
	Version PDF

Does relative melanoma distribution by body site 1960-2004 reflect changes in intermittent exposure and intentional tanning in the Swedish population?

European Journal of Dermatology. Volume 17, Number 5, 428-34, September-October 2007, Clinical report

DOI : 10.1684/ejd.2007.0242

Summary

Author(s) : Henrik Dal, Cecilia Boldemann, Bernt Lindelöf , Centre for Public Health, Stockholm County Council, Stockholm, Sweden, Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden, Department of Dermatology, Karolinska University Hospital, 17176 Stockholm, Sweden.

Summary : Intermittent exposure to UV-radiation at an early age is a known important factor in the aetiology of malignant melanoma. We surveyed data from the Swedish Cancer Registry for melanoma by body site for age and gender cohorts from 1960 to 2004, in an attempt to discern a reflection of major behavioural and societal changes in the relative distribution of melanoma by body site. The study comprised patients with malignant melanoma from the Swedish Cancer Registry, including information on body site of tumour (January 1, 1960 - December 31, 2004). In total, 46,337 malignant melanomas were diagnosed in 44,623 patients. Trends were assessed by incidence per site, and relative site distribution per age group and calendar period, and dividing body sites by exposure type to the sun: head (mostly continuous), trunk (mostly intermittent), and limbs (mixed exposure). Between calendar periods 1960-1964 and 2000-2004 melanomas increased most rapidly on the upper limbs (men 885%, women 1216%) on the trunk (men 729%, women 759%) and on the lower limbs (men 418%, women 289%) in both genders. The incidence increase of head tumors was slower. Across the life span, melanomas of the trunk and lower limbs dominate among patients <\; 70 years, whereas tumors of the head are most common among patients ≥ 70 years. Tumors of the trunk formed an increasing proportion of all melanomas during the period studied, particularly in females. The relative shift of melanomas from the head to the trunk with mostly intermittent UV exposure coincides with behavioral and societal changes with regard to sun exposure. This supports the hypothesis of a relationship between intentional exposure to ultraviolet radiation and malignant melanoma.

Keywords : melanoma, trends, proportion, incidence, body site, Sweden

Pictures

ARTICLE

Auteur(s) : Henrik Dal¹, Cecilia Boldemann^1,2, Bernt Lindelöf³

¹Centre for Public Health, Stockholm County Council, Stockholm, Sweden
²Department of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden
³Department of Dermatology, Karolinska University Hospital, 17176 Stockholm, Sweden

accepté le 23 Avril 2007

Intermittent exposure to ultraviolet (UV) radiation from the sun and sunbeds is an important factor in the etiology of malignant melanoma (hereafter melanoma) [1-4] besides the number of nevi per unit of body surface, freckling tendency and sun sensitivity (skin type), particularly if such exposure occurs at an early age [5, 6]. There is also epidemiological evidence of a dose-response relationship between common nevi in children and parents’ intentional tanning [7]. The low incidence of melanoma in Northern countries half a century ago seems to be the result of behavioral adjustment to UV intensity, probably also implying fewer nevi in the population at that time. Not much is known about induction times between exposure and manifest disease, but it is estimated to vary between a few years [8] to several decades, which complicates the track-down of a link to tanning behavior. Previous studies have observed an increase of melanoma incidence of the trunk and limbs, and thus indicated a behavioral genesis [9-14]. The increment of UV exposure by explicit and implicit intentional outdoor tanning (sun seeking behavior, aquatic and alpine sports, travel to seaside resorts), and the introduction of cosmetic sunbeds in 1978 (Sweden) may reflect in the relative body site distribution of melanoma. An increasing proportion of tumors, particularly with intermittently exposed body sites taking an increasing proportion, would further corroborate behavior as a major etiologic factor. The objective of this study was to establish the trends in relative body site distribution of malignant melanoma during a half century with regard to gender and age, in an attempt to discern behavioral and societal aspects in a descriptive fashion. We surveyed data from the Swedish Cancer Registry (SCR) of melanomas from 5-year calendar periods 1960-1964 to 2000-2004 by body site for age and gender cohorts and by intermittent, continuous and mixed UV exposure type, in an attempt to discern a reflection of major behavioral and societal changes in the relative distribution of melanoma by body site.

