The Development of Scotopic Sensitivity

Abstract

purpose. Test the hypothesis that the developmental increases in rod photoreceptor sensitivity and rod-mediated visual sensitivity at 10°, 20°, and 30° eccentric are concurrent. It is known that maturation of the parafoveal (10° eccentric) rod outer segments and visual sensitivity is delayed compared to that at 30° eccentric.

methods. Rod isolated electroretinographic (ERG) responses to full-field stimuli were obtained from dark-adapted subjects (n = 71), ranging in age from early infancy through middle age. Rod photoreceptor sensitivity was calculated by fitting a model of the activation of phototransduction to the a-wave response. Rod driven b-wave sensitivity was calculated from stimulus–response functions. A logistic growth model was used to summarize the developmental increases in sensitivity of the rod photoreceptors and the b-wave. Previously reported dark-adapted, rod-mediated visual sensitivities at 10°, 20°, and 30° eccentric, obtained using preferential looking procedures, were reanalyzed using the logistic growth model.

results. The logistic growth model accounted for 57% to 85% of the variance of each sensitivity parameter with age in normal subjects. The shape of the growth curve and the age at which sensitivity reaches 50% of the adult value is similar (10.0–13.5 weeks) for the rods, the b-wave, and peripheral visual sensitivity, but is significantly older, 19.5 weeks, for rod-mediated parafoveal visual sensitivity.

conclusions. Rod photoreceptor sensitivity and peripheral, rod-mediated visual sensitivity develop concurrently. A parsimonious explanation is that rod photoreceptor sensitivity determines dark-adapted, rod-mediated visual sensitivity during development.

The dark-adapted, rod-mediated visual sensitivity of young infants, according to most published data, is more than a log unit below that of adults. For instance, at ages 8 to 12 weeks, dark-adapted visual sensitivities, estimated using two alternative, forced-choice preferential looking procedures, are 1.1 to 1.7 log units lower than those of adults. ^{1 2 3 4 5} Over the same limited age range in early infancy, sensitivities derived from the electroretinographic (ERG) a- and b-waves are barely half a log unit less than that of adults. ^{4 6 7} The course of maturation of rod-mediated visual sensitivity, however, depends on the retinal region tested. The development of parafoveal (10° eccentric) sensitivity ^{8 9} and rod outer segments ¹⁰ is delayed compared to that in more peripheral retina. Furthermore, there is only a half log unit discrepancy between peripheral visual sensitivity for detecting 30° eccentric 2° diameter stimuli in 10-week-old and adult individuals. ^{8 9} In other words, for some stimulus conditions, the immaturities in ERG and visual sensitivity are the same. The rods are thought to account for immaturities of the ERG a- and b-wave. ^{7 11} Thus, despite some evidence to the contrary, ^{1 12 13} the immature rods, which have short outer segments ¹⁴ and low rhodopsin content ¹⁵ with consequent low probability of photon capture, cannot be dismissed as the primary determinant of infants’ low, dark-adapted, rod-mediated visual sensitivity. 

In developing mammalian retina, rod outer segments grow longer, ^{10 11 16 17 18 19 20 21 22} following a course of logistic growth, ²³ and finally asymptote at adult length with equal synthesis and disposal of outer segment discs. ¹⁸ In rats, ^{11 16} logistic growth ^{11 19 23} also describes the developmental increase in rhodopsin content (Fig. 1) during outer segment development. The developmental increase in rhodopsin of several species, ^{19 20 21 22} including human, ¹⁵ has also been summarized using a logistic growth model. By using this mathematical model of growth, the courses of development of the visual pigment and retinal responses could be compared. ¹¹ The courses of the developmental increase in rat rod photoreceptor and inner retinal sensitivity, derived from the electroretinographic (ERG) a- and b-waves, are indistinguishable from that of rhodopsin. ¹¹ That is, according to a logistic growth model, Age₅₀, the age at which a parameter is half the adult value, for rat rhodopsin, rod photoreceptor sensitivity, and b-wave sensitivity do not differ significantly. ¹¹  

