Subjects were children in grades K-2 (recruited during the 2003/04 school year) and children in grades 4–6 (recruited during the 2001/02 school year) who attended one of five elementary schools located on the Tohono O’odham Reservation in southern Arizona, and children at a sixth elementary school on the reservation who participated in a preliminary study during the 2000/01 and 2001/02 school years. Recruitment years for different grades were selected in order to minimize the possibility of recruiting children who participated in a previous eyeglass treatment study of Tohono O’odham preschool children (1997–2001, Miller, Dobson, Harvey, & Sherrill, 2000
; Dobson et al. 2003
; Harvey et al., 2004
). This population was chosen for the study because there is a high prevalence of astigmatism (Dobson, Miller, & Harvey, 1999
; Dobson, Miller, Harvey, & Sherrill, 1999
; Harvey, Dobson, & Miller, 2006
) and astigmatism-related amblyopia (Dobson et al., 1996
; Dobson et al., 2003
; Harvey et al., 2004
) among the Tohono O’odham.
The Institutional Review Board of the University of Arizona approved this study. Prior to each child’s participation, written informed consent was obtained from a parent or guardian, and written assent was obtained from children in grades 4, 5, and 6.
Each child was scheduled to participate in an initial eye examination, a baseline best-corrected vision testing session, a six-week follow-up best-corrected vision testing session, a one-year follow-up eye examination, and a one-year follow-up best-corrected vision testing session. Refractive error correction for the baseline and six-week follow-up vision testing sessions was determined at the initial eye examination, and refractive error correction for the one-year follow-up vision testing session was determined at the one-year follow-up eye examination.
At the eye examinations, each child underwent a complete eye examination including cycloplegic refraction, conducted by a pediatric ophthalmologist (JMM) at least 40 min after instillation of one drop of proparacaine (0.5%) and two drops of cyclopentolate (1%) separated by an interval of 5 min. Eyeglasses were prescribed for (a) children who had ≥ 2.00 diopters (D) of astigmatism in either eye, and (b) children who had uncorrected recognition acuity worse than 20/20 and significant refractive error (myopia ≥ 0.75 D in either meridian, hyperopia ≥ 2.50 D in either meridian, astigmatism ≥ 1.00 D in either eye, anisometropia ≥ 1.50 D spherical equivalent). Eyeglass prescriptions were determined by cycloplegic autorefraction (Nikon Retinomax K+, Nikon Inc, Tokyo, now manufactured by Righton Manufacturing Co., Tokyo), confirmed by retinoscopy and by subjective refinement (when possible). Correction of hyperopic refractive error was reduced by one-third or by 1.00 D, whichever was greater (Guyton, Miller, & West, 2003
The baseline vision testing session was conducted on a separate day approximately two to three weeks after the initial eye exam. The first follow-up vision testing session was conducted approximately six weeks after the baseline session, and the one-year follow-up vision testing session was conducted approximately two to three weeks after the one-year follow-up eye examination (approximately one year after the baseline vision testing session). Eyeglasses were prescribed only for children who met the above criteria, and these children were given their eyeglasses at the beginning of the baseline vision testing session. However, all children wore eyeglasses containing their refractive correction (with hyperopic refractive error reduced by one-third or by 1.00 D, whichever was greater) during the vision testing sessions. Thus, each child was tested with his/her best-correction, and testers were masked as to which children had been prescribed eyeglasses. Children who did not meet the prescribing criteria wore a pair of eyeglasses selected from a set of “stock” eyeglasses in which the right and left lens corrections were no more than 0.50 vector dioptric difference (Long, 1976
; Harris, 1990
; Harvey, Miller, Dobson, Tyszko, & Davis, 2000
) from the child’s refractive error.
