This cross-sectional study examines the association between sexual orientation and self-reported lifetime prevalence of skin cancer in the United States among individuals who have self-identified as being heterosexual or of a sexual minority population.
Key Points
Question
What is the association between sexual orientation and lifetime prevalence of skin cancer in the United States?
Findings
In this cross-sectional study of 845 264 adults, both gay and bisexual men had higher adjusted odds of lifetime prevalence of skin cancer compared with heterosexual men. Bisexual women, but not lesbian women, had lower odds of lifetime prevalence of skin cancer compared with heterosexual women.
Meaning
Patient education and community outreach programs targeting these populations may be helpful in reducing disparities in lifetime skin cancer prevalence.
Abstract
Importance
Sexual minority men have reported higher rates of both indoor tanning and skin cancer than heterosexual men, and sexual minority women have reported lower or equal rates of both indoor tanning and skin cancer compared with heterosexual women. Bisexual men, in particular, have reported higher rates of indoor tanning bed use than heterosexual men; however, no study has investigated skin cancer prevalence among gay, lesbian, and bisexual individuals as separate groups.
Objective
To evaluate the association between sexual orientation and lifetime prevalence of skin cancer.
Design, Setting, and Participants
This cross-sectional study analyzed data from the 2014-2018 Behavioral Risk Factor Surveillance System (BRFSS) surveys of a noninstitutionalized population in the United States that included 845 264 adult participants who self-identified as being heterosexual, gay, lesbian, or bisexual.
Main Outcomes and Measures
Self-reported lifetime history of skin cancer.
Results
The study included 845 264 participants, including 351 468 heterosexual men (mean age, 47.7; 95% CI, 47.5-47.8), 7516 gay men (mean age, 42.7; 95% CI, 41.9-43.5), 5088 bisexual men (mean age, 39.3; 95% CI, 38.2-40.4), 466 355 heterosexual women (mean age, 49.7; 95% CI, 49.6-49.9), 5392 lesbian women (mean age, 41.9; 95% CI, 40.7-43.2), and 9445 bisexual women (mean age, 32.7; 95% CI, 32.2-33.2). The adjusted odds ratios (AORs) of skin cancer prevalence were significantly higher among both gay (AOR, 1.26; 95% CI, 1.05-1.51; P = .01) and bisexual men (AOR, 1.48; 95% CI, 1.02-2.16; P = .04) compared with heterosexual men. The AORs of skin cancer were statistically significantly lower among bisexual women (AOR, 0.78; 95% CI, 0.61-0.99; P = .04) but not among gay or lesbian women (AOR, 0.97; 95% CI, 0.73-1.27; P = .81) compared with the AORs of skin cancer among heterosexual women.
Conclusions and Relevance
In this study, gay and bisexual men had an increased self-reported lifetime prevalence of skin cancer compared with the prevalence among heterosexual men. Patient education and community outreach initiatives focused on reducing skin cancer risk behaviors among gay and bisexual men may help reduce the lifetime development of skin cancer in this population. Continued implementation of the Behavioral Risk Factor Surveillance System’s sexual orientation and gender identity module is imperative to improve understanding of the health and well-being of sexual minority populations.
Introduction
Sexual minority groups include, but are not limited to, individuals who identify as gay, lesbian, or bisexual. Sexual minority men (SMM) have reported higher rates of both indoor tanning and skin cancer compared with the rate among heterosexual men, whereas sexual minority women (SMW) have reported lower or equal rates of both indoor tanning and skin cancer compared with the rate among heterosexual women.1,2,3 Bisexual men have reported more frequent use of indoor tanning beds compared with use among heterosexual men4; however, no study has investigated skin cancer prevalence among gay, lesbian, and bisexual individuals as separate groups. In this study, we evaluated whether lifetime prevalence of skin cancer differs among gay, lesbian, and bisexual individuals compared with heterosexual individuals.