Material and methods

Geographical data, study base and study population

Sweden, in Northern Europe, is situated between the 54^th and the 69^th parallels, with yearly ambient solar UVR of 270-420 kJ m² skin surface (horizontal surface) depending on the latitude [15]. The recorded meteorological spring period (diurnal mean temperature 0-10^o centigrade) is short, usually lasting from April to May [16], and summer (diurnal mean temperature >10^o) usually lasts from the end of May until the end of August. Episodes of ozone depletion in Sweden with marked downward deviations of long-range ozone values have also been recorded during the 90s, resulting in a higher episodic influx of UV-radiation during February-May [16].

During the period under study (1960-2004) the population has increased from 7 to 9 million inhabitants, and the percentage living in urban areas and south of the 60^th parallel has increased from 60 to 85% [17]. The population is mobile, with > 7 million registered airline departures to foreign destinations in 2004, of which 860,000 (12%) were departures to seaside resorts south of the 40^th parallel [18].

The cancer registry

In 1958, reporting to the SCR of cancers upon diagnosis became compulsory. Malignancies are reported by clinicians, pathologists and cytologists, with most cancers thus being reported twice [19]. The coverage of the SCR was reported to be 96% in the late 1970s, and close to 100% in 2000 [20]. The register is based on a 10-digit national registration number (NRN) unique for every Swedish resident. The NRN is made up of birth date (year, month, day), plus two digits indicating birthplace, one indicating gender, and a control number. The NRN enables linkage of exposure and outcome data and can be used to test specific etiologic hypotheses relevant for epidemiological surveillance and public health promotion.

Statistical methods

The melanoma cases, diagnosed from 1960 until 2004, and coded according to the 7^th revision of the International Classification of Diseases (ICD-7 code 190.0, 190.1-4 head sites, 190.5 trunk, 190.6 upper limbs, 190.7 lower limbs, 190.8-190.9 multiple parts or unspecified parts) were classified according to gender, age, and calendar period (5-year intervals). In situ lentigo maligna were excluded from analysis. The sites of the head and neck region were joined in one category: head-sites (190.1-4), as were multiple parts and parts unspecified: multiple parts/unspecified (190.8, 9). The site-specific age standardized incidences (European standard population) for each of the five defined anatomical sites were then calculated, including a total of 46,337 melanomas in 44,623 patients. The patients were then divided into the following age groups according to age at diagnosis: 0-29 years, 30-49 years, 50-69 years, and 70+ years. For clarity, melanomas of multiple and unspecified parts were excluded in the presentation of the relative site specific distribution. The patients were born between 1863 and 2001. Of the melanomas, 80% were diagnosed in patients born from 1900-1949. The relative site-specific distribution of diagnosed melanomas over time was calculated for each gender and age group, and thus trends assessed by incidence per site, relative site distribution per age group and calendar period. Further, a categorization was made of body sites by sun exposure type: head (face normally uncovered, exposure mostly continuous), trunk (normally covered, exposure mostly intermittent), and limbs (mixed exposure).

The relative distribution by body site was studied, comparing the youngest and the oldest age-at-diagnosis groups in the first (1960-1964), and the last (2000-2004) calendar periods, meaning that the oldest at diagnosis were born predominantly in the 1890s, and the youngest at diagnosis in the late 1970s and so on.

