PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Cancer Prev Res (Phila). Author manuscript; available in PMC 2013 March 1.
Published in final edited form as:
PMCID: PMC3294037
NIHMSID: NIHMS352290

Risk Factors for Malignant Melanoma in White and Non-White/Non-African American Populations: The Multiethnic Cohort

Abstract

It is unknown whether the established risk factors for malignant melanoma in whites influence malignant melanoma risk in non-whites. We examined the risk factors for melanoma among 39,325 whites and 101,229 non-whites/multiracials (Japanese American [47.5%], Latino American [34.8%], Native Hawaiian [2.1%] and multiracial [15.6%], excluding African Americans) in the Multiethnic Cohort study. With an average follow-up of 12.7 years, 581 invasive malignant melanoma (IMM) and 412 melanoma in situ (MIS) cases were identified, of which 107 (IMM) and 74 (MIS) were among non-whites/multiracials. The relative risks (RRs) and 95% confidence intervals (CIs) were estimated by Cox proportional hazards models using days from cohort entry as the underlying time variable. Among non-white/multiracial males, location of IMM tumors differed from those of white males (p<0.001); and non-white/multiracial females were more likely to be diagnosed with later stage of disease (p<0.001). After adjusting for potential confounders, age at cohort entry, male sex, higher education, and sunburn susceptibility phenotypes were associated with an increased risk of invasive malignant melanoma in non-whites/multiracials (p<0.05). The risk estimates for age at cohort entry and lighter hair and eye color were greater in non-whites/multiracials than in whites (p-heterogeneity=0.062, 0.016, and 0.005, respectively). For MIS risk, RRs between whites and non-whites/multiracials also differed for study location and education (p-heterogeneity ≤ 0.015). In conclusion, similar to whites, age at cohort entry, male sex, and susceptibility to sunburn phenotypes may be predictive of malignant melanoma risk in non-white populations excluding African-Americans.

Keywords: melanoma risk factors, non-whites

Introduction

Melanoma is one of the few cancers with increasing incidence worldwide (12). Although the incidence and its increase are predominantly found among non-Hispanic whites (3), populations commonly known for having lower incidence of disease, such as Japanese in Japan (45), Puerto Ricans (6), Hispanics in California (7) and Florida (8), and the non-whites in New Zealand (9) have also shown an increase in rates. The exact reasons for this are unknown. This increase has been hypothesized to be a result of the growing accessibility to recreational sun exposure (1011) and UV radiation (UVR) from tanning beds (1213); as well as increasing cancer surveillance and/or reporting (1). Most epidemiologic studies of malignant melanoma have been conducted in whites and they have identified these risk factors: exposure to UVR from the sun or artificial sources, older age, sunburns, and phenotypes that increase the risk of sunburns (e.g. fair skin color) (1415).

The incidence of malignant melanoma among non-whites is relatively low in the U.S. The respective age-adjusted incidence rates (per 100,000) for men and women are 4.0 and 3.9 in Hispanics and 1.6 and 1.3 in Asian/Pacific Islanders, compared to 30.9 and 19.7 in whites (16). Epidemiologic studies of malignant melanoma in non-European or African descendents are scarce, thus, risk factors in non-whites have not been well characterized. This population is often diagnosed at an advanced stage of disease compared to their white counterparts (8, 1719). Therefore, knowing risk factors in non-whites is necessary for melanoma prevention and the reduction of melanoma-related deaths.

We are unaware of any published cohort studies investigating risk factors for malignant melanoma in non-white or multiethnic populations. Using the Multiethnic Cohort Study (MEC) data, we examined whether tumor characteristics and known risk factors for malignant melanoma vary between white and non-white/multiracial (Japanese American, Latino, Native Hawaiian and multiracial, excluding African American) populations.

MATERIALS AND METHODS

Study population

The MEC is a prospective cohort study established to investigate the association of lifestyle and dietary factors with chronic diseases in a multiethnic population. Details of the study design have been previously published (20). The cohort is comprised of 215,251 men and women between the ages of 45 to 75 at recruitment, primarily belonging to one of these racial/ethnic groups: African Americans, Japanese Americans, Latinos, Native Hawaiians, and whites. Potential participants were identified in Hawaii and California (primarily Los Angeles County) through drivers’ license files, voter registration lists, and Health Care Financing Administration files. Between 1993 and 1996, each participant completed a mailed, self-administered questionnaire regarding demographic, dietary, lifestyle, and other exposure factors. The institutional review boards of the University of Hawaii, the University of Southern California, and the University of California, Los Angeles approved this study.

