|Home | About | Journals | Submit | Contact Us | Français|
Despite a rise in incidence of basal cell carcinoma (BCC) among young people and the ubiquity of indoor tanning in this population, few epidemiologic studies have investigated this exposure-disease relationship.
Evaluate the association between indoor tanning and early-onset BCC.
BCC cases (n=376) and controls with minor benign skin conditions (n=390) under age 40 were identified through Yale Dermatopathology. Participants provided information on ever indoor tanning, age of initiation, frequency, duration, burns while tanning, and type of tanning device during an in-person interview. We calculated odds ratios (OR) and 95% confidence intervals (CI) using multivariate logistic regression with never indoor tanners as the referent group.
Ever indoor tanning was associated with a 69% increased risk of early-onset BCC (95% CI=1.15-2.48). This association was stronger among women (OR=2.14, 95% CI=1.31-3.47), for multiple BCCs (OR=2.16, 95% CI=1.26-3.70), and for BCCs on the trunk and extremities (OR=2.81, 95% CI=1.57-5.02). Risk increased dose-dependently with years used regular indoor tanning devices (p-trend=0.003), number of overall burns (p-trend=<0.001) and burns to biopsy site (p-trend=<0.001) from indoor tanning. Approximately one-quarter (27%) of early-onset BCCs (or 43% among women) could be prevented if individuals never tanned indoors.
Potential recall bias of indoor tanning by cases and generalizability of the control population suggest replication in other studies is warranted.
Indoor tanning was a strong risk factor for early-onset BCC, particularly among women. Indoor tanning should continue to be targeted by both policy-based and behavioral interventions, as the impact on BCC-associated morbidity may be substantial.
In recent decades, the incidence of basal cell carcinoma (BCC), which comprises 80% of nonmelanoma skin cancers (NMSC),1-2 has been increasing.3-11 The rise has been striking among people under the age of 40,3, 9, 12 especially women,9, 12 pointing toward a corresponding change in environmental or lifestyle exposures. Because ultraviolet (UV) radiation is the primary environmental etiologic factor for BCC (reviewed in13-15), a logical hypothesis for the emergence of this malignancy among young people is increased exposure to UV.
Parallel trends of growing exposure to artificial UV from indoor tanning16-17 and increases in BCC incidence provide support at the ecologic level for the hypothesis that indoor tanning is leading to increases in BCC incidence rates among young people. Prevalence estimates of indoor tanning in developed countries vary widely (2.8%-47.0% tanned indoors in prior year).18 An estimated 30 million people tan indoors each year in the United States.16 Young women are the individuals most likely to engage in this behavior,18-19 lending additional support to indoor tanning playing a role in the changing patterns of BCC.
The International Agency for Research on Cancer (IARC) recently concluded there was “convincing evidence to support a causal association” between indoor tanning and melanoma and squamous cell skin cancer, but limited data for BCC did not support an association.17 Thus far, only one population-based case-control study of non-melanoma skin cancer has observed a significant 50% increased risk of BCC with ever indoor tanning;20 however, other research has been in populations of primarily older individuals with a low prevalence of indoor tanning.21-25 There is new interest in early-onset BCC with intriguing findings for indoor tanning as a risk factor, but this research has been limited by small sample sizes.26-27
Because the relationship between indoor tanning and BCC has been inconsistent and markedly understudied in relation to early-onset BCC, we evaluated this association in a large case-control study of individuals under age 40 in which indoor tanning was quite prevalent. In the context of the rising incidence of BCC among young people and indoor tanning being a modifiable risk factor, better understanding the relationship between this exposure and BCC could have a considerable impact on primary prevention.
The Yale Study of Skin Health in Young People is a case-control study of early-onset BCC conducted in Connecticut (July 2007-December 2010) described in detail elsewhere.28 BCC cases and controls with minor benign skin conditions diagnosed between July 1, 2006 and September 30, 2010 were identified through Yale University's Dermatopathology database. Eligible participants had to: be less than 40 years of age at the time of skin biopsy, reside in Connecticut, speak English, and themselves (or appropriate guardian for decisionally impaired individuals and those under age 18) be mentally and physically capable of completing all study components. Participants completed a structured in-person interview, self-administered questionnaires, and provided a saliva sample with Oragene®•DNA 2mL saliva collection kits (DNA Genotek Inc.; Ontario, Canada; http://www.dnagenotek.com/index.html). Yale University's Institutional Review Board approved the study (Protocol #0612002107, Approved: 02/02/2007) and study participants (or guardians) provided written informed consent.
