Prevention of the Adverse Photic Effects of Peripheral Light-Focusing Using UV-Blocking Contact Lenses

Abstract

purpose. Peripheral light-focusing (PLF) is an occult form of ultraviolet radiation (UVR) hazardous to the human eye. In PLF, obliquely incident light is refracted from the peripheral cornea to concentrated sites inside the anterior segment. In the current study, the directionality of this phenomenon for UVR and whether PLF is established in outdoor settings exposed to sunlight were investigated. The protection provided by a UV-blocking contact lens was also evaluated.

methods. UVA and UVB sensors were placed on the nasal limbus of an anatomically based model eye. The temporal limbus was exposed to a UV light source placed at various angles behind the frontal plane. PLF was quantified with the sensor output. The ensemble was mounted in the orbit of a mannequin head and exposed to sunlight in three insolation environments within the region of Sydney, Australia. PLF for UVA and UVB was determined with no eyewear or with sunglasses and commercially available soft contact lenses, with and without UV-blocking capability.

results. The intensity of UVA peaked at approximately 120° incidence, the level at which the UVB response was also at its maximum. The intensification of UVA was up to ×18.3. The intensity of PLF for UVA and UVB was reduced by an order of magnitude by a UV-blocking contact lens, whereas a clear contact lenses had a much lesser effect. Only the UV-blocking contact lens achieved a significant effect on UVA and UVB irradiance in the urban, beach, and mountain locales (P < 0.056).

conclusions. The results identify another type of sunlight hazard: the peripheral focusing of obliquely incident light. UVR from albedo (reflected ambient light) is capable of establishing PLF in the anterior segment, but this can be shielded by UV-blocking soft contact lenses. Sunglasses may be unable to shield oblique rays, unless side protection is incorporated. Contact lenses can offer UVR protection against all angles of incidence, including the peak-response angle. They can also protect the eye in settings in which the wearing of sunglasses is not feasible or convenient.

The biological duality of how light affects the eye is a paradoxical phenomenon. Although the visible wavelengths are vital to vision, the contiguous band of ultraviolet radiation (UVR) is implicated in many eye diseases. ^{1 2} Short-wavelength light is demonstrably hazardous and has great potential to induce phototoxic injury in ocular cells. A causative role of sunlight, particularly its UV component, has long been surmised in the etiology of the ophthalmohelioses (sun-related eye conditions) such as pterygium and cortical cataract. ³ However, whereas the hazard to the eye posed by UVR is generally accepted, the role of ultraviolet irradiance in the pathogenesis of these eye diseases is disputed. ^{4 5 6} The objections to the sunlight hypothesis are that it cannot explain the spatial specificity in some diseases, nor can it satisfy the requirement of photocytotoxic energy dosage. ⁷  

We have identified a mechanism that could mediate actinic damage in the anterior segment. ^{8 9} The extreme circumference of the corneal dome refracts oblique light to focal areas of concentrated light inside the eye. We have implicated this peripheral light-focusing (PLF) effect in the pathogenesis of pterygium ^{10 11} and age-related cortical cataract. ¹² Optical ray tracing has established that PLF can create areas of concentrated light in the nasal corneal limbus and lens cortex. ^{9 13} The intensity of light received by the distal limbus reaches levels 20 times higher than the intensity of temporally incident light. ¹³  

In theory, protective devices could protect against PLF. However, we have shown that PLF occurs over a range of incidence angles, including very oblique trajectories that originate from behind the eye’s frontal plane. ^{9 13} Although sunglasses are used to block UVR, most designs are inadequate for the protection of the eye from such side-on radiation. ^{3 14} The need for side protection of the eyes by sunglasses has been known for some time. In 1986 Silver ¹⁵ wrote: “A sunglass that permits significant transmission of toxic wavelengths does not fulfil its primary function, that is to protect the eye.” She went on to point out the importance of blocking unwanted light entering from the sides or above, creating a problem of distracting glare reflected off the back surface of a dark sunglass. ¹⁵ A useful alternative is a UV-blocking soft contact lens that our calculations indicate could attenuate side-on UVR and the PLF effect. ¹⁶  

In this study, we sought to confirm that PLF could be established under direct UV irradiation as well as in sunlight across a range of insolation environments. We used an anatomically based physical model eye to ascertain whether PLF has a directional preference, as indicated by our theoretical calculations ¹⁶ and whether a commercially available UV-blocking contact lens could provide effective protection from obliquely incident light. We also examined the relative efficacy of a standard sunglass design. 

