Prevalence and Correlates of Skin Cancer Screening Among Indoor Tanners and Nontanners

Key Points

Question

What are the prevalence and correlates of skin cancer screening among indoor tanners and nontanners?

Findings

In this survey study of 30 352 US adults who participated in the 2015 National Health Interview Survey, correlates for screening for tanners and nontanners were older age, higher income, seeking online health information, family skin cancer history, very high sun protection factor sunscreen use, and receipt of professional spray-on tans. Correlates for screening for nontanners only were white race, non-Hispanic ethnicity, email use, having a usual clinic or physician’s office, emergency department visits, previous cancer diagnosis, no worry about medical bills, sun protection, and sunless self-tanning.

Meaning

Indoor tanners have a greater tendency to be screened for skin cancer and to be screened at an earlier age than NTs. However, skin cancer screening is still underused in this high-risk population.

Abstract

Importance

The US Food and Drug Administration recommends that indoor tanners (ITs) be screened regularly for skin cancer (SC).

Objective

To investigate the association between indoor tanning and SC screening.

Design, Setting, and Participants

The 2015 National Health Interview Survey was a multistage, clustered, cross-sectional design with 30 352 US adults participating. The response rate for the sample adult data used in this study was 55.20% after excluding 1099 individuals who reported a history of SC and 2221 individuals with unknown SC screening or indoor tanning history. To examine the independent correlates of screening, we conducted multiple logistic regressions separately for ITs and nontanners (NTs), simultaneously including all preselected variables of interest as potential predictors. Formal interaction analyses were also performed to determine if the covariate effects differed significantly between ITs and NTs.

Exposures

Indoor tanning as well as sociodemographic, health care, and SC risk and sun protection factors.

Main Outcomes and Measures

The primary outcome was self-reported full-body SC screening by a physician. Univariable and multivariable analyses were conducted to determine the secondary outcome, correlates of SC screening among ITs and NTs.

Results

A total of 15 777 participants (51.98%) were female, and 23 823 (78.49%) were white; 4987 (16.43%) of the sample had indoor tanned, and 1077 (21.59%) of these had tanned last year. A total of 1505 ITs (30.18%) and 4951 NTs (19.52%) had been screened for SC. Correlates of screening for ITs and NTs were older age (ITs: odds ratio [OR], 4.29 [95% CI, 2.72-6.76]; NTs, OR, 5.14 [95% CI, 4.01-6.58], age ≥65 years vs 18-29 years), higher income (ITs: OR, 2.08 [95% CI, 1.50-2.88]; NTs: OR, 1.79 [95% CI, 1.51-2.12]; >$100 000 vs $0-34 999), seeking online health information (ITs, OR, 0.71 [95% CI. 0.56-0.91; NTs, OR, 0.65 [95% CI, 0.58-0.72], for not looking up health info online), family history of melanoma (ITs: OR, 1.92 [95% CI, 1.26-2.93]; NTs: OR, 1.58 [95% CI, 1.21-2.05]) or SC (ITs: OR, 1.59 [95% CI, 1.17-2.17; NTs: OR, 1.61 [95% CI, 1.33-1.94]), very high SPF sunscreen use (ITs: OR, 0.57 [95% CI, 0.42-0.78]; NTs: OR, 0.71 [95% CI, 0.61-0.82], use of SPF of 1-14 vs SPF of >50), and receipt of a professional spray-on tan (ITs: OR, 0.60 [ 95% CI, 0.41-0.88]; NTs: OR, 0.51 [95% CI, 0.32-0.81], for not receiving a salon spray-on tan). Correlates for NTs only were white race (blacks: OR, 0.45 [95% CI, 0.37-0.54], others: OR, 0.40 [95% CI, 0.33-0.48]), non-Hispanic ethnicity (Hispanics: OR, 0.42 [95% CI, 0.36-0.50]), email use (no email: OR, 0.67 [95% CI, 0.56-0.80]), having a usual clinic/or physician’s office (no usual place: OR, 0.56 [95% CI, 0.40-0.78]), emergency department visits (OR, 1.20 [95% CI, 1.06-1.35]), having had a previous cancer diagnosis (no cancer diagnosis: OR, 0.67 [95% CI, 0.57-0.79]), not being worried about medical bills (OR, 1.37 [95% CI, 1.15-1.63] vs very worried), sun protection (rarely/never: OR, 0.43 [95% CI, 0.34-0.56]), and sunless self-tanning (not using: OR, 0.62 [95% CI, 0.47-0.83]).

