Pacific UniversityCommonKnowledge
Faculty Scholarship (COO) College of Optometry
Fall 10-24-2014
Impact of excess light and yellow filters onaccumulation of lipofuscinJohn R. HayesPacific University
David K. GlabePacific University
Follow this and additional works at: http://commons.pacificu.edu/coofac
Part of the Optometry Commons
This Article is brought to you for free and open access by the College of Optometry at CommonKnowledge. It has been accepted for inclusion inFaculty Scholarship (COO) by an authorized administrator of CommonKnowledge. For more information, please [email protected].
Recommended CitationHayes, John R. and Glabe, David K., "Impact of excess light and yellow filters on accumulation of lipofuscin" (2014). FacultyScholarship (COO). Paper 30.http://commons.pacificu.edu/coofac/30
Impact of excess light and yellow filters on accumulation of lipofuscin
DescriptionPurpose. Previous reports suggest excess high energy, short wave light increase lipofuscin fluorophoreaccumulation in the retinal pigment epithelium layer. Oxidation of lipofuscin has been implicated in thegenesis of macular degeneration. By taking advantage of the increased exposure to light by optometrystudents, we tested whether optometry students accumulate more lipofuscin fluorophores than similarly agedallied health students and whether yellow filters alter lipofuscin accumulation.
Method. The sample consisted of 54 non-optometry students (Mean age27, 4.1SD; 63% Female), 62 first-yearoptometry students (Mean age 27, 4.9SD; 55% female), and 39 second-year students (Mean age 26, 3.8SD,54% female). First year practice patients were exposed primarily to anterior segment biomicroscopy, whilesecond year practice patients included posterior segment biomicroscopy with condensing lens and binocularindirect ophthalmoscopy sessions. A two distribution mixture model estimated the gray scale of thefluorescent lipofuscin ring around the macula.
Results. There was significantly less luminance intensity in optometry students (Mean 70.8 grayscale) relativeto non-optometry students (Mean 76.2 grayscale, F = 5.3, p=.024) which was opposite from our prediction.Covariates included ge (b=.9,p=.002) and baseline lipofuscin (b=1.1, p
Conclusion. Our results were more consistent with oxidation of lipofuscin fluorophores than accumulationfollowing the excess exposure to light as practice patients. The study revealed intriguing trends in a challengingenvironment that suggested the topic is worth further investigation in a more rigorous experimentalenvironment. The bottom line however is that we do not trust our results over time due to the systematiceffect of multiple camera flashes that were not controlled across patients. We sincerely believe this study needsreplicated with an autofluorescence camera that has a reference point to adjust for the physical conditions.
KeywordsLipofuscin; blue light; yellow filters; fundus autofluorescence; A2E
DisciplinesOptometry
CommentsThis original research manuscript has not undergone peer review.
RightsTerms of use for work posted in CommonKnowledge.
This article is available at CommonKnowledge: http://commons.pacificu.edu/coofac/30
Impact of excess light and yellow filters on accumulation of lipofuscin
John R. Hayes Ph.D.
Pacific University Oregon
David K. Glabe OD MS
Pacific University Oregon
Correspondence:
John R. Hayes
College of Optometry
Pacific University
2043 College Way
Forest Grove, Oregon
Fax: 503 352 2929
Email: [email protected]
Acknowledgement:
The study was funded by the Macular Degeneration Foundation, Las Vegas, NV
Contact lenses were supplied by CooperVision
Lipofuscin is a complex aggregate of indigestible lysosomal material which accumulates in postmitotic
cells over the lifetime of an individual.1 While lipofuscin has been considered a biological marker for the
aging of cells, abnormally high levels of lipofuscin within the retinal pigment epithelium (RPE) are
associated with retinal disease and RPE dysfunction, as in Stargardt’s and age-related macular
degeneration.2, 3 Laboratory evidence suggests4 that RPE lipofuscin pigments are unique from that of
other cells as they are derived from molecular components of the visual cycle and primarily form during
periods of excess light exposure with significant rhodopsin bleaching and elevated levels of all-trans-
retinal in the rod outer segment (ROS) disks.5-7 Under these circumstances, the rate of generation of all-
trans-retinal by photo-activation of rhodopsin exceeds the rate at which all-trans-retinal is reduced to
all-trans-retinol in the ROS, providing a substrate for random side reactions to the visual cycle. Various
fluorescent byproducts of the visual cycle are major components of RPE lipofuscin, including the di-
retinal conjugate A2E (N-retinylidene-N-retinylethanolamine), which can induce DNA damage and RPE
cell apoptosis through photooxidative processes upon exposure to short-wavelength (blue) light.8-10
These lipofuscin fluorophores have been shown to sensitize the RPE to blue light damage and are
thought to play a role in the pathogenesis of certain macular diseases.11-13 It has been proposed that A2E
and other lipofuscin fluorophores may be quantified via in vivo fundus autofluorescence imaging14.
Quantification of these molecules and methods of altering their accumulation may be of clinical interest
in assessing and altering risk of various maculopathies.15
Optometry students have unusual exposure to high levels of light as they serve as practice
patients for their colleagues in learning various ophthalmoscopic examination procedures over a four-
year period of professional education. Our first study objective was to determine if excess light
exposure in optometry students relative to non-optometry health profession students leads to greater
accumulation of lipofuscin fluorophores.
