Differences in Time Course and Visual Requirements of Ocular Responses to Lenses and Diffusers

Abstract

purpose. Myopia can be induced in chickens by having them wear either negative lenses (lens-compensation myopia [LCM]) or diffusers (form-deprivation myopia [FDM]), whereas positive lenses cause lens-compensation hyperopia (LCH). These three conditions were compared with respect to (i) their early time course and (ii) the effect of two manipulations of the lighting.

methods. Longitudinal changes in ocular dimensions and refractive error were measured in chicks maintained under three different conditions: (i) wearing either −15 D lenses or diffusers in a normal light/dark cycle; (ii) wearing either +15 D lenses, −15 D lenses, or diffusers with brief periods of stroboscopic lights at the beginning and end of the dark period; (iii) wearing either +6 D lenses, −6 D lenses, or diffusers with the nights interrupted by brief periods of white light. In addition, scleral and choroidal proteoglycan synthesis was measured in eyes that wore positive lenses, negative lenses, or diffusers for 3 hours followed by different periods of darkness.

results. (i) The time course of the changes in axial length over the first 72 hours was significantly faster in LCM than in FDM. Indeed, the diffusers did not begin to significantly affect the total length of the globe for 3 days, although the vitreous chamber had deepened after 9 hours, because the choroid thinned extremely rapidly (within 1 hour) with either diffusers or negative lenses. (ii) Scleral proteoglycan synthesis was higher in eyes with negative lenses than in those with diffusers at 11 hours, but the reverse was true at 27 hours. (iii) Brief periods of stroboscopic light attenuated FDM more than LCM. (iv) In contrast, interruption of the nights by brief periods of light attenuated LCM more than FDM. (v) Neither lighting manipulation affected LCH. (vi) Choroidal proteoglycan synthesis decreased similarly with 3 hours of wearing either diffusers or negative lenses.

conclusions. Although both negative lenses and diffusers cause similar increases in the rate of ocular elongation, the responses differ in time course and in the effect of manipulations of the daily lighting. The responses to positive lenses differ from both of these.

In the past three decades, research on experimental myopia in several animal species has increased our understanding of refractive development. All the animal species tested developed myopia under conditions of visual form deprivation (e.g., monkey, ¹ marmoset, ² tree shrew, ³ and chicken ⁴ ) or under the hyperopic defocus conditions imposed by negative lenses (e.g., monkey, ⁵ marmoset, ⁶ tree shrew, ⁷ and chicken ⁸ ); furthermore, the imposition of myopic defocus by positive lenses results in hyperopia (e.g., monkey, ⁹ marmoset, ⁶ and chicken ⁸ ). However, whether the eye uses the same mechanism to respond to diffusers and lenses is still open to question and is an issue of considerable importance. One possibility is that the responses to diffusers and to negative and positive lenses are all explicable by a simple rule: obscuring the visual image causes growth toward myopia, whereas sharp vision causes growth toward hyperopia. ¹⁰ According to this view, diffusers would obscure vision the most and hence would cause the strongest myopia, negative lenses would obscure vision less, and positive lenses would enhance the sharpness of the images (assuming that the chicks mostly looked at nearby objects and that without positive lenses, vision of near objects would be blurred because of accommodative lag). A test of this hypothesis would be to impose novel visual conditions and see if the responses to negative lenses and diffusers are similarly affected. 

On the basis of evidence from biochemical studies and from the effects of cutting the ciliary nerve or the optic nerve, Schaeffel et al. ¹¹ have proposed that the response of eye growth in chickens to diffusers and lenses reveals three different mechanisms. Thus, form-deprivation myopia (FDM) and lens-compensation hyperopia (LCH) are controlled locally within the retina, with only FDM being blocked by reserpine, whereas lens-compensation myopia (LCM) is centrally controlled. 

There have been three explicit comparisons of FDM and LCM that revealed differences between them. First, at the retinal level, the oscillatory potentials of the electroretinogram are reduced only in the FDM eyes, even though the other components of the electroretinogram are the same in FDM and LCM. ¹² Second, it is known that optic nerve section does not prevent the myopia and increased vitreous chamber depth of FDM. ^{13 14} Wildsoet ¹⁵ found that optic nerve section reduces the vitreous chamber elongation of eyes with LCM, but has no effect on eyes with FDM. Third, an abstract from Schmid and Wildsoet ¹⁶ asserts that continuous stroboscopic illumination attenuates FDM over a wider range of frequencies than that of LCM. 