Conclusions and Relevance

Few ITs have been screened for SC, although SC rates are higher than among NTs. It is not surprising that SC screening is associated with SC risk factors (eg, family history of SC and age) among ITs. However, some unscreened ITs may be putting themselves at even greater risk of SC by also being more likely to use low SPF sunscreen than ITs who have been screened for SC.

This survey study of 30 352 participants investigates the association between indoor tanning and skin cancer screening.

Introduction

Almost 5 million skin cancers are diagnosed annually in the United States, and skin cancer incidence has been rising over the past few decades.^1,2,3,4 Survival rates for skin cancers are high, particularly when detected early.⁵ Skin cancer screening is associated with detection of thinner tumors, which are associated with lower mortality rates.⁵ Research has not yet shown a direct mortality benefit for skin cancer screening, although screening is recommended for certain high-risk groups.^6,7

The millions of Americans who have tanned indoors are at much higher risk of melanoma and keratinocyte skin cancers than the general population.^8,9 Owing to accumulating evidence of the harms of indoor tanning, the US Food and Drug Administration (FDA) and similar bodies in other countries have become increasingly stringent about its regulation. In 2014, the FDA required manufacturers of indoor tanning devices to provide this warning statement to users: “Warning: Persons repeatedly exposed to UV radiation should be regularly evaluated for skin cancer.”¹⁰ In December 2015, the FDA proposed a rule to require users to sign a risk acknowledgment certification before use and every 6 months thereafter, stating that they were informed of health risks. One of the proposed required risk statements is “If you use indoor UV tanning devices and/or tan regularly outdoors, get regular skin cancer checkups from your physician because you are more likely to develop skin cancer.”¹¹ Because these risk statements are relatively new and do not seem to have been well publicized, it is doubtful that most primary care physicians or indoor tanners (ITs) are aware of them.

Despite the relatively large amount of literature on skin cancer screening and increasing attention to indoor tanning, few studies have investigated the intersection between the 2. Several European studies have found that, compared with nontanners (NTs), ITs are less likely to conduct skin self-examinations or receive a full-body examination by a clinician, receive early diagnosis, or survive melanoma.^12,13,14,15 A review of literature from Northern Europe also noted an association among sunbed use, lower socioeconomic status, later diagnosis, thicker tumors, and higher mortality.¹⁴

Using data from the 2005 National Health Interview Survey (NHIS), a national probability survey of the US adult population, our team previously found that the proportion of individuals who had tanned indoors in the past year was not significantly different between individuals who had ever been screened for skin cancer by a physician (7%-30%) vs those who had not been screened (8%-30%), based on age-stratified analyses.¹⁶ However, using 2010 NHIS data, Lakhani and colleagues¹⁷ found that individuals who had tanned indoors in the past year were more likely than NTs to have been screened for skin cancer. Using 2015 NHIS data, Fischer and colleagues¹⁸ did not find a difference in the skin cancer screening rate among recent ITs stratified by age or sex compared with NTs. However, the authors focused on frequency of indoor tanning in the past year rather than ever engaging in indoor tanning and did not assess screening correlates. The methods of these studies have varied and produced mixed findings, and the most recent study included only a subsample of the 2015 NHIS data. Therefore, the current study extends the investigation of the prevalence and correlates of skin cancer screening among ITs using the full 2015 NHIS sample.

Methods

Procedures

This study used data from the 2015 NHIS, a national health survey of the US adult population. The NHIS uses a multistage, clustered, cross-sectional design, with stratification at the state level and oversampling of Hispanic and black populations. The response rate for the sample adult data used in this study was 55.20%; the methodology of the 2015 NHIS is described in detail elsewhere.¹⁹ The current study used publically available deidentified archival data and, thus, did not involve human subjects and was exempt from review by the Fox Chase Cancer Center institutional review board. The sample included 30 352 participants after excluding 1099 individuals who reported a history of skin cancer and 2221 individuals with unknown skin cancer screening or indoor tanning history.