Methods
Study Design
This cross-sectional study analyzed data from the 2014-2018 Behavioral Risk Factor Surveillance System (BRFSS) annual questionnaires. The BRFSS is a national system of telephone surveys that collects demographic and health-related data on noninstitutionalized US adults. In 2014, the BRFSS introduced a sexual orientation and gender identity (SOGI) module, which was administered at least once in 37 states from 2014 to 2018.5 This study was deemed exempt from review by Partners Healthcare institutional review board because it used publicly available data without patient identifiers. All survey respondents agreed to have their data publicly released.
Study Population, Covariates, and Outcome Variable
The study population was stratified by self-reported sex as male or female. Self-reported sexual orientation was ascertained in the SOGI module. The analysis included respondents who identified as gay (for men), gay or lesbian (for women), bisexual, or heterosexual. The demographic characteristics of the study population are given in Table 1.
Table 1. Characteristics of Heterosexual, Gay or Lesbian, and Bisexual Respondents, Stratified by Sex.
| Characteristic | Men | P Valuea | Women | P Valuea | ||||
|---|---|---|---|---|---|---|---|---|
| Heterosexual | Gay | Bisexual | Heterosexual | Lesbian or Gay | Bisexual | |||
| Unweighted, No.b | 351 468 | 7516 | 5088 | NA | 466 355 | 5392 | 9445 | NA |
| Weighted, No. | 72 334 280 | 1 624 508 | 1 204 967 | NA | 78 191 198 | 1 126 745 | 2 364 343 | NA |
| Age, mean (95% CI), y | 47.7 (47.5-47.8) | 42.7 (41.9-43.5) | 39.3 (38.2-40.4) | NA | 49.7 (49.6-49.9) | 41.9 (40.7-43.2) | 32.7 (32.2-33.2) | <.001 |
| Region, % (95% CI) | ||||||||
| Northeast | 16.7 (16.5-16.8) | 19.7 (18.3-21.2) | 21.6 (19.6-23.5) | <.001 | 17.0 (16.8-17.1) | 17.8 (16.0-19.6) | 19.6 (18.3-20.9) | .01 |
| South | 39.1 (38.8-39.3) | 35.9 (33.4-38.3) | 36.5 (33.6-39.4) | 40 (39.8-40.2) | 38.1 (34.7-41.5) | 37.8 (35.7-40.0) | ||
| Midwest | 21.4 (21.3-21.5) | 16.7 (15.4-18.0) | 20.4 (18.6-22.2) | 21.2 (21.0-21.3) | 18.9 (17.0-20.8) | 21.3 (19.9-22.8) | ||
| West | 22.9 (22.6-23.1) | 27.7 (25.1-30.3) | 21.6 (18.6-24.6) | 21.9 (21.7-22.2) | 25.2 (21.4-28.9) | 21.2 (19.2-23.3) | ||
| Race/ethnicity, % (95% CI) | ||||||||
| Non-Hispanic White | 65.6 (65.2-65.9) | 63.9 (61.2-66.6) | 60.0 (57.0-63.0) | .01 | 65.4 (65.0-65.8) | 62.3 (58.6-66.0) | 62.6 (60.4-64.7) | .02 |
| Non-Hispanic Black | 10.9 (10.7-11.2) | 10.3 (8.5-12.1) | 11.2 (9.5-12.9) | 12.4 (12.2-12.6) | 15.3 (12.7-17.8) | 13.8 (12.0-15.6) | ||
| Hispanic | 8.0 (7.7-8.2) | 9.3 (7.2-11.3) | 10.7 (9.0-12.5) | 7.4 (7.2-7.7) | 9.2 (6.1-12.3) | 9.0 (7.9-10.2) | ||
| Otherc | 15.5 (15.2-15.9) | 16.5 (14.4-18.6) | 18 (15.1-21.0) | 14.8 (14.5-15.1) | 13.3 (10.6-15.9) | 14.6 (13.0-16.2) | ||
