University of Southern California Dissertations and Theses

Psychosocial and cultural factors in the primary prevention of melanoma targeted to multiethnic children

(USC Thesis Other)

Psychosocial and cultural factors in the primary prevention of melanoma targeted to multiethnic children

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

PSYCHOSOCIAL AND CULTURAL FACTORS IN THE PRIMARY PREVENTION OF
MELANOMA TARGETED TO MULTIETHNIC CHILDREN

BY
Kimberly A. Miller, MPH
____________________________________________________________

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PREVENTIVE MEDICINE: HEALTH BEHAVIOR RESEARCH)

December 2015

Copyright 2015 Kimberly A. Miller

ii
DEDICATION
To Beatrix, age 12,
who grew up with this research
and who always wears sunscreen.

iii
ACKNOWLEDGMENTS
I offer sincere thanks to the members of my committee, Drs. Myles Cockburn, Jimi Huh,
David Peng, Jean Richardson, and Jennifer Unger for their valuable guidance and support
throughout this dissertation. I am particularly grateful to Jean Richardson for her mentorship and
for the example of her own research, and to Myles Cockburn who encouraged me to pursue my
doctorate and in doing so, changed the course of my career. I owe them both a great professional
and personal debt.
I thank the USC Department of Preventive Medicine and the division of Health Behavior
Research for being an inspiring and supportive place to study. Thank you Marny Barovich,
tireless advocate for doctoral students, who has offered critical assistance at significant moments.
For their professional and personal support, I thank my USC colleagues Marlene Caldera and
Letech Caldera-Huerta along with fellow HBR students Lauren Martinez, Stephanie Dyal,
Myriam Forster, Tim Grigsby, Gillian O'Reilly Gentner, Eleanor Tate Shonkoff, and Cheng
Wen. I owe much to the members of the "Cockburn Lab Manuscript Support Group (MSG),"
especially Loraine Escobedo, Jennifer Truong, Kathy Wocjik, Ron Stewart, Lissette Ramirez,
Carrie Tayour and Amanda Goodrich, who provided a friendly forum to try out papers,
presentations, and ideas.
Special thanks to Susan Harris and Jeremy Myers for being extended NELA family and
driving morning carpool going on seven years. Friends and family members too numerous to
name provided kind support and encouragement throughout this research, but the friendship of
Joewon Yoon and Sarah Sullivan has been particularly important to me.

iv
I thank my mother Ann Miller, and remember my father David Miller, for their love and
support. Finally, my greatest love and gratitude goes to my husband Randolph Heard, daughter
Beatrix Heard, and dog Cookie, who give me the gift of a happy home.

v
TABLE OF CONTENTS
Dedication ii
Acknowledgments iii
List of Tables vii
List of Figures viii
Abbreviations ix
Abstract xi
Chapter 1: Introduction 1
Background and significance 1
Childhood UV exposure as a melanoma risk factor 5
Theoretical foundations and correlates of sun protection in children 8
Interventions to reduce UV exposure among children and adolescents 14
Limitations of research regarding skin cancer prevention and sun
protection in children 15
Introduction to the dissertation studies 16
Research hypotheses 18

Chapter 2: Methods 21
Data source and participants 21
Selection bias and generalizability among the studies 23
Measures 27
Statistical analysis for studies in this dissertation 29

Chapter 3: Study1 32
Correlates of sun protection behaviors in a sample of ethnic
minority children residing in a high UV environment
Abstract 33
Introduction 35
Methods 39
Results 44
Discussion 51
Conclusion 56

vi
Chapter 4: Study 2 57
Mediators of sunscreen use in a skin cancer prevention intervention among multi-ethnic
elementary schoolchildren
Abstract 58
Introduction 60
Methods 65
Results 69
Discussion 73
Conclusion 75

Chapter 5: Study 3 78
Patterns of sun protective behaviors among Hispanic children in a skin cancer prevention
intervention
Abstract 79
Introduction 80
Methods 82
Results 85
Discussion 91
Conclusion 94

Chapter 6: Discussion 96
Overall findings 96
Limitations and strengths of the current research 101
Implications for future research and public health impact 103

Bibliography 106

Appendix
Table of Intraclass Correlations 120

vii
LIST OF TABLES
Table 2-1 Sample baseline characteristics and between-group differences in Los
Angeles 4th and 5th graders participating in a skin cancer prevention program (SunSmart) .......26
Table 2-2 Selected measures used to evaluate a skin cancer prevention program in
Los Angeles 4th and 5th graders (SunSmart) ................................................................................28
Table 3-1 Demographic characteristics of sample at baseline, SunSmart .....................................45
Table 3-2 Prevalence of sun protective behaviors at baseline by race/ethnicity, SunSmart ..........46
Table 3-3 Prevalence of sunburn-related variables at baseline by race/ethnicity, SunSmart .......47
Table 3-4 Multivariate linear regression model for correlates of sun protective
behaviors among elementary schoolchildren, SunSmart ..............................................................48
Table 3-5 Bivariate associations between acculturation and sun protection among
the Hispanic subsample, SunSmart ...............................................................................................50
Table 3-6 Path coefficients and indirect effects for mediating variables between
acculturation and shade-seeking among the Hispanic subsample, SunSmart ................................50
Table 4-1 Demographic characteristics of sample at baseline, SunSmart .....................................70
Table 4-2 Coefficients and indirect effects with bias-corrected bootstrap confidence
intervals for change scores, Pre to Posttest ....................................................................................71
Table 4-3 Fit statistics for cross-lagged autoregressive models of mediators
of sunscreen use .............................................................................................................................72
Table 4-4 Coefficients and indirect effects with bias-corrected bootstrap confidence
intervals, cross-lagged analysis ......................................................................................................73
Table 5-1 Characteristics of sample at baseline, SunSmart ...........................................................86
Table 5-2 Model-fit indices for a latent class analysis of sun protective behaviors among
Hispanic schoolchildren, SunSmart ...............................................................................................87
Table 5-3 Four-latent-class model of sun protection behaviors: probabilities of
engagement for each subgroup, SunSmart .....................................................................................88
Table 5-4 Logit estimates and odds ratios for predictors of latent class
membership, Sunsmart ...................................................................................................................90

viii
LIST OF FIGURES
Figure 1-1 Theoretical Model for the Research .............................................................................18
Figure 2-1 SunSmart Intervention Design ....................................................................................22
Figure 2-2 Flow diagram of SunSmart RCT ..................................................................................23
Figure 4-1 Cross-lagged mediation model, SunSmart ...................................................................69
Figure 5-1 LCA among Hispanic schoolchildren, SunSmart ........................................................89

ix
ABBREVIATIONS
AHIMSA: Acculturation, Habits, and Interests Multicultural Scale for Adolescents
AIC: Akaike Information Criterion
ANOVA: One-way analysis of variance
API: Academic performance index
BIC: Bayesian Information Criterion
BRAF: Serine/threonine-protein kinase B-Raf oncogene
CDC: Centers for Disease Control and Prevention
CDKN2A: Cyclin-dependent kinase inhibitor 2A oncogene
CFI: Comparative fit index
FIML: Full-information maximum likelihood estimation
ICC: Intraclass correlation coefficients
IRB: Institutional review board
LCA: Latent class analysis
LMR: Lo-Mendell-Rubin likelihood test
MAR: Missing at random
MC1R: Melanocortin 1 receptor
MCAR: Missing completely at random
NHW: Non-Hispanic white
NMSC: Non-melanoma skin cancer
RMSEA: Root mean square error of approximation
SPF: Sun protection factor
UV: Ultraviolet

x
WHO: World Health Organization

xi
ABSTRACT
Malignant melanoma is the most serious form of skin cancer, accounting for more than
75% of all skin cancer deaths. In contrast with decreasing trends for most major cancer sites,
incidence rates of melanoma have been rising for the past several decades globally. Because
childhood sun exposure and sunburn have been associated with significantly increased risk of
melanoma in adulthood, children are particularly important targets for melanoma primary
prevention, which includes the reduction of solar overexposure through methods such as limiting
sun during peak hours, using sunscreen with adequate sun protection factor, using protective
clothing including long sleeves, long pants, and hats, and seeking shade.
This dissertation examines determinants of sun protection behaviors in a multiethnic
sample of children participating in a school-based skin cancer prevention intervention. To date,
interventions targeting children have been largely unsuccessful in improving UV-related
behaviors. In addition, data regarding the sun protective behaviors of non-white children are
scarce, and few skin cancer prevention interventions have been conducted amongst ethnically
diverse samples despite rising rates of melanoma in non-white populations, particularly among
U.S. Hispanics.
Findings from our studies indicate that ethnic minority children experience similar sun
protection correlates previously observed among non-Hispanic white (NHW) children. Strong
positive associations were found between parental variables and child sun protection across
studies. Mediation analysis between the intervention and sunscreen use found short-term effects
for psychosocial variables; however, only knowledge of sun protection was found to mediate
program effects over longer-term follow-up. Finally, latent class analysis revealed complex
patterns of sun protection behaviors for Hispanic children in the sample, with strata of

xii
increasingly risky UV behaviors. Children who reported high parental engagement with sun
protection were significantly more likely to be classified in high protective categories, as were
girls.
The studies presented here suggest considerable overlap between variables associated
with greater sun protection for NHW and multiethnic youth, as well as patterns that may be
unique to ethnic minority children. The results of this research will inform the design and
delivery of future interventions as well as aid in the development of more effective and tailored
strategies to improve sun protection practices in racial and ethnic minority groups.

1
CHAPTER 1: INTRODUCTION
Background & Significance
Etiology and epidemiology of melanoma. Malignant cutaneous melanoma is the most
serious form of skin cancer, accounting for more than 75% of all skin cancer deaths (Tarver,
2012). In the United States, melanoma incidence for non-Hispanic white (NHW) males is 24.1
per 100,000 and 15.3 per 100,00 for NHW females (U.S. Cancer Statistics Working Group,
2014). Data from National Cancer Institute’s Surveillance, Epidemiology, and End Results
(SEER) indicate that in 2015, approximately 73,800 new cases of melanoma will be diagnosed
and 9,940 people will die from the disease (National Cancer Institute). Five-year survival rates
for localized disease are high at 94%, but drop sharply to 17% for metastatic disease (National
Cancer Institute).
In contrast with decreasing trends for most major cancer sites, incidence rates of
melanoma have been rising for the past several decades globally (International Agency for
Research on Cancer, 2010). Melanoma is currently the 5th most common cancer in NHW men
and the 6th most common for NHW women, increasing by 1.9% annually for males and females
(U.S. Cancer Statistics Working Group, 2014). These increases are not simply attributable to
improved screening, as incidence rates have increased not only in thinner tumors which are more
likely to be detected through screening, but in thicker tumors, indicating a true increase in
disease (Jemal et al., 2011; Linos, Swetter, Cockburn, Colditz, & Clarke, 2009).
The etiology of cutaneous melanoma is multifactorial and includes genetic and
environmental causative factors. Phenotypic risk factors include fair skin, blue eyes, blonde or
red hair, freckles, presence of many, large, and or atypical moles, and skin that burns and does
not tan. These traits are in part determined by melanocortin 1 receptor (MC1R) which regulates
melanin synthesis and controls the type of melanin produced (Nelson & Tsao, 2009). Natural

2
variation in MC1R has been associated with the development of both melanoma and non-
melanoma skin cancers (Kanetsky et al., 2006). In addition, mutations in CDKN2A, a tumor-
suppressor gene, are responsible for the majority of familial melanoma risk; presence of MC1R
variants can also increase risk in individuals with CDKN2A mutations (Nelson & Tsao, 2009).
Mutations in the BRAF oncogene, which codes for a protein that regulates cell growth, have
been found in approximately half of melanomas (Davies et al., 2002; Maldonado et al., 2003).
Ultraviolet (UV) exposure is the primary environmental risk factor for melanoma in both
genetically susceptible and average risk individuals. Numerous studies have shown an
association between excessive UV exposure and increased risk for melanoma, with type of
exposure affecting risk of disease (Elwood & Jopson, 1997). A recent comprehensive meta-
analysis of UV exposure in adults reported pooled estimates for intermittent exposure, defined as
an intermittent pattern of relatively severe sun exposure, indicating a substantial risk of
melanoma for those with high intermittent exposures (RR=1.61; 95% CI: 1.31, 1.99) (Gandini et
al., 2005). In the same meta-analysis, chronic exposures indicated an inverse relationship
between high exposure and melanoma (RR=0.95; 95% CI: 0.87, 1.04), presumably as such
exposure promotes tanning thus increasing melanin and conferring some protective effect. For
individuals with a history of severe sunburn, pooled estimates revealed a substantial risk of
disease (RR = 2.03; 95% CI: 1.73) (Gandini et al., 2005). Another comprehensive meta-analysis
by Dennis et al. found strong associations between risk of melanoma and repeated sunburn at all
life-periods, with increased risk of disease associated with higher number of sunburns (Dennis et
al., 2008).

3
Non-melanoma skin cancer. The primary focus of the present research is on prevention
of melanoma skin cancer due to its severity and high mortality. However, non-melanoma skin
cancer (NMSC) is the most common form of malignancy in NHWs. While infrequently fatal,
NMSC causes significant morbidity and economic burden (Lomas, Leonardi-Bee, & Bath-
Hextall, 2012). While etiologies differ between melanoma and NMSC, thee are similarities with
respect to risk factors, and excessive exposure to UV radiation is the primary environmental risk
factor for both types of skin cancer (Gordon, 2013; Lomas et al., 2012). As with melanoma,
excessive UV exposure in childhood is associated with greater risk of NMSC (Stern, Weinstein,
& Baker, 1986). Further, like melanoma, the incidence of NMSC is increasing substantially
worldwide (Rogers, Weinstock, Feldman, & Coldiron, 2015); emerging research suggest that
NMSC is also rising among non-white populations such as U.S. Hispanics (Ortiz).
Prevention of both NMSC and melanoma are significant public health problems. Thus,
while the studies presented here focus primarily on melanoma prevention due to the increasing
burden of this aggressive and lethal disease, they apply equally to the reduction of incidence of
NMSC, which is currently epidemic in the U.S. population.

Melanoma in non-white populations. Melanoma is a relatively uncommon disease
among darker-skinned and ethnic minority populations, with age-adjusted incidence rates per
100,000 in the U.S. for males and females (respectively) of 1.3 and 1.05 in African-Americans,
1.6 and 1.3 in Asian/Pacific Islanders, and 4.8 and 4.8 in Hispanics (Park et al., 2012) versus
24.1 and 15.3 in NHW males and females, respectively. However, time trends show rising rates
of invasive melanoma in Hispanics, who have the second highest incidence of melanoma after
NHWs in the U.S. (Cockburn, Zadnick, & Deapen, 2006). In California and Florida, states with

4
highest proportions of Hispanics as well as high UV levels, rates of the thickest tumors have
increased markedly over the past 30 years in Hispanic populations (Hu et al., 2009).
Racial and ethnic minorities are diagnosed at later stages with more advanced disease and
have lower survival than NHWs (Pollitt et al., 2011). While it is not certain why such disparities
exist, factors influencing diagnosis at advanced stage in non-whites may include lower
awareness of melanoma leading to delayed diagnosis (Coups et al., 2014; Pichon, Corral,
Landrine, Mayer, & Adams-Simms, 2010) and socioeconomic factors such as disparities in
health care access or uninsured status (Pollitt et al., 2011). In addition, a greater proportion of
aggressive subtypes such as nodular and acral lentiginous melanomas in ethnic minorities may
contribute to survival disparities, as these subtypes have lower survival rates than superficial
spreading melanomas (Cormier et al., 2006; Cress & Holly, 1997).
Recent studies have shown similar and overlapping risk factors for certain non-white
ethnic groups such as Hispanics, Asians, and NHWs including male sex, older age, and sun-
sensitive phototype (Park et al., 2012). In addition, several studies investigating the role of sun
exposure on melanoma incidence in non-white populations have found positive associations,
indicating that UV is an etiologic factor in the development of disease in ethnic minority
populations (Hu, Ma, Collado-Mesa, & Kirsner, 2004; Pennello, Devesa, & Gail, 1999).
However, evidence regarding the role of UV exposure and melanoma in ethnic minority
populations is still unclear, as some studies have presented contrasting results, finding null
association between increased UV and melanoma incidence among non-white racial and ethnic
groups (Eide & Weinstock, 2005). However, these studies have been primarily ecological and
stronger study designs are needed to determine the actual relation between UV exposure and
melanoma in darker-skinned individuals.

5
Thus, despite limited evidence, based on rising rates of invasive melanoma among
Hispanics and disparities in disease profiles across racial and ethnic minority groups, researchers
have called for greater focus on primary and secondary prevention messages targeted to Hispanic
communities and other non-white racial and ethnic groups (Coups et al., 2012; C. Hernandez,
Calero, Robinson, Mermelstein, & Robinson, 2012; Pichon, Corral, Landrine, Mayer, &
Norman, 2010). Further, shifts in U.S. demographics that are resulting in the rapid increase in the
proportion of minorities in the population heightens the public health importance of the issue, as
the burden of disease may be considerable given the population magnitude in absolute terms.

Childhood UV exposure as a melanoma risk factor
Melanoma risk and childhood UV exposure. Across studies there has been a
consistent association between high sun exposure and sunburn in childhood with increased risk
of melanoma in adulthood (D. C. Whiteman, Whiteman, & Green, 2001). Although the exact
causal mechanism for childhood sun exposure as a period of susceptibility for adult melanoma is
not known, children’s skin may be more susceptible to the carcinogenic effects of UV exposure,
and/or UV exposure may confer risk by increasing the number of nevi (moles) (Green,
Wallingford, & McBride, 2011). One comprehensive meta-analysis found strong ecological
evidence that adults raised in high UV environments who experienced high levels of exposure in
childhood were at greater risk for development of melanoma in adulthood (D. C. Whiteman et
al., 2001). In case-control studies, early life UV exposure, independently and in tandem with
high adult exposures, has been found to be associated with a substantially increased risk of
melanoma in adulthood (Gandini et al., 2005; Green et al., 2011; D. C. Whiteman et al., 2001).
In a comprehensive meta-analysis, Gandini et al. reported a pooled estimate of RR =1.99 (95%

6
CI: 1.45, 2.7) for risk of melanoma associated with childhood sunburn (with childhood defined
as age less than 15 years old) (Gandini et al., 2005). Research is also emerging that cumulative
sun exposure and severe sunburn in childhood is associated with early onset of melanoma in
adulthood, before age 40 (Cust et al., 2011).
Thus, while excessive UV exposure throughout an individual’s entire lifetime may
increase the risk of melanoma, childhood and adolescence presents a critical time period for
melanoma prevention. Protecting youth from the carcinogenic effects of solar overexposure and
instilling sun protective habits at an early age may reduce lifetime risk of melanoma.

Sun protection and exposure behaviors in children. Children and adolescents are
subject to high levels of UV exposure and sunburn. Nearly 69% of adolescents from all races
reported sunburn during the summer in a nationally representative U.S. sample (Buller,
Cokkinides, et al., 2011). Broken down by race and ethnicity, in youths aged 14-17 years,
approximately half of NHWs had a sunburn during the preceding 12 months, followed by 22% of
Hispanics, 18% of Asians, and 7% of African-Americans (Centers for Disease Control and
Prevention (CDC), 1998). In another U.S. sample, 64% percent of parents of NHW children
under the age of 12 reported that their child spent more than one hour outdoors on sunny days
during peak hours (Hall, Jorgensen, McDavid, Kraft, & Breslow, 2001). This figure was 19% for
non-white children.
Children also practice low and often inadequate levels of sun protective behaviors. In a
study of more than 25,000 NHW children, only 20% reported “always” using some method of
sun protection, while 56% reported “sometimes” using sun protection (Coogan, Geller, Adams,
Benjes, & Koh, 2001). Twenty percent stated that they did not need sun protection.