In the present study, we tested the hypothesis that dark-adapted, rod photoreceptor sensitivity predicts dark-adapted, rod-mediated visual sensitivity during infancy. We have used contemporary ERG procedures to assess rod photoreceptor sensitivity, represented by the ERG a-wave, ^{24 25} and b-wave sensitivity, ²⁶ as well as the developmental increases in the saturated amplitudes of the rod photoreceptor response and rod driven b-wave. Additionally, we have reanalyzed the developmental increases in dark-adapted, rod-mediated visual sensitivity. ^{3 6 8 9 27 28 29 30} To facilitate comparison of the course of development of the several response parameters, a logistic growth model is used to summarize the variation of each response parameter with age. 

Methods

ERG Procedures

The left pupil was dilated with cyclopentolate, 1%, and the subject was dark-adapted for 30 minutes Parents stayed with infants and children throughout the procedure. After dark adaptation, in dim red light, proparacaine, 0.5%, was instilled, and a bipolar Burian–Allen electrode was placed on the left cornea. The ground electrode was placed over the left mastoid. 

Blue (Wratten 47B; λ < 510 nm) strobe stimuli (Novatron, Dallas, TX) were delivered through a 41-cm integrating sphere, controlled in intensity by calibrated neutral-density filters, and ranged from those that evoked a small b-wave (<15 μV) to those that saturated the a-wave amplitude and slope. Also, stimulus–response functions to red (Wratten 29; λ > 610 nm) flashes were obtained. To isolate the rod function, the responses to photopically matched red flashes were subtracted digitally from the responses to blue flashes. ³¹ The unattenuated flash, measured with a detector (S350; United Detector Technology, Orlando, FL) placed at the position of the subject’s cornea, was 3.82 logμ W/cm² per flash. Retinal illuminance varies directly with pupillary diameter and the transmissivity of the ocular media and inversely with the square of the posterior nodal distance. ⁶ The scotopic troland value ³² of the stimulus was calculated, taking each subject’s pupillary diameter and the average axial length ³³ into account. ^{3 6} Thus, the maximum intensity blue light produced about + 3.6 log scot td sec retinal illuminance in both infants and adults. If 1 scotopic troland produces ∼8.5 isomerizations/rod/flash, ³⁴ the maximum intensity blue flash produced ∼34,000 isomerizations per rod per flash. 

All responses were differentially amplified (bandpass 1–1000 Hz; gain: 1000), displayed on an oscilloscope, digitized, and stored on disc for analysis later. An adjustable voltage window was used to reject records contaminated by artifacts. Two to 16 responses were averaged in each stimulus condition. The interstimulus interval ranged from 2 to 60 seconds and was selected so that subsequent b-wave amplitudes were not attenuated. 

The rod photoresponse characteristics were calculated from the a-wave responses using the Hood and Birch ²⁴ formulation of the Lamb and Pugh ^{25 35} model of the biochemical processes involved in the activation of phototransduction. A curve-fitting routine (MATLAB, fmins) was used to determine the best fitting values of S, a sensitivity parameter, and R _mp3, the saturated response amplitude, in  

\[R(i,t){=}R_{\mathrm{mp3}}(1-\mathrm{exp}{[}-0.5S\ I(t-t_{\mathrm{d}})^{2}{]}),\]

where I is the flash in isomerizations/rod/flash, and t _d is a brief delay. Fitting of the model was restricted to the leading edge of the a-wave response or to a maximum of 20 msec after stimulus onset. All three parameters were free to vary. 