| Education level, % (95% CI) | ||||||||
| High school or less | 43.3 (42.9-43.7) | 29.6 (27.0-32.2) | 42.4 (39.5-45.3) | <.001 | 38.9 (38.6-39.3) | 34.8 (31.3-38.2) | 40.7 (38.5-42.8) | .02 |
| Some college or more | 56.7 (56.3-57.1) | 70.4 (67.8-73.0) | 57.6 (54.7-60.5) | 61.1 (60.7-61.4) | 65.2 (61.8-68.7) | 59.3 (57.2-61.5) | ||
| Employment status, % (95% CI) | ||||||||
| Unemployed | 35.4 (35.0-35.7) | 35.4 (33.0-37.9) | 42.7 (39.7-45.6) | <.001 | 50.3 (50.0-50.7) | 40.5 (36.9-44.0) | 43.2 (41.1-45.3) | <.001 |
| Employed | 64.6 (64.3-65.0) | 64.6 (62.1-67.0) | 57.3 (54.4-60.3) | 49.7 (49.3-50.0) | 59.5 (56.0-63.1) | 56.8 (54.7-58.9) | ||
| Insurance status, % (95% CI) | ||||||||
| No | 12.4 (12.1-12.6) | 11.7 (10.0-13.5) | 16.0 (13.7-18.4) | <.01 | 9.6 (9.4-9.8) | 11.9 (9.5-14.3) | 13.9 (12.4-15.3) | <.001 |
| Yes | 87.6 (87.4-87.9) | 88.3 (86.5-90.0) | 84.0 (81.6-86.3) | 90.4 (90.2-90.6) | 88.1 (85.7-90.5) | 86.1 (84.7-87.6) | ||
| Smoking history, % (95% CI) | ||||||||
| Never smoker | 53.1 (52.7-53.4) | 52.9 (50.4-55.4) | 54.4 (51.5-57.3) | <.001 | 64.2 (63.9-64.5) | 50.2 (46.8-53.7) | 54.6 (52.5-56.7) | <.001 |
| Former smoker | 29.1 (28.8-29.4) | 22.2 (20.3-24.1) | 22.5 (20.2-24.9) | 22.0 (21.7-22.3) | 24.6 (21.8-27.4) | 18.4 (16.9-20.0) | ||
| Current smoker | 17.8 (17.6-18.1) | 24.9 (22.6-27.1) | 23.1 (20.6-25.5) | 13.8 (13.6-14.1) | 25.1 (22.1-28.2) | 27.0 (25.1-28.8) | ||
| Alcohol use, % (95% CI)d | ||||||||
| Nondrinker | 40.8 (40.4-41.2) | 31.5 (29.2-33.7) | 41.2 (38.3-44.2) | <.001 | 52.9 (52.5-53.2) | 40.8 (37.4-44.2) | 40.0 (37.9-42.1) | <.001 |
| Light drinker | 43.3 (42.9-43.6) | 50.7 (48.2-53.2) | 42.0 (39.1-44.8) | 41.8 (41.4-42.1) | 47.8 (44.3-51.2) | 47.4 (45.3-49.6) | ||
| Moderate or heavy drinker | 15.9 (15.7-16.2) | 17.8 (16.0-19.6) | 16.8 (14.6-19.0) | 5.4 (5.2-5.5) | 11.5 (9.6-13.4) | 12.6 (10.9-14.2) | ||
| Other cancer diagnosis, % (95% CI) | ||||||||
| No | 94.3 (94.1-94.4) | 94.6 (93.7-95.6) | 95.7 (94.5-97.0) | .87 | 91.4 (91.3-91.6) | 92.1 (89.5-94.8) | 94.8 (94.0-95.6) | <.001 |
| Yes | 5.7 (5.6-5.9) | 5.4 (4.4-6.3) | 4.3 (3.0-5.5) | 8.6 (8.4-8.7) | 7.9 (5.2-10.5) | 5.2 (4.4-6.0) | ||
Abbreviation: NA, not applicable.
Analysis of variance used to assess the P value for age, χ2 tests used to assess remaining P values.
Unweighted sample excluded 121 714 (12.2%) individuals with missing data for sexual orientation, sex, and skin cancer. Those with missing data in any covariate were also excluded (32 676 [3.3%]).
Respondents who self-identified as American Indian or Alaska Native only, non-Hispanic; Asian only, non-Hispanic; Native Hawaiian or other Pacific Islander only, non-Hispanic; other race only, non-Hispanic; or multiracial, non-Hispanic.
