[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure 1.  Male to Female Age-Standardized Incidence Rate Ratios (IRRs) and 95% CIs for the 30 Countries With the Highest Incidence of Melanoma
Male to Female Age-Standardized Incidence Rate Ratios (IRRs) and 95% CIs for the 30 Countries With the Highest Incidence of Melanoma

Data reported by the Global Cancer Observatory.1

aData limited to individuals of white race.

Figure 2.  Male to Female Incidence Rate Ratios (IRRs) During 1982-2015 in 8 Countries
Male to Female Incidence Rate Ratios (IRRs) During 1982-2015 in 8 Countries

The IRRs were calculated from age-standardized rates for men and women (US 2000) and modeled using Joinpoint regression; annual percentage change for each trend is presented in eTable 1 in the Supplement.

aData limited to individuals of white race.

Figure 3.  Male to Female Incidence Rate Ratios (IRRs) for Melanomas by Anatomic Site in 1982-2015
Male to Female Incidence Rate Ratios (IRRs) for Melanomas by Anatomic Site in 1982-2015

The IRRs shown for melanomas developing on the head and neck (A), trunk (B), upper limbs (C), and lower limbs (D) were calculated from age-standardized rates for men and women (US 2000) and modeled using Joinpoint regression models.

aData limited to individuals of white race.

Figure 4.  Male to Female Incidence Rate Ratios (IRRs) by 5-Year Age Groups in 2015 for 8 Countries
Male to Female Incidence Rate Ratios (IRRs) by 5-Year Age Groups in 2015 for 8 Countries

The IRRs were modeled using Joinpoint regression models plotted on a log scale; annual percentage change for each trend is presented in eTable 2 in the Supplement.

aData limited to individuals of white race.

Figure 5.  Male to Female Age-Specific Incidence Rate Ratios (IRRs) for Melanomas in Men vs Women by Anatomic Site in 2015
Male to Female Age-Specific Incidence Rate Ratios (IRRs) for Melanomas in Men vs Women by Anatomic Site in 2015

The IRRs shown for melanomas developing on the head and neck (A), trunk (B), upper limbs (C), and lower limbs (D) were modeled using Joinpoint regression models.

aData limited to individuals of white race.