Results

Between the calendar periods 1960-1964 and 2000-2004, the most rapid incidence increase occurred for melanomas of the upper limbs (men 885%, women 1216%), and of the trunk (men 729%, women 759%), and of the lower limbs (men 418%, women 289%) in both genders. The standardized incidence increase of the upper limbs rose from 0.3 to 2.9 (men), and from 0.3 to 3.39 (women), of the trunk from 1.1 to 9.2 (men), and 0.6 to 5.2 (women), and of the lower limbs from 0.4 to 1.9 (men), and from 1.6 to 6.3 (women). The most frequent body sites of melanoma which have evolved during the studied time period are thus the trunk in both men and women, and lower limbs in women (figure 1A, B). The incidence increase of head tumors was slower and took a decreasing share of all tumors (figure 2A-D). Across the full life span, melanomas of the trunk and lower limbs dominated among patients < 70 years, whereas tumors of the head were most common among patients ≥ 70 years. Melanomas on the intermittently exposed trunk took an increasing proportion of all melanomas during the period studied except in 30-49-year-old men, and in women aged >70 years (figure 2A-D). Comparing the first calendar period of diagnosis 1960-1964 with the most recent one of 2000-2004, head tumors dominated in patients who were aged ≥ 70 years at diagnosis 1960-1964, whereas the trunk and limbs were the dominating body sites in patients aged < 70 years at diagnosis and diagnosed 2000-2004 (figure 2A-D, and table 1).
Table 1 Incident cases of melanoma by studied age group, body part, and most likely type of exposure Sweden 1960-1964, and 2000-2004, unspecified and multiple tumours excluded

	Men		Women
Diagnosed		Diagnosed
Age, body part, and most likely type of exposure	1960-1964 N (%)	2000-2004 N (%)	1960-1964 N (%)	2000-2004 N (%)
Head & neck (continuous)
0-29	10 (22)	16 (14)	11 (18)	17 (8)
30-49	22 (12)	63 (8)	23 (10)	67 (6)
50-69	50 (21)	212 (11)	52 (20)	158 (9)
70+	58 (44)	371 (22)	54 (42)	367 (25)
Trunk (intermittent)
0-29	20 (43)	57 (52)	10 (16)	107 (46)
30-49	113 (62)	446 (58)	65 (28)	411 (38)
50-69	111 (48)	1042 (57)	52 (20)	527 (31)
70+	25 (19)	870 (51)	18 (14)	270 (18)
Upper and lower limbs (mixed exposure)
0-29	16 (35)	37 (34)	41 (66)	107 (46)
30-49	47 (26)	258 (34)	142 (62)	600 (56)
50-69	71 (31)	579 (32)	154 (60)	1002 (60)
70+	48 (37)	467 (27)	57 (44)	844 (57)

Discussion

In terms of relative distribution by body site, the main finding in this study was the rapid incidence increase of melanomas of the trunk and limbs, and the decreased share of melanomas of the head and neck region of all melanomas. A half century ago tumors of the constantly exposed head and neck region dominated (though the incidence was low). Among tumors diagnosed in 2000-2004, intermittent sites dominated both in terms of incidence rise and by relative comparison with the other body sites.

Even though the occurrence of melanomas has not been observed to be related to chronically sun-exposed body sites in the same way as squamous cell cancer, this study may give a clue to an existing dose-response relationship between exposure site and risk of melanoma, as well as a relationship between exposed skin surface and risk of melanoma. Even though dysplastic nevi preferably occur on the trunk [21], there are no indications of an increased proportion of the population carrying a genetic disposition for dysplastic nevi. The rapid relative increase of melanomas of the trunk therefore makes a behavioral genesis even more likely. This has also been stated by a number of studies of white populations in Sweden [9], the United States [10], Canada [11], New Zealand [12], former East Germany [13], and Finland [14].

