomhall, J., Tripp, J. (1995). Visual deficits in children
bom at less than 32 weeks' gestation with and without major ocular pathology and
cerebral damage. British Journal of Qohthalmoloev. 79.447-452.
Ellemberg, D., Lewis, T.L., Liu, C.H., Maurer, D. (1999). Development of spatial and
temporal vision during childhood. Vision Research. 39. 2325-2333.
Epstein, W. (1972). Adaptation to uniocular image magnification: Is the underlying shift
proprioceptive?. Perception and Psvchophvsics. 11. 89-91.
Epstein, W., Davies, N. (1972). Modification of depth judgment following exposures to
magnification of uniocular image: Are changes in perceived absolute distance and
registered direction of gaze involved? Perception and Psvchophvsics. 12, 315-317.
Epstein, W., Morgan, C.L. (1970). Adaptation to uniocular image magnification: Modification
of the disparity-depth relationship. American Journal of Psvcholoev. 83. 322-329.
Fox, R., Patterson, R., Francis, E.L. (1986). Stereoacuity in young children. Investigative
Ophthalmoloev and Visual Science. 27. 598-600.
Freeman, R.D. (1975). Asymmetries in human accommodation and visual experience. Vision
Research. 15.483-492.
Freeman, R.D., Mitchell, D.E., Millodot, M. (1972). A neural effect of partial visual
deprivation in humans. Science. 175. 1384-1386.
Freeman, R.D., Thibos, L.N. (1975). Contrast sensitivity in humans with abnormal visual
experience. Journal of Phvsioloev. 247.687-710.
Furmanski, C.S., Engel, S.A. (2000). An oblique effect in human primary visual cortex.
Nature Neuroscience. 3, 535-536.
Gibson, E.J. (1963). Perceptual Leaming. Annual Review of Psychology. 14. 29-56.
Gibson, E.J. (1991). An Qdvssev in Learning and Perception. Cambridge, MA: MIT Press.
Gibson, E.J., and Pick, A.D. (2000). An ecological approach to perceptual leaming and
development. New York, NY: Oxford University Press.
164
Guyton, D.L. (1977). Prescribing cylinders: The problem of distortion. Survey of
Ophthalmology. 22. 177-188.
Gwiazda, J. (1987). Development of grating and vernier acuity and the detection of amblyopia
in human infants. In M. Lenk-Schafer (Ed.), Orthoptic Horizons - Transactions of the
Sixth International Orthoptic Congress (pp89-95). Harrogate, England
Gwiazda, J., Bauer, J., Thom, F., Held, R. (1986). Meridional amblyopia does result from
astigmatism in early childhood. Clinical Vision Science, i, 145-152.
Gwiazda, J., Mohindra, 1., Brill, S., Held, R. (1985). Infant astigmatism and meridional
amblyopia. Vision Research, 25, 1269-1276.
Gwiazda, J., Bauer, J., Thom, F., Held, R. (1997). Development of spatial contrast sensitivity
from infancy to adulthood; Psychophysical data. Optometry and Vision Science. lA, 785789.
Hamilton, J.E. (1976). Vision anomalies of Indian school children: The Lame Deer Study.
Journal of the American Optomeric Association. 47.479-487.
Harwerth, R.S., Smith, E.L., Duncan, G.C., Crawford, M.L.J., von Noorden, G.K. (1986).
Multiple sensitive periods in the development of the primate visual system. Science. 232.
235-238.
Held, R. (1965). Plasticity in sensory-motor systems. Scientific American. 213. 84-94.
Hess, R.F., Campbell, F.W., and Greenhalgh, T. (1978). On the nature of the neural
abnormality in human amblyopia; Neural aberrations and neural sensitivity loss. Pfliigers
Archives European Journal of Physiology. 377.201-207.
Hess, R.F., Wang, Y., Demanins, R., Wilkinson, F., Wilson, H.R. (1999). A deficit in
strabismic amblyopia for global shape detection. Vision Research. 39.901-914.
Hirsch, M.J. (1963). Changes in astigmatism during the first eight years of school - an interim
report from the OJAI Longitudinal Study. America Journal of Optometry American
Academy of Optometry. 40.127-132.
