People maintain their sun exposure behaviour in a 5–7-year follow-up study using personal electronic UVR dosimeters

Abstract

The main purpose of this study was to investigate whether people change their sun behaviour over a period of 7 years. Thus 32 volunteers, who had all participated in earlier sun exposure studies in 1999–2001, were enrolled in a follow-up study in 2006. They were selected to represent a previous low, medium and high UVR exposure. They participated for mean 121 days (range 65–157 days) wearing a personal, electronic wrist-borne UVR dosimeter and completed sun exposure diaries. No statistically significant differences were seen from year to year in the estimated annual UVR dose, mean UVR dose per day or mean percentage of ambient UVR. However, there was a person effect showing that participants maintained a high or low UVR dose lifestyle over the years. In 2006 the 32 participants received an estimated annual mean UVR dose of 221 SED and a median dose of 154 SED (range 25–1337 SED), while they received a mean for the previous participation years (1999–2001) of 236 SED (median 153 SED, range 24–980 SED). The estimated annual UVR dose for each of the previous sun years and the estimated annual dose for 2006 correlated significantly (R² = 0.51; p < 0.0001). Sun risk behaviour, expressed as days sunbathing, did not change either. The participants thus seem to have maintained their sun exposure behaviour over a 5–7-year period.

---

Introduction

Solar ultraviolet radiation (UVR) is known to be an important etiological factor in the development of cutaneous melanoma and non-melanoma skin cancer and sun related eye diseases.¹ Through prospective, continuous, objective studies in subgroups of the Danish population we have previously achieved a basic knowledge of factors important for UVR exposure patterns.^2,3 It was also demonstrated that there was stability in UVR data for persons taking part in two successive years.⁴ In addition, we found, in accordance with other studies, that some individuals received considerably higher or lower UVR doses than their peers.^5,6 Furthermore, a high UVR exposure dose was strongly correlated with risk behaviour defined as exposing one's upper body in the sun. We wanted to test the hypothesis that people maintain their sun behaviour throughout their lives. If this is true, it will be sufficient to investigate an individual's sun exposure behaviour over one season in order to be able to predict the individual's lifetime UVR exposure dose. We therefore conducted a follow-up study using the same protocol as in our previous studies performed from 5 to 7 years earlier.² The volunteers were selected to comprise persons who in the previous studies had either low, high or medium sun exposure.

Material and methods

Participants and procedure

The study took place in an urban population living in or near Copenhagen, Denmark, 56°N in May to October 2006. The solar zenith angle at noon was 46° on April 15th, 32° on June 21st and 64° on October 15th. We recruited 38 healthy, adult volunteers of Danish ancestry from among the 285 volunteers who had previously participated in UVR dosimeter studies in 1999–2001. The group was selected to comprise people (indoor and outdoor workers) with a previous low, medium or high UVR exposure, where “low” was defined as a UVR dose below the 25% quartile (<100 SED), “medium” as a UVR dose at the median level (100–300 SED) and “high” as a UVR dose above the 75% quartile (>300 SED). The volunteers (17 males and 21 females) had a mean age of 51 (range 31–71). As in the previous studies, the subjects wore a wrist-borne personal electronic UVR dosimeter and completed sun exposure diaries over the summer half of a year.^4,7 To be included in the data analysis the volunteers were required to have more than 30 days with both UVR dosimeter readings and corresponding diary data, of which 21 days or more had to be in June, July or August. 32 volunteers met the criteria for analysis. 13 of these volunteers had participated in all 3 previous years, 9 in 2 previous years and 10 in 1 previous year. This provided 99 sun-years for analysis (a sun-year is 1 subject participating in 1 6-month summer period). The 6 who did not meet the criteria for analysis were 3 males and 3 females. 2 were excluded due to insufficient compliance and the 4 others due to UVR dosimeter malfunction. The 2 excluded had a low UVR exposure dose in previous studies while the 4 dropouts due to dosimeter malfunction previously had a high UVR exposure dose. The study was approved by the Scientific Ethical Committees for Copenhagen and Frederiksberg (KF11-320779).

Personal electronic UV dosimeter, SunSaver

The dosimeters we used were developed and assembled in our department and includes a sensor and a data logger. It is mounted in a housing together with a digital watch. A Silicon Carbide Photodiode (JECF1-IDE; Laser Components; Olching, Germany) was chosen as a sensor only sensitive in the range 200–400 nm. The sensor has a built-in diffuser and cosine response with spectral response similar to the CIE erythema action spectrum.⁸ The data logger controls the sensor, which was set to measure every 8th second and to store the average of the last 75 measurements every 10 minutes together with the time. The measurement range of the dosimeter is 0.1 SED per hour to 23 SED per hour. The SunSaver is battery driven, can run for 145 days without maintenance, and the data can be transferred to a personal computer.

