Abstract
PURPOSE:
To assess test-retest repeatability of the accommodative response (AR) in children with and without amblyopia and adults using the Grand Seiko autorefractor.
DESIGN:
Prospective reliability assessment.
METHODS:
Test-retest of accommodation was obtained while participants viewed 20/150 sized letters at 33 cm using the Grand Seiko autorefractor in children 5 to <11 years with amblyopia (n=24) and without amblyopia (n=36), and adults 18 to <35 years (n=34). Bland-Altman 95% limits of agreement (LOA) and intraclass correlation coefficients (ICCs) were used to assess repeatability and reliability. The AR between the fellow and amblyopic eyes of children with amblyopia and eye 1 and eye 2 of the visually normal participants was assessed using group comparisons.
RESULTS:
The 95% LOA of the AR was greatest in the amblyopic eyes (−1.25 diopters [D], 1.62 D) of children with amblyopia. The 95% LOA were similar between the fellow eyes (−0.88 D, 0.74 D) of children with amblyopia and both eyes of the children without amblyopia (eye 1: −0.68 D, 0.71 D; eye 2: −0.59 D, 0.70 D) and the adults (eye 1: 95% LOA = −0.49 D, 0.45 D; eye 2: LOA = −0.66 D, 0.67 D). ICCs revealed the Grand Seiko autorefractor as a reliable instrument for measuring AR.
CONCLUSIONS:
The Grand Seiko autorefractor was more repeatable and reliable when measuring the AR in children and adults without amblyopia than in the amblyopic eye in children with amblyopia. It is recommended that multiple measures of the AR be obtained in amblyopic eyes to improve the precision of measures.
Open-field Autorefractors, Including The Grand Seiko WAM-5500, are reliable, valid, and have high repeatability for measuring refractive error, making these machines the gold standard for automated objective measures in visually normal children1 and adults.2-5 The Grand Seiko WAM-5500 is a US Food and Drug Administration–approved device6 that provides an unobstructed view of the fixation target, allowing fixation targets to be set at the preferred viewing distance in real space in front of the eye being measured (Figure 1).
FIGURE 1.
The experimental set-up consisting of the Grand Seiko WAM-5500 open-field auto refactor, the Grand Seiko near fixation rod, and the iPod positioned at 33 cm displaying the 20/150 letter target.
Prior reliability and validity studies of current and previous versions (ie, Shin-Nippon SRS-5000 autorefractor) of the Grand Seiko autorefractor have been performed in both children1 and adults4,5 over a wide range of refractive error7 by comparing the Grand Seiko measures to distance subjective refraction.2,3,8,9 To the authors’ knowledge, there has only been 1 reliability study of accommodative measures in visually normal children and adults (age 6-43 years, n = 41) using the Grand Seiko.10 Furthermore, some studies have suggested that there may be an increase in astigmatism during accommodation,11-13 but it is unknown if these changes are repeatable and reliable. Despite the limited data on the repeatability and reliability of the Grand Seiko for measures of accommodation, the Grand Seiko is used routinely in studies that evaluate the accommodative response.9,14-16
Understanding the repeatability and reliability of Grand Seiko autorefractor is particularly important for non–visually normal populations (i.e., individuals with amblyopia) and in populations who tend to have more variable accommodation (i.e., children),17 when interpreting measures of accommodation. Assessing the repeatability of measures of accommodation in children with unilateral amblyopia is important as studies have shown children with amblyopia have poor accommodation in their amblyopic eye under monocular viewing conditions relative to the fellow eye.18,19
The purpose of this study is to assess the test-retest reliability of accommodative measurements in children with amblyopia as well as visually normal children and adults using the Grand Seiko WAM-5500 open-field autorefractor.
METHODS
• STUDY PARTICIPANTS:
Our study followed the tenets of the Declaration of Helsinki and was approved by each institution’s Institutional Review Board. Participants younger than 18 years provided written informed assent in addition to a parent or legal guardian’s written informed consent. All participants 18 years or older provided written informed consent. The study consisted of a pediatric cohort who was recruited from the patient populations at Boston Children’s Hospital Department of Ophthalmology, Akron Children’s Hospital Department of Ophthalmology, and Stanford University School of Medicine’s Department of Ophthalmology. The adult cohort was recruited from staff and students naïve to the purpose of the study at the Boston Children’s Hospital Department of Ophthalmology and Stanford University School of Medicine Department of Ophthalmology.
The study enrolled children 5 to <11 years and adults 18 to <35 years. Participants were screened for the presence of exclusion criteria: history of ocular or systemic diagnosis that may affect accommodation, taking medications known to affect accommodation, history of developmental delay or behavioral diagnoses that may impact testing, previous intraocular surgery, current or previous known myopic refractive error correction > 1.50-diopter (D) spherical equivalent in either eye, or atropine penalization therapy within the last 30 days.
Child participants were required to wear refractive error correction for at least 2 weeks if they had ≥2.50 D spherical equivalent hypermetropia, >0.50 D astigmatism (in either eye), >0.50 D myopia, or >0.50 D of anisometropia. Refractive error correction (if required) was within 0.50 D of fully correcting the spherical equivalent anisometropia, hyperopia was not reduced more than 1.50 D spherical equivalent and was cut symmetrically between the 2 eyes, cylindrical power was within 0.50 D, and cylinder axis was within ±10 degrees if cylinder power was ≤1.00 D and within ±5 degrees if cylinder power was > 1.00 D.