1.
Ferlay  J, Ervik  M, Lam  F,  et al. Global Cancer Observatory. Cancer today: 2018. Accessed September 25, 2018. https://gco.iarc.fr/today
2.
Olsen  CM, Whiteman  DC. Clinical epidemiology of melanoma. In: Balch C, Atkins M, Garbe C, et al, eds. Cutaneous Melanoma. Springer; 2020.
3.
Anderson  WF, Pfeiffer  RM, Tucker  MA, Rosenberg  PS.  Divergent cancer pathways for early-onset and late-onset cutaneous malignant melanoma.  Cancer. 2009;115(18):4176-4185. doi:10.1002/cncr.24481PubMedGoogle ScholarCrossref
4.
Cho  E, Rosner  BA, Colditz  GA.  Risk factors for melanoma by body site.  Cancer Epidemiol Biomarkers Prev. 2005;14(5):1241-1244. doi:10.1158/1055-9965.EPI-04-0632PubMedGoogle ScholarCrossref
5.
Yuan  TA, Lu  Y, Edwards  K, Jakowatz  J, Meyskens  FL, Liu-Smith  F.  Race-, age-, and anatomic site-specific gender differences in cutaneous melanoma suggest differential mechanisms of early- and late-onset melanoma.  Int J Environ Res Public Health. 2019;16(6):pii:E908. doi:10.3390/ijerph16060908PubMedGoogle Scholar
6.
Linos  E, Swetter  SM, Cockburn  MG, Colditz  GA, Clarke  CA.  Increasing burden of melanoma in the United States.  J Invest Dermatol. 2009;129(7):1666-1674. doi:10.1038/jid.2008.423PubMedGoogle ScholarCrossref
7.
Thrift  AP, Gudenkauf  FJ.  Melanoma incidence among non-Hispanic whites in all 50 United States from 2001 through 2015.  J Natl Cancer Inst. 2019;djz153. doi:10.1093/jnci/djz153PubMedGoogle Scholar
8.
Liu-Smith  F, Farhat  AM, Arce  A,  et al.  Sex differences in the association of cutaneous melanoma incidence rates and geographic ultraviolet light exposure.  J Am Acad Dermatol. 2017;76(3):499-505.e3. doi:10.1016/j.jaad.2016.08.027PubMedGoogle ScholarCrossref
9.
Pérez-Gómez  B, Aragonés  N, Gustavsson  P, Lope  V, López-Abente  G, Pollán  M.  Do sex and site matter? different age distribution in melanoma of the trunk among Swedish men and women.  Br J Dermatol. 2008;158(4):766-772. doi:10.1111/j.1365-2133.2007.08429.xPubMedGoogle ScholarCrossref
10.
MacLennan  R, Kelly  JW, Rivers  JK, Harrison  SL.  The Eastern Australian Childhood Nevus Study: site differences in density and size of melanocytic nevi in relation to latitude and phenotype.  J Am Acad Dermatol. 2003;48(3):367-375. doi:10.1016/S0190-9622(03)70143-1PubMedGoogle ScholarCrossref
11.
Autier  P, Boniol  M, Severi  G, Pedeux  R, Grivegnée  AR, Doré  JF.  Sex differences in numbers of nevi on body sites of young European children: implications for the etiology of cutaneous melanoma.  Cancer Epidemiol Biomarkers Prev. 2004;13(12):2003-2005.PubMedGoogle Scholar
12.
Harrison  SL, Buettner  PG, MacLennan  R.  Body-site distribution of melanocytic nevi in young Australian children.  Arch Dermatol. 1999;135(1):47-52. doi:10.1001/archderm.135.1.47PubMedGoogle ScholarCrossref
13.
Gallagher  RP, McLean  DI, Yang  CP,  et al.  Anatomic distribution of acquired melanocytic nevi in white children—a comparison with melanoma: the Vancouver Mole Study.  Arch Dermatol. 1990;126(4):466-471. doi:10.1001/archderm.1990.01670280050008PubMedGoogle ScholarCrossref
14.
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):806-812. doi:10.1093/jnci/95.11.806PubMedGoogle ScholarCrossref
15.
Whiteman  DC, Green  AC, Olsen  CM.  The growing burden of invasive melanoma: projections of incidence rates and numbers of new cases in six susceptible populations through 2031.  J Invest Dermatol. 2016;136(6):1161-1171. doi:10.1016/j.jid.2016.01.035PubMedGoogle ScholarCrossref
16.
Olsen  CM, Green  AC, Pandeya  N, Whiteman  DC.  Trends in melanoma incidence rates in eight susceptible populations through 2015.  J Invest Dermatol. 2019;139(6):1392-1395. doi:10.1016/j.jid.2018.12.006PubMedGoogle ScholarCrossref
17.
National Cancer Institute. Surveillance, Epidemiology; End Results (SEER) Program. SEER*Stat Database. Incidence—SEER 9 Regs Research Data, Nov 2017 Sub (1973-2015) <Katrina/Rita Population Adjustment>: linked to county attributes—total US, 1969-2016 counties, released April 2018, based on the November 2017 submission. Published 2018. Accessed January 19, 2019. http://www.seer.cancer.gov
19.
Office for National Statistics. UK population single year 2012 to 2015. Accessed February 16, 2019. http://www.ons.gov.uk/ons/taxonomy/index.html?nsc1/4Population#tab-data-tables
20.
NORDCAN. Association of the Nordic Cancer Registries. Danish Cancer Society. Cancer incidence, mortality, prevalence and survival in the Nordic countries, version 7.0 (17.12.2014); 2010. Accessed February 14, 2015. http://www.ancr.nu
21.
Australian Institute of Health and Welfare (AIHW). Cancer data in Australia. Australian Cancer Incidence and Mortality (ACIM) books: melanoma of the skin. Cancer Incidence and Mortality (ACIM) books. Updated July 26, 2019. Accessed January 15, 2019. https://www.aihw.gov.au/reports/cancer/cancer-data-in-australia/
22.
Hudson  DJ.  Fitting segmented curves whose join points have to be estimated.  J Am Stat Assoc. 1966;61(316):1097-1129. doi:10.1080/01621459.1966.10482198Google ScholarCrossref
23.
Rosenberg  PS, Check  DP, Anderson  WF.  A web tool for age-period-cohort analysis of cancer incidence and mortality rates.  Cancer Epidemiol Biomarkers Prev. 2014;23(11):2296-2302. doi:10.1158/1055-9965.EPI-14-0300PubMedGoogle ScholarCrossref
24.
Szabo  G.  The number of melanocytes in human epidermis.  BMJ. 1954;1(4869):1016-1017. doi:10.1136/bmj.1.4869.1016PubMedGoogle ScholarCrossref
25.
Quevedo  WC  Jr, Szabó  G, Virks  J, Sinesi  SJ.  Melanocyte populations in UV-irradiated human skin.  J Invest Dermatol. 1965;45(4):295-298. doi:10.1038/jid.1965.131PubMedGoogle ScholarCrossref
26.
Gilchrest  BA, Blog  FB, Szabo  G.  Effects of aging and chronic sun exposure on melanocytes in human skin.  J Invest Dermatol. 1979;73(2):141-143. doi:10.1111/1523-1747.ep12581580PubMedGoogle ScholarCrossref
27.
Visconti  A, Ribero  S, Sanna  M, Spector  TD, Bataille  V, Falchi  M.  Body site–specific genetic effects influence naevus count distribution in women.  Pigment Cell Melanoma Res. 2019. PubMedGoogle Scholar
28.
Chang  YM, Barrett  JH, Bishop  DT,  et al.  Sun exposure and melanoma risk at different latitudes: a pooled analysis of 5700 cases and 7216 controls.  Int J Epidemiol. 2009;38(3):814-830. doi:10.1093/ije/dyp166PubMedGoogle ScholarCrossref
29.
Cockburn  M, Swetter  SM, Peng  D, Keegan  TH, Deapen  D, Clarke  CA.  Melanoma underreporting: why does it happen, how big is the problem, and how do we fix it?  J Am Acad Dermatol. 2008;59(6):1081-1085. doi:10.1016/j.jaad.2008.08.007PubMedGoogle ScholarCrossref
30.
Bulliard  JL, Cox  B.  Cutaneous malignant melanoma in New Zealand: trends by anatomical site, 1969-1993.  Int J Epidemiol. 2000;29(3):416-423.PubMedGoogle Scholar
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 583
    Citations 0
    Original Investigation
    March 25, 2020