Measures

Indoor Tanning, Demographics, and Skin Cancer Screening

Participants were asked if they had ever used an indoor tanning device, such as a sunlamp, sunbed, or tanning booth. Individuals who reported any history of indoor tanning were defined as ITs, and participants who had never indoor tanned were considered NTs. Participants indicated their sex, age, race/ethnicity, sexual orientation, marital status, internet use, email use, employment status, health insurance status, household income, and level of worry about paying monthly bills. Participants were asked whether they had ever had all of their skin from head to toe examined by a dermatologist or other kind of physician and the main reason for having the examination.

Health Information and Health Care Accessibility

Participants were asked about their level of worry about paying medical bills if sick or in an accident, having looked up health information online in the past 12 months, availability of a place to go when sick or in need of health advice, having a usual place to go for routine and/or preventive health care, and having visited the emergency department in the past 12 months.

Skin Cancer Risk and Protection Factors

Participants were asked whether they had ever had cancer (other than skin cancer) and whether any first-degree relatives had ever had melanoma, nonmelanoma, or an unknown type of skin cancer. Five items asked individuals to report how often they engaged in sun protection behaviors (on a 5-point scale from always [1] to never [5] as well as the option “don’t go in the sun”), including staying in the shade, wearing a wide-brimmed hat, wearing a long-sleeved shirt, wearing long pants, and using sunscreen when outside on a warm sunny day for over an hour. Because these items were highly correlated, they were averaged to create a sun protection score. The sun protection data were considered categorical rather than continuous for 2 reasons: (1) to avoid assuming a linear effect of the underlying ordinal Likert scales (1 = always, 5 = never) and (2) it is unclear how to include the response “don’t go in the sun” on a continuous scale. Other items asked participants how much they would burn if they went in the sun for an hour with no protection and how dark a tan they would get if they went in the sun repeatedly, such as every day for 2 weeks without protection. Participants also indicated what sun protection factor (SPF) sunscreen they used most often and how many times they had had a sunburn in the past 12 months (categorized for analyses into 0, 1, or ≥2). Participants were also asked whether they had used self-applied sunless tanning products and whether they had received a professional spray-on tan in the past 12 months.

Data Weighting and Statistical Analyses

Statistical analyses were conducted using Stata software (version 12; StataCorp) and accounted for the survey design using variables provided by the NHIS with the data file. Sample weights accounted for study design, ratios, and nonresponse, with poststratification adjustments for US Census-based estimates of age, sex, and race/ethnicity. Standard errors were calculated using first-order Taylor linearization (Huber-White sandwich variance estimate). All percentages reported in the tables and significance tests are weighted, and all sample sizes are unweighted.

Given the multiple associations examined and the large sample size, a cutoff of P < .01 was used to determine statistical significance. A series of χ² analyses was conducted to examine the extent to which each potential correlate was associated with skin cancer screening (binary indicator for ever vs never) among ITs and NTs. To examine the independent correlates of screening, we conducted multiple logistic regressions separately for ITs and NTs, simultaneously including all preselected variables of interest as potential predictors. This method was selected rather than including only variables significant in univariable analyses because a lack of significance would not necessarily indicate sufficient covariate balance, and simultaneous covariate effects would not be considered.²⁰

Formal interaction analyses were also performed to determine if the covariate effects differed significantly between ITs and NTs. We considered variables with P < .05 in either subgroup for interaction effects. Separate logistic regression models were conducted for each candidate interaction term because a single model with all potential interactions was intractable. To assess whether interaction effects were present, an F test was performed (jointly assessing all covariate levels for the interaction of interest).

Results

Sample demographic characteristics are provided in Table 1 and eTable 1 in the Supplement. The amount of missing data was small, with most variables having less than 1.00% missing, with a maximum proportion of 9.17% missing for household income. Indoor tanners comprised 4987 (16.43%) of the sample, and 1077 (21.59%) of these had tanned indoors in the past year. A total of 1505 (30.18%) of ITs, and 4951 (19.52%) of NTs had been screened for skin cancer. Of the individuals who reported having been screened for skin cancer, 4420 (68.47%) reported that this was part of a routine medical examination, 1553 (23.59%) that it was because of a problem, and 495 (7.66%) reported “other” as the reason.