Ambient UVR radiation exposure

Ambient solar UVR was measured with a UV-Biometer, model 501 (Solar Light Co. Inc., Glenside, PA) on the roof of our hospital. The UV-Biometer was calibrated with a reference detector traceable to the National Institute of Standards and Technology. The measurements are expressed in standard erythema doses (SED), where 1 SED = 100 J m⁻² normalized to 298 nm in accordance with the International Commission on Illumination (CIE) erythema action spectrum.^8–10

Statistical analysis

The volunteers participated for a mean of 121 days (range 65–157 days) in 2006. SunSaver measurements and sun exposure diary data were analysed for 12 053 participation days with both dosimeter and diary data. Of the 32 volunteers in 2006; 20 had participated in 2001; 26 in 2000 and 21 in 1999. To compare the individual UVR doses, we adjusted the observation period to a year, knowing that the received UVR during winter is almost negligible, except for winter holidays in sunny places and sun-bed use. The estimated annual UVR doses were calculated for each participant in each of the participation years on the basis of the individual measured daily doses in the actual year and for missing days as the same part of ambient UVR found on comparable days with measurements by separating days on/off work and being in/outside Denmark. In all calculations each sun-year weighed equally irrespective of the number of days per year a subject participated. We used mainly descriptive statistics presenting the mean and median of the continuous data. 2-way analysis of variance (ANOVA) was used to test differences in UVR exposure dose and behaviour from year to year also correlation for individual person effect. Spearman's rank correlation was used to investigate interactions between two continuous measurements. In each case a p-value of less than 0.05 was considered significant. We used IBM SPSS statistics version 19 for data analysis.

Results

Table 1 shows the baseline characteristics, such as group category, sex, age in 2006 and number of participation years. In addition, for each individual in increasing order, the mean estimated annual UVR dose of all participation years and the highest difference from this mean UVR dose in actual SED as well as in percent found in a participation year. There was a more than 30-fold difference in mean estimated annual UVR dose between the participant with the lowest to the participant with the highest mean annual dose.

Table 1 Baseline characteristics, such as group category^a, sex, age in 2006 and number of participation years. In addition, for each individual in increasing order, the mean estimated annual UVR dose for all participation years and the highest difference from this mean UVR dose in actual SED as well as in percent

No.	Group category^a	M = male F = female	Age in 2006	Years of participation	Mean estimated annual UVR in SED for all years	Highest actual difference from mean annual UVR in SED	Highest difference from mean annual UVR in %
a Group: I = Indoor workers, O = Outdoor workers (municipal gardeners), S = Sun worshippers and G = Golfers.
1	I	F	44	2	31	6	20
2	I	F	71	2	34	3	8
3	I	F	45	2	56	6	10
4	I	F	58	4	56	46	82
5	I	M	63	4	61	8	14
6	I	F	60	4	70	40	57
7	I	F	63	2	74	5	7
8	I	M	37	3	97	39	40
9	I	F	34	4	108	53	50
10	S	F	50	2	109	41	38
11	I	M	40	4	117	23	20
12	I	M	52	4	127	59	46
13	I	F	32	3	87	43	50
14	I	M	48	4	137	45	33
15	I	M	60	4	138	41	29
16	I	M	36	4	156	61	39
17	I	F	40	4	164	77	47
18	I	F	35	4	185	55	30
19	I	F	63	4	204	98	48
20	I	F	31	2	205	162	79
21	I	F	43	4	232	80	35
22	O	M	65	3	284	88	31
23	G	M	61	3	293	68	23
24	S	F	32	2	313	61	19
25	O	M	64	3	330	42	13
26	O	M	46	3	368	56	15
27	G	M	70	3	383	112	29
28	I	F	61	2	400	76	19
29	S	F	43	2	403	150	37
30	O	M	62	3	447	158	35
31	O	M	58	3	457	39	8
32	S	F	60	2	1158	178	15
Median			51	3	160	55	31
Mean			51	3	228	63	32
Min.			31	2	31	3	7
Max.			71	4	1158	178	82

---

When the last study year 2006 was compared with the former participation years, significant differences in the environmental UVR dose (ambient UVR) were found among the study years, a “year effect” (p < 0.001). The mean ambient UVR dose per day was thus significantly higher in 2006 than in all the previous study years (p < 0.001) (2006: 18.8 SED day⁻¹; 2001: 17.2 SED day⁻¹; 2000: 15.5 SED day⁻¹; 1999: 14.9 SED day⁻¹).

