Coups EJ, Stapleton JL, Hudson SV, Medina-Forrester A, Rosenberg SA, Gordon MA, Natale-Pereira A, Goydos JS. Linguistic Acculturation and Skin Cancer–Related Behaviors Among Hispanics in the Southern and Western United States. JAMA Dermatol. 2013;149(6):679-686. doi:10.1001/jamadermatol.2013.745
Author Affiliations: The Cancer Institute of New Jersey, New Brunswick (Drs Coups, Stapleton, Hudson, and Goydos, Mss Medina-Forrester and Gordon, and Mr Rosenberg); Departments of Medicine (Drs Coups and Stapleton), Family Medicine and Community Health (Dr Hudson), and Surgery (Dr Goydos), University of Medicine and Dentistry of New Jersey (UMDNJ)–Robert Wood Johnson Medical School, New Brunswick; Department of Health Education and Behavioral Science, UMDNJ–School of Public Health, Piscataway (Drs Coups and Hudson); and Department of Medicine, UMDNJ–New Jersey Medical School, Newark (Dr Natale-Pereira).
Objective To examine the association between linguistic acculturation (assessed using the Language Use and Linguistic Preference subscales from the Bidimensional Acculturation Scale for Hispanics) and skin cancer–related behaviors among US Hispanic adults to determine whether, compared with Hispanics denoted as Spanish-acculturated, English-acculturated Hispanics would report less frequent shade seeking and use of sun protective clothing and higher rates of sunscreen use, sunbathing, and indoor tanning.
Design Online survey study conducted in September 2011.
Setting Five southern and western US states.
Participants A population-based sample of 788 Hispanic adults drawn from a nationally representative web panel.
Main Outcome Measures Self-reported sunscreen use, shade seeking, use of sun protective clothing, sunbathing, and indoor tanning.
Results Multivariate regression analyses were conducted to examine predictors of the skin cancer–related behaviors. As hypothesized, English-acculturated Hispanics had lower rates of shade seeking and use of sun protective clothing and reported higher rates of sunbathing and indoor tanning than Spanish-acculturated Hispanics. English-acculturated Hispanics and bicultural Hispanics (ie, those with high Spanish and high English acculturation) reported comparably high rates of sunbathing and indoor tanning. Results suggested that bicultural Hispanics seek shade and wear sun protective clothing less often than Spanish-acculturated Hispanics but more often than English-acculturated Hispanics. Acculturation was not associated with sunscreen use.
Conclusions Hispanic adults do not routinely engage in behaviors that reduce their risk of skin cancer. Bicultural and English-acculturated Hispanics are particularly in need of skin cancer prevention interventions.
Hispanics are the fastest growing racial/ethnic group in the United States. From 2000 to 2010, the Hispanic population in the United States increased by 43%, which constituted more than half of the national population increase during that time period.1 By 2050, it is estimated that the Hispanic population will exceed 100 million and represent 24.4% of the total population.2 The incidence of melanoma among Hispanics has been rising for several decades. From 1992 to 2008, the annual age-adjusted melanoma incidence among Hispanics increased by 19%, from 3.95 to 4.70 per 100 000.3 Although Hispanics have a lower incidence of melanoma than non-Hispanic white individuals, Hispanics are more likely to be diagnosed as having melanoma at an earlier age and with more advanced disease that is less amenable to successful treatment.4 Hispanics are an important population to target for skin cancer prevention and control efforts.5
A large proportion of Hispanics do not routinely engage in skin cancer prevention practices, such as using sunscreen, staying in the shade, wearing sun protective clothing, and refraining from UV indoor tanning. In a recent study, the percentage of Hispanics who reported often or always engaging in sun protection behaviors when outside on a sunny day was as follows: using sunscreen, 24%; staying in the shade, 54%; wearing a hat, 32%; wearing a long-sleeved shirt, 24%; and wearing long pants, 58%.6 Approximately 12% of 18- to 29-year-old Hispanics reported indoor tanning in the past year.7
Relatively little is known about factors that are associated with skin cancer–related behaviors among Hispanics. A particularly important factor to consider is acculturation, which refers to individuals' adoption of attitudes, norms, and behaviors from multiple cultures. Several studies have examined the association between acculturation and skin cancer–related behaviors among US Hispanics. A study of 496 Hispanic adults drawn from the 2005 Health Information National Trends Survey found that greater acculturation to US cultural norms was associated with more frequent sunscreen use, less frequent shade seeking, and less use of sun protective clothing.6 Using data from 1676 adult Hispanic participants in the 2010 National Health Interview Survey, we similarly found that greater acculturation to US norms was linked with more frequent sunscreen use and less frequent use of sun protective clothing (Coups et al8). In addition, more acculturated individuals were more likely to report having a sunburn in the past year. These prior studies shed valuable light on the link between acculturation and skin cancer–related behaviors among Hispanics but are subject to limitations—primarily with regard to the assessment of acculturation—that are addressed in the current study.