Vision testing was conducted by a team of trained testers who were masked to each child’s refractive error and to results obtained at previous testing sessions. Each vision testing session included five tests: (1) monocular (right eye (RE) and left eye (LE)) distance (4 m) logMAR recognition acuity using 62- by 65-cm Early Treatment Diabetic Retinopathy Study (ETDRS) charts (Ferris, Kassoff, Bresnick, & Bailey, 1982
) mounted in an illuminator cabinet (Precision Vision, Inc., LaSalle, IL); (2) monocular (RE) grating acuity for V, H, and O gratings tested at 1.5 m using stimuli constructed from unmounted Teller acuity cards (Vistech Consultants, Inc., Dayton, OH) (Teller, McDonald, Preston, Sebris, & Dobson, 1986
); (3) monocular (RE) vernier acuity for V, H, and O stimuli tested at 1.75 m using stimuli that were generated using a computer program (Miller, Harvey, & Dobson, 2002
); (4) monocular (RE) contrast sensitivity tested at 3 m for low (1.5 cy/deg), middle (6 cy/deg), and high (18 cy/deg) spatial frequency V and H sinewave grating stimuli using stimuli constructed from unmounted VCTS6500 Contrast Sensitivity Charts (Vistech Consultants, Inc., Dayton, OH); and (5) stereoacuity tested at 40 cm using the Randot Preschool Stereoacuity Test (Stereo Optical Co., Chicago, IL) (Birch, Williams, Hunter, & Lapa, 1997
). Test order was counterbalanced across subjects but remained constant for each child across all testing sessions (baseline, six weeks, and one year). During monocular testing, the fellow eye was occluded with 5-cm wide adhesive paper tape (3M micropore, Minneapolis, MN). A detailed description of the tests and testing procedures is provided in another report (Harvey et al., 2007b
). Results of longitudinal recognition acuity testing are reported elsewhere (Harvey et al., 2007a
2.3. Resolution (Grating) Acuity
Grating acuity was assessed using a 3-alternative forced-choice (3AFC) procedure in which the subject’s task was to identify which one of three circles (number 1, 2, or 3) contained a grating. The remaining two circles on each trial contained gray stimuli constructed from the same Teller Acuity Card as the grating. Stimuli were organized into a test book that included grating spatial frequencies ranging from 0.86 to 38 cy/cm (2.3 to 99.5 cy/deg), ordered from lowest to highest spatial frequency, with V, H, and O gratings interleaved. Order of presentation of orientations within each spatial frequency was always the same for an individual child (at baseline, six weeks, and one year), but was counterbalanced across the five schools at which testing was conducted. Testing began with the 6.5 cy/cm (17 cy/deg) grating and continued until the subject could no longer identify the location of the grating on three of three or on three of four trials for an orientation. Grating acuity for each orientation was scored as the highest spatial frequency at which a subject could correctly locate the grating on at least three out of a maximum of four trials. For subjects who were judged unable to resolve the largest grating available (0.86 cy/cm, 2.3 cy/deg), a grating acuity corresponding to the next lower spatial frequency in the Teller Acuity Card set (0.64 cy/cm, 1.7 cy/deg) was assigned.
2.4. Vernier Acuity
Vernier acuity for V, H, and O lines was tested using a 3AFC procedure similar to that used to assess grating acuity. The subject’s task was to identify which one of three circles (number 1, 2, or 3) contained the “wiggly” line. The remaining two circles on each trial contained a straight line of the same width and length as the vernier stimulus. Stimuli were organized into a test book that included stimuli with offsets ranging from 80 to 5 arc sec, ordered from largest to smallest offset, with V, H, and O gratings interleaved. Order of presentation of orientations within each offset size was always the same for an individual child (at baseline, six weeks, and one year), but was counterbalanced across the five schools at which testing was conducted. Testing began with the 80 arc sec offset and continued until the subject could no longer identify the location of the vernier stimulus on three of three or on three of four trials for an orientation. Vernier acuity for each stimulus orientation was scored as the smallest vernier offset at which the child could correctly identify the vernier stimulus on three out of a maximum of four trials. Subjects who were judged unable to resolve the largest offset (80 arc sec) were assigned a vernier acuity 100 arc sec, i.e., 0.1 log unit larger than the largest level included in the test book.