Table 1. Characteristics of 30 352 Participants.

Variable	Full Sample %a
Female sex	51.98
Age, y
18-29	21.92
30-39	17.44
40-49	16.92
50-64	25.68
≥65	18.05
Race
White	78.49
Black	12.43
Other	9.08
Hispanic ethnicity	16.16
Employment status
Working	61.94
Not working	38.03
Family member with health insurance	95.44
Sexual orientation
Heterosexual	96.18
Not heterosexual	2.79
Missing	1.03
Marital status
Married or living with partner	60.15
Widowed/divorced/separated	17.18
Household income, $
0-34 999	27.63
35 000-74 999	27.35
75 000-99 999	11.46
≥100 000	24.40
Missing	9.17
Ever indoor tanned	16.43
Indoor tanned in past 12 mo (indoor tanners only)	21.59
Ever had a clinical skin examination	21.27

^aMissing data are reported only when ≥1% of responses were missing.

Results of the univariate analyses are provided in Table 2 and eTable 2 in the Supplement. We focus herein on the results of the multivariable logistic regressions. Significant (P < .01) sociodemographic and/or health care correlates of skin cancer screening for both ITs and NTs were older age (ITs: odds ratio [OR], 4.29 [95% CI, 2.72-6.76]; NTs: OR, 5.14 [95% CI, 4.01-6.58]; age ≥65 years vs 18-29 years), higher household income (ITs: OR, 2.08 [95% CI, 1.50-2.88]; NTs: OR, 1.79 [95% CI, 1.51-2.12]; ≥$100 000 vs $0-34 999), and looking up health information online in the past 12 months (ITs: OR, 0.71 [95% CI, 0.56-.91]; NTs, OR, 0.65 [95% CI, 0.58-0.72], for not looking up health info online). See Table 2 for variables significant in the multivariable analyses and eTable 2 in the Supplement for remaining multivariable and univariable results. Skin cancer risk and protective factors and behaviors related to these factors were examined in the multivariable logistic regressions as well. Significant (P < .01) correlates of skin cancer screening for both ITs and NTs were a history of melanoma (ITs: OR, 1.92 [95% CI, 1.26-2.93]; NTs: OR, 1.58 [95% CI, 1.21-2.05]) or skin cancer (ITs: OR, 1.59 [95% CI, 1.17-2.17]; NTs: OR, 1.61 [95% CI, 1.33-1.94]) in a first-degree relative, use of very high SPF sunscreen (ITs: OR, 0.57 [95% CI, 0.42-0.78]; NTs: OR, 0.71 [95% CI, 0.61-0.82]; use of SPF of 1-14 vs SPF of ≥50) and receipt of a spray-on tan at a salon in the past 12 months (ITs: OR, 0.60 [95% CI, 0.41-0.88]; NTs: OR, 0.51 [95% CI, 0.32-0.81] for not receiving a salon spray-on tan).

Table 2. Univariate and Multivariable (MV) Logistic Regression Analyses Examining Correlates of Skin Cancer Screening by Indoor Tanning Status^a.