Many ophthalmic instruments utilize light sources weighted toward shorter wavelengths more likely to
result in photo-oxidation of lipofuscin fluorophores such as A2E. The resulting products of lipofuscin
fluorophores may show increased fluorescence with early photo-oxidation;14 further oxidation and
photodegradation, however, may cause a decrease in autofluorescence above 540 nm.1, 16-20 Our second
study objective was to determine if blue light-blocking yellow filters would alter lipofuscin fluorophore
accumulation (as measured by fundus autofluorescence imaging) by having optometry students wear a
yellow contact lens on one randomly assigned eye during procedure practice sessions.
Method. This was a prospective cohort study for the comparison of fundus autofluorescence measures
of lipofuscin fluorophores in optometry and non-optometry students and a randomized controlled trial
exploring the effect of yellow filters on lipofuscin fluorophore accumulation within optometry students.
The research followed the tenets of the Declaration of Helsinki and informed consent was obtained from
subjects after explanation of the nature and possible consequences of the study. The study was
approved by the Pacific University Institutional Review Board.
Subjects. We recruited 61 first-year non-optometry students from the Pacific University College of
Health Professions who are in a three-year program, 62 first-year optometry students from the class of
2013 and 38 second-year students from the class of 2012. Participation was voluntary. Optometry
students were paid $10 per session and non-optometry students were paid $30 per session. The pay
differential was due to the requirement that non-optometry students had to drive from a different
campus (7 miles). Optometry students were also paid $90 for maintaining a practice log. The primary
outcome variable was digital luminance levels of serial fundus autofluorescence photographs. The study
was powered to detect a 0.5SD mean difference in luminance between optometry and non-optometry
students with a power of .9 at an alpha = .05 assuming a correlation between the baseline luminance
covariate and the follow-up of r = 0.7 with a 20% dropout.
Subjects had to have two healthy, normally functioning eyes. Those who reported any history of
eye disease or hereditary eye conditions were excluded. Individuals with visual correction were allowed
but must otherwise have had healthy functioning eyes. Optometry subjects had to be able to tolerate
contact lens wear, but were not required to be previous contact lens wearers. Biomicroscopy was
performed prior to administering mydriatic eye drops as typically done in routine eye examination to
assess risk of acute angle closure. Non-optometry subjects must have had von Herrick Grade 2 or larger
anterior chamber angles to ensure safety with dilation. Pregnant subjects were excluded (two non-
optometry students became pregnant during the study and were excluded from further participation
due to possible side effects of dilation).
Equipment and materials. A Topcon TRC-50DX retinal camera (Topcon, Tokyo) was used to collect data
from control and optometry subjects. The camera employs an excitation filter that produces a green
flash from 535nm to 595nm and a barrier filter for collection of flourescent light from 600nm to 730nm.
The camera was located in a completely dark room, and lights on the control panel were covered when
taking photographs to exclude any light other than the excitation flash.
CooperVision (Pleasanton, CA) supplied Edge III ProActive contact lenses (62% polymacon and
36% water) which were tinted yellow by SpecialtyTint (Escondido, CA). Subjects used the same lens
throughout the trial and were supplied with identical contact lens cleaner and cases. The contact lenses
exhibited the spectral profile of a longpass filter with cutoff at 470nm (above the 430 nm point of
oxidation for A2E). Contact lenses were removed prior to fundus imaging.
Optometry subjects completed a log sheet each time they served as a patient for practicing
ophthalmic procedures. Logs included the duration of light exposure to each eye and type of
ophthalmic equipment used (e.g. biomicroscope with or without condensing lens, binocular indirect
ophthalmoscopy). The file was monitored weekly for adherence to completion.
Procedure. Baseline fundus autofluorescence photographs were collected in January 2010 and follow-
up data in September 2010, yielding a nine month period of accumulation. Very little practice occurred
over the summer from June through August.
Subjects were recruited by a general email sent to first and second-year optometry students and first-
year professional students in all Pacific University’s College of Health Profession programs. Subjects
completed a brief questionnaire that included gender, age, ethnicity, refractive correction, and
nutritional supplementation.
Baseline photographs were taken by two photographers (optometry student research
assistants). Photographers alternated taking the first baseline photograph. Immediately following that
photograph the other photographer took a second photograph. For each photograph the camera was
refocused. Two photographs were also taken at follow-up but by the same photographer. The initial
baseline photographs were taken at flash intensity 50 watt seconds, but follow-up photographs were at
flash intensity 100 watt seconds. We chose the initial intensity for subject comfort, but concluded that a
higher intensity facilitated autofluorescence detection. The maximum for the camera was 350 watt
seconds Changing intensity prohibited direct measures of change from baseline; instead, baseline
measures were used as covariates to adjust for individual differences. Subsequent analysis
demonstrated that our method of computing luminance was accurate at either level of flash intensity.
Non-optometry students had their photographs taken on weekends and in the evening.
Optometry students had photographs scheduled at times during which they would be dilated for
practice or lab sessions, with images being acquired prior to light exposure as a practice patient. Eyes
were not bleached prior to the test photograph. Subject pupils were dilated at least 8 mm in diameter
before image acquisition as measured by a ruler prior to the photograph.
For optometry students, the yellow contact lens was randomized to either the left or right eye
at study entry and the same eye was filtered for the entire study. Optometry students wore the yellow
contact lens when exposed to light as a practice patient except for limited procedures requiring direct
contact of ophthalmic equipment to the cornea. During direct contact procedures, students utilized
yellow filters built into biomicroscopes when the eye randomized to the yellow lens was examined. If
biomicroscope filters were not available, then the student proceeded without filtering light.