    Evaluation of Sex-Specific Incidence of Melanoma

    Author Affiliations
    • 1Cancer Control Group, QIMR Berghofer Medical Research Institute, Brisbane, Queensland, Australia
    • 2Faculty of Medicine, The University of Queensland, Brisbane, Queensland, Australia
    • 3Melanoma Institute Australia, The University of Sydney, Sydney, New South Wales, Australia
    • 4The University of Queensland School of Public Health, Brisbane, Queensland, Australia
    JAMA Dermatol. Published online March 25, 2020. doi:10.1001/jamadermatol.2020.0470
    Key Points

    Question  Do observed differences in melanoma incidence between men and women vary by population, age, or anatomic site?

    Findings  In a cross-sectional study using data on predominantly fair-skinned populations in 8 countries, an analysis of sex- and site-specific temporal trends in melanoma incidence from 1982 to 2015 suggest that the rate at which melanoma develops differ consistently by body site and age, as well as geographic location. In addition, these apparent site-specific differences were modified by sex.

    Meaning  Sex differences in melanoma incidence patterns across time and latitude suggest etiologic heterogeneity for melanomas arising on different anatomic sites for men and women.

    Abstract

    Importance  Men and women develop melanoma at different rates on different body sites, with variation across countries, but explanations for these disparities remain elusive.

    Objective  To test whether observed differences in melanoma incidence between men and women vary by population, age, or anatomic site.

    Design  Cross-sectional analysis of sex- and site-specific temporal trends in melanoma incidence over 3 decades was conducted for men and women diagnosed with invasive melanoma in the US (limited to white race), Canada, Australia, New Zealand, the UK, Sweden, Norway, and Denmark. Using cancer registry data, male to female incidence rate ratios (IRRs) were calculated overall and by anatomic site, and Joinpoint regression models were used to estimate the annual percentage rate changes in sex- and site-specific incidence in each population. Incidence rates were standardized to the US 2000 population. Data on the incidence between January 1, 1982, and December 31, 2015, were obtained; analysis was conducted from March 1 to October 15, 2019.

    Main Outcomes and Measures  Male to female IRRs and annual percentage change in rates.

    Results  Total melanoma incidence was higher in men than women in US individuals (limited to white race), Canada, Australia, and New Zealand, but not in Denmark, the UK, Norway, and Sweden. In all populations, men had higher rates of melanoma of the head and neck and trunk than women (male to female IRR >1), but lower melanoma rates on the lower limbs (ie, male to female IRR approximately 0.5). The male to female IRR increased log linearly with age, with excess melanomas in women younger than 45 years in all populations (eg, IRR for 20-24 y age group, 0.3 in Denmark and 0.7 in Australia), and excess melanomas in men older than 69 years (eg, IRR for 70-74 y age group, 1.1 in Denmark and 2.1 in the US white population). The age at which the melanoma incidence in men exceeded the melanoma incidence in women differed by population, being achieved the earliest in Australia (45-49 years) and latest in Denmark (65-69 years).

    Conclusions and Relevance  In predominantly fair-skinned populations, melanoma incidence appears to differ systematically and consistently between men and women by age and anatomic site.

    Introduction

    Men and women develop melanoma at different rates in the US and other countries, but explanations for the observed disparities remain elusive and it is unclear to what extent the phenomena are due to innate biological differences (eg, cellular and hormonal) or different behaviors (eg, patterns of sun exposure, clothing, and sun protection). Global data on melanoma incidence for 2018 show a wide variation in the male to female incidence rate ratio (IRR); for the 30 countries with the highest melanoma incidence,1 the male to female IRR ranged from 0.68 (95% CI, 0.63-0.73) in Denmark up to 1.35 (95% CI, 1.33-1.37) in the US white population and 1.47 (95% CI, 1.42-1.52) in Australia (Figure 1). Data from North America, Europe, Australia, and New Zealand all suggest that the incidence is generally higher in women than men up to the age of approximately 50 years, after which higher rates prevail in men.2 The excess among men at older ages is most notable in the high-incidence populations of Australia and New Zealand, and data from the US suggest that the sex-specific pattern by age is similar across all histologic subtypes of melanoma.3