Inclusion and exclusion criteria

In preliminary analyses among non-whites, we found that heterogeneity of risk estimates were greater for some risk factors, such as ever-sunburned status, when including African Americans compared to when excluding African Americans (p=0.05 vs. 0.15). Therefore, to reduce heterogeneity and possible residual confounding within the non-white/multiracial group, African Americans and part-African Americans (n=28,119, with just 7 melanoma cases) were excluded. We also excluded participants who: i) did not belong to one of the five main racial/ethnic groups (n=13,488), ii) had an implausible dietary history (n=8,263), iii) had a prior history of melanoma or were missing non-melanoma skin cancer (NMSC) history (n=2,318), iv) had a prior history of cancer (other than NMSC) before the date of the baseline questionnaire (n=14,632), or vi) had missing data on variables of interest: education, natural hair and eye color, ever-sunburned status, tanning ability, and skin’s reactivity to acute sunlight (n=7,877).

After all exclusions, the eligible population for invasive malignant melanoma (IMM) included 67,521 men and 73,033 women. Although some melanoma in situ (MIS), if left untreated, may evolve into IMM, the risk of developing into an invasive form is unknown. For instance, in lentigo maligna, a type of MIS, the risk of progression has been reported to be as low as 5% and as high as 50% (21). Therefore, to examine the outcome of MIS, an additional 21 males and 26 females who had a previous history of MIS were also excluded from the at-risk eligible population. We performed a sensitivity analysis where we used a common at-risk population including participants without a history of IMM or MIS and found the associations did not differ. Here, malignant melanoma (MM) refers to both IMM and MIS. Excluded participants were slightly older (3.4 years) than those who remained in the analysis. A sensitivity analysis was conducted including those with prior non-melanoma cancers; findings were similar.

Follow-up and case identification

Participants’ follow-up began at the completion of the baseline questionnaire and continued until reaching one of these endpoints: 1) diagnosis of invasive malignant melanoma, 2) death, or 3) end of follow-up, December 31, 2007 and, in the instances where MIS was the outcome of interest, 4) diagnosis of MIS. All incident cases of malignant melanoma were identified by record linkage to the Hawaii Tumor Registry (HTR), the Cancer Surveillance Program (CSP) for Los Angeles County, and the California Cancer Registry (CCR). The Hawaii and California tumor registries participate in the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) program, estimated to have a >99% data completeness (22). Record linkages to the Hawaii and California registries are performed at least annually. The HTR and the CCR cover their respective state populations, while the CSP covers the Los Angeles County area. Linkage was performed by the tumor registries using personal identifiers of name, birthdate, sex, race and social security number. Partial matches were manually checked to identify further cases. Cases of malignant melanoma were defined by International Classification of Diseases for Oncology (ICD-O-3) histology codes 8720–8790 and site code C44.

We used annual linkages to state death certificate files in California and Hawaii, and periodical linkages to the National Death Index in order to identify deaths within the cohort. Cohort participants were followed for an average of 12.7 years, contributing to a total of 1,780,408 person-years. At the end of follow-up, 581 IMM cases and 412 MIS cases were identified among this eligible population of whites and non-whites/multiracials.

Data

Each participant self-reported one or multiple racial/ethnic categories for themselves and for each parent. This study was restricted to those who selected for themselves at least one of the following: non-Hispanic white, Latino, Japanese American, or Native Hawaiian. Here, race/ethnicity was defined by self-report of a single racial/ethnicity (e.g. single-race whites will be referred to as “whites”). Participants with more than one of the four above mentioned racial/ethnic groups were classified as multiracial, of which, the majority categorized themselves as having European-American (70.7%) and/or Native Hawaiian ancestry (65.3%). To account for white admixture in multiracials, which presumably would affect risk for melanoma, we also created a part-white status variable (yes and no). Due to the small number of malignant melanoma cases identified in Japanese American, Latino, Native Hawaiian, and multiracial populations, these groups were combined and labeled as “non-whites/multiracials” (non-whites or multiracials).

Baseline characteristics from the questionnaire were previously published (20). Participants answered questions regarding their natural hair color at the age of 20 (black, medium or dark brown, light brown, blonde, or red) and eye color (brown or black, blue, grey, and green). Over 99.5% of participants provided a single response. Among those with more than one response, traits were classified according to the greatest pigmented phenotype (i.e. “darkest”). Participants were asked about their skin’s tanning ability after repeated sun exposure and unprotected skin’s reactivity to one hour of acute sun exposure. Ever-sunburned status was defined as ever having a blistering sunburn. The age at which this first occurred and total lifetime number of sunburns were also reported.