Luminance calculation. The primary outcome measure was luminance intensity of the Topcon
autofluorescence photograph. The grayscale varied from black (0) to white (255). Figure 1a is a sample
autofluorescence photograph. A mixture model was developed using R (GNU, Free Software
Foundation). Figure 1b is the frequency distribution of the pixels along the grayscale and the gamma
distribution model of the distributions. The R function reading the TIFF files had automatic brightness
adjustment, so a small white square (10x10 pixel grayscale 255) was added to a black corner (grayscale
0) of each photograph in order to calibrate all pictures to the same brightness scale. The top one
percent of the brightness pixels and pixels less than a grayscale of four were eliminated from the
distribution to remove the black corners of the photograph and the extreme upper tail of brightness.
Our goal was to estimate the overall luminance of the brighter perifoveal ring of autofluorescence. The
smooth lines in Figure 1b show two modal points. The left distribution represents the distribution of
points primarily making up the pixels from the optic disc, macula, and main vessels. The right
distribution is the bulk of the lighter vessels and the gray of the entire photograph. We had two
research assistants independently use the open source graphics program GIMP to identify pixels that
best represented the “hyperfluorescent ring” around the macula. They used the Threshold function
within GIMP which sets all points black below a particular threshold and white above. Figure 1c is the
threshold view of Figure 1a. The small inset histogram identifies the threshold at grayscale 58 which
was approximately the mean of the higher distribution in Figure 1b. The researchers consistently chose
a point near the mean of the upper mixture distribution. This point best defined the bright ring around
the maculaFigure 1d shows the consistency in judgment (reliability) between the two research assistants
in estimating the mean autofluorescence luminance. In this figure, data are also shown on judgments
for major vessels. Two consecutive photographs were taken on each subject at baseline and follow-up.
Photograph 1 and 2 were significantly associated for both research assistant judge 1 (b=1.4, R2 = .92,
p<.001) and judge 2 (b=1.01, R2 = .90, p<.001) further supporting the reliability of the measure. Using
the subjective estimates as a guide for defining a rule for the computer, we averaged the pixel grayscale
above the mixture model mean of the second distribution as the estimate of the luminance of
autofluorescence.
The validity of our measure was calculated in several ways. There was a significant association
between baseline and follow-up measurements (R2 = 0.76, p<.001) demonstrating baseline measures
could predict luminance 9 months later. Age was significantly associated with baseline mean luminance
(R2 = 0.16, p<.001) as would be expected since lipofuscin accumulates over time. In testing the Topcon
camera on a blank piece of paper with the words “Can you see me now?”, we noted that the measured
luminance linearly increased with serial pictures every 10 seconds over a period of 40 trials (slope = .42
gray scale, R2 = .95, p<.001). We could see the text because the copy paper we used had added
fluorescence to make it brighter. The increase in luminance over time was likely due to heating of the
camera flash bulb. Considering the consecutive photographs taken during the study, the overall
baseline mean luminance was 45.50 gray scale for the first photograph and 45.91 for the second
photograph, a difference of .41. This compared favorably to the mean 10 second difference for plain
copy paper of .42 (copy paper autofluoresces due to fluorescent material added to the paper during
manufacturing).
Statistics. A between subjects analysis of covariance with baseline luminance and age covariates
assessed the first hypothesis of whether or not there was a difference between lipofuscin fluorophore
accumulation in optometry and non-optometry students as measured by fundus autofluorescence.21 A
within subjects analysis of covariance with baseline luminance and age covariates assessed the effect of
yellow filters on follow-up luminance. Figures are presented with 84% confidence intervals. These
confidence intervals mimic the results from unadjusted least significant difference tests at an p<.05.
Figure 1. Calculating luminance
Results.
Sample: Demographics of the sample are in Table 1.
student because of dropping out of school and
others due to an inability to locate.
57 (92%) first year students, 38 (97%) second year students, and 53(98%) of non
photographs were lost due to subjects dropping out of the study. Those remaining were 41 (72%) first
years, 34 (89%) second years, and 33 (62%) non
luminance at baseline for first year optometry students (45.4, 12.3SD versus 37.9, 13.8SD, t=2.93
p=.004). Second year and non-optometry dropouts tended to have higher baseline luminance means
but the differences were not statistically
versus 43.9, 16.2SD for second year and non
Demographics of the sample are in Table 1. At the outset of the study we lost 1 second year
student because of dropping out of school and two non-optometry students due to pregnancy and 5
others due to an inability to locate. Photographs of sufficient quality to be processed were available for
8 (97%) second year students, and 53(98%) of non-optometry. Follow
photographs were lost due to subjects dropping out of the study. Those remaining were 41 (72%) first
33 (62%) non-optometry students. Dropouts had significantly higher
luminance at baseline for first year optometry students (45.4, 12.3SD versus 37.9, 13.8SD, t=2.93
optometry dropouts tended to have higher baseline luminance means
but the differences were not statistically significant (49.9, 15.6SD versus 43.3, 12.3SD and 51.2, 20.1SD
versus 43.9, 16.2SD for second year and non-optometry respectively).
e lost 1 second year
pregnancy and 5
Photographs of sufficient quality to be processed were available for
optometry. Follow-up
photographs were lost due to subjects dropping out of the study. Those remaining were 41 (72%) first
gnificantly higher
luminance at baseline for first year optometry students (45.4, 12.3SD versus 37.9, 13.8SD, t=2.93
optometry dropouts tended to have higher baseline luminance means
significant (49.9, 15.6SD versus 43.3, 12.3SD and 51.2, 20.1SD
Table 1. Sample demographics at baseline.
