Classification of Human Senile Cataractous Change by

the American Cooperative Cataract Research Group

(CCRG) Method: III. The Association of Nuclear Color
(Sclerosis) with Extent of Cataract Formation,
Age, and Visual Acuity
Leo T. Chylack, J r . / f i Bernard J. Ransil,§ and Ophelia Whiref

Nineteen hundred and seventy-six immature human cataracts extracted intracapsularly were classified
according to the Cooperative Cataract Research Group (CCRG) method of cataract classification. Data
on cataract location and extent, nuclear color, preoperative visual acuity, age, and sex were organized
and stored in the PROPHET system. The data were examined for relationships between nuclear color
(sclerosis) and the age of the cataractous lens, the extent of opacification in seven anatomical regions
including the degree of nuclear opacification and the preoperative visual acuity. Nuclear color correlates
with age in a curvilinear manner. Nuclear yellowing increases gradually with increasing nuclear opa-
cification, but the color change is so slight as to be useless for the purposes of inferring the intensity
of nuclear opacification from the color of the nucleus. There is no association between the extent of
anterior cortical, equatorial cortical, posterior cortical, subcapsular or supranuclear opacification, and
nuclear color. Nuclear color impairs vision only for the range dark yellow through black. These data
justify the recommendation that nuclear color be abandoned as the single index of the severity of any
type of senile cataractous change. Invest Ophthalmol Vis Sci 25:174-180, 1984

In Parts I and II of this series of manuscripts,1'2 a
system for classifying cataractous change in human
lenses was described that differs significantly from other
systems of classification3"6 in which nuclear color is
used as the single or principal measure of the extent
of cataract formation. In order to estimate nuclear
color accurately, previous investigators 3 " 6 photo-
graphed extracted human lenses against a pure white
From the Howe Laboratory of Ophthalmology, Harvard Medical
School,* the Massachusetts Eye and Ear Infirmary, f the Brigham
and Women's Hospital,| the Charles A. Dana Research Institute
and the Harvard Thorndike Laboratory, Department of Medicine,
the Beth Israel Hospital and Harvard Medical School,§ Boston, Mas-
sachusetts.
Supported in part by NIH grants EY01276, EY03247, EY07063,
grant-RR-01032 from the General Clinical Research Centers Program
of the Division of Research Resources, NIH, and the Brigham Surgical
Group Foundation.
Data organization and analyses were performed on the PROPHET
system, a national computer resource sponsored by the Division of
Research Resources, NIH. BMDP programs on PROPHET originate
from the Department of Biomathematics, UCLA and were converted
for use on PROPHET by the University of Pittsburgh Computer
Center.
Submitted for publication: May 24, 1982.
Reprint requests: Leo T. Chylack, Jr., MD, Howe Laboratory,
243 Charles Street, Boston, MA 02114.
background or against a black and white grid.7 As we
have shown,' these lighting conditions, while enhancing
appreciation of nuclear color, often actually obscure
the features of the cataract itself. Although these ob-
jections to the photographic techniques of references3""6
are valid, they do not disqualify the use of nuclear
color as a measure of cataract extent. It is the purpose
of this study to demonstrate that while: nuclear color
correlates significantly with increasing age, the cor-
relation is not predictive for a given individual such
that the shade of nuclear color can be used as a measure
of the severity or the extent of cataract formation.
Large changes in the extent of cortical or the degree
of nuclear opacification can occur with only small
changes in nuclear color. These data suggest that con-
tinued reliance on nuclear color as an index of cataract
extent may result in misleading and possibly incorrect
correlations between laboratory data and cataract type.
We recognize that "nuclear sclerosis" is a misnomer
when used to designate nuclear color. However, it is
a widely used, clinically accepted, misnomer and was
used in other attempts at cataract classification (3-6)
as a measure of nuclear yellowing or brunescence. In
this paper we intend "nuclear sclerosis'" (NS) to mean
nuclear yellowing or nuclear color, and do not intend

174

Downloaded From: https://iovs.arvojournals.org/ on 07/27/2018

 No. 2 CATARACT CLASSIFICATION: III. NUCLEAR SCLEROSIS/Chylock er ol. 175

it to mean nuclear hardening. We are fully cognizant MEAN NS INDEX US AGE

of the objections to its use in this manner and wish

not to enter into or further a semantic controversy.
We use it simply to facilitate communication and to
provide a sense of continuity with earlier efforts at
cataract classification.3"6