In this article, we first compare LCM and FDM with respect to the early time course in ocular elongation, choroidal thinning, and proteoglycan synthesis. (In the case of proteoglycan synthesis, we examined the response to a brief “pulse” of lens or diffuser wear.) Next we ask whether they are similarly affected by two visual manipulations shown by Nickla ¹⁷ to affect FDM: (a) brief periods of stroboscopic light at dawn and dusk and (b) brief periods of light interrupting the night. 

Methods

Subjects

Newly hatched White Leghorn chicks (Gallus gallus domesticus) from Truslow Farms (Chestertown, MD) were raised in temperature-controlled brooders from 1 day of age. The chicks were fed Chick Starter Chow (Purina Mills, St. Louis, MO). White fluorescent lights were on from 8 AM to 10 PM every day. Care and use of the animals conformed to the ARVO Resolution on the Use of Animals in Ophthalmic and Vision Research. 

To induce myopia or hyperopia, one eye of each chick was covered by either white translucent plastic hemispheres (diffusers) or PMMA lenses with various refractive powers. To do this, a Velcro ring (Velcro USA Inc., Manchester, NH) was secured around the chick’s eye with Collodion (Fisher Scientific, Fairlawn, NJ). A mating ring of Velcro carried either a diffuser or a lens attached with Norland Optical Adhesive (Norland Products, Inc., New Brunswick, NJ). This method provides the convenience of being able easily to remove the lenses for cleaning or for ocular measurements. Lenses and diffusers were cleaned twice daily. In these experiments, we used positive and negative 6 and 15 diopter (D) lenses. From unpublished studies we know that the initial changes in axial length and choroidal thickness of 6 and 15 D lenses of the same sign are indistinguishable. 

Measurements

The chicks were anesthetized with halothane (0.8%; Halocarbon Laboratories, River Edge, NJ) during the refraction and ultrasound measurements. Typically, the chicks recovered from this inhalation anesthesia within minutes. For measurement of eyes during the period of lens wear, the diffusers or lenses were removed only in a darkened room after the chicks had been anesthetized and were put back immediately after the measurement. 

Refractometry

Refractometry was performed along the pupillary axis by using a Hartinger refractometer (Jena Coincidence Refractometer; Carl Zeiss, Jena, Germany) as described in previous articles. ^{18 19} This method yielded repeatable refractions and low interobserver variability (average SD for refraction in normal 4-week-old animals was within ±0.3 D). 

A-Scan Biometry

A-scan biometry was performed by using a high-frequency focused polymer transducer (30 MHz; Panametrics, Inc., Waltham, MA), the output of which was sampled by a computer at 100 MHz. The details of the measurement have been described previously. ^{18 20} The repeatability of the measurement was estimated to be approximately± 20 μm for all ocular components. ¹⁷ We used the same criteria as Nickla ¹⁷ in selecting the peaks representing different ocular layers. In contrast to the usual practice of reporting the axial length as the distance from cornea to retina, we use the total axial length, the distance from the anterior corneal surface to the posterior scleral surface. Also, we consider the anterior chamber dimension to be the distance from the anterior surface of cornea to the anterior surface of crystalline lens. 

Proteoglycan Synthesis

Scleral and choroidal proteoglycan synthesis was estimated by measuring incorporation of [³⁵S]sulfate into glycosaminoglycans (GAGs) in 6-mm punches from the posterior eye wall. ²¹ Sclera and choroid were separately cultured for 2 hours in N2 defined medium labeled with Na₂ ³⁵SO₄ and then digested with proteinase-K and centrifuged. GAGs were precipitated with cetyl pyridinium chloride, captured on filters, and scintillation-counted. Scleral GAG synthesis was normalized to the DNA content of the tissue determined by fluorometry using Hoechst 33258 dye; this was not done in the case of choroids because of the contaminating effect of nucleated red blood cells. 

Experimental Manipulations

Time-Course Experiment (Expt. I).

Two groups of seven 3-day-old birds, each wearing either a diffuser or a −15 D lens monocularly, were reared in 5000 cm², evenly lit, sound-proof chambers under normal lighting conditions (14 hours light/10 hours dark). We measured the ocular dimensions along the pupillary axis by using A-scan ultrasound at the start of the experiment and after 3, 9, 25, 48, and 71 hours of diffuser or lens wear. To obtain a more detailed view of the early time course of the choroidal response, we made hourly measurements of the ocular dimensions of other 2- and 3-day-old birds, wearing either a diffuser or a −15 D lens (n = 4 in each group). 