Methods

PLF of UVR was evaluated in a physical model eye. UV sensors were placed at the nasal limbus to quantify the concentration of short-wavelength light in two settings. First, the model eye was exposed to discrete angles of light from a laboratory light source emitting UVA and UVB. Second, the model eye was exposed to ambient sunlight at various locales. In both conditions, the attenuation effect of clear and UV-blocking contact lenses was appraised. 

Model Human Eye

A model eye was fabricated with an aspherical contact lens used to simulate the cornea. The custom-designed 14-mm diameter contact lens (Gelflex Laboratories, Osborne Park, Western Australia, Australia) had surface curvature radii of 7.8 mm anteriorly and 7.7 mm posteriorly, with an asphericity of Q = −0.25. It was glued concentrically to a polymethylmethacrylate (PMMA) cylinder with a diameter of 12 mm and 4-mm height, making the effective diameter of the model cornea also 12 mm (Fig. 1) . The sides of the cylindrical base were painted black (opaque to visible and UV light) to avoid a stray-light influence on the measurements. A small hole in the base enabled access to the anterior chamber cavity so that ocular irrigating solution (phosphate buffered saline, PBS) could be injected with a hypodermic needle. The measured refractive index of the PBS solution was 1.338 (λ = 589 nm; WY1A refractometer; Xintian Instrument Corp., Guiyang, China) which is similar to that of the aqueous humor. 

In other experiments, we used a physical model eye with a spherical cornea that consisted of a plano, 0.1-mm thick, 14-mm diameter PMMA contact lens. The radii of surface curvatures were 7.8 mm anteriorly and 7.7 mm posteriorly, thereby giving the cornea parallel surfaces. The model corneal material had a refractive index of 1.473 (λ = 589 nm; WY1A refractometer; Xintian Instrument Corp.) and acted as a UV cutoff filter with 100% absorption for wavelengths shorter than 270 nm, 50% at approximately 290 nm, and decreasing to 4% in the visible spectrum. No UV-absorber additives were present in the model cornea. 

Measurements at Specific Angles of Incidence

UV Sensors

Commercially available UV sensors were used to measure the intensity of incident light on the eye model and emerging light at the limbus. The UVA and UVB sensors were constructed from wide-bandwidth solid-state light detectors that were cannibalized from a UVA radiometer (Solarmeter 5.0; Solartec, Harrison Township, MI) and a UVB radiometer (Solarmeter 6.0; Solartec). The UVA light detector was a GaAsP photodiode with a stated response in the range 280 to 400 nm (UVA waveband: −3.5 dB at 320 nm and −9.6 dB at 400 nm for 0 dB at 370 nm) with a spectral sensitivity of 0.1 mW/cm². The active area was 0.8 × 0.8 mm. The UVB detector was an SiC photodiode with a stated response in the range of 280 to 320 nm (UVB waveband: 0 dB at 285 nm and −9.6 dB at 320 nm) with a spectral responsivity of 0.1mW/cm². The active area was 0.3 × 0.3 mm. The acceptance angle of both detectors was approximately 60°. The photodiodes were negatively biased. The relative irradiance was based on the photodiode output current converted with resistive circuitry into an output voltage that was measured to an accuracy of ±0.01 mV with a multimeter (model 87III; Fluke, Everett, WA). 

Light Source

For UV illumination of the anterior eye model, a mercury arc 350-W light source was used (model 6286; Thermo Oriel, Stratford, CT). This light source produced a collimated beam of light in a range of 200 to 2500 nm wavelength. An aperture diaphragm (10 × 0.8-mm slit) produced a beam of parallel light that was used to illuminate the anterior eye model from different angles of light incidence. The incident angle (measured from the sagittal plane) was varied from 90° to 130° in 5° steps. 

Layout

The eye model was placed upright on a turntable at its rotational axis and positioned in the optical path of the collimated beam from a fixed light source (Fig. 1) . The collimated beam was pointed at the rotational axis of the turntable to ensure that the model limbus was continuously illuminated by the light source at all incident angles. An angular scale on the turntable displayed the angle of rotation, configured to correspond with the angle of light incidence. 

