5. (a) Are limit values necessary for the irradiance of UVR from artificial sources, in particular from tanning devices for cosmetic purposes, with respect to health and safety? (b) Is it necessary to give different values for the irradiance of UVA, UVB and UVC radiation respectively? (c) If so, please specify the limit values for the irradiance of artificial UVR above which adverse health effects will occur. What are the uncertainties of these limit values?
From the above discussion on reciprocity, it is clear that acute clinical effects resulting from sunbed use (i.e. erythema) are likely to depend only on total dose and not dose rate. It is not possible to make any statements on the risk of skin cancer, especially melanoma. Since all tanning devices emit a broad UVR spectrum, the spectral profile of which determines the device’s effectiveness to elicit clinical effects, it is irrelevant to specify irradiance limits in different spectral wavebands, especially since the spectral regions UVA, UVB and UVC were originally based on the optical properties of different glasses (see section 2b). A more appropriate way to speak of tanning devices than using the terms UVA, UVB and UVC is to compare their erythemal power, as a percentage of total UVR power, with sunlight. This is expressed mathematically as:
E(λ) is the relative spectral power distribution of the UV source and ε(λ) is the erythemal effectiveness of radiation of wavelength λ nm (CIE 1998). An example of this calculation is given in Appendix A. For the 3 sources illustrated in the Figure 1, the erythemally effective percentages are 0.44%, 0.51% and 0.13% and for summer sunlight, the “Cleo Natural” and “Cleo Performance” lamps, respectively. Clearly, the “Cleo Natural” lamp more closely resembles sunlight “biologically” than the “Cleo Performance” lamp.
In specifying an upper limit of irradiance, the important quantity is the erythemally-weighted irradiance, obtained by weighting each spectral irradiance component of the lamp by its relative effectiveness to induce erythema and summing over all wavelengths present in the source spectrum (see equation above). To minimise the risk of timing errors, which might result in “sunburn” it is desirable that the prescribed sunbed exposure session should be no less than 10 minutes. The avoidance of “sunburn” may also reduce the risk of melanoma.
Depending on an individual’s phototype, the exposure in SED during this 10-minute period should not exceed the subject’s estimated indicative MED (see Table 1). The maximum erythemal-weighted irradiance should not exceed 11 SED/h (0.3 W/m2). This is equivalent to a UV index (UVI) of 12, which WHO describes as “extreme”. A noon UVI of 12 would be typical in summer in Darwin, Australia (13oS) and Colombo, Sri Lanka (13oN).
The main conclusion from this analysis is that the erythemally weighted properties of a given sunbed emission spectrum (as demonstrated in Appendix A) are more important than its physical properties per se. This is because UVB is orders of magnitude more erythemogenic than UVA (as shown in Figure 2) which, as shown in Figure 1, is the main UVR component of sunlight and tanning device spectra. At present, certainty can only be reasonably assured for acute effects such as erythema.
6 Please specify the limit values of total dose of artificial UVA, UVB and UVC radiation above which adverse health effects will occur, taking into account skin phototype, intensity of exposure, duration of exposure and associated uncertainties.
The clinical effects of UVR exposure can be either deterministic, where the magnitude of the effect is related to exposure and a threshold dose is possible, or stochastic, in which the probability of the effect is related to exposure and there is no threshold dose. Erythema is an example of a deterministic effect with a threshold and SCC is a stochastic effect without a threshold.
There is no need to specify total dose separately for UVA, UVB and UVC for the reasons given in Section 5. It is not possible to specify uncertainties for long-term effects.
For a single exposure on a sunbed it is important to avoid marked or severe erythema but necessary (for the desired cosmetic effect) to receive a sufficient UVR dose to stimulate melanogenesis. Experience has shown that an exposure at or just below that to induce a just perceptible MED 8 to 24 hours after exposure approximates the optimum.
A classification of skin phototypes based on susceptibility to sunburn in sunlight (WHO 2003), together with indicative MEDs that might be expected following exposure on unacclimatized skin, is given in Table 1. Depending on an individual’s phototype, the exposure in SED on any single occasion should not exceed their estimated indicative MED.
With a stochastic effect like SCC skin cancer there is no threshold dose below which the effect will not occur. Consequently, any recommendation about a limit value of total dose accumulated over a specific time period (e.g. year, lifetime) is arbitrary and subjective. A limit of 20 sessions per year (equivalent to an exposure of approximately 40 SED or 20 MED in melano- compromised individuals) was proposed by the British Photodermatology Group (BPG) in 1990 (Diffey et al 1990) and was subsequently adopted by the UK Health & Safety Executive (HSE 1995).
The International Electrotechnical Commission (1995) recommends that the maximum annual exposure should not exceed an erythemal-weighted dose of 15 kJ/m2 (150 SED), equivalent to around 50 MED in white-skinned people. Not surprisingly, it is this higher limit adopted by The Sunbed Association in the UK in their code of practice for operators (TSA 2004). A pragmatic defence of the limit of 20 sessions per year is that it discourages people from purchasing sunbeds to use at home where the temptation to use them several times a week or for prolonged time during a single tanning session is evident.
Other agencies have simply advised against sunbed use and not specified an “acceptable” maximum annual usage (AGNIR 2002; WHO 2003; ICNIRP 2003).