Blue-light filtering spectacle lenses have received increased attention in recent years because of their potential to reduce the effects of light scatter and chromatic aberrations, and the possible protection they offer against photo-oxidative effects induced by photons of relatively high energy.1–4 Indeed, the accumulation of the retinoid fluorophore N-retinylidene-N-retinylethanolamine within the retinal pigment epithelium is expected to increase the vulnerability of the retina to high-energy blue radiation.3–6 Spectacles, contact lenses, and intraocular lenses have all been proposed for filtering blue light. However, finding the balance between effectively reducing blue-light hazards and maintaining or improving visual functions remains a challenge, and one debate in the literature.6–18 Some authors,6–8,10,18 found visual acuity improvement and/or contrast sensitivity improvement when using filters which filter blue light both in normal subjects and in subjects affected by cataract, age-related macular degeneration, or other retinal diseases. Remarkably, some authors did not find any correlation between the CS improvement and the subjective preferences for the filters.7 Mahjoob et al.9 recently discussed the effect of a yellow filter on visual acuity and contrast sensitivity under glare for various ages, though the improvement was only significant in older subjects aged 51–60 years. Other papers reported neutral or negative effects of tinted lenses and filters on visual performances.11–16 In general, there is little objective and conclusive evidence that tinted lenses or filters improve visual function, and it is unclear whether lenses with specific spectral characteristics are better than any others. The only consistent reported effect is an improvement in dark adaptation,17 which would be expected intuitively since filters reduce the intensity of the incident light during exposure. In this context, blue-filtering Drive lenses (Hoya, Japan) were recently proposed as anti-glare filters for car drivers, with the main aim of reducing distracting glare from the dashboard, streetlights, and oncoming traffic when driving at night. At the same time, these lenses are expected to show an improvement of contrast and brightness perception, even in poor light and bad weather.

To gain further information on this issue, we performed a comparison between two Drive lenses and a clear lens, whose optical properties were also preliminary characterized. We used pshychophysical measurements to evaluate whether spectacle lenses incorporating these filters offered any real benefit.

In SVL group, 11 males and 9 females (age range: 24–55 years; mean±std dev: 47.8±14.7 years) were enrolled. In PAL group, 10 males and 10 females (age range: 46–73 years; mean±std dev: 59.5±7.2 years) were enrolled. Demographic characteristics of the whole sample, and the two groups separately, are reported in Table 1.

The optical transmittance spectra of three plano lenses (clear, STD, and PRO) are reported in Fig. 2. A cutoff wavelength can be defined as the wavelength corresponding to 50% transmittance. For the clear lens the cutoff is found at (405±2) nm, while the cutoff of STD and PRO is shifted to longer wavelength and it is found at (426±2) nm. Both the STD lens and the PRO lens have a transmittance that does not substantially depend on the thickness of the lens and, therefore, on the optical power. Indeed, the longpass character of these lenses with cutoff at (426±2) nm is due to a bulk internal layer of fixed thickness that the manufacturer has called “high-definition treatment”. In Fig. 2, the spectrum of human macular pigment is also shown as taken from the optical density reported in Werner et al.28 It will be commented in the discussion section.

For each type of lens (clear, STD, and PRO), the comparison between the data collected at baseline (1st day) and the corresponding data collected post-treatment (15th day) showed no significant differences for each condition. Since the measurements did not statistically change during the two weeks of wear, in the further analyses of the data only the values measured at the end of the second week were taken into consideration for each condition (results not showed). Since all the participants received the clear lens as the first lens (Fig. 1), after the third phase the visual analyses were repeated with the clear lens. As no differences were found between the data obtained with the clear lens during the first phase and the data obtained at the end of the project, only the results of the first phase were taken into consideration in the following. Fig. 3 shows the mean values of measured BCVAphotopic, BCVAmesopic, and BCVAglare in the three experimental conditions both for the PAL group (Fig. 3a) and the SVL group (Fig. 3b). No substantial differences were found in BCVAphotopic between Drive lenses (STD and PRO) and clear lenses for either the PAL group or the SVL group. The BCVAmesopic improved significantly only for STD lenses in the PAL group (from 0.15 to 0.12, p<0.05), while no substantial difference was observed for SVL group. Finally, BCVAglare improved with Drive filters compared to the clear one (p<0.05) from 0.03 to -0.02 (STD) and to -0.01 (PRO) for PAL and from -0.08 to -0.12 (STD and PRO) for SVL. Means of measured data of CS are reported in Fig. 4a (PAL group) and b (SVL group) as a function of the angular frequency. CS showed statistically significant improvements with the PRO lens respect to the clear lens only at 6 cycles/deg for the PAL group and at 1.5 cycles/deg for the SVL group (p<0.05). Conversely, the difference between STD and clear lenses become evident and statistically significant at 6, 12, and 18 cycles/deg in the PAL group (p<0.05) but only at 12 cycles/deg in the SVL group.