In separate groups of birds, the time course of scleral and choroidal proteoglycan synthesis was measured. Two-day-old chicks had one eye covered either by a diffuser or a −6 or −15 D lens for 3 hours or by a +6 D lens for 6 hours, followed by varying delays of up to 24 hours in the dark. 

Stroboscopic Light Experiment (Expt. IIa).

Birds in both the experimental group and the control group wore over one eye either a diffuser, a +15 D lens or a −15 D lens from days 6 to 11 (n ≥ 7 in each group), while housed as described above. The experimental group received stroboscopic light (15 Hz) for 30 minutes just before lights on (from 8:00 AM to 8:30 AM) and 30 minutes just after lights off (from 9:30 PM to 10:00 PM) from days 6 to 11. White fluorescent light (∼240 lux) was on from 8:30 AM to 9:30 PM. The control group received a normal light cycle, with the white fluorescent light on from 8 AM to 10 PM. Refractometry and A-scan ultrasonography were performed at the start of the experiment and after 1, 3, and 5 days. 

Interrupted Night Experiment (Expt. IIb).

The birds in both the experimental and the control group wore either a diffuser, a +6 D lens or a −6 D lens over one eye from days 2 to 5 (n = 8 in each group); they were housed as described above. The experimental group received 5 minutes of fluorescent light (∼240 lux) every 20 minutes during the 10 hours of night. The control group had normal dark nights. Refractometry and A-scan ultrasonography were performed on both eyes at the start of the experiment and on each successive day for 3 days. 

Analysis

Paired t-tests were used when comparing the eyes of individual birds within the same group; unpaired t-tests were used when comparing different groups of birds (Minitab Release 10.51 Xtra; Minitab Inc., State College, PA). Two-tailed t-tests were used except in the case of the stroboscopic lighting experiment because stroboscopic light has been shown to attenuate FDM in previous studies. ^{16 22 23} ANOVA was used for comparisons across experiments or across ages or to assess interactions between factors. 

In the time-course experiment (Fig. 2) , the developmental change in total axial elongation was expressed as the induced change in experimental eye (posttreatment − pretreatment) minus the change in fellow eye (posttreatment − pretreatment). In the visual-manipulation experiments (Figs. 6 and 7) , the parameter shown is the change in the experimental eyes from the preexperiment measurement. 

Scleral and choroidal proteoglycan synthesis data were expressed as ratios of experimental eye tissue to fellow eye tissue. The ratios were compared to a null hypothesis of mean = 1 by Student’s paired t-test and compared with each other by unpaired t-test. 

Results

Time-Course Experiment (Expt. I)

In general, both negative lenses and diffusers caused an increase in total axial length and a decrease in choroidal thickness, but the length changes were more rapid for eyes wearing negative lenses. To show the relationships among ocular dimensions, we plotted the changes in all axial components in experimental and fellow eyes wearing −15 D lenses and diffusers (Fig. 1) . 

Although arithmetically the length of the eye is the sum of its components, this is not a biologically sensible way to view eye growth. Rather, it is more useful to consider that the total axial length is the size of the globe, determined by the growth of the sclera and cornea, whereas the depth of the vitreous chamber is what remains after subtracting the thickness of the lens, retina, choroid, and so on. Thus, we would consider the change at 3 hours in Figure 1 as meaning that the total axial length did not change, but the choroid thinned, resulting in a corresponding increase in the vitreous chamber depth. 

Axial Length

Eyes wearing negative lenses elongate rapidly, being significantly longer than their fellow eyes by 25 hours (Fig. 1 , paired t-test, P < 0.05). By 48 hours, the rate of elongation relative to the fellow eye is five times greater than in eyes wearing diffusers, a significant difference (Fig. 2 ; unpaired t-test, P < 0.05). No significant difference was found between the fellow eyes of lens-treated eyes and the fellow eyes of form-deprived eyes at any time point. 

To show the ocular elongation relative to the fellow eyes, we plot together all the lens- and diffuser-wearing eyes of animals in normal visual environments from our three experiments (Fig. 2) . This includes the control animals for the strobe and interrupted night experiments. In each of these groups, by the second time point, negative lenses caused significantly more rapid elongation than did diffusers (t-test, P < 0.05), despite the differences in age and lens power. After 3 days the rate of elongation was no longer significantly greater in the eyes wearing −6 D lenses than that in the eyes wearing diffusers, presumably because the lens-wearing eyes had already compensated for the lenses. However, the eyes wearing −15 D lenses still maintained a greater rate of elongation than the eyes wearing diffusers (unpaired t-test, P < 0.01), presumably because compensation was still actively going on. 