First Years
(n=61)
Second Years
( n=39)
Non-Opt
(n=62)
Mean / % (SD) Mean / % (SD) Mean / % (SD)
Age (yrs) 27.3 (4.9) 26.5 (3.8) 26.8 (94.1)
Women (%) 54.8 53.8 63.0
Caucasian (%) 82.3 79.5 70.4
Asian (%) 17.7 15.4 24.1
Other (%) 5.1 5.6
Wore Correction (%) 72.6 64.1 44.4
Correct w/ contacts (%) 67.7 76.9 18.5
Slit-Lamp (minutes) 131.3 (71.6) 21.0 (24.1)
LED BIO (minutes) 1.4 (5.0) 14.9 (20.7)
Non-LED BIO (minutes) 0.7 (2.4) 142.1 (101.3)
High Plus (minutes) 13.5 (14.6) 141.6 (99.2)
Retinoscope (minutes) 34.1 (31.8) 2.2 (5.3)
Ophthalmoscope (minutes) 20.2 (22.3) 0.5 (2.0)
Other Exposure (minutes) 13.8 (4.4) 43.7 (11.6)
Total Exposure (minutes) 215.0 (106.9) 366.0 (183.8)
Hypothesis 1: Lipofuscin autofluorescence increases in optometry students as a function of
increased light exposure over non-optometry student controls (Figure 2). A between subject analysis of
covariance revealed significantly less luminance intensity (implied less lipofuscin) in optometry students
(Mean 70.8 grayscale) relative to non-optometry students (Mean 76.2 grayscale, F = 5.3, p=.024) which
was the opposite from our hypothesis. The effect size for this difference was -.52SD (DifferenceOpt-
NonOpt/SD = -5.4/10.5). The covariate influence on follow-up luminance in the model were age (b= .9
grayscale/age year, t=3.2, p=.002), baseline luminance (b=1.1 grayscale, t=15.1, p<.001), and replicate
photograph (b=1.7 grayscale for second photograph, t=2.2, p=.03). Figure 2 illustrates the adjusted
grayscale means for optometry and non-optometry students at nine month follow-up. Visual inspection
revealed no observable evidence of RPE damage in the follow-up autofluorescence photographs of
either group of students.
Table 2. Hypothesis 1: Model mean comparisons at follow-up luminance as a function of starting
baseline. Least significant t-test of optometry versus non-optometry students (Means square error =
99.6, 107df). Effect size is the difference between means divided by the root mean square error.
General linear model: Follow-up = 18.95 – 8.0(1st year) – 21.7 (2nd year) + 1.39 (Baseline) +
0.09(1st yr * Baseline) + 0.42 (2nd year * Baseline)
Group Quartile Baseline Follow-up
Difference
from Non-
Opt
t p Effect
Size (SD)
First Year Minimum 9.8 25.4 -7.1 -3.55 0.001 -0.71
25 32.1 58.3 -5.1 -2.55 0.012 -0.51
50 42.9 74.3 -4.1 -2.06 0.041 -0.41
75 51.3 86.7 -3.4 -1.69 0.095 -0.34
Maximum 91.2 145.6 0.2 0.11 0.914 0.02
Second Year Minimum 9.8 14.9 -17.6 -8.78 0.000 -1.76
25 32.1 55.3 -8.1 -4.06 0.000 -0.82
50 42.9 74.8 -3.6 -1.78 0.077 -0.36
75 51.3 90.0 0.0 -0.01 0.996 0.00
Maximum 91.2 162.2 16.9 8.42 0.000 1.69
Non-Optometry Minimum 9.8 32.5
25 32.1 63.4
50 42.9 78.4
75 51.3 90.1
Maximum 91.2 145.3
Figure 2. Luminance accumulation over 9 months
Hypothesis 2: Yellow filters will alter the
no effect of yellow filter for the first
filtered, t=.938/1.5, p=.532). The effect size for the sec
1.97/7.58), which was not statistically significant (mean = 72.0 unfiltered and 74.0 filtered, t = 1.5 , p=
.14). Age (b=.99, F=11.6, p=.001) and baseline luminance (b=1.34, F=309.5, p<.001) were significant
covariates, but replicate photograph (1.5 gray scale units greater on the second photo, F= 2.6, p=.11) did
not reach statistical significance.
After controlling for age and baseline luminance, there was no dose effect of total light exposure
follow-up luminance (p=.95). There was significantly more
class (Mean 366 minutes) than in the first year
exposure difference between filter and no filter (p= .574) or the group by filter interaction (p=.365).
. Luminance accumulation over 9 months
will alter the accumulation of lipofuscin fluorophores (Figure 3)
filter for the first-year class of optometry students (mean 69.4 unfiltered and 68.5
filtered, t=.938/1.5, p=.532). The effect size for the second-year optometry students was .26SD (
1.97/7.58), which was not statistically significant (mean = 72.0 unfiltered and 74.0 filtered, t = 1.5 , p=
6, p=.001) and baseline luminance (b=1.34, F=309.5, p<.001) were significant
covariates, but replicate photograph (1.5 gray scale units greater on the second photo, F= 2.6, p=.11) did
line luminance, there was no dose effect of total light exposure
up luminance (p=.95). There was significantly more practice light exposure in the second year
than in the first year (Mean 215 minutes, p<.001). However t
exposure difference between filter and no filter (p= .574) or the group by filter interaction (p=.365).
