Mechanisms of Myopia in Cohen Syndrome Mapped to

Chromosome 8q22
Paula Summanen,
1
Satu Kivitie-Kallio,
2
Reijo Norio,
3
Christina Raitta,
1,4
and Tero Kivela
1
PURPOSE. To analyze the mechanisms of myopia in Cohen
syndrome (Mendelian Inheritance in Man [MIM] no. 216550).
METHODS. A cross-sectional study of 22 Finnish patients (age
range, 257 years) with Cohen syndrome, which maps to
chromosome 8q22, was undertaken to record cycloplegic re-
fraction, keratometry (corneal power and radius of curvature),
biometry (anterior chamber depth [ACD], lens thickness [LT],
axial [AL] and vitreal length [VL]), and Hoffer Q-modeled lens
power. These components of refraction were correlated to age
and spherical equivalent (SE) at the corneal plane. Contribu-
tion to total myopia of refractive (corneal and lenticular) and
axial components was modeled by multiple linear regression
and by estimating the effect of deviation from population mean
values.
RESULTS. The mean SE in patients with Cohen syndrome older
than 10 years was 9.35 D; the mean cylinder power, 1.70 D;
and the mean anisometropia, 0.53 D. Relative to the em-
metropic eye of a young adult, the AL and VL (mean, 23.9 and
16.6 mm, respectively) and lens power (mean, 30.30 D) were
higher in 74% and 93% of patients, respectively, and the ACD
(mean, 2.5 mm) was smaller and the LT (mean, 4.9 mm) and
corneal power (mean, 45.63 D) higher than average in all
patients. Corneal power (r 0.513, P 0.021) increased with
age, but AL and VL (P 0.46 and 0.54, respectively) and lens
power (P 0.89) did not correlate with age. The lens power
decreased with AL (r 0.564, P 0.029) and tended to
increase with corneal power (r 0.475, P 0.074). Multiple
linear regression identied AL and corneal power as indepen-
dent predictors of SE. Based on deviation from population
means, the lens power explained 55%, corneal power 23%, and
AL 22% of total myopia. ACD decreased and LT increased
markedly with age, rendering angle-closure glaucoma a possi-
bility.
CONCLUSIONS. Myopia in Cohen syndrome is mainly refractive in
type and is due to high corneal and lenticular power, which is
otherwise rare in young patients. It may be superimposed on
axial myopia, probably related to polygenic factors that deter-
mine myopia in the general population. The refractive myopia
in Cohen syndrome may result from dysgenesis and atrophy of
the cornea, ciliary body, and iris, which in turn cause iridial
and zonular laxity and spherophakia. (Invest Ophthalmol Vis
Sci. 2002;43:16861693)
T
he Cohen syndrome (Mendelian Inheritance in Man [MIM]
No. 216550) described by Cohen et al. in 1973,
1
is an
autosomal recessive disease that maps to chromosome 8q22.
2,3
It is characterized by nonprogressive mental and motor retar-
dation; a sociable and cheerful disposition; microcephaly;
4
hypotony; dysmorphic features, including wave-shaped, often
down-slanting, lid openings; thick eyebrows and eyelashes; a
short philtrum with an inability to cover the upper teeth
5
;
granulocytopenia
57
; and retinochoroidal dystrophy.
5,6,811
Myopia is another frequent hallmark of Cohen syndrome,
and it is variously considered to be either a major or a minor
diagnostic criterion (for a review, see Refs. 11,12). Of 22
Finnish patients with Cohen syndrome, all but one 5-year-old
girl had myopia at the median age of 33 years. The myopia was
often of high grade with a median spherical equivalent of 11
D.
12
In other reports, myopia has mostly been moderate, but in
four it was of high grade.
5,9,10,13
The myopia in Finnish pa-
tients was progressive, with a median increase of 6.5 D
during an average follow-up of 15 years.
12
Myopia in general can be predominantly corneal, lenticular,
or axial, or it may represent a more complex imbalance be-
tween the total refractive power and the axial length of the
eye.
14
Preliminary observations suggest that myopia in some
patients with Cohen syndrome may be predominantly corneal
and lenticular, rather than axial.
12
Because this would be a
relatively unusual combination in young-adultonset myopia,
we analyzed in detail the components of refraction with kera-
tometry and biometry in patients of various ages with Cohen
syndrome and searched for secondary changes related to axial
myopia to better understand the mechanisms of myopia in
Cohen syndrome.
PATIENTS AND METHODS
Inclusion Criteria
A nationwide survey of Cohen syndrome in Finland, organized be-
tween 1994 and 1996 by the Department of Pediatrics, Helsinki Uni-
versity Central Hospital, ascertained from hospitals, pediatricians, and
clinical geneticists 29 patients who fullled the diagnostic criteria.
5,15
Patients in this population-based cohort were eligible for participation
in the present study. Informed consent was obtained from the guard-
ians of 22 patients (inclusion ratio, 76%). The median age of the 10
enrolled male and 12 female patients was 31 years (range, 257). The
Cohen syndrome gene (COH1) had been localized by linkage disequi-
librium and haplotype analysis to chromosome 8q22 in all the patients
except one, who did not undergo genetic analysis.
3
The study followed
the tenets of the Helsinki Declaration and was approved by the insti-
tutional review board.
