The most common behavioural method used to investigate infant vision by researchers is the forced choice preferential looking (FPL) test. The FPL technique was conceived by Davida Teller10 and has been used by various research groups such as Dobson et al.,11 Atkinson and Braddick,12 Banks and Salapatek13 and Gwiazda et al.14 In FPL the observer presents a display to the child, half of which is plain and the other half contains a pattern. They measure resolution acuity, using either a grating target as with the Teller cards (Figure 1A), or the vanishing optotype principle, as with the more recently developed Cardiff Acuity Cards15(Figure 1B). The child will tend to look at the pattern if she or he can resolve it. This technique becomes a “forced choice” method when the observer has to decide, based on their observation of the child's head and eye movements, where the stimulus is located. The threshold is usually defined as when the observer is correct 75% of the time. Operant preferential looking (OPL) is a modification of the forced-choice preferential looking test. In the OPL test, some kind of reward, such as the appearance of an animated toy, is given to the child when the observer is able to correctly determine the position of the target. When the observer makes a correct judgment, it implies that the child made a correct looking response. These methods were developed into the acuity card technique which is now used clinically. In the acuity card technique, rather than finding the 75% correct level, the threshold is defined as when the child makes a “clear” look (see figure 2 for the acuity cards being used in a clinical setting). There have been numerous studies of visual acuity development in infants using these techniques which are generally in agreement, showing that VA develops from about 1 cycle per degree (this is often taken to be equivalent to 6/180 Snellen=0.0333 decimal acuity) in the newborn to 2.6-12 cycles per degree (cpd) at one year.16,17

Figure 1.

A. Teller acuity cards. B. Cardiff acuity cards. Both are examples of resolution acuity.

(0.12MB).

Figure 2.

Teller acuity cards in use. The child is making a looking response to the left hand side of the card.

(0.13MB).

Atkinson and Braddick12 used OPL with older children and found that by the age of 3 years resolution acuity was very close, but not equal to adult values. Table 1 summarises the results of those studies of the development of VA that included 5-year old or older children. Mayer and Dobson,18 using OPL, showed that acuity develops systematically with age up to 40 cpd at 5 years. There was no statistical difference between the VA of the 4 and 5 year olds and the adults. Birch et al.,19 also using an operant procedure, found that visual acuity reached the adult asymptote at the age of 6. Heersema and van Hof-van Duin20 and Neu and Sireteeau,21 both using acuity cards, found that VA was not yet adult-like by 4 or by 6 years, respectively. Abramov et al.6 showed a difference between the 6-8 and the 18-40 year olds at the high end of the spatial frequency curve, but did not provide statistics. On the other hand, Ellemberg et al.3 found that acuity by 6-7 years was not significantly lower than that measured for adults.

There have been very few studies comparing recognition acuity in children and adults. Figure 3 shows some examples of recognition acuity. Although numerous matching tests using pictures (Figure 3A), shapes (Figure 3B) or letters (Figure 3B) have been developed for preschool children, many of these do not have published norms and many of the studies that do give normal data fail to compare this with an adult group.17,22 Stiers et al.23 measured both grating and Landolt-C acuity and found that neither had reached adult levels by the end of the fifth year. Atkinson and Braddick12, using Snellen acuity, found that crowded optotype acuity was 58% of that of adults at 5 years of age. Drover et al.24, using crowded optotypes, found a difference between 7 year olds and 8-10 year olds, but did not comment on whether there was a difference between 8-10 year olds and adults. De Vries and Spekreisje25 showed data that indicate that the trend line continues upward until the mid-teens, but there was no statistical analysis.

Figure 3.

Examples of recognition acuity. A. Kay pictures B. LEA symbols. C. Cambridge Crowding cards.

(0.19MB).

From table 1 it can be seen that, taking the studies together, there is some evidence that grating acuity matures earlier than recognition acuity. The mode seems to show that grating acuity has reached adult levels by 6-7 years (rather than before 6 or after 6). The evidence from studies of recognition acuity indicates that it matures sometime between 7 years and the mid-teens. There are a number of factors that may influence the different findings between studies, e.g. methods used or sample size. It must be remembered that these results include the influence of behavioural and cognitive differences between adults and children, which will be discussed later.

Objective methods used for the assessment of visual acuity development are mainly pattern steady-state VEP (ssVEP) and variations of pattern VEP such as sweep VEP (sVEP). Pattern VEP can be used to measure VA by measuring the response amplitude of the VEP at each frequency, for a range of spatial frequencies. The amplitude is plotted against spatial frequency and a regression fit is used to determine the point at which the response would become zero, which is used as an estimate of VA. Sweep VEP essentially performs the same operation, but the spatial frequencies are varied very quickly over time and the amplitudes are immediately plotted with respect to spatial frequency (or time). For example, to measure VA, the spatial frequency changes from low to high in about 10-20 seconds. The regression line of the response amplitude is extrapolated to zero, which gives a measure of the VA.8,26-33Figure 4A shows a video of the sweep VEP stimulus for VA testing and figure 4B is a plot of the resultant VEP amplitude against spatial frequency. Each measure of sweep VEP measures vision faster than regular pattern VEPs, so it is ideal for children who have shorter attention spans than adults.