Table 2 shows the differences in UVR exposure due to the study years called the “year effect” and the differences due to the behaviour of the individuals called the “person effect”. To identify possible “year effects”, the UVR data from year 2006 is compared with all previous years together (1999–2001). A significant “year effect” (p < 0.0001) was found in the participants’ daily mean UVR exposure hours (defined as hours with positive UVR dosimeter measurements). The individual daily exposure hours were mean (median) in 2006: 2.8 h (2.5 h); in 2001: 2.8 h (2.2 h); in 2000: 2.3 h (2.0 h) and in 1999: 1.8 h (1.8 h). Thus there were significantly more exposure hours in 2006 than in 1999 (p = 0.012) and 2000 (p < 0.001) but a similar amount to 2001 (p = 0.3). However, the higher numbers of exposure hours in 2006 did not result in a significant “year effect” as regard to the individual measured daily UVR exposure dose (p = 0.4), which were in mean (median): 2006: 1.3 SED (1.0 SED); in 2001: 1.2 SED (0.9 SED); in 2000: 1.4 SED (0.8 SED) and in 1999: 0.8 SED (0.8 SED). Neither was a significant “year effect” found (p = 0.7) when the personal measured daily UVR doses as percent of the ambient UVR dose were calculated as mean (median) in 2006: 6.0% (5.0%); in 2001: 6.4% (4.6%); in 2000: 8.5 (6.0%) and in 1999: 4.6% (3.6%). However, the participants received a significantly higher percent of ambient UVR dose in 2006 than in 2001 (p = 0.043) and 2000 (p < 0.015) but a similar amount to 1999 (p = 0.2). Even when comparing the estimated annual UVR exposure from the different years no “year effect” was found. In 2006 the 32 persons recruited for the follow-up study had an estimated annual UVR exposure dose of mean 221 SED, median 154 SED, (range 24–1336 SED), while the corresponding values for 1999–2001 were mean 236 SED and median 153 SED (range 36–980 SED). The mean and median values are thus almost identical from the previous years to 2006, and the IQR (interquartile range) shows a difference of ±35 SED or approx. ±30%. Only 6 persons had a difference in estimated annual UVR dose above ±50%; of these 2 were lower and 4 higher in 2006 than in the previous years.

Table 2 UVR exposure in 2006 vs. 1999–2001. Analysis of variance (ANOVA) to test differences in UVR exposure with both “Person” and “Year” as independent variables. The “Year Effect” was tested in 2006 versus all the years 1999–2001. The last three columns test the difference in the measured UV doses in SED and UV behaviour in days between 2006 and each individual year. A positive value means that the exposure was highest in 2006. Statistically significant differences are marked as follows: *p < 0.001; **p < 0.01; ***p < 0.05. Not marked differences were not statistically significant

	2006 – (1999–2001)		2006 – 1999, 2000 & 2001 individually
	Person effect	Year effect (p)	2006–1999	2006–2000	2006–2001
a Daily personal measured UVR dose (SED). b Daily personal measured UVR dose on days with risk behaviour (SED). c Daily personal measured UVR dose as percent of the corresponding ambient UVR dose. d Days with risk behaviour comprises “days with sunbathing to get a tan” and/or “days with exposing shoulders”.
Estimated annual UVR (SED)	Yes*	No (0.8)	−21	−0.6	−11
UVR per day^a (SED)	Yes*	No (0.4)	−0.03	+0.05	−0.16
UVR per risk day^b (SED)	Yes*	Yes**	−1	−1.9*	−1.2
Personal UVR in % of ambient^c UVR	Yes*	No (0.07)	−1%	−1.9%***	−1.7%***
Exposure hours per day (h)	Yes*	Yes*	+0.5 h***	+0.8 h*	+0.2 h
Risk behaviour^d (days)	Yes*	Yes*	+6	+12*	+7***
Exposing shoulders (days)	Yes*	Yes*	+6***	+13*	+8***
Sunbathing to get a tan (days)	Yes*	No (0.7)	−0.4	+1.8	+1
Sunburn episodes (days)	Yes*	No (0.5)	+0.7	+0.7	+0.4

---

As also shown in Table 2, a significant “person effect” expressed as statistically significant differences in UVR exposure among the participants was found for all the provided UVR exposure measures (p < 0.001), which means that some persons continuously received higher or lower UVR doses than their peer participants throughout the years in spite of the different weather conditions.