Ophthalmologic Examination
The patients underwent manual refraction with streak retinoscopy
while under cycloplegia induced by 0.5% cyclopentolate drops. Media
opacities precluded adequate refraction in three patients (ages, 34, 50,
and 57 years), including the two oldest ones. The corneal radius of
From the
1
Department of Ophthalmology and
2
Division of Child
Neurology, Department of Pediatrics, Helsinki University Central Hos-
pital, Helsinki, Finland; the
3
Department of Medical Genetics, The
Finnish Family Federation, Helsinki, Finland;
4
Deceased.
Supported by grants from The Eye Foundation of Finland; The Ulla
Hjelt Foundation, Finland; and the Helsinki University Central Hospital
Research Grant.
Submitted for publication May 31, 2001; revised November 9,
2001; accepted November 27, 2001.
Commercial relationships policy: N.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be marked advertise-
ment in accordance with 18 U.S.C. 1734 solely to indicate this fact.
Corresponding author: Paula Summanen, Department of Ophthal-
mology, Helsinki University Central Hospital, Haartmaninkatu 4 C, PL
220, FIN-00029 HUS, Helsinki, Finland; paula.summanen@hus..
Investigative Ophthalmology & Visual Science, May 2002, Vol. 43, No. 5
1686 Copyright Association for Research in Vision and Ophthalmology
curvature and corneal power in the two main meridians were mea-
sured by automated keratometry (RK; Canon Inc., Tokyo, Japan, 17
patients; and KM-500; Nidek Co., Gamagori, Japan, 3 patients). A-scan
biometry was performed in supine patients under cycloplegia to de-
termine the axial length, anterior chamber depth, and lens thickness
(Ultrasonic Biometer 810; Humphrey Instruments, San Leandro, CA, 18
patients; CompuScan L; Storz, St. Louis, MO, 2 patients). The mean of
two keratometry and ve biometry readings was used in statistical
analysis. No biometry (ages, 2 and 21 years) and keratometry readings
(ages 39 years) were obtained from two subjects, respectively, because
of fear of equipment and difculty in xation. A complete data set was
available for 15 patients.
Fundus changes related to axial myopia were determined in all eyes
with binocular indirect ophthalmoscopy.
14,16,17
A Goldmann three-
mirror contact lens examination was possible in 10 patients (age range,
1546 years). The presence or absence of myopic spectacle correction
before the age of 30 was determined in all parents who could be
contacted.
Statistical Methods
Computations were performed by computer (SPSS, ver 9.0.1; SPSS Inc,
Chicago, IL; Prism 3; GraphPad Software, San Diego, CA; and Stat-
Xact-3; Cytel Software, Cambridge, MA). Mean SD and median with
range are reported as summary statistics, which were calculated in
patients who were more than 10 years of age.
The age of the patient and all refractive components analyzed
fullled the assumption of normal distribution, as assessed by the
Kolmogorov-Smirnov test (Table 1).
18
Consequently, intereye correla-
tions and interrelationships between the age and refractive compo-
nents were analyzed with the parametric Pearson product moment
correlation and linear regression.
19
The intereye correlation was high
for all variables (Table 1), and one eye of each patient (except one who
had unilateral phthisis) was randomly chosen for statistical analysis on
the basis of random number tables.
Because many refractive errors were high, the refraction data were
converted to the cross-cylinder notation and vertexed to the corneal
plane.
20
The spherical equivalent and average keratometry reading
were calculated as the mean of the power in the two main meridians
and the cylinder power as the difference between powers in these
meridians. To obtain the mean cylinder power and axis, the cylinder
data were converted to the Cartesian coordinate system and a double-
angle, plus-cylinder axis plot was used to display the aggregate astig-
matism data.
20
The relative lens (center) position was calculated as the sum of the
anterior chamber depth and half of lens thickness, and the vitreous
length as the difference between the axial length and the sum of
anterior chamber depth and lens thickness. The ratio of axial length to
corneal radius of curvature was calculated. Lens power was modeled
by calculating the predicted power of an intraocular lens that would
produce the observed spherical equivalent of refraction by the Hoffer
Q theoretical formula.
21,22
The contribution to total myopia of refractive and axial compo-
nents was analyzed by two methods. The spherical equivalent was
predicted by multiple linear regression based on keratometry and
biometry readings.
19
The predicted change in spherical equivalent
caused by deviation of the corneal power and axial length from the
population means of 42.8 D and 23.5 mm, respectively, was calculated
(assuming that 0.45 mm corresponds to 1.0 D of axial myopia
14
). Any
deviation of predicted from observed spherical equivalent was attrib-
uted to lens power.
RESULTS
Cycloplegic Refraction
Of 19 patients who underwent refraction, 18 had myopia and
14 had at least 1 D of astigmatism (the latest known refraction
of the three patients with media opacities ranged from 8.5 to
13 D). The mean spherical equivalent at the corneal plane
was 9.35 3.35 D; Table 1, and the mean cylinder power
and axes of the right and left eyes were 1.71 D 107
and 1.70 D 81, respectively, vertexed to the corneal
plane (Fig. 1). The median anisometropia was 0.53 0.57 D;
Table 1.
The spherical equivalent tended to decrease (myopia
tended to increase) with age (Fig. 2A; r 0.443, P 0.057
Pearson product moment correlation), whereas the cylinder
power did not correlate with age (r 0.250, P 0.30).