Children with amblyopia (previously treated or untreated) had amblyopia associated with anisometropia (spherical equivalence difference of ≥1.00 D between eyes or astigmatism difference of ≥1.50 D between corresponding meridians in the 2 eyes), strabismus (presence of heterotropia on examination at distance or near fixation or documented history of strabismus which is no longer present), or combined mechanism (meeting both anisometropic and strabismus measures). Amblyopia was only included if the intraocular difference was ≥3 logMAR lines (ATS-HOTV) or > 15 letters (E-ETDRS). The visual acuity in the amblyopic eye was 20/40 to 20/125 (ATS-HOTV) or 47 to 70 letters (E-ETDRS). Visual acuity in the fellow eye was 20/32 (ATS-HOTV) or 73 to 77 letters (E-ETDRS) or better for 5 to <6 years, and 20/25 (ATS-HOTV) or 78 to 83 letters (E-ETDRS) or better for ≥6 years.
Children and adults without amblyopia had visual acuity in each eye of 20/32 (ATS-HOTV) or 73 to 77 letters (E-ETDRS) or better for 5 to <6 years, and 20/25 (ATS-HOTV) or 78 to 83 letters (E-ETDRS) or better for ≥6 years, intraocular difference <2 logMAR lines (ATS-HOTV) or <10 letters (E-ETDRS), no history of strabismus (intermittent or constant) or amblyopia, and normal binocular vision (defined as near point of convergence ≤6 cm, ≤10 prism diopters of exophoria, and ≤8 prism diopters of esophoria at near by prism alternating cover test, normal near random dot stereopsis using the Randot Preschool Stereoacuity Test20 of 200 arcsec or better for 5 to <6 years and 100 arcsec or better for 6 to <7 years, 60 arcsec or better for ≥7 years old).
• STUDY PROCEDURES:
The participants’ medical records were reviewed prior to study procedures to determine if they fit study eligibility criteria. If no medical record was available, oral history was obtained. For children, study procedures were performed with habitual spectacle correction (if required) that had been worn for at least 2 weeks (based on a cycloplegic refraction completed within 6 months prior to testing). For adults, study procedures were performed with habitual correction (cycloplegic refraction was not performed).
Eligibility criteria were confirmed at the beginning of the study visit. A standardized assessment of monocular distance visual acuity (HOTV for subject <7 years or E-ETDRS for participants ≥7 years) was assessed with the Electronic Visual Acuity (EVA) Tester or M&S SmartSystem (M&S Technologies, Inc).21 Stereoacuity was measured at 40 cm on the participants without amblyopia using the Randot Preschool Stereotest (Stereo Optical Company),20 and ocular alignment was assessed using cover uncover to rule out strabismus, and prism alternate cover test was performed to assess the ocular alignment magnitude (phoria or tropia).
Open-field autorefraction was measured monocularly with the Grand-Seiko WAM-5500, hereafter Grand Seiko, open-field autorefractor (Rexxam Co, Ltd). Participants were positioned where the tested eye viewed 20/150-sized letters displayed on an iPod Touch (6th-generation, 1136 × 640 pixels; Apple Inc) at a fixed position of 33 cm using the Grand Seiko near fixation rod (Figure 1). To keep the stimulus consistent across eyes, the larger letter size was used to ensure it was above threshold for each eye of all participants and has been shown to provoke an accommodative response.22,23
The letters were displayed using a preprogrammed Keynote slide presentation (Apple Inc). Each slide displayed one letter for 2 seconds before another letter was presented throughout the time of testing. The eye not being measured was occluded with an adhesive patch. The instrument was aligned with the visual axis of the eye being measured, and at least 5 measures were obtained for each eye for each trial. Each trial took less than 30 seconds to obtain the minimum 5 measures. Each eye was tested twice (ie, test and retest), and the eye order was counter-balanced for both the test and retest.
In between measures, participants were instructed to sit back as the adhesive patch was moved from one eye to another and then the participant was repositioned in the Grand Seiko to continue with testing. Total testing time was under 5 minutes. Participants were instructed to “look at the letters” and were not specifically instructed to keep the letters clear to evaluate habitual accommodation. Of the minimum of 5 measurements per trial, those that exceeded 1.00 D spherical value from the mode were excluded. If more than 5 measures were obtained and were within 1.00 D of the mode, the first and last measures were systematically removed until 5 measures remained such that all participants had an average of 5 measures per eye. Data were entered by the first author and then reentered and double-checked by the last author to ensure there were no inaccuracies or transcription errors when transferring measures from the Grand Seiko printouts into the study database.
• DATA ANALYSIS:
Autorefractor output included spherical and cylindrical components of the accommodative response. To ensure that off-axis measures were not included in the analysis, measures with >2 D of astigmatism were excluded.9 The accommodative measures obtained by the Grand Seiko were transformed using power vector notation (ie, spherical equivalent, J0 = cylindrical power at axis 0°, and J45 = cylindrical power at axis 45°).24 Each eye’s 5 measures were averaged for each trial. If participants wore spectacles, spectacle lens powers were also converted to vector notation.
All accommodative responses obtained from the Grand Seiko were referenced from the corneal plane, and thus, for participants wearing spectacles during testing, the averaged autorefractor responses (spherical equivalent, J0 and J45) were adjusted using the lens effectivity formula12 with a vertex distance of 12 mm and the respective spectacle lens power (spherical equivalent, J0, and J45). For participants without amblyopia, the eyes were randomly designated to eye 1 or eye 2 with the use of a random number generator.