Nondrinkers, 0 alcoholic beverages weekly; light drinkers, 1 to 7 alcoholic drinks weekly; moderate or heavy drinkers, more than 7 alcoholic beverages per weekly.
Covariates examined were sociodemographic and health care access variables in the BRFSS questionnaire that are either basic demographic information or factors associated with skin cancer.6 These included age, geographic region, race/ethnicity, education level, employment status, insurance status, smoking history, current alcohol consumption, and history of another cancer diagnosis.
History of skin cancer diagnosis was assessed by the following question: “Has a doctor, nurse, or other health professional ever told you had skin cancer?” Respondents who answered “don’t know,” “not sure,” or refused to answer any questions pertaining to sex, sexual orientation, or lifetime skin cancer diagnosis were excluded from the analysis.
Statistical Analysis
All analyses stratified participants by sex and sexual orientation, and heterosexual respondents of each sex were used as the reference group. Data were weighted according to BRFSS recommendations. Analysis of variance tests were used to compare age, and χ2 analyses were used to compare demographic characteristics between groups. The significance threshold was P < .05 and testing was 2-sided.
Univariate regression analyses were conducted to investigate associations between sexual orientation (Table 2) and other covariates (eTables 1 and 2 in the Supplement) and lifetime skin cancer prevalence. The direct method for age standardization was used to calculate age-adjusted lifetime skin cancer prevalence. Multivariate logistic regression analyses were conducted to calculate adjusted odds ratios (AORs) for lifetime skin cancer prevalence. No adjustment was made for multiple comparisons. All analyses were performed using R software, version 3.5.1 (R Core Team).
Table 2. Age-Adjusted Lifetime Prevalence, Univariate Odds Ratios, and Adjusted Odds Ratios of Skin Cancer, by Sexual Orientation and Sex.
| Variable | Age-Adjusted Lifetime Prevalence, % (95% CI)a | Univariate OR (95% CI) | P Value | AOR (95% CI)b | P Value |
|---|---|---|---|---|---|
| Men | |||||
| Heterosexual | 6.7 (6.5-6.9) | 1 [Reference] | NA | 1 [Reference] | NA |
| Gay | 8.1 (6.8-9.5) | 0.85 (0.72-1.01) | .06 | 1.26 (1.05-1.51) | .01 |
| Bisexual | 8.4 (6.3-11.0) | 0.81 (0.59-1.11) | .18 | 1.48 (1.02-2.16) | .04 |
| Women | |||||
| Heterosexual | 6.6 (6.5-6.8) | 1 [Reference] | NA | 1 [Reference] | NA |
| Lesbian or gay | 5.9 (4.8-7.3) | 0.66 (0.51-0.84) | <.001 | 0.97 (0.73-1.27) | .81 |
| Bisexual | 4.7 (3.8-5.7) | 0.29 (0.23-0.36) | <.001 | 0.78 (0.61-0.99) | .04 |
Abbreviations: AOR, adjusted odds ratio; NA, not applicable.
Age-adjusted prevalence was calculated using direct standardization with heterosexual respondents from the weighted sample, by sex, as the standard population group.
Adjusted for age, geographic region, race/ethnicity, education level, employment status, insurance status, current alcohol consumption, smoking history, and history of another cancer diagnosis.
Results
The total unweighted study sample of 845 264 individuals included 351 468 heterosexual men (mean age, 47.7 years; 95% CI, 47.5-47.8 years), 7516 gay men (mean age, 42.7 years; 95% CI, 41.9-43.5 years), 5088 bisexual men (mean age, 39.3 years; 95% CI, 38.2-40.4 years), 466 355 heterosexual women (mean age, 49.7 years; 95% CI, 49.6-49.9 years), 5392 lesbian women (mean age, 41.9 years; 95% CI, 40.7-43.2 years), and 9445 bisexual women (mean age, 32.7 years; 95% CI, 32.2-33.2 years) (Table 1).
Age-adjusted lifetime prevalence of skin cancer diagnosis was 8.1% (95% CI, 6.8%-9.5%) among gay men, 8.4% (95% CI, 6.3%-11.0%) among bisexual men, and 6.7% (95% CI, 6.5%-6.9%) among heterosexual men (Table 2). The adjusted odds ratio (AOR) of skin cancer diagnosis was statistically significantly higher among gay men (AOR, 1.26; 95% CI, 1.05-1.51, P = .01) and bisexual men (AOR, 1.48; 95% CI, 1.02-2.16, P = .04) than among heterosexual men.