Biometry and Keratometry Related to Age
The axial and vitreous lengths (Table 1) were longer than the
mean for the emmetropic eye of a young adult (23.5 and 16.2
mm, respectively)
23
in 14 of the 19 patients (74%; 95% con-
dence interval [CI], 4991) who underwent biometry and
were older than 10 years. The axial length (Fig. 2B; r 0.176,
P 0.46) and vitreous length (r 0.151, P 0.54) did not
TABLE 1. Age, Refraction, Biometry, and Keratometry in 20 Patients with Cohen Syndrome Mapped to 8q22
Descriptive Statistics
Intereye
Correlation r (P)*
Normality Test
Statistic (P) Mean SD Median (Range)
Age (y) 33 12.6 34 (1457) 0.14 (0.76)
Vertexed refraction (D)
Spherical equivalent 9.35 3.35 11.08 (2.65 to 15.92) 0.98 (0.001) 0.15 (0.78)
Anisometropia 0.53 0.57 0.36 (0.01.89) 0.23 (0.06)
Biometry (mm)
Axial length 23.9 1.38 24.0 (21.426.0) 0.82 (0.001) 0.14 (0.80)
Anterior chamber depth 2.5 0.46 2.4 (1.83.4) 0.89 (0.001) 0.15 (0.73)
Lens thickness 4.9 0.63 4.9 (4.26.0) 0.51 (0.036) 0.22 (0.29)
Relative lens position 4.9 0.32 4.9 (4.45.5) 0.65 (0.005) 0.15 (0.75)
Vitreous length 16.6 1.18 16.7 (14.518.4) 0.66 (0.004) 0.14 (0.84)
Keratometry
Mean corneal radius (mm) 7.3 0.27 7.3 (6.97.7) 0.71 (0.001) 0.15 (0.72)
Mean corneal power (D) 45.63 1.66 45.73 (43.1248.52) 0.72 (0.001) 0.15 (0.75)
Axial length/corneal radius (ratio) 3.28 0.20 3.35 (2.833.60) 0.88 (0.001) 0.15 (0.77)
Modeled lens power (D)
Hoffer Q power 30.30 4.60 28.62 (22.7537.00) 0.91 (0.001) 0.20 (0.54)
Patients were 10 years old or older. Data are for one randomly chosen eye.
*Pearson product moment correlation, two-sided.
Kolmogorov-Smirnov test, two-sided.
IOVS, May 2002, Vol. 43, No. 5 Myopia in Cohen Syndrome 1687
correlate with age. The anterior chamber depth was less and
the lens thickness more than the mean for the young adult eye
(3.6 and 3.6 mm, respectively)
23
in all 19 patients (95% CI,
82100). The anterior chamber depth decreased (Fig. 2C; r
0.727, P 0.001) and the lens thickness increased (Fig. 2D;
r 0.789, P 0.001) signicantly with age. The relative lens
position (Table 1) did not correlate with age (r 0.236, P
0.33).
The mean corneal power was higher and the radius of
curvature smaller (Table 1) than the average for the em-
metropic eye of a young adult (42.8 D and 7.79 mm, respec-
tively)
23
in all 18 patients who underwent keratometry and
were older than 10 years (100%; 95% CI, 81100). The corneal
power increased (radius decreased) with age (Fig. 2E; r
0.513, P 0.021). The ratio of axial length to corneal radius
was higher than 3.0 in all but one of 17 patients in whom it
could be calculated, and all but one of 20 patients who under-
went keratometry had more than 1.0 D of corneal cylinder
(Table 1), the mean power and axis of which were 2.57 D
93 and 3.09 D 92 in the right and left eyes, respectively
(Fig. 1). The corneal cylinder power did not correlate with age
(r 0.030, P 0.90).
The lens power modeled by the Hoffer Q formula ranged
from22.75 to 37.0 D (Table 1). It was higher than the mean
power for the emmetropic young adult lens (23.1 D)
23
in all
but one of the 14 patients in whom it could be calculated and
who were older than 10 years (93%; 95% CI, 66100). Lens
power was unrelated to age (Fig. 2F; r 0.038, P 0.89).
FIGURE 1. A double-angle, plus-cylinder plot of total (A, B) and cor-
neal (C, D) astigmatism in the right (A, C) and left (B, D) eyes of 19
Finnish patients with Cohen syndrome mapped to chromosome 8q22.
The open circle (centroid) shows the mean cylinder power and axis.
FIGURE 2. Correlation between age
and (A) vertexed spherical equiva-
lent, (B) axial length, (C) anterior
chamber depth, (D) lens thickness,
(E) average corneal power, and (F)
modeled lens power in 22 Finnish
patients with Cohen syndrome
mapped to chromosome 8q22 (two-
sided Pearson product moment cor-
relation). Lines are linear regressions
with 95% CIs. Age correlated in-
versely with anterior chamber depth
and correlated directly with lens
thickness and corneal power.
1688 Summanen et al. IOVS, May 2002, Vol. 43, No. 5
Correlation between Spherical Equivalent and
Refractive Components
The spherical equivalent decreased (myopia increased) with
increasing axial length (Fig. 3A; r 0.513, P 0.035),
vitreous length (r 0.592, P 0.016), and corneal power
(Fig. 3B; r 0.508, P 0.037). It did not correlate with
corneal cylinder power (r 0.042, P 0.86), anisometropia
(r 0.207, P 0.42), anterior chamber depth (r 0.273,
P 0.29), lens thickness (r 0.232, P 0.39), relative lens
position (r 0.158, P 0.56), and modeled lens power (r
0.329, P 0.23).