7
Sunscreen is the most frequently cited method of sun protection for NHW children, with
approximately 60% of children under 12 reporting any sunscreen use (Centers for Disease
Control and Prevention (CDC), 1998; Hall, Jorgensen, et al., 2001). For older adolescents,
approximately 1 in 10 high school students report routine use of sunscreen with sun protection
factor (SPF) 15 or higher (Buller, Cokkinides, et al., 2011), with slightly higher figures reported
for sunscreen use at the beach or pool (Cokkinides, Weinstock, Cardinez, & O'Connel, 2004).
Shade is the next most frequently reported method of sun protection, with approximately
22% of NHW children under the age of 11 and 27% of children ages 11-17 seeking shade when
outdoors for more than one hour on a sunny day (Cokkinides et al., 2001; Hall, Jorgensen, et al.,
2001). Lowest prevalence is reported for use of sun protective clothing, with only 9% of NHW
children under the age of 11 wearing long sleeves or long pants “always” or “often” when
outside on a sunny day, and only 4% wearing sun protective hats (Hall, Jorgensen, et al., 2001).
In adolescents age 11 to 17, 23% reported regularly wearing some type of protective clothing
when outdoors on a sunny day, and 5% wore sun-protective hats (Buller, Cokkinides, et al.,
2011).
Few data regarding sun protection are available for children and adolescents from racial
and ethnic groups other than NHWs. Estimates from two national samples range between 29%-
50% for any use of sunscreen by Latino children; for African-American children this figure is
approximately 10% (Centers for Disease Control and Prevention (CDC), 1998). Twenty-two
percent of African-American and 30% of other non-white children seek shade when playing
outdoors in the sun (Hall, Jorgensen, et al., 2001). Nationally representative data on protective
clothing or hat use are not readily available for U.S. non-white children and adolescents;

8
however, one study found that 12% of Hispanic high school students reported wearing sun
protective clothing “most of the time” or “always” (Ma, Collado-Mesa, Hu, & Kirsner, 2007).
Methods of sun protection and UV exposure behaviors also vary by race and ethnicity.
The use of sunscreen is more prevalent in NHW children whereas African-American and
Hispanic youth are more likely to practice shade-seeking and use umbrellas as primary sun
protective methods (Buller, Cokkinides, et al., 2011; Cokkinides et al., 2004; Ma et al., 2007).
However, at least one study found equal shade-seeking behavior between NHWs and Hispanic
children in a high UV environment, while finding overall a lower likelihood of use of any sun
protective method among Hispanic students than in their NHW counterparts (Ma et al., 2007).

Theoretical foundations and correlates of sun protection in children
Theoretical models. Theoretical models most frequently used to explain child sun
protection include the health belief model (Kristjansson, Branstrom, Ullen, & Helgason, 2003;
Mermelstein & Riesenberg, 1992), the theory of reasoned action and its extension, the theory of
planned behavior (Hunter et al., 2010); social cognitive theory (Buller, Reynolds, et al., 2006;
Buller, Taylor, et al., 2006); the transtheoretical model (Adams, Norman, Hovell, Sallis, &
Patrick, 2009; Olson, Gaffney, Starr, & Dietrich, 2008); and Fishbein’s integrative model
(Heckman & Coups, 2011). These individual-level theories ground children’s sun protection
behaviors in determinants such as personal perceived ability to use sun protective methods (self-
efficacy); decision-making abilities and intentions to practice sun protection; perceived personal
sense of susceptibility to adverse consequences as a result of too much sun exposure; perceived
severity of skin cancer and consequences of too much UV; and observational learning and

9
perceived normative influences as a result of the sun protective habits and norms of important
others (e.g. parents and peers).
In addition to the above theoretical frameworks, recent studies have also focused on
policy and environmental aspects of child sun protection (Buller et al., 2002; Buller, Reynolds, et
al., 2011; Glanz, Chang, Song, Silverio, & Muneoka, 1998) as well as the critical influence of
family habits on children’s protective behaviors (O'Riordan et al., 2009; Robinson, Rigel, &
Amonette, 2000). These studies have found that few schools and child recreation centers in the
U.S. have sun protection policies in place (Buller et al., 2002; Buller, Reynolds, et al., 2011).
Further, family-related variables such as parental insistence on use of sun protection for child,
parental role-modeling of sun protective practices, and parents’ own skin cancer risk factors for
and level of awareness of skin cancer are key determinants in the sun protection of children,
especially those under the age of 12. Thus, children’s environment and social context, both at
home and at school, greatly influence the amount of UV exposure they receive and the sun
protection methods practiced.
Overall, these theoretical frameworks highlight and endeavor to account for the complex
influences of individual and social influence motivating child sun protection. Despite this
research, some authors have called for greater theory development in the field of child sun
protection, stating that the field has relied primarily on descriptive research rather than in-depth
examination or testing of the psychosocial factors motivating child sun protective behaviors
(Glanz, Lew, Song, & Cook, 1999; Saraiya et al., 2004; Schuz, 2012). Moreover, these
theoretical frameworks may not adequately explain or appropriately generalize to sun protection
among ethnic minority children, as psychosocial factors associated with sun protection may be

10
different among these populations (Andreeva VA, 2008; Coups et al., 2014; Pichon, Corral,
Landrine, Mayer, & Adams-Simms, 2010).

Socio-demographic correlates. Socio-demographic correlates of greater sun protection
behaviors in children across all sun protective methods include non-Hispanic white race (Buller,
Cokkinides, et al., 2011; Centers for Disease Control and Prevention (CDC), 1998; Hall, Jones,
& Saraiya, 2001); fair and sun-sensitive skin (Buller, Cokkinides, et al., 2011); younger age of
child (Lower, Girgis, & Sanson-Fisher, 1998b); parental use of sun protection (Behrens,
Thorgaard, Philip, & Bentzen, 2013; Cokkinides et al., 2004; Dobbinson et al., 2012; O'Riordan,
Geller, Brooks, Zhang, & Miller, 2003); advice and information from health care providers,
friends and family to protect from the sun (Cokkinides et al., 2004); and family history of skin
cancer (Centers for Disease Control and Prevention (CDC), 1998). In addition, while younger
age of child has been associated with greater use of sun protection, younger children also
experience greater UV exposures with longer duration of time playing outdoors in the sun
(Buller, Cokkinides, et al., 2011).
While a few studies have shown contrasting results (Lower et al., 1998b), pre-adolescent
girls generally practice greater sun protection and hold more favorable attitudes towards
protective behaviors than boys (Alberg, Herbst, Genkinger, & Duszynski, 2002; Coogan et al.,
2001; Dixon, Borland, & Hill, 1999; A. C. Geller et al., 2002). Girls and boys also differ on their
method of sun protection, with sunscreen use higher in girls and use of sun protective clothing
including hats more prevalent in boys (Buller & Borland, 1999; Robinson, Rademaker,
Sylvester, & Cook, 1997). As children age into adolescence, gender-based tanning norms
emerge, and girls engage in intentional exposures (sunbathing and indoor tanning) at higher rates

11
than boys (Robinson et al., 1997), while teenage boys incur greater ambient UV exposure
through spending more time outdoors (Buller & Borland, 1998).
Correlates of sun protection among ethnic and racial minority youth have been found to
differ than those for NHWs (Buller, Cokkinides, et al., 2011; Ma et al., 2007), with acculturation
emerging as an important factor influencing the prevalence and patterns of sun protection among
persons of color. Acculturation, the process by which a cultural group is brought into contact
with and adopts the beliefs and behaviors of another culture (typically used to describe a
minority cultural group adopting the norms of dominant culture) (Berry, 1990; Negy, 1992), has
been shown to significantly influence sun protective behaviors in Hispanic adults, with more
Anglo-acculturated Hispanics adopting U.S. norms of sun protection (e.g. use of sunscreen) and
less acculturated Hispanics more likely to use sun protective clothing and hats and to seek shade
(Andreeva et al., 2009; Coups E.J., 2013; Coups et al., 2012). However, no studies to date have
specifically addressed the influence of acculturation on child sun protection behaviors.

Psychosocial correlates. Psychosocial factors that influence the use of sun protective
behaviors among children and adolescents include knowledge of sun protection concepts, with
greater knowledge associated with increased protective behaviors (Alberg et al., 2002)
(Andreeva VA, 2008); positive sun protection attitudes and perceived benefits of sun protection
(Cokkinides et al., 2001; Rouhani et al., 2009); perceived self-efficacy, or the confidence that
one can use sun protective methods to achieve protection (Alberg et al., 2002; Buller, Reynolds,
et al., 2006); and reduced barriers to use of sun protection, e.g. the availability and ease of use of
sunscreen, shade and protective clothing (Reynolds, Buller, Yaroch, Maloy, & Cutter, 2006).

12
As with socio-demographic predictors of child sun protection, psychosocial correlates are
often behavior-specific, with one study finding that positive protective attitudes and less
perceived benefits of tanning was associated with shade-seeking and sunscreen but not with the
use of sun protective clothing and hats (Cokkinides et al., 2001). Similarly, the strength of such
factors differ on gender, race, and ethnicity, with positive tanning attitudes found to be more
highly associated with sun protective behaviors among girls and NHW adolescents (Andreeva
VA, 2008) and knowledge of sun protection concepts less influential with regards to sun
protection in Hispanic and other minority youth (Rouhani et al., 2009). In addition, greater sun
protective attitudes and normative habits have also been inversely associated with higher UV
exposures; children with greater risk factors (e.g. white race, lighter skin) who have more
knowledge of sun protection methods as well as greater self-efficacy to use sun protection have
been found to report longer duration of time spent in the sun (Reynolds et al., 1996) and more
positive attitudes towards tanning (Andreeva VA, 2008). These findings suggest that children
and adolescents who take protective measures such as use of sunscreen or protective clothing
may extend their time outdoors as they perceive themselves to be protected. In addition, lighter-
skinned adolescents may hold greater pro-tanning attitudes than darker-skinned adolescents and
thus intentionally increase their UV exposures (Reynolds et al., 1996).
Perceived peer norms—the acquisition of knowledge, beliefs, and attitudes from friends
as well as children’s observations and perceptions of group social norms—are also strong
correlates of sun protective behaviors, particularly in older adolescence where tanning norms and
appearance motivations emerge (Arthey & Clarke, 1995; Guy et al., 2014; Holman & Watson,
2013). Such normative beliefs are particularly important for sunbathing and indoor tanning, as
adolescents who perceive their peer group to engage in such behavior often follow group norms

13
(Heckman & Coups, 2011). Normative beliefs and perceptions regarding sun protective practices
for adolescents and young adults have been found to vary by gender, with girls reporting stronger
influences from their mothers and from peers than boys (Abroms, Jorgensen, Southwell, Geller,
& Emmons, 2003). For younger children, parental and familial normative influence is often a
key correlate, with children of parents who show greater subjective norms regarding use of sun
protection for both themselves and their child reporting lower levels of UV exposure and greater
sun protective practices (Donavan & Singh, 1999; Thomson, White, & Hamilton, 2012; Turner
& Mermelstein, 2005).
Perceived susceptibility to skin cancer and skin damage caused by UV exposure, which
has been associated with higher sun protection in adults (Koh et al., 1997) has not been as widely
examined in the literature with regards to children and adolescents. However, several studies
have shown that perceived risk of UV-related consequences motivates protective behaviors in
children and adolescents. In Mermelstein et al., higher perceived susceptibility to skin cancer
was associated with increased use of sunscreen among a largely NHW sample of older
adolescents (Mermelstein & Riesenberg, 1992) while in Rouhani et al., perceived risk of
developing skin cancer was related to higher use of all protective measures among a multiethnic
sample of 3rd-5th grade children (Rouhani et al., 2009). While Ma et al. found an association
between perceived risk and sun protection among NHW high schoolers (Ma et al., 2007),
Hispanic students perceived themselves to be at lower risk and were less likely to use any sun
protection in comparison to NHW students regardless of perceived susceptibility. In a
mediational analysis of middle schools students, Reynolds et al. found that perceived
susceptibility was not associated with increased sun protection (Reynolds et al., 2006). Thus, the

14
extent to which perceived risk for long-term consequences of UV overexposure motivates sun
protective behaviors for children and adolescents remains unclear.

Interventions to reduce UV exposure among children and adolescents
Sun protection health promotion programs for children are recommended by the World
Health Organization (WHO), where they are characterized as having “strong evidence” to
support their effectiveness (World Health Organization, 2002). The U.S. Preventive Services
Task Force has assigned a grade of “B” to primary prevention behavioral counseling in a primary
care setting to fair-skinned children ages 10-24, indicating a high probability of moderate benefit
of preventing skin cancer (Lin, Eder, & Weinmann, 2011).
Despite these endorsements, systematic reviews have found inconsistent results for
school-based skin cancer prevention programs targeting children and adolescents. In a 1999
review of the literature, Buller and Borland found that while most interventions were able to
change children’s knowledge of sun protection, only multiunit presentations were effective at
changing sun protective behaviors (Buller & Borland, 1999). In addition, where results were
found, most interventions found greater effects for changes in knowledge than in attitudes or
behaviors. A more recent review by Hart et al. concurred, finding greatest intervention efficacy
among multicomponent programs, chiefly in primary school settings (Hart & Demarco, 2008).
In a systematic review of 33 school-based interventions targeted to children and
adolescents conducted by the Task Force on Community Preventive Services, Saraiya et al.
found sufficient evidence for only interventions in primary school settings over those targeting
older adolescents (Saraiya et al., 2004). Further, the Task Force concluded that studies showed
sufficient evidence for interventions to improve use of protective clothing only; evidence was not

15
sufficient to determine the effectiveness of interventions to change more objective measures of
UV exposures such as UV avoidance or sunburn reduction.

Limitations of research regarding skin cancer prevention and sun protection in children
Several limitations exist with respect to child sun protection research. As described, much
of the intervention efforts to date have been most successful in changing knowledge of sun
protection, but less able to effect change in UV-related behaviors and reduction of UV exposure.
While multicomponent interventions in primary schools have been comparatively effective,
magnitudes of their effects remain low to moderate (Buller & Borland, 1999; Saraiya et al.,
2004). These results suggests that to date, the motivations driving child sun protection may not
be adequately understood, or have not been appropriately targeted, or such behaviors are
resistant to change.
Further, the lack of standardized approaches to intervention assessment has hampered
comparability, as measurement of sun protective behaviors and objective measures of UV
exposure has varied across studies. While adequate reliability and accuracy has been found when
comparing self-reported UV behaviors to objective measures of UV (Hillhouse, Turrisi, Jaccard,
& Robinson, 2012), the use of different outcome measures and lack of “gold standard” for
measuring exposure has made it difficult to compare intervention effects or conduct meta-
analysis across studies. In addition, theory development has been relatively weak in skin cancer
prevention, and the lack of validated items and scales with construct validity to measure sun
protective behaviors has further hindered comparisons (Hart & Demarco, 2008; Saraiya et al.,
2004).

16
Another important limitation of research on child sun protection is the lack of ethnic
diversity, as most research to date has been confined to all or majority NHW populations. As
noted, while NHWs are at highest risk of skin cancer, incidence is rising among Hispanic
populations, and little in-depth information is known about the sun protection habits of this
population. Further, relatively high rates of sunburn have been found among Hispanics: 43.1% of
Hispanic adults and 22% of Hispanic adolescents report sunburn in the previous year (Centers
for Disease Control and Prevention (CDC), 1998; Coups et al., 2012). In addition, Hispanics
report lower perceived risk of consequences from UV overexposure or skin cancer than NHWs
(C. Hernandez et al., 2012). Very little data are available for Hispanic children under the age of
12 with regards to prevalence of sunburn, sun protective behaviors, and the psychosocial factors
motivating these behaviors.

Introduction to the dissertation studies
As incidence of melanoma and non-melanoma skin cancer continues to increase, large-
scale prevention efforts will be an important strategy to reducing rates of disease. Primary
prevention targeted to children, if effectively received, can yield long-term benefits, reducing
both individual burden of disease as well as the economic impact of skin cancer, which is
considerable in the United States (Seidler, Pennie, Veledar, Culler, & Chen, 2010). To aid in the
development of such efforts, the studies in this dissertation investigate determinants of children’s
sun protective practices in a unique dataset of largely ethnic minority children in a high UV
environment, Los Angeles County.
The dissertation comprises three studies, each of which examines a facet of the SunSmart
randomized controlled trial, a large-scale skin cancer prevention intervention for elementary

17
school students. Each study has as its overarching aim the generation of pertinent and useful data
to guide refinements and innovations for the design of future sun protection interventions
targeted to multiethnic children.
The theoretical model for the present research is presented in Figure 1-1 and draws from
Bandura’s social cognitive theory, in which individual, social, and environmental factors interact
to influence behaviors (Bandura, 1986). For the present research, children’s socio-cultural
background, family sun habits, and environmental context are understood to directly influence
sun protection. Several psychosocial constructs—knowledge, perceived self-efficacy, peer
subjective norms, perceived risk of the consequences of UV overexposure, and barriers to use of
sunscreen and sun protective clothing—mediate between individual and environmental variables
and sun protection behaviors (use of sunscreen, use of protective clothing, shade seeking, and
duration of sun exposure). Sun protective behaviors in turn predict UV exposure. Moderators
hypothesized to influence sun protective behaviors from the literature include gender, ethnicity,
acculturation, and skin phototype.

18
Figure 1-1. Theoretical Model for the Research

Research hypotheses
Three studies were conducted using data from SunSmart, a school-based randomized
controlled trial conducted in an ethnically diverse sample of 4th and 5th grade children in Los
Angeles County. As the trial is ongoing, two years of data (Years 1 and 2) were used (N=1,754).
In Study 1, associations between UV-related behaviors and a range of predictors (demographic,
psychosocial, and cultural/familial) in the baseline data were assessed. In Study 2, short-term and
long-term mediators between intervention effects and sunscreen use were examined. In Study 3,
a latent class analysis characterized unobserved subgroups in the sample based on distinct sun
protection behaviors. Specific hypotheses are as follows

Study 1
Hypothesis 1: Prevalence of sun protection behaviors would be lower for non-white
children than for ethnic minority children. However, sunburn rate would be comparable
Individual)&)
Environmental)
Factors)
Moderators)
Gender&
Ethnicity&
Accultura1on&
Skin&type&
Use&of&
sunscreen&
&
Use&of&
protec1ve&
clothing&
&
Shade;seeking&
&
&
&
&
&
Knowledge&&
Perceived&self;efficacy&
Perceived&risk&
Peer&subjec1ve&norms&
Barriers&to&sunscreen&use&
Barriers&to&use&of&protec1ve&
clothing&
&
&
&
&
&
Socio;cultural&
&
Family&sun&
protec1on&
&
Environmental&
barriers&&&
facilitators&
Cogni6ve)&)
Psychosocial)Factors)
Behaviors)
UV&&
Exposure&
Exposure)

19
particularly between Hispanics and NHWs, per research indicating equivalent rates of sunburn in
Hispanics.
Hypothesis 2: Higher levels of psychosocial and familial variables would be positively
associated with greater sun protective practices.
Hypothesis 3: Factors associated with sun safe behaviors would differ by level of
acculturation, with the use of sunscreen associated with greater acculturation to U.S. norms while
the use of protective clothing and shade would be more strongly associated with lower U.S.
acculturation, in line with recent research.
Hypothesis 4: Psychosocial variables would mediate the relationship between
acculturation and sun protection outcomes (e.g., greater acculturation would be associated with
higher levels of sun-protective psychosocial variables leading to greater sun protection).

Study 2
Hypothesis 4: Improvements in sun protection-related knowledge, perceived peer norms,
perceived risk (of long-term consequences from UV overexposure), perceived self-efficacy, and
barriers to use of sun protection methods would mediate program effects leading to improved
outcomes for students participating in both intervention conditions (dosimetry and curriculum).
Hypothesis 5: The dosimetry condition would yield stronger indirect effects than the
curriculum condition due to the innovative feedback-based approach.

Study 3
Hypothesis 6: Distinct subgroups that are homogeneous within each class would be
present in the sample.

20
Hypothesis 7: Subgroups will vary by hypothesized predictors, with female sex, fair skin
phototype, sunburn reactivity, and high family use of sun protection for child predicting
membership in higher protection classes. In alignment with studies regarding acculturation and
sun protection in Hispanic adults, lower acculturation to U.S. norms will predict membership in
higher protection classes.