For the rod driven b-wave, which represents mainly the activity of the on-bipolar cells, ^{26 36} the stimulus–response function  

\[V/V_{\mathrm{max}}{=}I/(I{+}{\varsigma})\]

was fit to the b-wave amplitudes of each subject using an iterative procedure that minimized the mean square deviation of the data from the equation. In this Equation 2 , V is the b-wave amplitude, V _max the saturated amplitude, I the stimulus in scot td sec, and ς the stimulus that evoked a half-maximum b-wave amplitude. Thus, 1/ς is a measure of sensitivity. The stimulus–response function was fit up to those higher flash intensities at which a-wave intrusion occurs. ³⁷  

Subjects

The main data are from 71 normal subjects recruited for study of rod photoreceptor function. Healthy, term-born infants, ages 23 through 100 days (n = 47) were recruited by mail. a-Wave data from seven were included in a previous report. ⁷ All had been born within 7 days of their due date and were in good general health. Normal child and adult control subjects (ages 8–52 years, n = 25) were also recruited; ERG data of 20 of these have been included as control data in a clinical report. ³⁸ No subject had a family history of eye or vision problems. Thorough ophthalmic examination disclosed no abnormalities. Written, informed consent was obtained from adult subjects and the parents of the infants and children. This study conformed to the tenets of the Declaration of Helsinki and was approved by the Children’s Hospital Committee on Clinical Investigation. 

The subjects described included few children. To obtain ERG data from children, we reviewed our clinical ERG database listing 1164 patients who had been referred for testing. We selected ERG data, obtained using the protocol described above, from the left eye of the 22 patients who were found by us to have perfectly normal eyes, including normal acuity for age, normal alignment and motility, normal pupillary responses, normal fundi, and little or no refractive error (spherical equivalents− 2 to +4 diopters). Their ages were 6 months to 15.1 years. Vision and ocular features have remained normal 1 to 6 years (median, 3 years) after electroretinography. None of the 22 had a family history of eye disease or visual disorders. 

Growth Curve Analysis

A logistic growth curve of the form  

\[y/y_{\mathrm{max}}{=}\mathrm{Age}^{n}/(\mathrm{Age}^{n}{+}\mathrm{Age}_{50}^{n}),\]

where Age₅₀ is defined as the age at which y is 50% of the adults’ value, y _max; and the exponent, n, indicating the steepness of the curve at Age₅₀, was fit to each ERG parameter, S, R _mp3 , V _max, and 1/ς (expressed as percent of the adult mean). A least squares criterion was used to determine the values that minimized the deviation of the data from Equation 3 . All parameters were free to vary. The b-wave response parameters, ς and V _max, from an additional 71 normal subjects that we had previously studied ³⁹ were included in the growth curve analysis. Also analyzed were S and R _mp3 values reported by Nusinowitz et al., ⁴⁰ in their Figure 2 , for infants and adults. In this analysis, y _max was fixed at the adult values of S and R _mp3 that were reported as means. ⁴⁰  

For comparison to the development of the ERG parameters, the logistic growth model was applied to the development of dark-adapted, rod-mediated visual sensitivity (the reciprocal of threshold). ^{3 6 8 9 27 28 29 30} These psychophysical data from 169 dark-adapted subjects had been obtained with a “fix and flash” preferential looking procedure. ^{9 41} All stimuli were 50 msec in duration. Some subjects (n = 43) were tested with 2° diameter spots presented 10° or 30° eccentric; the youngest subject tested with these small spots was 10 weeks old. ^{8 9 30} The majority (n = 126) were tested with 10° diameter spots presented 20° eccentric; the youngest subject tested with these large spots was 4 weeks old. ^{3 6 27 28 29} In each stimulus condition, adults’ mean threshold was designated 100%. For the growth curve analysis, sensitivity (the reciprocal of threshold) was expressed as percent of adults’ mean sensitivity for the particular stimulus. For instance, a threshold that is 0.3 log unit above the mean adult threshold is 50% of the adults’ mean sensitivity. Although nine infants’ data had been gathered longitudinally, ⁹ each sensitivity value was treated as an independent measure in the growth curve analysis. The variation in sensitivity with age was analyzed by fitting the logistic growth model to the data. For summaries of the development of rod-mediated visual sensitivity, data are also presented as log threshold, the familiar representation of visual sensitivity, and in these graphic displays, both the fitted growth curve (% sensitivity versus age) and transformed growth curve (threshold versus age) are shown. 