Materials and Methods 4. ••

Data on cataract location and extent (CCRG clas-

sification), nuclear color (nuclear sclerosis—NS), pre-
operative visual acuity (VA), age and sex for 1976
immature cataracts extracted intracapsularly at the
Massachusetts Eye and Ear Infirmary (MEEI) and the
Brigham and Women's Hospital (BWH) were orga-
nized and stored in the PROPHET system.8 Appro-
priate univariate and multivariate techniques were
employed to look for relationships between the nuclear
sclerosis index (NS = 1 to 8; NSi = colorless; NS 2
= very pale yellow; NS 3 = pale yellow; NS 4 = yellow;
NS 5 = dark yellow; NS 6 = very dark yellow; NS 7
= brown; NS 8 = black) and (1) the age of all lenses
with immature cataract, (2) the degree of nuclear opa-
cification (N = 1 to 4) in pure nuclear and mixed
corticonuclear (CXN), cataracts, (3) the extent of an-
terior cortical (CXA), equatorial cortical (CXE), pos-
terior cortical (CXP), supranuclear (SN), anterior sub-
capsular (SCA), and posterior subcapsular (SCP) opa-
cification in cataracts without nuclear (N) involvement,
and (4) the preoperative visual acuity. The methods
of measuring NS, cataractous involvement, and pre-
operative visual acuity are described in detail in ref-
erence 1.
Results
Figure 1 is a plot of the mean NS index calculated
for each year versus the age of the lens in years. The
20. 40. 60. 80. 180.
AGE
Fig. 1. Mean NS index of 1976 patients with immature cataracts
at the time of cataract extraction as a function of the annual age of
the subject. (NS| = colorless, NS2 = very pale yellow, NS 3 = pale
yellow, NS4 = yellow, NS 5 = dark yellow, NS6 = very dark yellow,
NS7 = brown, NS 8 = black).
plot shows that mean NS index increases in a positive
curvilinear fashion as a function of age in the form of
a growth curve, which may be S-shaped. The most
reliable points (30 to 100 subjects) occur in the middle
of the curve. The least reliable (5 to 10 subjects) occur
at both ends, giving more scatter. This finding is in
sharp contrast to that of Pirie1 in which no association
between nuclear color and patient age was detected,
but is consistent with the findings of Lerman and
Borkman.9
Figure 2 is a plot of the mean NS index versus the
degree of nuclear opacification, as denoted by N = 1
(faint nuclear opacification) to N = 4 (total nuclear
opacification), for all cataracts (solid circle), pure nu-
clear (N), and mixed cortico-nuclear cataracts. Figure

Mean NS vs. N, 1976

CO

Fig. 2. Mean NS index as a function of the severity of

nuclear opacification (N|_4) in pure nuclear (N), mixed
(CXN) cortico-nuclear and all (O) cataracts. 1976 immature lU
cataracts formed the basis for this graph.
U CORTICAL + N
• PURE N
O ALL CATARACTS

N - INDEX

Downloaded From: https://iovs.arvojournals.org/ on 07/27/2018

 176 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / February 1984 Vol. 25

Mean N vs. NS, I976

Fig. 3. Mean N index (intensity of nuclear haze, N o =

clear; N, = trace haze; N 2 = moderate haze; N 3 = intense
haze; N4 = opaque nucleus) as a function of nuclear color
Ui in pure nuclear (N), mixed cortico-nuclear (CXN) and all
(O) cataracts. 1976 immature cataracts formed the basis for
• CORTICO-NUCLEAR this graph.
A PURE N
O ALL CATARACTS