The repeatability of the accommodative measures (spherical equivalent, J0, and J45) from the Grand Seiko autorefractor was determined using Bland-Altman with 95% limits of agreement (LOA), and reliability (spherical equivalent) was assessed using intraclass correlation coefficient (ICC) based on mean-rating (k = 2), absolute-agreement, 2-way mixed effects analysis methods.25 ICC was classified as moderate (0.50-0.75), good (0.75-0.90), or excellent (>0.90) reliability.25 In participants with amblyopia, Pearson pairwise correlation analysis was performed to determine if there was a significant relationship between the visual acuity of the amblyopic eye and mean spherical equivalent accommodative response.
For within- and between-group differences of the accommodative response among the 2 eyes, the accommodative response (spherical equivalent, J0, and J45) between trial 1 and trial 2 were averaged. Within-participant differences in the mean spherical equivalent, J0, and J45 were determined using paired t test. Between-participant differences in spherical equivalent were evaluated using a 1-way analysis of variance with post hoc independent t test and Bonferroni correction. In children with amblyopia, linear regression was used to determine if there was an association between the difference in test and retest of the accommodative response (spherical equivalent) and visual acuity in the amblyopic eye. All analyses were performed in Stata (version 12.1; StataCorp LLC) and interpreted at a type I error rate of α = 0.05 statistical significance.
RESULTS
A total of 99 participants were recruited to participate in the study. Thirty-one children and 17 adults were recruited from Boston Children’s Hospital, 31 children were recruited from Akron Children’s Hospital, and 2 children and 18 adults were recruited from Stanford University. Five participants were excluded from data analysis: 1 child with amblyopia had visual acuity of 20/32 in the amblyopic eye, 1 child without amblyopia and 1 adult had >2 D of astigmatism on multiple accommodative measures, and 1 adult had visual acuity worse than 20/25. An additional child without amblyopia was excluded from the analysis as the difference in spherical equivalent between the first and second measures was >3 D, more than 5 standard deviations from the mean. Data analysis was performed with and without the child included and the removal of the child did not change the overall results of the study.
Summary statistics for the 94 participants included in the analysis are found in Table 1. Of the participants included in the analysis, 24 children had amblyopia (14 children with anisometropia, 4 with strabismus, and 6 with combined mechanism amblyopia) with a mean logMAR visual acuity of 0.48 ± 0.14 in the amblyopic eye and −0.05 ± 0.08 in the fellow eye. The mean logMAR visual acuity of the children without amblyopia (n = 36) was −0.04 ± 0.07 in eye 1 and −0.07 ± 0.08 in eye 2. In the 34 adults, the mean logMAR visual acuity was −0.13 ± 0.08 in eye 1 and −0.08 ± 0.07 in eye 2.
TABLE 1.
Summary Statistics for Children With Amblyopia, Children Without Amblyopia, and Adults
| Participant Group/Eye | Mean (SD) Visual Acuity (logMAR) |
Spherical Equivalent (D) |
J0 (D) |
J45 (D) |
||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Test | Retest | Difference Test-Retest |
Test | Retest | Difference Test-Retest |
Test | Retest | Difference Test-Retest |
||
| Mean (SD) |
Mean (SD) |
Mean ± SD (95% LOA) |
Mean (SD) |
Mean (SD) |
Mean ± SD (95% LOA) |
Mean (SD) |
Mean (SD) |
Mean ± SD (95% LOA) |
||
| Children with amblyopia (n = 24) | ||||||||||
| Fellow eye | −0.05 (0.08) | −1.78 (0.84) | −1.72 (0.68) | −0.07 ± 0.41 (−0.88, 0.74) | −0.02 (0.18) | −0.01 (0.22) | −0.01 ± 0.15 (−0.29, 0.28) | 0.03 (0.20) | 0.02 (0.22) | 0.02 ± 0.07 (−0.12, 0.16) |
| Amblyopic eye | 0.48 (0.14) | −0.99 (0.88) | −1.18 (0.96) | 0.18 ± 0.73 (−1.25, 1.62) | 0.02 (0.27) | 0.09 (0.31) | −0.07 ± 0.17 (−0.41, 0.26) | 0.07 (0.25) | 0.04 (0.27) | 0.03 ± 0.13 (−0.22, 0.29) |
| Children without amblyopia (n = 36) | ||||||||||
| Eye 1 | −0.07 (0.08) | −1.71 (0.54) | −1.77 (0.47) | 0.06 ± 0.33 (−0.59, 0.70) | −0.02 (0.20) | −0.01 (0.24) | −0.01 ± 0.18 (−0.37, 0.34) | 0.04 (0.13) | 0.04 (0.14) | 0.00 ± 0.10 (−0.20, 0.19) |
| Eye 2 | −0.04 (0.07) | −1.72 (0.52) | −1.73 (0.44) | 0.01 ± 0.36 (−0.68, 0.71) | −0.01 (0.23) | 0.03 (0.26) | −0.03 ± 0.14 (−0.30, 0.24) | −0.03 (0.17) | −0.04 (0.16) | −0.01 ± 0.15 (−0.27, 0.30) |
| Adults (n = 34) | ||||||||||
| Eye 1 | −0.13 (0.08) | −2.02 (0.48) | −2.02 (0.40) | 0.00 ± 0.34 (−0.66, 0.67) | 0.03 (0.30) | 0.05 (0.27) | −0.02 ± 0.17 (−0.36, 0.31) | −0.01 (0.18) | −0.03 (0.13) | 0.00 ± 0.07 (−0.12, 0.13) |
| Eye 2 | −0.08 (0.07) | −2.08 (0.34) | −2.06 (0.28) | 0.02 ± 0.24 (−0.49, 0.45) | −0.03 (0.25) | −0.04 (0.24) | 0.01 ± 0.16 (−0.30, 0.31) | 0.02 (0.14) | 0.02 (0.14) | 0.03 ±0.11 (−0.19, 0.24) |
D = diopters, LOA = limits of agreement.