2.5. Contrast Sensitivity
Assessment of contrast sensitivity was conducted using a test design similar to that used to test grating and vernier acuity: each trial was a 3AFC task (V, tilted clockwise, or tilted counter-clockwise; or H, tilted clockwise, or tilted counter-clockwise), and the subject had to correctly identify the grating orientation by holding up a pen, and matching the orientation of the pen to the orientation of the grating. For each of the three spatial frequencies tested, the test included eight levels of contrast (max-min/max+min) ranging from 0.33 to 0.006 for 1.5 cy/deg, 0.20 to 0.004 for 6 cy/deg, and 0.25 to 0.011 for 18 cy/deg, with V and H stimuli interleaved within the test book. Order (V or H first) was constant across sessions for each child, but was counterbalanced across schools. Contrast sensitivity for each grating orientation for each spatial frequency was scored as the lowest contrast level on which the child was able to correctly identify the orientation of the grating on at least three out of a maximum of four trials. Order of testing across the three spatial frequencies was always the same for an individual child, was randomly selected by the tester prior to the child’s first test session, and was counterbalanced across subjects.
A contrast sensitivity threshold one step larger than the highest contrast level included in the test book (average step size was 0.2 log unit) was assigned for subjects who were judged unable to resolve the highest contrast stimulus (contrast threshold values of 0.52, 0.32, and 0.39 were assigned for 1.5, 6.0, and 18.0 cy/deg stimuli, respectively).
The Randot Preschool Stereoacuity Test includes six levels of retinal disparity that range from 800 to 40 arc sec. Subjects wore test-specific polarized glasses over their eyeglasses. Stereoacuity was recorded as the smallest disparity at which the subject could correctly identify two of three shapes in the random dot display. For subjects who were judged unable to resolve the largest disparity level, a stereoacuity of 1600 arc sec was assigned.
2.7. Encouraging and Monitoring Treatment Compliance
Eyeglasses were initially dispensed at the baseline vision testing session. At the end of the testing session, children who required eyeglasses were given one pair and were instructed to wear them all the time. Each child’s teacher was given a spare pair of eyeglasses to keep in the classroom. Classroom eyeglasses were for use on days when children did not bring their glasses, or their glasses were lost or broken. Teachers were asked to provide the children with their spare pair when needed, and to try to collect them at the end of the day so that the child would always have a pair at school. Children were given the classroom spare to take home over the summer vacation, and were instructed to use them if their other pair became lost or broken. A study staff member made periodic visits to each classroom to check on glasses and to encourage eyeglass wear. This staff member carried a spare pair for each child so that they could be dispensed as soon possible whenever a child’s glasses became lost, broken, or badly scratched. As soon as this spare was dispensed, a replacement spare was ordered. Daily log books were given to each teacher to record whether children were or were not wearing their eyeglasses. The log books also served as a reminder to teachers to dispense the classroom pair if the child did not have his/her eyeglasses.
2.8. Data Analysis
Threshold values for each measure were transformed to log values for data analyses. Subjects were assigned to astigmatism groups based on the cycloplegic refraction results of the initial eye examination. The non-astigmatic control (NonA) group included children with little or no astigmatism (< 0.75 D in the RE and the LE), and the astigmatic group included children with ≥ 1.00 D RE with-the-rule (plus cylinder axis 90±15 deg) astigmatism. Subjects in the astigmatic group were further divided into two subgroups: (a) children with hyperopic astigmatism (HA), sphere (plus cyl) ≥ 0, and (b) children with myopic or mixed astigmatism (M/MA), sphere (plus cyl) < 0. Subjects were also categorized by age cohort. The younger cohort (YC) included children < 8 years of age, and the older cohort (OC) included children ≥ 8 years of age on the day of baseline best-corrected vision testing (and glasses dispensing).