Lifestyles related to urbanization quickly became established during the period of 1960-2000 when the percentage of the Swedish population living in densely populated areas rose from 60 to 85% [16]. The shift from rural to urban living may have resulted in a higher proportion of intermittent exposure, as it involves lifestyles with an increased risk of such exposure. During the same period the number of airline departures to seaside resorts south of the 40^th parallel increased by 1,229% from 70,000 in 1962 to 860,000 in 2004 [18]. During the same period (1978) cosmetic tanning devices (sunbeds) were introduced on the Swedish market, the use of which became widespread within a short period of time, particularly in the younger population [22-24]. It is reasonable to assume that sun seeking behavior patterns and the use of sunbeds involve intermittent exposures to larger skin surfaces [25] and also increase intermittent exposure on body sites other than the head, compared to unintentional continuous exposure related to occupational exposure such as fishing and farming. Intentional exposure may be closely related to intermittent exposure, the more so as indoor work has increased and thereby continuous, suberythemal UV exposure, possibly protective against melanoma [26], may have declined. It also must be pointed out that continuous sun exposure to the head does not protect against MM. Intentional tanning also involves a considerable increase in horizontal exposure at a right angle, resulting in a further increase of UV exposure of the trunk and legs [27, 28]. The relative increase of melanomas on the arms may also reflect little outdoor stay other than during the summer and/or on holiday, and that – when outdoors- the limbs are covered up not because of high UV-intensity, but because of a cold climate.

The use of sunscreens, which at an early stage were promoted as imperative means of protection, has been questioned due to scanty application and the probable intention of the user to prolong sun exposure [29]. Also, during the first decades of the register period, sunscreens on the market mostly provided protection against UVB, and less against UVA. This may have resulted in an increase of sun exposure on body sites other than the head if sunscreens were used on the trunk and the limbs.

Cosmetics with or without sun protective agents used on the face may have yielded protection, mainly for women. On the other hand, other cosmetics may have photosensitizing properties. Even if the trend of the use of various cosmetics is largely unknown it is reasonable to assume that the use of cosmetics both with and without sunscreens has increased during the studied period.

Divergent etiological pathways for melanoma have been discussed, including etiological factors not related to UV exposure, such as melanocytic proliferation [30]. However, melanomas on intermittently and chronically exposed body sites have been observed to be correlated to an over-expression of p53 [31, 32]. A further aspect is the possible increase in aquatic and alpine sports involving chilled skin in combination with high exposure to ultraviolet radiation, and the implication this has for DNA-repair, as there is evidence of poorer DNA-repair in chilled skin (1,000 dimer repairs/sec vs. 10,000 dimer repairs/sec) [33].

Considering ozone depletion, it has been estimated that a 1% loss of stratospheric ozone would result in a global increase of skin cancers by 1-2% [34]. Episodes of ozone depletion (deviation of long range mean total ozone by mean value per month) have been recorded in Sweden throughout the 90s, with peaks of depletion occurring in the spring months of 1993, 1996, and 2000 [16]. Even though it is unlikely that any cases of melanoma in Sweden so far could be ascribed to ozone depletion, it may in the future have implications for both unintentional and intermittent sun exposure to the head and intentional sun exposure all over the body.

Finally, it can be suspected that in a long-term perspective the increase of melanoma may be diluted due to an upsurge in the use of sunbeds, and hence intermittent exposure of the total skin surface. In terms of anthropometric measures, the rising trends and relative increase of melanoma on the trunk may also be due to the obesity pandemic as the skin surface may have increased and hence more skin cells on these body sites are at risk [35]. A hormonal impact, reflected in increased body height, has also been observed, at least in men [36].

The information available from the SCR includes site, histological type of tumor, hospital and pathology department, specimen number and the year when the specimen was taken. Thus, subtypes of melanoma are not registered with specific codes. The classification of MMs into different subgroups would require an analysis of all the original histopathological reports from all reporting laboratories which is beyond the scope of this study. Furthermore, the site distribution of the decades before 1960 is unknown.

The highest relative melanoma increase in non-head sites was observed in the oldest age-group 70+ years (figure 2A-D) whereas in the younger age groups the proportion of melanomas on non-head sites was already high. If this shift in site-distribution is caused by behavioral changes due to indoor occupations, more leisure time with outdoor activities, intentional tanning, including sunbed use, it would be plausible to assume that these changes – at least in part – were already taking place among younger age groups prior to 1960. Behavioral and societal changes as listed above would be expected to affect the younger population first, so what we observe in the oldest age group is partly the end of “old times” and partly a cohort-effect, as young people keep their habits when growing older. Thus, it could be that major site changes in melanoma affected the younger population in Sweden during a time of increasing employment in industry, and the introduction of vacation regulated by law (two weeks, 1938), and when holidays abroad were rare.