Nasal Limbus Irradiation

The photodiode was placed adjacent to the nasal limbus of the anterior eye model and pointed toward the opposite temporal limbus where the light was incident. Figure 1 illustrates the layout of the eye model, light source, and photodiode. Because of the surface curvature of the eye model, there was a gap between the photodiode and the limbus of between 1.3 and 1.5 mm (UVA sensor) and 3 mm (UVB sensor). As a consequence, the sensors, which had an acceptance angle α = 60°, recorded not only light emerging at the limbus but also light scattered and refracted by the cornea, especially at lower angles of incidence. This was minimized with black masking tape. The relative intensity at selected angles was calculated as the ratio between the photodiode output voltage measured at the nasal limbus and the voltage measured at the point of incidence at the temporal cornea. 

Contact Lenses

The effect of contact lenses on PLF was determined with commercial soft contact lenses. Designs of the same material in the same dioptric power were used. Both a conventional non-UV-blocking contact lens (modified etafilcon A; Vistakon/Johnson & Johnson, Jacksonville, FL) and the same material with UV-blocking additives (etafilcon A; Vistakon/Johnson & Johnson) were used. The spectral transmission curve of the UV-blocking lenses (supplied by John Enns, Vistakon/Johnson & Johnson) has a very low transmission of approximately 2% to 3% in the spectral range between 280 and 370 nm. The transmission increases steadily from 3% to 95% in the wavelength range of 370 to 380 nm. There is a relative peak of 40% transmission in the UVC range at approximately 260 nm, but this peak is not of concern to ocular health, because solar UVC is mostly absorbed by the atmosphere. The non-UV-blocking contact lens has a UVA transmission of 96% to 94% (320–400 nm) and a UVB transmission of 94% to 96% (280–320 nm). ¹⁷  

The contact lenses had to be modified to avoid double-passage through the material due to the position of the UV sensor, which was close to the nasal limbus but mounted external to the cornea. Using a sharp steel blade, we truncated a segment from the nasal margin of the contact lens, which was then placed on the anterior eye model so that the photodiode faced the cut edge of the lens. The measurements were performed rapidly to minimize dehydration of the contact lens. Lenses had an optical power of −1.00, −2.50, and −3.00 D. Four determinations were made for each angle and averaged. 

Qualitative Demonstration of UV Blocking in Porcine Eyes

In qualitative experiments, porcine eyes were used to demonstrate the effect of UV-blocking and non-UV-blocking contact lenses on the PLF phenomenon. The porcine eye was used because its anterior eye geometry is a useful approximation to that in the human eye. ¹⁸ Porcine eyes were obtained fresh from the abattoir and transported to the laboratory on ice. The 350-W mercury arc light source was used to illuminate the eyeball mounted in a holder. The total irradiated power within the wavelength range of 280 to 400 nm of the collimated beam of light was calculated to be approximately 3 W. The beam of light was then successively passed through a UV-transmitting glass filter (U340; T _max at 340 nm, full-width half-maximum of 85 nm; Hoya, Tokyo, Japan) and through a single or triple slit of width 0.2 to 0.6 mm to detect green-stimulated luminescence. The sheet of light was directed parallel and perpendicular to the frontal plane of the porcine eye. The anterior chamber of the model eye was injected with a mixture of 14 mg/mL sodium hyaluronate (Healon GV; Pharmacia, Peapack, NJ) and sodium fluorescein (10⁻⁶–10⁻⁵ g/mL). The cornea and anterior chamber were observed, and the fluorescence recorded with a digital camera (model DSC-F505; Sony Corp., Tokyo, Japan). We subsequently recorded the effect of placing contact lenses on the passage of UV light through the anterior chamber. 

Measurements of PLF in Sunlight

Mounting System

A mannequin head is a reasonable approximation of in vivo exposure. ^{19 20} A mannequin head made from white Styrofoam (Dow Chemical Co., Midland, MI), approximating white facial characteristics, was mounted on a camera tripod set to a height of 1.60 m. The UVA and UVB light detectors were the same photodiodes as those used in the previous measurements and had an angle of acceptance of α = 120°. The photodetector selected for either UVA or UVB measurements was attached to the model eye, which was inserted into the left eye socket of the mannequin head (Fig. 2) . The model eye-photodetector ensemble was mounted so that it could be rotated around the optical z-axis. The nasal limbal area and photodiode were covered with black masking tape to shut out stray light. 