    Reports in the literature also point to differences between men and women with respect to the anatomic distribution of melanoma, with melanomas more likely to arise on the trunk in men and on the lower limbs in women.4 The observed differences in anatomic distribution by sex have commonly been attributed to behavioral differences between men and women in relation to sun exposure. However, there are few data comparing incidence temporal trends across populations, with previous studies conducted either in single populations,3,5-7 at single points in time8 or restricted to particular anatomic sites.9 An alternative explanation for male to female differences in melanoma incidence may be that men and women differ in the susceptibility of melanocytes on different body sites. In support of the latter hypothesis, the body site distribution of melanoma in men and women accords with the distribution of nevi in male and female children,10-13 and molecular and genetic studies provide evidence for divergent biological pathways to melanoma development that are site related.14

    Data have been published on trends in melanoma incidence for the US population limited to white race (hereafter, US white individuals) and for 7 other populations across a broad range of latitudes and ambient UV radiation exposure levels.15,16 One study and its update examined trends in melanoma incidence from January 1, 1982, to December 31, 2015 in 4 populations comprising mainly descendants of migrants from Europe and other continents (US white individuals, Canada, Australia, and New Zealand) and 4 populations comprising European countries where the populations have been largely static for many centuries (the UK, Norway, Sweden, and Denmark).15,16 The results showed generally increasing incidence rates in all countries except New Zealand and Denmark; however, these trends were not examined according to sex or anatomic site.

    To examine why men and women develop melanomas at different rates, and why the male to female IRR varies from country to country, we explored the potential association of age, time period, and anatomic site with sex-specific melanoma incidence. We performed this analysis using long-term melanoma incidence data from the same 8 populations as in the previous study by Olsen et al16 that reside in regions that differ widely in their ambient solar radiation and seasonal temperature ranges. We aimed to test whether the observed differences in melanoma incidence between men and women vary by population, age, or anatomic site.

    Methods

    We obtained age- and sex-specific data on incident invasive, histologically confirmed melanoma cases from population-based cancer registries in the US, the UK, Norway, Sweden, Denmark, Australia, and New Zealand from January 1, 1982, to December 31, 2015. Incidence data were unavailable for Canada (except Quebec) before 1992. United States data were obtained from the Surveillance, Epidemiology, and End Results program of the National Cancer Institute (9 registries, covering approximately 9.4% of the US population).17 For the UK, we obtained melanoma incidence data from cancer registries in Scotland, Wales, and Northern Ireland, and from the Office for National Statistics for England18; for all countries in the UK we obtained population denominators from the Office for National Statistics.19 Melanoma incidence data and population denominators for Sweden, Norway, and Denmark were sourced from NORDCAN.20 Australian melanoma incidence and population data were obtained from the Australian Institute of Health and Welfare.21 All data for Canada were obtained from Statistics Canada (all provinces except Quebec), and for New Zealand via request from Statistics New Zealand.

    We obtained anatomic site-specific incidence data from 5 jurisdictions. Site-specific data for Australia as a whole were not available; therefore, we obtained data for the state of Queensland (1982-2015) from the population-based Queensland Cancer Registry. Queensland is the most northerly state, with a latitude range of 10° to 28° S; it experiences the highest melanoma incidence of all Australian states and territories. We obtained site-specific data for white individuals in the US for 1982-2015 from the Surveillance, Epidemiology, and End Results database. The cancer registries of Norway and Sweden provided site-specific incidence data for 1982-2017, and Statistics New Zealand provided data for 1995-2015. We examined 4 aggregated anatomic sites: head and neck, trunk, upper limbs, and lower limbs. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies.

    Approvals from the respective cancer registries were obtained before the data were released. No potentially identifying information was released; thus, this study was exempt from approval and informed patient consent under the ethical review policies of the QIMR Berghofer Medical Research Institute-Human Research Ethics Committee.

    Statistical Analysis

    Data analysis was conducted from March 1 to October 15, 2019. To examine population trends in age-standardized incidence rates according to sex, we used Joinpoint regression models. We calculated the average annual percentage rate change (APC) in the incidence of invasive melanoma for men and women separately using the Joinpoint Regression Program, version 4.0.4 (National Cancer Institute), with Hudson’s22 continuous fitting algorithm. We first examined sex-specific incidence trends overall for all 8 countries and then sex- and site-specific incidence trends for US white individuals, Norway, Sweden, New Zealand, and Queensland. All incidence rates were standardized to the US 2000 population. We calculated male to female IRR as the ratio of the age-standardized rates comparing men and women for each year of the study period. We assessed the trend in IRRs using Joinpoint regression analyses.