Correlation between Refractive Components
The axial length did not correlate with corneal power (Fig. 3C;
r 0.104, P 0.68), whereas the modeled lens power
decreased with increasing axial length (Fig. 3D; r 0.564,
P 0.029). The anterior chamber depth (r 0.155, P 0.51)
and the lens thickness (r 0.266, P 0.27) were also unre-
lated to the axial length. The anterior chamber depth de-
creased with increasing lens thickness (Fig. 3E; r 0.746,
P 0.001) and with decreasing corneal radius of curvature
(Fig. 3F; r 0.694, P 0.001), but it did not correlate with the
other variables studied. The lens thickness correlated inversely
with the corneal radius of curvature (Fig. 3G; r 0.542, P
0.025). The modeled lens power increased with decreasing
relative lens position (anterior shift of the lens center; r
0.621, P 0.018), and it also tended to increase with corneal
power (Fig. 3H, r 0.475; P 0.074). It did not correlate with
lens thickness (r 0.224, P 0.44).
A graph of the contribution of anterior chamber depth, lens
thickness, and vitreal-to-axial length in patients of increasing
FIGURE 3. Correlation between re-
fractive components in 22 Finnish
patients with Cohen syndrome
mapped to chromosome 8q22. (A)
Axial length and (B) average corneal
power against spherical equivalent;
axial length against (C) average cor-
neal power and (D) modeled lens
power; (E) lens thickness and (F) av-
erage corneal radius of curvature
against anterior chamber depth; (G)
lens thickness against average cor-
neal radius of curvature; and (H)
modeled lens power against average
corneal power (two-sided Pearson
product moment correlation). Lines
are linear regressions with 95% CIs.
Spherical equivalent correlated in-
versely with axial length and corneal
power, axial length with lens power,
and anterior chamber depth with
lens thickness, whereas corneal ra-
dius of curvature correlated directly
with anterior chamber depth and in-
versely with lens thickness.
IOVS, May 2002, Vol. 43, No. 5 Myopia in Cohen Syndrome 1689
age with Cohen syndrome (Fig. 4) emphasizes that age-related
change took place in the anterior segment, in which lens
thickness increased at the expense of decreasing anterior
chamber depth, without systematic change in axial or vitreous
length.
Multiple Linear Regression of
Spherical Equivalent
Multiple linear regression indicated that axial length (P
0.001) and corneal power (P 0.001) were strong indepen-
dent predictors of the spherical equivalent (Table 2). A multi-
variate model including these two variables was estimated to
explain 72% of observed spherical equivalent (R
2
0.724; Fig.
5A). The lens thickness (P 0.23), anterior chamber depth
(P 0.72), and relative lens position (P 0.25) did not
improve prediction signicantly (R
2
0.7270.785). Modeled
lens power could not be legitimately entered into the model,
because it was a derivative of the axial length and corneal
power.
Analysis of Refractive versus Axial Myopia
Analysis of deviation from the mean of normal young adults
suggested that among patients with Cohen syndrome with a
low-to-moderate grade myopia of 1 to 10 D, a short or relatively
normal axial length is associated with a disproportionately high
corneal and lens power, making the myopia mainly refractive
(Figs. 5BD). In high-grade myopia of more than 10 D, the axial
length is often moderately increased but is never compensated
for by decreased corneal and lens power. Thus, myopia was
predicted by a refractive and an axial component in this sub-
group (Figs. 5BD).
The quantitative contribution to total myopia of corneal and
lens power and axial length was modeled by calculating the
predicted change in refraction caused by deviation from mean
values for the young adult human eye (Fig. 6). In this analysis,
the mean proportion of total myopia explained by dispropor-
tionately high lens power was 55% (95% CI, 4465), whereas
disproportionately high corneal power and axial length ex-
plained 23% (95% CI, 1630) and 22% (95% CI, 1134) of total
myopia, respectively. Thus, a mean of 78% (95% CI, 6689) of
total myopia was refractive. Whereas lens and corneal power
always contributed to myopia, a short axial length decreased it
in three patients and the contribution of axial length to total
myopia was negligible (10%) in three other patients (Fig. 6).
Increased axial length accounted for a major proportion
(40%) of total myopia in four eyes (Fig. 6).
Family History of Early-Onset Myopia
Presence or absence of juvenile and young-adultonset myopia
was known in parents of 13 patients (59%) from 10 families.
The mother in one family and both parents in two families had
had myopia of 2.0 to 6.0 D since the ages of 18 to 20 years.
Of their offspring, one was estimated to have 46% axial myopia
(Fig. 6; patient 3), one had a negligible axial component (pa-
tient 4), and the axial length of the third patient could not be
reliably measured (patient 1).
Retinal Changes Related to Axial Myopia
None of the patients had any central (myopic crescent, lacquer
cracks, Fuchs spots, posterior staphyloma) or peripheral retinal
disease (lattice degeneration, white without pressure, retinal
breaks, retinal dialysis) associated with myopia. One patient
had a blind eye due to long-standing retinal detachment, attrib-
uted to injury.
DISCUSSION
Finnish patients with Cohen syndrome mapped to chromo-
some 8q22, who had had moderate to high myopia from
childhood, had eyes that showed a higher corneal and lentic-
ular power, a shallower anterior chamber, and a thicker lens
than the average eye of a young adult, but their axial length
differed little from that of an emmetrope of similar age.