Data from subjects who did not meet the criteria for any of the three astigmatism groups (NonA, HA, M/MA), subjects with anisometropia (≥1.50 D difference in spherical equivalent between eyes), subjects with ocular abnormalities other than refractive error, astigmatic subjects whose uncorrected RE recognition acuity was 20/20 or better, and subjects who did not participate in best-corrected vision testing at baseline, six weeks, and one year were excluded from analyses.
Amblyopia is a clinical term, and while it is generally defined as reduced best-corrected vision in the absence of ocular causes, the specific deficits that indicate the presence of amblyopia, both in clinical and research settings, can vary. In order to avoid any confusion with this term, for the purpose of the present report we define astigmatism-related amblyopia as significantly reduced vision in astigmats, relative to vision in an age-matched non-astigmatic control group. Similarly, we define meridional amblyopia as a significant difference in vision across stimulus orientation in astigmats, relative to meridional differences observed in a non-astigmatic age-matched control group. In the present study amblyopia is examined in astigmats as a group (e.g., mean acuity in astigmats vs. non-astigmats), rather than for individuals (i.e., data on deficits in individual astigmatic subjects are not reported).
The first set of analyses was aimed at determining if there were significant deficits at baseline for the HA and M/MA groups relative to the NonA group (significantly reduced best-corrected vision and/or significant meridional amblyopia). We have previously published a detailed report of baseline measurements of visual performance in the three astigmatism groups (Harvey et al., 2007b
). However, the present report includes only those subjects followed for a full year after baseline, i.e., a subset of the 805 subjects for whom we reported results of baseline testing. Therefore, we present analyses here to document whether or not baseline deficits were significant in this smaller sample. Separate analyses of variance (ANOVAs) compared baseline measures of grating acuity for V, H, and O stimuli, vernier acuity for V, H, and O stimuli, contrast sensitivity for V and H stimuli (with separate analyses for low, middle, and high spatial frequency stimuli), and stereoacuity across astigmatism group (HA, M/MA, NonA) and age cohort (YC vs. OC). In order to evaluate meridional amblyopia, analyses (ANOVAs) also compared vertical-horizontal (V-H) grating acuity, vernier acuity, and contrast sensitivity across astigmatism group and age cohort.
The second set of analyses focused on determining if there was a significant effect of treatment, i.e., whether the amount of change in best-corrected visual performance over time in the HA and M/MA groups was greater than that observed in the NonA group. Separate repeated measures analyses of variance (RM-ANOVAs) compared mean change over time in resolution (grating) acuity for V, H, and O stimuli, in vernier acuity for V, H, and O stimuli, in contrast sensitivity for V and H stimuli (with separate analyses for low, middle, and high spatial frequency stimuli), and in stereoacuity across age cohort (YC vs. OC) from baseline to six weeks to one year. Analyses (RM-ANOVAs) also evaluated change over time in meridional amblyopia (the difference between performance for V and H stimuli) for grating acuity, vernier acuity, and contrast sensitivity for the astigmatic groups, relative to the NonA group. All significant main effects and interactions are reported. Main effects of age cohort are reported, and when significant, always reflect better vision in the older cohort unless otherwise noted. Planned post hoc comparisons (t-tests with Bonferroni correction for multiple comparisons) were conducted only on significant interactions that included both the “time” and
the “astigmatism group” variables, as these effects were pertinent to the primary aims of the study: To determine if there was significantly greater improvement in the astigmatic groups over time, relative to the non-astigmatic group. Because a previous detailed report of baseline data showed no evidence that visual deficits were associated with presence of astigmatic anisometropia or with previous eyeglass wear in this sample (Harvey et al., 2007b
), these variables were not entered into the present analyses.
Finally, the last set of analyses were aimed at determining if significant baseline deficits for the HA and M/MA groups relative to the NonA group (significantly reduced best-corrected vision and/or significant meridional amblyopia) remained significant after one year of treatment. The same analyses conducted on baseline data (see summary of first set of analyses above) were conducted on one-year data for measures that yielded significantly reduced performance for astigmatic children at baseline.