    To simultaneously explore time-dependent associations between men and women for melanomas arising at different anatomic sites, we constructed age-period cohort models using the National Institutes of Health National Cancer Institute web tool (available at https://analysistools.nci.nih.gov/apc/).23 These analyses were restricted to US white individuals, Queensland, Norway, and Sweden—populations with contrasting ambient UV environments and for which site-specific data were available for the entire study period. We examined various APC parameters, including net drift, local drifts, and longitudinal fitted age-at-onset curves. Net drift describes the estimated APC for all age groups adjusted for cohort effects. Local drifts describe the age-specific estimated APCs and provide information about significant cohort effects. If temporal patterns are the same in every age group, local drifts do not differ from the net drift. The fitted longitudinal (ie, cohort specific) age-at-onset curve extrapolates from observed age-specific rates of the full range of birth cohorts to estimate past, current, and future rates for a referent cohort (ie, 1955 in this study). The fitted curve is therefore adjusted for both calendar period and birth-cohort associations and summarizes the age-associated natural history of the disease.

    To examine population trends in age-specific incidence rates according to sex, we first calculated IRRs for all 8 countries and for the most recent year of data available (2015). We examined the trend in the IRRs according to age, using Joinpoint regression analyses. For US white individuals, Queensland, Norway, and Sweden, we then examined fitted longitudinal age-at-onset curves derived from APC models for melanoma of different body sites. All statistical tests were 2-sided, and P values <.05 were considered statistically significant.

    Results

    During 1982-2015, the age-standardized incidence of invasive melanoma for all body sites combined was higher in men than women in the populations of US white individuals, Canada, Australia, and New Zealand (eFigure 1 in the Supplement). In contrast, melanoma rates were higher in women than men in Denmark for the entire study period, in the UK from 1982 to 1998, and in Norway from 1982 to 1989 (eFigure 2 in the Supplement). Melanoma incidence rates were similar for men and women in Sweden for the entire study period and for the more recent time periods in the UK and Norway (eFigure 2 in the Supplement). In general, over the 3 decades of data examined, the male-to-female melanoma IRRs were higher (ie, IRR>1) and increased in magnitude to a greater extent in the US white, Canada, Australia, and New Zealand populations than in Europe (Figure 2); the IRRs were highest in US white individuals and Australia across the entire time period. In all countries except Sweden and Denmark there was a steep increase in the IRR up to the early- to mid-1990s (APC ranging from 1.7% to 3.5%), followed by a stabilization or more gradual increases to 2015 (Figure 2; eTable 1 in the Supplement).

    We examined temporal trends in sex-specific melanoma incidence by anatomic site in the 5 populations for which historical site-specific data were available (US white individuals; Queensland, Australia; New Zealand; Norway; and Sweden) (eFigure 3 in the Supplement). In all jurisdictions, men had consistently higher rates of melanoma of the head and neck and trunk than women (ie, IRR>1) (Figure 3A and B). The highest male to female IRRs were observed in US white individuals (approximately 3-fold excess) and in Australia and New Zealand (approximately 2-fold excess) for melanomas of the head and neck, and in Australia and New Zealand (approximately 2.5-fold excess) for melanomas arising on the trunk. We observed consistently lower rates of melanoma developing on the lower limbs in men than women in all populations (ie, IRR approximately 0.5) (Figure 3D). Melanomas of the upper limbs occurred with similar frequency in men and women in all countries (Figure 3C).

    We observed a general trend of increasing male to female IRRs with increasing age, with IRRs less than 1 (ie, female excess of melanoma) in all populations younger than 45 years, and IRRs greater than 1 in those older than 69 years. The rate of change in IRR with age was most notable in US white individuals, New Zealand, and Canada (Figure 4). The excess of melanoma in men compared with women increased at a rate of 2.9% to 3.4% per 5-year intervals of age in US white individuals, New Zealand, and Canada, compared with 2.0% to 2.5% per 5-year intervals of age in the European countries (eTable 2 in the Supplement). The age at which the IRR equaled 1 differed by population, being achieved earliest in Australia (45-49 years), followed by New Zealand (50-54 years); US white individuals, Canada, the UK, and Norway (55-59 years); Sweden (60-64 years); and Denmark (65-69 years).

    When we examined the male to female IRRs at different ages by anatomic site, we found that increases in IRR in each population were associated largely with changes in the incidence of melanomas of the head and neck and trunk sites (Figure 5A and B). In contrast, we found that melanomas occurred more frequently on the lower limbs in women than men in all populations at all ages (Figure 5D), and that this ratio remained essentially constant over the observation period.

    In addition, we derived age-at-onset curves from APC models for melanomas arising at different anatomic sites (eFigure 4 in the Supplement). These analyses compiled age-, sex-, and site-specific incidence data across consecutive periods for each population to derive a comprehensive overview of patterns of trends. The analyses showed that the male excess for trunk melanoma occurred at a younger age than for head and neck melanoma (mean, 45.0 vs 52.5 years), and the female excess of melanoma on the lower limbs was apparent even earlier (mean, 27.5 years).

    Discussion

    To explore the questions of why men and women develop melanomas at different rates and why the male-to-female IRR varies from country to country, we examined temporal trends in melanoma incidence by sex for 8 populations. The findings suggest that (1) women have higher rates of melanoma than men in early life in all countries, (2) men have higher rates of melanoma than women in late life in all countries, and (3) these patterns are due to sex-specific differences in melanoma incidence at specific anatomic sites (ie, higher rates of lower limb melanoma in early life in women and higher rates of head and neck melanoma in men in later life).