Due to the high correlations among sunburn susceptibility phenotypes: hair color, eye color, tanning ability (deeply, moderately, lightly, not at all), and skin’s reactivity to acute sunlight (no effect or tans, mild sunburn then tans, severe sunburn, severe sunburn with blistering), we created a sunburn susceptibility phenotype index that categorized the summation of the four above-mentioned propensity to sunburn phenotypes. The categories of hair color, tanning ability, and skin’s reactivity to acute sunlight each had four levels (0 to 3), scaled from lowest to highest risk. We did not detect a gradient of risk within blue, grey, and green eye colors, therefore, the measure of eye color was given two levels: dark=0 and light=1. The sum of all four phenotypes assumes equal weight of each of the four phenotypic factors on an additive scale (0 to 10), where 0 is having dark hair, dark eyes, deeply tans, and has no reaction or tans in reaction to sunlight and 10 is having red hair, light eyes, does not tan at all, and has blistering sunburn reaction to sunlight. Subsequently, this summed value was categorized into a five-category sunburn susceptibility phenotype index (0–1: “lowest risk,” 2–3: “low risk,” 4–5: “medium risk,” 6–7: “high risk”, >7: “highest risk”).

Statistical analyses

Tests for difference in the distribution of stage, histology and tumor location between whites and non-whites/multiracials were conducted using the chi-square test and Fisher’s exact test when expected values for one of the cells was less than 5.

Due to differing etiological factors, acral lentiginous melanoma (ALM) cases were excluded from calculations of relative risks (RR). The other more common histologic subtypes remained in the analysis, since ~58% of cases had not otherwise specified (NOS) histologies. RRs and 95% confidence intervals (CIs) were calculated using Cox proportional hazards models and stratified by race: whites and non-whites/multiracials (Japanese Americans, Latinos, Native Hawaiians, and multiracials). The underlying time variable was days of follow-up, date of cohort entry to date of exit at one of the previously mentioned endpoints. For the IMM analysis, incident MIS cases were followed until death or end of follow-up. For the MIS analysis, incident IMM cases were censored at diagnosis of IMM. All variables of interest met the proportional hazard assumption.

In the minimally-adjusted model, we adjusted for age at cohort entry, sex, race/ethnicity, part-white status, study site and education. For associations between sunburn susceptibility phenotypes and MM, the full regression model included the covariates from the minimally-adjusted model with further adjustment for ever-sunburned status, family history of melanoma and personal history of NMSC. For associations between all other risk factors of interest and MM, a sunburn susceptibility phenotype index (five-levels, explained above) was also included in the full regression model. The results were consistent whether adjusting for all four sunburn susceptible phenotypes (hair and eye color, ability to tan, or reactivity to acute sun exposure), one of the four phenotypes, or the sunburn susceptibility phenotype index. To maximize precision, the index score was used in place of all or one of these phenotypic measures.

Tests for heterogeneity by racial/ethnic groups were evaluated using the likelihood ratio test, which compared the full regression model including the interaction term, product of the race/ethnicity and variable of interest, and a main effects model. Interaction terms were created using the categories described above. The test of heterogeneity between MIS and IMM was performed using competing risk techniques, where each outcome was a different event. A Wald test was used to compare the parameters between outcomes. All statistical analyses were performed in SAS version 9.2 (SAS Institute, Cary, NC).

RESULTS

Whites were more likely than non-whites/multiracials to have: college attendance, phenotypes that increase susceptibility to sunburn (sunburn susceptibility phenotype index >7), blistering sunburns after acute sun exposure, ever-sunburn, a family history of melanoma and a personal history of NMSC (Table 1). Japanese Americans had the greatest homogeneity in hair and eye color (Supplementary Table 1).

Table 1
Percent distribution of baseline characteristics, constitutional factors, and sunburn history among eligible participants of the Multiethnic Cohort Study, 1993– 1996.

The tumor characteristics for IMM and MIS among MEC participants stratified by sex and race/ethnicity are presented in Table 2. Approximately, 42% of cases had specific tumor histology information. Non-white/multiracial females were more likely to be diagnosed with later stage disease (p<0.001). This difference was not found in males. The anatomical location of IMM lesions differed between white and non-white/multiracial males; white males were more likely to have lesions located on the trunk (44.8% vs. 25.8%) and non-white/multiracial males were more likely to have such lesions on the lower limbs, including hips (19.7% vs. 6.3%). In women, the anatomical locations of IMM did not differ by race. However, non-white/multiracial women were more likely to present with MIS on the face (41.2% vs. 21.7%).

Table 2
Invasive malignant melanoma and melanoma in situ tumor characteristics among participants recruited to the Multiethnic Cohort Study, 1993–1996, stratified by sex.

Table 3 presents both the minimally-and fully-adjusted RRs for IMM among whites and non-whites/multiracials. We found that the direction of associations were consistent between the minimally-and fully-adjusted models. In non-whites/multiracials, the association with ever-sunburned status was slightly attenuated where the association was no longer significant in the fully adjusted model (p=0.305). With the exception of age at cohort entry (p-heterogeneity=0.062) and hair and eye color (p-heterogeneity≤0.016), we found that the other IMM risk factors did not differ between whites and non-whites/multiracials (p-heterogeneity≥0.10). Age at cohort entry was suggestive of greater IMM risk among non-white/multiracials and lighter hair and eye colors were associated with greater IMM risk among non-white/multiracials.