In respect of sun behaviour, Table 2 shows no statistically significant “year effect” regarding number of days in which people engaged in risk behaviour expressed as “days sunbathing to get a tan”. However, a “year effect” was seen in number of days with risk behaviour expressed as “exposing shoulders”, which was significantly higher in 2006 than in all three previous years, probably because the summer was sunnier in 2006 than in the other years.

A significant correlation was found between the estimated UVR dose for 2006 and “the mean estimated annual UVR dose for 1999–2001”, Spearman's r = 0.83, p < 0.001. Fig. 1 shows that there is also a clear correlation (R² = 0.51, p < 0.001) between the estimated annual UVR dose a participant received in 2006 and the estimated annual UVR dose for each of the previous years 1999–2001 in which that person had participated.

Fig. 1 The points mark the estimated annual UVR dose in SED for all previous years from 1999–2001 versus the estimated annual UVR dose in SED for 2006 for each participant. Same coloured points on a vertical line represent the same person in different years. The red line is y = x. This means that a point above the line shows that the estimated annual UVR dose was previously higher than in 2006, while a point below the red line shows an estimated annual UVR dose that was previously lower than in 2006. NB: One person included in the analysis is not shown in Fig. 1. Her mean estimated annual UVR dose was 980 SED in 2000 and 1336 SED in 2006 – far beyond the rest of the group.

Discussion

Among all 285 former participants in our earlier studies comprising 346 sun years in 1999–2001 an estimated annual UVR exposure dose of mean 199 SED, median 173 SED, (range 17–980 SED) was found,² which was comparable with the follow-up study in 2006. However, it should be kept in mind that the 32 participants were selected to include both low and high exposure extremes, to make it easy to establish whether the participants had maintained their behaviour throughout the years.

When looking solely at the 32 participants’ individual mean estimated annual UVR doses for all participation years as shown in Table 1, the 30-fold difference from the lowest to the highest mean annual UVR dose is striking. However, even though the highest difference in annual UVR of 82% (46 SED) is found in one of the female indoor workers with a low mean annual UVR dose of 56 SED, her UVR dose in the year of highest UVR dose is still only 102 SED, far below the mean value for the whole group. While, in contrast, the female sun worshipper with the highest mean annual UVR dose of 1158 SED had a difference of only 15%, which represented 178 SED. This difference alone is approximately the mean annual UVR dose for the whole group, which emphasizes the importance of trying to identify persons with a high UVR load.

The main purpose of this study was, however, to present differences between the last participation year, 2006, and the former participation years 1999–2001, in order to assess if this gap in years gave rise to altered sun exposure behaviour.

The 32 participants had previously participated in 67 sun years during 1999–2001; 13 of them had taken part in all 4 years. The follow-up year 2006 had the warmest summer with the greatest number of sun hours in a century, which also resulted in a significantly higher ambient UVR per day in the study period. In addition, the participants had more daily exposure hours in 2006 than in the previous years. In spite of this the participants did not receive a significantly higher UVR dose in 2006. In that year the participants had more days with sun risk behaviour expressed as days exposing shoulders or upper body outdoors, which was probably due to the hot weather, but not more days sunbathing to get a tan. The fact that the mean UVR dose of 3.1 SED per risk behaviour day in 2006 was lower than in the previous years although significantly lower only than in 2000 (p = 0.001) strengthens this observation. Furthermore, the participants did not experience significantly more sunburn episodes in 2006 than in the previous years. It appears therefore that people adhere to their sun habits, at least over a period of 7 years.