Johnson, C.A., and Leibowitz, H.W. (1979). Practice effects for visual resolution in the
periphery. Perception and Psvchophvsics. 25.439-442.
Julesz, B. (1971). Foundations of Cvclopian Perception. Chicago, IL: University of Chicago
Press.
165
Keech, R.V., & Kutschke, PJ. (1995). Upper age limit for the development of amblyopia.
Journal of Pediatric Ophthalmology and Strabismus. 32. 89-93.
Kerschner, R. and Brick, D. (1984). Prevalence of high corneal astigmatism in Papago school
children. Investigative Ophthalmology and Visual Science. 25.217.
Kuman, I.G., Fedorov, S.N., Navikova, L.A. (1984). The sensitive period in the development
of the human viusal system. Soviet Psychology. 22. 35-46.
Levi, D.M., Carkeet, A. (1993). Amblyopia: A consequence of abnormal visual development.
In K. Simons et al. (Eds.), Earlv Visual Development: Normal and Abnormal, (pp. 391408). New York, NY: Oxford University Press.
Levi, D.M., Klein, S. (1982). Hyperacuity and amblyopia. Nature. 298.268-270.
Levi, D.M., Polat, U., Hu, Y.S. (1997). Improvement in vemier acuity in adults with
amblyopia: Practice makes better. Investigative Ophthalmology and Visual Science. 38,
1493-1510.
Lewis, T.L., Ellemberg, D., Maurer, D., Wilkinson, F., Wilson, H.R., Dirks, M., Brent, H.P.
(2002). Sensitivity to global form in Glass patterns after early visual deprivation in
humans. Vision Research. 42,939-948.
Lewis, T.L., Maurer, D., and Brent, H.P. (1986). Effects on perceptual development of visual
deprivation during infancy. British Journal of Ophthalmology. 70.214-220.
Maurer, D., Lewis, T.L. (1993). Visual outcomes after infantile cataract. In K. Simons et al.
(Eds). Early Visual Development: Normal and Abnormal (pp. 454-484). New York, NY:
Oxford University Press.
Mayer, D.L., Dobson, V. (1982). Visual acuity development in infants and young children, as
assessed by operant preferential looking. Vision Research. 22.1141-1151.
Miller, J.M., Dobson, V., Harvey E.M., and Sherrill D.L. (2001). Comparison of preschool
vision screening methods in a population with a high prevalence of astigmatism.
Investigative Ophthalmology & Visual Science. 45.917-924.
Miller, J.M., Dobson, V., Harvey E.M., and Sherrill, D.L. (2000). Astigmatism and
Amblyopia among Native American Children (AANAC): design and methods..
Ophthalmic Epidemiology. 7, 187-207.
166
Miller, J.M., Harvey, E.M., and Dobson, V. (2002). A practical method for testing vernier
acuity in children. 2002 Annual Meeting Abstract and Program Planner [on CD-ROM].
Association for Research in Vision and Ophthalmology. Abstract 2665.
Mintz-Hittner, H.A., Fernandez, K.M. (2000). Successful amblyopia therapy initiated after
age 7 years.. Archives of Ophthalmology. 118.1535-1541.
Mitchell, D.E., Freeman, R.D., Millodot, M., Haegerstrom, G. (1973). Meridional amblyopia:
Evidence for modification of the human visual system by early visual experience. Vision
Research, 13. 535-558.
Mitchell, D.E., Wilkinson, F. (1974). The effect of early astigmatism on the visual resolution
of gratings. Journal of Phvsiolopv. 243. 739-756.
Mohindra, I., Jacobson, S.G., Held, R. (1983). Binocular visual form deprivation in human
infants. Documenta Ophthalmoleica. 55. 237-249.
Mohindra, I., Nagaraj, S. (1977). Astigmatism in Zuni and Navajo Indians. American Journal
of Optometric and Phsiological Optics. 54. 121-124.
Morris, R.J., Scott, W.E., Dickey, C.F. (1993). Fusion after surgical alignment of
longstanding strabismus in adults. Ophthalmology. 100. 135-138.