(Figure 3). There was
69.4 unfiltered and 68.5
year optometry students was .26SD (-
1.97/7.58), which was not statistically significant (mean = 72.0 unfiltered and 74.0 filtered, t = 1.5 , p=
6, p=.001) and baseline luminance (b=1.34, F=309.5, p<.001) were significant
covariates, but replicate photograph (1.5 gray scale units greater on the second photo, F= 2.6, p=.11) did
line luminance, there was no dose effect of total light exposure on
exposure in the second year
. However there was no
exposure difference between filter and no filter (p= .574) or the group by filter interaction (p=.365).
Figure 3. Yellow filters v unfiltered
Discussion.
Prior research has demonstrated that light exposure results in the accumulation of RPE
lipofuscin fluorophores. We hypothesized that optometry students’ exposure to excess light while
serving as practice patients would lead to more lipofuscin
baseline autofluorescence luminance (lipofuscin proxy), optometry students had significantly less
accumulation at nine months follow
difference.
While most previous research has supported the concept of increased lipofuscin with
light exposure, Morgan et al have shown a reduction in autofluorescence in
macaques after light exposure at 568 nm.
levels of light exposure. Subsequent analysis suggests that the processes of photoisomerization,
photooxidation and photodegradation may explain this phenomenon.
of light exposure led to permanent RPE cell dysfunction as viewed with an adaptive optics scanning laser
ophthalmoscope (AOSLO). We noted a decline in
not detect damage to RPE cells, although we lacked an AOSLO.
The second hypothesis tested whether yellow filters could alter lipofuscin accumulation. We
found no scientifically relevant trend of filter for first years, effect size equal to .12SD. However, we
Prior research has demonstrated that light exposure results in the accumulation of RPE
lipofuscin fluorophores. We hypothesized that optometry students’ exposure to excess light while
serving as practice patients would lead to more lipofuscin autofluorescence. After adjusting for age and
luminance (lipofuscin proxy), optometry students had significantly less
accumulation at nine months follow-up. The effect was a scientifically meaningful .5 standard deviation
le most previous research has supported the concept of increased lipofuscin with
light exposure, Morgan et al have shown a reduction in autofluorescence in
macaques after light exposure at 568 nm.18 They found transient decreases in autofluorescence
levels of light exposure. Subsequent analysis suggests that the processes of photoisomerization,
photooxidation and photodegradation may explain this phenomenon. In Morgan's study, higher levels
exposure led to permanent RPE cell dysfunction as viewed with an adaptive optics scanning laser
ophthalmoscope (AOSLO). We noted a decline in autofluorescence in our optometry students but did
not detect damage to RPE cells, although we lacked an AOSLO.
The second hypothesis tested whether yellow filters could alter lipofuscin accumulation. We
found no scientifically relevant trend of filter for first years, effect size equal to .12SD. However, we
Prior research has demonstrated that light exposure results in the accumulation of RPE
lipofuscin fluorophores. We hypothesized that optometry students’ exposure to excess light while
. After adjusting for age and
luminance (lipofuscin proxy), optometry students had significantly less
up. The effect was a scientifically meaningful .5 standard deviation
le most previous research has supported the concept of increased lipofuscin with increased
ound transient decreases in autofluorescence at low
levels of light exposure. Subsequent analysis suggests that the processes of photoisomerization,
In Morgan's study, higher levels
exposure led to permanent RPE cell dysfunction as viewed with an adaptive optics scanning laser
autofluorescence in our optometry students but did
The second hypothesis tested whether yellow filters could alter lipofuscin accumulation. We
found no scientifically relevant trend of filter for first years, effect size equal to .12SD. However, we
noted a small trend for second year students (.26SD) which was not significant. We could not rule out
chance as a reasonable alternative to the differences found.
One possible explanation for our findings is that excess light exposure led to increased oxidation
and photodegradation of lipofuscin fluorophores such as A2E within eyes of optometry students relative
to non-optometry students. Oxidation of lipofuscin fluorophores decreases fluorescence above 540nm
(our study camera’s barrier filter collects fluorescent light from 615-715nm). Oxidation of lipofuscin
occurs more readily at short wavelengths, and thus may have been reduced by the presence of a yellow
filter, as has previously been demonstrated to occur in vitro. Light exposure in second-year optometry
students was greater as they were exposed to procedures involving more direct illumination of the
fundus (high plus and biomicroscopy (BIO)), while first years were exposed primarily to anterior segment
BIO. This may explain why the protective effect of yellow filters was present only in second-year
optometry students. The filter effects were trends only and were not significantly different.
We acknowledge a number of limitations to our study which should provoke a cautious
interpretation of results. While we were confident in our ability to quantify autofluorescence with the
Topcon camera, the device was not designed for this purpose. The quantification method of analysis
was by computer program and there was no subjective component. The difference between
consecutive photographs was detectable and similar to the difference between consecutive
photographs of plain copy paper with added fluorescence caused by the heating of the instrument’s
flash bulb.
We lost access to our Topcon camera after nine months, at which point a Heidelberg Spectralis
with BluePeak SLO/OCT (Heidelberg, Carlsbad, CA) was acquired and utilized in following the subjects for
another year. The Spectralis was also not designed to quantify lipofuscin. The Spectralis proved more
difficult for our photographers to operate, and blue light laser exposure time was variable for the
subject as the instrument required multiple photographs of a certain quality. The Spectralis could not
be run in a dark room, allowing uncontrolled ambient light during photography. We were not able to
demonstrate reliability or validity for quantifying luminance with the Spectralis in any of the
measurements described above for the Topcon camera. Subsequent analysis showed no correlation
between the Topcon baseline and Spectralis follow-up, or even an appreciable correlation between first
and second consecutive photographs within a Spectralis imaging session. There was no correlation with
age. A reliable and valid method of quantifying lipofuscin has been more recently designed for the
Spectralis, but the new modification of the camera was unavailable to us at the time of our study.14, 22
The study suffered from attrition. For many of the first year optometry students, managing a practice
log and using the contact lens was too much of a distraction to continue in the study. The non-
optometry students were from a different campus in a different town and the inconvenience of the trip
may have contributed to a number of subjects choosing not to participate in the follow-up visit. Future
studies will need to consider ways to reduce these barriers to study completion.
