The Course of Maturation of Rod-Mediated Visual Thresholds in Infants

Abstract

purpose. To measure the developmental course of infants’ rod-mediated thresholds.

methods. Thresholds for detecting stimuli (2° diameter, 50 msec duration) presented at 10° (parafoveal site) or 30° (peripheral site) from a central fixation target were estimated using a preferential-looking method. Nine infants were tested at both stimulus positions at ages 10, 18, and 26 weeks.

results. At 10 weeks, infants’ thresholds at both sites were significantly higher than those of adults. The infants’ average threshold at 10° was 0.5 log unit higher than the infants’ average threshold at 30°. Adults’ thresholds at the two sites were equal. Thresholds of all infants decreased with age until by age 26 weeks the parafoveal and peripheral thresholds were equal and were the same as those of adults. The rate of change of parafoveal thresholds was significantly faster than the rate of change of peripheral thresholds.

conclusions. Although postreceptoral factors cannot be ruled out, the results suggest that developmental increases in rod outer segment length and rhodopsin density account for most of the threshold changes during infancy.

The dark adapted, rod-mediated visual thresholds of young infants are significantly higher than those of adults. ^{1 2 3 4 5 6 7 8 9} For example, cross-sectional data show that the average threshold is 1.4 log unit higher than that of adults at 4 weeks, 1.1 log unit higher at 10 weeks, and 0.65 log unit higher at 18 weeks. ⁷ Thus, although thresholds decrease, at age 18 weeks they remain significantly above those of adults. ⁶ The age at which infants’ thresholds become equal to those of adults is unknown. Furthermore, we are unaware of any report that describes the course of maturation of normal scotopic thresholds in individual infants. 

We undertook a longitudinal study of rod-mediated threshold development and elected to test parafoveal (10° eccentric) and peripheral (30° eccentric) retinal sites, because development is nonuniform across the retina. For instance, anatomic studies show parafoveal rod photoreceptor outer segment growth is delayed relative to peripheral outer segment growth, ^{10 11 12 13} despite the axiom that the central retina matures earlier than the peripheral retina. Psychophysical study shows that 10-week-old infants’ thresholds at a parafoveal site were significantly elevated relative to their thresholds at a peripheral site, whereas adults’ thresholds at these sites were equal. ¹⁴ In this longitudinal study, we tested the hypothesis that thresholds measured at the parafoveal site change more rapidly than those at the peripheral site. 

Methods

Stimuli

Stimuli were 50 msec, 2° diameter, blue spots (Wratten 47B,λ < 510 nm) presented on a rear projection screen, 10° or 30° to the right or the left, of a 30-min arc diameter red LED fixation target flickering at 1 Hz. Stimulus intensity was controlled by calibrated neutral density filters. Calculation of the retinal illuminance produced by the stimuli was based on luminance measurements made with a calibrated photodiode (UDT S-350; United Detector Technology, Orlando, FL) placed in the position of the subject’s eyes. At the beginning and end of each session, the subject’s pupillary diameter was estimated by direct observation with an infrared viewer. Pupillary diameter was determined by comparison with the diameter of the cornea which is 11 ± 0.5 mm in infants from term to 6 months of age. ¹⁵ Retinal illuminance varies directly with pupillary diameter and the transmissivity of the ocular media and inversely with the square of the posterior nodal distance. ⁷ The scotopic troland value of the stimulus ¹⁶ was calculated taking each subject’s measured pupillary diameter and the average axial length ¹⁷ into account. ^{6 7} The correction for light losses in the ocular media was based on previous results in infants. ¹⁸  

Procedure

Thresholds were estimated using a two-alternative, forced-choice, preferential-looking method ¹⁹ that incorporated a fix-and-flash procedure. ⁹ After the subject dark adapted for 30 minutes, an adult held the infant 50 cm in front of the center of the screen. The flickering red LED fixation target attracted the infant’s gaze to the center of the screen. A second adult watched the infant with an infrared viewer and reported when the infant was alert and looking at the fixation target. The fixation target was extinguished, and a test flash presented. The observer reported stimulus location, right or left, based on the infant’s head and eye movements and received feedback on every trial. Threshold was measured with a transformed up–down staircase that estimated the 70.7% correct point of the psychometric function. ²⁰ Control experiments with adults indicated that the observer could reliably detect a horizontal deviation of 3° or more from the fixation target. ¹⁴ Thus, a reliable response to the 10° eccentric stimulus was expected. 