The conclusion is that the shift in relative body-site distribution observed during the past four decades, of the relative increase of melanomas, is congruent with societal changes regarding UV exposure, with migration from rural to urban areas, the introduction of sunbeds, and the upsurge in leisure time involving aquatic sports, and airline travel to seaside resorts. But finding certain evidence of a particular behavior as a cause of MM requires more specific studies of behavior over time.

Acknowledgements

Financial support: none. Conflict of interest: none.

References

1 Veierod MB, Weiderpass E, Thörn M, et al. A prospective study of pigmentation, sun exposure, and risk of cutaneous malignant melanoma in women. J Natl Cancer Inst 2003; 95(20): 1530-8.

2 Walter S, King WD, Marrett LD. Association of cutaneous malignant melanoma with intermittent exposure to ultraviolet radiation: results of a case-control study in Ontario, Canada. Int J Epidemiol 1999; 28(3): 418-27.

3 Westerdahl J, Ingvar C, Masback A, Jonsson N, Olsson H. Risk of cutaneous malignant melanoma in relation to use of sunbeds: further evidence for UV-A carcinogenicity. Br J Cancer 2000; 82(9): 1593-9.

4 Walter S, Marrett L, From L, Hertzmann C, Shannon H, Roy P. The association of cutaneous malignant melanoma with the use of sunbeds and sunlamps. Am J Epidemiol 1990; 131(2): 232-43.

5 Gandini S, Sera F, Cattaruzza MS, et al. Meta-analysis of risk factors for cutaneous melanoma: II. Sun exposure. Eur J Cancer 2005; 41(1): 45-60.

6 MacKie RM, Freudenberger T, Aitchison TC. Personal risk-factor chart for cutaneous melanoma. Lancet 1989; 26; 2(8661): 487-90.

7 Rodvall Y, Wahlgren CF, Ullen H, Wiklund K. Common melanocytic nevi in 7-year-old schoolchildren residing at different latitudes in Sweden. Cancer Epidemiol Biomarkers Prev 2007; 16(1): 122-7.

8 MacKie RM, Aitchison T. Severe sunburn and subsequent risk of primary cutaneous malignant melanoma in Scotland. Br J Cancer 1982; 46(6): 955-60.

9 Thörn M, Bergström R, Adami HO, Ringborg U. Trends in the incidence of malignant melanoma in Sweden, by anatomic site, 1960-1984. Am J Epidemiol 1990; 132(6): 1066-77.

10 Chen YT, Zheng T, Holford TR, Berwick M, Dubrow R. Malignant melanoma incidence in Connecticut (United States): time trends and age-period-cohort modeling by anatomic site. Cancer Causes Control 1994; 5(4): 341-50.

11 Elwood JM, Gallagher RP. Body site distribution of cutaneous malignant melanoma in relationship to patterns of sun exposure. Int J Cancer 1998; 78(3): 276-80.

12 Bulliard JL, Cox B. Cutaneous malignant melanoma in New Zealand: trends by anatomical site, 1969-1993. Int J Epidemiol 2000; 29(3): 416-23.

13 Stang A, Stabenow R, Eisinger B, Jöckel KH. Site- and gender-specific time trend analyses of the incidence of skin melanomas in the former German Democratic Republic (GDR) including 19351 cases. Eur J Cancer 2003; 39(11): 1610-8.

14 Stang A, Pukkala E, Sankila R, Söderman B, Hakulinen T. Time trend analysis of the skin melanoma incidence of Finland from 1953 through 2003 including 16,414 cases. Int J Cancer 2006; 119(2): 380-4.

15 www.smhi.se (continuous update).

16 Josefsson W. Five years of solar UV-radiation monitoring in Sweden. Norrköping. Sweden: SMHI; RMK-report No 71, 1996.

17 Ylander H. Urbanization and densification in Sweden (Urbanisering och tätortsutveckling, in Swedish). Atlas of Sweden’s classification of counties, municipalities, parishes and built-up areas. Sweden: Statistics, 1992.