Graded numerical depictions of test-retest of the absolute difference in accommodative response for spherical equivalent, J0, and J45 are found in Table 2. Graphical depictions of the difference in test-retest for spherical equivalent are shown in Figure 2. In children with amblyopia, there is a wider distribution of the difference between test-retest in the amblyopic eye than in the fellow eye (Figure 2, A). The distribution of the difference between test-retest of spherical equivalent of eye 1 and eye 2 are similar for the children without amblyopia and the adults (Figure 2, B and C). The distributions of the differences in test-retest for J0 and J45 are similar for children with and without amblyopia, and in the adults (Table 2).
TABLE 2.
Absolute Differences in Test-Retest of the Mean Spherical Equivalent, J0, and J45
| Children With Amblyopia (n = 24) |
Children Without Amblyopia (n = 36) |
Adults (n = 34) |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Fellow Eye |
Amblyopic Eye |
Eye 1 |
Eye 2 |
Eye 1 |
Eye 2 |
|||||||
| Absolute Difference in Test-Retest | n (%)* | Cumulative %* | n (%)* | Cumulative %* | n (%)* | Cumulative %* | n (%)* | Cumulative %* | n (%)* | Cumulative %* | n (%)* | Cumulative %* |
| Spherical equivalent | ||||||||||||
| <0.25 D | 13 (54.2) | 54.2 | 10 (41.7) | 41.7 | 21 (58.3) | 58.3 | 21 (58.3) | 58.3 | 25 (73.5) | 73.5 | 26 (76.5) | 76.5 |
| 0.25 to <0.50 D | 7 (29.2) | 83.3 | 4 (16.7) | 58.3 | 12 (33.3) | 91.6 | 11 (30.5) | 88.8 | 6 (176) | 91.1 | 7 (20.6) | 97.1 |
| 0.50 to <0.75 D | 3 (12.5) | 95.8 | 5 (20.8) | 79.2 | 2 (5.6) | 97.2 | 2 (5.6) | 94.4 | 1 (2.9) | 94.0 | 0 (0) | 97.1 |
| 0.75 to <1.00 D | 0 (0) | 95.8 | 2 (8.3) | 87.5 | 0 (0) | 97.2 | 1 (2.8) | 97.2 | 1 (2.9) | 96.9 | 1 (2.9) | 100 |
| 1.00 to <1.25 D | 0 (0) | 95.8 | 0 (0) | 87.5 | 1 (2.8) | 100 | 1 (2.8) | 100 | 0 (0) | 96.9 | 0 (0) | 100 |
| 1.25 to <1.50 D | 1 (4.2) | 100 | 1 (4.2) | 91.6 | 0 (0) | 100 | 0 (0) | 100 | 1 (2.9) | 100 | 0 (0) | 100 |
| 1.50 to <1.75 D | 0 (0) | 100 | 0 (0) | 91.6 | 0 (0) | 100 | 0 (0) | 100 | 0 (0) | 100 | 0 | 100 |
| 1.75 to <2.00 D | 0 (0) | 100 | 2 (8.3) | 100 | 0 (0) | 100 | 0 (0) | 100 | 0 (0) | 100 | 0 | 100 |
| J0 | ||||||||||||
| <0.25 D | 22 (91.7) | 91.7 | 21 (87.5) | 87.5 | 33 (91.6) | 91.6 | 30 (83.3) | 83.3 | 29 (85.3) | 85.3 | 29 (85.3) | 85.3 |
| 0.25 to <0.50 D | 2 (8.3) | 100 | 3 (12.5) | 100 | 3 (8.3) | 100 | 6 (16.7) | 100 | 4 (11.7) | 97.0 | 5 (14.7) | 100 |
| 0.50 to <0.75 D | 0 (0) | 100 | 0 (0) | 100 | 0 (0) | 100 | 0 (0) | 100 | 1 (2.9) | 100 | 0 (0) | 100 |
| J45 | ||||||||||||
| <0.25 D | 24 (100) | 100 | 23 (95.8) | 95.8 | 32 (88.9) | 88.9 | 35 (100) | 100 | 33 (971) | 97.1 | 34 (100) | 100 |
| 0.25 to <0.50 D | 0 (0) | 100 | 1 (4.2) | 100 | 4 (11.1) | 100 | 0 (0) | 100 | 1 (2.9) | 100 | 0 (0) | 100 |
D = diopter.
Owing to rounding, percentages may not sum to 100 or their respective cumulative percentages.
FIGURE 2.
Histograms representing the difference in test and retest of the spherical equivalent (in diopters [D]) of the accommodative response in the fellow eye and amblyopic eye of (A) participants with amblyopia (fellow and amblyopic eyes), (B) participants without amblyopia (eye 1 and eye 2), and (C) adult participants (eye 1 and eye 2). In participants without amblyopia and adults, eye 1 and eye 2 were randomly assigned. Mean difference ± SD of test-retest of the accommodative response and 95% limits of agreement (LOA) are displayed in the top right corner of each graph.
Bland-Altman plots of test-retest are shown in Figure 3. For spherical equivalent, the 95% LOA were greatest in the amblyopic eye of the children with amblyopia (95% LOA: −1.25 D, 1.62 D). The 95% LOA were similar in the fellow eye (95% LOA = −0.88 D, 0.74 D) of children with amblyopia and in eye 1 and eye 2 in children without amblyopia (eye 1: 95% LOA = −0.68 D, 0.71 D; eye 2: LOA = −0.59 D, 0.70 D). The 95% LOA was smallest for adults (eye 1: 95% LOA = −0.49 D, 0.45 D; eye 2: LOA = −0.66, 0.67).
FIGURE 3.
Bland-Altman plot of the spherical equivalent (in diopters [D]), J0, and J45 of the accommodative response of (A) participants with amblyopia, (B) participants without amblyopia, and (C) adult participants. The fellow eye in children with amblyopia and eye 1 in children without amblyopia and adults are shown in yellow. The amblyopic eye in children with amblyopia and eye 2 in children without amblyopia and adults are shown in blue. Eye 1 and eye 2 were randomly assigned in participants without amblyopia and adults. Dashed lines represent the mean of difference in test-retest measures and the solid lines represent the upper and lower 95% limits of agreement of difference in test-retest measures.
ICC analyses revealed that the Grand Seiko autorefractor has good to excellent reliability in measuring accommodation (spherical equivalent) in children with and without amblyopia and in adults without amblyopia (Table 3). Children with amblyopia have good reliability in their amblyopic eye (ICC 0.81, 95% CI 0.57, 0.92) and excellent reliability in their fellow eye (ICC 0.92, 95% CI 0.82, 0.97). Children without amblyopia and adults displayed an equally good reliability in both eye 1 (ICC children without amblyopia = 0.85, 95% CI 0. 70, 0.92; ICC adults = 0.83, 95% CI 0.66, 0.92) and eye 2 (ICC children without amblyopia = 0.88, 95% CI 0.77, 0.94; ICC adults = 0.83, 95% CI 0.66, 0.92).
TABLE 3.
Results of Intraclass Correlation Coefficient Calculation Using 2-Way Mixed Effects With Absolute Agreement
| Intraclass Correlation | 95% CI |
||
|---|---|---|---|
| Lower Bound | Upper Bound | ||
| Children with amblyopia | |||
| Amblyopic eye average | 0.81 | 0.57 | 0.92 |
| Fellow eye average | 0.92 | 0.82 | 0.97 |
| Children without amblyopia | |||
| Eye 1 average | 0.85 | 0.70 | 0.92 |
| Eye 2 average | 0.88 | 0.77 | 0.94 |
| Adults | |||
| Eye 1 average | 0.83 | 0.66 | 0.92 |
| Eye 2 average | 0.83 | 0.66 | 0.92 |
Intragroup comparisons of the accommodative responses between eyes are shown in Table 1. In children with amblyopia, the amblyopic eye had a significantly lesser mean accommodative response (ie, greater accommodative lag) than the fellow eye (mean difference = 0.67 D, 95% CI 0.26, 1.06, P = .002). A significant difference in mean accommodative response between the 2 eyes was not detected in the children without amblyopia (mean difference = 0.02 D, 95% CI −0.06, 0.09, P = .62) or in the adults (mean difference = −0.05 D, 95% CI −0.15, 0.06, P = .37).
There was no difference in mean J0 or mean J45 in the children with amblyopia (J0 mean difference = 0.07 D, 95% CI −0.05, 0.18, P = .25; J45 mean difference = 0.03 D, 95% CI −0.14, 0.19, P = .74), J0 in the children without amblyopia (J0 mean difference = 0.03 D, 95% CI −0.03, 0.09, P = .35), or J45 in the adults (J45 mean difference = 0.042 D, 95% CI −0.03, 0.11, P = .22). There was a small yet significant difference in J45 in children without amblyopia (J45 mean difference = −0.08 D, 95% CI −0.15, −0.01, P = .04) and in J0 in adults (J0 mean difference = −0.07 D, 95% CI −0.13, −0.01, P = .02). Correlation analysis revealed a mild (r = 0.27), but nonsignificant (P = .20), association between amblyopic eye visual acuity and the mean spherical equivalent accommodative response of the amblyopic eye.
Intergroup comparisons using 1-way analysis of variance revealed a significant difference in mean accommodative response for the amblyopic eye/eye 2 across groups (P < .001), with the amblyopic eye in children with amblyopia being significantly less than both eye 2 of children without amblyopia (mean difference = 0.64 D, P < .001) and eye 2 in the adults (mean difference = 0.98 D, P < .001). Eye 2 of children without amblyopia also had a significantly lesser accommodative response than the eye 2 of adults (mean difference = −0.34 D, P = .03). However, a significant difference in mean accommodative response was not detected between groups for the fellow eye in children with amblyopia and eye 1 in children without amblyopia and eye 1 in adults (P = .06).
Linear regression analysis was performed to determine if the differences in test-retest were associated with amblyopic eye visual acuity in the children with amblyopia. Scatter plots of the signed difference and absolute differences in test-retest are shown in Figure 4. As shown in Figure 4A, the greatest differences in test-retest were found in children with lower levels of amblyopia. Figure 4B illustrates that the magnitude of the absolute difference in test-retest was greatest in amblyopic eyes with better visual acuity and decreased as the visual acuity worsened (95% CI −0.23, −0.03, r = −0.49, P = .016).
FIGURE 4.
Scatterplots illustrating the relationship between amblyopic eye visual acuity. (A) difference in signed test-retest of the accommodative response and (B) the absolute difference in test-retest of the accommodative response.
DISCUSSION
This is the first study to objectively measure the reliability of accommodative responses (spherical equivalent, J0, and J45) in children with amblyopia using the Grand Seiko WAM-5500 open-field autorefractor. We have also included visually normal children and adults as control participants. The average difference of the test-retest mean spherical equivalent accommodative response of the amblyopic eye in children was greater than twice as much as that of the fellow eye and in eye 1 and eye 2 of the children without amblyopia (Table 1). As seen in Table 2, only 14 of 24 (58%) of amblyopic eyes had an absolute mean difference of <0.50 D, in comparison to 20 of 24 (83%) of the fellow eyes, 33 of 36 (92%) of eye 1 eyes, and 32 of 36 (89%) of eye 2 eyes for the children without amblyopia (Table 2, Figure 2). However, 21 of 24 (87.5%) of amblyopic eyes had a difference in test-retest <1.00 D, with only 3 of 24 (13.5%) of amblyopic eyes having >1.25 D difference in test-retest.
The Bland-Altman and the ICC analysis confirmed that the repeatability and reliability of accommodative measures in the amblyopic eye of children with amblyopia is less than the fellow eye and less than children without amblyopia. This difference is most apparent in the Bland Altman plots as the 95% LOA of the amblyopic eye was 1.25 D wider than the 95% LOA of the fellow eye and 1.50 D wider than the 95% LOA of eye 1 and eye 2 in the children without amblyopia. The wider LOA in the amblyopic eye is attributable to the 3 previously mentioned eyes with >1.25 D differences on test-retest.
Our test-retest measures of accommodation (spherical equivalent) for the fellow eye of children with amblyopia, and the children and adults without amblyopia, are similar to reliability studies comparing subjective refraction to Grand Seiko autorefractor2,3,8 as well as the one study that previously assessed reliability of accommodation in visually normal children and adults,10 and another study that assessed accommodative amplitude repeatability in visually normal adults with the TONOREF III autorefractor (NIDEK Co, Ltd).26 It is also important to note that the 95% LOA of the fellow eye closely resembles the 95% LOA of eye 1 and eye 2 of the children without amblyopia suggesting that, as expected, the reliability accommodative measures of the fellow eye is unaffected by the amblyopic eye.
Furthermore, the 95% LOAs for the children without amblyopia are similar to the adults without amblyopia positing that accommodative repeatability using the Grand Seiko does not differ between children and adults. For completeness, we also assessed the reliability of J0 and J45 and found similar reliability to other studies that compared the J0 and J45 from the Grand Seiko to subjective refraction,2,4,5 suggesting that the Grand Seiko obtains reliable measures of astigmatism during an accommodative task. Thus, differences in spherical equivalent of test-retest are not likely due to differences in the cylindrical component of the accommodative response.
Similar to other studies,18,19,27,28 we found that the amblyopic eye exhibited a lesser accommodative response than the fellow eye (mean difference of 0.67 D). In our sample population, the difference between the 2 eyes was not associated with the visual acuity of the amblyopic eye. Unlike Singman and associates who found that the accommodative response increased with better visual acuity,19 our results were in agreement with Manh et. al as we did not find a relationship between the magnitude of the accommodative response and visual acuity.18 However, we did find that the difference in test and retest was greatest in amblyopic eyes with better visual acuity (Figure 4).
One may have hypothesized that because amblyopia is associated with a greater depth of focus,29 the response may vary more between test and retest in eyes with poorer visual acuity as an increase in depth of focus is associated with an increase in variability of the accommodative response.30-32 However, as shown in Figure 4A, the amblyopic eyes with worse visual acuity had the smallest differences in test and retest of the accommodative response, suggesting the children with worse visual acuity did not get any benefit by varying their accommodation, and thus, they may tend to repeatedly accommodate to a particular location within the depth of focus. It should be noted that only 6 of 24 (25%) of our participants with amblyopia had visual acuity worse than 0.5 logMAR, and thus, our sample may not be representative of the accommodative responses in children with more moderate or severe amblyopia.
The increased difference in test-retest in children with amblyopia found in this study may have clinical implications. Our results show that as a group, test-retest was more variable in amblyopic eyes; however, 14 of 24 (58%) of amblyopic eyes had an absolute mean difference of <0.50 D. Thus, it may be beneficial for the eye care provider to obtain multiple measures of accommodation in children with amblyopia during monocular viewing (ie, during occlusion) as the measures may provide insight into the variability of the accommodative response in a given patient, which may facilitate various treatment approaches (eg, bifocal lens). Further work is needed in this area, to identify the underlying variables that contribute to these differences in variability between and within individuals, which may result in an improved treatment approach.
Our study is not without limitations. We used a large accommodative target that was well above threshold for the fellow eye as well as both eyes of the participants without amblyopia. However, to keep the stimulus consistent across eyes and individuals, a large stimulus was required for the amblyopic eye. Additionally, our stimulus size has been shown to be adequate to provoke an accommodative response.
Notably, an increase in accommodative response may have been found in the participants without amblyopia if a stimulus size closer to threshold was used. However, the repeatability and reliability of the accommodative response in our study was similar to previous studies of distance refraction, suggesting the letter size used in this study was unlikely to affect repeatability and reliability of the accommodative response.
Additionally, participants were allowed minimal undercorrection of refractive error, which may have impacted both the accuracy and reliability of the measures. However, the combination of letter size and the allowance of minimal undercorrection of refractive error represented a more real-world habitual viewing environment, and thus our results are more likely representative of the habitual accommodative response.
We also only assessed same-day repeatability and, thus, future studies are needed to understand the repeatability of the accommodative responses over larger time intervals. Lastly, we had a small number of children with anisometropic, strabismic, and mixed mechanism moderate amblyopia and thus we were not powered to analyze the data by amblyopia classification type. Thus, our results may not be generalizable to particular classification types of amblyopia or to children with more mild or more severe amblyopia. Future studies should evaluate the repeatability of the accommodative response in various types of amblyopia, as well as varying degrees of severity (moderate and severe).
In conclusion, accommodative measures obtained using the Grand Seiko autorefractor have high repeatability and reliability in both children and adults without amblyopia. However, test-retest reliability is slightly reduced in children with amblyopia, and thus, a single measure of accommodation in the amblyopic eye may not suffice when evaluating the accommodative response. Implementing repeated measures of accommodation for children with amblyopia may provide clinicians with a more precise assessment of the accommodative performance.
Funding/Support:
National Institutes of Health / National Eye Institute R01EY029307 (TLR), National Eye Institute P30-EY026877 (TLR-Stanford University Core Grant), Research to Prevent Blindness, Inc. (TLR-Stanford University Department of Ophthalmology Unrestricted Grant), Children’s Hospital Ophthalmology Foundation Faculty Discovery Award (AR).
Footnotes
Financial Disclosures: The authors indicate no financial support or conflicts of interest. All authors attest that they meet the current ICMJE criteria for authorship.
Contributor Information
RYAN N. CHINN, Department of Ophthalmology, Boston Children’s Hospital/Harvard Medical School, Boston, Massachusettes
APARNA RAGHURAM, Department of Ophthalmology, Boston Children’s Hospital/Harvard Medical School, Boston, Massachusettes.
MOLLY K. CURTISS, Akron Children’s Hospital, Akron, Ohio
ALYSSA M. GEHRING, Akron Children’s Hospital, Akron, Ohio
ANA JURIC DE PAULA, Akron Children’s Hospital, Akron, Ohio.
TAWNA L. ROBERTS, Akron Children’s Hospital, Akron, Ohio; Spencer Center for Vision Research, Byers Eye Institute at Stanford University, Palo Alto, California.
REFERENCES
- 1.Chat SWS, Edwards MH. Clinical evaluation of the Shin-Nippon SRW-5000 autorefractor in children. Ophthalmic Physiol Opt. 2001;21(2):87–100. doi: 10.1046/j.1475-1313.2001.00584.x. [DOI] [PubMed] [Google Scholar]
- 2.Davies LN, Mallen EAH, Wolffsohn JS, Gilmartin B. Clinical evaluation of the Shin-Nippon NVision-K 5001/Grand Seiko WR-5100K autorefractor. Optom Vis Sci. 2003;80(4):320–324. doi: 10.1097/00006324-200304000-00011. [DOI] [PubMed] [Google Scholar]
- 3.Sheppard AL, Davies LN. Clinical evaluation of the Grand Seiko Auto Ref/Keratometer WAM-5500. Ophthalmic Physiol Opt. 2010;30(2):143–151. doi: 10.1111/j.1475-1313.2009.00701.x.. [DOI] [PubMed] [Google Scholar]
- 4.Mallen EAH, Wolffsohn JS, Gilmartin B, Tsujimura S. Clinical evaluation of the Shin-Nippon SRW-5000 autorefractor in adults. Ophthalmic Physiol Opt. 2001;21(2):101–107. doi: 10.1046/j.1475-1313.2001.00585.x . [DOI] [PubMed] [Google Scholar]
- 5.Mallen EAH, Gilmartin B, Wolffsohn JS, Tsujimura S. Clinical evaluation of the Shin-Nippon SRW-5000 autorefractor in adults: an update. Ophthalmic Physiol Opt. 2015;35(6):622–627. doi: 10.1111/opo.12254. [DOI] [PubMed] [Google Scholar]
- 6.Establishment Registration & Device Listing. Accessed March 1, 2021. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfrl/rl.cfm?lid=265085&lpcd=HKO [Google Scholar]
- 7.Wolffsohn JS, O’Donnell C, Charman WN, Gilmartin B. Simultaneous continuous recording of accommodation and pupil size using the modified Shin-Nippon SRW-5000 autorefractor. Ophthalmic Physiol Opt. 2004;24(2):142–147. doi: 10.1111/j.1475-1313.2004.00186.x. [DOI] [PubMed] [Google Scholar]
- 8.Win-Hall DM, Ostrin LA, Kasthurirangan S, Glasser A. Objective accommodation measurement with the Grand Seiko and Hartinger coincidence refractometer. Optom Vis Sci. 2007;84(9):879–887.doi: 10.1097/OPX.0b013e3181559ace. [DOI] [PubMed] [Google Scholar]
- 9.Anderson HA, Stuebing KK. Subjective vs objective accommodative amplitude: preschool to presbyopia. Optom Vis Sci. 2014;91(11):1290–1301.doi: 10.1097/OPX.0000000000000402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.McClelland JF, Saunders KJ. The repeatability and validity of dynamic retinoscopy in assessing the accommodative response. Ophthalmic Physiol Opt. 2003;23(3):243–250. doi: 10.1046/j.1475-1313.2003.00113.x . [DOI] [PubMed] [Google Scholar]
- 11.Tsukamoto M, Nakajima T, Nishino J, Hara Y, Uozato H, Saishin M. Accommodation causes with-the-rule astigmatism in emmetropes. Optom Vis Sci. 2000;77(3):150–155. doi: 10.1097/00006324-200003000-00014. [DOI] [PubMed] [Google Scholar]