Choroidal Thickness

One of the surprising findings of the time-course experiment was that the choroidal thinning was maximal by the first time point at 3 hours after fitting either negative lenses or diffusers (Fig. 1) . To see the effect of form deprivation and negative lenses on choroidal thickness over even shorter times, we measured the choroid thickness by ultrasonography every hour for 4 hours in 2- and 3-day-old chicks (n = 4 in each group). This experiment confirmed that the choroidal thinning occurred very rapidly, being nearly maximal within 1 or 2 hours of either diffuser or negative-lens wear (Fig. 3) . The difference between the effect of negative lenses and diffusers on choroidal thinning is not statistically significant. Negative lenses cause significant choroidal thinning at both ages tested (one-way ANOVA, P < 0.05 for each age group), and diffusers were significant only in older birds (one-way ANOVA, P = 0.19 for 2-day-old and P < 0.01 for 3-day-old birds). The magnitude of choroidal thinning after 4 hours agrees well with that shown in Figure 1 . No significant difference in choroidal thickness was observed in fellow eyes of lens-treated eyes compared with fellow eyes of form-deprived eyes. 

Relationships among Components

We confirmed that the refractive error differences that we observed were attributable to axial differences by plotting refractive error versus vitreous chamber depth (Fig. 4) . The data from all the treatment groups fell on one line, indicating that the differences in refractive error among groups can be largely accounted for by changes in vitreous chamber depth (Fig. 4A ; slope = −26.8 D/mm; R ² = 0.84, P < 0.05). Thus, our treatments all appear to modulate ocular elongation, rather than some treatments having effects on corneal curvature or lens power. 

Furthermore, we find that changes in the vitreous chamber depth in the positive-lens wear eyes can be largely accounted for by changes in choroidal thickness (Fig. 4B ; slope = −0.71; R ² = 0.79, P < 0.001), whereas in the eyes wearing negative lenses or diffusers, changes in choroid thickness are not correlated with changes in vitreous chamber depth (except during the first day; Fig. 1 ). Instead, there is a rapid and uniform choroidal thinning shown in Figure 1 and shown by the near-zero slopes in Figure 4B . This asymmetry results because with both +6 and +15 D lenses the eyes rapidly stop elongating, leaving the choroidal expansion to produce the differences in vitreous chamber depth. In contrast, with negative lenses, the choroid rapidly thins maximally with either power of lens, leaving differential ocular elongation to produce the differences in the vitreous chamber depth. 

Proteoglycan Synthesis

It has been previously shown that chick eyes wearing either diffusers or negative lenses increase the net synthesis of proteoglycans in the sclera. ^{24 25} This increased total synthesis includes increased synthesis in the cartilaginous layer and decreased synthesis in the fibrous layer; positive lenses have the opposite effect. ²¹ We find that the time course of the increase in scleral proteoglycans is more rapid with negative lenses than with diffusers (Fig. 5) . After 3 hours of negative-lens wear, there was a sharp rise in synthesis between 4 and 8 hours later (time = 7–11 hours), so that by the 11-hour time point, the rate of scleral proteoglycan synthesis had nearly doubled relative to that in the untreated fellow eyes, a significant change (ratio = 1.86, n = 14, P < 0.05), and this level was maintained for 16 more hours in the dark. In contrast, after 3 hours of diffuser wear, a sharp rise occurred between 18 and 24 hours later (at time = 27 hours, ratio = 2.8, n = 12, P < 0.0001). Furthermore, immediately after diffusers were worn for 3 hours, the scleral proteoglycan synthesis was slightly but significantly reduced (ratio = 0.86, n = 27, P < 0.05); this did not occur after 3 hours of wearing either −6 D lenses (ratio = 1.21, n = 8, P > 0.05) or −15 D lenses (ratio = 1.18, n = 32, P = > 0.05). These ratios were significantly different (by two-sample t-tests, P < 0.05 either for −6 D versus FD or for −15 D versus FD). 

Therefore, the eyes wearing negative lenses synthesized more proteoglycan than did the eyes wearing diffusers immediately after the 3 hours of visual experience (140% as much, P < 0.05) and also after 8 hours of darkness (144% as much, P = 0.001 by two-sample t-test) but synthesized less proteoglycan after 24 hours of darkness (64% as much, P = 0.039). Choroidal proteoglycan synthesis was not different between eyes wearing diffusers and negative lenses; in both cases it was reduced significantly by the 7-hour time point. 

