Ординатура / Офтальмология / Английские материалы / Progress in Lens and Cataract Research_Hockwin_2002

Fig. 3. Definition of MTD2.3:16. Dose-response function for UVR-300 nm radiation induced cataract ( ) and 1 standard deviation ( ) and 2 standard deviations above (—).

In each individual rat, it is expected that there is a 16% chance to find a difference between the exposed and contralateral side greater than 1 standard deviation above the doseresponse function at any dose. If figure 1 and figure 2 are combined, figure 3 is obtained.

The MTD then may be defined as the dose corresponding to the crossover between 2 standard deviations above no difference of light scattering at zero dose, and 1 standard deviation above the dose-response curve for the difference of light scattering between the exposed and contralateral nonexposed side.

From figure 3 it is seen:

2 k(MTD2.3:16)2

or

Results

The sensitivity, k (equation 1), was estimated to be 7.17 10 3 (kJ/m2)2 and the residual standard deviation was estimated to be 9.56 10 2 tEDC. MTD2.3:16 was therefore estimated to be 3.65 kJ/m2.

Discussion

In the present study, we developed a strategy for the estimation of toxicity of UVR to the ocular lens with small sample experiments.

The interpretation of the current finding of MTD2.3:16 of 3.65 kJ/m2 is that there is a 16% probability that an individual exposed to the MTD will have a

difference of intensity of forward light scattering between the exposed and the nonexposed contralateral lens exceeding the level found in 97.7% of eyes from individuals that have not been exposed to UVR. The currently found MTD provides a limit for avoidance of cataract that is very close to the threshold limit of 5 kJ/m2 for permanent lens damage that was previously published by Pitts et al.

[10]based on a binary response event model.

The strategy for MTD estimation can be generalized to all continuous

response events. However, depending on the specific dose-response curve, the formula for calculation will vary. Further, the probability levels may be chosen differently but that will then also modify the formula for the calculation of the MTD.

In the current strategy, it is assumed that the residual standard deviation is constant regardless of the difference of intensity of light scattering recorded. In some cases, there may be a functional relationship between the residual standard deviation and the difference of intensity of light scattering. If this is the case, it has to be considered.

We are here assuming that the square root of the ratio between the residual standard deviation and sensitivity as estimated from the regression is a correct estimation of the expected value for the square root of the ratio between the residual standard deviation and the real population sensitivity (equation 3). The uncertainty of the estimation of the MTD may be expressed e.g. as a confidence interval. For this, it is necessary to derive the expression that describes the estimation of the standard deviation for MTD. This expression is currently not available.

One of the most significant drawbacks of current safety limits is that these have been derived from acute experiments. Those results are then extrapolated to long-term exposure. The currently derived strategy can also be used for the determination of toxicity in long-term experiments. With such experiments it will be possible to predict safety levels for long-term exposures.

Acknowledgments

The present work was supported by Karolinska Insitutets forskningsfonder, Carmen och Bertil Regnérs Fond för Forskning inom området ögonsjukdomar, Kronprinsessan Margaretas Arbetsnämnd för synskadade, Swedish Society of Medicine, the Swedish Radiation Protection Institute, and Sandqvists stiftelse.

References

1Hiller R, Giacometti L, Yuen K: Sunlight and cataract: An epidemiologic investigation. Am J Epidemiol 1977;105:450–459.

2Hiller R, Sperduto RD, Ederer F: Epidemiologic associations with cataract in the 1971–1972 National Health and Nutrition Examination Survey. Am J Epidemiol 1983;118:239–249.

3Taylor HR, West SK, Rosenthal FS, Munoz B, Newland HS, Abbey H, et al: Effect of ultraviolet radiation on cataract formation. N Engl J Med 1988;319:1429–1433.

4West SK, Duncan DD, Munoz B, Rubin GS, Fried LP, Bandeen-Roche K, et al: Sunlight exposure and risk of lens opacities in a population-based study: The Salisbury Eye Evaluation project. JAMA 1998;280:714–718.

5Cruickshanks KJ, Klein BE, Klein R: Ultraviolet light exposure and lens opacities: The Beaver Dam Eye Study. Am J Public Health 1992;82:1658–1662.

6Widmark J: Über die Durchlässigkeit der Augenmedien für ultraviolette Strahlen. Beitr Ophthalmol Stockh 1891;460–502.

7Söderberg PG: Acute cataract in the rat after exposure to radiation in the 300 nm wavelength region. A study of the macro-, microand ultrastructure. Acta Ophthalmol (Copenh) 1988;66: 141–152.

9Verhoeff FH, Bell L, Walker CB: The pathological effects of radiant energy upon the eye. Proc Am Acad Art Sci 1915/1916;51:629–818.

10Pitts DG, Cullen AP, Hacker PD: Ocular effects of ultraviolet radiation from 295 to 365 nm. Invest Ophthalmol Vis Sci 1977;16:932–939.

11Merriam J, Löfgren S, Michael R, Söderberg PG, Dillon J, Zheng L, et al: An action spectrum for UV-B radiation and the rat lens. Invest Ophthalmol Vis Sci 2000;41:2642–2747.