Figure 3.

Mean values of measured binocular BCVAphotopic, BCVAmesopic, and BCVAglare for the three types of lenses in PAL group (a) and in SVL group (b). Asterisks indicate that the observed difference compared to clear lens is statistically significant (p<0.05, grey asterisks for STD compared to clear lenses, black asterisks for PRO compared to clear lenses).

(0.13MB).

Figure 4.

Mean values of binocular contrast sensitivity for the three types in PAL group (a) and in SVL group (b). Asterisks indicate that the observed difference compared to clear lens is statistically significant (p<0.05, grey asterisks for STD compared to clear lenses, black asterisks for PRO compared to clear lenses).

(0.17MB).

Table 3 shows the corresponding data for each type of lens, including also standard deviations and p-values. Concerning the comparison between the two types of Drive lenses (STD and PRO), both for PAL wearers and for SVL wearers, the 95% confidence interval of the differences was evaluated for each measured variable (conditioned to the value observed with the clear lens). The estimated differences were found to show a 95% confidence interval which contains the null value. The results of the particpants’ subjective choice are reported in Table 4. Most participants in the PAL group chose Drive lenses (15 subjects out of 19, 79%) with only four subjects electing to keep the clear PAL lenses. Drive lenses were also the first choice of participants in the SVL group (12 subjects out of 20, 60%). The difference in the frequency distribution between PAL group and SVL group was not significant (χ2=2.31; n.s.).

The first part of this discussion concerns the comparison between the optical properties of the Drive lenses and the clear lens. Three main aspects are discussed: (i) the lower number of transmitted photons in the visible range by the Drive lenses, (ii) the lower chromatic aberration of the Drive lenses and, mostly, the decrease of the chromatic aberration of the eye, and (iii) the decrease of intraocular scattering. Concering the number of transmitted photons in the visible range, a quantitative evaluation depends both on the shape of the transmittance spectrum of the filter under investigation and also on the spectral distribution of the emission of the light source. For Drive filters, this reduction of transmitted light compared to the clear lens can be evaluated as follows. If S(λ) is the emission spectrum of a light source (given in number of photons as a function of wavelength λ), the ratios

Concering chromatic aberration, the shift of the cutoff from (405±2) nm to (426±2) nm is expected to reduce the chromatic aberration of the human eye. Longitudinal chromatic aberration of the eye is reported by many authors to be of the order of about 2 diopters when considering the whole visible range.31–35 Based on these data reported in the literature, longitudinal chromatic aberration is reduced of about 0.3 diopters when reducing the spectral range form 405–750nm (range of transmission of the clear lens) to 426–750nm (range of transmission of the Drive lenses). Concerning the aberration of the lens itself, the longitudianal chromatic aberration of the clear lens and that of the Drive lenses can be evaluated taking into consideration their Abbe number V and their refractive index at a specific wavelength (Table 1). The difference between clear and Drive lenses was found to be more than one order of magnitude lower than the reduction (∼0.3 diopters) of the longitudinal chromatic aberration of the eye.

Intraocular light scattering can be generated by different components of the eye and it can show different characteristics.19–21,36,37 For example, backward light scattering in the crystalline lens was found to vary with wavelength λ into the nucleus as the reciprocal of the fourth power (λ−4) as predicted by the Rayleigh model.37 Conversely, in the superior layers and for forward directions in the nucleus, light scattering was found to be less dependent on wavelength.37 Intraocular scattering was also found to depend strongly on pigmentation of the eye and age, scattering showing nearly perfect λ−4 dependence only in young and well-pigmented eyes.21 Even if the overall dependence on wavelength can deviate from the perfect λ−4 law, intraocular light scattering is known to typically increase at the shortest wavelengths in the visible range. The effects begin to manifest at the lowest wavelengths in the age period corresponding approximately to the SVL group of this work and then extend to longer wavelength with age. Therefore the Drive lenses, which attenuate light below (426±2) nm, are expected to play a role in reducing intraocular scattering, mainly in the PAL group, but eventually also in the SVL group. The PRO lenses show a weak additional optical band centred at 450nm (Fig. 2), which does not appear in the STD spectrum. This band is due to a surface treatment that differs from the STD lens. Interestingly, this band is very close to the maximum of absorption of the macular pigment of the human eye (also reported in Fig. 2 for comparison), which peaks at about 460nm.28,37,38 Although the function of macular pigment remains uncertain, it is suspected of acting as a natural optical filter, absorbing part of the blue light, thus partially reducing chromatic aberration and intraocular scattering.39–43 In the spectral range of absorption of the macular pigment, between about 430nm and 500nm, it appears that the eye naturally protects itself from scattering and aberration, at least for young and healthy subjects. However, it could be interesting to investigate whether the PRO filter, with its additional band at about 450nm, has positive effects in case of a possibe decrease of macular pigment density in elderly or pathological subjects.