14,23
Closer analysis indicated that three quarters of the myopia was
refractive, especially lenticular. Correlation between refraction
and age in cross-sectional analysis suggested that the myopia
was progressive, which has been conrmed by longitudinal
analysis.
12
Progression was mainly due to age-related increase
in corneal power. This is contrary to the rule that juvenile and
early-onset adult myopia are axial in type.
14,22
In juvenile
myopia, the corneal power is greater than average, but axial
length and anterior chamber depth should also be higher, and
the lens power should not differ from that of emmetropes.
23
In
contrast, late-adultonset myopia often is refractive.
14
Cohen
syndrome represents an unusual combination of refractive my-
opia and young age.
Frequent anterior segment abnormalities in patients with
Cohen syndrome point to the possibility that, in them, myopia
may result from dysgenesis or degeneration of the iris, ciliary
body, zonules, and lens. Iridodonesis,
12
together with in-
creased thickness, apparent spherophakia, and anterior shift of
the lens, suggests that zonular laxity and lens subluxation may
act as mediators. This may be due to generalized atrophy of the
uvea or a specic molecular defect of the zonules. In Marfan
syndrome, zonular laxity is caused by abnormal brillin, a 350
kDa glycoprotein that is a major constituent of the lens capsule
and zonules.
24
This leads to stunted lens growth, spheropha-
kia, iridodonesis, and cataracts,
24,25
which seem to be part of
Cohen syndrome as well. However, in Marfan syndrome the
TABLE 2. Multiple Linear Regression of the Spherical Equivalent in
Cohen Syndrome
Coefcient
(SE) 95% CI t P
Constant 107.81 (21.75) 60.42 to 155.20 4.96 0.001
Axial length 1.54 (0.37) 2.36 to 0.72 4.10 0.001
Mean corneal
power 1.77 (0.42) 2.68 to 0.87 4.27 0.001
R
2
0.724.
FIGURE 4. Contribution of anterior chamber depth, lens thickness,
and vitreous length to axial length in 19 Finnish patients with Cohen
syndrome mapped to chromosome 8q22. Lens thickness increased
with age at the expense of decreasing anterior chamber depth.
1690 Summanen et al. IOVS, May 2002, Vol. 43, No. 5
cornea attens rather than steepens, and frank luxation of the
lens is typical, unlike in Cohen syndrome.
12,25
Iris atrophy, which had been observed in eight of our
patients,
5,12
possibly contributes to the increased lens thick-
ness and the anterior shift of the lens diaphragm. Experiments
with surgically aniridic rhesus monkeys show that the nonac-
commodated lens becomes thicker and is shifted anteriorly if
the iris is removed and no longer in contact with the lens.
26
The investigators speculated that an intact iris diaphragm is
necessary to keep the lens in position and to atten it.
The corneal radius of curvature, which was always smaller
than average, decreased with age, and the anterior chamber
paradoxically became progressively shallower with decreasing
corneal radius. The anterior chamber depth should correlate
with the axial length,
14
but this was not true in Cohen syn-
drome. Perhaps a concomitant decrease in relative diameter of
the ciliary ring took place, which may have added to zonular
laxity and anterior shift of the lens.
27
A notable age-related
change was an increase in lens thickness, which added to
decreasing anterior chamber depth. In myopia, the anterior
chamber is usually deep,
14
but Cohen syndrome has the po-
tential to cause primary angle-closure glaucoma in high myo-
pia, as happened in both eyes of one of our patients, a 42-year-
old woman with myopia of 12.0 D.
12
Cohen syndrome shares these features with retinopathy of
prematurity (ROP), another condition characterized by fre-
quent myopia. Several studies agree that an anterior chamber
that is more shallow and a lens that is thicker than average
characterize ROP, especially if it has advanced to stage 3.
28,29
Presumably, scarring related in part to treatment causes ante-
rior shift of the lens diaphragm and relaxation of the
zonules.
29,30
However, in contrast to Cohen syndrome, no
consistent abnormality of corneal power has been docu-
mented,
29
and progression of myopia usually ends before 3
years of age.
30,31
FIGURE 5. Refractive and axial com-
ponents of myopia in 15 Finnish pa-
tients with Cohen syndrome mapped
to chromosome 8q22. (A) Observed
spherical equivalent versus that pre-
dicted by multiple linear regression
of average corneal power and axial
length. (B) Axial length and average
corneal power, (C) axial length and
modeled lens power (D), and aver-
age corneal power and modeled lens
power relative to mean values in a
young-adult emmetropic eye, divided
by grade of myopia. Corneal and lens
power were always higher than aver-
age, whereas axial length varied from
short to long.
FIGURE 6. Estimated contribution of corneal power, lens power and
axial length to the spherical equivalent in 14 Finnish patients with
Cohen syndrome mapped to chromosome 8q22. Observed spherical
equivalent is the arithmetic sum of the three elements; when all three
components contribute to myopia, the observed spherical equivalent
can be read directly.
IOVS, May 2002, Vol. 43, No. 5 Myopia in Cohen Syndrome 1691
Some patients with Cohen syndrome have had microcor-
nea,
1,5,32,33
suggesting corneal dysgenesis, and microphthal-
mia.