Age-adjusted lifetime prevalence of skin cancer diagnosis was 5.9% (95% CI, 4.8%-7.3%) among lesbians, 4.7% (95% CI, 3.8%-5.7%) among bisexual women, and 6.6% (95% CI, 6.5%-6.8%) among heterosexual women. The AOR of skin cancer diagnosis was lower among bisexual women (AOR, 0.78; 95% CI, 0.61-0.99; P = .04) but not among gay or lesbian women (AOR, 0.97; 95% CI, 0.73-1.27, P = .81) compared with that among heterosexual women (Table 2).
Discussion
Gay men and bisexual men were more likely than heterosexual men to have reported a skin cancer diagnosis. Bisexual women were less likely than heterosexual women to have reported a skin cancer diagnosis. Although prior studies showed an increased lifetime prevalence of skin cancer among SMM compared with heterosexual men, that increased prevalence persisted in this study when examining the data on gay and bisexual men separately.
Increased lifetime prevalence of skin cancer among gay and bisexual men likely reflects at least in part the increased indoor UV exposure among both gay and bisexual populations.4,7 The primary motivators for indoor tanning among SMM have been shown to be concerns about appearance and community pressures,7,8 and a recent study showed that indoor tanning salons are more likely to be located near neighborhoods with higher concentrations of male-male partnered households.9 Consistent with US Preventive Services Task Force recommendations for the general population, health care professionals should counsel SMM patients against using indoor tanning beds10 or to consider sunless tanning, which has proven effective in reducing indoor tanning bed use among women.11 Patient education and community outreach initiatives centered on the increased rate of skin cancer diagnosis among SMM may reduce skin cancer risk behaviors, as fear of developing skin cancer was identified as a primary motivation for stopping indoor tanning among SMM.8
Findings in the present study are also consistent with prior findings that SMW have a decreased prevalence of skin cancer and skin cancer risk behaviors compared with heterosexual women, although in this study prevalence was low among only bisexual women.1 Although it is possible that this indicates a decreased prevalence of skin cancer or skin cancer risk behaviors among bisexual women, it could also reflect evidence that bisexual women, for cost-related reasons, are less likely to seek medical care.12
Limitations
This study has limitations, and it must be interpreted in the context of the study design. First, BRFSS data are based on self-reported, unvalidated skin cancer diagnoses, which may be inaccurate. Second, BRFSS did not collect information on important potential confounders, including UV exposure, photoprotective behaviors, Fitzpatrick skin type, or HIV and immunosuppression status, which may vary by sexual orientation. Third, because the BRFSS study did not collect information on health care use, the results may be affected by surveillance bias, as same-sex couples are less likely than opposite-sex couples to undergo total body skin examination.13 In addition, because the BRFSS SOGI module has been implemented in only 37 states, the study population may not be fully generalizable to the entire United States.
The Centers for Disease Control and Prevention (CDC) recently considered stopping implementation of the SOGI module for future BRFSS surveys,14 which would preclude further data collection on this vulnerable population through the BRFSS. The data in this report and our corresponding article15 reflect the current sum of national data collected by the CDC on this population, and we strongly advocate for both continued implementation of the BRFSS SOGI module by the CDC and for more states to implement this module annually.
Conclusions
To our knowledge, this was the largest study to examine lifetime prevalence of skin cancer among sexual minorities and the first to examine gay or lesbian individuals and bisexual individuals separately. Patient education and community outreach initiatives focused on reducing skin cancer risk behaviors among SMM may help reduce lifetime development of skin cancer in this population. Future advocacy efforts should focus on the continued implementation of the BRFSS SOGI module to improve understanding of the health and well-being of sexual minority populations.