We recognize that talking about non-significant trends is hazardous. This study should be
replicated once a reliable and validated measure of quantifying autofluorescence is available. Our
student population appeared to be a reasonable working sample for this type of study; however, the age
and health of this study population may limit application of results to older patients more at risk for
retinal disease. Replication of these findings has important clinical implications for understanding the
role of light exposure in lipofuscin fluorophore measures and its possible relation to retinal disease.
References
1. Sparrow JR, Gregory-Roberts E, Yamamoto K, Blonska A, Ghosh SK, Ueda K, Zhou J. The bisretinoids of
retinal pigment epithelium. Prog Retin Eye Res Mar 2012;31(2):121-35. Available from:
http://sfxhosted.exlibrisgroup.com/pacificu?sid=OVID:medline&id=pmid:22209824&id=doi:10.1016%2F
j.preteyeres.2011.12.001&issn=1350-9462&isbn=&volume=31&issue=2&spage=121&pages=121-
35&date=2012&title=Progress+in+Retinal+%26+Eye+Research&atitle=The+bisretinoids+of+retinal+pigm
ent+epithelium.&aulast=Sparrow&pid=%3Cauthor%3ESparrow+JR%3BGregory-
Roberts+E%3BYamamoto+K%3BBlonska+A%3BGhosh+SK%3BUeda+K%3BZhou+J%3C%2Fauthor%3E%3C
AN%3E22209824%3C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E. Accessed 20120220.
2. Sparrow JR, Fishkin N, Zhou J, Cai B, Jang YP, Krane S, Itagaki Y, Nakanishi K. A2E, a byproduct of the
visual cycle. Vision Res Dec 2003;43(28):2983-90.
3. Delori FC, Staurenghi G, Arend O, Dorey CK, Goger DG, Weiter JJ. In vivo measurement of lipofuscin in
stargardt's disease--fundus flavimaculatus. Invest Ophthalmol Vis Sci Oct 1995;36(11):2327-31.
4. Wolf G. Lipofuscin and macular degeneration. Nutr Rev Oct 2003;61(10):342-6.
5. Ben-Shabat S, Parish CA, Vollmer HR, Itagaki Y, Fishkin N, Nakanishi K, Sparrow JR. Biosynthetic studies
of A2E, a major fluorophore of retinal pigment epithelial lipofuscin. J Biol Chem Mar 1 2002;277(9):7183-
90. Available from:
http://sfxhosted.exlibrisgroup.com/pacificu?sid=OVID:medline&id=pmid:11756445&id=doi:&issn=0021-
9258&isbn=&volume=277&issue=9&spage=7183&pages=7183-
90&date=2002&title=Journal+of+Biological+Chemistry&atitle=Biosynthetic+studies+of+A2E%2C+a+majo
r+fluorophore+of+retinal+pigment+epithelial+lipofuscin.&aulast=Ben-Shabat&pid=%3Cauthor%3EBen-
Shabat+S%3BParish+CA%3BVollmer+HR%3BItagaki+Y%3BFishkin+N%3BNakanishi+K%3BSparrow+JR%3C
%2Fauthor%3E%3CAN%3E11756445%3C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E. Accessed
20020225.
6. Katz ML, Drea CM, Eldred GE, Hess HH, Robison WG,Jr. Influence of early photoreceptor degeneration
on lipofuscin in the retinal pigment epithelium. Exp Eye Res Oct 1986;43(4):561-73.
7. Sparrow JR, Vollmer-Snarr HR, Zhou J, Jang YP, Jockusch S, Itagaki Y, Nakanishi K. A2E-epoxides
damage DNA in retinal pigment epithelial cells. vitamin E and other antioxidants inhibit A2E-epoxide
formation. J Biol Chem May 16 2003;278(20):18207-13.
8. Zhou J, Kim SR, Westlund BS, Sparrow JR. Complement activation by bisretinoid constituents of RPE
lipofuscin. Invest Ophthalmol Vis Sci Mar 2009;50(3):1392-9. Available from:
http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=19029031;
http://sfxhosted.exlibrisgroup.com/pacificu?sid=OVID:medline&id=pmid:19029031&id=doi:&issn=0146-
0404&isbn=&volume=50&issue=3&spage=1392&pages=1392-
9&date=2009&title=Investigative+Ophthalmology+%26+Visual+Science&atitle=Complement+activation
+by+bisretinoid+constituents+of+RPE+lipofuscin.&aulast=Zhou&pid=%3Cauthor%3EZhou+J%3BKim+SR
%3BWestlund+BS%3BSparrow+JR%3C%2Fauthor%3E%3CAN%3E19029031%3C%2FAN%3E%3CDT%3EJo
urnal+Article%3C%2FDT%3E.
9. Schutt F, Bergmann M, Holz FG, Dithmar S, Volcker HE, Kopitz J. Accumulation of A2-E in
mitochondrial membranes of cultured RPE cells. Graefes Arch Clin Exp Ophthalmol Mar
2007;245(3):391-8.