Positive lens wear caused changes in proteoglycan synthesis in both sclera and choroid that were in the opposite direction from those caused by negative lenses and diffusers but differed from both in requiring a longer duration of lens wear: 3 hours of lens wear did not produce significant effects, although 6 hours of lens wear did (Fig. 5) . 

Lighting Manipulations (Expt. II)

In brief, we found that the brief stroboscopic illumination attenuated the myopia resulting from the diffusers more than that resulting from the negative lenses, whereas giving periods of light during the night did the reverse (two-way ANOVA, interaction between treatment (lenses/diffusers) and lighting conditions, P < 0.05 for refractive error, vitreous chamber depth, and total axial length). In this analysis and in the ones presented in the following two sections, we deal with the results from the experimental eyes alone under these conditions. The complete data from these experiments can be found in Kee. ²⁶  

Effects of Stroboscopic Light on Form-Deprivation Myopia and Lens-Compensation Myopia (Expt. IIa).

Brief twice-daily stroboscopic illumination significantly attenuated the myopic refractive error as well as the increases in vitreous chamber depth and ocular elongation resulting from 3 days of form deprivation (Fig. 6) . If we consider the changes over the 3 days of form deprivation under normal lighting to be an effect of 100%, the changes under stroboscopic light are as follows: refractive error, 20%; total axial length, 45%; vitreous chamber, 30% (P < 0.05 in all cases; one-tailed t-test). This attenuation was much less and was not significant in the negative-lens group (relative to the effect of negative lenses under normal lighting as 100%): refractive error, 81%; total axial length, 82%; vitreous chamber, 94%; or in the positive-lens group (refractive error, 78%; total axial length, 78%; vitreous chamber, 97%). The fact that strobe had a greater effect on the eyes wearing diffusers despite a 0.6 log unit attenuation of the light transmission argues that the difference is not likely to have resulted from differences in light intensity. 

The effect of the lighting condition was mainly on the vitreous chamber depth and on the rate of ocular elongation of the eyes wearing diffusers; the choroid thickness was unaffected by stroboscopic illumination (choroidal thickness changes over the same period in normal lighting versus stroboscopic lighting: negative lenses:− 0.034 ± 0.012 versus −0.055 ± 0.012 mm; positive lenses: 0.231 ± 0.06 versus 0.243 ± 0.020 mm; diffusers:− 0.045 ± 0.016 versus −0.036 ± 0.013 mm). 

To evaluate whether the analysis just presented is compromised by our considering only the experimental eyes, we subtracted the change over the 5 days of the experiment in the fellow eye from that of the experimental eye, both from the experiment presented here and from an unpublished experiment identical with the one presented here except that −6 D lenses were used (experiment conducted by Marc Howlett of University of Newcastle, Newcastle, NSW, and James Mertz of New England College of Optometry, Boston, MA). Across the two experiments, stroboscopic light attenuated the ocular elongation significantly more in the eyes wearing diffusers than in the eyes wearing negative lenses (Table 1 ; one-way ANOVA, P < 0.05). 

Therefore, stroboscopic light attenuates the changes in refractive error and ocular elongation caused by form deprivation more than those caused by negative- or positive-lens wear. 

Effects of Interrupted Night on Lens-Compensation Myopia and Form-Deprivation Myopia (Expt. IIb).

The interrupted night condition significantly attenuated the refractive error change and the vitreous chamber elongation caused by 3 days of wearing negative lenses (Fig. 7) , but had less effect on the changes resulting from form deprivation. If we consider the changes over the 3 days of lens wear under normal lighting to be an effect of 100%, the changes under interrupted night are: refractive error: 47%; total axial length: 42%; vitreous chamber depth: 34% (unpaired t-test; P < 0.05). The form-deprived eyes showed changes (relative to form-deprived eyes in normal lighting) of: refractive error: 89%; total axial length: 64%; vitreous chamber depth: 77% (unpaired t-test; P > 0.05). The eyes wearing positive lenses showed changes of: refractive error: 130%; vitreous chamber depth: 80%; unpaired t-test; P > 0.05. These results are generally similar to those obtained by Nickla. ¹⁷ She found that interrupted night conditions reduced the effect of wearing diffusers on total axial length by 60% compared to our value of 64%. Neither result is significant. In contrast, the lens-wearing eyes in our experiments under the interrupted night conditions were the same length as their fellow eyes, that is, the increased ocular elongation was completely blocked. Nickla’s data on refractive error differ in degree from ours in that her diffuser-wearing animals reared with normal illumination were more myopic than ours, perhaps because her animals wore the diffusers longer. This resulted in a greater difference between the interrupted night and normal illumination condition. We cannot dismiss the possibility that there is a real inconsistency between our two sets of results. 

