This study examined the refractive and biometric profiles of a
group of albino subjects that included both oculocutaneous and ocular
subtypes. We found no significant differences between the individual
groups for the parameters examined. As a group, these albino subjects
exhibited a bias toward hyperopia in their refractive errors, although
there was significant variability among them, with high myopia as well
as high hyperopia being encountered. These subjects also exhibited
abnormally high levels of refractive astigmatism that was consistently
with-the-rule. The following questions arising from these data are
taken up in the following: 1. How representative are these data of
albino populations? 2. What is the significance of these results in the
context of emmetropization and ocular growth regulation?
How do these data compare with previous studies of albinos? Although it
is typically reported that albinos are myopic, the opposite trend was
seen in the present study, with hyperopia being more common than
myopia. A similar trend is evident in a related study by Dickerson and
Abadi
10 ; although no mean data are provided, analysis of
data shown graphically revealed 15 of their 25 subjects to be
hyperopic, with 3 cases having mean refractive errors greater than 5 D.
Hyperopic biases also appear to predominate overall (see
Table 4 ). Nonetheless, high myopia is generally represented among refractive
errors greater than 5 D, contributing to the typically large SDs
reported in studies of albino subjects. In relation to the effect of
albino subtype on refractive status, Kasmann and
Ruprecht
23 describe a loose, albeit not statistical
significant, association between myopia and Ty− albinism, and
hyperopia and Ty+ albinism; this trend is not borne out by our data,
which include an disproportionate number of high hyperopes to high
myopes in the Ty− group (3:1); also, in our study, refractive error
differences between the various groups were not statistically
significant. Reported means for refractive astigmatism range from 1.07
D
22 to 2.58 D,
12 with the prevalence of
with-the-rule astigmatism ranging from 50%
22 to
100%.
3 19 These data are comparable with our findings of
2.37 D for mean refractive astigmatism and 100% for the prevalence of
with-the-rule refractive astigmatism.
The present study included a biometric investigation of the components
contributing to observed refractive errors. The observed high
correlation between “best sphere” refractive errors and axial
lengths implies an axial origin to the refractive errors. Although this
conclusion is at odds with that drawn by Harris and
Heyman
24 based on more limited data, it is also indirectly
supported by our corneal data. Specifically, it was found that the
corneal radius of curvature increased in parallel with eye size; this
trend is opposite to that required for a corneal contribution to
refractive errors.
25
What do refractive data indicate in terms of emmetropization and ocular
growth regulation in albino subjects? As already noted, high
with-the-rule astigmatism was characteristic of this group, and as
infantile astigmatism tends to show an against-the-rule bias (see
review by Lyle
15 ), this argues against it being a product
of arrested emmetropization. That visual acuity had no bearing on the
magnitude of astigmatism (
r = 0.01,
P > 0.05) also supports this conclusion. However, that with-the-rule
astigmatism is also characteristic of idiopathic
nystagmus
10 20 raises the alternative possibility of an
etiologic link with nystagmus. The suggestion by
Grosvenor
26 among others that corneal molding by the lids
might give rise to astigmatism offers a potential explanation here, if
the effect of the accompanying nystagmus is to lower corneal rigidity,
thereby rendering it more moldable. The predominantly corneal
origin
20 of the refractive astigmatism in these cases is
consistent with this interpretation. Other indirect support for this
interpretation is contained in studies showing acute lid-induced
changes in astigmatism: Gray and Yap
27 report an increase
in with-the-rule refractive astigmatism with narrowing of the palpebral
aperture, whereas Masci
28 and Wilson et al.
29 report that lifting the lid decreases with-the-rule astigmatism.
Although demonstration of a reduction in astigmatism with lid
retraction in nystagmus subjects would provide direct proof of the
above hypothesis, their astigmatism is likely to be less reversible
because of the chronic nature of the condition.
30
What do the spherical refractive errors indicate about emmetropization
in albinos? If the congenital nature of albinism were to preclude any
emmetropization, i.e., arrest development, one might predict a
refractive error distribution not unlike those reported for neonates
(e.g., Cook and Glasscock
31 ). This prediction is at least
partly borne out by our refractive data, which have a broad
distribution and include both high hyperopia and high myopia. Indeed,
32% of subjects had refractive errors greater than 5 D, which may have
been present neonatally. Also overall, our subjects displayed a
hyperopic bias as characteristic of normal infants. This bias also is
consistent with the more general association made by Nathan et
al.
12 between hyperopia and visual impairments developing
within the first 3 years of life.
In the absence of longitudinal data, it is impossible to exclude the
possibility that emmetropization was disrupted rather than arrested.