Among the 665 potentially eligible BCC cases identified, 17 (2.6%) were ineligible upon initial contact: 14 lived out of state and 3 could not complete all study components. Of the remaining 648 individuals, 114 (17.6%) could not be contacted directly (no telephone number, non-working telephone number, only spoke to other person in household, left message only). Among the 534 cases we were able to directly reach and determine full eligibility, 389 enrolled (participation rate=72.8%) and 145 (27.2%) declined to participate. Cases were classified into single (only one BCC, n=242) or multiple (two or more BCCs, n=147) BCC under the age of 40 based on the Yale Dermatopathology database (data from 1990 on) and participant self-report.
Randomly sampled controls were frequency matched to BCC cases on age at biopsy (5 year age groups), gender, and biopsy site (head/neck, trunk, extremity). A variety of diagnoses were determined ineligible for sampling, including skin cancers/precancers (e.g., melanoma, squamous cell carcinoma, T-cell lymphomas, actinic keratoses), potentially UV-related benign conditions (e.g., solar lentigo, abnormal nevus), erythematous conditions associated with photosensitivity or aggravated by UV exposure (e.g. lupus erythematous, erythema multiforme, rosacea), dermal conditions treated with UV therapy (e.g., psoriasis), and pigment disorders (e.g., vitiligo). Among the 1,102 potentially eligible controls, 60 (5.4%) were ineligible upon initial contact (39 lived out of state, 10 non-English speakers, 2 did not recall having a skin biopsy, 1 hearing impaired, 1 hospitalized) or during the interview (7 self-reported a BCC). Of the remaining 1,042 individuals, 288 (27.6%) could not be contacted. Among the 754 potential controls we directly reached and determined full eligibility, 458 controls enrolled in the study (participation rate=60.7%) and 296 (39.3%) declined to participate. Our dermatologist (DJL) reviewed the individual diagnoses of all enrolled controls to ensure eligibility criteria were met. The most common control conditions were cyst (16.4%), seborrheic keratosis (16.2%), and wart (11.4%). All other conditions were present among less than 10% of controls.
The structured interview contained questions on sociodemographics, outdoor UV exposure (incidental exposure, intentional sunbathing), history of sunburns, sunscreen use, melanoma and non-melanoma skin cancer among first degree relatives, height, weight, alcohol intake, smoking status, and self-reported phenotype characteristics (eye, skin, and hair color; skin reaction to strong sunlight for the first time in the summer for one hour without sunscreen; skin reaction after repeated and prolonged exposure to sunlight; freckles on face; moles on the back ≥ 5 mm). Occupational UV exposure was gathered with a self-administered questionnaire. Questionnaires were adapted from those used by other recent epidemiologic studies20 to facilitate future data pooling. Interviewers were blinded to case-control status until the end of the interview, when participants were asked about their personal history of cancer.
Participants were also asked about their indoor tanning history (using established instruments) and were provided color photos of different types of tanning beds/booths as visual aids. We queried ever use of indoor tanning (regular tanning beds/booths, high speed/high intensity tanning beds/booths, high pressure tanning beds/booths), age first indoor tanned, and number of burns (any part of the body, skin biopsy site) from indoor tanning. Across four specified age periods (ages 11-15, 16-20, 21-30, and 31 plus), we obtained frequency of use and average length of tanning sessions. Participants were also asked the total number of years they had used each type of tanning bed/booth.
DNA was isolated from the saliva samples based on the manufacturer's protocol and variants in MC1R were obtained via sequencing, with the full methodology described previously.28 Sequencing was conducted at W. M. Keck Facility at Yale University using Applied Biosystems 3730 capillary instruments. Sequencing reactions utilized fluorescently-labeled dideoxynucleotides (Big Dye Terminators) and Taq FS DNA polymerase in a thermal cycling protocol. The sequence was analyzed using Sequencher 4.9 (Gene Codes Corporation) comparing the query sequence to the standard sequence with no variants in MC1R (NM_002386.3). MC1R variants were classified into synonymous and non-synonymous variants. All laboratory personnel were blinded to case-control status.
Our sample was limited to non-Hispanic Whites; 380 (97.7%) cases and 390 (85.2%) controls. One participant missing indoor tanning data and an additional three BCC cases with Gorlin Syndrome, which predisposes individuals to multiple BCCs early in life,29 were further excluded, leaving an analytic population of 376 cases and 390 controls.