Results

Sample ERG records and model fits show all four ERG parameters, S, R _mp3, ς, and V _max, are lower in the infant than in the adult (Fig. 2) . Each of the four parameters is significantly smaller (Table 1A ) in 10-week-old infants (63–77 days; n = 27) than in the children and adults (n = 25). Consistent with a previous report, ⁴² there is little change in the ERG parameters between age 8 and 52 years. The ERG parameters are plotted as a function of age in Figures 3 and 4 . The parameters of the logistic growth model fit to the developmental increase in each parameter are listed in Table 1B . 

If logistic growth is assumed to summarize the normal developmental increase in human rod photoreceptor sensitivity, the age at which S is half the adults’ mean value is 10.7 weeks (95% confidence interval [CI], 8.9–12.5 weeks ) (Fig. 3A ; Table 1B ). Equation 3 accounts for 76% of the variation of S with age in normal subjects (n = 71). The age at which R _mp3 (Fig. 3B) is half the adults’ value is 11.2 weeks (95% CI, 9.4–13.0 weeks). The shape of the growth curve, as indicated by the exponent, n (Table 1B) , is similar for S and R _mp3. The values of S and R _mp3 at age 10 weeks, estimated by the growth curve analysis (Table 1B) , agree well with the mean values at age 10 weeks (Table 1A) , and values of y _max agree well with the mean observed values in adults (Table 1A) . Inclusion of the data from the 22 patients with normal eyes in the analysis has little effect on the growth curves for S and R _mp3 (Figs. 3A 3B ; Table 1B ). The Age₅₀ values for S and R _mp3 (Table 1B) are in reasonable agreement with those calculated for data reported by Nusinowitz et al. ⁴⁰ for which Age₅₀ of S is 9.4 weeks (95% CI: 7.8–11 weeks) (r ² = 0.39) and of R _mp3 is 12.5 weeks (95% CI: 9.4–15.6 weeks) (r ² = 0.18). Thus, the Age₅₀ values for the S and R _mp3 data reported by Nusinowitz et al. ⁴⁰ overlap one another and those for the subjects studied herein (Table 1B) . 

Assuming logistic growth describes the developmental increase in b-wave sensitivity, Age₅₀ is 11.0 weeks (95% CI, 9.3–12.7 weeks) (Fig. 4A ; Table 1B ). The logistic growth curve fit to saturated b-wave amplitude, V _max (Fig. 4B ; Table 1B ), reaches half of the adults’ mean value at 10.0 weeks (95% CI, 8.3–11.7 weeks). The values of log ς and V _max estimated from the growth curve analysis at age 10 weeks are in reasonable agreement with the observed values; y _max and adult means are also in reasonable agreement (Tables 1A 1B) . 

The mean dark-adapted visual threshold of 10-week-old infants is significantly higher (sensitivity lower) than that of adults at all eccentricities (Table 2A ); the highest infant threshold is at 10° eccentric (parafoveal). In Figure 5 , for each eccentricity, the sensitivities (left) and corresponding thresholds (right) ^{8 9 30} are plotted as a function of age. Assuming logistic growth of peripheral visual sensitivity (Fig. 5A , left), the calculated value of Age₅₀ is 13.5 weeks (95% CI, 12.2–14.8 weeks) (Table 2B) . This interval overlaps those for rod photoreceptor sensitivity, S, and b-wave sensitivity (Table 1B) . By this analysis, the course of development of peripheral, dark-adapted, rod-mediated visual sensitivity, as defined here, is concurrent with the development of rod photoreceptor (Fig. 3A) and b-wave (Fig. 4A) sensitivity. 