Each infant was tested first at age 10 weeks, the age at which previous research has shown that thresholds are significantly higher for the stimuli at 10° than those at 30°, and thresholds at both eccentricities are significantly higher than those of adults. ¹⁴ The second visit was scheduled at age 18 weeks because an earlier study showed that thresholds for 50 msec, 10° diameter stimuli presented 20° from fixation remained significantly higher than those of adults. ⁶ As the 18-week-old infants’ results were collected, it was clear that thresholds for the 2° stimuli remained above those of adults at both sites. 

Additional sessions at 8-week intervals were planned to continue until thresholds at both sites reached the adult troland value. Consequently, the infants were next scheduled to return at age 26 weeks. At every session, thresholds were obtained from each infant at both eccentricities (10° and 30°). At each age, five of the subjects were tested first at 10°, and four were tested first at 30°. 

Normal adult control subjects were tested using the same 2° diameter stimuli at 10° and 30° eccentricities in conjunction with the staircase procedure. Adults named the position (left or right) of the stimuli. 

Subjects

Fifteen infants aged 69 to 81 days (median, 71 days) at the first session participated. Nine completed testing at all three ages (10, 18, and 26 weeks); their data are the basis for this report. Six did not complete testing, and their data are not included. Two infants did not complete the first session, three completed only the first session, and one completed the first and second sessions, but not the third. The thresholds obtained from these infants are within the range of those in infants who completed the longitudinal study. 

All infants were born within 10 days of term, were in good general health, and had normal eyes documented on thorough ophthalmic examination. Fifteen adults (aged 19–35 years) with normal eyes were tested; their results have been reported previously. ^{14 21} The study conformed to the tenets of the Declaration of Helsinki and was approved by the Children’s Hospital Committee on Clinical Investigation. Written, informed consent was obtained from the infants’ parents and from adult subjects before each session. 

Results

The thresholds of individual infants are shown in Figure 1 . At age 10 weeks, all thresholds at both eccentricities were higher than any of the adults’ thresholds. The difference between the median thresholds of infants and adults at 10° was 1.06 log unit (Mann–Whitney = 0; P < 0.01) and at 30° was 0.58 log unit (Mann–Whitney = 0; P < 0.01). Previous studies of infants’ and adults’ spectral sensitivities indicate that thresholds with these troland values are rod mediated. ⁷ These results agree well with the parafoveal and peripheral thresholds in eleven 10-week-old infants previously reported. ¹⁴  

Thresholds decreased with age at both 10° and 30° sites in every subject (Fig. 1 , upper and middle panels). At age 18 weeks, thresholds remained immature; the median thresholds at 10° and 30° were 0.62 and 0.29 log unit higher than the median threshold of adults. By age 26 weeks, the threshold of every infant at both 10° and 30° was within the adult range. Analysis of variance with repeated measures on two factors (age and target location) showed significant effects for age (F = 296; df 2,8; P < 0.01) and target location (F = 131; df 1,8; P < 0.01) and a significant interaction of age and target location (F = 53; df 2,8; P < 0.01). 

The rate of change in threshold was more rapid at 10° than at 30° (Fig. 1) . Between ages 10 and 26 weeks, thresholds decreased, an average of 1 log unit at the 10° eccentricity. Assuming a linear decrease in log threshold, at 10° the median rate of change, determined by linear regression of each infant’s data, was 0.067 log unit per week (range, 0.053–0.080 log unit). At 30° the median rate was 0.033 log unit per week (range, 0.023–0.044 log unit). The rate of change at 10° was significantly faster than at 30° (Wilcoxon matched-pairs signed-rank test; T = 0; P < 0.01). 

The difference between thresholds at 10° and 30° (Δ_10–30) decreased after age 10 weeks (Fig. 1 , bottom). By age 26 weeks all infants had Δ_10–30 values within the range found in adults. 

Discussion

The rates of change in parafoveal and peripheral thresholds were not equal. Parafoveal thresholds, which at age 10 weeks are less mature than peripheral thresholds, changed at twice the rate of peripheral thresholds (Fig. 1) . Dark-adapted thresholds for small, brief stimuli obtained in this within-subject study, were unequal at parafoveal and peripheral sites in 10-week-old infants. This difference decreased systematically in every infant and vanished by age 26 weeks. Moreover, by 26 weeks, thresholds at both sites equalled those of adults. 

The developmental course for the parafoveal (10° eccentric) thresholds (Fig. 1 , top) was similar to that for the cross-sectional data. ^{1 5 6 8 9} The cross-sectional studies were performed with stimuli more than 2° in diameter presented at approximately 20° eccentricity. At age 8 to 10 weeks, the dark adapted thresholds are 1.1 to 1.5 log unit higher than those of adults. ^{1 4 6 8} All these previous studies were performed with larger diameter stimuli than the 2° used in the present study. In the present study the median parafoveal threshold at 10 weeks was 1.06 log unit higher than the median adult threshold. At 18 weeks, Hansen et al. ⁶ found that dark-adapted thresholds are elevated by 0.65 log unit, which agrees well with the median threshold elevation (0.62 log unit) found in the present study. 