3 4 5
NS INDEX

Table 1. NS vs N (frequency analysis) 3 plots the converse relationship, mean nuclear index
versus NS index, for the same populations.
Both the NS and N indices are ordinal scale data,
A. FREQUENCY (NUMBER):
4 8 TOTAL
where the scales are nonlinear with respect to nuclear
N~-—-NS 2 3 5 6 7
color and degree of opacification, respectively. Con-
1 52 109 174 127 25 14 0 501 sequently, the mean indices do not represent reliable
2 32 103 219 269 99 51 3 776 central measures of underlying normal distributions;
3 9 35 75 75 34 52 1 281 the consequence of this is that the two plots do not
4 13 25 22 22 16 45 3 146 coincide (as they should if the N and NS populations
were normally distributed) when scaled appropriately
TOTAI 106 272 490 493 174 162 7 17D4
and overlaid.
Despite the shortcomings of the measures, both plots
B. RELATIVE FREQUENCY (X): tell us that NS on the average increases with increasing
nuclear opacification (Fig. 2) and vice versa (Fig. 3).
1 49 40 36 26 14 9 0 But while nuclear opacification increases with nuclear
2 30 38 45 55 57 31 43 sclerosis over the range, especially for NS > 5, the
3 9 13 15 15 20 32 14 relationship is of little clinical utility, because the net
4 12 9 4 4 9 28 43 change in mean nuclear opacification is too small and
MN INDEX 1.84 1.91 1.89 1.98 2.24 2,79 3.00 only 2 units or less over the entire NS range are of no
predictive value for a specific individual or group of
individuals. It is, therefore, not possible to infer ac-
C. CUMULATIVE FREQUENCY ( X ) : curately the intensity of nuclear opacification from an
estimate of nuclear color. The plots tell us further that
1 49 40 36 26 14 9 0 the pure nuclear cataracts show more opacification
79 78 81 81 71 40 43 and less nuclear sclerosis on the average than the mixed
IS)

3 88 91 96 96 91 72 57 cataracts.
4 100 100 100 100 100 100 100 The data of Figure 3 for the CXN cataracts are
AREA UNDER presented in tabular form (Table 1) as frequency values,
CUMULATIVE 242* 239' 245' 240" 219* 166* 150'
FREQUENCY together with their corresponding relative frequency
and cumulative frequency values. The mean N index
Frequency data of Figure 3 in tabular form of frequency, relative frequency
plotted in Figure 3 for each group of cataracts (CXN,
and cumulative frequency from which the mean N index is derived for plotting. pure N, total) is readily computed from the appropriate
If the cumulative frequency (%) is plotted against N (1-4) and the area under
the curve is calculated, one gets a more statistically rigorous measure of the
relative frequency tables by multiplying each index by
association between the two variables N and NS. It was this measure that was its appropriate relative frequency and summing.
used to calculate the significance of the differences between the groups of
corticonuclear, pure nuclear, and all cataracts (differences were not significant
Table 2 is a typical contingency table for CXA extent
at any point at P < 0.05). versus the NS index, for all lenses without nuclear