    From 1982 to 2015, the male-to-female IRR increased in US white individuals, Canada, Australia, New Zealand, the UK, and Norway, but remained stable in Sweden and Denmark. When we examined the trends according to body site of melanoma, the increase in the male-to-female IRR in US white individuals, Australia, and New Zealand appeared to be associated with higher rates of head and neck melanomas and, to a lesser extent, trunk melanomas in men compared with women.

    An examination of sex-specific trends in age-specific incidence rates suggested that, in all 8 populations, women in the younger age groups have higher rates than men, but men take the lead at approximately mid-life. The age at which this switch occurs varies by population; it develops earliest in the high ambient UV radiation environment of Australia, between ages 45 and 49 years, and latest in Denmark, between ages 65 and 69 years. These trends appear to reflect sex differences in melanoma incidence by anatomic site, with rates of head and neck and upper limb melanoma diverging between men and women at older ages compared with melanoma of the trunk and lower limbs. The female predominance in younger age groups appears to reflect higher rates of melanoma of the lower limbs.

    Others have reported male-to-female differences in melanoma incidence either in single populations,3,5-7 at a single point in time,8 or for particular sites.9 To our knowledge, the analyses presented herein constitute the first comprehensive comparison across multiple populations over an extended period, with detailed statistical interrogation of incidence trends according to sex and body site.

    Although sex differences in melanoma incidence across different anatomic sites have often been attributed to behavioral differences, such as those related to clothing preferences, there are also credible biological explanations. The distribution of melanocytes is known to differ by anatomic site,24,25 being generally most numerous per unit of body surface area on the head and neck, intermediate on the limbs, and lowest on the trunk.26 Sex differences in the anatomic distribution of melanoma resemble the sex-specific distributions of nevi observed in children and adolescents.10-13 For example, Autier and colleagues11 found that in European schoolchildren, boys had more nevi on the head and neck and trunk than girls, but fewer nevi than girls on the upper and lower limbs. Gallagher and colleagues13 observed the same patterns in Canadian adolescents. Recently published data from the TwinsUK cohort have suggested that the sex difference in nevus distribution on the lower limbs persists throughout life and has strong heritability (being the strongest of all body sites for women).27

    Our findings suggest that there are differences in the rates at which melanoma develops on various body sites and that these differences are modified by sex. As such, these observations accord with the so-called divergent pathway hypothesis for melanoma development.14 Melanoma of the trunk in men and of the lower limbs in women may be predominantly genetically determined and less related to sun exposure, occurring more often in younger age groups. In contrast, melanoma of the head and neck and upper limbs may be more related to cumulative sun exposure, occurring more often in older age groups. The increase in male-to-female incidence in populations comprising mainly descendants of migrants from Europe and other continents over time is associated largely with higher rates of head and neck melanoma in men, and probably reflects higher levels of cumulative exposure to high ambient UV radiation. This finding is consistent with other research reporting that cumulative sun exposure is more strongly associated with melanoma at low latitudes.28

    Strength and Limitations

    This study has both strengths and limitations. A strength of this study was our ability to examine incidence patterns between sexes and among anatomic sites simultaneously using registry-captured, histologically confirmed melanoma incidence data. A potential limitation may be differential underreporting of melanoma diagnoses to cancer registries across the countries included in these analyses, the extent of which is difficult to determine. Underreporting of melanoma has been documented in some Surveillance, Epidemiology, and End Results registries at different times29 and was known to occur in New Zealand prior to 199430; however, underreporting is unlikely to be differential across sex or anatomic site. Our analyses were restricted to broad anatomic sites, and it is possible that the incidence of melanomas arising at different subsites of the trunk (ie, back, chest, and abdomen), lower limb (ie, leg, thigh, and buttock), and head and neck (scalp, face, and neck) regions also differ by sex. A comparison of the sex-specific incidence of melanoma of these sites across all populations may provide a greater understanding of the trends that we have reported.

    Conclusions

    The temporal patterns of melanoma incidence by sex, age, and anatomic site across populations appear to be consistent with a complex interplay of innate and external factors influencing melanoma development. Specifically, we observed higher rates of melanoma in women before mid-life in all countries, mostly owing to higher rates of lower limb lesions. We observed an excess in men in all countries following mid-life, owing largely to higher rates of melanoma of the head and neck, and most notable in the highest incidence populations of Australia and New Zealand that experience high ambient UV radiation. We believe future research should focus on finer subdivisions of anatomic site. Further insights may be gained through the molecular characterization of melanomas arising at different body sites, not only in terms of disease source, but also tumor biological characteristics and prognosis. To date, this research has been limited by the number and diversity of primary tumors and does not have a global reach. In addition, understanding the genetic basis of nevus and melanoma body site distribution in men and women may assist in explaining differences in sex- and site-specific melanoma incidence.

    Back to top
    Article Information

    Accepted for Publication: February 5, 2020.