Directionality of Sensors

To gauge the directionality of the UVA and UVB detectors, ^{21 22} we exposed them to sunlight at various angles of elevation and orientation to the sun. The direction normal to the meter surface was elevated from angles of 0°, 30°, 45°, 60°, and 90° to the horizontal plane. At each position the detectors were rotated counterclockwise around a rotation angle θ = 0° to 270° where angle θ = 0° corresponds to the detector being pointed toward the sun, and θ = 180° corresponds to the detector pointed away from the sun. Also determined at the time of measurement were the solar zenith angle ξ_s defined as the angle between the sun and the vertical, and ξ_s = 90° − γ_s where γ_s is the solar elevation defined as the angle of the sun above the horizon. 

Nasal Limbus Irradiance

For measurement of PLF, the mannequin head was positioned in such a way that the sun illuminated the temporal limbus (Fig. 2) , and PLF was established at the nasal limbus on the photodetector face. The photodetector axis, initially oriented horizontally, was adjusted so that the photodiode output was at its maximum. This procedure ensured that the maximum PLF effect was achieved. 

Protective Eyewear Tested

Light-focusing in the model eye was determined under four conditions. The mannequin head was either unprotected, or wearing non-wraparound sunglasses, clear non-UV-absorbing contact lenses, or UV-blocking contact lenses. The contact lenses were the same design as used previously. Measurements were performed in three insolation environments: in an urban area, the campus of The University of New South Wales, with grass-covered ground surrounded by trees and three-story buildings, within 10 km of the Pacific Ocean; at a white-sand seaside beach, Cronulla Beach, 30 m from the Pacific Ocean; and in the mountains, Blue Mountains, 700 m above sea level, eucalyptus-covered hillsides, 120 km from the Pacific Ocean. All environments were situated within Sydney, Australia, and its hinterland (latitude 34.0° south, longitude 151.0° east). 

Results

PLF at Specific Incidence Angles

Aspheric Eye Model.

The UVA response increased with larger angles of incidence, peaking at an incidence angle of 120° with a significant falloff for larger angles (Fig. 3) . The UVB response showed a monotonic increase with larger angles of incidence but did not display a peak in the range tested (Fig. 4) . At approximately 105° incidence, the estimated UVA intensification factor was calculated to be ×18.3. 

Spherical Cornea Eye Model.

Comparable results were found in the spherical model eye. At the angle of peak intensity, there was approximately an order of magnitude reduction in the UV concentration when the model cornea was covered by the UV-blocking contact lens (Acuvue; Vistakon/Johnson & Johnson). The UVA response peaked at 120° and the UVB response, although somewhat flatter, peaked at 110° (data not shown). The UVA amplification factor was found to be ×15. 

PLF in Sunlight

The greatest intensity of UVA or UVB light was measured when the detectors were pointed directly upward (90° normal to the horizontal plane), regardless of the insolation environment (data not shown). In the horizontal position, as the sensor was rotated, the output decreased the most in the first 90° by similar amounts for the urban (U) and beach (B) sites, and less for the mountain environment (M). The UVA irradiance decreased by 40% ± 4% (U and B) and 5% (M); the UVB declined by 36% ± 2% (U and B) and 1% (M). Further rotation had only a slight effect on UV irradiance. The variation at other angles of elevation was larger (data not presented), but we used only the horizontal position for PLF readings. The measurements were made between 11:00 AM and 3:00 PM Eastern Daylight-Saving Time, east coast of Australia (coordinated universal time [UTC] + 10 hours). During these times, the solar zenith angle ξ_s was between 26° and 13° on cloud-free days. Under these conditions with the sensor pointed horizontally, scattered light was the main form of light detected, because the solar elevation angle γ_s (64−77°) was outside the acceptance half-angle (60°) of the UV detectors. 