The troland value of the 2°, 50 msec stimulus takes into account pupillary diameter, posterior nodal distance, and light losses in the media of the eye. ^{6 18} With such large stimuli, in dark-adapted conditions, scotopic visual efficiency is high and is limited by photoreceptors and postreceptoral pooling rather than by optical or preretinal factors. ²²  

Changes in rod outer segment (ROS) length and rhodopsin density can account for the improvement in rod-mediated thresholds at parafoveal and peripheral sites. Anatomic development of the retina continues during the ages over which thresholds have been measured. ROSs increase in length. ¹³ Thus, the increased probability of photon capture must be considered among the possible explanations for the observed decreases in thresholds. Are thresholds proportional to ROS length? 

Measurement of peripheral rod outer segments illustrated in Figures 41-13 and 41-14 of Hendrickson ¹³ indicate that peripheral ROS length in adults are 2.3 times longer than those of 5-day-old infants, predicting a threshold difference of approximately 0.4 log unit. In the present study, the median 10-week-old infant threshold at the peripheral site was 0.58 log unit (range, 0.4–0.85) higher than the median adult threshold. Also of note, the peripheral retinal sensitivity of 10-week-old infants, estimated with the full-field scotopic electroretinogram, is also approximately 0.5 log unit less than that of adults. ²³ Elevations of 0.5 and 0.58 log unit are a bit higher than expected from ROS lengths alone. However, in developing ROSs, at least in infant rats, ²⁴ ROS length overestimates the amount of rhodopsin available for photon capture. 

At the parafoveal site, ROS length in an adult is nine times that of a 5-day-old infant, ¹³ which predicts a threshold difference of nearly a log unit. At age 10 weeks, the difference between the median infant and adult thresholds at the parafoveal site is 1.06 log unit (range, 0.85–1.3). At age 11 months, the next age studied anatomically, ¹³ the parafoveal and peripheral ROS lengths of infants are 69% and 67% of adult ROS. These ROS lengths predict thresholds differences of only 0.16 and 0.17 log unit, which may not be detectable across subjects. The present psychophysical data show that no infant has a threshold that is more than 0.1 log unit different from the median adult threshold by age 6 months. Thus, at both peripheral and parafoveal sites, the increased probability of photon capture associated with increased ROS length helps explain threshold changes during infancy. 

ROS lengths in the parafovea and periphery ¹³ were also compared. At age 5 days after term, parafoveal ROS lengths are only approximately a quarter of those in more peripheral retina. This predicts that the parafoveal threshold is four times, or 0.6 log unit, higher than the peripheral threshold. In 10-week-old infants, the median parafoveal threshold was 0.5 log unit higher than at the peripheral site (Fig. 1) . At age 11 months after term, both parafoveal and peripheral ROS have elongated and are equal in length, ¹³ predicting equal thresholds at these sites by age 11 months. In fact, based on the thresholds (Fig. 1) , we suspect parafoveal and peripheral ROS lengths are equal by age 6 months. 

In summary, during development, the rate of change of rod thresholds is significantly higher in parafoveal than peripheral retina. The differences between infant and adult scotopic thresholds are reasonably well accounted for by photoreceptor outer segment immaturities. Although immature postreceptoral processes such as the functional organization of receptive fields, which are thought to underlie spatial summation, ^{5 9 25 26} and the gain of transmission from photoreceptors to second-order neurons ²⁷ are acknowledged, the hypothesis that photoreceptor immaturities explain the scotopic thresholds of infants cannot be rejected. Studies of background adaptation can test this hypothesis by comparing shifts of the eigengrau and dark-adapted thresholds of increment thresholds measured in parafoveal and peripheral retina. ^{1 7 28 29}  

 Reprint requests: Ron Hansen, Department of Ophthalmology, Children’s Hospital and Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115.

 Supported by Grant EY10597 from the National Institutes of Health and the William Randolph Hearst Foundation.

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, 1996.

 Submitted for publication October 14, 1998; revised January 5, 1999; accepted January 25, 1999.

 Proprietary interest category: N.

 

Figure 1.

 

Thresholds for the 10° parafoveal (top) and 30° peripheral (middle) retinal positions are shown as a function of age. The bottom panel shows the difference (Δ_10–30) between the parafoveal and peripheral thresholds as a function of age. Thresholds from individual infants are connected by lines. The median adult threshold is represented by the triangle in each panel. The range of adult thresholds is indicated by the dashed lines.

The authors thank Terri Halperin for her assistance during the experiment.