Downloaded From: https://iovs.arvojournals.org/ on 07/27/2018

 No. 2 CATARACT CLASSIFICATION: III. NUCLEAR SCLEROSIS/Chylock er ol.
involvement (N = 0). In terms of frequencies, it shows
how the extent of CXA opacification changes with the
NS index. A mean CXA index may be derived for
each NS index value from the corresponding relative
frequency table (not shown) as described above and
the results plotted as in Figure 4a to determine if mean
extent of cataract involvement correlates with NS.
Similar tables were constructed for CXE, CXP, SN,
SCA, and SCP (Figs. 4b-f). In only two cases, Figures
4a and f (CXA and SCP vs. NS) are mean extent and
NS positively correlated. However, in neither case is
the correlation statistically significant. In the remaining
regions, there is either no relationship between the
mean extent of opacification and the NS index, or the
mean extent of opacification decreases, on the average,
with increasing NS. This may in part be due to the
obscuration of nuclear color by extensive cortical
opacification, much like that which occurs in mature
(M) and hypermature (H) cataracts, and that precludes
an estimate of NS index in these cataracts. The results
demonstrate conclusively that NS is not a measure of
the extent of cataract involvement in this population
of cataracts.
In Figure 5, the mean VA index of the operated eye
measured prior to cataract extraction is plotted against
the NS index. The VA index denotes a scale that is
both quantitative and qualitative, both ordinal and
nominal. It is neither linear with respect to what it
measures nor normally distributed. Consequently, as
for the N and NS indices, a mean VA value is not a
reliable central measure and should not be interpreted
as such. We observe that the mean VA index exhibits
a gradual linear increase (ie, visual acuity deteriorates)
as NS progresses from pale yellow through deepening
shades of yellow (NS = 2 through 6). As NS progresses
to brown and black (NS = 7, 8), the deterioration in
acuity accelerates. From this behavior, we infer that
nuclear color significantly impairs vision only for large
NS index values (dark yellow through black).
These analyses have been confined to immature cat-
aracts; in mature (M) and hypermature (H) cataracts,
it is not possible to assess nuclear color accurately even
against a bright white background due to the excessive
scatter of light from the opacified envelope of lens
cortex. Intact M and H cataracts that appear to be
pale yellow in color may actually have dark brown
nuclei. This is apparent when the nucleus is dissected
out of the lens.
Discussion
The degree of nuclear yellowing or browning has
served as a measure of the severity of cataract formation
for many scientists3"6 in the European research corn-
Downloaded From: https://iovs.arvojournals.org/ on 07/27/2018
177
Table 2. Contingency table of CXA extent versus NS
index, controlled for CXA = 1 to 4 and N = 0
NS INDEX
CXA EXTENT 1 2 3 4 5 6 7 8
TOTALS
1 1 7 12 11 10 3 0 0 44
2 0 5 10 6 3 1 2 0 22
3 0 1 1 1 0 0 0 0 3
4 0 3 8 7 2 3 1 0 24
TOTALS 1 16 31 25 15 7 3 0 98
Contingency table of severity of anterior cortical opacification (CXA) versus
the NS index in a population of pure cortical (but not pure CXA) cataracts.
Similar tables were constructed but not presented here for the other anatomical
regions of opacification.
munity and at least one7 member of the American
community. In spite of Pirie's3 original recommen-
dation that the nuclear color scale be used only as an
index of insoluble protein formation, nuclear sclerosis
has been used alone 3 4 6 or together with a measure of
the "osmotic" change5 to facilitate the subdivision of
a large single population of cataracts into several
smaller groups, each of which shares certain similarities.
These groups have been analyzed for Na + , K + , protein,
and other biochemical constituents.10"12
Several scientists have suggested that nuclear color
may not be a suitable index of the severity of cataract
formation. Weale,13 in discussing the physical changes
in the lens due to aging and cataract formation, pointed
out that in pure absorption (ie, no light scattering nuclei
present), spectral and spatial frequencies do not in-
teract. He emphasized that the resolving power of a
yellow lens might be unimpaired. Lerman and Bork-
man 9 stated their clinical impression that "the normal
human lens also shows an increasing yellow coloration
with age up to and including the hazel brown color
without any impairment of visual acuity and as such
should not be confused with a truly cataractous lens,
ie, one that seriously and significantly impairs visual
acuity to the point where it requires surgical interven-
tion." They also demonstrated that two specific age-
related fluorogens are responsible for the yellow color
of the normal lens nucleus which becomes readily ap-
parent by the second decade of life." Furthermore,
they demonstrated that the intensity of the nuclear
color may significantly reduce the lenticular trans-
mission of visible light, a finding that might explain
the reduction in visual acuity that we noted in the
 178 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / February 1984 Vol. 25

o.CXA b. CXE
4
4
3 -
Uj 3
Q
1 2 __#

I 8
1
0
0 2
i i
4 6
i
8
1
0 2 4
NS INDEX NS INDEX

d.SN
4r
Figs. 4a-f. Mean extent a,
CXA; b, CXE; c, CXP; d, SN;
3" e, SCA; f, SCP of regional
opacification versus the NS
index. 1976 immature cata-
2" racts formed the basis for this
table.
1 •