Influence of UV-Blocking Contact Lenses

The application of UV-blocking contact lenses in the aspheric model eye had a profound effect on PLF of both UVA and UVB, reducing the PLF intensities by an order of magnitude or more (Figs. 3 4 , lower curves). In contrast, the presence of a clear, non-UV-blocking soft contact lens had only a slight effect on the focusing intensity of both UVA and UVB (Figs. 3 4 , middle curves). The UV-blocking characteristics suggest there is an angular shift in the peak response of up to approximately 5°, due to the application of the contact lens to the model cornea. In the spherical corneal model, the maximum limbal illumination for all UV-blocking lenses tested (optical powers: −1.00, −2.50, and −3.00 D) was achieved at an angle of incidence approximately 3° to 5° lower than without a UV-blocking lens. This is illustrated by the results for the −2.50-D lenses shown in Figure 5 for the combined UVA and UVB concentrations at the nasal limbus. There was also a shift in the incident angle of peak intensity when the non-UV-blocking contact lens was worn. 

The contact lens UV-blocking effect was confirmed by two additional methods. First, in the isolated porcine eye ex vivo, a non-UV-blocking contact lens placed on the eye did not alter the amount of anterior chamber fluorescence. However, the UV-blocking contact lens dramatically attenuated the UV light and extinguished all fluorescence from the anterior chamber. Second, based on the emittance spectrum of the light source, the transmission characteristics of the various optical media, and the spectral response of the detectors, we estimate that the UV-blocking contact lens can reduce the focusing effect by at least 69% for UVA and 96% for UVB. 

UV-Blocking Contact Lenses and Sunlight

Because of the presence of significant albedo, the intensity of peripherally focused UVA light in the mountain and beach environments exceeded that in the urban area by 67% and 83%, respectively (Fig. 6) . The UVB light in the mountain and beach environments exceeded that in the urban area by 100% and 240%, respectively (Fig. 7) . In all environments, sunglasses, and clear contact lenses provided little or no protection from peripherally focused UV light (Figs. 6 7) . However, the presence of a UV-blocking contact lens significantly reduced the intensity of UVA PLF at the nasal limbus (ANOVA; P < 0.02). This difference was maintained across all three environments tested. Similarly, the UV-blocking contact lens reduced the intensity of UVB light-focusing (ANOVA; P < 0.056) in all three environments. The solar zenith angle ξ_s was 13° at the beginning and 24° by the end of the urban and beach measurements and 15° at the beginning and 22° by the end of the mountain measurements. 

Discussion

Our measurements confirm theoretical predictions of PLF in the human eye. ^{9 13 16} Using an anatomically based model eye, we found significant focusing in the nasal corneal region of temporal light incident at oblique angles beyond the frontal plane (Figs. 3 4 5) . We also confirmed the directionality of PLF with a peak of intensity as foreshadowed by model eye computations. ^{13 16} We found a peak effect for UVA at an incidence angle of 118 ± 3° and at 122 ± 3° for UVB. This compares with intensity peaks predicted at 104° for UVA (λ = 365 nm) and 107° for visible light (λ = 555 nm). ¹⁶ To check the validity of the physical model eye, we compared our results with the theoretical intensification factor of peripherally focused UV light. The intensification factor for the physical measurements were of the same order of magnitude as the theoretical calculations. The intensification factor of peripheral light at the nasal limbus is calculated to be ×22.5 for visible light and ×8.5 for UVA (365 nm). ¹⁶ In the physical model eye without a contact lens, we found UVA intensification factors at the nasal limbus of ×15 to ×18.3. 

Our measurements demonstrate that UV-blocking contact lenses are effective in significantly reducing PLF. This was found in tests with a light source at various oblique incidence angles as well as outdoors with the model eye exposed to sunlight. The attenuation of PLF was around an order of magnitude for UVA and UVB light for oblique angles beyond the frontal plane (incidence >90°). The oblique light path through a contact lens is calculated to increase absorption by 6% for UVA. The optical powers tested did not significantly affect the attenuating properties of the UV-blocking contact lens. The reduction of UV-induced PLF was confirmed when a UV-blocking soft contact lens extinguished fluorescence from the porcine anterior chamber, indicating substantial blocking of obliquely incident UVR. 