Morrison, L.C. (1972). Further studies on the adaptation to artificially-induced aniseikonia.
British Journal of Physiological Optics. 27. 84-101.
Myers, V.S., Gidlewski, N., Quinn, G.E., Miller, D., Dobson, V. (1999). Distance and near
visual acuity, contrast sensitivity, and visual fields of 10-year-old children. Archives of
Ophthalmology. 117.94-99.
Pettigrew, J.D. (1978). The paradox of the critical period for striate cortex. In C.W. Cotman
(Ed.), Neuronal Plasticity (pp, 311-330). New York, NY: Raven Press.
Polat, U., Ma-Naim, T., & Belkin, M. (2001). Treatment of adult amblyopia by perceptual
learning. Investigative Ophthalmology and Visual Science, 42, S400.
Ramachandran, V.S. (1976). Leaming-like phenomena in stereopsis. Nature. 262. 382-384.
Ramachandran, V.S., Braddick, 0. (1973). Orientation-specific learning in stereopsis.
Perception. 2, 371-376.
Reinke, K.S. (1999). Visual and Neural Plasticity: A study of line orientation discrimination.
Dissertation Abstracts International: Section B: The Sciences & Engineering. 59.6088.
167
Riggs, L.A. (1965). Visual Acuity. In C.H. Graham (Ed), Vision and Visual Perception (PP.
321-349). New York, NY: Wiley.
Rosenfield, M. (1998). Refractive status of the eye. to W.J. Beniamin (Ed.) Borish's Clinical
Refraction (pp. 2-29). Philadelphia, PA: W.B. Saunders Company.
Scharre, J.E., Cotter, S.A., Block, S.S., and Kelly, S.A. (1990). Normative contrast sensitivity
data for young children. Optometry and Vision Science. 67. 826-832.
Simmers, A.J., Gray, L.S., McGraw, P.V., Winn, B. (1999). Functional visual loss in
amblyopia and the effect of occlusion therapy, tovestigative Ophthalmology and Visual
Science. 40. 2859-2871.
Skoczenski, A.M. (2002). Limitations on visual sensitivity during infancy: Contrast
sensitivity, vernier acuity, and orientation processing, to Fagen, J.W. and Hayne, H.
(Eds.) Progress in tofancv Research. Vol. 2 (pp. 169-214). Mahwah, N.J.: Lawrence
Erlbaum Associates.
Sowden, P., Davies, I., Rose, D. (1996). Perceptual learning of stereoacuity. Perception. 25.
1043-1052.
Sowden, P.T., Rose, D., Davies, I.R.L. (2002). Perceptual learning of luminance contrast
detection: Specific for spatial frequency and retinal location but not orientation. Vision
Research. 42.1249-1258.
Teller, D.Y., Allen, J.L., Regal, D.M., Mayer, D.L. (1978). Astigmatism and acuity in two
primate infants, tovestigative Ophthalmology & Visual Science. 17. 344-349.
Wallach, H. (1968). Adaptation based on cue discrepancy, to R. Held, E. Taub, and S.J.
Freeman (Eds.), Neuropsychology of spatially oriented behavior. Homewood, IL:
Dorsey Press.
Welch, R.B. (1978). Perceptual Modification: Adapting to altered sensory enviroments. New
York, NY: Academic Press.
Wick, B., Crane, S. (1976). A vision profle of American todian children. Journal of
Optometric and Physiological Optics. 53.34-40.
Woodruff, M.E. (1986). Vision and refractive status among grade 1 children of the Province
of New Brunswick. American Journal of Optometry and Physiological Optics. 63. 545552.
168
Zadnik, K., and Mutti, D.O. (1998). Incidence and distribution of refractive anomalies. In
W.J. Benjamin (Ed.) Borish's Clinical Refraction (pp. 30-46). Philadelphia, PA: W.B.
Saunders Company.
Zanker, J., Mohn, G., Weber, U., Zeitler-Driess, K., Fahle, M. (1992). The development of
vernier acuity in human infants. Vision Research. 32. 1557-1564.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement