Author Affiliations: Department of Ophthalmology, Children’s Hospital and Harvard Medical School, Boston, Mass.
To study the development of visual acuity in term-born children with cerebral visual impairment and a history of neonatal hypoxic-ischemic encephalopathy.
We studied 19 term-born children, aged 6 months to 6 years, with moderate to severe neonatal hypoxic-ischemic encephalopathy and behaviors indicative of cerebral visual impairment. Longitudinal measures of grating acuity were obtained using preferential looking (PL) and visual evoked potential (VEP. procedures. Visual acuities at first and last visits were compared. The courses of acuity development in the 9 children who underwent both VEP and PL acuity assessment at 4 or more ages were compared with normal development.
All children had measurable PL and VEP acuity, despite poor visual behavior. In nearly all, both PL and VEP acuity were below normal for age. For both PL and VEP measures, acuity at the last visit was, on average, 1 octave better than at the first visit, with a rate of improvement lower than normal. Although parallel courses of PL and VEP development occurred in many, substantial disparities in PL and VEP acuity were observed in others.
Modest increases in PL and VEP grating acuity occur during early childhood in many of these patients. The rate of increase is lower than normal.
Little is known about the course of development of acuity in cerebral visual impairment (CVI), the most common cause of bilateral visual impairment in infants and children.1 Typically, the ocular structures are healthy and the pupillary responses are brisk. In short, the ocular findings do not explain the child’s visual impairment. Anomalous behaviors in children with CVI include avoidance of faces, light gazing, and preference for moving rather than stationary objects and familiar rather than novel objects. Visual response to complex visual displays is reduced, delayed, or suppressed, especially in the presence of competing auditory or tactile stimuli.1- 6
Preferential looking (PL) and visual evoked potential (VEP) measures of visual acuity are feasible and deemed reliable and valid in infants and children with CVI.7- 9 Both measures are used in clinical practice. Visual evoked potential acuity depends on the integrity of the pathway from the eye to the visual cortex. Preferential looking acuity is also dependent on this pathway, but additional factors such as attention and ocular motor abilities affect the child’s response and thus the examiner’s judgment of the child’s response. It is well established that normal grating acuity, whether measured by PL or VEP, increases during early childhood.10
Reports of young children with CVI often present the acuities as cross-sectional data and include patients with diverse causes of CVI.9,11- 14 Serial measurements of visual acuity in individual children, all with the same cause of CVI, have seldom been reported.9,11,13- 15
The aim of this study is to determine whether a developmental increment in visual acuity occurs in children with neonatal hypoxic-ischemic encephalopathy and poor visual behavior consistent with CVI. Herein we report longitudinal measures of visual acuity in children with moderate to severe neonatal hypoxic-ischemic encephalopathy who were born at term and were referred because of CVI-like behaviors. Every patient underwent PL and VEP acuity evaluations as a function of age.
All patients were referred because of parental concern about visual impairment and behaviors consistent with CVI.2- 5 all of the patients had major disabilities in addition to the concerns about vision (Table 1). Clinical inclusion criteria for this report were a history of birth at term, antenatal or perinatal distress or both, moderate to severe neonatal encephalopathy with seizures, and magnetic resonance imaging evidence of brain injury in a pattern typical of hypoxic-ischemic encephalopathy. Each had assessment of his or her developmental quotient using the Bayley Scales of Infant Development.16 we excluded patients with metabolic disorders, hypoglycemia, trauma, abnormal ocular structures (except mild optic atrophy), or malformations of the brain.
The magnetic resonance imaging findings showed diffuse signal abnormality in the subcortical gray matter, including the thalamus and basal ganglia, and in the cerebral gray and white matter, including portions of the optic radiations and visual areas of the occipital cortex. Neurological features are listed in Table 1. Every patient had some persistent motor abnormality; only 3 became ambulatory during the period of observation. Six patients had seizures after the neonatal period. All but 2 patients had a developmental quotient below 50. A normal developmental quotient16 is 100 (SD, 15). Head circumference was below the second percentile in most of the patients. We considered unfavorable neurological features to be seizures after the neonatal period, a developmental quotient less than 50, and acquired microcephaly with head circumference below the second percentile. Each of these features was scored, with 1 indicating unfavorable and 0, favorable. The severity of neurological outcome was taken as the sum of these scores (table 1). Thus, those with the most severe overall neurological outcome had a score of 3, and those with less severe disease had a lower score.
Ophthalmic features (Table 2) included mild disc pallor in 7 patients. All but 2 patients had strabismus. Most patients were exotropic, typically with variable angle. All strabismus was alternating. Four patients had nystagmus. The median spherical equivalent at the time of the first visual acuity test was +1.50 diopters (D) (range, –1.63 to +4.63 D). None had anisometropia greater than 1 D. Throughout their course, all except patient 19 had spherical equivalents within the 95. prediction interval of normal for age.17 at 36 months of age, the spherical equivalents for patient 19 were +4.38 D OD and +4.63 D OS. Visual acuity testing in patient 19 was performed with correction in place.
We included only patients who met the inclusion criteria described in the previous section, had both PL and VEP measurements of acuity at the same session on at least 1 date, and had visual acuity (PL or VEP or both) measured at more than 1 session. Visits occurred during an 8-year period from March 1996 through February 2004. The median age at the first acuity test was 15 months (range, 6-36 months). For the 9 patients who had both PL and VEP acuity tests in the same session at 4 or more ages, the median duration of follow-up was 29 months (range, 24-76 months), during which time as many as 16 PL and 6 VEP acuities were measured. Grating acuities were obtained with binocular viewing.
A clinical variant of the Teller Acuity Card procedure was used to measure PL acuity.18 The acuity cards were presented horizontally or vertically, lateral to, or central to the gaze direction to ensure the child could view each grating position and the observer could judge the child’s detection of the grating. Such modifications minimize the effects of visual field defects and eye movement abnormalities. The stimuli were 12.5-cm squares of high-contrast (83%), black-and-white square-wave gratings (range, 0.23-26.0 cycles/cm, in approximately 0.5-octave steps) on rectangular cards. The gratings had the same space-averaged luminance as the background of the gray cards. The luminance of the cards was at least 10 candela (cd)/m2. The tester, unaware of the right-left position of the stripes, presented the cards at a distance of 38 to 55 cm from the patient. Based on the patient’s looking behavior, the tester judged the finest grating that the child detected. This was taken as the acuity, expressed in cycles per degree (cpd). The patient’s acuities were compared with normal binocular PL acuity for age.19 Although test-retest reliability is poorer in infants with an abnormal perinatal history20 or developmental delay,21,22 reliability was clinically acceptable for the present sample. For the sample reported herein, between- and within-tester variations were no greater than 1 card interval (0.5 octave) in most of the patients.
Sweep VEPs were recorded using the NUDiva system (Smith-Kettlewell, San Francisco, Calif).7,23,24 The spatial frequency of a high-contrast (80%) vertical square-wave grating (average luminance, 76 cd/m2) alternating at 5.5 Hz (11 reversals per second) was swept from low to high spatial frequency during a 10-second trial. To accommodate the large range of acuities, the test distances were 50 to 150 cm. This provides nominal test fields of approximately 42° × 32° to 14° × 11°. Thus, because the VEP represents the function of the central 5° to 8° of the field,10 all subjects received adequate macular stimulation. Electrodes were placed 3 cm above the inion (Oz) and 3 cm to the left (O1) and right (O2), with a reference electrode at the vertex and a ground electrode on the forehead. The electroencephalogram was amplified (Grass P10 preamplifier [Astro-Med, Inc, West Warwick, RI]; gain, 20 000; bandpass, 1-100 Hz) and monitored continuously during the session. The average of 5 or more sweeps was used. Acuity was estimated by linear extrapolation to estimate the spatial frequency that produced a zero microvolt VEP. For VEP acuity, test-retest agreement is somewhat lower in children with CVI than in healthy children.7,25 For the present sample, VEP thresholds were within 0.5 octave in 8 of 10 patients who had repeated measures within 6 months.
Visual acuity (log2 cpd, ie, an octave scale) was plotted as a function of age and compared with the normal course of development. An octave indicates a doubling of spatial frequency; in Snellen nomenclature, an improvement from 20/80 to 20/40, for example, is an improvement of an octave.
Normal PL and VEP acuity increases rapidly from birth to 6 months of age (Figure 1) and then more gradually thereafter.10 Courage and Adams19 reported that during normal development from 6 months to 3 years of age, average binocular PL acuity increased 1.66 octaves from 5.9 to 18.6 cpd, or, on average, 0.66 octave/y. Mayer and colleagues26 reported that in healthy children aged 6 through 48 months, monocular PL acuity increased 2.12 octaves from 5.7 to 24.8 cpd, or 0.61 octave/y. Birch27 found that from 6 to 18 months of age, average normal binocular VEP acuity increased from 12.82 to 18.27 cpd, or 0.51 octave/y. An acuity of 18.27 cpd is only 0.19 octave less than the mean normal VEP acuity in adults tested in our laboratory (20.91 cpd; n = 11).
Figure 1. All visual acuity measurements in our patients. A, Preferential looking (PL) values; B, Visual evoked potential (VEP) values. Serial measurements in an individual are connected by lines. The 95% prediction intervals for normal PL and VEP acuities are indicated by the dotted and dashed lines, respectively. The ordinates are a log base 2 scale; cpd indicates cycles per degree.
For all 19 patients, the first and last acuities were compared and expressed as octave change per year. For those 9 patients who had 4 or more sessions at which both PL and VEP acuities were measured, the rate of change in acuity per year was determined by linear regression.
All PL and VEP acuities of the 19 patients are plotted as a function of age in Figure 1. Nearly all visual acuities were below normal for age. Individual children underwent 2 to 16 measurements (median, 8 measurements) of PL acuity and 1 to 7 measurements (median, 4 measurements) of VEP acuity.
Inspection of the course of PL acuity development in each child (n = 19) shows that between the first and last measurement, acuity improved 1 octave on average (range, −2.1 to 5.0 octaves). One octave is often considered a clinically significant change. Three patients had a decrease in PL acuity. Two of the 3 (patients 2 and 13) had decreases of more than 1 octave in PL acuity; both had intractable seizures. In the third patient (patient 1), the decrease was only 0.5 octave. The median period of observation was 2.7 years (range, 6 months to 6 years), and for individuals the average rate of change in PL acuity per year was +0.18 octave/y (range, −0.90 to +1.80 octaves/y). The final PL acuity was related to the neurological score (Spearman ρ = −0.70; p<.001); the more favorable the score, the better the acuity. The rate of change in PL acuity was independent of the neurological score.
Improvement in VEP acuity (Figure 1) was also on average 1 octave (range, −0.22 to 3.09 octaves). During the period of observation (median, 2 years; range, 6 months to 6 years), the average rate of change in VEP acuity was +0.41 octave/y (range, −0.3. to +1.30 octaves/y). For individuals, the rate of change in VEP acuity did not differ significantly from that for PL acuity (paired t test, t13 = 0.49; p = .63). The rate of VEP acuity change and the final VEP acuity were not related to the neurological score. Of note, fewer VEP than PL test results were obtained, and the median period of observation for VEP tests was shorter than that for the PL tests.
The courses of visual acuity in the 9 children with measures of both PL and VEP acuity at 4 or more ages are plotted in Figure 2. As in normal development, VEP acuity was always better than PL acuity. For this subset of 9 children, as in the whole sample, most acuities were below normal, but some were not far below (patients 3, 5, and 9). Only patient 14 had normal PL and VEP acuity. Linear regression analysis indicates an average improvement in PL acuity of +0.26 octave/y. The individual slopes determined by linear regression (range, −0.05 to +0.55 octave/y) were within the range of estimated rate of change for the whole sample. The rate of increase in normal PL acuity is more than +0.60 octave/y.19,26 for VEP acuity, linear regression indicated an improvement of +0.25 octave/y (range, −0.10 to +1.52 octave/y). Although 2 patients had substantial declines in PL acuity, none of the VEP courses showed substantial decline.
figure 2. The courses of the patients who had 4 or more measurements of both preferential looking (PL) and visual evoked potential (VEP) acuities. The lower limits of the normal interval for PL and VEP acuities are indicated by the dotted and dashed lines, respectively. The patients are described in table 1. The ordinates are a log base 2 scale; cpd indicates cycles per degree.
