Background Exposure to ultraviolet (UV) radiation from sunlight is recognized as the principal cause of skin cancer. Moreover, sunbeds have been classified as carcinogenic by the International Agency for Research on Cancer. Despite this, there is a shortage of objective data on UV exposure levels in sunbeds in England.
Objectives We set out to measure UV emission levels in sunbeds at sites around England, and to compare these levels with both current standards and natural sunlight.
Methods Between October 2010 and February 2011, UV spectra were measured on site from a total of 402 artificial tanning units in England. Measurement instrumentation was calibrated, traceable to the National Physical Laboratory. Compliance with the relevant British and European standard was determined, and a skin-cancer weighting factor was used to compare the carcinogenic potential of sunbeds with that of sunlight.
Results For compliance with the European standard, erythemal-effective irradiance should not exceed 0·3 W m−2. The values that we measured ranged between 0·10 and 1·32 W m−2 with a mean of 0·56 ± 0·21 W m−2. Only 10% of sunbeds surveyed were within the recommended limit. Application of the skin-cancer weighting factor produced values that varied from 0·17 to 2·52 W m−2 with a mean of 0·99 ± 0·41 W m−2. The comparable value for Mediterranean noonday sun was 0·43 W m−2.
Conclusions Nine out of 10 sunbeds surveyed throughout England emitted levels of UV radiation that exceed the maximum levels contained within the European standard. Moreover, the skin cancer risk for comparable times of exposure was up to six times higher than that for Mediterranean sunlight. This situation is unacceptable and stricter control measures must be put in place.
The International Agency for Research on Cancer, an agency of the World Health Organization, has classified sunbeds as a Group 1 carcinogen, which is the highest risk category.1 This followed a meta-analysis that found a 75% increase in the risk of melanoma when indoor tanning started during adolescence or young adulthood.2 An update of this meta-analysis has now revised this figure upwards to an 87% increased risk of melanoma with first use of sunbeds before the age of 35 years, with the risk increasing with the number of sunbed sessions.3
Surprisingly, there is a shortage of data from England on the levels of ultraviolet (UV) radiation received and the detailed spectrum to which sunbed customers are exposed. Previous studies carried out in Scotland revealed high levels of UVB found in new high-power sunbeds.4–6 These studies also showed that the estimated cancer risk from sunbeds had increased by a factor of three in the last 10 years due to the use of high-power sunlamps. A British and European Standard, introduced in 2003,7 set limits on the UV emission of sunbeds. No study has been performed to investigate compliance with the standard in England. Sunbed operators in England are not obliged to provide advice on health risks to customers; however, they are in Scotland under the Public Health Scotland Act 2008. They can also operate unmanned premises, equipped with coin-operated sunbeds available to anyone of any age, despite a report that half of all 15–17 year olds in Liverpool and Sunderland have used a sunbed.8 In various regions of England, tanning outlets are required to operate under licence from their local authority, but UV emission levels are not measured due to the cost and complexity of obtaining reliable data.
In 2006 the European Union (EU) Scientific Committee on Consumer Products published an opinion stating that UV tanning equipment should not exceed an erythema-weighted irradiance level of 0·3 W m−2, and that equipment above this level would be considered unsafe for tanning devices used for cosmetic purposes.9 All EU member states, including the U.K., agreed to introduce this level from 1 April 2009 for all new and traded tanning devices. The classification scheme for sunbeds is shown in Table 1. Unless intended for medical use, the erythemal-effective UVA and UVB output should not exceed 0·15 W m−2, and the erythema-weighted total UV irradiance should not exceed 0·3 W m−2. UV type 3 is the only class of sunbed intended for unskilled use, and therefore suitable for general use in commercial tanning outlets.
Table 1. Limits on erythemal irradiance according to BS EN 60335-2-27:2003
Ultraviolet (UV) type
Erythemal irradiance, W m−2
UVB, λ = 280–320 nma
UVA, λ = 320–400 nma
aPlease note that the definitions of UVA and UVB given here differ from the more common International Commission on Illumination definitions.
The aims of the present study were to measure the actual intensity of UV at each wavelength produced by sunbeds in England with a geographical spread from north to south, to apply a skin-cancer risk model to the results, and to assess compliance with the British and European standard on sunbed emissions.
Materials and methods
During the period from October 2010 to February 2011 the UV spectra were measured on site from a total of 402 artificial tanning units. Local databases were searched to ensure adequate numbers of sunbeds, and within each area all sunbeds were surveyed to avoid selection bias. Initial contact was made with the local Environmental Health Departments. Letters were drafted and sent to the sunbed operators informing them of the survey and agreeing that no individual sunbed location would be identified. Access to the sunbed establishments was authorized with cooperation from local Environmental Health Officers.