0 - \ i i i
2 4 6 8
0 2 4 8 °c
NS WDEX NS INDEX

e.SCA 1: S C P
10C
s 100
80
)-
8C)-

60 6C)-

40 V 4C)- #

20 2C»- / N
i

0 2 4 8 0 2 4 6 £
NS INDEX NS INDEX

very dark yellow-black lenses. Nordmann,14 in dis-
cussing another of Lerman and Borkman's papers,15
stated his concurrence with Lerman's finding that "a
coloured nucleus by itself does not automatically cat-
egorize a lens as being cataractous."
However, in none of the above studies was the as-
sumption tested that nuclear color is indeed a valid
measure of the extent of senile cataract formation.
Another problem that seriously affects thefinalclas-
sification of Duncan4 and Marcantonio5 is their reli-
ance on freezing to preserve the lens prior to classi-
fication. In 1981, we published a study16 showing that
freezing the human lens at — 15°C for as little as 15
min leads to significant clarification of cortical but not
nuclear opacification. Freezing, apart from its destruc-
tive effect on lens membranes, leads to a change in
cataract morphology that would lead the classifier to
underestimate the number and extent of cortical opac-
ities, while overestimating the relative amount of nu-
clear opacification.
In this study, the cataracts were not frozen but were
photographed immediately after intracapsular cataract
extraction with a photographic technique that sepa-
rately identifies the features of the opacity (black back-
ground) and the color of the nucleus (white back-
ground). This separation not only accurately captures