eTable 1. Univariate Odds Ratios for Lifetime Diagnosis of Skin Cancer Among Men
eTable 2. Univariate Odds Ratios for Lifetime Diagnosis of Skin Cancer Among Women
References
- 1.Mansh M, Katz KA, Linos E, Chren M-M, Arron S. Association of skin cancer and indoor tanning in sexual minority men and women. JAMA Dermatol. 2015;151(12):1308-1316. doi: 10.1001/jamadermatol.2015.3126 [DOI] [PubMed] [Google Scholar]
- 2.Gao Y, Arron ST, Linos E, Polcari I, Mansh MD. Indoor tanning, sunless tanning, and sun-protection behaviors among sexual minority men. JAMA Dermatol. 2018;154(4):477-479. doi: 10.1001/jamadermatol.2018.0003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Yeung H, Baranowski MLH, Chen SC. Skin cancer risk factors and screening among sexual minority and heterosexual women. J Am Acad Dermatol. 2019;81(4):1015-1018. doi: 10.1016/j.jaad.2019.02.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Yeung H, Chen SC. Sexual orientation and indoor tanning device use: a population-based study. JAMA Dermatol. 2016;152(1):99-101. doi: 10.1001/jamadermatol.2015.2038 [DOI] [PubMed] [Google Scholar]
- 5.CDC - BRFSS Annual Survey Data https://www.cdc.gov/brfss/annual_data/annual_data.htm. Published August 29, 2019. Accessed November 9, 2019.
- 6.Diepgen TL, Mahler V. The epidemiology of skin cancer. Br J Dermatol. 2002;146(suppl 61):1-6. doi: 10.1046/j.1365-2133.146.s61.2.x [DOI] [PubMed] [Google Scholar]
- 7.Blashill AJ, Rooney BM, Wells KJ. An integrated model of skin cancer risk in sexual minority males. J Behav Med. 2018;41(1):99-108. doi: 10.1007/s10865-017-9879-2 [DOI] [PubMed] [Google Scholar]
- 8.Admassu N, Pimentel MA, Halley MC, et al. . Motivations among sexual-minority men for starting and stopping indoor tanning. Br J Dermatol. 2019;180(6):1529-1530. doi: 10.1111/bjd.17684 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Chen R, Hipp JA, Morrison L, Henriksen L, Swetter SM, Linos E. Association of number of indoor tanning salons with neighborhoods with higher concentrations of male-male partnered households. JAMA Netw Open. 2019;2(10):e1912443-e1912443. doi: 10.1001/jamanetworkopen.2019.12443 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.US Preventive Services Task Force Final recommendation statement: skin cancer prevention: behavioral counseling. https://www.uspreventiveservicestaskforce.org/Page/Document/RecommendationStatementFinal/skin-cancer-counseling2. Accessed December 19, 2019.
- 11.Pagoto SL, Schneider KL, Oleski J, Bodenlos JS, Ma Y. The sunless study: a beach randomized trial of a skin cancer prevention intervention promoting sunless tanning. Arch Dermatol. 2010;146(9):979-984. doi: 10.1001/archdermatol.2010.203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Blosnich JR, Farmer GW, Lee JG, Silenzio VM, Bowen DJ. Health inequalities among sexual minority adults: evidence from ten U.S. states, 2010. Am J Prev Med. 2014;46(4):337-349. doi: 10.1016/j.amepre.2013.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Heckman CJ, Darlow S, Manne SL, Kashy DA, Munshi T. Correspondence and correlates of couples’ skin cancer screening. JAMA Dermatol. 2013;149(7):825-830. doi: 10.1001/jamadermatol.2013.515 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.CDC to roll back data collection on the health and well-being of LGBT people. The Williams Institute. https://williamsinstitute.law.ucla.edu/press/cdc-data-rollback-brfss/. Published May 17, 2018. Accessed November 15, 2019.
- 15.Singer S, Tkachenko E, Hartman RI, Mostaghimi A. Gender identity and lifetime prevalence of skin cancer in the United States [published online February 5, 2020]. JAMA Dermatol. doi: 10.1001/jamadermatol.2019.4197 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Univariate Odds Ratios for Lifetime Diagnosis of Skin Cancer Among Men
eTable 2. Univariate Odds Ratios for Lifetime Diagnosis of Skin Cancer Among Women