In contrast to the stroboscopic illumination, which had a significant effect even after 1 day, the interrupted night effects began on the second day. On the first day, negative lenses in both the interrupted-night and control groups induced significant amounts of myopia (paired t-test with fellow eyes, P < 0.05). This myopia decreased (but not significantly) in the following days in the interrupted night group but increased in the control group (see Fig. 7 ). As with the stroboscopic illumination, the interrupted night condition had no significant effect on refractive and ocular dimensional changes in compensation for positive lenses, nor did it have significant effects on choroidal thickness (data not shown). Under interrupted night conditions, even the fellow eyes elongated more slowly than did those of birds under normal lighting conditions (Axial elongation in the fellow eyes to the eyes wearing diffusers and negative lenses after 3 days: normal lighting, 0.28 ± 0.02 mm; Interrupted night, 0.16 ± 0.03 mm; unpaired t-test, P < 0.01. All data are shown in Kee ²⁶ ). 

Discussion

We have reported here two principal results: (a) Eyes wearing negative lenses show more rapid increases in ocular elongation and in synthesis of scleral proteoglycans than do eyes wearing diffusers. (b) Eyes wearing negative lenses or diffusers are differentially sensitive to brief periods of stroboscopic lighting or of light during the night. Both lighting manipulations affect only the rate of ocular elongation and, perhaps for this reason, do not affect compensation for positive lenses, during which the eyes are not elongating very much. 

Development of FDM and LCM

In three separate experiments comparing the effects of wearing diffusers with negative lenses of two powers, we found that the eyes wearing lenses elongate more rapidly than those wearing diffusers (Figs. 1 and 2) . Indeed, eyes wearing diffusers did not elongate more than their fellow eyes until 3 days of diffuser wear. In contrast, eyes wearing negative lenses started to show a stable and significant increase after only 1 day (Fig. 1) . The more rapid response to negative lenses, which permit form vision, than to diffusers, which do not, argues that the degree of increase in ocular elongation is not simply proportional to the degree of image degradation. Rather, it argues for a more specific effect of negative lenses. 

The delayed effect of diffusers on ocular elongation contrasts with the rapid effect on vitreous chamber depth, as shown here (Fig. 1) and by others. We attribute the difference to the fact that our measurements of total axial length are from anterior cornea to posterior sclera and therefore are not influenced by the thickness of the choroid, whereas measurements of vitreous chamber depth as well as conventional measurements of axial length (from cornea to retina) can show an increase if the choroid thins, as happens rapidly when diffusers or lenses are worn (Figs. 1 and 3) . Thus, the steady increase in vitreous chamber depths from the first day of diffuser wear can be attributed to immediate choroidal thinning followed by subsequent ocular elongation. 

Form-deprived eyes transiently reduce their scleral proteoglycan synthesis below that of fellow eyes (P < 0.05, relative to fellow eyes, Fig. 5 ). We also find a hint of a transient slowing of the ocular elongation (Figs. 1 and 2 , first data point for eyes wearing diffusers, P > 0.05) before it starts to increase. Similarly, macaque monkeys wearing diffusers showed a transient hyperopic shift before the axial myopia develops. ²⁷  

Consistent with the ocular elongation results, scleral proteoglycan synthesis increased more rapidly in eyes wearing negative lenses than in those wearing diffusers. Eight hours after a 3-hour “pulse” of lens or diffuser wear, the synthesis was higher in the eyes that had worn lenses, and 24 hours later it was higher in the eyes that had worn diffusers. These more rapid growth responses to lenses than to diffusers suggest that lenses may provoke different biochemical processes than do diffusers, rather than that the degree of growth stimulation is proportional to the degree of image degradation. The biochemical results are not entirely parallel with the ocular elongation results. Lenses produce an enduringly greater ocular elongation than do diffusers, whereas the proteoglycan response to lenses (at least to a single brief period of lens wear) is only transiently greater. 