Indeed, visual acuity is reported to be near normal in albino infants
over the first 12 months of life,
19 and thus some
emmetropization might be expected. This situation can be expected to
change as visual acuity subsequently declines. Furthermore, that higher
refractive errors were associated with greater visual impairment
(poorer visual acuity) is consistent with greater disruption of
emmetropization in these cases. Although there is on-going debate over
to what extent the reduced visual acuity reflects the underlying
pathology versus nystagmus and/or amblyopia
4 32 33 34 ;
nonetheless, the lower than normal cone density in the foveae of albino
eyes
35 alone is likely to increase the eye’s depth of
focus and so impair emmetropization. The generalized nature of the
ocular hypopigmentation problem in albinism and the further
observation that rods are more affected than cones
36 implies that the peripheral retina is also abnormal, although the
influence of latter on emmetropization for humans is currently not well
understood. In this context, the significance of the apparent disparity
between albinism, where hyperopia predominates, and other conditions
encompassing either peripheral or peripheral plus central anomalies,
where myopia predominates,
12 is unclear, although it adds
weight to the case for arrested development in albinism.
The preceding discussion and conclusion that emmetropization is either
arrested or impaired in albinism is based on considerations of best
sphere (average) refractive error data. However, a closer inspection of
these refractive data in terms of their meridional components raises
the possibility of “meridional emmetropization” for our hyperopic
subjects. Specifically for this group, the mean refractive error for
the vertical meridian, while still hyperopic, was closer to emmetropia
than was that for the horizontal meridian. Refractive data from a
related study of albino subjects by Dickerson and Abadi
10 show a similar trend. To evaluate the plausibility of this hypothesis
and understand why myopic subjects do not follow the same trend, one
needs to consider first, the effect of nystagmus on vision and second,
the growth processes underlying emmetropization.
Consider first, the effect of nystagmus on vision. Although not the
primary limiting factor of visual acuity in albinos, reductions in
visual acuity and contrast sensitivity attributable to nystagmus have
been documented.
8 32 It is the meridional differences in
the influence of nystagmus, as implied by meridional performance
differences
3 4 that is of greatest relevance to the issue
at hand. Specifically, if as to be expected, horizontal details (e.g.,
vertically elongated objects) are degraded (smeared), whereas vertical
details are relatively well preserved, then one might also predict
emmetropization to be preserved for the vertical meridian. Hyperopic
albinos appear to follow this prediction. Inherent in this
interpretation are two necessary assumptions: (1) that the observed
refractive astigmatism is not in itself a product of emmetropization (a
point taken up again later) and (2) that emmetropizaton dominates over
any influence of image degradation on the other meridian. Thus, in this
model, the refractive error of the orthogonal (horizontal) meridian is
not directly regulated but will be “dragged along” with the other
meridian, with a superimposed influence of corneal molding. As a
proviso here, it should be noted that only evidence of active
emmetropization in the form of longitudinal refractive data can make
the distinction between the model described here and the alternative
possibility that the observed refractive pattern is an artifact of the
combination of hyperopia and with-the-rule astigmatism.
Neither the myopic subjects described in the study by Dickerson and
Abadi
10 nor those from the present study showed the“
meridional emmetropization” ascribed to hyperopic subjects.
Although one cannot rule out the possibility that the myopic subjects
represent a separate subgroup in which there are other, unidentified
factors at work, their profiles also can be explained in terms of the
mechanisms underlying emmetropization. The potentially significant
difference between hyperopic and myopic eyes in relation to
emmetropization is that the former must increase eye growth, whereas
the latter must slow their growth. Consequently, the capacity to
emmetropize is limited for hyperopic eyes only by the capacity of eyes
to grow, but is limited for myopic eyes by the capacity of corneas to
flatten after eye growth has ceased.
13 25 If the high
with-the-rule corneal astigmatism encountered in albino eyes implies
that the amount of developmental flattening is restricted specifically
in the vertical meridian, then this would have the effect for myopic
subjects of precluding “meridional emmetropization.”
As an aside to the preceding discussion, an underlying assumption
throughout has been that the observed with-the-rule astigmatism occurs
independently of, and is neither correctable by, or a product of,
emmetropization. Direct tests of this assumption through animal studies
involving imposed astigmatic errors are generally
supportive.
37 38 39 40 41 42
Finally, in the context of emmetropization, the significance of the
high interocular correlations that were observed for various ocular
parameters warrants some consideration. These relationships imply that
whatever the influences on eye growth in these albino subjects, the two
eyes of individual subjects are similarly affected. The high
correlation between the two eyes in terms of visual acuity suggests
that the severity of the underlying pathology, which is presumably
genetically determined, is a significant contributing factor. However,
if the properties of the emmetropization mechanism, e.g., its gain, are
also genetically determined,
43 then as observed, similar
refractive outcomes for the two eyes can be expected where there is
high symmetry in the “perturbing” pathology.
In conclusion, the refractive profile of albino subjects, which is
typified by high refractive errors with an overall bias toward
hyperopia, as well as high with-the-rule astigmatism, suggests that
normal emmetropization is impaired. However, the data for hyperopic
eyes open the further possibility that some capacity for
emmetropization is retained in the vertical meridian, where visual
function is likely to be less affected by the accompanying nystagmus.
That myopic eyes do not follow the same pattern may simply reflect
differences in operating physical constraints.