Using multivariate logistic regression, we calculated odds ratios (ORs) and 95% confidence intervals (CIs) for the associations between indoor tanning and early-onset BCC. Continuous variables were categorized into tertiles based on the distribution of the exposures in controls who had tanned indoors; never indoor tanners served as the referent group. Our multivariate models included variables that altered the risk estimates by at least ten percent or were significantly associated with disease status in our population: skin color, family history of melanoma and/or non-melanoma skin cancer, first exposure of the season to one hour of summer sun, prolonged exposure to the sun, and MC1R non-synonymous variants. All models were also adjusted for the frequency matching variables of age at biopsy, body site of skin biopsy, and gender. Inclusion of the following exposures did not significantly alter risk estimates: education, eye color, hair color, moles ≥ 5 mm on back, freckles on face, body mass index (BMI), regular use of sunscreen, alcohol consumption, smoking status, incidental outdoor sun exposure, outdoor activities, sunburns, sunbathing sessions, and outdoor employment.
Trend tests were based on ordinal categorical variables representing the referent (never indoor tanners) and the tertiles of each exposure. We evaluated the linear trend using variables scored as the median of the tertiles, but due to the skewed nature of the exposures (e.g., tanning sessions, tanning hours), the ordinal scores appeared more appropriate and are presented here. This was supported by a goodness of fit test (χ2 distribution with k-2 degrees of freedom) taking the difference between the χ2 statistic from the model with k-1 variables for each exposure (where k=number of exposure categories) and 1) the χ2 statistic for a model with the ordinal categorical variable; and 2) the χ2 statistic for a model with the median scored variable.30
We tested interactions by body site of biopsy, skin color, MC1R variants, age at biopsy, and gender by including cross-product terms in the multivariate models. We calculated population attributable risk for case-control studies: P(E|D)(1-1/RR); where P=the proportion of cases exposed (E=exposure, D=disease) and RR=relative risk approximated by the OR based on the rare disease assumption.30 All descriptive and multivariate analyses were conducted using SAS Version 9.2 (SAS, Cary, NC) and reported p-values, except for tests of trend, are two-sided.
Of the 766 participants, 69.2% were female and the median age at skin biopsy was approximately 36 years. BCC cases were more likely to have fairer pigment-related characteristics, a family history of skin cancer, regularly used sunscreen on the body site of their skin biopsy, spent more time outdoors during warm months, and sunburned more frequently than controls (Table I). Cases were also more likely to have never smoked, have normal BMIs, and have attained higher education levels compared to controls.
Ever indoor tanning was associated with a 69% increased risk of early-onset BCC (OR=1.69, 95% CI=1.15-2.48) (Table II). This association was stronger for multiple BCC case status (OR=2.16, 95% CI=1.26-3.70) than single BCC cases (OR=1.46, 95% CI=0.96-2.22). In a sensitivity analysis removing controls with the three most common conditions one at a time, there was little impact on the association (data not shown).
Indoor tanning frequency was positively associated with early-onset BCC, with evidence for statistically significant increased risk across all three tertiles of sessions and the top two tertiles of hours spent indoor tanning (Table II). BCC risk was slightly higher for the youngest age of initiation (≤ 16 years OR=1.83, 95% CI=1.12-2.97) as compared to the upper tertiles. Dichotomizing at the median of 17.4 years elapsed between first indoor tanning and skin biopsy, we observed a slightly stronger association between indoor tanning and BCC with longer (OR=1.77, 95% CI=1.13-2.80) versus shorter (OR=1.63, 95% CI=1.07-2.51) time elapsed, although both were statistically significant.
Having been burned while indoor tanning (OR=1.87, 95% CI =1.17-2.97), particularly burning at the site of the skin biopsy (OR=2.72, 95% CI =1.57-4.69), was strongly associated with early-onset BCC (Table II). The number of overall burns (p-trend=<0.001) and number of burns specifically to the biopsy site (p-trend=<0.001) also showed a positive linear relationship with BCC.
Risk of early-onset BCC was significantly increased with ever use of each type of tanning bed/booth, with stronger associations for high speed/high intensity and high pressure tanning devices (Table II). Years of use of regular tanning beds/booths showed a positive linear relationship with BCC risk (p-trend=0.003), with those who tanned indoors six or more years greater than two-fold more likely to have BCC than never indoor tanners (OR=2.16, 95% CI=1.34-3.48). Years of use of high speed/high intensity (p-trend=0.030) and high pressure (p-trend=0.004) tanning beds/booths were also positively linearly associated with BCC (data not shown).