18 Luftfartsverket (Swedish Civil Aviation Administration). Non-scheduled passenger statistics, 1962-2000. Norrköping, 2006 (www.lfv.se) 24 May 2006.

19 Mattsson B, Wallgren A. Completeness of the Swedish Cancer Register. Non-notified cancer cases recorded on death certificates in 1978. Acta Radiol Oncol 1984; 23: 305-13.

20 National Board of Health and Welfare. The Cancer Registry: Cancer Incidence in Sweden. Stockholm: Centre for Epidemiology, 2004.

21 Stierner U, Augustsson A, Rosdahl I, Suurkula M. Regional distribution of common and dysplastic naevi in relation to melanoma site and sun exposure. A case-control study. Melanoma Res 1992; 1(5-6): 367-75.

22 Boldeman C, Jansson B, Nilsson B, Ullén H. Sunbed use in Swedish urban adolescents related to behavioural characteristics. Prev Med 1997; 26: 114-9.

23 Brandberg Y, Ullen H, Sjöverg L, Holm LE. Sunbathing and sunbed use related to self-image in a randomized sample of Swedish adolescents. Eur J Cancer Prev 1998; 7(4): 321-9.

24 Boldeman C, Bränström R, Dal H, et al. Tanning habits and sunburn in a Swedish population age 13-50 years. Eur J Cancer 2001; 37(18): 2441-8.

25 Wester U, Boldemann C, Jansson B, Ullén H. Population UV-dose and skin area – do sunbeds rival the sun? Health Phys 1999; 77(4): 436-40.

26 White E, Kirkpatrick CS, Lee JA. Case-control study of malignant melanoma in Washington State. I. Constitutional factors and sun exposure. Am J Epidemiol 1994; 139(9): 857-68.

27 Diffey BL, Kerwin M, Davis A. The anatomical distribution of sunlight. Br J Dermatol 1977; 97(4): 407-10.

28 Parisi AV, Kimlin MG, Lester R, Turnbull D. Lower body anatomical distribution of solar ultraviolet radiation in the human form in standing and sitting postures. J Photochem Photobiol 2003; 69: 1-6.

29 Autier P, Dore JF, Négrier S, et al. Sunscreen use and duration of sun exposure: a double-blind, randomized trial. J Natl Cancer I 1999; 91(15): 1304-9; [see comments].

30 Whiteman DC, Watt P, Purdie DM, Hughes MC, Hayward NK, Green AC. Melanocytic nevi, solar keratoses, and divergent pathways to cutaneous melanoma. J Natl Cancer Inst 2003; 95(11): 806-12.

31 Cho E, Rosner BA, Colditz GA. Risk factors for melanoma by body site. Cancer Epidemiol Biomarkers Prev 2005; 14(5): 1241-4.

32 Kim YC, Lee MG, Cho SH, Lee JH, Lee DH, Ahn BK. The relation between p53 expression and physical site of melanoma: an immunohistochemical study. Br J Dermatol 2003; 149(6): 1299-300.

33 Hjertvik M, Erixon K, Ahnström G. Repair of DNA damage in mammalian cells after treatment with UV and dimethyl sulphate: discrimination between nucleotide and base excision repair by their temperature dependence. Mutat Res 1998; 407(2): 87-96.

34 de Gruijl FR, Van der Leun JC. Estimate of the wavelength dependency of ultraviolet carcinogenesis in humans and its relevance to the risk assessment of a stratospheric ozone depletion. Health Phys 1994; 67(3): 1-7.

35 Gallus S, Naldi L, Martin L, Martinelli M, La Vecchia C, Oncology Study Group of the Italian Group for Epidemiologic Research in Dermatology (GISED). Anthropometric measures and risk of cutaneous malignant melanoma: a case-control study from Italy. Melanoma Res 2006; 16(1): 83-7.

36 Shors AR, Solomon C, McTiernan A, White E. Melanoma risk in relation to height, weight, and exercise. Cancer Causes Control 2001(7): 599-606; (United States).