According to the logistic growth model, the parafoveal visual threshold (Fig. 5B , right) is 0.3 log unit above (or sensitivity is half) that of an adult’s threshold at 19.5 weeks (95% CI, 17.2–21.8 weeks). This is significantly older than Age₅₀ for peripheral thresholds (Fig. 5A ; Table 2B ), and rod photoreceptor (Fig. 3A ; Table 1B ) or b-wave (Fig. 4A , Table 1B ) sensitivity. In Figure 5C (right), the dark-adapted, rod-mediated thresholds for detection of large, 10° diameter spots, presented 20° eccentric, are shown. According to the logistic growth curve, the age at which the threshold is 0.3 log unit above (or sensitivity is half) that of an adult’s, is 33.0 weeks (95% CI, 26.2–39.8 weeks). 

Discussion

The developmental courses of rod photoreceptor sensitivity, b-wave sensitivity, and peripheral rod-mediated visual sensitivity are statistically indistinguishable. Even though the ERG and psychophysical data were not gathered following a within-subject design, the 95% CIs overlap for the ages at which rod photoreceptor, b-wave, and peripheral visual sensitivity reach half the adult’s value (Fig. 6) . Additionally, the shape of the summarizing growth curves, defined by the exponent n, is similar (Tables 1B 2B) . The developmental course for visual sensitivity mediated by parafoveal retina is relatively delayed, as is rod outer segment maturation in the parafoveal region. ^{10 14} Sensitivity measured with large (10° diameter) spots centered at the rod ring ¹⁴ is also relatively delayed. In infants, the large spots must test retina-containing rod outer segments in various states of maturation. ^{10 14}  

In developing rat retina, it has been possible to demonstrate that the developmental increase in rhodopsin content, rod photoreceptor, and b-wave sensitivity follow the same course. There is not broad overlap of the course of development of human rhodopsin and rod photoreceptor sensitivity (Fig. 6) . As the large confidence interval for human rhodopsin indicates, the content of rhodopsin in human eyes is quite variable. ¹⁵ The variability is due not only to bleaching, but also likely depends on numbers of rods per retina and perhaps on long-term adaptations to ambient light. ^{15 22}  

The logistic growth curve has been fit empirically to the data herein. In other species, this model provides a satisfactory summary of developmental increases in rhodopsin (as shown, e.g., in Fig. 1 ) and other proteins involved in phototransduction processes as well as growth of outer segment length. ^{11 19 22} Furthermore, in developing rats, the amount of rhodopsin available for capture of light and the numbers of channels in the outer segment membranes available for closure by light are proportional to the rod photoreceptor response parameters, S and R _mp3. ^{11 43} In other words, during development, rod photoreceptor sensitivity is scaled by the amount of rhodopsin present, and the amount of rhodopsin present is proportional to the number of channels in the receptor membrane that are available for closure by light. As the present data for human development show, the rat b-wave response parameters, perhaps representing mainly the on-bipolar cells, ^{26 36} which take their input from the rods, have the same developmental course as the rod photoreceptor response parameters. ¹¹  

Immaturities of rod-mediated, human retinal function may depend on similar rod outer segment immaturities, although the anatomic data are sparse. ^{8 9 10 14} Infants’ higher visual thresholds at parafoveal than peripheral sites were reasoned to be consistent with shorter parafoveal rod outer segments. ⁸ The courses of maturation of the visual thresholds at parafoveal and peripheral sites appear to be a consequence of a later course of maturation of the parafoveal than peripheral rod outer segments. ⁹  

Other results are also consistent with rods as the main determinant of 10-week-old infants’ lower scotopic visual sensitivity. Sensitivities derived from scotopic ERG and VEP responses to full-field stimuli were examined. ⁴ The differences between infants’ and adults’ rod (ERG a-wave), inner retinal (ERG b-wave), and VEP sensitivities were all the same, ⁴ and, in fact, similar to the infant–adult difference subsequently found for peripheral visual sensitivity at age 10 weeks, ⁴ namely approximately 0.5 log unit. Thus, it seemed likely that the determination of 10-week-old infants’ scotopic, dark-adapted visual sensitivity was primarily in the rods. 