Table 3
Relative Risks (RR) for invasive malignant melanoma (IMM) Among whites and non-whites/multiracials recruited to the Multiethnic Cohort Study, 1993–1996

Table 4 presents both the minimally-and fully-adjusted RRs for MIS among whites and non-whites/multiracials. We found that the RRs for MIS between whites and non-whites/multiracials also differed for geographical location and education (p-heterogeneity≤0.015), where both of these factors were not associated with MIS in non-whites/multiracials. The comparison of risk factors between IMM and MIS within whites and within non-whites/multiracials showed no difference in associations (p-heterogeneity>0.10) (data not shown), although the RRs for age at cohort entry and sunburn susceptible phenotypes appear stronger for IMM than MIS.

Table 4
Relative Risks (RR) for melanoma in situ (MIS) Among whites and non-whites/multiracials recruited to the Multiethnic Cohort Study, 1993–1996

DISCUSSION

To our knowledge, this is the first prospective study to investigate tumor characteristics and risk factors for malignant melanoma in non-whites other than African-Americans and to compare these factors between this population and whites.

We found that non-white/multiracial males had a lower proportion of IMM on the anatomical trunk and non-white/multiracial females were more likely to present with later stage disease. Previous studies using cancer registry data have also reported that non-whites were more likely to present with tumors on the lower extremities, of ALM histological subtype and at later stage of disease (1720, 2328). Based on these observations, it has been suggested that UVR exposure may play less of an etiologic factor in melanoma development in non-whites.

To examine this, first, we confirmed the associations for established risk factors (1415, 2932): latitudinal location (Hawaii), higher education, male sex, age, lighter hair and eye color, propensity to sunburn phenotypes, ever-sunburned status, personal history of NMSC and family history of melanoma, with risk of IMM and MIS in our white population. Here, we showed that the association of these risk factors for IMM did not differ between whites and non-whites/multiracials. All factors, with the exception of latitudinal location and higher education, were associated with an increased risk of MIS in non-whites/multiracials. These findings suggest that, similar to whites phenotypes exhibiting greater susceptibility to sunburn are associated with an increased risk of MM in non-white/multiracial populations. Supporting this suggestion, experimental studies have shown that the amount of UVR-related DNA damage did not differ among those with darker skin phototype compared to whites (33). The difference in disease incidence may be partially explained from the observations that among non-whites, the UVR-related DNA damage may have a higher rate of repair (33) and/or may be restricted to the upper layers of skin (34).

Two sun exposure pathways to malignant melanoma have been postulated (3537), acute and chronic sun exposure. The former is characterized by younger age at diagnosis, lesions often located on the trunk for men and lower limbs for females, and superficial spreading histological subtype (35, 3845). Alternatively, tumors that may arise from chronic sun exposure have been associated with history of NMSC, older age, lesions found on the face and neck, and the lentigo maligna melanoma (LMM) histological subtype (35, 3840, 42, 44). In non-whites/multiracials, the distribution of tumor characteristics, associations found with age and no clear association with ever-sunburned status suggests that in this population, chronic sun exposure may be more predictive of MM risk than acute exposure. However, due to the sizable number of unspecified tumor histologies, we were unable to confirm the proportion of LMM cases to support the chronic sun exposure hypothesis. In addition, test for heterogeneity between whites and non-whites/multiracials showed no difference in risk estimates for ever-sunburned status, suggesting that the role of acute sun exposure cannot be ruled out. The sample size may have been too small to detect a modest effect with ever-sunburned status, or heterogeneity of risk estimates between these two racial groups may have been insufficient. Studies with measures of chronic sun exposure, such as degree of solar keratoses (36) or tumor characteristics that have been associated with chronic exposure, such as TP53-positive status (46), would help to clarify the association of sun exposure patterns with malignant melanoma risk in non-whites/multiracials.

In contrast, chronic sun exposure may promote epithelial thickening and protect melanocytes from UVR damage (44, 47): in occupational studies, workers with chronic sun exposure have lower risk of melanoma (37, 48) than those with intermittent exposure (41, 49). Although occupational information exists for our participants, the sample size was too small to evaluate a potential difference in MM incidence among non-whites/multiracials with occupational variations of sun exposure (e.g. office vs. farm work). Further study to investigate the association with cumulated lifetime chronic sun exposure is needed.