All pairs of PL and VEP data obtained at the same session are shown in Figure 3. The range of PL acuity (6.0 octaves) was greater than that for VEP acuity (4.5 octaves). In almost every case, VEP acuity was better than PL acuity. The magnitude of the disparity was not related to the neurological score. For more than half of the points, the disparity between VEP and PL acuity exceeded an octave, with better VEP acuity. The discrepancy between PL and VEP acuity was greatest at lower acuity levels, as previously reported by Orel-Bixler and coworkers28 in patients with developmental disabilities and by Good7 in children with CVI due to several different causes. Also, as found in the study by Orel-Bixler et al,28 the age of our patients did not account for the larger discrepancy at lower acuities.
Figure 3. The relationship of preferential looking (PL) and visual evoked potential (VEP) acuities. Every patient is represented, and individual patients contribute 1 to 5 points. The diagonal lines have a slope of 1.0. Data would lie on the solid line if PL and VEP acuity values were in perfect agreement. The dashed lines are 1 octave above and below the solid line. The ordinate and the abscissa are a log base 2 scale. cpd indicates cycles per degree.
All of these children with persistent CVI-like behaviors had measurable visual acuity, although most remained below normal. Furthermore, the typical developmental course indicated a modest improvement in visual acuity, but at a slower rate than normal. Very few of the patients had a rate of improvement that was as fast as normal. On the other hand, only 2 with a seizure disorder had substantial declines in acuity, and that was in their PL but not their VEP acuity. Thus, the seizure disorder may have interfered with the child’s looking behavior on which the PL response depends.
In these young children with CVI, the overall improvement in PL and VEP acuities was about the same, ie, 1 octave. Typically, in individuals, the rates of PL and VEP improvement were similar, and in some of the patients with longitudinal data (figure 2), the courses of PL and VEP acuity were approximately parallel. In a few patients (figure 2), the course was suggestive of a delay in development of visual responses. Visual acuity tests at earlier ages would be informative about this matter.
In many of these children, the PL and VEP results yielded a similar picture of visual development, with VEP acuity better than PL acuity by approximately 1 octave, as is the case in normal development (Figure 1 and figure 2). However, large discrepancies between PL and VEP acuity occurred in more than half of the test sessions (Figure 3). The reasons for these discrepancies must lie in differences between PL and VEP procedures and in abnormal processes in the brain with a history of diffuse injury. Unlike the VEP response, the PL response requires other systems, such as the ocular motor system, and attention.
The luminous stimulus field and the apparent motion of the swept gratings distinguish the VEP from the static black-and-white stripes on the PL cards and could conceivably be more salient to the child with CVI. Differences in PL and VEP acuity are incompletely explained by differences in stimulus variables.29 imprecise retinotopic mapping might be postulated as the basis for the larger VEP stimulus field producing much better VEP than PL acuity in some of our patients (Figure 3). Other potential explanations for substantially better VEP than PL acuity in some patients are the larger number of trials inherent in the sweep VEP than in the PL procedure and different scoring of PL and VEP responses.28,29
In instances where PL and VEP measures give parallel information about visual development, both tests are not needed. Our experience suggests that the VEP acuity test gives little additional information in children who have PL acuities within 1 octave of normal for age, who have a PL acuity of 4 cpd or better, or who during the PL test have consistent, repeatable responses. For such children, we find ourselves comfortably relying on the fast and relatively inexpensive PL test alone. Children who are difficult to engage, make only fleeting eye contact, if that, and have a PL acuity of 4 cpd or less are likely to have much better VEP acuity. For such visually inattentive children, we find that the VEP provides a reliable measure of acuity, but acuity may be astonishingly good for the child’s level of visual behavior. Thus, although clinicians find acuity a familiar measure, more specific tests of visual cortical function and attention13,14 warrant consideration for valid assessment of vision in children such as those included in this study.
Correspondence: Anne B. Fulton, MD, Department of Ophthalmology, Children’s Hospital and Harvard Medical School, 30. Longwood Ave, Boston, MA 02115 (Anne.Fulton@childrens.harvard.edu).
Submitted for Publication: March 30, 2004; final revision received December 15, 2004; accepted January 13, 2005.
Financial Disclosure: None.
Funding/Support: This study was supported in part by grants from the Blind Children Center, Los Angeles, Calif, and United Cerebral Palsy Foundation, Washington, DC.
Thank you for submitting a comment on this article. It will be reviewed by JAMA Ophthalmology editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 16
Customize your page view by dragging & repositioning the boxes below.
The Rational Clinical Examination
The Rational Clinical Examination
All results at
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.