Downloaded From: https://iovs.arvojournals.org/ on 07/27/2018

 No. 2 CATARACT CLASSIFICATION: III. NUCLEAR SCLEROSIS/Chylock er ol.
Mean VA vs. NS , I976
Uj 3 0 asymptotically, or is there a more abrupt appearance
2 would suggest that, although the human lens has a
MEAN VA
1
0 1 2 3 4 5 6 7 8 years in our population of lenses), the search for the
NS INDEX
Fig. 5. Mean preoperative visual acuity index (VA) at the time
of cataract extraction for 1976 patients with immature cataracts as
a function of the NS index. Visual acuity index 1 = 20/20-20/30;
2 = 20/40-20/50; 3 = 20/60-20/80; 4 = 20/100-20/200; 5 = 20/
400-CF; 6 = hand movements-light projection; 7 = light perception;
8 = no light perception.
the anatomic and color detail, but permits one to care-
fully evaluate the relationship between nuclear color
and the extent (severity) of cataract formation.
By means of this methodology, we observed a grad-
ual increase in mean nuclear color, on the average,
with increasing nuclear opacification, but the color
change was so slight and the nuclear color index range
so broad as to make the degree of color change virtually
useless for the classification of nuclear opacification.
Further, mean extent data linking the severity of CXA,
CXE, CXP, SN, and SCP opacification with the NS
index show no correlations that might serve as the
basis for classifying the extent of cortical cataract for-
mation by the degree of nuclear yellowing.
In view of these findings, it appears reasonable to
recommend that nuclear color (sclerosis) be abandoned
as the single index of the severity of any type of senile
cataract formation.
The data on the relationship between nuclear color
and visual acuity are consistent with the impression
held by ophthalmologists for two centuries that yel-
lowing of the lens alone does not cause a significant
decrease in visual acuity. However, when the color
progresses from dark yellow through black, these is a
accelerated loss of vision. This finding supports use of
the term "cataracta nigra"17 and "cataracta brunes-
cens" 18 in the literature of the 18th and 19th centuries,
respectively. Sir Stewart Duke-Elder continued the use
of this term in his Textbook of Ophthalmology19 and
System of Ophthalmology.20
The dependency of nuclear color on age is also con-
sistent with long-term clinical experience, but its doc-
Downloaded From: https://iovs.arvojournals.org/ on 07/27/2018
179
umentation and the determination of its functional
form depends on the definition and careful application
of a systematic classification scheme such as the one
employed in these studies. That the form of the NS
versus Age curve is a growth curve is to be expected.
Of interest is its behavior between 0 and 40 years of
age, a range excluded by our data. Does NS appear
gradually in the early years so that the curve approaches
at some specifiable age? Lerman and Borkman's data 9
pale yellow color at birth, the color increases signifi-
cantly by the second decade. Lacking this crucial in-
formation (behavior of the NS index from 0 to 40
best "model" by curve-fitting is considered premature.
Parenthetically, and in closing, the experience gained
from these studies and their analysis convince us that
human lenses should be classified before freezing, and
the effect of freezing on the parameters being studied
must be measured, rather than assumed to be insig-
nificant. Further, to maximize photographic detail,
cataracts should be photographed against a flat black
background and nuclear color estimated with the lens
against a white background as described in Part I of
this series. Finally, due to the inability to estimate
consistently true nuclear color through an opacified
lens cortex, mature and hypermature lenses should not
be included in studies of nuclear color unless the nu-
cleus can be removed from the lens for photography.
This is due to the inability to estimate consistently
true nuclear color through an opacified lens cortex.
Key words: nuclear color (sclerosis), not a reliable index of
cataractous change
Acknowledgments
The authors would like to acknowledge the technical and
photographic expertise of Dr. Marjorie Lee, the excellent
secretarial support provided by Ms. Ann Piccolomini, and
the continuing support of the operating room nurses and
surgeons of the MEEI and the BWH.
References
1. Chylack LT Jr, Lee MR, Tung WH, and Cheng HM: Classifi-
cation of human senile cataractous change by the American
Cooperative Cataract Research Group (CCRG) Method I. In-
strumentation and technique Invest Ophthalmol Vis Sci 24:424,
J983.
2. Chylack LT Jr, White O, and Tung WH: Classification of human
senile cataractous change by the American Cooperative Cataract
Research Group (CCRG) Method. II. Staged simplification of
cataract classification. Invest Ophthalmol Vis Sci 25:166-173,
1984.
3. Pirie A: Color and solubility of the proteins of human cataracts.
Invest Ophthalmol 7:634, 1968.
4. Duncan G and Bushell AR: Relationships between colour, so-
 180 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / February 1984
dium and protein content in individual senile human cataractous
lenses. Ophthalmic Res 11:397, 1979.
5. Marcantonio JM, Duncan G, Bushell AR, and Davies PD: Clas-
sification of human senile cataracts by nuclear colour and sodium
content. Exp Eye Res 31:227, 1980.
6. Duncan G, editor: Mechanisms of Cataract Formation in the
Human Lens. London, Academic Press, 1981, pp. 1-5.
7. Cotlier E: Senile cataracts: evidence for acceleration by diabetes
and deceleration by salicylate. Can J Ophthalmol 16:113, 1981.
8. Rubin AD and Risley JF: The PROPHET System—an exper-
iment in providing a computer resource to scientists. MEDINFO,
Shires/Wolf, editors. 1F1P, North Holland Publishing Co, 1977,
pp. 77-81.
9. Lerman S and Borkman R: Spectroscopic evaluation and clas-
sification of the normal, aging, and cataractous lens. Ophthalmic
Res 8:335, 1976.
10. Truscott RJW and Augusteyn RC: Changes in human lens pro-
teins during nuclear cataract formation. Exp Eye Res 24:159,
1977.
11. Van Heyningen R: The human lens. I. A comparison of cataracts
extracted in Oxford (England) and Shikarpur (W. Pakistan). Exp
Eye Res 13:136, 1972.
12. Van Heyningen R: The human lens. III. Some observations on
the post-mortem lens. Exp Eye Res 13:155, 1972.
Downloaded From: https://iovs.arvojournals.org/ on 07/27/2018
Vol. 25
13. Weale RA: Physical changes due to age and cataract. In Mech-
anisms of Cataract Formation in the Human Lens, Duncan G,
editor. London, Academic Press, 1981, p. 60.
14. Nordmann J: Gerontological aspects of eye: research: selected
papers. In Interdisciplinary Topics in Gerontology, Vol. 13,
Hockwin O, editor. Basel, Switzerland, S. Karger, 1978, p. 182.
15. Lerman S and Borkman RF: Gerontological aspects of eye re-
search: Selected papers. In Interdisciplinary Topics in Geron-
tology, Vol. 13, Hockwin O, editor. Basel, Switzerland, S. Karger,
1978, p. 154.
16. Chylack LT Jr, Bettelheim FA, and Tung WH: Studies on human
cataracts I. Evaluation of techniques of human cataract pres-
ervation after extraction. Invest Ophthalmol Vis Sci 20:327,
• 1981.
17. Wenzel. Abhan. v.d. Star, Nurenberg, 1878.
18. Becker O. Zur Anatomie der Gesunden und Kranken Linse.
Unter Mitwirkung Von JR Da Gama Pinto und H. Schafer.
Wiesbaden, Bergmann, 1883, xv, p. 219.
19. Duke-Elder WS: In Textbook of Ophthalmology, Vol. 3. St.
Louis, CV Mosby, 1941, p. 3189.
20. Duke-Elder WS: Diseases of the lens and vitreous, glaucoma
and hypotony. In System of Ophthalmology, Volume 10. St.
Louis, CV Mosby, 1969, p. 128.