- 12.Mutti DO, Enlow NL, Mitchell GL. Accommodation and induced with-the-rule astigmatism in emmetropes. Optom Vis Sci. 2001;78(1):6–7. doi: 10.1097/00006324-200101010-00004. [DOI] [PubMed] [Google Scholar]
- 13.Ukai K, Ichihashi Y. Changes in ocular astigmatism over the whole range of accommodation. Optom Vis Sci. 1991;68(10):813–818. doi: 10.1097/00006324-199110000-00011. [DOI] [PubMed] [Google Scholar]
- 14.Altoaimi BH, Kollbaum P, Meyer D, Bradley A. Experimental investigation of accommodation in eyes fit with multifocal contact lenses using a clinical auto-refractor. Ophthalmic Physiol Opt. 2018;38(2):152–163. doi: 10.1111/opo.12438. [DOI] [PubMed] [Google Scholar]
- 15.Ramasubramanian V, Glasser A. Prediction of accommodative optical response in prepresbyopic subjects using ultrasound biomicroscopy. J Cataract Refract Surg. 2015;41(5):964–980. doi: 10.1016/j.jcrs.2014.12.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Poltavski DV, Biberdorf D, Petros TV. Accommodative response and cortical activity during sustained attention. Vision Res. 2012;63:1–8. doi: 10.1016/j.visres.2012.04.017 . [DOI] [PubMed] [Google Scholar]
- 17.Roberts TL, Manny RE, Benoit JS, Anderson HA. Impact of cognitive demand during sustained near tasks in children and adults. Optom Vis Sci. 2018;95(3):223–233. doi: 10.1097/OPX.0000000000001186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Manh V, Chen AM, Tarczy-Hornoch K, Cotter SA. Rowan Candy T. Accommodative performance of children with unilateral amblyopia. Investig Ophthalmol Vis Sci. 2015;56(2):1193–1207. doi: 10.1167/iovs.14-14948. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Singman E, Matta N, Tian J, Silbert D. Association between accommodative amplitudes and amblyopia. Strabismus. 2013;21(2):137–139. doi: 10.3109/09273972.2013.786737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Birch E, Williams C, Drover J, et al. Randot® preschool stereoacuity test: normative data and validity. J AAPOS. 2008;12(1):23–26. doi: 10.1016/j.jaapos.2007.06.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Drover JR, Felius J, Cheng CS, Morale SE, Wyatt L, Birch EE. Normative pediatric visual acuity using single surrounded HOTV optotypes on the electronic visual acuity tester following the Amblyopia Treatment Study protocol. J AAPOS. 2008;12(2):145–149. doi: 10.1016/j.jaapos.2007.08.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Heath GG. The influence of visual acuity on accommodative responses of the eye. Optom Vis Sci. 1956;33(10):513–524. doi: 10.1097/00006324-195610000-00001. [DOI] [PubMed] [Google Scholar]
- 23.Charman WN, Tucker J. Dependence of accommodation response on the spatial frequency spectrum of the observed object. Vision Res. 1977;17(1):129–139. doi: 10.1016/0042-6989(77)90211-5. [DOI] [PubMed] [Google Scholar]
- 24.Thibos LN, Wheeler W, Horner D. Power vectors: An application of Fourier analysis to the description and statistical analysis of refractive error. Optom Vis Sci. 1997;74(6):367–375. doi: 10.1097/00006324-199706000-00019. [DOI] [PubMed] [Google Scholar]
- 25.Koo TK, Li MY. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med. 2016;15(2):155–163. doi: 10.1016/j.jcm.2016.02.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Weng CC, Hwang DK, Liu CJL. Repeatability of the amplitude of accommodation measured by a new generation autorefractor. PLoS One. 2020;15(1). doi: 10.1371/journal.pone.0224733. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ciuffreda KJ, Hokoda SC, Hung GK, Semmlow JL. Accommodative stimulus/response function in human amblyopia. Doc Ophthalmol. 1984;56(4):303–326. doi: 10.1007/BF00155676. [DOI] [PubMed] [Google Scholar]
- 28.Ciuffreda KJ, Hokoda SC, Hung GK, Semmlow JL, Selenow A. Static aspects of accommodation in human amblyopia. Am J Optom Physiol Opt. 1983;60(6):436–449. doi: 10.1097/00006324-198306000-00004. [DOI] [PubMed] [Google Scholar]
- 29.Wang B, Ciuffreda KJ. Depth-of-focus of the human eye: Theory and clinical implications. Surv Ophthalmol. 2006;51(1):75–85. doi: 10.1016/j.survophthal.2005.11.003 . [DOI] [PubMed] [Google Scholar]
- 30.Yao P, Lin H, Huang J, Chu R, Jiang BC. Objective depth-of-focus is different from subjective depth-of-focus and correlated with accommodative microfluctuations. Vision Res. 2010;50(13):1266–1273. doi: 10.1016/j.visres.2010.04.011. [DOI] [PubMed] [Google Scholar]
- 31.Kotulak JC, Schor CM. The accommodative response to subthreshold blur and to perceptual fading during the Troxler phenomenon. Perception. 1986;15(1):7–15. doi: 10.1068/p150007. [DOI] [PubMed] [Google Scholar]
- 32.Roberts TL, Stevenson SB, Benoit JS, Manny RE, Anderson HA. Blur detection, depth of field, and accommodation in emmetropic and hyperopic children. Optom Vis Sci. 2018;95(3):212–222. doi: 10.1097/OPX.0000000000001177. [DOI] [PMC free article] [PubMed] [Google Scholar]