Only 4 persons differed by more than 200 SED from their previously estimated mean annual UVR exposure dose. This could be explained by altered situations rather than by altered attitudes toward UVR exposure. One person took part in 1999 and 2006 only. In 1999 as a student she went on a 2-week trip to Southern Europe, had 22 days with risk behaviour and had an estimated annual UVR dose of 367 SED, 121 of which were received during risk behaviour. In 2006 she had graduated and had her second child, and the family moved to a new house, where the summer holiday was spent on indoor refurbishment. Her estimated annual UVR dose fell to 43 SED and she had no risk behaviour at all, which must be considered extremely low for her. Another person participated in 2000 and 2006 and received 979 SED and 1336 SED in the respective years. She was an extreme sun worshipper and tried to get all the sun possible in both years. In 2006 she was 60 years old, had retired and lived in a small house with a garden, where she stayed outdoors most of the time and took advantage of the greater number of sun hours in 2006, which explained the increase in UVR dose. A third person was a gardener who participated in 2000 (521 SED), 2001(530 SED) and 2006 (290 SED). He was a keen mountain climber and had several tours in the mountains in 2000 and 2001. The fall in his UVR dose in 2006, when he was 62 years old, was due to not climbing mountains and a change in his work schedule resulting in fewer working hours outdoors. The fourth person was a sun worshipper, who in 2000 received 553 SED in estimated annual UVR dose. Risk behaviour accounted for 387 SED, 300 of which were received during a beach holiday on a Greek island. In 2006 her estimated annual UVR dose was 252 SED. Risk behaviour accounted for 179 SED, of which, 65 SED were received during a 4-week backpacking trip to Vietnam. As the trip was focused on sightseeing rather than sunbathing, she received considerably less UVR in 2006.

Risk behaviour

More than anything else it is risk behaviour, especially “sunbathing to get a tan”, which defines the UVR dose level of indoor workers.⁴ Thus, 16 persons with an estimated annual UVR dose above 150 SED received a mean 155 SED and a median 140 SED during risk behaviour. Significant correlations were found between the estimated annual UVR dose and: “numbers of risk behaviour days” (r = 0.72, p < 0.001), “sum of UVR dose during risk behaviour” (r = 0.82, p < 0.0001), and “UVR dose per risk behaviour day in Southern Europe” (r = 0.81, p < 0.0001). Additionally, if the UVR dose is received during risk behaviour, a great part of the body is irradiated which will probably give rise to more skin damage, than if only a small part of the body is exposed.

Outdoor work

4 out of 5 gardeners received a high level UVR dose in 2006 and all 5 also did so in their previous years of participation. As outdoor workers they had a special exposure pattern, which is also found in other studies.^11–14 All 5 had almost twice as much UVR exposure (mean 4.8 h versus 2.5 h per day) as that received by the total group, but received their UVR dose more evenly distributed over the period. The gardeners received their main UVR dose during workdays without risk behaviour.

Outdoor sport

2 of the participants were keen golfers (males) and had a high level of estimated UVR doses in all their 6 years of participation. Their exposure hours (mean 4.8 h per day) were at the same level as the gardeners’. They received the main part of their UVR exposure during “days off without risk behaviour”. This was probably due to the golfers’ dress code, which required a minimum of polo shirt and knee-length shorts, while golf caps are commonly used.

Conclusion

It needs collaboration among specialists to set guidelines for realistic, acceptable sun exposure.^4,15–18 This study shows that people maintain their sun exposure habits, whether they have previously been at a low, medium or high level. We do not know of other studies objectively investigating possible changes in UVR doses over a period of years. Studies based on questionnaires have shown greater awareness of the benefits from siesta and other sun prevention methods, but UVR doses do not seem to be lowered accordingly.¹⁹ Although this study was conducted among Danes at a relatively high latitude, we see no reason why at least indoor workers in other climate zones would act differently from the Danes, since several dosimeter studies from lower latitudes show UVR exposure doses of the same order.^20–23 This is possibly because in warmer climates the heat itself keeps people out of the sun, especially around midday, and thereby counterbalances the higher ambient UVR. Further studies investigating people's sun behaviour objectively over longer periods in different populations are needed to obtain a better understanding of the different sun exposure patterns and thereby make it possible to reduce the UVR dose among people receiving medium and high levels of UVR.

Acknowledgements

This study was supported by funding from the European Community Environment and Climate 1994–1998 Work Programme Contract no. ENV4-CT97-0556. We thank all the participants and the staff of the Department of Dermatology D92.