    Corresponding Author: David C. Whiteman, MBBS, Cancer Control Group, QIMR Berghofer Medical Research Institute, 300 Herston Rd, Herston, QLD 4006, Australia (david.whiteman@qimrberghofer.edu.au).

    Published Online: March 25, 2020. doi:10.1001/jamadermatol.2020.0470

    Author Contributions: Drs Olsen and Whiteman had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Olsen, Thompson, Whiteman.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: All authors.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Olsen, Pandeya, Whiteman.

    Obtained funding: Olsen, Whiteman.

    Administrative, technical, or material support: Whiteman.

    Supervision: Whiteman.

    Conflict of Interest Disclosures: Dr Thompson reported receiving honoraria for Advisory Board participation from Merck Sharp & Dohme Australia and Bristol-Myers Squibb Australia. He reported receiving honoraria and travel expenses from GSK and Provectus Inc. Dr Whiteman reported receiving research funding from the National Health and Medical Research Council of Australia. No other disclosures were reported.

    Funding/Support: This work was supported by National Health and Medical Research Council of Australia grants APP1073898 and APP1058522 to Dr Whiteman.

    Role of the Funder/Sponsor: The National Health and Medical Research Council of Australia had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Disclaimer: The interpretation and reporting of these data are the sole responsibility of the authors, and no endorsement by the Cancer Registry of Norway is intended or should be inferred.