10. Sparrow JR, Nakanishi K, Parish CA. The lipofuscin fluorophore A2E mediates blue light-induced
damage to retinal pigmented epithelial cells. Invest Ophthalmol Vis Sci Jun 2000;41(7):1981-9.
11. Schutt F, Davies S, Kopitz J, Holz FG, Boulton ME. Photodamage to human RPE cells by A2-E, a
retinoid component of lipofuscin. Invest Ophthalmol Vis Sci Jul 2000;41(8):2303-8.
12. Sparrow JR, Zhou J, Ben-Shabat S, Vollmer H, Itagaki Y, Nakanishi K. Involvement of oxidative
mechanisms in blue-light-induced damage to A2E-laden RPE. Invest Ophthalmol Vis Sci Apr
2002;43(4):1222-7.
13. Delori F, Greenberg JP, Woods RL, Fischer J, Duncker T, Sparrow J, Smith RT. Quantitative
measurements of autofluorescence with the scanning laser ophthalmoscope. Invest Ophthalmol Vis Sci
2011;52(13):9379-90. Available from: http://alliance-
primo.hosted.exlibrisgroup.com/openurl/PU/pu_services_page?sid=OVID:medline&id=pmid:22016060
&id=doi:10.1167%2Fiovs.11-8319&issn=0146-
0404&isbn=&volume=52&issue=13&spage=9379&pages=9379-
90&date=2011&title=Investigative+Ophthalmology+%26+Visual+Science&atitle=Quantitative+measure
ments+of+autofluorescence+with+the+scanning+laser+ophthalmoscope.&aulast=Delori&pid=%3Cautho
r%3EDelori+F%3BGreenberg+JP%3BWoods+RL%3BFischer+J%3BDuncker+T%3BSparrow+J%3BSmith+RT
%3C%2Fauthor%3E%3CAN%3E22016060%3C%2FAN%3E%3CDT%3EComparative+Study%3C%2FDT%3E.
Accessed 20111214.
14. Kim SR, Jang YP, Sparrow JR. Photooxidation of RPE lipofuscin bisretinoids enhances fluorescence
intensity. Vision Res Mar 31 2010;50(7):729-36. Available from: http://alliance-
primo.hosted.exlibrisgroup.com/openurl/PU/pu_services_page?sid=OVID:medline&id=pmid:19800359
&id=doi:10.1016%2Fj.visres.2009.09.015&issn=0042-
6989&isbn=&volume=50&issue=7&spage=729&pages=729-
36&date=2010&title=Vision+Research&atitle=Photooxidation+of+RPE+lipofuscin+bisretinoids+enhances
+fluorescence+intensity.&aulast=Kim&pid=%3Cauthor%3EKim+SR%3BJang+YP%3BSparrow+JR%3C%2Fa
uthor%3E%3CAN%3E19800359%3C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E. Accessed
20100316.
15. Feldman TB, Yakovleva MA, Dontsov AE, Ostrovsky MA. Fluorescence emission and excitation
spectra of fluorophores of lipofuscin granules isolated from retinal pigment epithelium of human
cadaver eyes. Russian Chemical Bulletin 2010;59(1):276 <last_page> 283.
16. Hunter JJ, Morgan JI, Merigan WH, Sliney DH, Sparrow JR, Williams DR. The susceptibility of the
retina to photochemical damage from visible light. Prog Retin Eye Res Jan 2012;31(1):28-42. Available
from:
http://sfxhosted.exlibrisgroup.com/pacificu?sid=OVID:medline&id=pmid:22085795&id=doi:10.1016%2F
j.preteyeres.2011.11.001&issn=1350-9462&isbn=&volume=31&issue=1&spage=28&pages=28-
42&date=2012&title=Progress+in+Retinal+%26+Eye+Research&atitle=The+susceptibility+of+the+retina+
to+photochemical+damage+from+visible+light.&aulast=Hunter&pid=%3Cauthor%3EHunter+JJ%3BMorg
an+JI%3BMerigan+WH%3BSliney+DH%3BSparrow+JR%3BWilliams+DR%3C%2Fauthor%3E%3CAN%3E22
085795%3C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E. Accessed 20111219.
17. Morgan JI, Hunter JJ, Merigan WH, Williams DR. The reduction of retinal autofluorescence caused by
light exposure. Invest Ophthalmol Vis Sci Dec 2009;50(12):6015-22. Available from:
http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=medl&AN=19628734.
18. Sparrow JR, Yoon KD, Wu Y, Yamamoto K. Interpretations of fundus autofluorescence from studies of
the bisretinoids of the retina. Invest Ophthalmol Vis Sci Sep 2010;51(9):4351-7. Available from:
http://alliance-
primo.hosted.exlibrisgroup.com/openurl/PU/pu_services_page?sid=OVID:medline&id=pmid:20805567
&id=doi:10.1167%2Fiovs.10-5852&issn=0146-
0404&isbn=&volume=51&issue=9&spage=4351&pages=4351-
7&date=2010&title=Investigative+Ophthalmology+%26+Visual+Science&atitle=Interpretations+of+fund
us+autofluorescence+from+studies+of+the+bisretinoids+of+the+retina.&aulast=Sparrow&pid=%3Cauth
or%3ESparrow+JR%3BYoon+KD%3BWu+Y%3BYamamoto+K%3C%2Fauthor%3E%3CAN%3E20805567%3
C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E. Accessed 20100831.