12ACGIH 1995–1996: Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposure Indices. Cincinnati, American Conference of Governmental Industrial Hygienists, 1996.

13Finney DJ: Probit Analysis. Cambridge, Cambridge University Press, 1971.

14Michael R, Söderberg PG, Chen E: Dose-response function for lens forward light scattering after in vivo exposure to ultraviolet radiation. Graefes Arch Clin Exp Ophthalmol 1998;236:625–629.

15Söderberg PG: Development of light dissemination in the rat lens after exposure to radiation in the 300 nm wavelength region. Ophthalmic Res 1990;22:271–279.

16Michael R, Söderberg PG, Chen E: Long-term development of lens opacities after exposure to ultraviolet radiation at 300 nm. Ophthalmic Res 1996;28:209–218.

17Söderberg PG, Chen E: An objective and rapid method for the determination of light dissemination in the lens. Acta Ophthalmol (Copenh) 1990;68:44–52.

P.G. Söderberg, Research Department, St. Erik’s Eye Hospital, Karolinska Institutet, Fleminggatan 22, SE–112 82 Stockholm (Sweden) Tel. 46 8672 3098, Fax 46 8672 3352, E-Mail per.soderberg@ste.ki.se

suggest exposure to visible light can damage the retinal pigment epithelium [5]. Population-based studies have provided equivocal data, with no relationship in one study [6] and suggestive evidence in two others [7, 8]. In the Waterman Study, no association between UVB and early or severe AMD was found [9] but severe AMD was associated with exposure to visible light late in life [7]; the data were based on only seven cases of severe AMD in that study, but the results suggest further investigation is warranted. Cruickshanks et al. [9] analyzed the relationship between retinal pigment abnormalities and self-reported outside exposure in summer and found a significant association in men.

Exposures to visible light were not explicitly modeled in any of the previous studies. Therefore, in order to carry out our determination of the relationship between cumulative ocular exposure to visible radiation and the prevalence of AMD, a model for ocular exposure to visible light was created. The work reported herein represents a generalization of our previous work in estimating UVB exposures among a general population living in Salisbury, Md. [10–12].

Methods

Our objective is to estimate an individual’s lifelong ocular exposure in the visible portion of the sunlight spectrum. As shown elsewhere [10], our conceptual model of cumulative exposure for a single day is the following:

season); Teye fixed factor (between 0 and 1) that reflects the diminution conferred by the use of eyewear; G a geographic correction factor that relates the total yearly ambient exposures seen in the Maryland area to those experienced at locations elsewhere in the world; Ft(ti) the fraction of time spent outdoors in the ith period of the day (can be variable by month), and Ha(ti) the global ambient exposure during this day (variable by month and hour of day).

Units of exposure and exposure rate are, respectively, the (photopically weighted) energy per unit area and power per unit area. The ocular-ambient exposure ratio (OAER) depends heavily upon the reflectivity of the surface over which an individual works. This dependence is subsumed in the actual numerical value which is used. Therefore, equation 1 contains no explicit dependence on the reflectivity of the work surface.

The conceptual model presented in equation 1 is implemented with the following formula:

Table 1. Monthly and hourly visible ambient exposures for Chesapeake Bay Area (units are milli-VMSY)

VMSY 3,770 V h, photopic weighting.

where m index that runs over the months; and t index that runs over the hours of the day from 5 a.m. (5 h) to 7 p.m. (19 h).

As indicated in equation 2, the model accounts for the possibility that one’s geographical location and OAER may change with the season. Fractional time spent out of doors is allowed to vary according to personal habit with the month and time of day. Ambient exposure varies with month and time of day. Diminutions conferred by the use of eyewear are fixed. However, by allowing a variation with season, we account for variability in use.

The ambient exposure levels were obtained from data bases derived from a 2-year measurement program near Easton, Md. at the Foundation for the Advanced Research in the Medical Sciences (38° 43.9 N, 76° 8.4 W) [12]. Instrumentation at this site included aYankee Environmental Systems Model TSRP-1 photopic pyranometer (Yankee Environmental Systems, Turner Falls, Mass., USA). The monthly and hourly visible ambient exposure [Ha(ti )] for the Chesapeake Bay area is shown in table 1. Entries in this table are expressed in milli-visible Maryland sun years (VMSY), where a VMSY 3,770 V h. Although the pyranometer possesses a photopic response, it is uncalibrated in an absolute sense.

The numeric value of the geographic modifier, G, is the result of a semi-empirical model developed at NASA [13]. This model uses reduced resolution narrow band radiance (600 and 11,000 nm) measurements made by the imaging radiometers on five geostationary satellites (METEOSAT, INSAT, GMS, GOES-EAST and GOES-WEST) and at least one polar orbiting NOAA Satellite [14–16]. These data along with a radiative transfer model [17, 18], allow the estimate of the photosynthetically active radiation (PAR; unweighted integration over the 0.4- to 0.7- m band) yearly dose (corrected for cloud cover) at any point on the globe, excluding the polar regions (latitudes greater than 64.5°). The geographic correction factor is

Table 2. Geographic correction factor relative to Maryland for exposure in the visible

shown in table 2. Unless otherwise noted in table 2, the location of the state capital was used in the computation of this parameter. Although a PAR action spectrum was employed, we demonstrate in Appendix A that the numeric value of the geographic correction factor is insensitive to details of the particular action spectrum.