In contrast, clear contact lenses (non-UV-blocking) had only a slight effect on PLF, mainly affecting the intensity at lower angles of incidence. The small, almost insignificant amount of protection provided by clear contact lenses may be attributable to the thickness and shape of the periphery of the contact lens. The present findings and calculations presented elsewhere ¹⁵ indicate that the lens periphery alters the nature of light-focusing but does not attenuate the PLF effect. Our measurements suggest that peripheral lens thickness and contact lens design differences may shift the incidence angle of peak focus intensity, depending on the different refractive powers. We point out that the focused light emergent from the nasal limbus was measured after a double passage through the cornea, rather than after a single passage followed by partial absorption. ^{13 23} As a consequence, our results would underestimate light concentration in the living human eye. However, this was assumed not to affect the comparison of the relative effect of contact lens wear, because the incident light traversed the cornea twice in both non-lens- and lens-wearing conditions. 

Standard sunglasses (front-face only, non-wraparound type) provided no protection at all from PLF in sunlight (Figs. 6 7) . These findings expose a deficiency in UVR protection by sunglasses and spectacles that is not widely recognized. ²⁴ Even for front-on irradiance, the lens material may have inadequate absorption for UV protection, ^{15 25 26} and sunglasses must be correctly worn for optimum performance. ²⁷ However, most sunglasses even when properly fitted are ineffective against obliquely incident irradiation. ^{3 24} Our outdoor measurements indicate that diffuse UVR can generate significant levels of PLF, and we conclude that traditional sunglass designs offer limited protection against PLF from this unsuspected source of UVR. We measured UV irradiance on cloudless days during the midday to midafternoon at a time of year when the solar zenith angle was small. In eastern Australia where our measurements were taken, these conditions tend to favor lower diffuse UV irradiation. ²⁸ On cloudy days, Kimlin et al. ²⁹ found that UV irradiation on the nose and face increases as a result of scattering of UV by cloud cover, although this finding may apply only in subtropical latitudes (27.5° south, 151.9° east) Therefore, our results may underestimate PLF under cloudy conditions. 

It is worth noting that Sakamoto et al. ³⁰ used a mannequin head with an anatomically average Japanese bone structure and exposed it to direct solar UVR. They found that the total UV surface irradiance in the temporal eyelid area was 12% higher than in the nasal portion. Between 12:10 and 3:02 PM, the temporal eyelid irradiance increased by 36%, whereas the nasal eyelid irradiance remained fairly constant. The imposition of clear glass spectacles reduced the temporal, central, and nasal surface intensities by only 1.5%, 6.4%, and 4.6%, respectively. In contrast, our mannequin head mimicked white facial anatomy (Fig. 2) . Dosimetry studies of the white facial erythemal UV distribution show that, although the orbital area is relatively spared during outdoor exposure to direct UVR, ^{19 20 24 29 31} the temporal eyelid is relatively more exposed with an irradiance more than twice the nasal side level. ^{30 31} This explains why the temporal cornea has the highest UV irradiance when contact lens dosimetry is used. ³² In our system, the main contributor to PLF was diffuse UVR, which predominates above 30° solar elevation angle ³³ (our γ_s was 64–77°) and the photodetector was aligned horizontally with half-angle of acceptance 1/2α < γ_s. To test the effectiveness of direct irradiance, we rotated the mannequin head by 90° to face the sun and confirmed a similar PLF effect for UVA and UVB (data not presented). This agrees with the measurements of Narayanan et al., ³⁴ who used a similar apparatus. Increased albedo from highly reflective ground, such as sand, can increase irradiance of the face and orbit ^{14 19 24} and possibly increase PLF. It appears that both direct and diffuse UVR can concentrate light at the nasal limbus. 

Our findings indicate that UV-blocking contact lenses may provide a supplementary but important means of protecting eyes from chronic exposure to high levels of UV light. Consequently, the possibility arises that the risk of eye diseases such as pterygium and early cortical cataract may be reduced by wearing UV-blocking contact lenses. We confirm theoretical predictions that the UV-blocking contact lens can attenuate the amount of peripherally focused UV light by up to an order of magnitude, although our results are likely to be underestimates of the effect of PLF in vivo. Our data indicate that the UV-blocking lens was effective against UVA and UVB in different environments, even in the mountain area, where UV irradiance was highest, presumably from high levels of scattered UVR. ³⁵ Contact lenses may have an additional advantage when the wearer is under tree shade where the diffuse UVA irradiation is still significant ^{36 37} and there is a higher proportion of UVB than in direct sunlight. ³⁸  