Females who tanned indoors were approximately two times more likely to have a BCC compared to female never indoor tanners (OR=2.14, 95% CI=1.31-3.47), whereas among men this association did not reach statistical significance (OR=1.16, 95% CI=0.60-2.25) (Table III). There was little evidence of an association between indoor tanning and BCCs located on the head/neck, yet there was an approximately two-fold and four-fold increased risk for BCCs on the trunk and extremities, respectively. When trunk and extremities were combined, body site significantly modified the effect of ever indoor tanning (pinteraction=0.012; OR=2.81, 95% CI=1.57-5.02). We observed stronger associations for ever indoor tanning among women by body site; non-significant 35% increased risk for BCC on the head/neck (95% CI=0.69-2.64) and statistically significant associations for BCC on the extremities (OR=6.55, 95% CI=1.87-22.95) and trunk (OR=2.89, 95% CI=1.08-7.65).
The adverse effect of indoor tanning was primarily observed in persons with one or more non-synonymous MC1R variants (OR=1.99, 95% CI=1.28-3.09), although the gene-environment interaction was not significant (pinteraction=0.194) (Table III). Indoor tanning also appeared to be more harmful in persons who had very fair skin (OR=2.26, 95% CI=1.08-4.74), as compared to fair or olive skin (OR 1.56, 95% CI=0.99-2.46), but this interaction was also not significant (pinteraction=0.730). The association between indoor tanning and BCC was not modified by age at biopsy (data not shown).
Based on calculations of population attributable risk, approximately 27% of early-onset BCC cases could be prevented if individuals never tanned indoors. Among women under age 40, the proportion of preventable cases was even higher, with 43% of BCCs avoided if females did not tan indoors.
In this case-control study of early-onset BCC, we observed a 69% increased risk of disease with ever indoor tanning. The indoor tanning association was stronger for cases with multiple BCCs and more pronounced for women, as female indoor tanners were two times more likely to have BCC than women who had never tanned indoors. Indoor tanning was also more strongly associated with BCCs located on the trunk and extremities, body sites likely to be exposed predominantly when tanning, as compared to lesions on the head/neck, which receive considerable incidental UV exposure.
Prior to this investigation, research on indoor tanning and BCC (summarized in Table IV) had largely been in older populations,20-25 with only two small studies of early-onset BCC.26-27 The prevalence of indoor tanning in studies of BCC cases of all ages has been quite low, and in combination with limited sample sizes, may have hindered the ability to detect associations if they existed. Several other studies, also with limited power and lack of quantitative measures, have evaluated indoor tanning in relation to multiple skin cancer types combined, with a 24% non-significant increased risk for skin malignancies on the head/neck31 and null results for NMSCs in two hospital case-control studies.32-33
The summary association between indoor tanning and melanoma from a meta-analysis was statistically significant, but of fairly modest effect size (OR= 1.15, 95% CI=1.00-1.31).17 However, much of the melanoma literature and many studies of indoor tanning and nonmelanoma skin cancer suffer from important limitations, including lack of sun exposure data, low prevalence of indoor tanning, and no quantitative information to examine dose-response relationships. Recent studies in melanoma, done in younger and more highly exposed populations, that addressed many of these limitations observed much stronger associations of melanoma with ever indoor tanning, as well as dose-response relationships.34-35 In our study among a highly exposed population with extensive data on indoor tanning and sun exposure, the risk estimate for indoor tanning in relation to BCC was very similar to newer findings for melanoma,34-35 highlighting the importance of study design and population exposure in interpreting findings regarding health effects of indoor tanning.
Age at initiation of indoor tanning may be an important component of skin cancer risk, as younger age at initiation has been more strongly associated with both melanoma overall17 and early-onset melanoma,35 with a suggestive trend for BCC.20 However, other evidence suggests a similar melanoma risk regardless of the age at initiation.34 The latter observation is consistent with our findings of increased risk of early-onset BCC across all ages of initiation of indoor tanning, but the range of age of when individuals first tanned indoors was fairly narrow in our young population; 95% started tanning indoors when they were 25 years of age or less and half reported their first use at age 17 or under.
In our sample, indoor tanning was much more common and frequent among women, and our population being predominantly female, limited our ability to examine the association in males. The stronger effect of indoor tanning in women is likely due to a number of factors, including earlier age at initiation, greater number of tanning sessions (more individuals with greater exposure), and a larger proportion of women with tumors located on the trunk and extremities, which were more strongly related to indoor tanning in our data. While exposure differences are the most likely explanation for the gender difference, some of the observed effects could be due to other unidentified factors and should be investigated in future larger or pooled studies. Of note, among men we saw the same pattern by body site, with elevated, though non-significant, associations for indoor tanning in relation to trunk and extremity BCCs. The differences we observed by body site are important, as they highlight that for those body parts that are less likely to be exposed to incidental solar UV, the effect of indoor tanning may be more pronounced. Consideration of body site in future studies may be necessary to accurately characterize risk. Our finding of an increased risk of UV from indoor tanning on BCC among individuals with at least one non-synonymous variant in MC1R suggests potential interactions between this gene and UV exposure that should be explored in larger studies.