Variable	Ever Indoor Tanners		Nontanners
	% Screenedb	OR (95% CI)c MV Adjusted	% Screenedb	OR (95% CI)c MV Adjusted
Age, y
18-29	17.05	1 [Reference]	7.92	1 [Reference]
30-39	25.95	1.44 (0.99-2.09)	12.40	1.49 (1.19-1.85)
40-49	33.48	2.12 (1.49-3.01)	16.82	1.97 (1.58-2.47)
50-64	36.94	2.38 (1.68-3.38)	25.04	3.15 (2.51-3.95)
≥65	46.80	4.29 (2.72-6.76)	33.24	5.14 (4.01-6.58)
Race
White	30.58	1 [Reference]	23.10	1 [Reference]
Black	17.49	0.63 (0.19-2.07)	8.46	0.45 (0.37-0.54)
Other	19.52	0.77 (0.49-1.19)	9.30	0.40 (0.33-0.48)
Hispanic
Yes	16.79	0.61 (0.37-1.01)	7.13	0.42 (0.36-0.50)
No	30.73	1 [Reference]	22.35	1 [Reference]
Email use
Yes	31.38	1 [Reference]	22.14	1 [Reference]
No	22.01	1.01 (0.69-1.49)	14.24	0.67 (0.56-0.80)
Household income, $
0-34 999	17.23	1 [Reference]	12.81	1 [Reference]
35 000-74 999	25.73	1.37 (1.02-1.85)	16.90	1.14 (1.00-1.30)
75 000-99 999	31.59	1.48 (1.03-2.10)	21.83	1.38 (1.14-1.67)
≥100 000	40.96	2.08 (1.50-2.88)	29.81	1.79 (1.51-2.12)
Worried about paying medical bills if sick or accident
Very worried	21.72	1 [Reference]	10.33	1 [Reference]
Somewhat worried	26.96	0.86 (0.61-1.22)	17.30	1.20 (1.02-1.35)
Not at all worried	34.27	1.16 (0.81-1.65)	23.35	1.37 (1.15-1.63)
Looked up health info on internet in past 12 mo
Yes	30.83	1 [Reference]	24.75	1 [Reference]
No	21.05	0.71 (0.56-0.91)	14.96	0.65 (0.58-0.72)
Usual place to go for routine/preventative health care
No routine/preventative care	13.92	0.80 (0.47-1.35)	5.38	0.56 (0.40-0.78)
Clinic/physician’s office	31.96	1 [Reference]	21.59	1 [Reference]
Hospital	8.77	0.18 (0.04-0.95)	9.26	0.70 (0.40-1.23)
Other/no usual place	24.22	1.13 (0.47-2.70)	11.65	1.00 (0.68-1.45)
Emergency department visit in past 12 mo
Yes	25.26	0.95 (0.75-1.19)	19.33	1.20 (1.06-1.35)
No	31.38	1 [Reference]	20.37	1 [Reference]
Ever diagnosed with cancer (other than skin cancer)
Yes	41.35	1 [Reference]	38.24	1 [Reference]
No	29.61	0.80 (0.54-1.17)	18.30	0.67 (0.57-0.79)
History of melanoma in first-degree relative
Yes	47.56	1.92 (1.26-2.93)	39.01	1.58 (1.21-2.05)
No	29.31	1 [Reference]	19.07	1 [Reference]
History of nonmelanoma or unknown skin cancer type in first-degree relative
Yes	44.65	1.59 (1.17-2.17)	39.38	1.61 (1.33-1.94)
No	28.86	1 [Reference]	18.61	1 [Reference]
Sun protection
1-1.9 (always)	31.31	1 [Reference]	36.01	1 [Reference]
2-2.9 (most of the time)	45.97	2.16 (1.02-4.58)	27.50	0.73 (0.58-0.92)
3-3.9 (sometimes)	33.33	1.55 (0.74-3.26)	19.68	0.54 (0.43-0.68)
4-5 (rarely, never)	22.5	1.22 (0.57-2.61)	13.85	0.43 (0.34-0.56)
Don’t go out in the sun	28.11	0.78 (0.14-4.37)	17.04	0.45 (0.30-0.67)
Sun protection factor used most often
1-14	15.48	0.57 (0.42-0.78)	12.65	0.71 (0.61-0.82)
15-29	30.91	0.97 (0.71-1.33)	21.78	0.82 (0.67-0.99)
30-49	37.89	1.21 (0.96-1.53)	27.07	0.98 (0.85-1.12)
≥50	31.04	1 [Reference]	25.33	1 [Reference]
Don’t go out in the sun	35.87	2.51 (0.41-15.23)	17.11	0.81 (0.48-1.21)
Used self-applied sunless tanning products in past 12 mo
Yes	37.63	1 [Reference]	38.81	1 [Reference]
No	28.54	0.78 (0.62-0.99)	19.00	0.62 (0.47-0.83)
Received spray-on tan at a salon or other business in past 12 mo
Yes	38.51	1 [Reference]	39.27	1 [Reference]
No	29.57	0.60 (0.41-0.88)	19.38	0.51 (0.32-0.81)

Abbreviation: OR, odds ratio.