This study of peripheral light-focusing highlights the need to review current standards of UV protection by sunglasses and contact lenses. The inadequacy of most sunglass designs against PLF and the protection provided under different insolation conditions merit further study. The facial anatomy ³⁰ and skin pigmentation ²⁴ may predispose some wearers to a higher risk of PLF. Some groups of outdoor workers such as sailors ³⁹ are at risk of chronic overexposure to UVR. ⁴⁰ Spectacles are not feasible for some outdoor activities such as surfing, ⁴¹ and contact lenses may be the only effective protection against diseases such as pterygium. ⁴² Our results imply that diffuse UVR can initiate PLF in both the UVA and UVB wavebands and that corneoscleral optics is an effective modulator of light concentration. The preferential focusing found in the current study may be wavelength dependent, and although the action spectrum of PLF is unknown, protective contact lenses must have a specific formulation to block UVR, because simple tints are inadequate. ⁴³ Sliney ^{24 35} has pointed out that dark sunglasses may undermine UV protection by disabling the eye’s natural reactions to sunlight such as squinting and may initiate counterproductive responses such as pupil dilation. Standard sunglasses reduce direct UV irradiance of the eye but do not eliminate the inferonasal concentration of light. ³³ To counter UV-induced PLF, sunglasses should have adequate side protection to act as a horizon shade, as proposed by Urbach. ¹⁹ There is a need to revisit Sliney’s concept ³³ of protective eyewear as “spatial filters” and to reevaluate sunglass design to incorporate protection against PLF. Further study of the effectiveness of protective visual devices against PLF due to direct and scattered UVR is indicated. 

 Supported in part by grants from Vistakon/Johnson & Johnson, the Australian Government through the Cooperative Research Centre Scheme, and the Ophthalmic Research Institute of Australia.

 Submitted for publication April 16, 2002; revised October 17, 2002; accepted November 4, 2002.

 Commercial relationships policy: F (VAK, AH, MTC); N (LSK).

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Measurement of the limbal focusing intensity in the model eye (not to scale). Collimated light strikes the temporal cornea (left) at an angle to the sagittal plane (at 110° in this example) and is refracted across the anterior chamber to a focus at the nasal limbus (right). A UV detector measures the relative intensities at different angles of incidence, which are indicated by a turntable underneath the model eye (not shown). The anterior chamber is filled with PBS through a hole in the cylindrical base behind the cornea. Measurements were repeated with a contact lens fitted onto the anterior surface of the cornea.

Figure 2.

 

Mannequin head fitted atop a camera tripod showing the model eye and UV sensor mounted in the left orbital area.

Figure 3.

 

Intensity of peripheral light (arbitrary units) focused at the distal limbus of the aspheric model cornea, measured with a UVA detector (▪). A conventional −1.50-D contact lens had a slight effect at lower angles (○). The intensity was significantly reduced in the presence of a contact lens of the same −1.50-D optical power and material but with a UV-blocking additive (▴).

Figure 4.

 

Intensity of peripheral light (arbitrary units) focused at the distal limbus, measured with a UVB detector (▪). A conventional −1.50-D contact lens had a slight effect (○). A contact lens of the same −1.50-D optical power and material but with a UV-blocking additive (▴) caused a significant reduction in the intensity but did not show a peak.

Figure 5.

 

The combined UVA and UVB concentration at the distal limbus of the spherical model cornea (▪). At the angle of peak intensity, the UV-blocking contact lens (▴) caused a reduction in the UV concentration of approximately an order of magnitude. The non-UV-blocking contact lens (○) caused a shift in the incident angle of peak intensity compared with the non-lens-wearing situation.

Figure 6.

 

Relative intensity of peripherally focused UVA measured using the mannequin head model in three insolation conditions. In all environments, the UV-blocking contact lens produced significantly lower intensities than no eyewear (*P < 0.02). Bar, SEM.

Figure 7.

 

Relative intensity of peripherally focused UVB measured using the mannequin head model in three insolation conditions. In all environments, the UV-blocking contact lens produced significantly lower intensities than no eyewear (*P < 0.056).

The authors thank Therese Pham for technical assistance.