The optometric findings are now discussed. A first result of this study shows that no substantial differences were found in BCVAphotopic between Drive lenses (STD and PRO) and clear lenses for either the PAL group or the SVL group (see Fig. 3 and Table 3). A comparison with data of visual acuity on healthy subjects reported in the literature is not trivial because the optical characteristics of the filters discussed in the literature are different. For example, Zigman et al.6 found visual acuity improvement in healthy subjects when using a filter with absorption below the cutoff of 480mm. The filters under investigation in this work, which are found to produce no substantial visual acuity improvement, show a cutoff at much shorter wavelength than 480nm (426±2nm, Fig. 2). However, this should not lead to the conclusion that the improvement occurs when the cutoff increases. Indeed, Barron and Waiss11 measured visual acuity differences using a Corning CPF527 filter and a clear lens. Notwithstanding the relatively long cutoff wavelength, no significant differences were observed by these authors in average visual acuity compared to clear lenses. They concluded that a possible subjective impression of better vision was not equivalent to the conventional definition of visual acuity as measured with a high contrast chart. Other studies concern cataract or low-vision subjects. Also in these cases, different types of filters were studied, mixed results can be found in the literature, and there is little conclusive evidence that blue-filtering lenses improve visual acuity.6,7,10,11 Concering the Drive lenses under investigation in this work, the reduction of chromatic aberration and intraocular scattering due to the shift of the cutoff (discussed above) are not accompanied by a significant improvement in BCVAphotopic. Also the additional band at 450nm of the PRO lens does not seem to produce detectable effects in BCVAphotopic.

The other optometric analysis carried out in photopic condition is the CS assessment. The first point concerning CS is the comparison between clear lens (black diamonds in Fig. 4) and STD (gray circles and gray asterisks in Fig. 4). In general, as can also be observed in Table 3, all mean data of CS taken with STD type are equal or better compared to the corresponding value for the clear lens. No statistically significant improvement at the lowest angular frequencies is found in both groups (PAL and SVL). In the PAL group the differences become evident and statistically significant at 6, 12, and 18 cycles/deg (p<0.05; Fig. 4a and Table 3). In the SVL group the difference between STD and clear lens is significant only at 12 cycles/deg (Fig. 4b). Data for the other Drive lens (PRO) show that all mean data of CS taken with PRO type are equal or better compared to the corresponding value for the clear lens (Table 3, Fig. 4). CS is statistically better than CS with clear lenses at 6 cycles/deg in the PAL group (black asterisk in Fig. 4a), and at 1.5 cycles/deg in SVL group (black asterisk in Fig. 4b). In the literature, mixed results are reported. For example, Zigman found that a yellow filter improved significantly CS of subjects with healthy eyes in the 3–12 cycles/deg range of spatial frequency.8 On the contrary, blue-light filtering spectacle lenses did not significantly affect the contrast sensitivity in a recent study of Leung et al.16 In the present work, a certain degree of improvement of photopic CS by Drive lenses can be inferred compared to the clear lens. Compared to BCVAphotopic, the measurement of CS is a more complete assessment of visual performance,44,45 thus making the effects due to the shift of the cutoff detectable, at least at some angular frequency. This improvement can be mainly attributed to a reduction of intraocular scattering, which is known to play a major role in CS reduction.46,47 On the other hand, the reduction (about 0.3 diopters) of the eye chromatic aberration with the Drive lenses is not expected to play a role in improvimng CS. Indeed, the spectral luminous efficiency of the human eye in photopic condition is relatively low at the shortest wavelengths in the visible range.30 Experimental evidence was provided by Kruger et al.,33 who found no differences when measuring CS with or without a neutralizing lens in place. The neutralizing lens was designed to focus all wavelengths approximately in the same plane in the eye so that there was no longitudinal chromatic aberration. As the effect on CS is absent by neutralizing the aberration, it is expected to be absent even chromatic aberration is partially reduced (∼0.3 diopters), as in the case of Drives lenses. As can be observed for the PAL group in Fig. 4a, it happened at several angular frequencies that the STD lens was better than the clear lens (three grey asterisks in Fig. 4a) compared to how many times it happened with the PRO lens (one black asterisk in Fig. 4a). However, the direct comparison between the two lenses did not support any significant difference between STD and PRO. Indeed, when comparing the two Drive lenses, based on 95% confidence interval for CS data, there was no statistical evidence of difference between them. Therefore, the possible positive effects of the additional band at 450nm is not sufficiently marked to generate, on average, a detectable variation in SC tests, even in the older subjects of the PAL group.