Three types of artificial tanning units were included in this research; they are referred to as ‘vertical’, ‘horizontal’ and ‘high-pressure’ sunbeds, as seen in Figure 1. Many establishments use the horizontal sunbeds, which consist of an upper canopy and lower base bench arrayed with a total of 40–50 lamps of power range 80–250 W. The vertical units are usually known as ‘stand-up units’ or ‘sunshowers’. The customer stands inside the cabin and is irradiated from 48–60 lamps, either in a circumference or sometimes split up into equal banks. In some machines, high-pressure metal halide lamps are utilized. They provided more UVA, less UVB and overall less total erythemal irradiance.
Measurements were performed on site using a Maya 2000 Pro Ocean Optics Spectrometer (Ocean Optics, Dunedin, FL, U.S.A.) with grating tuned to 230–440 nm, and an optical fibre and cosine corrector with Spectralon diffusing material (CC-3-UV-S). Angular response was measured over a field of view of 120°. Technical difficulties prevented measurement over the full 180°. The measured angular response was used to calculate the cosine error,10f2 = 6·69%, which is below the 10% maximum recommended for commercial UV radiometers.11 All on-site sunbed lamps were allowed 2 min to warm up and stabilize outputs. The time for each measurement varied between 10 and 50 s. In order to increase the dynamic range, measurements were taken over short and long time periods, using the technique described by Gaigalas et al.12 The calibration of the device, traceable to the National Physical Laboratory, was carried out in the Optical Calibration Laboratory of the Photobiology Unit, University of Dundee. The Laboratory is accredited by the United Kingdom Accreditation Service. Measured values were corrected for stray light using a xenon arc lamp and cut-off filter.
The measurements of vertical sunbeds were taken without the cabin being occupied. Folding plastic crates were stacked in vertical tanning units to mimic an occupied cabin. This had the effect of placing a barrier behind the collecting optics to absorb UV radiation in a similar manner to a client standing within the cabin. The diffuser and optical fibre were placed in a custom holder and mounted on top of a camera tripod at a height of 160 cm. The holder was clamped in place and the face was directed parallel to a bank of lamps 22 cm away using a spacer, as this approximated the position of someone standing in the cabin. The door was closed and the optical fibre transmitted the UV radiation to the spectrometer, which was placed outside the cabin to avoid heating.
For horizontal sunbeds, measurements were taken of the canopy and lower lamps separately. To measure the UV from the lower lamps, a specially designed custom holder was used to hold the front end of the collecting optics in close proximity to the lamps. The design of the holder prevented light entering from the upper canopy, which would normally be blocked by the client’s body. The mean of three readings was calculated from a central 15-cm zone. The holder was then flipped over to measure the UV from the canopy (mean of three readings). Used this way, the collecting optics were raised 20 cm above the Perspex acrylic surface at a similar position to that occupied by someone lying on the sunbed (Fig. 2). A black sheet was used to cover the lower lamps in order to block both light from the lower lamps and also reflections that would not be present in an occupied sunbed. The outputs from the canopy and lower lamps were usually similar. However, the higher of the two readings was used, as clients do not generally turn over during treatment and so one side will receive the higher dose.
Erythemal irradiance and compliance with BS EN 60335
The effect of applying an erythemal weighting factor13 is demonstrated in Figure 3. Although the emission from the sunbed is mainly in the UVA region, when it is weighted by the erythemal action spectrum, the significance of the short wavelength UVB becomes apparent.
Skin cancer and comparison with natural sunlight
Another method of analysing the sunbed spectra is the application of a skin-cancer weighting factor, called Skin Cancer Utrecht Philadelphia-human (SCUP-h).14 It is useful to compare the UV emission from a sunbed to that occurring naturally from sunlight. Data of the spectral irradiance of the sun was obtained from the Laboratory of Atmospheric Physics at the University of Aristotle in Greece (40° 39′N, 22° 58′E) as recorded on 18 July 2009 at 10:36 UT (i.e. nearest to local noon when the irradiance is at a maximum).
Sunbed measurements were conducted in the North, Midlands, South West and London.
The majority (60%) of sunbeds were located in tanning centres or beauty salons (Fig. 4), although significant numbers were also located in hairdressers (17%) and fitness centres (15%). There were between one and eight sunbeds at each site, with 63% being the vertical type of unit.
Erythemal irradiance and compliance with BS EN 60335
Total erythemal irradiance ranged between 0·10 and 1·32 W m−2, with a mean of 0·56 ± 0·21 W m−2 (n = 402) (Fig. 5). For compliance with BS EN 60335 (the British and European standard), a UV type-3 appliance should not exceed 0·3 W m−2. The mean erythemal irradiance of the 402 sunbeds tested was almost double the compliance level. Only 10% were within this limit.