1,32,34
In one patient described by Cohen himself, the
microphthalmia was associated with a coloboma that involved
the anterior and posterior uvea,
1
and one patient in another
series had bilateral colobomatous microphthalmia.
32
None of
the Finnish patients had either typical microphthalmia or a
coloboma, even if they had small anterior segments.
12
It is
possible that different mutations of COH1 will be found to
account for different phenotypes of Cohen syndrome,
9,11
or
that microphthalmia is a secondary defect in Cohen syndrome.
Modeled lens power tended to be lower when axial length
was longer, suggesting some degree of emmetropization. Ap-
parently, the lens also compensated for approximately one half
of corneal astigmatism. Biomicroscopy and lens opacitometry
showed frequent incidence of early nuclear sclerosis in our
patients with Cohen syndrome.
12
The lens thickness did not
predict myopia by multiple linear regression, and lenticular
myopia in Cohen syndrome probably is related more to an
increase in the lens radius of curvature and the refractive
index.
14
Unexpectedly, lens power was unrelated to age in our
series.
By comparing components of refraction to population av-
erages, we could conclude that increased axial length was
clinically signicant in the myopia of Cohen syndrome in only
4 of 14 patients. When axial length was long, the myopia was
always high. Notwithstanding the predominance of refractive
myopia, axial length correlated with spherical equivalent, and
the axial length in addition to the corneal power was a strong
independent predictor of the spherical equivalent by multiple
linear regression. Possibly as evidence of attempted em-
metropization, shorter than average axial length compensated
for increased corneal and lenticular power in three eyes. In
most patients, the cornea was steep in relation to the axial
length, however, and caused increased myopia.
Long axial lengths can be induced in animals subjected to
visual deprivation.
35
In Cohen syndrome, vision usually re-
mains normal until the age of 10, and even when visual elds
constrict and impair distance vision, reasonable reading vision
is retained until the age of 30 to 40 years.
12
Visual deprivation
is not likely to explain the myopia of Cohen syndrome. Tape-
toretinal degeneration and chorioretinopathy associate with
axial myopia in many syndromes, such as gyrate atrophy,
36
progressive bifocal chorioretinal atrophy,
37
long-chain 3-hy-
droxyacyl-CoA-dehydrogenase deciency,
38
and several types
of retinitis pigmentosa.
39
The retinal pigment epithelial mot-
tling, choroidal atrophy, narrow retinal arterioles, and an iso-
electric electroretinogram in Cohen syndrome bear resem-
blance to symptoms in retinitis pigmentosa.
12
However, none
of our patients had retinal changes associated with axial myo-
pia, thought to be the only reliable method of distinguishing
eyes in which the posterior segment has expanded beyond that
of normal growth.
14,16
The fairly normal range of axial lengths in our series sug-
gests that elongation of the eye is probably independent of
mutated COH1 and is determined by other factors. Similarly, it
is postulated that axial myopia in premature infants with ROP
may be unrelated to retinopathy. Presumably, the refractive
error of parents inuences the axial component. This nds
some support in the fact that many siblings with Cohen syn-
drome have similar refractions.
1,10,34,4048
In our series, the
parents of two siblings who had myopia of 11 to 12 D had
myopia of 2.0 to 6.0 D, and one of the siblings had notable
axial myopia. The refraction of the parents of other children
who had notable axial myopia remained unknown. A family
has been reported in which the child had myopia of 14.0 D
and the parents of 5.0 and 7.0 D.
13
Intuitively, parental
hyperopia may lead conversely to a shorter than average axial
length and may counterbalance myopia in Cohen syndrome.
Hypermobility of joints, skeletal and postural anomalies
such as long and slender hands and ngers, cubitus valgus,
genu valgum, pes planovalgus, scoliosis, and kyphoscoliosis are
common in Cohen syndrome, and mitral valve prolapse and
decreased left ventricular function are occasionally ob-
served.
15,34
These features hint that there is a generalized
disorder of connective tissue in Cohen syndrome.
34,49,50
Myo-
pia and cataract are common to several collagen mutations,
such as Stickler syndrome, Wagner disease, and Kniest dyspla-
sia.
51,52
However, the myopia in these connective tissue syn-
dromes is axial in type and is often associated with an in-
creased risk of retinal detachment,
53
which is not the case in
Cohen syndrome.
12
All evidence taken together, the myopia in Finnish patients
with Cohen syndrome mapped to chromosome 8q22 is pre-
dominantly refractive and due to a disproportionately high
power of the cornea and the lens, but it may be modulated by
an axial component that is independent of COH1, the Cohen
syndrome gene. The refractive components in Cohen syn-
drome resemble those in late-adultonset myopia,
14
a fact that
may be related to dysgenesis, degeneration, and premature
aging of the anterior segment of the eye in this syndrome.
Future identication of the COH1 gene is likely to reveal a
specic molecular defect that can lead to nonaxial high myo-
pia. Regardless of its exact pathogenesis, correction of myopia
is indicated in Cohen syndrome. Our patients enjoyed wearing
spectacles as a rule, and correction of myopia and astigmatism
affected their development positively.
12,15
Acknowledgments
The authors thank the patients and their parents and guardians for their
collaboration in making this study possible and the Department of
Ophthalmology, University of Oulu, for providing the facility for ex-
amination of three of the patients.
References
1. Cohen MM Jr, Hall BD, Smith DW, Graham CB, Lampert KJ. A new
syndrome with hypotonia, obesity, mental deciency, and facial,
oral, ocular, and limb anomalies. J Pediatr. 1973;83:280284.