21
Chapter 2: Methods
Data source and participants
The present study utilized data from SunSmart, a randomized controlled trial designed to
improve sun protection knowledge, attitudes, and behaviors in 4
th
and 5
th
grade elementary
school students in Los Angeles County. All program activities took place during the Spring
semester and were integrated into normal classroom time. Schools were recruited as a
convenience sample based on proximity to the university and willingness to participate in the
intervention. All were Title I public schools in Los Angeles County, and all were located in
urban areas. The percentage of students eligible to participate in the free/reduced lunch program
ranged from 56.9% to 94.6%, with a median of 86.9%.
Eleven schools were randomized to one of three experimental conditions: Curriculum,
Curriculum + Dosimetry ("Dosimetry"), or Observation Only ("Observation") (Figure 2-1) (see
Chapter 4 for a detailed description of the intervention). Data were collected at baseline before
program implementation and at two follow-up time points: immediately following program
activities (week 4 for Curriculum and Observation; week 5 for Dosimetry); and at 3 months from
baseline for all conditions. At present, the trial is ongoing; the present research utilized data
collected in Year 1 and Year 2 of SunSmart (Spring 2013-2014).

22
Figure 2-1. SunSmart Intervention Design

Data were collected in accordance with the University of Southern California’s
Institutional Review Board (IRB) approved Human Subjects Research practices. Informed
consent was waived for students due to the educational nature of the program. Paper and pencil
surveys were administered in classrooms by trained data collectors with teachers present. In
total, 1,754 students from 11 schools and 62 classrooms participated in the trial in the first two
years, with 1,248 completing all three waves of data collection.

23
Figure 2-2. Flow diagram of SunSmart RCT

Selection bias and generalizability among the studies
Studies 1 and 3 used baseline data in cross-sectional examinations of correlates and
patterns of sun protection behavior among the sample. For these studies, the primary threat to
validity is the potential difference between children who participated in data collection at
baseline (pretest) versus those who were not present. While our baseline sample size comprised
1,646 children, our total size over the three waves of data collection was 1,754. Thus, 94% of the
total sample was represented in the baseline data, and no differences were found between
students who were present for the baseline survey versus those not present with respect to
Recruited to SunSmart: 11 schools
N=1,754
Dosimetry
N=826
Curriculum
N=622
Observation
N=198
Completed
Baseline
Dosimetry
N=815
Curriculum
N=607
Observation
N=198
Completed
Post-test
Dosimetry
N=631
Curriculum
N=586
Observation
N=187
Completed
Post-post
test
Randomized

24
demographic characteristics and primary covariates considered in the analysis. Further, we
would not expect a differential relationship of predictors to outcome between completers and
non-completers (e.g., the relationship between psychosocial, familial, or cultural variables and
sun protection outcomes would not likely differ between children who were present versus
absent at baseline).
In addition, informed consent was waived due to the educational content of the
intervention, which had the advantage of mitigating potential selection bias. Because active
assent and consent, and particularly active parental consent, has been found to create differences
among students who participate versus those who do not participate in interventions (e.g. higher-
risk students with less healthy behaviors or those less receptive to intervention) (Anderman et al.,
1995), the recruitment of all students within schools and grades that opted to participate reduced
the probability of this type of selection bias.
For Study 2, which used a longitudinal design to compare effects between treatment
conditions, baseline values for demographic characteristics and key sun protection variables were
compared across the three conditions (Table 2-1). Chi-square tests and one-way analysis of
variance (ANOVA) were used to evaluate any significant differences. Significant differences
were found for grade level, with a greater number of mixed 4th and 5th grade classes in the
curriculum condition and no mixed classes in the observation condition. Race and ethnicity were
also significantly different, with a greater proportion of students self-reporting as
Hispanic/Latino in the dosimetry and observation conditions compared to the curriculum
condition, and a slightly higher number of students reporting as Asian/Pacific Islander in the
dosimetry and curriculum conditions. In addition, level of acculturation was significantly
different, with a slightly higher mean score for the curriculum condition compared to the other

25
two conditions. Finally, duration of time outdoors at school differed significantly between
conditions, with a greater proportion of students in the curriculum condition reporting being
outdoors for 31 minutes to 1 hour on an average school day compared to the other conditions.
These systematic differences between conditions may have introduced bias specifically
related to the imbalance in ethnicity, with the curriculum condition less Hispanic/Latino and
more U.S.-acculturated than either the dosimetry or observation groups. Because Hispanics may
have lower sun protection and/or less receptive to intervention based on perceived lower risk,
estimates may be biased for comparisons of the curriculum versus the observation group.
However, as detailed in Study 1 (Chapter 3), sun protection behaviors did not substantially differ
by race/ethnicity in this sample nor was skin phototype a significant correlate of sun protection,
suggesting that such differences may have minimal impact in this multiethnic sample. In
addition, Study 2 utilizes a subsample of data that did not differ significantly by ethnicity, and
adjusted for grade level based on subsample comparisons (see Chapter 4 for further details).

26

Dosimetry
(N=876)
N(%)
Curriculum
(N=665)
N(%)
Observation
(N=213)
N(%)
Between-Group
Differences
Demographic Variables
Grade level
4th 410 (46.80) 288 (43.31) 81 (38.03)
5th 443 (50.57) 256 (38.50) 132 (61.97)
Mixed 4th/5th 23 (2.63) 121 (18.20) 0
Gender
Female 409 (46.69) 308 (46.32) 95 (44.60)
Male 416 (47.49) 308 (46.32) 103 (48.36)
Missing 51 (5.82) 49 (7.37) 15 (7.04 )
Race/Ethnicity
Hispanic/Latino 567 (64.73) 329 (49.47) 161 (75.59)
Asian/Pacific Islander 55 (6.28) 54 (8.12) 3 (1.41)
Non-Hispanic white 21 (2.40) 38 (5.71) 11 (5.16)
African-American/Black 57 (6.51) 85 (12.78) 4 (1.88)
Mixed race/other 104 (11.87) 84 (13.99) 19 (8.92)
Missing 72 (8.22) 66 (9.92) 15 (7.04)
Skin phototype
Lighter skin (Type I-II) 139 (15.87) 126 (18.95) 43 (20.19)
Darker skin (Type III-V) 683 (77.97) 484 (72.78) 155 (72.77)
Missing 54 (6.16) 55 (8.27) 15 (7.04)
Level of acculturation
(AHIMSA scale)* 2.8 (2.21) 3.18 (2.46) 2.68 (2.25)
F(2, 1,567)=5.56,
p=0.0039
UV-related Variables
Duration of time outdoors at
school
30 minutes or less 583 (66.55) 346 (52.03) 132 (61.97)
31 minutes to 1 hour 197 (22.49) 238 (35.79) 59 (27.70)
2+ hours 42 (4.79) 36 (5.41) 7 (3.29)
Missing 54 (6.16) 45 (6.77) 15 (7.04)
Duration of time outdoors at
home
30 minutes or less 301 (34.36) 235 (35.34) 86 (40.38)
31 minutes to 1 hour 199 (22.72) 152 (22.86) 48 (22.54)
2+ hours 327 (37.32) 233 (35.03) 64 (30.04)
Missing 49 (5.59) 45 (6.77) 15 (7.04)
Ever sunburned
Yes 455 (51.94) 348 (52.33) 123 (57.75)
No 370 (42.24) 267 (40.15) 75 (35.21)
Missing 51 (5.82) 50 (7.52) 15 (7.04)
Sunburn in past month
0 times 523 (59.70) 407 (61.20) 127 (59.62)
1-3 times 256 (29.92) 176 (26.46) 61 (28.64)
>3 times 46 (5.25) 31 (4.66) 10 (4.7)
Missing 51 (5.82) 51 (7.67) 15 (7.04)
Sunburn in prior summer
0 times 403 (46.0) 407 (61.20) 83 (38.97)
1-3 times 323 (36.87) 176 (26.46) 89 (41.78)
>3 times 99 (11.30) 31 (4.66) 26 (12.20)
Missing 51 (5.82) 51 (7.67) 15 (7.04)
*M±SD
Table 2-1. Sample baseline characteristics and between-group differences in Los Angeles 4th and 5th graders
participating in a skin cancer prevention program, SunSmart (N=1,754)
χ2= 9.467;
p=0.304
χ2=157.327;
p<0.0001
χ2= 82.409;
p<0.0001
χ2=4.3765;
p=0.112
χ2= 47.917;
p<0.0001
χ2=4.402;
p=0.622
χ2=3.158;
p=0.206
χ2=10.20;
p=0.251
χ2=0.245;
p=0.885

27
Finally, all measures used in the following studies were self-reported, with the potential
for introducing social desirability bias, potentially altering the magnitude or direction of effect.
However, recent studies have shown that self-reported sun protection behaviors have adequate to
good accuracy among adolescents (Hillhouse et al., 2012; Lower, Girgis, & Sanson-Fisher,
1998a).
Generalizability: Schools elected to participate in the intervention. Therefore, schools
represented in these studies may be more motivated to adopt health promotion programs and
subsequently have students with greater exposure to such programs. Therefore, the sample may
not represent typical sun protection behaviors and thus may not be generally representative of the
school-based population. However, school demographic data do not reveal any substantial
distinctions between the Title I schools represented in the study on school-level characteristics
such as race/ethnicity, achievement as measured by Academic Performance Index (API) scores,
or socioeconomic status in comparison to the district average. In addition, while the sample size
is large, it is predominantly Hispanic and reflects the unique ethnic composition of Southern
California. Therefore, these results should generalize to Western states with multiethnic and
large Latino populations.
Measures
SunSmart utilized a questionnaire with 67 measures and additional demographic
variables. All measures were self-reported. Sun protection behavior items and psychosocial
variables were primarily adapted from the work of Buller et al. (Buller, Taylor, et al., 2006;
Reynolds et al., 2006) and Glanz et al. (Glanz et al., 2008).

28

Variable Reference Description Waves Used
Outcomes Sunscreen use Glanz et al.
Three items measuring sunscreen use at school, outside of
school, and general frequency on a four-point frequency
Likert-type scale (1=Often; 4=Never) All three waves
Sun protective
clothing use Glanz et al.
Four items assessing the use of long sleeves and long pants
at school and outside of school on a four-point frequency
Likert-type scale (1=Often; 4=Never) All three waves
Sun protective
hat use Glanz et al.
Two items assessing use of sun protective hats at school and
outside of school on a four-point frequency Likert-type
scale (1=Often; 4=Never). All three waves
Shade-seeking Glanz et al.
Two items assessing shade-seeking behaviors at school and
outside of school on a four-point frequency Likert-type
scale (1=Often; 4=Never). All three waves
Psychosocial
predictors
Knowledge of
sun protection Buller et al.
Nine variables with response options“true”, “false”, and “I
don’t know.” Correct responses were summed into a single
variable with higher scores reflecting greater knowledge of
sun protective concepts. ["I don't know" coded as 0.] All three waves
Barriers to
sunscreen use Buller et al.
Four Likert-type items rated on a 1-4 scale frequency
(1=untrue; 4=true). The scale was averaged into a single
summary variable and reverse-coded so that higher scores
reflected lower barriers. All three waves
Barriers to use of
sun protective
clothing Buller et al.
Three Likert-type items rated on a 1-4 scale frequency
(1=untrue; 4=true). The scale was averaged into a single
summary variable and reverse-coded so that higher scores
reflected lower barriers. All three waves
Perceived peer
norms Buller et al.
Four Likert-type items rated on a 1-4 frequency scale
(1=untrue; 4=true). The scale was averaged into a single
summary variable with higher scores reflecting greater
perceived peer norms. All three waves
Perceived risk Buller et al.
Two Likert-type items rated on a 1-4 frequency scale
(1=untrue; 4=true). The scale was averaged into a single
summary variable with higher scores reflecting higher
perceived risk. All three waves
Perceived self-
efficacy Buller et al.
Five Likert-type items rated on a 1-4 frequency scale
(1=I’m sure I can’t; 4=I’m sure I can). The scale was
averaged into a summary variable with higher scores
reflecting higher perceived risk. All three waves
Family and
cultural
predictors
Family use of sun
protection for
child
Developed for
SunSmart
Three Likert-type items rated on a 1-4 frequency scale
(1=often; 4=never). Each item assessed the use of sunscreen
and protective clothing by parent for child. The items were
averaged into a single summary variable, with higher scores
reflecting greater use of sun protection by parent for child. All three waves
Availability of
sunscreen at
home
Developed for
SunSmart
Single item with responses including “yes, definitely”
“maybe,” and “no, definitely.” "Maybe" and "no" coded as
0. All three waves
Family
communication
regarding sun
protection
Developed for
SunSmart
Two items asking if children had spoken to their family
about sunscreen or protective clothing (long sleeves, long
pants, hats) within the past two weeks; possible responses
were “yes” and “no.” All three waves
Acculturation Unger et al.
Measured with the Acculturation, Habits, and Interests
Multicultural Scale for Adolescents (AHIMSA) scale, an 8-
item scale measuring cultural preferences. See Chapter 3:
Study 1 for details. Wave 2 only
Socio-
demographic
covariates Skin phototype
Fitzpatrick et
al.
Five-level item ranging from 1=Very Fair to 5=Very Dark,
adapted from skin types I-V of the Fitzpatrick skin
phototype scale (Fitzpatrick, 1988). Children were provided
with a visual graphic to aid in skin type matching. Wave 1 only
Table 2-2. Selected measures used to evaluate a skin cancer prevention program in Los Angeles 4th and 5th graders, SunSmart

29
Items and scales were coded so that higher scores reflected more protective behaviors.
All measures used in the dissertation studies are described in Table 2-2, including source and
measurement wave.

Statistical analysis for studies in this dissertation
The primary aim of all three studies was to explore associations and psychosocial
mechanisms promoting greater sun protective knowledge, attitudes, and distinct behaviors in a
unique sample of predominantly minority children, utilizing both baseline data (Studies 1 and 3)
and measurements over the course of the four-month intervention (Study 2). Preliminary analysis
for all studies consisted of univariate and bivariate analyses and descriptive statistics. Because of
the study design of the SunSmart trial, which recruited students from multiple schools, data were
examined for potential effects of clustering. Intraclass correlation coefficients (ICCs) were
calculated for all variables, and in each case, found to be <0.02. As any effects of clustering
appeared negligible in the sample (ICC near zero; see Appendix), multilevel analyses were not
conducted.
For Study 1, multiple linear regression was used to evaluate associations between
demographic, familial and cultural variables and sun protection. Path analysis was conducted
using maximum likelihood estimation to identify psychosocial mediators between acculturation
and sun protection behaviors in single mediator models.
For Study 2, autoregressive path analyses were used to identify significant mediators of
intervention effects on sunscreen use as an outcome. We followed contemporary
recommendations for mediation analysis which focus on the product of path coefficients a
(independent variable to mediator) and b (mediator to outcome) (D. MacKinnon, 2008), and used

30
bias-corrected bootsrap confidence intervals to determine significance level of the mediated
effect (Preacher & Hayes, 2004).
Study 3 used latent class analysis, a person-centered empirically-driven statistical method
to detect unobserved subgroups or “classes” in a population using a set of observed variables (B.
Muthén & Muthén, 2000). Using maximum likelihood estimation, individuals are categorized
into mutually exclusive classes determined by their responses on indicator variables to find the
smallest number of classes that fit the data. The model’s parameters include class membership
probabilities (the likelihood of being in a class) and item-response probabilities (the likelihood of
endorsing or fulfilling each indicator based on class membership). Model fit is determined by
information criteria including the Akaike Information Criterion (AIC), Bayesian Information
Criterion (BIC), and the sample-size adjusted BIC that compare the results of model with k
classes to the model with k-1 classes. The model that best fits the data with the smallest number
of classes and yields interpretable results is then chosen (Collins & Lanza, 2010).
All statistical analysis was performed with Stata version 12.1 (StataCorp, 2011) and
Mplus version 6 (L. K. Muthén, & Muthén, B. O., 1998-2011).

Treatment of missing data
Because no differences were found between students who completed all waves of data
versus those who completed only one or two waves on key variables, missing data were
considering ignorable (missing completely at random [MCAR] or missing at random [MAR])
(Little & Rubin, 1987). For Study 1 using baseline data, a complete case approach was used,
excluding observations with item nonresponse on pertinent variables. Because the percentage of

31
missing data was low (<2.0%) for item non-response, robust sample size and statistical power
were maintained.
For Studies 2 and 3, full-information maximum likelihood estimation (FIML) was used.
FIML estimates model parameters and standard errors using available data across all
observations and measurement waves and is the optimal approach to ignorable missing data
(Collins & Lanza, 2010). However, FIML cannot accommodate missing data on covariates, and
therefore individuals with data missing on covariates were excluded from analysis. For Studies 2
and 3, this represented a low percentage of individuals (for Study 2, <10%; for Study 3, less than
10 observations), and therefore did not reduce sample size or statistical power.

32
Chapter 3: Study 1

Correlates of sun protection behaviors in a sample of ethnic minority children residing in a
high UV environment

Kimberly A. Miller, MPH
1
, Jimi Huh, PhD
1
, Jennifer B. Unger, PhD
1
, Jean L. Richardson,
DrPH
1
, Martin W. Allen, PhD
2
, David H. Peng, MD
3
, Myles G. Cockburn, PhD
1,3

1
Department of Preventive Medicine, Keck School of Medicine of the University of Southern
California, Los Angeles, CA
2
MacDiarmid Institute for Advanced Materials and Nanotechnology, Department of Electrical
and Computer Engineering, University of Canterbury, Christchurch, New Zealand
3
Department of Dermatology, Keck School of Medicine of the University of Southern California,
Los Angeles, CA.

Contact:
Kimberly A. Miller
Department of Preventive Medicine, Keck School of Medicine
2001 N Soto St, Suite 318-A
Los Angeles, CA 90032
Phone: (626) 864-6433 / Fax: (323) 865-0095
Email: kim.miller@med.usc.edu

33
Abstract
Exposure to ultraviolet (UV) light is the primary environmental risk factor for skin
cancers. Excessive UV and sunburn in childhood are associated with increased risk of
developing melanoma and non-melanoma skin cancer in adulthood. While non-Hispanic white
(NHW) populations are at greatest risk of melanoma, rates of invasive melanoma are rising in
ethnic minorities, particularly among Hispanics living in high UV environments in the United
States. We examined factors related to sun protection among a large sample of ethnic minority,
predominantly Hispanic elementary schoolchildren using baseline data from a skin cancer
prevention randomized trial (N=1,488). Associations between multiple predictors from different
domains (psychosocial, familial and cultural) and four distinct sun protection outcomes (use of
sunscreen, protective clothing, hats, and shade) were examined in multivariate linear regression
models. We also examined the mediating effects of familial and psychosocial variables on the
relationship between acculturation and sun protection outcomes.
Hispanics and NHW children had comparable prevalence of sunburn. Female gender and
Hispanic ethnicity were both associated with greater use of sunscreen, while being female was
associated with higher shade-seeking behavior. Psychosocial correlates associated with greater
sun protection behaviors included perceived peer norms, perceived self-efficacy, and lower sun
protection barriers. Family-related variables, including parent's use of sun protection for child,
sunscreen availability at home, and discussion of sun protection were significantly associated
with greater sun protection across all sun protection outcomes. Only one mediated effect was
found in which family sun habits for child mediated the relationship between acculturation and
child’s shade-seeking.

34
Findings from the study show that ethnic minority children experience similar sun
protection correlates previously observed among NHW children. This overlap in risk and
protective factors suggests that public health messages to improve sun protection may generalize
across cultural contexts and could be scaled into sun safety health promotion for diverse youth
audiences. The strong positive associations between parental variables and child sun protection
suggest that family habits are instrumental to children’s sun safe behaviors and interventions that
incorporate family and caretakers may be most effective for school-aged children. Future
research should continue to investigate the extent to which culturally tailored versus standard
messages are required, and focus on the meaningful incorporation of family caretakers into sun
protection education for children.