The mean UVB erythemal irradiance was 0·35 ± 0·17 W m−2 (n = 402) (Fig. 6), more than double the compliance limit of 0·15 W m−2. Only 15% of the results were below this compliance value. According to BS EN 60335, any device with UVB erythemal irradiance above 0·15 W m−2 should be used only following medical advice.
The UVA erythemal irradiance is shown in Figure 7. The mean value was 0·21 ± 0.06 W m−2 (n = 402), which is above the compliance limit of 0·15 W m−2; only 26% of the tanning units were below the compliance value. The maximum UVA erythemal irradiance was 0·44 W m−2.
Skin cancer and comparison with natural sunlight
The total SCUP-h effective irradiance for the tanning units varied between 0·17 and 2·52 W m−2, with a mean of 0·99 ± 0·41 W m−2 (n = 402). For comparison, the effective SCUP-h for Mediterranean noon sun was 0·43 W m−2. Hence, the average artificial tanning unit has a carcinogenic risk per minute of exposure that is 2·3 times that of midday Mediterranean sun.
The maximum SCUP-h value was 2·52 W m−2, measured from a vertical unit containing 250 W lamps; this value was six times that of Mediterranean sun. The lower limit of the upper decile level was 1·53 W m−2; that is, 10% of sunbeds emitted carcinogenic-weighted irradiance levels at least 3·6 times greater than Mediterranean sun.
Results from sunbeds surveyed around England are shown in Table 2. There were 76 sunbeds surveyed in the metropolitan borough of North Tyneside and Newcastle-upon-Tyne in North-East England. This area includes the unlicensed areas of Wallsend, North Shields, Whitley Bay, Forest Hall, Camperdown, Dudely, West Allotment and Gosforth. In total 56 tanning units were measured across Nottinghamshire and Derbyshire. These included the licensed areas of Chesterfield, Mansfield, Ashfield and Ripley, and Amber Valley. Six London boroughs were visited during the survey. These were the licensed boroughs of Barnet (n = 48), Bromley (n = 32), Islington (n = 47) and Sutton (n = 33), and the unlicensed Newham (n = 24) and Bexley (n = 41). In the South West of England measurements were conducted in Cheltenham, Coleford and Newton Abbot in the Teignbridge district of Devon. These were all licensed regions.
Table 2. Ultraviolet (UV) levels measured from sunbeds in the different locations surveyed during the investigation
Total erythemal irradiance, W m−2 (SD)
UVB erythemal irradiance, W m−2 (SD)
UVA erythemal irradiance, W m−2 (SD)
SCUP-h irradiance, W m−2 (SD)
SCUP-h, Skin Cancer Utrecht Philadelphia-human (skin-cancer weighting factor).
Cheltenham, Coleford and Newton Abbot
Nottingham, Chesterfield & Ripley
At the time of the study, North Tyneside and the London boroughs of Bexley and Newham were unlicensed. A two-group t-test was implemented between the licensed and unlicensed London boroughs. The difference between the mean values of sunbed irradiance for the two groups was not statistically significant (P = 0·237). When the two-group t-test was performed between the unlicensed North East and the licensed South West, there was a statistically significant difference in the mean values of the two regions (P < 0·001).
In this study, considerable care was taken to ensure that UV measurements were accurate. The spectrometer used in this study has a charge-coupled device (CCD) detector. One of the major technical limitations of CCD arrays is the high level of stray light. For this reason, stray light corrections were carried out. Another error can arise when measurements are being taken in an unoccupied sunbed. A person standing in or lying on a sunbed absorbs incident UV radiation; when the sunbed is unoccupied, radiation from one bank of lamps strikes the opposite lamps and reflectors, and reflects on to the detector. The methodology adopted ensured that this error did not occur. These two measurement artefacts would both lead to an inappropriately high radiation output value, and so considerable care was exercised to ensure that the data collected were reliable.
To date there have been only a small number of sunbed studies carried out in which full spectral measurements were performed on site. Spectral data are required to give accurate information on the UVA and UVB content of sunbeds and to allow the use of weighting factors in the analysis. The importance of measuring on site is that the spectrum and intensity of UV radiation depend on the type, condition, age, stacking density and temperature of the lamps, the design and condition of the reflectors, and the material and condition of the plastic lamp protector.
A hazard assessment of 38 artificial tanning units carried out in Scotland in 1998 found that SCUP-h irradiances were comparable with U.K. summer sunlight.4 These results were confirmed in a similar study published shortly afterwards.15 A follow-up study of 133 sunbeds in Scotland in 2007 found that the median of the SCUP-h irradiance of all lamps was then equivalent to Mediterranean sunlight.6 This was due to the widespread use of ‘fast tan’ lamps. Although the new British and European standard was in place, 83% of sunbeds exceeded the UVB limit set out in the new standard.