In line with the broader research literature on acculturation and health,9 studies of skin cancer–related behaviors among Hispanics have typically used a unidimensional approach to assess acculturation on a continuum from unacculturated to acculturated.10 In the current study, we used a bidimensional approach that assessed acculturation with regard to each of US and Hispanic cultures. This approach better reflects the complexity of acculturative processes and outcomes experienced by US Hispanics.9,11 Thus, we were able to categorize the study participants as being Spanish-acculturated (ie, high Spanish acculturation and low English acculturation), English-acculturated (ie, high English acculturation and low Spanish acculturation), or bicultural (ie, high Spanish and high English acculturation). We focused specifically on linguistic acculturation in the present study. Language use and preference have been used widely in prior studies of acculturation and health-related behaviors among Hispanics11,12 and have been found to be the best indicators of acculturation.13,14 We hypothesized that, compared with Hispanics denoted as Spanish-acculturated, English-acculturated Hispanics would have lower rates of shade seeking and use of sun protective clothing and higher rates of sunscreen use, sunbathing, and indoor tanning. Given the lack of relevant prior research, we did not specify a priori hypotheses regarding the rates of skin cancer–related behaviors among Hispanics denoted as bicultural. A greater understanding of the relationship between acculturation and skin cancer preventive and risk behaviors will shed light on the need for dermatologists and other practitioners to consider acculturation-related issues and processes when treating Hispanic individuals. As secondary research questions, we examined whether the prevalence of skin cancer–related behaviors among Hispanics varied according to their demographic characteristics (ie, sex, age, education level, latitude of residence, and Hispanic heritage) and objective risk for melanoma. Overall, the current study represents one of the most comprehensive examinations to date of acculturation and other correlates of skin cancer–related behaviors among US Hispanics.
The study participants were drawn from KnowledgePanel Latino, a nationally representative web panel of US Hispanic adults administered by the research company Knowledge Networks (www.knowledgenetworks.com). Hispanic individuals at least 18 years old are recruited to join KnowledgePanel Latino using a combined address-based sampling and random-digit dial method, which covers approximately 97% of Hispanic households in the United States. Panel members complete online surveys and are provided with a cost-free laptop and Internet access, if necessary. (Additional information about KnowledgePanel Latino is available elsewhere.15) Panel members residing in Arizona, California, Florida, New Mexico, or Texas were randomly selected and invited via e-mail to take part in the current study. We focused specifically on Hispanic individuals living in these 5 southern and western US states, because they have high percentages of Hispanic residents (varying from 22.5% to 46.3%, compared with a national average of 16.3%)16 and relatively high UV indexes. Individuals who reported never having been diagnosed as having skin cancer were eligible to complete the approximately 25-minute online survey. Study respondents received $5 for participating in the study. All recruitment and survey materials were available to participants in English or Spanish, based on their preference. Survey items were drawn from the skin cancer prevention literature17- 20; measures not already available in Spanish were translated professionally and further refined for plain language adaptation by several bilingual research staff members. Participant recruitment and survey completion occurred on September 14-26, 2011. Participants provided informed consent online before completing the survey. This study received institutional review board approval.