35
Introduction
Exposure to ultraviolet (UV) light is the major environmental risk factor for both
melanoma and non-melanoma skin cancer, and excessive UV and sunburn in childhood are
associated with increased risk of developing melanoma in adulthood (Gandini et al., 2005; D. C.
Whiteman et al., 2001). Primary prevention of melanoma, achieved by reduction of excessive
UV exposure, is thus appropriately targeted to children. While fair-skinned, non-Hispanic white
(NHW) populations are at greatest risk of melanoma, ethnic minorities such as Hispanics,
Asians, and African-Americans experience more advanced disease at diagnosis and poorer
disease outcomes than NHWs (Cress & Holly, 1997; Hu et al., 2009). In addition, rapidly rising
rates of invasive melanoma have been observed among Hispanics living in high UV
environments in the United States (Cockburn et al., 2006; Rouhani et al., 2010). Despite these
trends, few if any melanoma primary prevention public health efforts have been targeted to
Hispanic or other ethnic minority children (Saraiya et al., 2004) who are at risk of UV
overexposure, increasing their risk of melanoma in adulthood.
Because of racial and ethnic disparities in melanoma outcomes, researchers have begun
to examine the sun protection practices and psychosocial correlates of sun protection in non-
white adults. While sun protection is poorly practiced in the U.S. population in general (Buller,
Cokkinides, et al., 2011), non-white and darker-skinned adults have lower prevalence of sun
protection behaviors and low awareness and perceived risk of susceptibility to skin cancer,
compared with NHWs (Coups et al., 2012; C. Hernandez et al., 2012; Pichon, Corral, Landrine,
Mayer, & Norman, 2010; Pipitone, Robinson, Camara, Chittineni, & Fisher, 2002). Data are
sparse regarding the sun protection behaviors of ethnic minority children. In the few studies that
exist, ethnic minority youth have been found to practice less sun protection and possess less

36
knowledge about skin cancer than their NHW counterparts (Hall, Jones, et al., 2001; Ma et al.,
2007). In addition, ethnic minority and NHW children may have different sun protection
patterns, as studies have shown that Hispanic and African-American adolescents wear sunscreen
less frequently but use protective clothing and seek shade more often than NHWs (Rouhani et
al., 2009).
Data are also lacking with regards to the cognitive and psychosocial correlates of sun
protection among ethnic minority youth. Among NHW children, a sizeable literature has
identified correlates of sun protection, including knowledge of sun protection concepts (Alberg
et al., 2002; Andreeva VA, 2008); positive attitudes towards and perceived benefits of sun
protection (Cokkinides et al., 2001); perceived self-efficacy to use sun protection (Alberg et al.,
2002; Reynolds et al., 2006); reduced barriers to use of sun protection (Reynolds et al., 2006);
perceived peer norms (Arthey & Clarke, 1995; Guy et al., 2014; Holman & Watson, 2013);
parental normative influence (Donavan & Singh, 1999) (Thomson et al., 2012; Turner &
Mermelstein, 2005); and perceived risk (Mermelstein & Riesenberg, 1992; Rouhani et al., 2009).
However, it is unclear whether these constructs are relevant for and generalize to ethnic and
racial minority youth.

Sun protection and Hispanics
Due to substantial increases in melanoma incidence among Hispanic populations who
have the highest incidence rate of melanoma after NHWs in the U.S. (Cress & Holly, 1997),
there is a need for specific focus on sun protection education for this ethnic group. Studies have
identified disparities with respect to sun protective behaviors, beliefs, and attitudes between
Hispanics and NHWs, with both Hispanic adults and older adolescents demonstrating less use of

37
sun protection, less awareness of skin cancer, and less access to both sun protection information
and sunscreen than NHWs (Hay, Coups, Ford, & DiBonaventura, 2009; C. Hernandez et al.,
2012; Ma et al., 2007; Pipitone et al., 2002). In addition, sunburn prevalence is high amongst
Hispanic adults, with rates comparable to or exceeding those of NHWs (CDC, 2007).
Recent studies have focused on the potential role of acculturation on sun protection
behaviors of U.S. Hispanic. Acculturation, the process by which a cultural group is brought into
contact with and adopts the beliefs and behaviors of another culture (Berry, 1990; Negy, 1992;
Unger, 2002) has been associated with both risk and protective practices with regards to sun
protection in Hispanics. Higher levels of acculturation to the United States culture have been
associated with greater use of sunscreen among Hispanic adults, in alignment with norms of sun
protection among U.S. NHWs who use sunscreen as their chief form of sun protection (Hall,
May, Lew, Koh, & Nadel, 1997). However, higher acculturation for Hispanics has also been
associated with risk behaviors such as sunbathing and indoor tanning, practices which are again
typically more prevalent amongst NHWs (Andreeva et al., 2009; Coups E.J., 2013). By contrast,
lower U.S. acculturation among Hispanics has been associated with greater use of physical
barriers such sun protective clothing and hats as well as shade seeking, possibly in accordance
with traditional forms of sun protection practiced in Hispanic/Latino countries (Andreeva et al.,
2009; Coups E.J., 2013; C. Hernandez et al., 2012).
While such studies shed light on acculturation and sun protection among Hispanic adults,
few have focused on the role that acculturation may play in the sun protection attitudes, beliefs
and behaviors of Hispanic children. While a recent study found greater perceived benefits of UV
exposure for higher versus lower U.S.-acculturated Hispanic teens (Heckman & Cohen-Filipic,
2012), more research is required to examine the influence of acculturation on the practices of

38
Hispanic early adolescents, who are at risk of substantial UV exposures and are in the formative
phases of developing sun protection attitudes in order to design effective and tailored
interventions for this population.

The present study
This study examined factors related to sun protection among a sample of ethnic minority,
predominantly Hispanic elementary schoolchildren. Our primary aim was to explore correlates of
sun protection among this unique sample and to investigate whether factors associated with
greater sun protection were similar with those identified among NHW children. Such
information is an important step in developing sun safety interventions broadly targeted to
multiethnic children.
We examined associations between psychosocial, familial, and cultural predictors
previously in a sample of 4
th
and 5
th
graders participating in a sun safety education program in
Los Angeles (n=1,488). Because recent studies have suggested that sun protection behaviors
have distinct motivators and barriers (Saraiya et al., 2004), we assessed four separate sun
protection outcomes: use of sunscreen, sun protective clothing (long sleeves and long pants), use
of sun protective hats, and shade seeking. Psychosocial variables included knowledge of sun
protection concepts, perceived risk for UV-related consequences, perceived self-efficacy to use
sun protection, perceived peer descriptive norms for sun protection, and perceived barriers to sun
protection. Family predictors included parental use of sun protection for child, availability of
sunscreen in the home, and family communication regarding sun protection. The primary cultural
factor examined was acculturation. Covariates included gender, skin type, and sunburn ability.
Due to the large sample of Hispanic children and smaller numbers of other races, comparisons

39
were limited to estimating differences between Hispanic versus non-Hispanic in statistical
models. However, prevalence of sun protection behaviors and sunburn was examined by each
ethnic and racial category to examine differences in proportions between these groups.
We hypothesized that prevalence of sun protection would be lower for non-white children
than for NHWs. However, we expected that sunburn rate would be comparable between
Hispanics and NHWs, per research indicating equivalent rates of sunburn between Hispanics and
NHWs (CDC, 2007). Extrapolating from the literature on NHWs and findings regarding
psychosocial predictors of sun protection among Hispanic adults (Coups et al., 2014), we
anticipated that higher levels of psychosocial and familial variables would be positively
associated with sun protection outcomes. We additionally hypothesized that higher levels of U.S.
acculturation would be positively associated with sunscreen use, but negatively associated with
the three other sun protection outcomes. Finally, we conducted an exploratory analysis among
the Hispanic sample only (n=972) in which we examined the mediating role of select
psychosocial constructs between acculturation and sun protection outcomes.

Methods
Sample and measures
Baseline data were used from the SunSmart randomized controlled trial (N=1,646).
Because acculturation was measured at posttest of the intervention, the sample was restricted to
the 1,488 students who completed pretest and were present to complete acculturation measures at
posttest. No significant differences were found between completers of both pre- and posttest and
those completing pretest only on demographic or sun protection items. For mediation analyses,

40
the sample was restricted to students who completed both pretest and posttest and who self-
reported as Hispanic or Latino (N=972).

Outcomes
We examined four sun protection outcomes: use of sunscreen, use of protective clothing
(long sleeves and long pants), use of sun protective hats, and shade-seeking. All variables were
coded so that higher scores reflected higher protective behaviors.
Sunscreen use: Three items assessing the use of sunscreen at school, outside of school,
and general sunscreen use on a four-point frequency Likert-type scale (1=Never; 4=Often) were
averaged into a single variable (Cronbach's α=0.74).
Use of long sleeves and long pants: Four items on a four-point frequency Likert-type
scale (1=Never; 4=Often) assessed the use of long sleeves and long pants at school and outside
of school. These were averaged into a single summary variable (Cronbach's α=0.74).
Use of hats for sun protection: Two items on a four-point frequency Likert-type scale
(1=Never; 4=Often) assessed use of sun protective hats at school and outside of school. These
were averaged into a single summary variable (Spearman's ρ=0.42).
Shade-seeking: Two items on a four-point frequency Likert-type scale (1=Never;
4=Often) assessed shade-seeking behaviors at school and outside of school. These were averaged
into a single summary variable (Spearman's ρ=0.40).

41
Predictors
Predictors were selected for their theoretical relevance to sun protection:
Psychosocial Predictors
Knowledge of sun protection concepts: Nine variables assessed students’ knowledge of
sun protection concepts. Response options included “true”, “false”, and “I don’t know.” Correct
responses were summed across the nine questions into a single summary variable with higher
scores reflecting greater knowledge of sun protective concepts. The response "I don't know" was
coded as an incorrect response.
Barriers to use of sunscreen: Barriers to sunscreen were assessed with four Likert-type
items rated on a 1-4 scale frequency (1=untrue; 4=true). The items were averaged into a single
summary variable with higher scores reflecting lower barriers.
Barriers to use of protective clothing: Barriers to protective clothing were assessed with
three Likert-type items rated on a 1-4 scale frequency (1=untrue; 4=true). The items were
averaged into a single summary variable with higher scores reflecting lower barriers.
Perceived peer norms: Perceived peer norms were assessed with four Likert-type items
rated on a 1-4 frequency scale (1=untrue; 4=true). The items were averaged into a single
summary variable with higher scores reflecting greater perceived peer norms.
Perceived risk: Perceived risk of consequences from solar overexposure was assessed
with two Likert-type items rated on a 1-4 frequency scale (1=untrue; 4=true). The items were
averaged into a single summary variable with higher scores reflecting higher perceived risk.

42
Perceived self-efficacy: Perceived self-efficacy to use sun protection was assessed with
five Likert-type items rated on a 1-4 frequency scale (1=I’m sure I can’t; 4=I’m sure I can). The
items were averaged into a summary variable with higher scores reflecting higher perceived risk.

Family and cultural predictors
Family use of sun protection for child: Family use of sun protection for child was
measured with three Likert-type items rated on a 1-4 frequency scale (1=never; 4=often). Items
assessed the use of sunscreen, protective clothing, and hat use by parent for child. The items
were averaged into a summary variable, with higher scores representing greater protective
behaviors variable.
Availability of sunscreen in the home: Availability of sunscreen in the home was
measured with a single item with responses including “yes/maybe/no.” The item was
dichotomized as 1=“yes” versus 0=“maybe/no.”
Family communication regarding sunscreen/protective clothing: Family communication
regarding sunscreen and protective clothing was measured with two single items asking if
children had spoken to their family about sunscreen or protective clothing (long sleeves, long
pants, hats) within the past two weeks; possible responses were “yes/no.”
Acculturation: Level of acculturation was measured with the Acculturation, Habits, and
Interests Multicultural Scale for Adolescents (AHIMSA) scale (Unger, 2002). The AHIMSA
scale is an 8-item scale measuring cultural preferences related to friends, food, media, etc. The
scale is multidimensional, with 4 response options for each preference: “the U.S.”, “the country
my family is from,” “both,” and “neither.” Each score ranges from 0-8 reflecting distinct

43
orientations of assimilation, separation, integration, and marginalization. For this study, the U.S.
orientation scale (assimilation) was used. To score this scale, the number of times students
responded to questions with “the U.S.” were summed to generate a single score ranging from 0-
8, with higher scores reflecting higher levels of assimilation to U.S. norms (Cronbach's α= 0.78
at baseline).

Socio-demographic covariates
Covariates included gender, ethnicity (Hispanic versus non-Hispanic), skin phototype,
and ever sunburned (yes/no).
Skin phototype: Skin type was assessed with a five-level item ranging from 1=Very Fair
to 5=Very Dark, corresponding to skin types I-V of the Fitzpatrick skin phototype scale
(Fitzpatrick, 1988); children were provided with a visual graphic to aid in skin type matching.
Skin type ranged from 1=very fair to 5=very dark.

Mediators
Variables hypothesized to mediate the relationship between acculturation and sun
protection outcomes included perceived risk, perceived self-efficacy, perceived peer norms, and
family sun protection habits for child. The mediation analysis was conducted on the Hispanic
sample only (N=971).

Statistical analysis
Descriptive statistics were used to characterize the sample overall and to examine the
prevalence of sun protection behaviors between ethnic and racial groups. Cronbach's alpha or

44
Spearman's rho were calculated for those items presumed to indicate a single underlying
construct (Cronbach, 1951). Multiple linear regressions were performed with all variables
entered simultaneously into the model. Path analysis was conducted using maximum likelihood
estimation to identify significant psychosocial mediators between acculturation as the exogenous
variable and outcomes in single mediator models (Kline, 2011).
The level of significance was set to alpha <0.05, and all analyses were performed in Stata
version 12 (StataCorp, 2011). Approval for the study was obtained by the University of Southern
California Institutional Review Board.

Results
Descriptive statistics
Table 3-1 shows baseline socio-demographic characteristics of children participating in
SunSmart.

45

Participants were balanced with regards to gender and grade level, and the largest
ethnic/racial category was Hispanic (65%). The majority of children (62%) reported a skin
phototype of “light brown.”

Prevalence of sun protection behaviors and sunburn
Table 3-2 and Table 3-3 show the prevalence of sun protection behaviors and sunburn-
related variables at baseline.
Mean age (SD) 9.84 0.71
n %
Gender
Female 746 50.1
Male 735 49.4
Missing 7 0.5
Grade Level
4th 649 43.6
5th 719
48.3
Mixed 4th/5th 120
8.1
Race/Ethnicity
Hispanic 972 65.3
Asian/Pacific Islander 111 7.5
Non-Hispanic White 62 4.2
African-American 117 7.9
Native American/American Indian 29 2.0
Mixed race/other 158 10.6
Missing 39 2.6
Skin phototype
Very fair 16 1.1
Fair 261 17.5
Light brown 924 62.1
Dark brown 256 17.2
Very dark 17 1.1
Missing 14 0.9
Table 3-1. Demographic characteristics of sample at
baseline, SunSmart (N=1,488)

46

Variable
Hispanic/Latino
n=972
Asian/Pacific
Islander
n=111
Non-Hispanic
White
n=62
African-
American
n=117
Mixed
race/other
n=187 χ2
Sunscreen at school
Often/sometimes 360 (37.15) 32 (28.83) 17 (27.42) 33 (28.21) 41 (21.92)
Rarely/never 609 (62.85) 79 (71.17) 45 (72.58) 84 (71.79) 146 (78.07)
Sunscreen at home
Often/sometimes 330 (33.95) 35 (31.53) 16 (25.81) 26 (22.22) 48 (25.81)
Rarely/never 642 (66.05) 76 (68.47) 46 (74.19) 91 (77.78) 138 (74.19)
Long sleeves at school
Often/sometimes 664 (68.45) 88 (79.28) 34 (54.84) 80 (68.38) 121 (64.71)
Rarely/never 306 (31.55) 23 (20.72) 28 (45.16) 37 (31.62) 66 (35.29)
Long sleeves at home
Often/sometimes 487 (50.36) 67 (60.36) 29 (46.77) 54 (46.15) 76 (40.86)
Rarely/never 480 (49.64) 44 (39.64) 33 (53.23) 63 (53.85) 110 (59.14)
Long pants at school
Often/sometimes 788 (81.40) 97 (87.39) 50 (80.65) 92 (78.63) 141 (75.81)
Rarely/never 180 (18.60) 14 (12.61) 12 (19.35) 25 (21.37) 45 (24.19)
Long pants at home
Often/sometimes 487 (50.36) 67 (60.36) 29 (46.77) 54 (46.15) 76 (40.86)
Rarely/never 480 (49.64) 44 (39.64) 33 (53.23) 63 (53.85) 110 (59.14)
Shade at school
Often/sometimes 554 (57.05) 66 (60.00) 37 (59.68) 55 (47.41) 94 (50.27)
Rarely/never 417 (42.95) 44 (40.00) 25 (40.32) 61 (52.59) 93 (49.73)
Shade at home
Often/sometimes 591 (60.87) 77 (69.37) 41 (66.13) 62 (52.99) 111 (59.36)
Rarely/never 380 (39.13) 34 (30.63) 21 (33.87) 55 (47.01) 76 (40.64)
Hat at school
Often/sometimes 554 (57.05) 66 (60.00) 37 (59.68) 55 (47.41) 94 (50.27)
Rarely/never 417 (42.95) 44 (40.00) 25 (40.32) 61 (52.59) 93 (49.73)
Hat at home
Often/sometimes 591 (60.87) 77 (69.37) 41 (66.13) 62 (52.99) 111 (59.36)
Rarely/never 380 (39.13) 34 (30.63) 21 (33.87) 55 (47.01) 76 (40.86)
0.05
0.48
*Race/ethnicity data were missing for 39 children; item non-response varied across variables.
0.004
0.008
0.023
0.072
0.075
Table 3-2. Prevalence of sun protective behaviors at baseline by race/ethnicity, SunSmart (N=1,488)*
N(%)
0.012
0.112
<0.0001

47

While prevalence of sun protection behaviors across the different ethnic and racial
categories were relatively similar, Hispanic children reported significantly greater use of
sunscreen at school "often/sometimes" than non-Hispanics. Asian and Asian-American students
reported significantly greater use of protective clothing (long sleeves and long pants)
"often/sometimes" for sun protection both at school and at home than other racial/ethnic
categories.
Significant differences were found for sunburn-related variables, with a greater
proportion of non-Hispanic white and Hispanic children reporting "ever" experiencing sunburn
than other ethnic/racial categories. In addition, Hispanic children had higher prevalence of
sunburn in the past month and past summer compared to non-Hispanic children.

Variable
Hispanic/Latino
n=972
Asian/Pacific
Islander
n=111
Non-Hispanic
White
n=62
African-
American
n=117
Mixed
race/other
n=187 χ2
Ever sunburned
Yes 603 (62.23) 51 (46.36) 42 (67.74) 41 (35.34) 91 (48.66)
No 366 (37.77) 59 (53.64) 20 (32.26) 75 (64.66) 96 (51.34)
How many times in
the past month were
you sunburned?
0 times 575 (59.28) 91 (82.73) 48 (77.42) 82 (70.69) 134 (72.04)
1 times 205 (21.13) 12 (10.91) 7 (11.29) 18 (15.52) 27 (14.52)
2-4 times 165 (17.01) 5 (4.55) 6 (9.67) 13 (11.21) 19 (10.22)
>4 times 25 (2.58) 2 (1.82) 1 (1.61) 3 (2.59) 6 (3.22)
How many times last
summer were you
sunburned?
0 times 401 (41.51) 67 (60.91) 23 (37.10) 75 (64.66) 104 (55.61)
1 times 274 (28.36) 22 (20.00) 22 (35.48) 28 (24.14) 37 (19.79)
2-4 times 239 (24.75) 17 (15.46) 14 (22.58) 7 (6.03) 33 (17.65)
>4 times 52 (5.38) 4 (3.64) 3 (4.84) 6 (5.17) 13 (6.95)
Table 3-3. Prevalence of sunburn-related variables at baseline by race/ethnicity, SunSmart (N=1,488)*
N(%)
*Race/ethnicity data were missing for 39 children; item non-response varied across variables.
<0.0001
<0.001
<0.0001

48
Sun protection behavior outcomes
Table 3-4 shows the results of the regression analyses for each sun protection behavior
outcome.