We found no differences in the choroidal response to negative lenses and diffusers. Both decreased after only 1 hour of lens or diffuser wear (Fig. 3 ; statistically significant for 3-day-old birds), suggesting that this thinning might be the result of contraction of the nonvascular smooth muscle that spans the choroid ²⁸ rather than the result of the fluid movements probably responsible for choroidal thickening. ^{29 30} It is clear from Figure 4B that the defocus imposed by positive lenses caused a dramatic increase in choroid thickness of several hundred micrometers, accounting for 71% of the decrease in vitreous chamber depth of the experimental eye over the 5-day experiment. In contrast, negative lenses or form deprivation caused hundreds of micrometers of vitreous chamber elongation after 24 hours, of which only a minor component was the choroidal thinning (Fig. 1) , presumably because the choroid can only thin by approximately 100μ m and it has fully thinned during the first day. 

Effect of Altered Illumination on FDM and LCM: Possible Circadian Factors

If negative lenses and diffusers influence eye growth by the same retinal signals, then visual disturbances that affect one should affect the other similarly. However, our results show that stroboscopic light at dawn and dusk attenuated the myopia much more in the form-deprivation group than in the lens-compensation group; in contrast, light pulses interrupting the night affected lens-compensation myopia much more than form-deprivation myopia. In both cases the effects were on the rate of ocular elongation. This result indicates that the retinal mechanisms leading to these two forms of experimental myopia are likely to be different, although the resulting anatomic changes may be similar. What might account for the complementary effects of these two lighting manipulations on the two forms of myopia? 

Normal circadian rhythms are kept synchronized with the external day/night cycle by the phase-shifting effect of light delivered at particular times. Thus, light at the start of the night tends to shift circadian rhythms forward (phase-advance) and light at the end of the night tends to shift them back (phase-delay). It is, therefore, plausible that frequent light pulses throughout the night would have a disruptive effect on the synchronization of circadian rhythms. On the other hand, the presence of strobe at dawn and dusk may enhance the synchronization of circadian rhythms with the external day/night cycle. Indeed, animals kept in darkness except for light at dawn and dusk (a“ skeleton” photoperiod) can maintain fully synchronized circadian rhythms. ³¹ We speculate, first, that normal circadian rhythms are necessary for normal ocular growth and also for the enhanced ocular elongation that is the major cause of negative lens compensation. Second, we speculate that form deprivation disturbs the ocular circadian rhythms, and this partly contributes to the resulting myopia. Thus, the dawn and dusk strobe light may normalize the circadian rhythms of the form-deprived eyes, thereby reducing their myopia but have less effect on lens-wearing eyes, because their circadian rhythms are already normal. Conversely, the interrupted night would interfere with lens compensation by disrupting the birds’ circadian rhythms but would have less effect on form-deprived eyes, because their circadian rhythms are already disrupted. 

Our first speculation is supported by the fact that under interrupted night conditions, not only the eyes wearing lenses, but also the untreated fellow eyes elongate significantly more slowly than in normal lighting. 

Our second speculation is supported by evidence of altered circadian function in form-deprived eyes. First, the diurnal rise in retinal dopamine is markedly attenuated in chicks and monkeys under form deprivation. ^{32 33} Second, the daily cycle of ocular elongation is disrupted under form deprivation. ^{20 34} (It is not the case that the cyclicity of ocular elongation is lost under form deprivation as originally suggested ³⁴ but rather that the phase was shifted, causing the morning and evening measurement times to coincide with the times of equality of the axial length in these phase-shifted eyes. ²⁰ ) 

Multiple Output Pathways from the Retina

In addition to the differences in the retinal mechanisms underlying FDM and LCM suggested by our results, we also found a dissociation between the two main ocular components determining the refractive status. Both of our illumination-altering manipulations affected the rate of ocular elongation, without affecting the choroid. Furthermore, neither manipulation prevents the inhibition of ocular elongation in the case of positive lenses. This pattern of results suggests that there may be more than one output pathway from the retina—one controlling choroidal thickness and one controlling overall ocular elongation—and that these pathways have differential susceptibility to visual manipulations. Recent experiments also point in this direction. When chicks wear lenses under conditions of brief episodes of illumination (the rest of the time in darkness), we find that if the episodes are frequent, the eye makes appropriate compensatory responses of both choroid and total axial length to both positive and negative lenses. With infrequent episodes, however, only two of the four compensatory responses occur: the choroidal response of the eyes wearing negative lenses and the ocular elongation response of eyes wearing positive lenses. This pattern of results argues that several distinct signals must be involved. ³⁵ Furthermore, if a weak diffuser is added to positive lenses, the choroidal response is diminished without diminishing the ocular elongation. ³⁶  

In conclusion, although FDM and LCM show similar anatomic changes, their differences in early time course and in the effect of exposure to two visual manipulations as well as the evidence referred to in the Introduction imply that compensation for negative lenses is not a special case of form deprivation. Rather, lens compensation operates by rules of its own. 