Since 1983, all tanning models in Norway have needed approval before being sold, and type-3 limits were applied from late 1992 onwards.16,17 However, on-site inspections in 1998–99 revealed that only 28% of tanning devices were equipped with the correct lamps (i.e. the same type of sunlamps as approved), though this had increased to 59% in 2003. This highlights another problem, which is ensuring that replacements lamps are the same as or similar to the original. In recent studies in Norway, mean erythema-weighted UVB and UVA irradiances were 0·194 ± 0·010 W m−2 and 0·156 ± 0·008 W m−2, respectively.16,17 They found that only 23% of solaria were within the limits of a type-3 sunbed.
In the present study, on-site spectral measurements were performed on 402 sunbeds in England. The mean erythema-weighted UVB and UVA irradiances were 0·35 ± 0·17 W m−2 and 0·21 ± 0·06 W m−2, respectively, and the mean erythemal total UV irradiance was 0·56 ± 0·21 W m−2. Only 10% of sunbeds tested complied with the type-3 limit.
There is a trend that may be identified from reviewing the reports on recent measurements of sunbeds. UV emissions are increasing with the development of new high-power sunlamps. Moreover, the publication of a European Standard in 2003 does not seem to have arrested this development. Sunbeds with erythema-weighted UVB irradiance exceeding 0·15 W m−2 should be used only following medical advice, according to BS EN 60335. Yet 85% of sunbeds tested exceeded this level.
The SCUP-h action spectrum indicates the relative effectiveness for induction of nonmelanoma skin cancer.14 Its merit is in facilitating a quantitative comparison between artificial tanning units and sunlight. In the present study, the mean SCUP-h irradiance was 2·3 times that of Mediterranean sunlight. The maximum value recorded was six times higher than Mediterranean sunlight.
Sunbeds tested in North-East England had significantly higher UV emissions than those in the South West (0·70 ± 0·21 W m−2 vs. 0·45 ± 0·25 W m−2). Sunbeds are licensed in the South West but not in the North East. It may be that licensing was a contributory factor in the difference observed, although there may be other influences such as competition in the market place. In this respect, there was no significant difference between the London boroughs that had licensing and those that did not, and so in this more closely comparable area, licensing per se did not appear to influence compliance with the European standard. This is not surprising, as the local authorities do not have the capability for making the necessary UV spectral measurements. A possible solution would be if only type-3 sunbeds were permitted to be sold and used within the U.K., and units were inspected to show that only the correct lamps were fitted.
Five of the tanning centres visited were unmanned and operated on a coin or credit card system. Anyone of any age or skin type can turn up and use them, and can have as many treatments as they wish. Until recently, these were in widespread use, but they are now banned in Scotland.
Individual studies examining the association between sunbeds and skin cancer may have contradictory findings. Although Elliot et al.18 did not show a link between melanoma and sunbed use, Lazovich et al.19 found a dose–response relationship for years during which sunbeds were used. A meta-analysis combines results from all investigations, taking due account of the strength of evidence in each study. In a recent meta-analysis, results from 25 studies were combined.3 This showed that use of sunbeds increased the risk of melanoma by 20%. The risk of melanoma was almost doubled when use started before the age of 35 years. It was estimated that 3438 cases (5·4%) of melanomas in Western Europe were related to sunbed use. Calculations showed that 99 deaths each year in the U.K. were attributable to sunbed use.
Most people use sunbeds for purely aesthetic reasons. If the sunbed were a cosmetic product it would have been withdrawn years ago. However, legislation has been very slow in coming within the U.K. The bill regulating sunbeds in Scotland was passed in 2008 and came into force in 2009. This banned unmanned salons and prevented the use of sunbeds by those under the age of 18 years. It also required prescribed information on the health risks to be supplied to users. The other governments in the U.K. followed Scotland’s example, and now England, Wales and Northern Ireland all have legislation in place prohibiting the use of sunbeds by people under the age of 18 years. Supporting measures still need to be introduced in the legislation in England, which is the only nation in the U.K. to allow the use of unmanned tanning salons.
The present study, covering 402 sunbeds spread throughout England, indicates clearly that the current situation is very unsatisfactory, and much more needs to be done in England to discourage the use of sunbeds. This investigation clearly shows that 90% of sunbeds emit levels of UV radiation that exceed the limits allowed by the British and European standard. The standard is intended to safeguard the public but it is being largely ignored by the sunbed industry. Stricter control measures must be put in place along with continued programmes of education. Otherwise, the melanoma burden will continue to increase.
We acknowledge the invaluable assistance provided by local Environmental Health Officers from Cheltenham, Derbyshire, London Barnet, London Bexley, London Bromley, London Islington, London Newham, London Sutton, Newton Abbot, Nottinghamshire and North Tyneside. Their support and collaboration were key to the success of this project. We would also like to thank all owners, managers and operators of the tanning establishments for their cooperation. Thanks also to Professor Alkiviadis Bais for the provision of natural sunlight spectral data from Thessaloniki, Greece.