Correlates of sunscreen use: Demographic factors significantly associated with greater
use of sunscreen included female gender and Hispanic ethnicity. Psychosocial factors associated
with greater use of sunscreen included higher perceived norms of peer sun protection as well as
lower barriers to use of sun protection. Significant family-related factors included greater use of
sun protection by parent for child, discussion of sunscreen use with family in the past two weeks,
and having sunscreen available at home.
Sunscreen use
Protective
clothing Hat use Shade seeking
b (se)
1
b (se)
1
b (se)
1
b (se)
1
Girls (ref: boys) 0.10 (0.04)
**
0.14 (0.04)
**
−0.19 (0.04)
**
0.37 (0.04)
**
Hispanic (ref: non-Hispanic) 0.16 (0.04)
**
−0.01 (0.04) −0.05 (0.05) −0.04 (0.04)
Lighter skin phototype −0.02 (0.03) 0.003 (0.03) 0.007 (0.03) 0.09 (0.03)
**
Ever sunburned (ref: never) 0.05 (0.04) −0.09 (0.04)
*
0.10 (0.04)
*
0.05 (0.04)
Knowledge of sun protection^ 0.01 (0.01) 0.009 (0.01) −0.003 (0.01) 0.01 (0.01)
Barriers to sunscreen use^ 0.17 (0.03)
**
0.02 (0.03) −0.08 (0.03)
*
0.01 (0.03)
Barriers to protective clothing^ −0.04 (0.02) 0.04 (0.03) 0.32 (0.03)
**
−0.04 (0.03)
Perceived norms^ 0.14 (0.03)
**
0.05 (0.03) 0.02 (0.03) 0.17 (0.03)
**
Perceived self-efficacy^ 0.004 (0.03) 0.16 (0.03)
**
0.04 (0.04) 0.08 (0.04)
*
Perceived risk^ −0.02 (0.02) −0.01 (0.02) −0.02 (0.02) 0.02 (0.02)
Higher U.S. acculturation 0.007(0.01) −0.004 (0.01) 0.004 (0.01) −0.01 (0.01)
Family sun protection for child^ 0.31 (0.03)
**
0.19 (0.03)
**
0.17 (0.03)
**
0.15 (0.03)
**
Sunscreen available at home
(ref: maybe/not available) 0.27 (0.04)
**
−0.03 (0.04) −0.005 (0.05) 0.009 (0.04)
Family sunscreen discussion
(ref: did not discuss) 0.32 (0.05)
**
−0.003 (0.06) 0.10 (0.06) 0.09 (0.06)
Family clothing discussion
(ref: did not discuss) 0.02 (0.05) 0.11 (0.05)
*
0.14 (0.06)
*
0.08 (0.06)
Intercept 0.43 (0.16)
**
1.72 (0.18)
**
0.92 (0.19)
**
1.76 (0.19)
**
1
Unstandardized coefficients
^Higher scores denote protective responses
Table 3-4. Multivariate linear regression model for correlates of sun protective behaviors among
elementary schoolchildren, SunSmart (N=1,488)
P <
*
<0.05; P <
**
<0.01

49
Correlates of protective clothing: Female gender and higher self-efficacy to use sun
protection were significantly associated with greater use of protective clothing in the sample.
Ever having a sunburn was significantly associated with lower use of protective clothing. Both
use of family sun protection for child and family discussion regarding sun protective clothing in
the past two weeks were significantly associated with greater use of protective clothing.
Correlates of sun protective hats: Ever having sunburn, lower barriers to use of
protective clothing, greater use of family sun protection for child, and family discussion
regarding sun protective clothing in the past two weeks were significantly associated with greater
use of hats for sun protection. Female gender and lower barriers to sunscreen use were
significantly associated with less frequent hat use.
Correlates of shade seeking: Female gender, lighter skin phototype, greater perceived
peer sun protection norms, greater self-efficacy to use sun protection, and greater use of family
sun protection for child were significantly associated with greater use of shade for sun
protection.

Mediation analysis
We observed no significant direct association between acculturation and the sun
protection outcomes. However, the relationship between shade seeking and acculturation was
borderline significant (p=0.06) (Table 3-5). Therefore, analyses were restricted to the shade-
seeking outcome, and psychosocial variables were entered into the model as single mediators.

50

Table 3-6 shows the results of the path analyses for the mediating effects of psychosocial
factors on the relationship between acculturation and shade seeking.

A significant indirect effect was found for only one variable, family sun habits for child.
Higher level of acculturation was significantly negatively associated with higher family sun
habits, while higher family habits were significantly positively associated with increased shade-
b (se) p-value
Sunscreen use -0.008 (0.01) 0.51
Protective clothing use -0.002 (0.01) 0.83
Hat use 0.003 (0.01) 0.81
Shade seeking -0.02 (0.01) 0.06
Table 3-5. Bivariate associations between acculturation and sun
protection among the Hispanic subsample, SunSmart (N=972)
Mediating variable Unst. SE z p
Perceived risk
Accult −> Perceived risk -0.01
0.01 -0.98 0.33
Perceived risk −> shade seeking 0.05
0.03 1.92 0.05
Accult −> shade seeking -0.02
0.01 -1.77 0.08
Indirect effect -0.0001
0.001 -0.87 0.39
Shade self-efficacy
Accult −> shade self-efficacy -0.01
0.01
-0.55 0.58
Shade self-efficacy −> shade seeking 0.26
0.02
12.74 <.0001
Accult −> shade seeking
-0.02 0.01 -1.72 0.09
Indirect effect
-0.002 0.003 -0.55 0.59
Shade peer norms
Accult −> shade peer norms
-0.01 0.01 -0.62 0.54
Shade peer norms −> shade seeking
0.16 0.02 10.69 <.0001
Accult −> shade seeking
-0.02 0.01 -1.59 0.11
Indirect effect
-0.001 0.002 -0.63 0.53
Family sun habits for child
Accult −> family habits
-0.02 0.01 -2.15 0.03
Family habits −> shade seeking
0.24 0.02 10.11 <.0001
Accult −> shade seeking
-0.02 0.01 -1.42 0.16
Indirect effect
-0.005 0.002 -2.09 0.04
Table 3-6. Path coefficients and indirect effects for mediating variables between
acculturation and shade-seeking among the Hispanic subsample, SunSmart (N=972)

51
seeking. The indirect effect was significant and represented a negative association, suggesting
that higher acculturation negatively influences the relationship between family sun protection
habits for child and child’s shade-seeking behavior. The direct association between acculturation
and shade-seeking (c' path) was not significant, suggesting full mediation.

Discussion
We examined correlates of sun protection in a sample of predominantly ethnic minority,
largely Hispanic schoolchildren. Overall, we found that ethnic minority children experience
similar sun protection correlates previously observed among NHW children. Female gender, for
example, was significantly positively associated with use of sunscreen, protective clothing and
shade seeking but negatively associated with hat use in the present study, a pattern found in prior
studies amongst NHW children (Buller, Cokkinides, et al., 2011; Robinson et al., 1997).
Psychosocial constructs such as lower barriers, higher perceived descriptive norms, and greater
perceived self-efficacy were all significantly associated with greater sun protection in our study,
as in prior studies among NHW children (Alberg et al., 2002; Reynolds et al., 2006; Turner &
Mermelstein, 2005). In addition, although Hispanic children in this sample reported using
sunscreen at school more often than other ethnic and racial categories including NHWs, we
found high reported rates of sunburn among Hispanic children.
Overall, the most consistent predictor across all sun protection methods in the current
study was family use of sun protection for child, similar to studies conducted among NHWs that
have demonstrated the strong influence of parental variables on children’s sun protection
(Thomson et al., 2012; Turner & Mermelstein, 2005). This finding suggests that while
racial/ethnic tailoring might still be important for child sun protection, intervention strategies that

52
focus on family engagement and target parental norms for children’s sun protection may be
effective for and relevant to broad multi-ethnic audiences.
In addition, sunscreen availability at home was significantly associated with greater use
of sunscreen by child. Promoting sunscreen to parents, then, may boost its application for young
children, a finding with implications for Hispanics and other ethnic minority groups where
sunscreen use has been found to be low despite high sunburn rates (Coups et al., 2012).
However, accessibility to sunscreen may differ between ethnic minorities and NHWs; several
studies have found reduced availability of sunscreen for Hispanics and other ethnic minority
groups compared to NHW neighborhoods (Dubas & Adams, 2012; C. Hernandez et al., 2012).
Thus, further research on the influence of sunscreen availability on child sun protection should
examine both immediate household and neighborhood accessibility among ethnic minority
populations to determine the most effective approaches for increasing sunscreen use. In addition,
intervention and policy strategies to increase accessibility of sunscreen within Hispanic
neighborhoods are warranted.
Despite similarities with prior research, we note the following differences. In contrast to
other studies, we observed a stronger association between Hispanic ethnicity and sunscreen use
in comparison to non-Hispanics (Ma et al., 2007). This finding may have resulted from the
comparison group, which comprised all non-Hispanic children including 8% African-American
and 11% mixed race children, groups that may use less sunscreen than Hispanics (Pichon,
Corral, Landrine, Mayer, & Norman, 2010). However, prevalence rates of frequent sunscreen
were highest for Hispanic children in this sample than for all other ethnic/racial groups,
including NHWs.

53
Unlike prior studies among NHW samples where lighter skin phototype and skin
sensitivity to sunburn have been found to be predictors of greater sun protection (Clarke,
Williams, & Arthey, 1997; Thieden, Philipsen, Sandby-Moller, & Wulf, 2005), neither physical
attribute was strongly associated with increased protective behaviors across all of the four
outcomes. Lighter skin phototype was associated only with higher shade-seeking among children
in the sample, while ever being sunburned was associated with higher use of sun protective hats
only. While these findings do indicate some behavioral differences among children with greater
phenotypic risk factors, they suggest suboptimal sun protection behaviors among these groups
for both UV overexposure and sunburn. Subsequently, it may be important to target at-risk
subgroups among ethnic minority children (e.g., lighter skin phototype and more sun reactive
skin) when developing sun protection risk communication and intervention strategies.
In support of our hypotheses, we found distinct predictors for the four different sun
protection methods. While lower barriers to use of sunscreen was positively associated with
greater sunscreen use, lower barriers to protective clothing use was associated only with the use
of sun-protective hats (and not with use of long sleeves/pants). It is probable that wearing hats
for sun protection presents greater challenges for children, as the use of hats is less normative
than other barrier methods of clothing and requires greater effort. As hat use is an important
method of sun protection that has been particularly difficult to change in the intervention setting
particularly outside of school (Hunter et al., 2010), future research might focus on identifying
salient barriers and targeting these for greater intervention efficacy in increasing this effective
barrier method.
In addition, greater perceived descriptive norms were associated with more frequent use
of sunscreen and shade seeking only, and not with protective clothing or hat use. Sunscreen use

54
and shade seeking may be closely associated to peer social networks and friendship proximity
(e.g., children may use sunscreen with peer at school, or may seek shade when playing together).
Future research, which more closely investigates peer norms and social networks, may yield
further insight into the specific mechanisms by which peers exert influence on child sun
protection.
Although an important predictor in several other studies regarding sun protection among
Hispanics (Andreeva et al., 2009; Coups E.J., 2013), acculturation was not significantly
associated with any of the sun protection outcomes in the present study. Nor did our mediation
analysis reveal substantial findings regarding the indirect influence of acculturation on
psychosocial and familial predictors save for one indirect effect in which higher acculturation
negatively influenced shade-seeking mediated by family habits. This lack of effect could be
explained in part by low variability in the data, as the majority of children were relatively low-
acculturated, with a low mean score on the AHIMSA scale. However, as studies that have found
differences in sun protection by level of acculturation have been conducted exclusively amongst
Hispanic adults, the effects of acculturation could differ for children. Moreover, as acculturation
is a dynamic and often family-centered process, its influence on child sun protection may be
more effectively examined within a familial and generational context and assessed over time.

Limitations and strengths of the research
The use of cross-sectional data limited causal inference, as only associations among
variables can be asserted. Self-reported measures of sun protection, though the only feasible
method for capturing data in the present study, may not accurately represent children’s actual

55
usage. However, studies have suggested that self-reported sun protection behaviors provide
sufficiently valid reports compared to objective data (O'Riordan et al., 2009).
In this study, “Hispanic” is considered a uniform category; however, Hispanic ethnic
origin is heterogeneous, with varying subgroups, and recent studies have indicated that sun
protection may differ by Hispanic subpopulation (Coups E.J., 2013). Due to demographic data
for Los Angeles, we assume that the majority of students who self-reported as “Hispanic/Latino”
in the sample were Mexican-American based on available census data (~75% of Hispanics in LA
County are Mexican-American) (United States Census Bureau, 2013). However, further research
is needed to examine potential differences in children’s sun protection among Hispanic
subgroups.
Despite these limitations, strengths of the study include a large sample of predominantly
ethnic minority children, an understudied population in sun protection. In addition the utilization
of previously validated core questionnaire items and inclusion of several domains (psychosocial,
familial, cultural) is a unique strength of the study.

Conclusion
In conclusion, the findings from the present study indicate both significant overlap in
factors associated with sun protection between NHWs and ethnic minority youth, with several
differences that attest to the variability of child sun protection and the potential influence of
ethnic and cultural factors on children’s sun behaviors. In addition, Hispanic children reported
rates of sunburn higher than NHW children in this sample, indicating their high-risk status with
respect to UV overexposure.

56
These findings are particularly important given the changing demographics of the United
States. Notably, exponential increases in the U.S. Hispanic population (Ennis, Ríos-Vargas, &
Albert, 2011) accompanied by rising rates of skin cancer among Hispanics and disparities in
outcomes will necessitate greater public health focus on primary prevention strategies that extend
broadly to multi-ethnic audiences. The present study provides evidence that previously identified
correlates of sun protection may generalize across cultural contexts to be scaled into future sun
safety health promotion for diverse youth audiences. However, future research should continue
to investigate the extent to which culturally specific messages and strategies might most
effectively and durably improve child sun protection in the intervention setting.

57
Chapter 4: Study 2

Mediators of sunscreen use in a skin cancer prevention intervention among multi-ethnic
elementary schoolchildren

Kimberly A. Miller, MPH
1
, Jimi Huh, PhD
1
, Jennifer B. Unger, PhD
1
, Jean L. Richardson,
DrPH
1
, Martin W. Allen, PhD
2
, David H. Peng, MD
3
, Myles G. Cockburn, PhD
1,3

1
Department of Preventive Medicine, Keck School of Medicine of the University of Southern
California, Los Angeles, CA
2
MacDiarmid Institute for Advanced Materials and Nanotechnology, Department of Electrical
and Computer Engineering, University of Canterbury, Christchurch, New Zealand
3
Department of Dermatology, Keck School of Medicine of the University of Southern California,
Los Angeles, CA.

Contact:
Kimberly A. Miller
Department of Preventive Medicine, Keck School of Medicine
2001 N Soto St, Suite 318-A
Los Angeles, CA 90032
Phone: (626) 864-6433 / Fax: (323) 865-0095
Email: kim.miller@med.usc.edu

58
Abstract
In contrast to other cancer types, incidence rates of melanoma are increasing. Because
excessive exposure in childhood is an important risk factor for development of melanoma in
adulthood, children and adolescents represent a priority population for primary prevention
interventions. While sun safety interventions have proliferated in the past twenty years, few
studies have formally examined mediators of intervention effects.
In this study we assessed five psychosocial constructs hypothesized to be mediators
between program effects and sunscreen use in a large, multi-ethnic sample of 4th and 5th grade
children participating in a sun safety intervention. Two intervention conditions were assessed: 1)
classroom education only and 2) classroom education + UV feedback/experiential learning.
Mediators included knowledge, perceived risk, perceived self-efficacy, perceived peer norms,
and reduced barriers to sunscreen use.
Mediation was tested using change score analysis from Time 1 (pretest) to Time 2
(posttest), and with autoregressive cross-lagged models to assess changes over a longer follow-
up period, from Time 1 to Time 3 (3 months from baseline). In panel change score analyses,
indirect effects of reduced sunscreen barriers, perceived peer norms, and perceived self-efficacy
were observed. In cross-lagged analyses, these effects were attenuated and no longer significant,
and only knowledge emerged as a mediator for longer-term follow-up for the education + UV
feedback condition.
We found evidence that although constructs such as peer norms, barriers, and self-
efficacy may mediate program effects in the short-term, as time from the intervention passes
these influences appear to diminish. However, in longer-term follow-up, knowledge of learned

59
concepts may prove to exert increasing influence over children's sun protection behaviors. The
findings of this study add to the knowledge of the psychological mechanisms motivating
children's use of sunscreen and sun protection, an understudied area.

60
Introduction
In contrast to many other cancer types, incidence rates of malignant melanoma, the most
lethal form of skin cancer, are increasing (International Agency for Research on Cancer, 2010).
Melanoma prevention, achieved by both primary prevention (reduction of excessive exposure to
ultraviolet radiation) and secondary prevention (skin examination by self or physician) is
therefore a key public health issue. Children and adolescents represent a priority population for
melanoma primary prevention, as high ultraviolet exposures and sunburn in early life have been
associated with increased risk of melanoma in adulthood (D. C. Whiteman et al., 2001).
Because of the link between childhood sun exposure and risk of melanoma, sun safety
programs directed towards children and adolescents have proliferated over the past 20 years,
particularly in school settings where children spend a large amount of time outdoors during peak
UV hours. In a review by the Task Force on Community Preventive Services, sufficient evidence
for sun safety intervention was found in the primary school setting for improving sun protective
behaviors such as use of protective clothing and sunscreen use (Saraiya et al., 2004), although
results of interventions continue to be inconsistent in changing behaviors. The need for effective
interventions continues for this age group, and new approaches are being tested to improve and
to sustain sun protective behaviors over the long-term to reduce skin cancer risk.
While many school-based studies have evaluated the direct effects of sun protection
interventions on child sun protection behaviors, few have examined mediators of program
effects. The systematic understanding of mediators—variables that intervene in and transmit
effects between the independent and dependent variables—provides evidence of the utility and
salience of the theoretical frameworks of interventions and enables the identification of
psychosocial targets for behavior change (D. P. Mackinnon & Dwyer, 1993). A better

61
understanding of the psychological mechanisms involved in child sun protection is an important
and necessary step in creating more effective and durable interventions.
While a few studies have examined general mediators of sunscreen use among children
and adolescents in cross-sectional analyses (Andreeva VA, 2008; Schuz, 2012), longitudinal
analysis of mediation in the context of intervention is uncommon in the children's sun protection
literature. In a comprehensive study by Reynolds et al., a range of psychosocial mediators were
examined from pretest to posttest in the context of a large-scale sun protection intervention for
middle school children (Reynolds et al., 2006). The study found significant indirect effects for
perceived barriers to use of sunscreen, knowledge of sun protection, and perceived self-efficacy
to use sun protection in single mediator models. However, the authors acknowledged that the
these effects may not hold with longer follow-up.
In a 24-month study, Adams et al. tested decisional balance as a mediator of program
effects in a primary care-delivered sun safety intervention targeting adolescents ages 10-16 years
(Adams et al., 2009). The study measured whether changes in adolescent evaluation of the pros
of sun protective behaviors mediated intervention outcomes, hypothesizing that improved
decisions based on an understanding of the consequences of UV overexposure would lead to
greater protective behavior. The authors found a significant mediated effect, suggesting that
improving adolescent decision-making skills may be an important component for sun protection
intervention in teens.
Existing studies indicate that psychosocial constructs may play an important mediating
role in child sun protection interventions, leading to improved outcomes. However, further
research is needed to formally evaluate mediation in the intervention context to determine which
constructs are most salient to target, and which might lead to sustained increases in targeted

62
behavior over time. Such information may enable the design of interventions that more
effectively improve sun behavior in children, with long-lasting outcomes.