 Supported by National Institutes of Health Grants EY-02727 and RR-03060.

 Submitted for publication October 27, 1999; revised May 22, September 5, and November 10, 2000; accepted November 22, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

The change in individual ocular components of eyes wearing negative lenses or diffusers. The lens or diffuser wear began at 3 days of age. Each bar represents the average change in the ocular components from the time the lens or diffuser was fitted: AC, anterior chamber; Lens, crystalline lens; VC, vitreous chamber; Ret, retina; Cho, choroid; Scl, sclera. At each duration of treatment, × indicates the bar for the experimental eye and F for the fellow control eye. Bars above and below the y = 0 line show an increase and decrease, respectively, in mean thickness from the pretreatment value. Arrow, average change in total axial length. *Significant change in total axial length of the experimental eye compared with the fellow eye (paired t-test, P < 0.05).

Figure 2.

 

Mean change in total axial length relative to the fellow eye of eyes wearing lenses and diffusers in three experiments: Expt I, the time-course experiment (days 3–6) and control groups from the stroboscopic light experiment: Expt. IIa, Control, days 6–11, and the interrupted night experiment: Expt. IIb, Control, days 2–5. In each experiment, eyes wearing lenses are shown by open symbols and dashed lines; eyes wearing diffusers are shown by solid symbols and solid lines. *Significant difference occurred between LCM and FDM in that experiment (unpaired t-test, *P < 0.05, **P < 0.01, ***P < 0.001). Error bars, SEM.

Figure 3.

 

Mean choroidal thickness change over the first 4 hours of wearing −15 D lenses or diffusers by 2- and 3-day-old birds. Error bars, SEM.

Figure 4.

 

The effects of lens and diffuser wear on changes in refractive status, vitreous chamber depth, and choroidal thickness of eyes from the control groups of the two visual manipulation experiments (stroboscopic light experiment, days 6–11; interrupted night experiment, days 2–5). (A) Refractive status change is correlated with vitreous chamber depth change in eyes wearing either lenses or diffusers (P < 0.05). (B) In the positive-lens-wear group, a decrease in vitreous chamber depth is accompanied by an increase in choroidal thickness (P < 0.001); in the negative-lens- and diffuser-wear groups, there is no association between these variables. The data plotted are the changes from the pretreatment measurement from individual eyes (not interocular differences). The time interval over which the changes were calculated ranged from 1 to 5 days.

Figure 5.

 

The effects of briefly wearing lenses and diffusers on subsequent proteoglycan synthesis in sclera (top) and choroid (bottom). The lens or diffuser wear began at time 0 (2 days of age) and lasted for the duration shown in the inset; the animals were then kept in darkness until the time shown on the x-axis. By 11 hours scleral proteoglycan synthesis was significantly greater with negative lenses (filled circles) than with diffusers (triangles; two-sample t-test). Top: all data points are significantly (P < 0.05) different from 1, except those marked n.s.; bottom: significant points are shown by asterisks. The number of eyes for each data point is shown in parentheses, except for the 3- and 7-hour diffuser scleral time points, which were 27 and 18, respectively. Error bars, SEM.

Figure 6.

 

The effects of twice daily periods of stroboscopic light on the consequences of wearing diffusers and lenses (days 6–11). Stroboscopic light inhibits form-deprivation myopia development, but it does not inhibit the lens compensation. *Significant difference (P < 0.05) occurred between the experimental group (stroboscopic light) and control group (normal light). The data shown are from the experimental eyes only. Error bars, SEM.

Table 1.

 

Effect of Brief Stroboscopic Lighting on Axial Length

	Mean Difference (μm) (Experimental Eye-Fellow Eye)
		Diffusers	Negative Lenses
Present experiment	−44 ± 305 (7)	301 ± 177 (8)
Howlett/Mertz experiment	12 ± 108 (5)	227 ± 275 (5)

 Values are means ± SD, with number of animals in parentheses.

Figure 7.

 

The effect of interrupted night on the consequences of wearing diffusers or lenses (days 2–5). The interrupted night inhibits the myopic response to the negative lenses, but it does not inhibit the effect of diffusers or of positive lenses. *Significant difference occurred between the experimental group (interrupted night) and control group (normal night). *P < 0.05,*** P < 0.001. The data shown are from the experimental eyes only. Error bars, SEM.

The authors thank Jonathan Winawer for reading of the manuscript.