Of the 1717 individuals invited to take part in the survey, 25 were ineligible because they reported a personal history of skin cancer, 904 declined to participate, and 788 completed the survey (52.3% in English, 47.7% in Spanish), giving a completion rate of 46.6%. We compared the (unweighted) demographic characteristics of the study participants and the decliners using independent samples t tests and χ2 tests. The study participants were slightly older than the decliners (mean age, 42.3 years vs 38.0 years; P < .001), included a higher percentage of men (49.9% vs 44.8%; P = .04), and had a higher level of education (15.2% vs 9.7% completed at least a Bachelor's degree; P = .002). The study participants and decliners did not differ with regard to their state (P = .55) or latitude of residence (P = .56). As described in the “Statistical Analysis” subsection, the statistical analyses we conducted were weighted to adjust for demographic differences between study participants and decliners.
The English and Spanish language surveys used in this study are available from the first author (E.J.C.).
Participants reported their sex, age, level of education, and Hispanic heritage. The latitude of residence (ie, degrees north of the equator) was denoted for each participant based on the home address.
Participants completed questions regarding 8 risk factors for melanoma. Drawing on prior research,17,19 we categorized individuals as to whether they had each of the risk factors as follows: had naturally red or blonde hair; had blue, green, or gray eyes; had at least a few freckles; had very fair or fair untanned skin color; would get a severe or moderate sunburn if exposed to midday summer sun without protection; had had a severe sunburn with blisters; had at least 1 mole larger than a pencil eraser (about 0.6 cm); or had at least 1 first-degree relative diagnosed as having melanoma. We calculated the total number of melanoma risk factors (0-8) for each participant.
Participants completed the 6-item Language Use (4-point response scale from 1 [almost never] to 4 [almost always]) and 12-item Linguistic Proficiency (4-point response scale from 1 [very poorly] to 4 [very well]) subscales from the Bidimensional Acculturation Scale for Hispanics.21 The Language Use subscale assesses how often individuals used English and Spanish languages, and the Linguistic Proficiency subscale assessed their ability to use English and Spanish. We created separate indexes of English and Spanish acculturation for each participant by averaging responses to the English-oriented (eg, “How often do you speak English?”; “How well do you speak English?”) (α = .98) and Spanish-oriented (eg, “How often do you speak Spanish?”; “How well do you speak Spanish?”) items (α = .97), respectively. Following established guidelines,21 we used a cutoff of 2.5 on both the English and Spanish indexes to determine low vs high acculturation and assigned each participant to 1 of 3 groups: Spanish-acculturated (high Spanish acculturation, low English acculturation); English-acculturated (high English acculturation, low Spanish acculturation); or bicultural (high Spanish acculturation, high English acculturation). Three participants had scores of 2.5 or lower on both the English and Spanish indexes and were assigned to an acculturation group based on their language of survey completion.
Participants completed recommended survey items regarding multiple skin cancer preventive (sunscreen use, shade seeking, and use of sun protective clothing) and risk behaviors (sunbathing and indoor tanning).18,20 With regard to skin cancer preventive behaviors, participants used a 5-point response scale (1 [never] to 5 [always]) to indicate how often they use sunscreen (1 item), stay in the shade (1 item), and wear sun protective clothing (3 items: long-sleeved shirt, long pants or other clothing that reaches the ankles, wide-brimmed hat) when outside on a warm sunny day.18 Responses to the 3 clothing items were averaged to create a sun protective clothing index (α = .63). Individuals who reported any use of sunscreen indicated the sun protection factor (SPF) of the sunscreen they used most often as well as the usual timing of their sunscreen application. With regard to skin cancer risk behaviors, single items assessed how often participants sunbathed (using a 5-point response scale from never to always)18 and whether they had ever engaged in UV indoor tanning.20 For analytic purposes, we coded the sunbathing variable according to whether individuals ever sunbathed.
We conducted an analysis of variance to examine whether the number of melanoma risk factors differed according to individuals' linguistic acculturation. We used correlation coefficients to examine the associations among the 5 skin cancer preventive and risk behavior variables. For each of the behaviors, we conducted a separate multivariate linear regression analysis (for the preventive behaviors of sunscreen, shade seeking, and sun protective clothing) or logistic regression analysis (for the risk behaviors of sunbathing and indoor tanning). Demographic factors, the melanoma risk factor score, and the linguistic acculturation variable were included as independent variables in each regression analysis. For all of the statistical analyses, the data were weighted using a variable that adjusted for multiple factors, including the probability of panel selection, Spanish language use, and potential poststratification nonresponse and noncoverage biases in both the overall panel (ie, before the sample for the current study was selected) and in the study sample. The poststratification adjustment variables included age, sex, education level, state of residence, metropolitan area, Internet access, and primary language by census region. (Additional information regarding the statistical weighting is available elsewhere.22) For all analyses, a cutoff of P < .05 was used to determine statistical significance.