    References
    1.
    Ferlay  J, Ervik  M, Lam  F,  et al. Global Cancer Observatory. Cancer today: 2018. Accessed September 25, 2018. https://gco.iarc.fr/today
    2.
    Olsen  CM, Whiteman  DC. Clinical epidemiology of melanoma. In: Balch C, Atkins M, Garbe C, et al, eds. Cutaneous Melanoma. Springer; 2020.
    3.
    Anderson  WF, Pfeiffer  RM, Tucker  MA, Rosenberg  PS.  Divergent cancer pathways for early-onset and late-onset cutaneous malignant melanoma.  Cancer. 2009;115(18):4176-4185. doi:10.1002/cncr.24481PubMedGoogle ScholarCrossref
    4.
    Cho  E, Rosner  BA, Colditz  GA.  Risk factors for melanoma by body site.  Cancer Epidemiol Biomarkers Prev. 2005;14(5):1241-1244. doi:10.1158/1055-9965.EPI-04-0632PubMedGoogle ScholarCrossref
    5.
    Yuan  TA, Lu  Y, Edwards  K, Jakowatz  J, Meyskens  FL, Liu-Smith  F.  Race-, age-, and anatomic site-specific gender differences in cutaneous melanoma suggest differential mechanisms of early- and late-onset melanoma.  Int J Environ Res Public Health. 2019;16(6):pii:E908. doi:10.3390/ijerph16060908PubMedGoogle Scholar
    6.
    Linos  E, Swetter  SM, Cockburn  MG, Colditz  GA, Clarke  CA.  Increasing burden of melanoma in the United States.  J Invest Dermatol. 2009;129(7):1666-1674. doi:10.1038/jid.2008.423PubMedGoogle ScholarCrossref
    7.
    Thrift  AP, Gudenkauf  FJ.  Melanoma incidence among non-Hispanic whites in all 50 United States from 2001 through 2015.  J Natl Cancer Inst. 2019;djz153. doi:10.1093/jnci/djz153PubMedGoogle Scholar
    8.
    Liu-Smith  F, Farhat  AM, Arce  A,  et al.  Sex differences in the association of cutaneous melanoma incidence rates and geographic ultraviolet light exposure.  J Am Acad Dermatol. 2017;76(3):499-505.e3. doi:10.1016/j.jaad.2016.08.027PubMedGoogle ScholarCrossref
    9.
    Pérez-Gómez  B, Aragonés  N, Gustavsson  P, Lope  V, López-Abente  G, Pollán  M.  Do sex and site matter? different age distribution in melanoma of the trunk among Swedish men and women.  Br J Dermatol. 2008;158(4):766-772. doi:10.1111/j.1365-2133.2007.08429.xPubMedGoogle ScholarCrossref
    10.
    MacLennan  R, Kelly  JW, Rivers  JK, Harrison  SL.  The Eastern Australian Childhood Nevus Study: site differences in density and size of melanocytic nevi in relation to latitude and phenotype.  J Am Acad Dermatol. 2003;48(3):367-375. doi:10.1016/S0190-9622(03)70143-1PubMedGoogle ScholarCrossref
    11.
    Autier  P, Boniol  M, Severi  G, Pedeux  R, Grivegnée  AR, Doré  JF.  Sex differences in numbers of nevi on body sites of young European children: implications for the etiology of cutaneous melanoma.  Cancer Epidemiol Biomarkers Prev. 2004;13(12):2003-2005.PubMedGoogle Scholar
    12.
    Harrison  SL, Buettner  PG, MacLennan  R.  Body-site distribution of melanocytic nevi in young Australian children.  Arch Dermatol. 1999;135(1):47-52. doi:10.1001/archderm.135.1.47PubMedGoogle ScholarCrossref
    13.
    Gallagher  RP, McLean  DI, Yang  CP,  et al.  Anatomic distribution of acquired melanocytic nevi in white children—a comparison with melanoma: the Vancouver Mole Study.  Arch Dermatol. 1990;126(4):466-471. doi:10.1001/archderm.1990.01670280050008PubMedGoogle ScholarCrossref
    14.
    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):806-812. doi:10.1093/jnci/95.11.806PubMedGoogle ScholarCrossref
    15.
    Whiteman  DC, Green  AC, Olsen  CM.  The growing burden of invasive melanoma: projections of incidence rates and numbers of new cases in six susceptible populations through 2031.  J Invest Dermatol. 2016;136(6):1161-1171. doi:10.1016/j.jid.2016.01.035PubMedGoogle ScholarCrossref
    16.
    Olsen  CM, Green  AC, Pandeya  N, Whiteman  DC.  Trends in melanoma incidence rates in eight susceptible populations through 2015.  J Invest Dermatol. 2019;139(6):1392-1395. doi:10.1016/j.jid.2018.12.006PubMedGoogle ScholarCrossref
    17.
    National Cancer Institute. Surveillance, Epidemiology; End Results (SEER) Program. SEER*Stat Database. Incidence—SEER 9 Regs Research Data, Nov 2017 Sub (1973-2015) <Katrina/Rita Population Adjustment>: linked to county attributes—total US, 1969-2016 counties, released April 2018, based on the November 2017 submission. Published 2018. Accessed January 19, 2019. http://www.seer.cancer.gov
    19.
    Office for National Statistics. UK population single year 2012 to 2015. Accessed February 16, 2019. http://www.ons.gov.uk/ons/taxonomy/index.html?nsc1/4Population#tab-data-tables
    20.
    NORDCAN. Association of the Nordic Cancer Registries. Danish Cancer Society. Cancer incidence, mortality, prevalence and survival in the Nordic countries, version 7.0 (17.12.2014); 2010. Accessed February 14, 2015. http://www.ancr.nu
    21.
    Australian Institute of Health and Welfare (AIHW). Cancer data in Australia. Australian Cancer Incidence and Mortality (ACIM) books: melanoma of the skin. Cancer Incidence and Mortality (ACIM) books. Updated July 26, 2019. Accessed January 15, 2019. https://www.aihw.gov.au/reports/cancer/cancer-data-in-australia/
    22.
    Hudson  DJ.  Fitting segmented curves whose join points have to be estimated.  J Am Stat Assoc. 1966;61(316):1097-1129. doi:10.1080/01621459.1966.10482198Google ScholarCrossref
    23.
    Rosenberg  PS, Check  DP, Anderson  WF.  A web tool for age-period-cohort analysis of cancer incidence and mortality rates.  Cancer Epidemiol Biomarkers Prev. 2014;23(11):2296-2302. doi:10.1158/1055-9965.EPI-14-0300PubMedGoogle ScholarCrossref
    24.
    Szabo  G.  The number of melanocytes in human epidermis.  BMJ. 1954;1(4869):1016-1017. doi:10.1136/bmj.1.4869.1016PubMedGoogle ScholarCrossref
    25.
    Quevedo  WC  Jr, Szabó  G, Virks  J, Sinesi  SJ.  Melanocyte populations in UV-irradiated human skin.  J Invest Dermatol. 1965;45(4):295-298. doi:10.1038/jid.1965.131PubMedGoogle ScholarCrossref
    26.
    Gilchrest  BA, Blog  FB, Szabo  G.  Effects of aging and chronic sun exposure on melanocytes in human skin.  J Invest Dermatol. 1979;73(2):141-143. doi:10.1111/1523-1747.ep12581580PubMedGoogle ScholarCrossref
    27.
    Visconti  A, Ribero  S, Sanna  M, Spector  TD, Bataille  V, Falchi  M.  Body site–specific genetic effects influence naevus count distribution in women.  Pigment Cell Melanoma Res. 2019. PubMedGoogle Scholar
    28.
    Chang  YM, Barrett  JH, Bishop  DT,  et al.  Sun exposure and melanoma risk at different latitudes: a pooled analysis of 5700 cases and 7216 controls.  Int J Epidemiol. 2009;38(3):814-830. doi:10.1093/ije/dyp166PubMedGoogle ScholarCrossref
    29.
    Cockburn  M, Swetter  SM, Peng  D, Keegan  TH, Deapen  D, Clarke  CA.  Melanoma underreporting: why does it happen, how big is the problem, and how do we fix it?  J Am Acad Dermatol. 2008;59(6):1081-1085. doi:10.1016/j.jaad.2008.08.007PubMedGoogle ScholarCrossref
    30.
    Bulliard  JL, Cox  B.  Cutaneous malignant melanoma in New Zealand: trends by anatomical site, 1969-1993.  Int J Epidemiol. 2000;29(3):416-423.PubMedGoogle Scholar
    ×