The MM risk estimates for sunburn susceptible phenotypic score and hair and eye color differed between non-whites/multiracials and whites, where MM risk was greater among non-whites/multiracials with lighter hair and eye color. This may be due to the differences within this group. For instance, in our study, the risk estimates for hair and eye color likely reflects a comparison of Latinos and multiracials, who have fairer hair and eye colors, to Japanese Americans, who have uniform dark hair and eyes. A study conducted by Wagner JK, et al. found that in Hispanics (n=45) and East Asians (n=15), melanin content, as measured by Adjusted Melanin Index (AMI), was similar between these two populations (p=0.371); however, tanning response, as measured by Melanogenic Dose-Response (MDR), was higher among East Asians than Hispanics (p<0.001) (50), suggesting that East Asians may be a lower risk of disease than Hispanics. In our study, we did not detect heterogeneity of sunburn susceptibility phenotypes within non-whites/multiracials. However, in future studies, quantitative measurements such as these, may improve accuracy and comparability within a multiethnic study population. In addition, adjustment for ancestral markers and/or investigating the association of polymorphisms in MC1R, which has been found to be associated with melanoma risk independently of these constitutional factors (5153), may decrease residual confounding by race and improve risk predictability in populations that present with homogenous susceptibility to sunburn phenotypes. However, these markers were not available for this study.

Studies have found that non-whites are more likely to be diagnosed with later stage disease (1718, 27) and thereby have decreased disease survival. Risk factors for MIS may represent risk for early stage disease, as well as associations with screening behaviors and accessibility to medical care. A case-only study using CCR data found that when accounting for socioeconomic status (SES), as derived from principle component analysis of census block-level data, survival for Asians (p=0.745) and Hispanics (p=0.296) did not differ from that for whites (54). In our study, no clear association was found between education and risk of MIS among non-whites/multiracials, suggesting that either higher education, as a proxy for SES, did not improve early detection of disease or delayed diagnosis may have occurred for other reasons, such as perceived lower susceptibility of disease. Further evaluation of SES, education and screening behaviors would helpful.

In addition, the risk profile of MIS and IMM may be very different; not all MIS progress to IMM (21, 55). MIS is a form of radial growth phase melanomas; therefore, in situ cases have not been identified in melanomas without a radial growth phase, such as nodular melanoma (55). The molecular pathology of these lesions have shown that MIS have lower frequencies of genetic mutations, such as NRAS and BRAF, and it has been suggested that MIS may acquire additional mutations to progress to invasive disease (5657). Our study had the potential to elucidate lifestyle factors that are associated with melanoma screening and risk factors for radial growth phase melanomas among non-whites/multiracials. Although it appeared that the sun susceptibility variables were stronger for IMM than for MIS, when we used the competing risk model, no differences in risk estimates was detected for these two outcomes.

This study had some limitations. We were unable to adjust for length of sun exposure and freckling patterns or moles (58). Information regarding tumor Breslow index was unavailable, which is regrettable since studies have found that Hispanics in California and Pacific peoples in New Zealand may present with greater tumor thickness (7, 9). Residual confounding by constitutional factors may be present. However, the results were similar in models that included a single phenotypic factor, the sunburn susceptible phenotype index, or all phenotypic factors. This suggests that if such confounding was present, it was not captured in the baseline questionnaire data. Due to the limited number of cases in our California and non-white/multiracial populations we were unable to stratify by state of recruitment. With longer follow-up and additional cases, we will have improved power in analysis of non-whites. Hair and eye color, skin’s tanning ability, and skin’s reaction to acute sun exposure may also have differential misclassification depending on perceived self-identity and self-image. The high proportion of melanoma histological subtypes coded as unspecified is a limitation, although rather consistent with what has been reported by other tumor registries (17, 59). Lastly, since we were unable to account for the cancer status of participants who moved out of both Hawaii and California, we censored out-of state migrants at the end of follow-up. This may have introduced bias to our analysis; however, this out-migration has been found to be extremely low (20).

Study strengths include the ability to investigate a range of potential risk factors for malignant melanoma within non-whites/multiracials, the prospective design, and a greater accuracy in measures of race/ethnicity compared to registry data.

In conclusion, this study presents the challenges in identifying risk factors for malignant melanoma in populations with a low incidence of disease. Here, age and susceptibility to sunburn phenotypes were associated with an increased risk of malignant melanoma in non-whites/multiracials. Awareness of these potential risk factors and screening among individuals who are older or have phenotypes that are susceptible to sunburns may help with disease prevention.

Supplementary Material

Acknowledgments

Funding statement: This work was supported by National Cancer Institute grants: R37 CA54281 (to the Multiethnic Cohort study), T32 CA09142 (to S.L.P.) and Career Development Award (K07) CA116543 (to V.W.S).

We are indebted to all the Multiethnic Cohort members for their participation and commitment.