19. Yamamoto K, Zhou J, Hunter JJ, Williams DR, Sparrow JR. Toward an understanding of bisretinoid
autofluorescence bleaching and recovery. Invest Ophthalmol Vis Sci Jun 2012;53(7):3536-44. Available
from: http://alliance-
primo.hosted.exlibrisgroup.com/openurl/PU/pu_services_page?sid=OVID:medline&id=pmid:22570342
&id=doi:10.1167%2Fiovs.12-9535&issn=0146-
0404&isbn=&volume=53&issue=7&spage=3536&pages=3536-
44&date=2012&title=Investigative+Ophthalmology+%26+Visual+Science&atitle=Toward+an+understand
ing+of+bisretinoid+autofluorescence+bleaching+and+recovery.&aulast=Yamamoto&pid=%3Cauthor%3E
Yamamoto+K%3BZhou+J%3BHunter+JJ%3BWilliams+DR%3BSparrow+JR%3C%2Fauthor%3E%3CAN%3E2
2570342%3C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E. Accessed 20120611.
20. SPSS [computer program] - version 20. Chicago, IL. IBM corp; 2013.
21. Sparrow JR, Blonska A, Flynn E, Duncker T, Greenberg JP, Secondi R, Ueda K, Delori FC. Quantitative
fundus autofluorescence in mice: Correlation with HPLC quantitation of RPE lipofuscin and
measurement of retina outer nuclear layer thickness. Invest Ophthalmol Vis Sci 2013;54(4):2812-20.
Available from: http://alliance-
primo.hosted.exlibrisgroup.com/openurl/PU/pu_services_page?sid=OVID:medline&id=pmid:23548623
&id=doi:10.1167%2Fiovs.12-11490&issn=0146-
0404&isbn=&volume=54&issue=4&spage=2812&pages=2812-
20&date=2013&title=Investigative+Ophthalmology+%26+Visual+Science&atitle=Quantitative+fundus+a
utofluorescence+in+mice%3A+correlation+with+HPLC+quantitation+of+RPE+lipofuscin+and+measurem
ent+of+retina+outer+nuclear+layer+thickness.&aulast=Sparrow&pid=%3Cauthor%3ESparrow+JR%3BBlo
nska+A%3BFlynn+E%3BDuncker+T%3BGreenberg+JP%3BSecondi+R%3BUeda+K%3BDelori+FC%3C%2Fau
thor%3E%3CAN%3E23548623%3C%2FAN%3E%3CDT%3EJournal+Article%3C%2FDT%3E. Accessed
20130418.
Appendix. R program to calculate mean luminance from autofluorescence photos. The
program used a mixture to compute the proportion of pixels, mean, and standard deviations of
two distributions within the file. The luminance attributed to lipofuscin was based on the mean
luminance of the brighter distribution calculating the mean luminance of the pixels of the pixels
greater than the mean of the second distribution.
R code
library(pixmap)
library(rtiff)
library(mixtools)
library(mixdist)
path<-c(
'd:/blue light/study data/Class 2012/Sept 10/Left/'
) */ For example
np<-length(path) # Number of folders to review. Each subject class has their own folder for
each eye
p=1
i=1
while (p<=np) # Go through each folder
{
eyepic=dir(path=path[p] ,pattern="*.tif",ignore.case=TRUE)
sort(eyepic)
ni<-length(eyepic)
###
pic_mixture<-function(i){
filename<-paste(path[p],sep="",eyepic[i])
#filename<-'d:/blue light/new baseline unfiltered/126 af 2.tif' # for example
getDescription(filename)
pic<-readTiff(filename)
plot(pic,main=path[p],sub=eyepic[i])
mat<-getChannels(pic)
mat<-sort((mat))
hist(mat)
mat<-mat*255
mat<-round(mat)
q1<-quantile(mat,probs=c(.05,.5,.999),na.rm=TRUE)
mat<-ifelse(mat>4&mat<q1[3],mat,NA) #Eliminate black corners and text on the photo
mat<-na.omit(mat)
q<-quantile(mat,probs=c(.05,.5,.9),na.rm=TRUE)
pi<-c(.05,.95)
mu<-c(structure(c(q[1],q[3]),names=NULL))
sigma<-c(.3*mu[1],.2*mu[2])
parms<-data.frame(pi,mu,sigma)
matdata=mixgroup(mat,breaks=200)
matmix<-as.mixdata(matdata)
fitmat1<-
mix(matmix,mixpar=parms,dist="gamma",constr=mixconstr(consigma="SFX",fixsigma=c(FALSE,
FALSE),conmu="MFX",fixmu=c(FALSE,FALSE)), iterlim=150)
x<-fitmat1[[1]]
dev.next()
plot(fitmat1)
plotname<-c(paste(substr(filename,1,nchar(filename)-4),sep="",".jpg"))
dev.copy(jpeg,plotname)
dev.off()
mat2<-ifelse(mat>x[[2,2]],mat,NA) #/select pixels > mean of the upper distribution
lipo<-mean(mat2,na.rm=TRUE)
id<-substr(eyepic[i],1,3)
eye<-1
#
line<-c(id,eye,lipo,q,x[[1]],x[[2]],x[[3]],fitmat1[[5]],fitmat1[[6]],filename)
#line<-c(id,eye,picnum,lipo,q,x[[1]],x[[2]],x[[3]],fitmat1[[5]],fitmat1[[6]],filename)
line
out<-paste( 'd:/blue light/lipo 20120510/',sep="","2012Sept10Left.txt")
write.table(t(line),out,append=TRUE,quote=FALSE,sep=",",row.names=FALSE,col.names=FALSE)
print(q)
fitmat1<-""
mat<-""
pic<-""
return(TRUE)
}
Top Related