^aA separate category for missing data was included for each categorical variable with missing data. These categories are not reported in this Table.

^bAssociations with skin cancer screening significant at P < .01 in unadjusted univariate models are denoted using bold font (ie, the proportion of individuals who were screened differed significantly among categories within a variable).

^cAssociations with skin cancer screening significant at P < .01 in adjusted multivariable models are also denoted using bold font. This Table presents only variables found to have significant associations in indoor tanners or nontanners in the multivariable analyses. Results for nonsignificant variables in the multivariable analyses are presented in eTable 2 in the Supplement. Nonsignificant variables included sex, employment, health insurance, sexual orientation, marital status, internet use, worry about bills, availability of a place to receive health care when sick, sunburn history, and skin reaction to sun (short and long term).

Remaining variables were significantly associated with screening in NTs only. Significant sociodemographic correlates were white race (blacks: OR, 0.45 [95% CI, 0.37-0.54]; others: OR, 0.40 [95% CI, 0.33-0.48]), non-Hispanic ethnicity (Hispanics: OR, 0.42 [95% CI, 0.36-0.50]), and using email (no email: OR, 0.67 [95% CI, 0.56-0.80]). Significant health care correlates were having a usual clinic or physician’s office to go for routine and preventive health care (no usual place: OR, 0.56 [95% CI, 0.40-0.78]), having visited an emergency department in the past 12 months (OR, 1.20 [95% CI, 1.06-1.35]), and not being worried about paying medical bills (OR, 1.37 [95% CI, 1.15-1.63] vs very worried). Significant skin cancer risk or protective correlates were a cancer diagnosis other than skin cancer (no cancer diagnosis: OR, 0.67 [95% CI, 0.57-0.79]), always using sun protection (rarely or never: OR, 0.43 [95% CI, 0.34-0.56]), and use of self-applied sunless tanning products in the past 12 months (not using: OR, 0.62 [95% CI, 0.47-0.83]). Although more variables were significant correlates of skin cancer screening for NTs compared with ITs, no significant (P < .01) interactions were found (results not shown).

In secondary analyses examining recency of indoor tanning, it was found that compared with NTs, those who had tanned more than 12 months ago were more likely to have been screened for skin cancer (OR, 1.27; 95% CI, 1.12-1.45; P < .001), but indoor tanning in the past 12 months was not significantly associated with screening (OR, 1.07; 95% CI, 0.86-1.33; P = .56). Of individuals who had indoor tanned more than 1 year ago, 1257 (32.14%) had been screened for skin cancer, compared with 248 individuals (23.07%) who had tanned indoors in the past 12 months, and 4951 NTs (19.52%).

Discussion

Regarding the prevalence of skin cancer screening, only 30.18% of ITs and 19.52% of NTs had ever received a full-body skin cancer examination by a physician. Among ITs and NTs, screening was significantly associated with older age, family skin cancer history, higher household income, looking up health information online, use of very high SPF sunscreen, and receiving a professional spray-on tan in the past 12 months. Additional variables associated with screening in NTs only were white race, non-Hispanic ethnicity, using email, having a usual clinic or physician’s office to go for routine and/or preventive health care, having visited an emergency department in the past 12 months, not being worried about paying medical bills, a cancer diagnosis other than skin cancer, always using sun protection, and use of self-applied sunless tanning products in the past 12 months.

This study adds to the small quantity of literature on the relationship between indoor tanning and skin cancer screening. Although the FDA recommends that individuals repeatedly exposed to UV radiation be screened regularly for skin cancer,^10,11 less than a third of ITs had ever been screened. The current findings contrast with those of several European studies in which NTs were more likely to have been screened than ITs.^12,13,14,15 That ITs were more likely to get skin cancer screening than NTs in the current study is similar to the findings of Lakhani and colleagues¹⁷ using 2010 NHIS data. However, the current study seems to be the first to use NHIS data focusing on ever, rather than recent, indoor tanning. It was found that more ITs are getting screened at younger ages than NTs, which is interesting given that the association between indoor tanning and skin cancer screening is primarily driven by those whose most recent indoor tanning event was more than 12 months ago. The association of screening with past, rather than recent, indoor tanning might indicate that former ITs are more likely than recent ITs to be aware of their increased risk and seek skin cancer screening accordingly. Two prior NHIS studies did not find a difference in the screening rates among ITs compared with NTs, but one study used data that is more than 10 years old,¹⁶ and the other sample included only non-Hispanic white adults ages 18 to 60 years.¹⁸