References

1. B. K. Armstrong and A. Kricker, The epidemiology of UV induced skin cancer, J. Photochem. Photobiol., B, 2001, 63, 8–18.
2. E. Thieden, P. Philipsen, J. Heydenreich and H. C. Wulf, UV radiation exposure related to age, sex, occupation, and sun behavior based on time-stamped personal dosimeter readings, Arch. Dermatol., 2004, 140, 197–203.
3. E. Thieden, P. Philipsen, J. Sandby-Møller, J. Heydenreich and H. C. Wulf, UV radiation exposure Proportion of Lifetime UV Dose received by children, teenagers and adults based on time-stamped personal dosimetry, J. Invest. Dermatol., 2004, 123, 1147–1150.
4. E. Thieden, Sun exposure behaviour among subgroups of the Danish population. Based on personal electronic UVR dosimetry and corresponding exposure diaries, Dan. Med. Bull., 2008, 55, 47–68.
5. B. L. Diffey, C. J. Gibson, R. Haylock and A. F. McKinlay, Outdoor ultraviolet exposure of children and adolescents, Br. J. Dermatol., 1996, 13, 1030–1024.
6. E. Thieden, P. A. Philipsen and H. C. Wulf, Ultraviolet radiation exposure pattern in winter compared with summer based on time-stamped personal dosimeter readings, Br. J. Dermatol., 2006, 154, 133–138.
7. J. Heydenreich and H. C. Wulf, Miniature personal electronic UVR dosimeter with erythema response and time-stamped readings in a wristwatch, Photochem. Photobiol., 2005, 81, 1138–1144.
8. A. F. McKinlay and B. L. Diffey, A reference spectrum for ultraviolet induced erythema in human skin, CIE J., 1987, 6, 17–22.
9. H. C. Wulf, J. Lock-Andersen and The Scandinavian photodermatology reseach group, Standard erythema dose, Skin Res. Technol., 1996, 4, 192.
10. B. L. Diffey, C. T. Jansen, F. Urbach and H. C. Wulf, The standard erythema dose: a new photobiological concept, Photodermatol., Photoimmunol. Photomed., 1997, 13, 64–66.
11. C. D. J. Holman, I. M. Gibson, M. Stephenson and B. K. Armstrong, Ultraviolet-irradiation of human-body sites in relation to occupation and outdoor activity – field studies using personal UVR dosimeters, Clin. Exp. Dermatol., 1983, 8, 269–277.
12. H. P. Gies, C. R. Roy, S. Toomey, R. MacLennan and M. Watson, Solar UVR exposures of three groups of outdoor workers on the Sunshine Coast, Queensland, Photochem. Photobiol., 1995, 62, 1015–1021.
13. M. G. Kimlin, A. V. Parisi and J. C. F. Wong, Quantification of personal solar UV exposure of outdoor workers, indoor workers and adolescents at two locations in Southeast Queensland, Photodermatol., Photoimmunol. Photomed., 1998, 14, 7–11.
14. P. Gies and J. Wright, Measured solar ultraviolet radiation exposures of outdoor workers in Queensland in the building and construction industry, Photochem. Photobiol., 2003, 78, 342–348.
15. E. Thieden, P. Philipsen, J. Sandby-Møller and H. C. Wulf, Sunscreen use related to UV exposure, age, sex and occupation based on time-stamped personal dosimetry and sun behaviour dairies, Arch. Dermatol., 2005, 141, 967–973.
16. E. Thieden, P. Philipsen, J. Sandby-Møller and H. C. Wulf, Sunburn related to UV radiation exposure, age, sex, occupation, and sun bed use based on time-stamped personal dosimetry and sun behavior diaries, Arch. Dermatol., 2005, 141, 482–488.
17. E. Thieden, S. M. Collins, P. A. Philipsen, G. M. Murphy and H. C. Wulf, Ultraviolet exposure patterns of Irish and Danish gardeners during work and leisure, Br. J. Dermatol., 2005, 153, 795–801.
18. M. Naylor and J. K. Robinson, Sunscreen, sun protection, and our many failures, Arch. Dermatol., 2005, 141, 1025res–1027.
19. B. Køster, C. Thorgaard, A. Philip and I. H. Clemmensen, Prevalence of sunburn and sun-related behaviour in the Danish population: a cross-sectional study, Scand. J. Pub. Health, 2010, 39, 548–552.
20. D. Vishvakarman, J. C. F. Wong and B. W. Boreham, Annual occupational exposure to ultraviolet radiation in central Queensland, Health Phys., 2001, 81, 536–544.
21. C. Y. Wright and A. I. Reeder, Youth solar ultraviolet radiation exposure, concurrent activities and sun-protective practices: a review, Photochem. Photobiol., 2005, 81, 1331–1342.
22. D. E. Godar, S. P. Wengraitis, S. P. J. Shreffler and D. H. Sliney, UV doses of Americans, Photochem. Photobiol., 2001, 73, 621–629.
23. D. E. Godar, UV doses worldwide, Photochem. Photobiol., 2005, 81, 736–749.

---

Footnote

† This article is published as part of a themed issue on current topics in photodermatology.

---