Descriptive statistics for the study variables are shown in Table 1. Half of the study participants were female, and 11.5% had a college degree. The most common states of residence were California and Texas, and 70.9% of participants reported being of Mexican heritage. The mean latitude of 32.2° is the approximate latitude of San Diego (California), Dallas (Texas), and Tucson (Arizona). Participants had an average of 2.0 melanoma risk factors, and almost one-third of the sample had 3 or more risk factors. The most commonly reported risk factors were: ever had a severe sunburn with blisters, 45.3%; would get a severe or moderate sunburn if exposed to midday summer sun without protection, 42.5% (specifically, participants reported the following: severe sunburn, 15.2%; moderate sunburn, 27.4%; mild sunburn, 22.6%; turn darker without sunburn, 22.8%; nothing would happen to the skin, 12.1%); have at least 1 mole larger than a pencil eraser (about 0.6 cm), 36.7%; and have very fair or fair untanned skin, 35.9% (specifically, participants reported their untanned skin color as follows: very fair, 6.0%; fair, 29.9%; olive, 13.9%; light brown, 42.9%; dark brown, 7.1%; very dark, 0.2%). With regard to linguistic acculturation, 35.6% of the participants were denoted as Spanish-acculturated, 44.9% were denoted as bicultural, and 19.5% were English-acculturated. Individuals' number of melanoma risk factors differed according to their linguistic acculturation (F = 4.66; P = .01). English-acculturated Hispanics had significantly more risk factors (mean, 2.2) than those who were Spanish-acculturated (mean, 1.8). Bicultural Hispanics had an average of 2.0 melanoma risk factors. On a 1 to 5 scale, the mean values for sunscreen use, shade seeking, and use of sun protective clothing were 2.8, 3.4, and 2.8, respectively. Among individuals who reported using sunscreen, 71.6% indicated that they usually used a sunscreen with an SPF of at least 15, although 22.6% of participants did not know the SPF of their sunscreen. In terms of the timing of sunscreen application, 79.4% of participants reported applying sunscreen before going out in the sun, with almost all other individuals indicating that they apply it as soon as they get in the sun. Sunbathing was reported by 39.4% of participants, and 5.3% reported ever indoor tanning.
The correlations among the skin cancer–related behaviors are shown in Table 2. There were small to moderate positive correlations among the 3 skin cancer preventive behaviors (sunscreen use, shade seeking, and use of sun protective clothing). There was a small positive correlation between sunbathing and indoor tanning. Sunscreen use was not significantly associated with sunbathing or indoor tanning. Shade seeking and use of sun protective clothing each had small inverse associations with sunbathing and indoor tanning.
Results of the linear regression analyses examining correlates of the skin cancer protective behaviors are shown in Table 3. Factors associated with less frequent sunscreen use included being male, having a lower education level, residing at a lower latitude (ie, closer to the equator), and having fewer melanoma risk factors. Age, Hispanic heritage, and linguistic acculturation were not associated with sunscreen use. The frequency of shade seeking was lower among men, younger individuals, those reporting “other” Hispanic heritage compared with those of Mexican heritage, individuals with fewer melanoma risk factors, and English-acculturated Hispanics compared with those denoted as Spanish-acculturated or bicultural. Shade seeking was not associated with education level or latitude of residence. Factors associated withless frequent use of sun protective clothing included being female, younger age, having a higher level of education, residing at a lower altitude (ie, closer to the equator), having fewer melanoma risk factors, and English acculturation (compared with both Spanish and bicultural acculturation). Hispanic heritage was not associated with the use of sun protective clothing.