Several of the multivariable associations between skin cancer screening and sociodemographic and skin cancer risk and protective factors are consistent with findings of prior research.²¹ Although some variables were significant correlates of screening in NTs but not in ITs, significant interactions were not found. In addition, there were no variables that were significant for ITs but not NTs. This suggests that differences in significance were primarily driven by the larger and more heterogeneous sample of NTs, and not subgroup-specific effects. For example, ITs are more likely to be non-Hispanic whites than NTs. Thus, there is less power to demonstrate a difference in screening among ITs by race or ethnicity. Therefore, the correlates that were statistically significant for both groups are highlighted here. First, consistent with prior research,²¹ a history of skin cancer among first-degree relatives was associated with skin cancer screening. Greater awareness of family risk and its implications, particularly among ITs, could help motivate prevention and detection. However, some individuals are not knowledgeable about the skin cancer history of their family, for example—whether the skin cancers were melanomas or keratinocyte skin cancers.²² Second, variables related to health information or potential access to health care (ie, looking up health information online and higher household income) were associated with screening. Notably, prior research found an association of lower socioeconomic status with thicker and more advanced melanomas and higher mortality.^14,23 Third, some skin cancer protective factors were associated with being screened for skin cancer. Similar to prior research in the general population,²¹ use of low-SPF sunscreen (compared with very high SPF sunscreen) was associated with a lower likelihood of skin cancer screening among ITs. Thus, some individuals may be putting themselves at risk for skin cancer in multiple ways (eg, by having tanned indoors and using ineffective sunscreen) and are also underusing skin cancer detection strategies. Finally, we found no prior study that had investigated the association between sunless tanning and skin cancer screening. In addition to having access to disposable income, the association between receiving a professional spray-on tan in the past 12 months and skin cancer screening suggests a general concern about the health and/or appearance of one’s skin among a subgroup of both ITs and NTs.

Limitations

Limitations of the current study include the narrow focus of the items (ie, only 1 or a few items per topic and limited response options) and their cross-sectional nature. The order of occurrence of behaviors such as indoor tanning, UV-free (sunless) tanning, and skin cancer screening was not explored in the current study but should be in future research. In addition, it is difficult to interpret associations between not going out in the sun and skin cancer screening, because skin protection is not necessary when indoors, and it is unknown whether individuals who do not go in the sun do so to protect their skin. However, only 5.09% to 8.08% of the sample reported not going out in the sun. Future research exploring the relationships among these behaviors in more depth could be useful.

Conclusions

Indoor tanners have a greater tendency to be screened for skin cancer and to be screened at an earlier age than NTs. However, skin cancer screening is still underused in this high-risk population, even among former ITs. The FDA recommends that individuals repeatedly exposed to UV radiation be screened regularly for skin cancer, but it is likely that ITs and their general practitioners are unaware of this recommendation, which should be disseminated more widely. The current data suggest that providing the public with greater access to health information and health care (eg, waiving copays) should be investigated as potential strategies to increase skin cancer screening rates, especially among high-risk groups, such as ITs. Interventions could target ITs who are younger than 40 years or with a household income of less than $100 000 per year, who are less likely to be screened than others. It is not surprising that skin cancer screening is associated with skin cancer risk factors such as family history of skin cancer and older age among ITs. However, some unscreened (compared with screened) ITs may be putting themselves at greater risk of skin cancer, for example, by also being more likely to use low-SPF sunscreen, thus offering multiple potential avenues for intervention.

Supplement.

eTable 1. Participant characteristics

eTable 2. Univariable and multivariable logistic regression analyses examining correlates of skin cancer screening by indoor tanning status