The present study evaluates sensitivity from early infancy through middle age. The concurrent courses of development of rod sensitivity, b-wave sensitivity, and peripheral visual sensitivity suggest that the maturation of the rod outer segments and quantum catch govern the developmental increase in dark-adapted scotopic visual sensitivity. Although recognizing the influence of receptive field immaturities on infants’ scotopic visual responses, ^{2 29 41 44} the data herein as well as the results of analysis of the eigengrau in infants’ parafoveal and peripheral retina ⁴⁵ are evidence that rod outer segment immaturities, with consequent low probability of quantum capture, are critical determinants of dark-adapted infants’ low scotopic visual sensitivity. 

 Supported in part by Grant R01-EY10597 from the National Institutes of Health, Bethesda, Maryland.

 Submitted for publication August 23, 1999; revised December 9, 1999; accepted December 16, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

The points plot the percent of the mean adult rhodopsin content per eye as a function of age during rat rod photoreceptor development. The data are taken from Table 2 of Bonting et al. ¹⁶ The smooth curve is a logistic growth curve of the form y/y _max = Age ⁿ /(Age ⁿ + Age₅₀ ⁿ ), where Age₅₀ is the age at which y is 50% of y _max, the adults’ value; the exponent n indicates the steepness of the function about Age₅₀. According to the logistic growth curve fit to the points, Age₅₀ is 20 days, and the exponent n is 3.3. The fitted curve accounts for 99% of the variance of rhodopsin with age.

Figure 2.

 

Sample ERG records and model fits of a-wave and b-wave responses in a 33-day-old infant and a 35-year-old adult. The numbers to the left of the ERG traces indicate the stimulus in log scot td sec. Top right: in these two panels the first 40 msec of the ERG responses are shown by the continuous lines and the model (Eq. 1) fit to the a-wave by the dashed lines. The calculated values of S, R _mp3, and t _d are as indicated. Bottom right: in these two panels, the triangles plot b-wave amplitude as a function of stimulus intensity. The smooth curves represent Equation 2 fit to the points, up to those representing responses to higher intensity stimuli at which the a-wave intrudes. ³⁷ The calculated values of ς and V _max are as indicated.

Table 1.

 

Scotopic ERG

	A. Comparison of ERG Parameters in 10-Week-Old Infants and Children and Adults
		ERG Parameter	10-Week-Old Infants (n = 27)*	Children and Adults (n = 25)*
a-Wave (Eq. 1)	S (scot td sec)	0.59 ± 0.19	1.20 ± 0.19
	R mp3 (μV)	161 ± 51	383 ± 75
	t d (ms)	3.6 ± 0.7	3.4 ± 0.5
b-Wave (Eq. 2)	Log ς (scot td sec)	−0.38 ± 0.18	−0.84 ± 0.10
	V max (μV)	194 ± 62	379 ± 59

ERG Parameter	B. Growth Curve Parameters (Eq. 3)
		No. of Subjects	y max (μV)	Age50 (weeks), †	Exponent, n	r 2	Value at 10 Weeks
S (scot td sec)	71	1.20 (1.12–1.28)	10.7 (8.9–12.5)	1.2	0.76	0.57
	93	1.17 (1.10–1.24)	10.0 (8.1–11.9)	1.4	0.66
R mp3 (μV)	71	383 (357–409)	11.2 (9.4–13.0)	1.26	0.73	177
	93	388 (365–411)	11.8 (9.4–14.2)	1.26	0.68
Log ς (scot td sec)	142	−0.85 (−0.82–−0.88)	11.0 (9.3–12.7)	2.7	0.57	−0.35
V max (μV)	142	384 (366–402)	10.0 (8.3–11.7)	1.2	0.68	218

*  Values are means ± SD.

†  Values in parentheses are 95% CI.

Figure 3.