Figure 4.

A. Shows a video of the stimulus for sweep VEP measurement of VA (link to movie in electronic PDF version, available at: www.journalofoptometry.com). B. A sVEP plot showing the amplitude in μV plotted against spatial frequency (1 to 40 cpd) used to estimate visual acuity. The recording was taken from the Oz electrode (mid-line just above the visual cortex) and the open circles show the background noise at each spatial frequency. The pink line is the regression line extrapolated to zero amplitude, which is usually taken as the threshold. The vertical dotted lines show the data that was included to fit the regression line. The recording was at 2F i.e. two times the fundamental temporal frequency.

(0.1MB).

Most investigations using VEP to measure VA have studied infants. VEP-measured VA tends to be higher than OPL acuity in the first year of life.34-36 The difference decreases with age until about 12 months. By one year of age, VEP estimates of visual acuity are closer to adult levels than those estimates based on behavioural techniques. Norcia and Tyler,26 using sVEP, showed that visual acuity increased from a mean of 4.5 cpd during the first month to about 20 cpd at 8-13 months of age. By 8 months the authors stated that it was not “reliably different” from adults (24.3 cpd), but this difference was not tested statistically. In 1989, Hamer et al.37 reported similar data. Note that, in adults, VEP often gives lower sensitivity values than psychophysical methods.36,38 In a later study, Norcia et al.38 reported that visual acuity was 16.4 cpd at 7 months. Skoczenski and Norcia39 showed that VEP VA at one year is within about one octave of adult acuity; i.e. it is closer to adult's acuity than PL, but still not adult-like. Sokol et al.35 also showed results that indicate that VEP acuity at one year is still different from that measured on adults. It is important to recognise that the exact measure of VA obtained with sVEP or ssVEP depends on many parameters; e.g. whether data from electrodes is averaged or the best value is taken,37 whether a checkerboard or a grating is used, and the exact method for estimating the acuity threshold. Thus, when comparing infants and children with adults, the same techniques should be used on both. Of the studies mentioned above, Norcia and Tyler,26 Sokol et al.35 and Skoczenski and Norcia39 included control adult subjects.

There are almost no data on visual acuity development using either sVEP or ssVEP techniques in children above the first year of life. In fact, the only study of visual acuity itself is that of De Vries and Spekreisje25 using a pattern VEP. Their plotted results showed that the threshold clearly decreases until 8 years of age, reaching a plateau somewhere in the teens. Gordon and McCulloch2 used a VEP technique to investigate parallel visual pathway development in primary school age children. They found that the magnocellular (M) pathway develops earlier than parvocellular (P) pathway and that there was some evidence that the P pathway is not developed by 5 years. The P pathway is considered to be the main carrier of high contrast, color perception and high spatial frequency (detail) information, especially at lower temporal frequencies (slowly moving or static targets); i.e., visual acuity information.

Thus, the studies using subjective/behavioural techniques indicate that there is some variability in the exact age at which VA is adult-like in children. It appears that VA reaches adult levels somewhere between 6 and 10 years of age. There is even less data using VEP, but those that do exist seem to suggest an even later age for full development.

Similar to its use for acuity thresholds, VEP can be used to measure contrast thresholds. In this case contrast is the parameter that is varied, and contrast thresholds are determined by extrapolation to zero amplitude. Figure 5A shows a video of the sweep VEP stimulus for CS testing and figure 5B is a plot of the resultant VEP amplitude against contrast. Sweep VEP has been developed for this purpose38 and a ssVEP may also be used. Sweep VEP has been used to study CS development in infants38,49 and has shown that CS develops rapidly for both low and high frequencies in the first 3-4 months of life. Norcia et al.38 found that there are two phases in the development of contrast sensitivity and acuity. Between the 4th and the 9th week, overall contrast sensitivity increased by a factor of 4-5 at all spatial frequencies. Beyond 9 weeks, contrast sensitivity at low spatial frequencies remained relatively constant, while sensitivity increased systematically at higher spatial frequencies.

Figure 5.

A. Shows a video of the stimulus for sweep VEP measurement of contrast sensitivity. (link to movie in electronic PDF version, available at: www.journalofoptometry.com). B. shows a similar plot to Figure 4B but for contrast sensitivity measurement. In this case amplitude is plotted against % contrast.

(0.07MB).

As for VA, there are very few studies which have used objective, VEP techniques to study CS development in childhood. De Vries and Spekreisje25 showed that the trend line was upward until 7 years, but this relates to the optimum checkerboard size, rather than contrast sensitivity per se.

It can be seen that there is considerable variation between the various studies; while some showed that CS develops by 7 years, others gave an age of maturation as high as 10-19 years.