Results of the logistic regression analyses examining correlates of the skin cancer risk behaviors are shown in Table 4. Sunbathing was more commonly reported by younger individuals, those of Puerto Rican, Cuban, South American, or “other” Hispanic heritage compared with those of Mexican heritage, and among bicultural and English-acculturated Hispanics compared with those denoted as Spanish-acculturated. Sex, level of education, latitude, and the number of melanoma risk factors were not associated with sunbathing. Indoor tanning was more prevalent among women, individuals of Cuban or “other” Hispanic heritage compared with those of Mexican heritage, individuals with more melanoma risk factors, and among bicultural and English-acculturated Hispanics compared with those denoted as Spanish-acculturated.
In this study, we examined the prevalence and correlates of skin cancer protective and risk behaviors in a population-based sample of adult Hispanic individuals living in southern and western US states. There was considerable variation with regard to the prevalence of the skin cancer–related behaviors among the study participants. With regard to the skin cancer protective behaviors, more participants reported routinely (ie, often or always) staying in the shade (53.2%) compared with using sunscreen (30.9%) and wearing sun protective clothing (24.4%). This is consistent with prior findings among US Hispanics6 (Coups et al8) as well as US adults in general.23 In terms of the skin cancer risk behaviors, 39.4% of participants reported sunbathing, but only 5.3% had ever tanned indoors. These findings shed light on the skin cancer–related behaviors for which there is the most need for intervention among Hispanic individuals.
The primary objective of the current study was to examine the association between linguistic acculturation and skin cancer preventive and risk behaviors among Hispanic adults. Consistent with our hypotheses, compared with Hispanics denoted as Spanish-acculturated, English-acculturated Hispanics reported being less likely to seek shade and wear sun protective clothing and were more likely to report sunbathing and indoor tanning. Contrary to our expectations, linguistic acculturation was not significantly associated with sunscreen use. Prior studies have found acculturation among Hispanics to be associated with sunscreen use in univariate but not multivariate analyses6 (Coups et al8). Future research is warranted to further examine the association between acculturation and sunscreen use among Hispanics.
Of particular importance in the current study, our use of a bidimensional measure of acculturation allowed us to compare the skin cancer–related behaviors of Hispanics denoted as bicultural with those of persons designated as English-acculturated or Spanish-acculturated only. An interesting pattern of results emerged in that regard. In terms of preventive behaviors, results suggested that bicultural Hispanics seek shade and wear sun protective clothing less often than Spanish-only–acculturated Hispanics but more often than English-only–acculturated Hispanics. However, with regard to risk behaviors, bicultural Hispanics engaged in sunbathing and indoor tanning at high rates comparable with those of English-only–acculturated Hispanics. Overall, the picture that emerges is that high English acculturation among Hispanics—regardless of individuals' level of Spanish acculturation—is associated with greater engagement in skin cancer risk behaviors. However, the relatively low rates of shade seeking and wearing sun protective clothing among highly English-acculturated Hispanics are somewhat offset among bicultural Hispanics.
A number of demographic factors were significantly associated with the skin cancer–related behaviors, and these results were largely consistent with those of prior research among Hispanics6 (Coups et al8) as well as the general US population.7,23,24 Hispanic men had lower rates of indoor tanning and reported using sunscreen and seeking shade less often than women but were more likely to report wearing sun protective clothing. Younger Hispanics reported seeking shade and wearing sun protective clothing less often than older individuals and were also more likely to report sunbathing. Participants' level of education was associated positively with sunscreen use and inversely with wearing sun protective clothing. There was some indication that Hispanics of Mexican heritage were more likely to seek shade and less likely to sunbathe and tan indoors compared with individuals with heritages from several other countries. However, further research is needed to examine potential differences in skin cancer–related behaviors, attitudes, and beliefs among individuals of varying Hispanic origins. It is encouraging that Hispanics with a greater number of objective melanoma risk factors reported more frequently engaging in skin cancer preventive behaviors. However, it is a cause for concern that indoor tanning was more commonly reported among Hispanics with more melanoma risk factors. Interventions to discourage indoor tanning among Hispanics should target individuals with a greater number of risk factors for melanoma. Overall, the study results highlight the importance of developing culturally appropriate, tailored interventions to reduce the risks of skin cancer among Hispanic individuals of varying backgrounds and levels of acculturation. Future research is also needed to examine skin cancer–related behaviors among particular subgroups of Hispanics, such as those who work outdoors in the sun.