2. Tahvanainen E, Norio R, Karila E, et al. Cohen syndrome gene
assigned to the long arm of chromosome 8 by linkage analysis. Nat
Genet. 1994;7:201204.
3. Kolehmainen J, Norio R, Kivitie-Kallio S, Tahvanainen E, de la
Chapelle A, Lehesjoki AE. Rened mapping of the Cohen syn-
drome gene by linkage disequilibrium. Eur J Hum Genet. 1997;5:
206213.
4. Kivitie-Kallio S, Autti T, Salonen O, Norio R. MRI of the brain in the
Cohen syndrome: a relatively large corpus callosum in patients
with mental retardation and microcephaly. Neuropediatrics. 1998;
29:298301.
5. Norio R, Raitta C, Lindahl E. Further delineation of the Cohen
syndrome: report on chorioretinal dystrophy, leukopenia and con-
sanguinity. Clin Genet. 1984;25:114.
6. Warburg M, Pedersen SA, Horlyk H. The Cohen syndrome: retinal
lesions and granulocytopenia. Ophthalmic Pediatr Genet. 1990;
11:713.
7. Kivitie-Kallio S, Rajantie J, Juvonen E, Norio R. Granulocytopenia in
Cohen syndrome. Br J Hematol. 1997;98:308311.
8. Resnick K, Zuckerman J, Cotlier E. Cohen syndrome with bulls
eye macular lesion. Ophthalmic Pediatr Genet. 1986;7:18.
9. Kondo I, Nagataki S, Miyagi N. The Cohen syndrome: does mottled
retina separate a Finnish and a Jewish type? Am J Med Genet.
1990;37:109113.
10. Fryns JP, Legius E, Devriendt K, et al. Cohen syndrome: the clinical
symptoms and stigmata at a young age. Clin Genet. 1996;49:237
241.
1692 Summanen et al. IOVS, May 2002, Vol. 43, No. 5
11. Horn D, Krebsova A, Kunze J, Reis A. Homozygosity mapping in a
family with microcephaly, mental retardation, and short stature to
a Cohen syndrome region on 8q21.38q22.1: redening a clinical
entity. Am J Med Genet. 2000;92:285292.
12. Kivitie-Kallio S, Summanen P, Raitta C, Norio R. Ophthalmologic
ndings in Cohen syndrome: a long-term follow-up. Ophthalmol-
ogy. 2000;107:17371745.
13. O

zturk B, Weber HP. Cohen-Syndrom. Eigene Beobachtung und

Literaturubersicht. Monatsschr Kinderheilkd. 1991;139:844848.
14. Curtin BJ. The Myopias. Basic Science and Clinical Management.
Philadelphia: Harper & Row; 1985.
15. Kivitie-Kallio S. Cohen Syndrome: A Clinical Study of 29 Finnish
Patients. Helsinki: University of Helsinki; 1999:166 http://
ethesis.helsinki./julkaisut/laa/kliin/vk/kivitie-kallio/. Thesis.
16. Karlin DB, Curtin BJ. Peripheral chorioretinal lesions and axial
length of the myopic eye. Am J Ophthalmol. 1976;81:625635.
17. Gozum N, Cakir M, Gucukoglu A, Sezen F. Relationship between
retinal lesions and axial length, age and sex in high myopia. Eur J
Ophthalmol. 1997;7:277282.
18. Altman DG. Practical Statistics for Medical Research. London:
Chapman & Hall; 1991.
19. Fox J. Applied Regression Analysis, Linear Models, and Related
Methods. Thousand Oaks, CA: SAGE Publications; 1997.
20. Holladay JT, Dudeja DR, Koch DD. Evaluating and reporting astig-
matism for individual and aggregate data. J Cataract Refract Surg.
1998;24:5765.
21. Hoffer KJ. The Hoffer Q formula: a comparison of theoretic and
regression formulas (published correction J Cataract Refract
Surg.1994;20:677) J Cataract Refract Surg. 1993;19:700712.
22. Fledelius HC. Adult onset myopia-oculometric features. Acta Oph-
thalmol Scand. 1995;73:397401.
23. Grosvenor T, Scott R. Three-year changes in refraction and its
components in youth-onset and early adult-onset myopia. Optom
Vis Sci. 1993;70:677683.
24. Mir S, Wheatley HM, Hussels IE, Whittum-Hudson JA, Traboulsi EI.
A comparative histologic study of the brillin microbrillar system
in the lens capsule of normal subjects and subjects with Marfan
syndrome. Invest Ophthalmol Vis Sci. 1998;39:8493.
25. Maumenee IH. The eye in the Marfan syndrome. Birth Defects.
1982;18:515524.
26. Crawford KS, Kaufman PL, Bito LZ. The role of the iris in accom-
modation of rhesus monkeys. Invest Ophthalmol Vis Sci. 1990;31:
21852190.
27. Strenk SA, Semmlow JL, Strenk LM, Munoz P, Gronlund-Jacob J,
DeMarco JK. Age-related changes in human ciliary muscle and
lens: a magnetic resonance imaging study. Invest Ophthalmol Vis
Sci. 1999;40:11621169.
28. Kent D, Pennie F, Laws D, White S, Clark D. The inuence of
retinopathy of prematurity on ocular growth. Eye. 2000;14:2329.