Present study
The present study examined potential mediators of sunscreen use in the context of a sun
safety intervention targeted to elementary school-aged children (SunSmart). We assessed the
short terms effects of mediation from pre- to posttest, as well as longer effects over three waves
of measurement (approximately 3 months from baseline). The aim of the study was to identify
constructs important in the path between the intervention and sun protection outcomes, and to
evaluate whether such changes were sustained several months post-intervention. Our study
sample comprised a unique multi-ethnic cohort of children residing in a high UV environment
(Los Angeles County).
In the SunSmart intervention, two strategies to increase sun protection in elementary
school-aged children were implemented: 1) classroom-based sun protection education
(curriculum condition); and 2) education plus a customized feedback intervention and
experiential learning laboratory using dosimeters to measure ground-level UV (dosimetry
condition). Children randomized to the curriculum condition received three one-hour lessons
during regular class time, covering the hazards of excessive UV exposure and methods to protect
themselves from risky exposure. Lessons were classroom-based and primarily didactic, with
interactive components to engage children’s interest. Children randomized to the dosimetry
condition received the curriculum condition plus an additional one-hour lesson in which they
measured UV in small teams on their playground, downloaded data, and read and interpreted
graphs of their UV measurements. The graphs, which were calibrated to the UV Index, provided

63
a visual presentation of areas of high UV exposures (e.g., on the blacktop) and lower exposures
(e.g., in the shade/under trees). For the third program condition, observation, children received
assessment with the same schedule as the other two conditions only.
While improvements in sun protection behaviors were hypothesized for both intervention
conditions, stronger effects were expected for the dosimetry condition due to the novel
feedback/experiential learning component. Feedback-based interventions that provide objective
data to participants have been found to be effective in changing health behaviors in the context of
skin cancer prevention among adults (Emmons et al., 2011). Further, the hands-on, student-led
learning style of the dosimetry intervention was designed to make sun safety a tangible and
relevant issue for children, enabling the visualization of both high-risk and safe exposures in
their daily environment leading to greater effects across all measures. We thus hypothesized that
the curriculum plus dosimetry intervention would yield stronger indirect effects than the
curriculum only condition.
Our outcome for the present study was sunscreen use. While sun protection behavior in
past studies has been frequently analyzed as a composite variable (Adams et al., 2009; Buller,
Reynolds, et al., 2006; Reynolds et al., 2006), recent studies suggest that different sun protection
methods represent distinct behaviors (Tripp et al., 2003). Exploratory factor analysis of the data
indicated a separate factor for sunscreen use and therefore, we limited our outcome to use of
sunscreen due to measurement specificity. In addition, although recent studies have emphasized
the greater efficacy of barrier methods of sun protection such as long sleeves, long pants, and
hats, sunscreen remains an important and effective component of sun protection (Viros et al.,
2014), particularly in high UV regions such as California where the present study was conducted.

64
Hypothesized mediators were drawn both from the sun protection literature as well as
from the theoretical basis of the intervention, which engaged key constructs from the Integrated
Behavioral Model and Social Cognitive Theory (Bandura, 1986; Malotte et al., 2000).
Psychosocial variables believed to be influential included knowledge of sun protection concepts,
perceived risk for long-term consequences of UV overexposure, perceived peer norms for sun
protection, reduced barriers to sunscreen use, and self-efficacy to use sun protection. Of these,
knowledge, self-efficacy to use sun protection, and barriers to sunscreen were identified as
mediators of program effects improving sun protection (Reynolds et al., 2006).
As direct effects of social norms on the sun protection practices of children have been
found across several studies (Abroms et al., 2003; Arthey & Clarke, 1995; Cody & Lee, 1990;
Lowe, Balanda, Gillespie, Del Mar, & Gentle, 1993), we also examined perceived peer norms for
sunscreen use. In contrast to past studies that have primarily measured injunctive norms, we
focused on descriptive norms, in which children observe, evaluate, and imitate popular
behaviors, particularly if they share affiliation or affinity with the referent group exhibiting the
behaviors. Because injunctive norms may be weak for sun protection in this age group, we
hypothesized that descriptive norms may be a more salient mediator of program effects for
elementary school-aged children (Cialdini, Reno, & Kallgren, 1990; Rimal, Lapinski, Cook, &
Real, 2005).
In addition, we examined perceived risk or susceptibility to long-term UV consequences,
which has been shown to play a mediating role in determining adult sunscreen use (Craciun,
Schuz, Lippke, & Schwarzer, 2012). While such findings may not generalize to child sun
protection, we hypothesized that program exposure would increase susceptibility to harm from
excessive UV leading to increased sunscreen use, and in particular the dosimetry intervention

65
would increase perceived risk through the visual presentation of personalized information (e.g.,
students' own exposures).

Methods
Sample
Data from the SunSmart study was used. Details regarding the recruitment and
randomization of the sample are detailed in Chapter 2: Methods. Briefly, participants were
recruited from 11 public schools in Los Angeles County for a school-based randomized
controlled trial to improve sun safe behaviors. Students were assessed at three time points: at
baseline before the start of intervention activities; directly after the intervention at one month;
and again three months from baseline. Students were assessed during regular classroom time by
trained research staff using pencil-and-paper surveys. The same procedures and timing were used
for all program conditions.
The entire SunSmart sample consisted of 1,754 students. Because one school was not
able to complete measurement at Time 3 due to scheduling issues (N=185) the present study was
conducted with the 10 schools that completed three waves of measurement (N=1,569). No
differences were found between the 185 children who did not complete measures at Time 3 and
those who completed all waves of measurement on demographic characteristics or key baseline
sun protection variables.

66
Measures
Outcome
Sunscreen was modeled as a continuous (interval ratio) variable comprising an average of
three items assessing the use of sunscreen at school, outside of school, and general sunscreen
use. Response options ranged from 1="never" to 4="often," with higher values indicating greater
use of sunscreen (Cronbach's alpha: pretest α= 0.73; posttest α=0.82; post-posttest α=0.85).

Mediators
Knowledge:
Nine variables assessed students’ knowledge of sun protection concepts.
Response options included “true”, “false”, and “I don’t know.” Correct responses were summed
across the nine questions into a single summary variable with higher scores reflecting greater
knowledge of sun protective concepts.
Perceived Risk: Perceived risk of UV-related consequences was assessed with two
averaged items that measured students' perceived susceptibility to skin cancer and skin damage
from too much sun when older. Response options ranged from 1="untrue" to 4="true," with
higher values indicating greater perceived risk.
Barriers to use of sunscreen: Barriers to sunscreen were assessed with four averaged
items that measured potential attitudinal barriers to students' use of sunscreen (e.g., too messy,
too difficult, too difficult too chose, too bothersome to skin). Items were reverse-coded so that
higher scores reflected lower barriers to sunscreen use. Response options ranged from
1="untrue" to 4="true."

67
Perceived peer norms for sunscreen use: Perceived peer norms were was assessed with a
single item that measured students' observations that their friends used sunscreen. Response
options ranged from 1="untrue" to 4="true," with higher scores indicating higher perceived peer
norms.
Perceived self-efficacy to use sunscreen: Perceived self-efficacy to use sunscreen was
assessed with a single item that measured students' sense of self-to use sunscreen with SPF 15 or
above. Response options ranged from 1="I’m sure I can’t" to 4="I’m sure I can," with higher
values indicating higher self-efficacy.

Statistical analysis
Initial univariate and bivariate analysis and descriptive statistics were performed to
investigate means and frequencies. Program condition was compared for any significant
difference between groups. Because grade level (4th, 5th, or mixed 4th/5th) was found to be
significantly different (p<.0001), with a greater proportion of mixed 4th/5th grade classes in the
curriculum condition and a greater proportion of 5th grade classes in the observation condition,
grade level was adjusted for in all models. In addition, intraclass correlation coefficients (ICCs)
were calculated for all variables, and found to be <0.02. Because the effect of clustering
appeared negligible in the sample (ICC near zero), a multilevel analysis was not performed.
To assess mediation from Time1 to Time 2 of the intervention, change scores were
calculated by subtracting the baseline from posttest follow-up for mediators and outcome. Path
analyses were conducted to examine the indirect effect of the intervention on the outcome via
each mediator. Exposure to the intervention program was dummy-coded, and each program
condition (dosimetry and curriculum) was compared to the observation group.

68
Following contemporary recommendations for mediation analysis, we relaxed the
requirement for the prior significance of the direct effect of x on y (path c’), focusing on the
product of path coefficients "a," or independent variable to mediator, and "b," (mediator to
outcome) (D. MacKinnon, 2008). Bias-corrected bootstrap confidence intervals were used for
significance testing of the indirect effect as they correct for the potential non-normality of the
distribution of the mediated effect (Preacher & Hayes, 2004). Mediation was established if
confidence intervals for the indirect effect (calculated as the product of path coefficients "a," or
independent variable to mediator, and "b," mediator to outcome) did not overlap zero.
To examine the data over three waves of assessment, autoregressive cross-lagged path
models with observed variables were conducted to evaluate the relationships between mediators
and sunscreen use over time (Cole & Maxwell, 2003). Such models are able to capture temporal
precedence between the mediator and the outcome, where the mediator measured at the prior
time point may be associated with the outcome at a subsequent time point, controlling for prior
levels of each variable and allowing measures to covary at each wave to reflect potential
contemporaneous relationships between the variables (D. MacKinnon, 2008) (see Figure 4-1).
We assessed the indirect effect of the intervention program on the outcome measured at Time 3
via the indirect effect of the putative mediator measured at Time 2.

69

For the cross-lagged mediation model, model fit was evaluated with the χ2 goodness-of-
fit statistic, the comparative fit index (CFI), and the root mean square error of approximation
(RMSEA). A non-significant chi-square statistic, CFI > .90, and RMSEA < .05 are generally
considered to represent adequate fit of the model (Kline, 2011).
All coefficients presented were standardized. All analyses were conducted using Stata
version 12 (StataCorp, 2011) and Mplus version 6 (L. K. Muthén, & Muthén, B. O., 1998-2011).

Results
Characteristics of the sample are presented in Table 4-1.
Treatment condition
Dosimetry vs. observation
Curriculum vs. observation
a’
c’
b’
Mediator))
Time)1)
Mediator))
Time)2)
Mediator))
Time)3)
Sunscreen)use)Time)1)
Sunscreen)use)Time)2) Sunscreen)use)Time)3)
Fig. 4-1: Cross-lagged mediation model, SunSmart

70

The results of the path analyses for mediation using change scores from pre- to posttest
are presented in Table 4-2. For both the dosimetry and the curriculum condition, exposure to the
program significantly increased all mediators from pretest to posttest (path a’); effects were
comparable for both groups, with slightly greater increases for the curriculum group in all
mediators except for reduced sunscreen barriers. For both groups, reduced sunscreen barriers and
perceived peer norms were significantly associated with increased sunscreen use (path b’); for
the curriculum group, self-efficacy to use sunscreen was also significantly associated with the
outcome. Indirect effects with non-overlapping confidence intervals were found for both
conditions for reduced sunscreen use and perceived peer norms; an indirect effect for self-
efficacy to use sunscreen was marginally significant for the curriculum condition only.

Overall % Dosimetry Curriculum Observation
Mean age (SD) 10.85 10.86 10.81 10.91
% % % %
Gender
Female 46.4 47.0 46.3 44.6
Male 47.2 47.8 46.3 48.36
Missing 6.4 5.2 7.4 7.04
Grade Level
4th 43.7 45.9 43.3 38.03
5th
47.1 50.8 38.5 61.97
Mixed 4th/5th
9.2 3.3 18.2 0
Race/Ethnicity
Hispanic 59.6 64.4 49.5 75.59
Asian/Pacific Islander 6.5 6.5 8.1 1.41
Non-Hispanic White 4.3 2.8 5.7 5.16
African-American 8.2 5.8 12.8 1.88
Native American/American Indian 2.1 1.6 2.6 2.35
Mixed race/other 10.6 11.1 11.4 6.57
Missing 8.6 7.81 9.92 7.04
Table 4-1. Demographic characteristics of sample at baseline, SunSmart (N=1,569)

71

Table 4-3 presents fit statistics for the cross-lagged mediation path analyses. For each
mediator model, model fit was observed to be mediocre to adequate. While all models achieved
>0.90 on the CFI, none of the path models achieved RMSEA >0.05 and all had a significant chi-
square.
Dosimetry vs. observation
Knowledge 0.368*** 0.025 0.073* 0.009 ˗0.026, 0.045
Perceived risk 0.231*** ˗0.008 0.085* -0.002 ˗0.018, 0.015
Reduced sunscreen barriers 0.177*** 0.134*** 0.06^ 0.024** 0.008, 0.040
Peer normsǂ 0.23*** 0.082* 0.019 0.019* 0.001, 0.038
Perceived self-efficacyǂ 0.153*** 0.031 0.033 0.005 ˗0.008, 0.017
Curriculum vs. observation
Knowledge 0.398*** 0.077 0.012 0.031 ˗0.009, 0.070
Perceived risk 0.272*** 0.057 0.03 0.016 ˗0.008, 0.039
Reduced sunscreen barriers 0.118** 0.14** 0.029 0.017* 0.002, 0.031
Peer normsǂ 0.271*** 0.104** ˗0.005 0.028* 0.004, 0.053
Perceived self-efficacyǂ 0.121** 0.11* 0.01 0.013^ ˗0.001, 0.027
ǂ=Categorical variable
Coefficients are standardized; models adjust for grade level.
^=p<.10; * = p<.05; **=p<.01; ***p<.001
Table 4-2. Coefficients and indirect effects with bias-corrected bootstrap confidence intervals for change scores, Pre to
Posttest (N=1,569)
Mediating variable
Program effect on
mediator
Mediator on
sunscreen use
Direct effect
from program
to outcome
Indirect
effect
95% CI for
indirect effect

72

The results of the cross-lagged path analyses are presented in Table 4-4. As with the
change score model from pre- to posttest, exposure to the SunSmart program significantly
increased each mediating construct with the exception of perceived peer norms in both treatment
groups. Only knowledge of sun protection in the dosimetry group was significantly associated
with sunscreen use at Time 3; accordingly, only knowledge was found to be a significant
mediator between program exposure and sunscreen use for the dosimetry group only.
Variable χ2(d.f.) CFI RMSEA
Dosimetry vs. observation
Knowledge 70.509 (7)*** 0.95 0.10
Perceived risk 67.952 (7)*** 0.93 0.09
Reduced sunscreen barriers 67.034 (7)*** 0.95 0.09
Peer normsǂ 58.917 (13)*** 0.97 0.06
Perceived self-efficacyǂ 86.988 (13)*** 0.96 0.08
Curriculum vs. observation
Knowledge 39.397 (7)*** 0.97 0.07
Perceived risk 25.997 (7)*** 0.97 0.06
Reduced sunscreen barriers 79.414 (7)*** 0.94 0.11
Peer normsǂ 113.512 (13)*** 0.95 0.09
Perceived self-efficacyǂ 61.024 (13)*** 0.97 0.07
ǂ=Categorical variable
Table 4-3. Fit statistics for cross-lagged autoregressive models of mediators of
sunscreen use (N=1,569)
CFI = Comparative Fit Index; RMSEA = Root Mean Square Error of Approximation
* = p<.05; **=p<.01; ***p<.001
Overall model fit

73

Discussion
We assessed short-term and longer-term mediators in a school-based sun safety
intervention targeting 4th and 5th grade schoolchildren. The results of the study found evidence
for the mediating influence of reduced sunscreen barriers, perceived peer norms for sunscreen
use, and perceived self-efficacy to use sunscreen for the shorter follow-up period from pre- to
posttest. However, over longer-term follow-up, these effects were not sustained. In the cross-
lagged analysis that provides clearer temporal precedence over a longer time frame, only
knowledge emerged as a mediator for children, and in the dosimetry condition only. Thus, while
our findings suggest that these variables are modifiable by intervention and that they can in turn
increase sunscreen use, their influence to motivate protective behaviors appears to wane as time
from the intervention increases.
Dosimetry vs. observation
Knowledge 0.409*** 0.086* -0.039 0.035*
(0.012, 0.059)
(0.012, 0.059)
Perceived risk 0.251*** 0.041 0.005 0.01 (−0.006, 0.027)
Reduced sunscreen barriers 0.184*** -0.009 0.018 -0.002 (-0.020, 0.016)
Peer normsǂ 0.068 0.04 0.019 0.003 (˗0.023, 0.028)
Perceived self-efficacyǂ 0.255*** 0.014 0.016 0.004 (−0.013, 0.020)
Curriculum vs. observation
Knowledge 0.415*** 0.042 -0.001 0.017 (−0.006, 0.041)
Perceived risk 0.277*** 0.068 0.05 0.044 (-0.014, 0.102)
Reduced sunscreen barriers 0.151*** -0.019 0.032 ˗0.003 (˗0.203, 0.058)
Peer normsǂ 0.067 0.213 0.03 0.014 (-0.019, 0.052)
Perceived self-efficacyǂ 0.24*** 0.007 0.035 0.002 (-0.01, 0.004)
ǂ=Categorical variable
^=p<.10; * = p<.05; **=p<.01; ***p<.001
Direct effect
from program
to outcome
95% CI for
indirect effect
Table 4-4. Coefficients and indirect effects with bias-corrected bootstrap confidence intervals, cross-lagged analysis Pre
to Post-Posttest (N=1,569)
Mediating variable
Program effect on
mediator
Mediator on
sunscreen use
Indirect
effect
Coefficients are standardized; models adjust for grade level.

74
Knowledge of sun protection, however, which was not a mediator from pre- to posttest,
emerged as a significant mediator in the cross-lagged model for the dosimetry condition only.
This "delayed" effect may result from the time required for students to operationalize learned
concepts with a resultant strengthening of the mediation pathway. That this relationship was
significant for the dosimetry group only suggests that the feedback-based, experiential learning
component may be more influential than the classroom-only approach over the long-term, as this
component may boost learned concepts and more readily motivate children to put theoretical
understanding into practice (Wurdinger & Carlson, 2010).
Overall, while the SunSmart intervention was effective in modifying psychosocial
constructs, increases in these variables were not strongly linked to sunscreen use at either short-
and longer-term follow-up. This finding may indicate that more complex causal pathways are
involved in children's sunscreen use, and/or that other variables not included in the analysis
might serve to directly or indirectly influence child sun protection, e.g. parental or environmental
factors. In addition, our sample was multi-ethnic and predominantly Hispanic. While we
identified psychosocial constructs from the existing sun protection literature, it is possible that
these variables, which have been derived from majority non-Hispanic white populations, do not
generalize to Hispanic and/or multiethnic cohorts. While our findings are consistent at least in
part with one prior study conducted in a predominantly NHW sample of middle-school aged
children (Reynolds et al., 2006), further research is needed to determine if mediators of child
sunscreen use differ by ethnicity and race.

75
Limitations and strengths of the research
While inference from cross-lagged panel mediation models has advantages over both
cross-sectional and panel mediation model, measurement error due to self-reported measures
may have compromised model adequacy, attenuating or overestimating the true magnitude of
mediated effects (D. MacKinnon, 2008). Our cross-lagged model fit was only mediocre; while
modification indices were examined, items for model fit improvement that were suggested were
not conceptually appropriate, and the original models were maintained. Inclusion of additional
variables in more complex multiple mediator path models may yield improved fit in future
analyses. In addition, while our study design with three waves of measurement improved over
past panel analyses, the follow-up period was still relatively brief, comprising one school
semester (February-May). As with all studies, results must be replicated, and generalization is
limited to elementary schoolchildren of similar racial/ethnic composition.
Strengths of the study include a longitudinal design allowing for temporal precedence of
the predictors relative to the mediators, as well as a large sample size required for asymptotically
robust models required by the cross-lagged design. Further, our study is one of very few existing
studies in the children's sun protection literature assessing mediation within the context of a large
sun safety intervention, and contributes to the literature on this understudied and important issue.

Conclusion
Overall, the findings of the present study suggest that modifying psychosocial variables
for elementary schoolchildren may have limited and short-term significance in improving
behaviors after intervention effects. Boosters may be required that reinforce these constructs if
they are to continue to exert influence on sun protection behaviors as distance from the program

76
increases. In addition, the targeted constructs presented here, though amenable to improvement
by intervention, may not be the critical ones related to child sun protection, and/or child sun
protection may be shaped more directly by parental factors, availability of sun protection, or
other environmental/social prompts. Further research should be undertaken to include a wider
array of potential variable to examine their direct and indirect effects on child sun protection
practices.
Consistent with several other studies, knowledge emerged as an important mediator
between exposure to the intervention and improved sunscreen use in longer-term follow-up.
While direct effects of knowledge have been noted in prior studies among adolescents (Alberg et
al., 2002; Andreeva VA, 2008), our study indicates that improving knowledge of sun protection
in children may grow in influence over time as children put sun protection concepts into practice;
further, our results suggest that sun protection interventions which emphasize hands-on,
experiential learning that "shows" rather than "tells" students may improve knowledge-related
outcomes leading to desired behaviors (Wurdinger & Carlson, 2010). Indeed, interventions in
other health prevention contexts for adolescents such as oral health education, healthy eating,
and diabetes self-management have shown that experiential learning substantially boosts
intervention efficacy over didactic comparators (Angelopoulou, Kavvadia, Taoufik, & Oulis,
2015; Dudley, Cotton, & Peralta, 2015; Spencer, Cooper, & Milton, 2013). Thus, future
interventions might incorporate active and experiential learning strategies to increase and sustain
behavior change in the field of child sun protection. Further research might also investigate
multiple mediation models to better understand the complex causal pathways involved in
children's sun behaviors.