Abbreviations

IMM
invasive malignant melanoma
MIS
melanoma in situ
MM
malignant melanoma
RR
relative risks
CI
confidence intervals
IR
incidence rates
NOS
not otherwise specified
NM
Nodular Melanoma
LMM
Lentigo Maligna Melanoma
SSM
Superficial Spreading Melanoma
ALM
Acral Lentiginous Melanoma
NMSC
non-melanoma skin cancer

References

1. Linos E, Swetter SM, Cockburn MG, Colditz GA, Clarke CA. Increasing Burden of Melanoma in the United States. J Invest Dermatol. 2009;129:1666–74. [PMC free article] [PubMed]
2. Lens MB, Dawes M. Global perspectives of contemporary epidemiological trends of cutaneous malignant melanoma. British Journal of Dermatology. 2004;150:179–85. [PubMed]
3. Fast Stats: An interactive tool for access to SEER cancer statistics. Surveillance Research Program, National Cancer Institute; [cited June 10, 2011]; Available from: http://seer.cancer.gov/faststats.
4. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin. 2005;55:74–108. [PubMed]
5. Ishihara K, Saida T, Otsuka F, Yamazaki N. Statistical profiles of malignant melanoma and other skin cancers in Japan: 2007 update. Int J Clin Oncol. 2008;13:33–41. [PubMed]
6. Valentin SM, Sanchez JL, Figueroa LD, Nazario CM. Epidemiology of melanoma in Puerto Rico, 1987–2002. P R Health Sci J. 2007;26:343–8. [PubMed]
7. Cockburn MG, Zadnick J, Deapen D. Developing epidemic of melanoma in the Hispanic population of California. Cancer. 2006;106:1162–8. [PubMed]
8. Hu S, Parmet Y, Allen G, Parker DF, Ma F, Rouhani P, et al. Disparity in Melanoma: A Trend Analysis of Melanoma Incidence and Stage at Diagnosis Among Whites, Hispanics, and Blacks in Florida. Arch Dermatol. 2009;145:1369–74. [PubMed]
9. Sneyd MJ, Cox B. Melanoma in Maori, Asian, and Pacific peoples in New Zealand. Cancer Epidemiol Biomarkers Prev. 2009;18:1706–13. [PubMed]
10. Bentham G, Aase A. Incidence of Malignant Melanoma of the Skin in Norway, 1955–1989: Associations with Solar Ultraviolet Radiation, Income and Holidays Abroad. Int J Epidemiol. 1996;25:1132–8. [PubMed]
11. Westerdahl J, Olsson H, Ingvar C, Brandt L, Jonsson PE, Moller T. Southern travelling habits with special reference to tumour site in Swedish melanoma patients. Anticancer Res. 1992;12:1539–42. [PubMed]
12. Lazovich D, Vogel RI, Berwick M, Weinstock MA, Anderson KE, Warshaw EM. Indoor Tanning and Risk of Melanoma: A Case-Control Study in a Highly Exposed Population. Cancer Epidemiology Biomarkers & Prevention. 2010;19:1557–68. [PMC free article] [PubMed]
13. El Ghissassi F, Baan R, Straif K, Grosse Y, Secretan B, Bouvard V, et al. A review of human carcinogens--part D: radiation. Lancet Oncol. 2009;10:751–2. [PubMed]
14. Markovic SN, Erickson LA, Rao RD, Weenig RH, Pockaj BA, Bardia A, et al. Malignant melanoma in the 21st century, part 1: epidemiology, risk factors, screening, prevention, and diagnosis. Mayo Clin Proc. 2007;82:364–80. [PubMed]
15. Gruber SB, Armstrong BK. Cutaneous and Ocular Melanoma. In: Schottenfeld D, Fraumeni JF, editors. Cancer Epidemiology and Prevention. New York: Oxford Press; 2006. pp. 1196–217.
16. Howlader N, Noone A, Krapcho M, Neyman N, Aminou RWW, et al. SEER Cancer Statistics Review, 1975–2008. 2011 [cited June 10, 2011]; Available from: http://seer.cancer.gov/csr/1975_2008/, based on November 2010 SEER data submission, posted to the SEER web site.
17. Cress RD, Holly EA. Incidence of cutaneous melanoma among non-Hispanic whites, Hispanics, Asians, and blacks: an analysis of california cancer registry data, 1988–93. Cancer Causes Control. 1997;8:246–52. [PubMed]
18. Cormier JN, Xing Y, Ding M, Lee JE, Mansfield PF, Gershenwald JE, et al. Ethnic differences among patients with cutaneous melanoma. Arch Intern Med. 2006;166:1907–14. [PubMed]
19. Hemmings DE, Johnson DS, Tominaga GT, Wong JH. Cutaneous melanoma in a multiethnic population: is this a different disease? Arch Surg. 2004;139:968–72. [PubMed]