 

Photoreceptor response parameters (S and R _mp3 in Eq. 1 ) plotted as a function of age and mean adult (±SD) values (filled squares). (A) Rod photoreceptor sensitivity, S. The circles represent the 71 main subjects in this study; the open circles indicate those data previously reported. ^{7 38} The solid line is the logistic growth curve (Eq. 3) fit to the circles. Age₅₀ is 10.7 weeks (95% CI, 8.9–12.5 weeks). See also Table 1B . The dashed curve is Eq. 3 fit to the circles and inverted, filled triangles, which represent the 22 patients with normal eyes. (B) Saturated amplitude of the rod response, R _mp3. The symbols are as in (A). The solid line is a logistic growth curve fit to the circles (n = 71); Age₅₀ is 11.2 weeks (95% CI, 9.4–13.0 weeks). See also Table 1B . The dashed line is the growth curve fit to the circles and inverted triangles (n = 93).

Figure 4.

 

b-Wave response parameters (ς and V _max in Eq. 2 ) plotted as a function of age and mean adult (±SD) values (filled squares). In both panels, the circles represent normal subjects (n = 71), as in Figure 3 , and open squares (n = 71) replot the data of normal subjects from ^{Ref. 39} . The solid line is Eq. 3 fit to the data from all normal subjects (n = 142). Also shown are the data from patients with normal eyes; the inverted, filled triangles are for individuals whose a-wave results are plotted in Figure 3 , and the open diamonds are for those patients with normal eyes whose b-wave data were reported in ^{Ref. 39} . (A) Log ς, the flash producing a half maximum b-wave amplitude. Age₅₀ for normal subjects is 11.0 weeks (95% CI, 9.3–12.7 weeks); see also Table 1B . Note that this is a log-log display, with the rising limb linear, rather than S-shaped, as in the linear-log display shown in (B). (B) Saturated b-wave amplitude, V _max. Age₅₀ for normal subjects is 10.0 weeks (95% CI, 8.3–11.7 weeks); see also Table 1B .

Table 2.

 

Dark Adapted Scotopic Visual Thresholds

Stimuli, 50-msec Duration			A. Comparison of Thresholds in 10-Week-Old Infants and Adults
				Mean Thresholds (log scot td sec)*			Infant–Adult Difference					P
Diameter	Eccentricity	10-Week-Old Infants	Adults	Log Unit	t-Value	df
2°	30°	−2.65 ± 0.15 (25)	−3.21 ± 0.14 (18)	0.56	12.16	41	<0.01
2°	10°	−2.15 ± 0.16 (25)	−3.21 ± 0.13 (18)	1.06	23.27	41	<0.01
10°	20°	−2.68 ± 0.27 (68)	−3.90 ± 0.12 (26)	1.22	22.3	92	<0.01

Stimuli, 50-msec Duration			B. Growth Curve Parameters (Eq. 3)
Diameter	Eccentricity	No. of Thresholds	y max (%)	Age50 (weeks), †	Exponent, n	r 2	Threshold at 10 Weeks (log scot td sec)
2°	30°	62	100	13.5 (12.2–14.8)	2.7	0.66	−2.68
2°	10°	62	100	19.5 (17.2–21.8)	2.7	0.78	−2.36
10°	20°	126	100	33.0 (26.2–39.8)	2.9	0.85	−2.36

*  Values in parentheses are number of subjects.

†  Values in parentheses are 95% CI.

Figure 5.

 

For stimuli at 30°, 10° (parafoveal), and 20° eccentric in (A), (B), and (C), respectively, the dark-adapted visual sensitivity (left) or threshold (right) is shown as a function of age. On the left, the smooth curve is the logistic growth model fit to the data expressed as percent of adults’ mean sensitivity and displayed on log-linear coordinates. On the right, the data are plotted as log threshold, and the fitted growth curve is transformed for the log-log displays. In each panel, the square and error bars indicate the mean ± SD value in adults.

Figure 6.

 

Summary of developmental courses of rhodopsin, retinal and visual sensitivity. The results of the growth curve analyses presented in Figures 3 4 and 5 are shown as the Age₅₀ values along with the 95% CIs. Also shown is the Age₅₀ value, 5 weeks (95% CI, 0–10 weeks), reported for human rhodopsin. ¹⁵ The CIs for rhodopsin and the ERG and peripheral visual sensitivity parameters overlap.