77
In conclusion, our study is one of the few to formally and prospectively examine
mediators of child sun protection. These analyses have identified short-term and longer-term
program effects, and may serve as a guide to the conception and design of future interventions to
improve sun safety in children, a priority population for melanoma prevention.

78
Chapter 5: Study 3

Patterns of sun protective behaviors among Hispanic children in a skin cancer prevention
intervention

Kimberly A. Miller, MPH
1
, Jimi Huh, PhD
1
, Jennifer B. Unger, PhD
1
, Jean L. Richardson,
DrPH
1
, Martin W. Allen, PhD
2
, David H. Peng, MD
3
, Myles G. Cockburn, PhD
1,3

1
Department of Preventive Medicine, Keck School of Medicine of the University of Southern
California, Los Angeles, CA
2
MacDiarmid Institute for Advanced Materials and Nanotechnology, Department of Electrical
and Computer Engineering, University of Canterbury, Christchurch, New Zealand
3
Department of Dermatology, Keck School of Medicine of the University of Southern California,
Los Angeles, CA.

Contact:
Kimberly A. Miller
Department of Preventive Medicine, Keck School of Medicine
2001 N Soto St, Suite 318-A
Los Angeles, CA 90032
Phone: (626) 864-6433 / Fax: (323) 865-0095
Email: kim.miller@med.usc.edu

Status: In press at Preventive Medicine

79
Abstract
Invasive melanoma is becoming more common in U.S. Hispanics. More research is
needed regarding the sun protection behaviors of Hispanics, particularly children and adolescents
who incur high UV exposures, to develop tailored skin cancer prevention interventions. We used
latent class analysis to examine patterns of sun protective behaviors in a cross-sectional survey of
Hispanic elementary students participating in a sun safety intervention in Los Angeles from
2013-2014 (N=972). Five behavior indicators in two environments (school and home)
representing multiple methods of sun protection were selected for the model.
Results suggested a four-class model best fit the data. Classes were labeled in order of
increasing risk as multiple protective behaviors (28%), clothing and shade (32%), pants only
(15%), and low/inconsistent protective behaviors (25%). Children who reported high parental
engagement with sun protection were significantly more likely to be classified in high overall
protective categories (OR=4.77). Girls were more likely than boys to be classified in the highest
protecting class (OR=3.46), but were also more likely to be in the “pants only” class (OR=2.65).

Sensitivity to sunburn was associated with less likelihood of being in the “clothing and shade”
class (OR=0.53).
The differences amongst these classes and their predictors reveal the heterogeneity and
complexity of Hispanic children’s sun protective behaviors. These findings have implications for
the design and delivery of future sun protection interventions targeting Hispanic children, as
strategies tailored to specific subgroups may be more effective in achieving meaningful
behavioral changes.

80
Introduction
Melanoma is a public health concern of increasing significance for U.S. Hispanics.
Although non-Hispanic whites (NHWs) have highest incidence of the disease, the proportion of
melanomas presenting at a late stage is increasing in Hispanics at a rate exceeding NHWs
(Cockburn et al., 2006; Rouhani et al., 2010). With Hispanics currently comprising 17% of the
U.S. population (approximately 53 million), a figure projected to grow to 31% by 2060 (CDC),
the melanoma burden in the near future for this population may be considerable.
Hispanics are therefore an important population to target for primary prevention efforts
that aim to reduce melanoma incidence through reduction of excessive exposure to ultraviolet
(UV) radiation, the primary environmental risk factor for melanoma (Elwood & Jopson, 1997; D.
Whiteman & Green, 1994). As high levels of UV exposure and sunburn in childhood increase
melanoma risk in adulthood (Gandini et al., 2005; D. C. Whiteman et al., 2001), primary
prevention interventions conducted among Hispanic children and adolescents are particularly
needed (Buller, Taylor, et al., 2006; A. Geller, Rutsch, Kenausis, & Zhang, 2003; Saraiya et al.,
2004). However, primary prevention interventions have largely targeted NHW children.
Studies of sun protection behaviors in Hispanic late adolescents and adults have found
low awareness and perceived risk of skin cancer and low prevalence of sun protective behaviors
in comparison to NHWs (Buller, Cokkinides, et al., 2011; Coups, Manne, & Heckman, 2008; Ma
et al., 2007; Pipitone et al., 2002). Despite perceptions that Hispanics do not sunburn as easily as
NHWs, Hispanics have high rates of sunburn comparable to or exceeding rates for NHWs (CDC,
2007; Coups et al., 2012). In addition, recent studies have suggested that acculturation, the
process by which a cultural group encounters and selectively adopts the beliefs and behaviors of
another culture (Negy, 1992), may influence sun protection for Hispanics. More U.S.-

81
acculturated Hispanics tend to adopt U.S. norms of sun protection such as use of sunscreen, and
less acculturated Hispanics are more likely to use sun protective clothing and hats and seek shade
(Andreeva et al., 2009; Coups E.J., 2013; Coups et al., 2012).
Studies regarding Hispanics and sun protection have primarily examined adult sun
behaviors and thus data regarding Hispanic children are limited. To address this research gap, the
present study explored sun protection among Hispanic children using latent class analysis
(LCA). This approach, which examines the heterogeneity within a seemingly homogeneous
sample, yields insights into the sun protection patterns and risk profiles of Hispanic children in
order to contribute to what is known about UV behaviors in this group and guide the
development of tailored interventions.
Latent class analysis is a person-centered statistical method for detecting unobserved
subgroups in a population using a set of observed variables (B. Muthén & Muthén, 2000).
Individuals are categorized into mutually exclusive classes determined by their responses on
indicator variables. The model’s parameters estimate the probabilities of identified classes and
probabilities of response for each indicator, conditional on class membership. Model fit is
determined by a set of statistical criteria that compare the results of model with n classes to the n-
1 model. The model that best fits the data with the smallest number of classes and yields
interpretable results is chosen (Lanza, Collins, Lemmon, & Schafer, 2007). Subsequently,
pertinent covariates can be used to predict the likelihood of belonging to a particular latent class
(Collins & Lanza, 2010).
We hypothesized that distinct subgroups, perhaps representing groups that might be more
effectively targeted with tailored interventions, would be present in the sample. Extrapolating
from the literature among primarily NHW populations, we additionally hypothesized that

82
Hispanic girls would be categorized in more protective classes (Alberg et al., 2002; Coogan et
al., 2001; Dixon et al., 1999; A. C. Geller et al., 2002). Consistent with studies among NHW
populations that have shown greater sun protection behaviors for children whose parents
establish standards for sun protection (O'Riordan et al., 2003; Turner & Mermelstein, 2005), we
hypothesized that Hispanic children who reported greater family sun protection would be
categorized in more protective classes.
In addition, following recent research about the influence of acculturation on sun
protection behaviors in Hispanic adults (Andreeva et al., 2009; Coups E.J., 2013), we
hypothesized that greater acculturation to U.S. norms would predict membership in classes
characterized by the use of sunscreen, a sun protection method more prevalent among NHWs,
while lower acculturation would predict membership in classes characterized by the use of
protective clothing and shade. Finally, we included skin phototype and sunburn sensitivity as
covariates in the model, anticipating that these variables would predict membership in more
protective categories.

Methods
Sample and Measures
Baseline data were used from the SunSmart study (N=1,646), a school-based sun safety
intervention conducted in Los Angeles in 2013-2014. Schools were recruited within proximity to
the University of Southern California (USC), and those selected were representative of school
district demographics. The USC Institutional Review Board (IRB) approved the study. The
analytic sample was restricted to students who self-reported as Hispanic or Latino (N=1,057). All
questionnaire items were administered before implementation of intervention activities.

83
However, because acculturation was measured at posttest and 85 students were absent the day of
that questionnaire, the sample was restricted to students who were present for the follow-up
questionnaire (N=972). No significant differences were found between the 972 students who
completed both pre-and posttest in comparison to the 85 who completed pretest only on
demographic characteristics and sun protection practices. Further, due to missing values on
covariate variables, the final analytic sample for the model including covariates comprised 967
students.
Latent class indicators were selected to represent multiple dimensions of children’s sun
protection. Five sun protective behaviors measured in two environments, at school and outside of
school (10 total) pertaining to the use of sunscreen, long sleeves, long pants, sun protective hats,
and shade seeking were used. Each question was scored on a 4-level Likert-type scale with
responses including “Never” “Rarely” “Sometimes” and “Often.” To create binary indicators for
the purposes of LCA, each item was dichotomously coded as 0 for “never/rarely” implementing
the behavior vs. 1 for “sometimes/often” implementing the behavior. This classification method
enabled the differentiation between higher risk students who rarely or never engaged in
prevention objectives, versus lower risk students who met prevention objectives at least some of
the time when outdoors.

Covariates
Five covariates were selected for their theoretical significance to sun protection in
Hispanic children: gender, acculturation, skin phototype, sunburn sensitivity, and family sun
protection habits for child. Acculturation was assessed using the U.S. orientation (assimilation)
subscale of the Acculturation, Habits, and Interests Multicultural Scale for Adolescents

84
(AHIMSA) scale (Unger, 2002). A single score was used ranging from 0-8, with higher scores
reflecting higher levels of U.S. acculturation, measured as assimilation to U.S. norms.
Skin phototype was assessed with a five-level item adapted from the Fitzpatrick skin
phototype scale and was entered into the model as a continuous variable ranging from 1=Very
Fair to 5=Very Dark (Fitzpatrick, 1988). Sunburn sensitivity was assessed with one yes/no item
in which students were asked if they had ever experienced sunburn. Family sun protection habits
for child comprised an average of three variables that asked students if parents asked them to
wear sunscreen, sun protective clothing, and a hat when outdoors on a sunny day. Responses
were on a 4-point Likert-style frequency scale and ranged from “never” to “often,” with higher
scores indicating higher sun protection.

Statistical analysis
Latent class analysis was conducted using Mplus Version 6.0 (L. K. Muthén, & Muthén,
B. O., 1998-2011) The number of classes examined began at 1 and was increased incrementally.
Information criteria indices including the Akaike Information Criteria (AIC), Bayseian
Information Criteria (BIC) and sample size adjusted BIC (SS-BIC) were used to determine
model fit. The results of a Lo-Mendell-Rubin likelihood-ratio test which compares n vs. n−1
class models were also evaluated to reject the null hypothesis that n–1 class is better (Lanza et
al., 2007). Lower values on the AIC, BIC, and SS-BIC suggest superior fit, and a significant Lo-
Mendell-Rubin (LMR) test indicates that a model of k+1 classes fits significantly better than a
model with only k classes. We stopped increasing the class size when there was no substantial
decrease in information criteria and a non-significant LMR test. An entropy summary statistic
also aided in assessing the quality of classification with values ranging from 0-1; values near 1

85
indicate good discrimination between classes (B. Muthén et al., 2002). In addition, model
interpretability and theoretical meaningfulness were considered in selecting the best fitting
model (Collins & Lanza, 2010). Subsequently, covariates were incorporated into the final latent
class model simultaneously to examine associations between latent classes and observed
covariates. Multinomial logistic regression was used to determine the relationship between
predictor variables and latent classes, and ORs with p-values were reported.

Results
Characteristics of the sample are summarized in Table 5-1. Response categories for
students reporting “sometimes” or “often” for each sun protective behavior included high
endorsement of use of long sleeves and long pants, particularly at school. More than half of
students reported seeking shade on sunny days when at home and at school. Smaller proportions
of students used sunscreen or hats in either environment.

86

Mean age (SD) 9.84 0.71
n %
Gender
Female 510 52.47
Male 460 47.33
Missing 2 0.21
Grade Level
4th 407 41.87
5th 486
50.00
Mixed 4th/5th 79
8.13
Skin Tone
Very fair 7 0.72
Fair 167 17.18
Light brown 647 66.56
Dark brown 143 14.71
Very dark 3 0.31
Missing 5 0.51
Ever Sunburned
Yes 603 62.04
No 366 37.65
Missing 3 0.31
Sun behavior characteristics
a
Hat use at school 204 20.98
Hat use at home 358 36.83
Long sleeves at school 664 68.32
Long sleeves at home 487 65.13
Long pants at school 788 81.07
Long pants at home 645 66.36
Use of sunscreen at school 360 37.03
Use of sunscreen at home 330 33.96
Seeking shade at school 554 57.00
Seeking shade at home 591 60.80
Table 5-1. Characteristics of sample at baseline,
SunSmart (N=972)
a
Percent responding "often" or "sometimes"
Note: Percentages based on N responding to each
question. The amount of missing data varied across
item responses.

87
Description of classes
Results of the LCA led to a selection of a four-class model. Table 5-2 presents fit
statistics for the four-class solution, which was determined by a non-significant LMR test for the
5-class model and negligible decreases in the information criteria. In addition, class sizes were
substantial and yielded conceptually meaningful interpretation (Collins & Lanza, 2010).

Table 5-3 and Figure 5-1 presents the distribution of the latent classes estimated by the
model. Classes were arranged in approximate order of sun protection and comprised two higher-
protecting classes (Classes 1 and 2) and two lower-protecting classes (Classes 3 and 4). Class 1
(“multiple protective behaviors”; 28%) comprised a response pattern characterized by high
probabilities of the use of nearly all sun protective methods and the highest probabilities for
sunscreen use. Class 2 (“clothing and shade”; 32%) was the most prevalent class and was
characterized by high probabilities of the use of long sleeves and long pants as well as the
second-highest probabilities for shade use after Class 1. Classes 3 and 4 were similar, with both
classes demonstrating lower sun protection aside from moderate probabilities of the use of shade.
Variable 1 2 3 4 5
No. of parameters 10 21 32 43 54
Log likelihood -6073.233 -5723.239 -5648.497 -5600.967 -5560.82
AIC 12166.466 11488.478 11360.993 11287.935 11229.64
BIC 12215.260 11590.944 11517.133 11497.747 11493.125
N-adjusted BIC
12183.500 11524.249 11415.501 11361.179 11321.621
Lo–Mendell–Rubin
2 vs. 1 3 vs. 2 4 vs. 3 5 vs. 4
testing the null hypothesis
LMR probability
<0.0001 0.004 0.009 0.21
Entropy 0.69 0.65 0.68 0.68
Table 5-2. Model-fit indices for a latent class analysis of sun protective behaviors among
Hispanic schoolchildren, SunSmart (N=972)
No. of classes
Notes: No. = number; AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion.

88
However, Class 3 (“pants only”; 15%) was characterized by high probabilities of the use of long
pants only both at home and at school in contrast to other sun protection methods. One quarter
of the sample was likely to be classified as Class 4 (“low/inconsistent protective behaviors”;
25%); this class was characterized by the low use of most sun protective behaviors aside from
moderate use of long sleeves at school and moderate shade.

1
Multiple
protective
behaviors
2
Clothing and
shade
3
Pants only
4
Low/inconsistent
protective
behaviors
n=264 (28%) n=308 (32%) n=136 (15%) n=264 (25%)
Use of sunscreen at school 0.77 0.22 0.22 0.22
Use of sunscreen at home 0.78 0.16 0.12 0.23
Hat use at school 0.30 0.14 0.11 0.26
Hat use at home 0.54 0.31 0.18 0.37
Long sleeves at school 0.93 0.89 0.24 0.42
Long sleeves at home 0.76 0.82 0.00 0.13
Long pants at school 0.95 0.96 1.00 0.38
Long pants at home 0.89 0.89 0.77 0.07
Seeking shade at school 0.79 0.52 0.45 0.47
Seeking shade at home 0.82 0.61 0.40 0.49
Table 5-3. Four-latent-class model of sun protection behaviors: probabilities of engagement for each
subgroup, SunSmart (N=972)
Latent class

89
Figure 5-1

Covariates predicting latent classes
The 4-class model was then fit with the five covariates simultaneously. Table 5-4 shows
ORs for predictors of class categories, presenting alternative parameterizations designating each
of the two lower-protective classes as a reference category (i.e., Classes 3 and 4). Highest ORs
were observed for students with high family engagement in sun protection who were
significantly more likely to be classified in the two high protection categories relative to the
lower protection classes. Relative to Class 3 (pants only), students with high family sun
protection were significantly more likely to be classified in Class 1, the most protective class
(multiple protective behaviors (OR=
4.77; p<.0001). Relative to Class 4 (low/inconsistent
protective behaviors), students with high family sun protection were also significantly more
likely to be classified in Class 1 (OR=4.01; p<.0001) as well as the second most protective
category, Class 2 (clothing and shade) (OR=2.34; p<.0001).
0.00#
0.50#
1.00#
Sunscreen#at#school#
Sunscreen#at#home#
Hat#at#school#
Hat#at#home#
Sleeves#at#school#
Sleeves#at#home#
Pants#at#school#
Pants#at#home#
Shade#at#school#
Shade#at#home#
Probability*
Fig.%5'1:%LCA%among%Hispanic%schoolchildren,%SunSmart%
1##
Mul8ple#protec8ve#
behaviors#
(28%)#
2#
Clothing#and#shade#
(32%)#
3#
Pants#only##
(15%)#
4#
Low#and#inconsistent#
protec8ve#behaviors#
(25%)#

90

Girls were significantly more likely to be classified in the most protective category, Class
1 (multiple protective behaviors) in comparison to Class 4 (low/inconsistent protective
1
Multiple
protective
behaviors
2
Clothing and
shade
3
Pants only
4
Low/inconsistent
protective
behaviors
n=263 (27%) n=219 (23%) n=256 (26%) n=231 (24%)
Intercept -0.78 0.75 Ref -0.09
Family sun habits for child
Logit 1.56 1.02 Ref 0.17
OR 4.77
***
2.77
***
Ref 1.18
Female
Logit 0.27 -2.18 Ref -0.97
OR 1.31 0.11
***
Ref 0.38
**
Acculturation
Logit -0.05 0.04 Ref -0.05
OR 0.96 1.04 Ref 0.95
Sunburn reactive skin
Logit 0.001 -0.45 Ref 0.19
OR 1.00 0.64 Ref 1.21
Skin phototype
Logit 0.19 0.14 Ref 0.21
OR 1.20 1.15 Ref 1.23
Intercept -0.70 0.84 0.08 Ref
Family sun habits for child
Logit 1.39 0.85 -0.17 Ref
OR 4.03
***
2.35
***
0.85 Ref
Female
Logit 1.24 -1.21 0.97 Ref
OR 3.46
***
0.30
**
2.65
***
Ref
Acculturation
Logit 0.01 0.09 0.05 Ref
OR 1.01 1.09 1.06 Ref
Sunburn reactive skin
Logit -0.19 -0.64 -0.19 Ref
OR 0.83 0.53
**
0.83 Ref
Skin phototype
Logit -0.02 -0.07 -0.21 Ref
OR 0.98 0.93 0.81 Ref
Table 5-4. Logit estimates and odds ratios for predictors of latent class membership, SunSmart
(N=967)
Boldface indicates statistical significance (p< *0.1; p< **<0.05; p < ***<0.01)

91
behaviors) (OR=3.46; p=0.002). Girls were also more likely to be classified in Class 3 (pants
only) (OR=2.64; p=0.003) relative to Class 4. However, girls were significantly less likely to be
classified in Class 2 (clothing and shade) relative to Class 3 (OR=
0.11; p=0.001) or Class 4
(OR=0.30; p=0.01) Relative to Class 3, girls were also significantly less likely to be classified in
Class 4 (OR=0.38; p=0.01).
Students who reported sunburn sensitivity were significantly less likely to be classified in
Class 2 (clothing and shade) relative to Class 4 (low/inconsistent protective behaviors)
(OR=0.53; p=0.04). Notably, neither skin phototype nor level of acculturation were significantly
associated with class membership.