In addition to BCVAphotopic, BCVAmesopic has been studied in this work and, although clinically small, a statistically significant improvement compared to the clear lens was observed. BCVAmesopic changed in PAL group from 0.15 logMAR to 0.12 logMAR in the case of STD (p<0.05) and to 0.13 logMAR in the case of PRO. No substantial differences were observed between the three lenses for BCVAmesopic in the case of SVL group. It is well known that the curve of the spectral efficiency of the human eye depends on the environmental conditions. In mesopic condition, the maximum of the curve shifts toward the blue compared to the maximum in photopic condition (Purkinje shift). For this reason, intraocular scattering of blue light, especially in older subjects, acquires more relevance in mesopic conditions. This might explain the positive role of the Drive lenses in improving BCVAmesopic of the PAL group. Although it is expected both a reduction of the intraocular scattering and a reduction of the chromatic aberration of the eye with the Drive lenses, it is reasonable to assume that the improvement of BCVAmesopic is mainly attributable to the reduction of the scattering.

In this work, the main improvement was found under glare conditions. Both STD and PRO showed better BCVAglare performances than clear lenses and the improvement was statistically significant, about 0.05 logMAR, for both PAL group and SVL group. Concerning the direct comparison between the two Drive lenses, no statistical evidence of differences in BCVAglare was found both in PAL group and SVL group. A reasonable explanation for the improvement of BCVAglare by a blue-filtering lens is again the attenuation of intraocular scattering, which is known to play a relevant role under glare.19–21,48 It is known that intraocular scattering increases gradually with age.19–21 As already discussed, the effects begin to manifest at the lowest wavelengths in the age period corresponding approximately to the SVL group. In older subjects, the effect at the lowest wavelengths increases and it also extends to longer wavelengths. The zero transmittance of the Drive lenses in the blue region up to (426±2) nm attenuates intraocular scattering. Since this spectral range corresponds to a rerion of scattering even in the age range between about 30 and 50 years, the Drive lenses are also effective in reducing intraocular scattering in the SVL group, not only in the the PAL group. Also Mahjoob et al.9 recently reported the positive effect of a yellow filter on visual acuity under glare, though the improvement was only significant in subjects aged 51–60 years. However, the only information available about the spectral characteristics of this yellow filter is the total light transmission (80%). It is not specified whether the reduction in transmittance is due to the presence of a specific absorption band centred somewhere in the blue spectral region or to the shift of the cutoff as occurs in longpass filters.

The subjective preference of the participants (Table 4) confirms, in some way, the visual performance outcomes. Most participants chose Drive lenses (STD: 18, PRO: 9), in agreement with the psychophysical tests where the Drive lenses actually produced overall better results compared to the clear lens. One could argue that this preference may be affected by a sort of recency effect i.e. the tendency for people to be most influenced by what they have last experienced. It is worth mentioning that only 6 out of 18 who chose STD lenses and 5 out of 9 subjects who chose the PRO lens had used these lenses in the third phase. Notwithstanding this, it cannot be excluded that only a minority chose the clear lens because of the order in which the three filters were worn. As shown in Fig. 1, in fact, all the participants wore the clear lens for the first fifteen days. The optometric data were acquired again at the end of the project with the clear lens and there were no differences with respect to phase 1, but the order may have influenced the subjective choice.

The effects of colored filters on visual performance is dependent on the balance between the possible reduction of chromatic aberration and intraocular light scattering vs. the undesirable though unavoidable reduction of transmitted light intensity. STD and PRO Drive lenses are longpass transmittance filters with cutoff at (426±2) nm, shifted to longer wavelength of about 20nm compared to a clear control lens. The only difference between STD and PRO is an additional band centred at about 450nm in the PRO spectrum, in the same spectral region of the absorption of the human macular pigment.

Both Drive lenses are found to maintain or improve some visual functions compared to the clear lens. The improvement of photopic contrast sensitivity at some angular frequencies, mesopic visual acuity, and visual acuity under glare is mainly attributed to the reduction of intraocular light scattering as a consequence of the total light attenuation in the blue spectral range below the cutoff. The improvement is observed in older subjects (age range: 46–73 years, PAL group). For younger subjects (age range: 26–55 years, SVL group), it is detectable for all but mesopic visual acuity.

Hoya Corporation Vision Care (Tokyo) partly funded this study. Natalia Vlasak and Silvano Larcher are fully employed in Hoya Vision Care. All the other authors report no conflicts of interest and have no proprietary interest in any of the materials mentioned in this article.