Burns from indoor tanning were strongly related to risk of early-onset BCC; with evidence of a dose-response effect. Potential recall bias could be particularly relevant to reporting burns specifically to the biopsy site. Conversely, social desirability bias may have also been an issue, with BCC cases possibly under-reporting overall indoor tanning. Although the impact of these potential biases on our results are unknown, the percentage of cases and controls, 28% and 23% respectively, who reported burns from indoor tanning, was similar to the approximately 20% of individuals in general population samples who reported burns from tanning indoors in the previous 12 months.36-38
Our study had several important strengths including adequate power, particularly among women, to examine the relationship between BCC and indoor tanning in an extremely relevant population, as well as assessment of numerous exposures as potential confounders. We were also able to evaluate dose-response relationships and as these were statistically significant, lend strength to our findings. Our study design limited the potential for interviewer bias, and because controls had also undergone skin biopsy, the potential for differential recall of behaviors by case status may have been minimized. In addition, by identifying cases and controls from a centralized dermatopathology facility serving many dermatologists in Connecticut, our controls represent the source population of our cases; that is, young people who see a dermatologist for a skin condition. Because study participants were all under age 40 at the time of skin biopsy, their reporting of indoor tanning was less subject to poor recall than older populations. Our sensitivity analyses removing individual control diagnoses indicated our findings were not driven by the inclusion of any particular benign condition.
In addition to the potential biases mentioned earlier, as in any observational study, it is possible that the association we observed is due, in part, to other unmeasured factors or residual confounding. Arguing against this, we considered known correlates of indoor tanning18 and evaluated numerous characteristics as potential confounders, including incidental and intentional sun exposure. In addition, there is a chance that participants in our study were not representative of all individuals under age 40 in Connecticut seen by dermatologists. Another limitation is related to our control group being individuals who saw a dermatologist for biopsy of a benign skin condition and the potential for their indoor tanning behaviors to differ from a more general population sample of people under age 40. Our controls may be very aware of their skin health and therefore less likely to tan indoors than the general population. Alternatively, our control group may be enriched with individuals highly focused on their appearance who utilize indoor tanning to a greater extent. While population-based controls are often sought in case-control studies, because our cases either sought out or were referred for dermatologic evaluation for a lesion under age 40, which could track with factors like socioeconomic status or insurance status that may also be correlated with tanning, population-based controls could also have introduced bias.
Indoor tanning was a strong risk factor for BCC in a population of individuals under age 40. We observed stronger associations in women, for BCCs located on the trunk and extremities, and for multiple BCCs. With a lack of epidemiologic data on indoor tanning and BCC risk in any age group, this research adds substantially to our understanding of this relationship. Our findings are in line with and extend the recent conclusions of IARC classifying UV-emitting tanning devices as Group I carcinogens.39 While additional replication in studies with different control populations and/or in studies with prospectively collected exposure data on indoor tanning are necessary to confirm the positive association we observed between indoor tanning and BCC, our results fulfill many of the criteria for causality including biologic plausibility, strength of the association, dose-response effects, specificity of effect to the body sites most uniquely exposed to indoor tanning, and coherent findings with melanoma studies. The increased prevalence of indoor tanning, especially in young women, parallels an increase in BCC, which is also more pronounced in young women. We thus conclude that indoor tanning is a risk factor for early-onset BCC and appears to be causally contributing to the increasing incidence of this malignancy. Both policy-based and behavioral interventions, to restrict or reduce indoor tanning in young people, are needed to alter the increasing incidence of this most common human malignancy.
This work was supported by the Yale SPORE in Skin Cancer funded by the National Cancer Institute grant number 1 P50 CA121974 (R. Halaban, PI). We would like to acknowledge the following individuals for their overall support and assistance with the coordination of this project: Dr. Jennifer McNiff, Robert Criscuolo, and James Platt from Yale Dermatopathology; Dr. Valencia Thomas from Yale Dermatology (now at University of Texas Medical School at Houston); and James McCusker from the Biostatistics/Bioinformatics Core of the Yale SPORE. We would also like to recognize and thank our interviewers, Carol Gordon and Lisa Lyon, for their dedication and skill in recruiting and interviewing the study participants. Finally, we are indebted to the individuals who participated in this study.
Financial Support: This work was supported by the National Cancer Institute of the National Institutes of Health (1P50 CA121974 and 1F32 CA144335) and a CTSA Grant from the National Center for Research Resources of the National Institutes of Health (UL1 RR024139).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest: The authors have no conflict of interest to declare.
Prior Presentation: Results have not been presented previously in any format.