20. Kolonel LN, Henderson BE, Hankin JH, Nomura AM, Wilkens LR, Pike MC, et al. A multiethnic cohort in Hawaii and Los Angeles: baseline characteristics. Am J Epidemiol. 2000;151:346–57. [PubMed]
21. Cohen LM. Lentigo maligna and lentigo maligna melanoma. Journal of the American Academy of Dermatology. 1995;33:923–36. [PubMed]
22. SEER: Surveillance, Epidemiology, and End Results. 2011 [cited; Available from: http://seer.cancer.gov/
23. Sng J, Koh D, Siong WC, Choo TB. Skin cancer trends among Asians living in Singapore from 1968 to 2006. J Am Acad Dermatol. 2009;61:426–32. [PubMed]
24. Bergfelt L, Newell GR, Sider JG, Kripke ML. Incidence and anatomic distribution of cutaneous melanoma among United States Hispanics. J Surg Oncol. 1989;40:222–6. [PubMed]
25. Black WC, Goldhahn RT, Jr, Wiggins C. Melanoma Within a Southwestern Hispanic Population. Arch Dermatol. 1987;123:1331–4. [PubMed]
26. Shoo BA, Kashani-Sabet M. Melanoma arising in African-, Asian-, Latino- and Native-American populations. Semin Cutan Med Surg. 2009;28:96–102. [PubMed]
27. Chang AE, Karnell LH, Menck HR. The National Cancer Data Base report on cutaneous and noncutaneous melanoma: a summary of 84,836 cases from the past decade. The American College of Surgeons Commission on Cancer and the American Cancer Society Cancer. 1998;83:1664–78. [PubMed]
28. Shin S, Palis BE, Phillips JL, Stewart AK, Perry RR. Cutaneous melanoma in Asian-Americans. Journal of Surgical Oncology. 2009;99:114–8. [PubMed]
29. Green A, Whiteman D, Frost C, Battistutta D. Sun exposure, skin cancers and related skin conditions. J Epidemiol. 1999;9:S7–13. [PubMed]
30. Naldi L, Altieri A, Imberti GL, Giordano L, Gallus S, La Vecchia C. Cutaneous malignant melanoma in women. Phenotypic characteristics, sun exposure, and hormonal factors: a case-control study from Italy. Ann Epidemiol. 2005;15:545–50. [PubMed]
31. Veierod MB, Weiderpass E, Thorn M, Hansson J, Lund E, Armstrong B, et al. A prospective study of pigmentation, sun exposure, and risk of cutaneous malignant melanoma in women. J Natl Cancer Inst. 2003;95:1530–8. [PubMed]
32. Gandini S, Sera F, Cattaruzza MS, Pasquini P, Zanetti R, Masini C, et al. Meta-analysis of risk factors for cutaneous melanoma: III. Family history, actinic damage and phenotypic factors. Eur J Cancer. 2005;41:2040–59. [PubMed]
33. Sheehan JM, Cragg N, Chadwick CA, Potten CS, Young AR. Repeated Ultraviolet Exposure Affords the Same Protection Against DNA Photodamage and Erythema in Human Skin Types II and IV but is Associated with Faster DNA Repair in Skin Type IV. 2002;118:825–9. [PubMed]
34. Del Bino S, Sok J, Bessac E, Bernerd F. Relationship between skin response to ultraviolet exposure and skin color type. Pigment Cell Research. 2006;19:606–14. [PubMed]
35. Green A. A theory of site distribution of melanomas: Queensland, Australia. Cancer Causes Control. 1992;3:513–6. [PubMed]
36. Whiteman D, Watt P, Purdie D, Hughes M, Hayward N, Green A. Melanocytic nevi, solar keratoses, and divergent pathways to cutaneous melanoma. J Natl Cancer Inst. 2003;95:806–12. [PubMed]
37. Rivers JK. Is there more than one road to melanoma? The Lancet. 2004;363:728–30. [PubMed]
38. Elwood JM, Gallagher RP. Body site distribution of cutaneous malignant melanoma in relationship to patterns of sun exposure. International Journal of Cancer. 1998;78:276–80. [PubMed]
39. Whiteman D, Stickley M, Watt P, Hughes M, Davis M, Green A. Anatomic site, sun exposure, and risk of cutaneous melanoma. J Clin Oncol. 2006;24:3172–7. [PubMed]
40. Holman CD, Armstrong BK, Heenan PJ. Relationship of cutaneous malignant melanoma to individual sunlight-exposure habits. J Natl Cancer Inst. 1986;76:403–14. [PubMed]