The OAER is defined as the quotient of the exposure striking the tangent plane of the eye and the global ambient exposure on a horizontal plane. To determine this ratio [12], we instrumented a series of individuals for measurement of visible radiation striking the plane of the face. These volunteers also completed an interview form detailing the amount of time spent out of doors in the 12-hour period for which the monitors were worn, the use of hats and glasses, eye color, and whether or not they considered themselves to be sensitive to high light levels. The protocol was approved by the Joint Committee on Clinical Investigation of the Johns Hopkins University Medical Institutions. The protocol follows the tenets of the Declaration

Table 3. OAER for visible light by season

of Helsinki, and informed consent was obtained after the nature of the study was explained. Simultaneously, we made measurements of the ambient radiation using identical instrumentation. These measurements allowed calculation of the numeric values of the OAER shown in table 3 [10]. We found no strong evidence of variation in Roa by job category, although some of the sample sizes were small. As noted previously, one of the principal factors which determines the value of the OAER is the reflectivity of the surface over which the individual works [19]. None of the job categories for this general population was associated with a characteristic work surface, such as for a waterman [19]. Rather, within each job category, the reflectivity of the work surfaces was highly variable. It was not surprising therefore, that no job dependency was found for this general population.

The fractional time spent outdoors, the use of eyewear, and the geographic modifier were established for an individual through the use of a personal history interview that reached back to the age of 30 years. Pilot studies demonstrated unacceptable reliability in recall in our population below about age 26. Another factor in this choice of 30 years was the length of the interview and the fact that from ages 16–29, people in this age group on the Eastern Shore typically changed jobs often. The age range for people in our population-based study is from 65 to 84 years. To implement the exposure model, we divided the day into two intervals: 10 a.m. to 4 p.m., and its complement (before 10 a.m. and after 4 p.m.). In each case, the total times reported outdoors or driving a car were apportioned uniformly over the corresponding interval. For instance if an individual reported spending 3 h outdoors between 10 a.m. and 4 p.m., we assumed that he was outside for one half hour in each of the 1-hour intervals between 10 a.m. and 4 p.m.

The effect due to glasses was based on photopically weighted transmission measurements made on a variety of sunglasses (prescription and otherwise; Teye 0.23, n 20, SEM 0.04) and regular prescription glasses (Teye 0.95, n 6, SEM 0.02). We found no statistically significant diminution in ocular exposure to visible light due to the use of hats [10].

Results

The cumulative ocular exposure in this population ranged from zero to 3.82 VMSY, with a median of 0.388 VMSY (fig. 1). Those for whom we calculated no exposure were those reporting less than 1 h per day outside in jobs or leisure activities for their lifetime between age 30 to the present. Although this group obviously has a finite exposure, our program categorizes them as zero.

Fig. 1. Distribution of cumulative ocular visible exposures. Total sample size was 2,453. MSY Maryland sun years.

Fig. 2. Distribution of average annual ocular visible exposures. Total sample size was 2,453. MSY Maryland sun years.

Table 4. Quartiles (milli-VMSY) of average annual ocular exposure to visible for population aged 65–84 in Salisbury, Md.

The average annual ocular exposure to visible radiation ranged from zero to 0.0931 VMSY, with a median of 0.009 VMSY (fig. 2). Quartiles of annual exposure are listed in table 4.

For the population aged 65–84, the cumulative exposure continues to increase with age by 0.06 VMSY per every additional year (p 0.0001) (fig. 3), as expected if exposure to visible light continues with aging. Average annual exposures, (fig. 4), however, are unchanged with age (p 0.885). Note that this is a cross-sectional view of our population rather than a statement of exposure progression.

Fig. 3. Cumulative ocular visible exposures by age group. MSY Maryland sun years. Fig. 4. Average annual ocular visible exposures by age group. MSY Maryland

sun years.

The distribution of average annual exposures is different among AfricanAmericans than among whites (fig. 5). While the median values are similar (0.009 for whites vs. 0.010 for African-Americans), the distribution in the upper quartile suggests higher exposures for African-Americans. The boundary for the upper quartile among whites is 0.017 vs. 0.021 for African-Americans. Racial differences in average annual visible exposures were statistically significant (p 0.001).

Females have substantially lower exposure than males (fig. 6). Their median value of average annual ocular exposure, 0.006, is significantly lower than that for males, 0.016 (p 0.0001). The ageand race-adjusted differences by gender are also significantly different.

Fig. 5. Average annual ocular exposure by race. MSY Maryland sun years.

Fig. 6. Average annual ocular visible exposures by gender. MSY Maryland sun years. Fig. 7. Average annual ocular visible exposures by education. MSY Maryland sun

years.

Fig. 8. Average annual ocular visible exposures by expressed photophobia. MSY Maryland sun years.