29. Choi MY, Park IK, Yu YS. Long term refractive outcome in eyes of
preterm infants with and without retinopathy of prematurity:
comparison of keratometric value, axial length, anterior chamber
depth, and lens thickness. Br J Ophthalmol. 2000;84:138143.
30. Quinn GE, Dobson V, Siatkowski R, et al. Does cryotherapy affect
refractive error? Results from treated versus control eyes in the
cryotherapy for retinopathy of prematurity trial. Ophthalmology.
2001;108:343347.
31. Quinn GE, Dobson V, Kivlin J, et al. Prevalence of myopia between
3 months and 5 1/2 years in preterm infants with and without
retinopathy of prematurity. Cryotherapy for Retinopathy of Pre-
maturity Cooperative Group. Ophthalmology. 1998;105:1292
1300.
32. Carey JC, Hall BD. Conrmation of the Cohen syndrome. J Pediatr.
1978;93:239244.
33. Moreno-Montanes DJ, Garcia GJ, Barrera VV, Palomar GA. Une
nouvelle observation du syndrome de Cohen. J Fr Ophtalmol.
1988;11:197200.
34. Friedman E, Sack J. The Cohen syndrome: report of ve new cases
and a review of the literature. J Craniofac Genet Dev Biol. 1982;
2:193200.
35. Wiesel TN, Raviola E. Increase in axial length of the macaque
monkey eye after corneal opacication. Invest Ophthalmol Vis Sci.
1979;18:12321236.
36. Peltola KE, Nanto-Salonen K, Heinonen OJ, et al. Ophthalmologic
heterogeneity in subjects with gyrate atrophy of choroid and
retina harboring the L402P mutation of ornithine aminotransfer-
ase. Ophthalmology. 2001;108:721729.
37. Godley BF, Tifn PA, Evans K, Kelsell RE, Hunt DM, Bird AC.
Clinical features of progressive bifocal chorioretinal atrophy: a
retinal dystrophy linked to chromosome 6q. Ophthalmology.
1996;103:893898.
38. Tyni T, Kivela T, Lappi M, Summanen P, Nikoskelainen E, Pihko H.
Ophthalmologic ndings in long-chain 3-hydroxyacyl-CoA dehy-
drogenase deciency caused by the G1528C mutation: a new type
of hereditary metabolic chorioretinopathy. Ophthalmology. 1998;
105:810824.
39. Horneber M, Gottanka J, Milam AH, Lutjen-Drecoll E. Alterations in
anterior segment dimensions in eyes with retinitis pigmentosa.
Graefes Arch Clin Exp Ophthalmol. 1996;234:7178.
40. Kousseff BG. Cohen syndrome: further delineation and inheri-
tance. Am J Med Genet. 1981;9:2530.
41. Doyard P, Mattei JF. Syndrome de Cohen chez deux soeurs. Ann
Pediatr. 1983;30:777781.
42. Ferre P, Fournet JP, Courpotin C. Le syndrome de Cohen, une
affection autosomique recessive? Arch Fr Pediatr. 1982;39:159
160.
43. Goecke T, Majewski F, Kauther KD, Sterzel U. Mental retardation,
hypotonia, obesity, ocular, facial, dental, and limb abnormalities
(Cohen syndrome): report of three patients. Eur J Pediatr. 1982;
138:338340.
44. North C, Patton MA, Baraitser M, Winter RM. The clinical features
of the Cohen syndrome: further case reports. J Med Genet. 1985;
22:131134.
45. Zetler S, Romke C, Aksu F. Cohen-Syndrom bei zwei Brudern. Klin
Padiatr. 1987;199:5557.
46. Arcas MJ, Garcia PJJ, Ramos LJ, Diaz GC, Pascual CI. Sindrome de
Cohen: presentacion de dos casos de gemelas. Anal Esp Pediatr.
1991;34:8385.
47. Steinlein O, Tariverdian G, Boll HU, Vogel F. Tapetoretinal degen-
eration in brothers with apparent Cohen syndrome: nosology with
Mirhosseini-Holmes-Walton syndrome. Am J Med Genet. 1991;41:
196200.
48. North KN, Fulton AB, Whiteman DA. Identical twins with Cohen
syndrome. Am J Med Genet. 1995;58:5458.
49. Mehes K, Kosztolanyi G, Kardos M, Horvath M. Cohen syndrome:
a connective tissue disorder? Am J Med Genet. 1988;31:131133.
50. Kivitie-Kallio S, Eronen M, Lipsanen-Nyman M, Marttinen E, Norio
R. Cohen syndrome: evaluation of its cardiac, endocrine and ra-
diological features. Clin Genet. 1999;56:4150.
51. Wilkin DJ, Mortier GR, Johnson CL, et al. Correlation of linkage
data with phenotype in eight families with Stickler syndrome.
Am J Med Genet. 1998;80:121127.
52. Wilson MC, McDonald-McGinn DM, Quinn GE, et al. Long-term
follow-up of ocular ndings in children with Sticklers syndrome.
Am J Ophthalmol. 1996;122:727728.
53. Robertson JE, Meyer SM. Hereditary vitreoretinal degenerations.
In: Ogden TE, eds. Basic Science and Inherited Retinal Disease.
St. Louis: CV Mosby; 1989:469479.
IOVS, May 2002, Vol. 43, No. 5 Myopia in Cohen Syndrome 1693