41. Nelemans PJ, Groenendal H, Kiemeney LA, Rampen FH, Ruiter DJ, Verbeek AL. Effect of intermittent exposure to sunlight on melanoma risk among indoor workers and sun-sensitive individuals. Environ Health Perspect. 1993;101:252–5. [PMC free article] [PubMed]
42. Armstrong B. Epidemiology of malignant melanoma: intermittent or total accumulated exposure to the sun? J Dermatol Surg Oncol. 1988;14:835–49. [PubMed]
43. Dennis LK, Vanbeek MJ, Beane Freeman LE, Smith BJ, Dawson DV, Coughlin JA. Sunburns and Risk of Cutaneous Melanoma: Does Age Matter? A Comprehensive Meta-Analysis Annals of Epidemiology. 2008;18:614–27. [PMC free article] [PubMed]
44. Elwood JM, Gallagher RP, Hill GB, Pearson JC. Cutaneous melanoma in relation to intermittent and constant sun exposure--the Western Canada Melanoma Study. Int J Cancer. 1985;35:427–33. [PubMed]
45. Elwood J, Jopson J. Melanoma and sun exposure: an overview of published studies. Int J Cancer. 1997;73:198–203. [PubMed]
46. Whiteman D, Parsons P, Green A. p53 expression and risk factors for cutaneous melanoma: a case-control study. Int J Cancer. 1998;77:843–8. [PubMed]
47. Gandini S, Sera F, Cattaruzza MS, Pasquini P, Picconi O, Boyle P, et al. Meta-analysis of risk factors for cutaneous melanoma: II. Sun exposure Eur J Cancer. 2005;41:45–60. [PubMed]
48. Walter S, King W, Marrett L. Association of cutaneous malignant melanoma with intermittent exposure to ultraviolet radiation: results of a case-control study in Ontario, Canada. Int J Epidemiol. 1999;28:418–27. [PubMed]
49. Pukkala E, Aspholm R, Auvinen A, Eliasch H, Gundestrup M, Haldorsen T, et al. Incidence of cancer among Nordic airline pilots over five decades: occupational cohort study. BMJ. 2002;325:567–71. [PMC free article] [PubMed]
50. Wagner JK, Parra EJ, Norton HL, Jovel C, Shriver MD. Skin Responses to Ultraviolet Radiation: Effects of Constitutive Pigmentation, Sex, and Ancestry. Pigment Cell Research. 2002;15:385–90. [PubMed]
51. Kanetsky PA, Panossian S, Elder DE, Guerry D, Ming ME, Schuchter L, et al. Does MC1R genotype convey information about melanoma risk beyond risk phenotypes? Cancer. 2010;116:2416–28. [PMC free article] [PubMed]
52. Dwyer T, Stankovich JM, Blizzard L, FitzGerald LM, Dickinson JL, Reilly A, et al. Does the addition of information on genotype improve prediction of the risk of melanoma and nonmelanoma skin cancer beyond that obtained from skin phenotype? Am J Epidemiol. 2004;159:826–33. [PubMed]
53. Palmer JS, Duffy DL, Box NF, Aitken JF, O’Gorman LE, Green AC, et al. Melanocortin-1 receptor polymorphisms and risk of melanoma: is the association explained solely by pigmentation phenotype? Am J Hum Genet. 2000;66:176–86. [PubMed]
54. Zell JA, Cinar P, Mobasher M, Ziogas A, Meyskens FL, Jr, Anton-Culver H. Survival for Patients With Invasive Cutaneous Melanoma Among Ethnic Groups: The Effects of Socioeconomic Status and Treatment. J Clin Oncol. 2008;26:66–75. [PubMed]
55. Liu V, Mihm MC. Pathology of malignant melanoma. Surg Clin North Am. 2003;83:31–60. [PubMed]
56. Clark WH, Jr, Elder DE, Guerry Dt, Epstein MN, Greene MH, Van Horn M. A study of tumor progression: the precursor lesions of superficial spreading and nodular melanoma. Hum Pathol. 1984;15:1147–65. [PubMed]
57. Greene VR, Johnson MM, Grimm EA, Ellerhorst JA. Frequencies of NRAS and BRAF Mutations Increase from the Radial to the Vertical Growth Phase in Cutaneous Melanoma. J Invest Dermatol. 2009;129:1483–8. [PMC free article] [PubMed]
58. Gandini S, Sera F, Cattaruzza MS, Pasquini P, Abeni D, Boyle P, et al. Meta-analysis of risk factors for cutaneous melanoma: I. Common and atypical naevi. Eur J Cancer. 2005;41:28–44. [PubMed]
59. Pollitt RA, Clarke CA, Swetter SM, Peng DH, Zadnick J, Cockburn M. The expanding melanoma burden in California hispanics. Cancer. 2011;117:152–61. [PMC free article] [PubMed]