There are several limitations to the current study. Most participants reported being of Mexican heritage, which, while consistent with national statistics,2 limits the conclusions that can be drawn regarding associations between Hispanic heritage and skin cancer–related behaviors. The cross-sectional design of the study does not provide insight on the timing and nature of acculturative processes that influence Hispanic individuals' engagement in skin cancer–related behaviors. Future research is needed to examine longitudinal changes in skin cancer–related behaviors that may be influenced by acculturation. We identified differences in the demographic characteristics of the recruited participants and those who declined to participate. However, the potential effects of sampling bias are attenuated by our use of statistical weights that controlled for multiple factors, including sample nonresponse. We used a cutoff of P < .05 to determine statistical significance for all analyses. Thus, the family-wise α rate across the entire study is greater than .05.
The results of the present study indicate that Hispanic adults in the United States do not sufficiently engage in behaviors that reduce their risk of developing skin cancer. Clinicians should prioritize discussions of skin cancer prevention and risk behaviors among their Hispanic patients who are bicultural or English-acculturated. These individuals are more likely than Spanish-acculturated Hispanics to be at risk for skin cancer based on their objective melanoma risk factor profile and their engagement in skin cancer–related behaviors. Many bicultural Hispanics may be receptive to interventions in English or Spanish,25 although individuals' preferred language may vary according to the intervention modality (eg, in person, written materials, computer-based). Of additional relevance is the need to deliver interventions in a culturally proficient manner.25 This highlights the need to include issues related to culture and race/ethnicity in dermatology training programs, as well as the importance of racial/ethnic diversity in the dermatology workforce.5,26 Most sunscreen users in the current study indicated that they used sunscreen with at least SPF 15 and applied it before going out in the sun, which is in line with recommended practices. However, there is still considerable room for clinicians to educate Hispanic patients regarding sunscreen use because more than one-third (43.2%) of participants reported using sunscreen never or rarely, and almost one-quarter (22.6%) of the sunscreen users did not know the SPF of their sunscreen. The relatively low correlations observed among the skin cancer–related behaviors suggest that the engagement in a specific skin cancer protective or risk behavior by a Hispanic individual is not necessarily indicative of whether he or she also engages in other such behaviors. Dermatologists, public health practitioners, and researchers should consider this issue with regard to skin cancer prevention initiatives targeting Hispanics and seek to promote a wide range of protective behaviors. Future studies should further examine associations between acculturative processes and outcomes and skin cancer–related behaviors, attitudes, and beliefs among Hispanics. Additional research is also warranted to develop, test, and disseminate culturally appropriate behavioral interventions to reduce the risks of skin cancer among the rapidly growing US Hispanic population.
Accepted for Publication: July 3, 2012.
Published Online: April 17, 2013. doi:10.1001/jamadermatol.2013.745
Correspondence: Elliot J. Coups, PhD, The Cancer Institute of New Jersey, 195 Little Albany St, Room 5567, New Brunswick, NJ 08901 (firstname.lastname@example.org).
Author Contributions: Dr Coups had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Coups, Stapleton, Hudson, Medina-Forrester, Rosenberg, Gordon, Natale-Pereira, and Goydos. Acquisition of data: Coups. Analysis and interpretation of data: Coups, Hudson, Medina-Forrester, Rosenberg, Gordon, Natale-Pereira, and Goydos. Drafting of the manuscript: Coups. Critical revision of the manuscript for important intellectual content: Coups, Stapleton, Hudson, Medina-Forrester, Rosenberg, Gordon, Natale-Pereira, and Goydos. Statistical analysis: Coups. Obtained funding: Coups.
Conflict of Interest Disclosures: None reported.
Funding/Support: This research was supported by a Cancer Prevention and Control Pilot Award from The Cancer Institute of New Jersey (Dr Coups) and by National Cancer Institute grants K07CA133100 (Dr Coups) and K01CA131500 (Dr Hudson).
Role of the Sponsors: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.
Additional Contributions: Kristina Tatum, MS, provided comments on an earlier draft of this manuscript, Ciara Rivera, BA, helped with study logistics and translation of survey items, and Knowledge Networks assisted with data collection.