Discussion
We identified four latent classes of levels of sun protection behaviors among Hispanic
schoolchildren in the Los Angeles area, supporting our hypothesis that distinct subgroups existed
in the sample. The sample was divided between higher protectors, who used multiple methods of
sun protection (Classes 1 and 2), and lower protectors, who used few or moderate levels of
protection methods (Classes 3 and 4).
Although protective clothing and shade were the most prevalent methods of sun
protection, only one class had high probabilities of sunscreen use. This distinction might have
emerged due to low use of sunscreen among Hispanics, a result of less awareness and perceived
risk of skin cancer or limited access to sunscreen in Hispanic communities (C. Hernandez et al.,
2012). While sunscreen offers only partial protection in the prevention of skin cancer, it remains
an important method to use in tandem with physical barriers such as protective clothing and
shade, particularly for children in high UV environments (Viros et al., 2014). In our study, while

92
one class of Hispanic children reported frequent use of sunscreen (Class 1), children who were
otherwise high protectors reported low probabilities of sunscreen use (Class 2) as did the two
low-protecting classes (Classes 3 and 4). Accordingly, understanding the facilitators and barriers
of and increasing sunscreen use among Hispanic children is a priority area for future research.
The strongest predictor of membership in high protecting classes was family habits for
children’s sun protection. While the important influence of family variables on the health
practices of Hispanics has been extensively researched (Katiria Perez & Cruess, 2014), our
results are in alignment with studies among NHW samples where parental sun protection habits
have been shown to exert strong influence on child sun protection (Robinson et al., 2000; Turner
& Mermelstein, 2005). For elementary school-aged children, sun protection may be shaped less
by independent health decisions than by familial norms or provision of resources (sunscreen or
protective clothing). Thus, broad strategies that incorporate families in sun protection might
generalize cross-culturally as a critical component of effective primary prevention for school-
aged children.
Contrary to our hypotheses, neither acculturation nor skin phototype was associated with
class membership. One explanation for this lack of association may be low variability in the data,
as the sample included relatively low-acculturated children (with a mean score 2.6 out of 8 on
the assimilation AHIMSA subscale). In addition, prior research on acculturation and sun
protection has been conducted among adults and its influence on children’s protective behaviors
may differ (Andreeva et al., 2009; Coups E.J., 2013). Children’s sun protective practices may
reflect parental level of acculturation given the strong family influence regarding sun protection;
thus, further research that captures the dynamics of acculturation as a family-level phenomenon
is warranted with respect to its influence on children’s sun protection.

93
Skin phototype was not significantly associated with class membership; this variable also
lacked variability, with the majority of children self-reporting light brown skin. However,
sunburn sensitivity was significantly associated with less likelihood of membership in a high-
protection class. It is possible that Hispanic children who use little sun protection are more likely
to report being sunburned as a result of time spent outdoors unprotected. This finding is of
concern, as sensitivity to sunburn has been associated with greater sun protection for NHW
adolescents (Buller, Cokkinides, et al., 2011), and even one severe sunburn in childhood can
double the rate of melanoma in adulthood (Armstrong & Kricker, 2001; Gandini et al., 2005).
For Hispanic children and their families, sunburn sensitivity may not motivate increased use of
sun protection due possibly to lower perceived risk or less physician counseling about harmful
UV overexposure and sunburn avoidance (Ma et al., 2007; Pipitone et al., 2002). Future
intervention strategies might strengthen recognition of sunburn as a sign of skin damage for
Hispanic children and families, particularly for children in less protective subgroups.
The role of gender in relation to sun protection subgroups was complex in this study.
While several studies have affirmed the greater use of sun protection by NHW preadolescent
girls than boys (Dixon et al., 1999; A. C. Geller et al., 2002) and Hispanic girls in the current
study were more likely to be categorized in high-protecting classes, it is unclear in our study why
girls were also classified in the lower-protecting classes (in particular Class 3, pants only). It is
possible that additional subgroups exist that further delineate high from low protecting Hispanic
girls, potentially due to peer norms or culturally specific values such as modesty (Juckett, 2013;
2007). More research is required with a wider array of variables relevant to sun protection to
further distinguish subgroups among Hispanic preadolescent girls.

94
Limitations and strengths of the research
Because results of the LCA are highly data-driven, generalizability of the findings are
limited and should be replicated. There is some potential for misclassification regarding the
dichotomous variable to distinguish endorsement vs. non-endorsement of behaviors as middle
values (“rarely” and “sometimes”) were included. Our measures were self-reported, which could
introduce misclassification or social desirability bias. However, recent studies have shown
adequate to good accuracy for self-reported sun protection estimates (Glanz & Mayer, 2005;
Hillhouse et al., 2012).
In addition, Hispanics represent a heterogeneous ethnicity, and our analysis does not
identify Hispanic subgroup. Based on recent census data for Los Angeles County (United States
Census Bureau, 2013), our assumption is that the majority of children participating in the study
were Mexican-American followed by Central American. However, as differences have been
found among Hispanic subgroups with respect to sun protection (Coups et al., 2012), further
research with a more diverse sample is required to examine whether our analysis would hold for
other Hispanic heritage groups (South American, Cuban, Puerto Rican, etc.).
Strengths of the research include a large sample of Hispanic preadolescents and an
innovative approach to profiling and delineating the heterogeneity of sun protective behaviors
within this sample. To our knowledge, the current study is the first attempt to use LCA to
identify distinct typologies of sun protection among Hispanic children.

Conclusion
Our analysis used a novel, person-centered approach and is the first paper to employ
latent class analysis among Hispanic children in the sun protection literature. Results from the
study demonstrate that while there are subgroups of high protecting Hispanic children, the

95
majority belonged to classes distinguished by suboptimal sun protection. Furthermore, the most
salient predictor for being in the high protecting classes was the use of sun protection by parent
for child. Therefore, our findings suggest that for Hispanic children, a culturally appropriate
family-based intervention might most effectively target subgroups at highest risk. In addition,
subgroups not using sunscreen might be encouraged to adopt its use in addition to physical
barrier methods, while sunburn avoidance might be more strongly emphasized for Hispanic
families. Such prevention steps, in addition to future research among Hispanic children and
adults, may contribute to the long-term stabilization and/or reduction of rates of melanoma in
this at-risk population.

96
CHAPTER 6: DISCUSSION
Overall findings
The research conducted in this dissertation provides novel insights into the motivators
and barriers of sun protection in elementary school-age children, in particular among
predominantly ethnic minority children, an understudied population with respect to skin cancer
prevention. The wide range of psychosocial and cultural measures used in these studies enabled a
thorough exploration of variables not previously explored in depth in this age range, nor with
ethnically and racially diverse samples.
We observed significant associations between psychosocial constructs such as perceived
self-efficacy, barriers to sun protection, perceived peer norms, and sun protection behaviors, in
accordance with our original hypotheses. These constructs have been found to be significant
predictors of improved sun protection in prior studies conducted among majority NHW
populations, both in adults and children (Andreeva VA, 2008; Buller, Taylor, et al., 2006;
Mermelstein & Riesenberg, 1992). In contrast to other studies that have found perceived risk
associated with greater sun protection (Koh et al., 1997) perceived risk did not emerge as a
significant variable in these studies. While perceived susceptibility/risk has not been a consistent
predictor of sun protection for NHWs, it is possible that perceived risk was not salient in these
studies due to low perceived awareness of and risk for skin cancer among non-white populations
and in particular, in Hispanics (C. Hernandez et al., 2012). However, the age of participants in
these studies may have also been a factor, as preadolescents and adolescents may not readily
respond to risk-based messaging regarding skin cancer, particularly due to the distal nature of the
disease. Thus, perceived risk to long-term consequences of UV exposure may not be as strong an

97
influence in prevention practices for younger and multiethnic children; future studies might focus
on testing whether proximal risks (e.g., of sunburn-related pain, redness, or peeling) may carry
more influence.
Variables that were associated with greater sun protection were not uniformly associated
with each protective method, a finding that been noted in prior research (Cokkinides et al., 2001)
but not widely studied. That different sun protective methods are unique behaviors, with distinct
psychosocial facilitators and barriers, may prove to be increasingly important in future
intervention efforts. In our studies, self-efficacy to use sun protection, for example, was
associated with use of protective clothing and shade seeking, but not with sunscreen or hat use,
whereas perceived norms were associated with use of sunscreen and shade seeking, but not with
protective clothing or hat use. Such differential associations challenge broad-based and more
generic approaches to sun safety interventions, as the complex of behaviors that comprise sun
protection may have unique correlates. Our findings thus reinforce recent calls for the
recognition of and more research regarding sun protection methods as a set of distinct practices
requiring behavior-specific analyses and intervention approaches (Saraiya et al., 2004; Tripp et
al., 2013).
Strongly associated with increased sun protection across all behavioral methods,
however, was family's use of sun protection for child. Family influence on children's health
practices is important across a spectrum of behaviors, including physical activity (Davison &
Birch, 2002), dietary habits (Birch & Fisher, 1998), and substance use (L. Hernandez, Rodriguez,
& Spirito, 2015). In the field of sun protection in prior studies among non-Hispanic white
children, parental habits and standards have been associated with greater child sun protection,
particularly by parent for child (O'Riordan et al., 2003; Robinson et al., 2000). Parental

98
insistence that children use sun protection, parental modeling of sun protective behaviors, and
lack of barriers have been most salient in influencing greater sun protection among children
(Cokkinides et al., 2004).
Our studies corroborate these findings among a different, predominantly Hispanic and
racial minority sample, suggesting that the influence of family on child sun protection is an
important area of overlap between NHW and non-white children. Accordingly, future
interventions that incorporate family and caregiver components may be most effective in
improving and sustaining child sun protection. Our findings also indicate that while availability
of sun protection methods is an important factor (e.g. sunscreen availability in the home),
variables such as parental standards of sun protection for children and family communication
about sun protection are as important, if not more significant, influences. In other words,
availability of protection methods, while necessary, is not sufficient to motivate children's sun
protection: a familial environment that is aware of and promotes sun protection may be a critical
component for establishing durable protective habits for children across cultural contexts.
Thus, while school-based interventions have been the norm for sun safety targeted to
children (due in part to the substantial time children spend at school, the UV exposures incurred
during this time, and the feasibility of program delivery), other intervention contexts and
frameworks where the family is involved, for example the pediatric primary care setting or
family-based counseling in the home, may yield greater effect. Therefore, future research is
needed to develop and test novel and scalable interventions that incorporate family involvement
to effect change in child sun safety behaviors if rates of skin cancer are to be impacted over the
long term.

99
Given the unique composition of the sample, we had originally hypothesized in Studies 1
and 3 the importance of acculturation, as prior studies among Hispanic adults suggest that greater
acculturation and assimilation to U.S. norms result in both protective behaviors (greater
sunscreen use) and risk behaviors (greater duration of UV exposure and less use of protective
clothing), whereas lower U.S. acculturation is associated with greater use of sun protective
clothing and shade-seeking (Coups et al., 2012). In contrast with those studies, our findings did
not suggest a strong influence of acculturation on sun protection among this predominantly
Latino and minority sample. However, our failure to find an influence of acculturation on sun
protection may have to do with the age of the sample, as previous studies were conducted with
adults. It is possible that the effects of acculturation increase with age, as children become more
autonomous and cultural adaptation and integration becomes more salient (Phinney, 1990). In
addition, the sample was relatively low acculturated, particularly Hispanic children in the
sample, and lack of variability in the data may have precluded finding hypothesized
relationships.
Despite our observation of higher levels of psychosocial constructs associated with
greater sun protection in the baseline data, evidence for the mediating role of these variables
between effects of the intervention and protective behaviors was relatively weak. While the
SunSmart intervention was able to increase levels of psychosocial variables for both treatment
conditions, these increases did not strongly influence use of sunscreen at both short- and longer-
term follow-up. These findings extend the very few other studies (notably, that of Reynolds et
al., 2006) that have examined mediators on a short-term basis (several weeks) and which have
found evidence for the mediating effect of psychosocial variables between program effects and
sun protection outcomes (Reynolds et al., 2006). However, over our three-month time span, we

100
found that the few variables that may serve as mediators of the intervention attenuated as time
from the intervention increased, suggesting more limited, short-term influence on children's sun
behavior. Our findings, then, raise the question of the utility of extensively targeting individual-
level psychosocial constructs if such constructs may only have effect of limited duration on
child's sun protection habits.
In our studies, only knowledge of sun protection emerged as a significant mediator over
the longer follow-up period. Of note, knowledge has proven to be one of the constructs most
amenable to change in the context of child and adolescent sun protection interventions (Saraiya
et al., 2004); however, the connection between knowledge and sun safe behaviors has been
inconsistent in studies, particularly among older adolescents (Marks & Hill, 1988; Saraiya et al.,
2004). In the mediation study by Reynolds et al., while knowledge met the criteria for mediation
in single mediator models, it did not meet the criteria for mediation in multiple mediator models
and had lower reliability than other mediating variables. While the authors of that study suggest
that this indicates that the association of knowledge is accounted for by inclusion of other
mediators in the model, our study suggests that knowledge may be a unique mediator,
particularly in the case of the SunSmart dosimetry intervention, which emphasized child-led
experiential knowledge over classroom-based didactic instruction. In educational contexts,
studies have shown that the more immediate and personal approach of experiential knowledge
may build longer-term reflection, resulting in knowledge deficits in the short-term compared to
didactic approaches, but increasing retention over the longer-term as learners process their
experiences to generate new knowledge (Beames, Higgins, & Nicol, 2012). Thus, the innovative
addition of an experiential learning component in the SunSmart intervention may prove to be an
important mediator of children's sun protection over more common instructional approaches,

101
which have comprised the majority of school-based sun safety interventions. However, these
effects may not be immediate, and studies with more extended follow-up are needed to capture
and determine the potential efficacy of experiential-based approaches such as the SunSmart
dosimetry intervention in increasing child knowledge of sun protection leading to improve sun
protective behaviors.

Limitations and strengths of the current research
Limitations
Despite the unique nature of our largely ethnic and racial minority sample, there was not
enough power to conduct comparisons between different races and ethnicities, as the majority of
children participating in the study were Hispanic (64%). Thus, analyses were limited to
differences between Hispanic/non-Hispanic. While the sample may be representative within its
frame, a more informative approach would compare distinct ethnic and racial groups with respect
to research hypotheses. Additionally, we were unable to distinguish between Hispanic heritage
subgroups; however, as noted in Study 3, our assumption, given census and school
demographics, is that the majority of students in the sample were of Mexican-American
background.
Our analyses were limited by low variability in the data for several variables, including
skin phototype and acculturation, limiting patterns and/or findings related to study hypotheses.
These results, however, may be indicative of the sample frame (children living in urban areas of
Los Angeles County) and may nevertheless contribute to future prevention efforts in the region
and in areas with similar demographics.

102
Longitudinal follow-up for Study 2 was relatively short at three months; however, this
period greatly improves upon many studies among children and adolescents the majority of
which have follow-up period of one month or less (Saraiya et al., 2004). In addition, model fit
for Study 2 was mediocre to adequate for our mediation analyses. While this issue most likely
was due to the examination of single mediator models (which omit a range of covariates and
additional variables), further studies might investigate models with multiple mediators and
mediational pathways, as the influence and interplay of psychosocial constructs on sun
protection may prove to be more complex.
Strengths
The current research is, to our knowledge, one of the first and largest studies set among
predominantly Hispanic and minority samples within this age group with respect to skin cancer
prevention. As such, these studies contribute importantly to knowledge regarding an
understudied group in the field of sun safety and will contribute to future prevention efforts
among this population.
The inclusion of a wide range of psychosocial constructs widely used and validated in
prior research (Reynolds et al., 2006) allowed breadth in examining correlates and predictors of
sun protection outcomes and enhances comparability with past studies. Our use of separate
behavioral measures for sun protection provided insights into associations between correlates
and specific protective behaviors, adding to an emerging literature that recognizes the distinction
of UV-related behaviors and their different motivators and barriers (Tripp et al., 2013).
Finally, our studies are the first to examine acculturation and sun protection among
elementary-school aged children. While we did not observe a significant impact of acculturation

103
in these studies, further research is required in this area. Our research may spur further interest in
and provide valuable context for future studies.

Implications for future research and public health impact
The findings of this research suggest the complex nature of children's sun protection
behaviors. Because of rising rates of melanoma and non-melanoma skin cancer among non-white
populations and particularly among U.S. Hispanics (Cockburn et al., 2006), our work has
focused on a predominantly ethnic minority, largely Latino sample in a high-risk region, Los
Angeles County. Our studies have found overlap between the facilitators and barriers of sun
protection between ethnic minority children and prior research on NHW children, and highlight
the important influence of family influence. However, more work is needed to extend the
findings here and to work to translate them into effective interventions to modify and improve
child sun protection behaviors.
It is clear from these studies that effective education for children's sun protection must
address the multifaceted nature of child's sun protection behaviors which includes recognizing
these behaviors as distinct, identifying the underlying behavioral patterns that characterize
samples, and engaging the social and environmental influences that govern such behaviors (peer
and particularly, family norms). Such interventions may require novel strategies that incorporate
these identified influences, e.g. family-based or social network approaches rather than
approaches that rely solely on individual-level change which have comprised the majority of sun
safety programs for children to date.
In addition, while evidence for the sustained influence of psychosocial mediators was
relatively weak in this research, further research is needed into experiential learning as a

104
potentially powerful way to build more sustained cognitive reflection for children's practice of
sun protection, and additional longitudinal research is needed to both capture this potential and to
examines other interconnected mediational pathways. Furthermore, different mediating
constructs for multiethnic and ethnic minority children may be more relevant than those
examined here. In-depth research, both qualitative and quantitative, is required to identify salient
constructs and formulate effective intervention approaches for these children.
Finally, due to the nature of sun protection, which comprises practices driven by factors
other than the intention to protect one's skin from high UV exposure (e.g., shade-seeking or use
of protective clothing may be motivated less from the desire to protect one's skin from the sun
and more by heat-related discomfort, fashion or peer norms), broad strategies are needed that
might incorporate such factors, and/or that may incorporate other levels aside from individual
behavior change. Policy or community-based approaches that create environments that enable
UV protection without reliance on individuals' motivation and compliance (e.g. shading parks
and school playgrounds, or school district hat policies in high UV regions) may be essential to
reducing excessive UV overexposure. Such strategies might be modeled for example on efforts
to reduce tobacco use among youth, which has largely adopted regulatory and policy driven
approaches and in so doing, has met with substantial declines in rates of smoking (Centers for
Disease & Prevention, 2002).
In conclusion, the melanoma burden continues to increase worldwide (International
Agency for Research on Cancer, 2010). Increases in racial and ethnic demographics shift in the
U.S. with greater phenotypic diversity (Perez & Hirschman, 2009) along with increases in skin
cancer rates in non-white populations necessitate new understanding and attention to sun
protection. The present research has provided unique insights into the sun protection practices of

105
predominantly ethnic minority children who are important targets for sun protection intervention.
Despite limitations, the present research offers descriptive utility for future investigative and
intervention-oriented research.

106
Bibliography
Abroms, L., Jorgensen, C. M., Southwell, B. G., Geller, A. C., & Emmons, K. M. (2003).
Gender differences in young adults' beliefs about sunscreen use. Health Educ Behav,
30(1), 29-43.
Adams, M. A., Norman, G. J., Hovell, M. F., Sallis, J. F., & Patrick, K. (2009).
Reconceptualizing decisional balance in an adolescent sun protection intervention:
mediating effects and theoretical interpretations. Health Psychol, 28(2), 217-225. doi:
10.1037/a0012989
Alberg, A. J., Herbst, R. M., Genkinger, J. M., & Duszynski, K. R. (2002). Knowledge, attitudes,
and behaviors toward skin cancer in Maryland youths. J Adolesc Health, 31(4), 372-377.
Anderman, C., Cheadle, A., Curry, S., Diehr, P., Shultz, L., & Wagner, E. (1995). Selection Bias
Related To Parental Consent in School-Based Survey Research. Evaluation Review,
19(6), 663-674. doi: 10.1177/0193841X9501900604
Andreeva VA, R. K., Buller DB, Chou C-P, Yaroch AL. . (2008). Concurrent psychosocial
predictors of sun safety among middle school youth. . Journal of School Health, 78(7),
374–381. doi: 10.1111/j.1746-1561.2008.00317.x.
Andreeva, V. A., Unger, J. B., Yaroch, A. L., Cockburn, M. G., Baezconde-Garbanati, L., &
Reynolds, K. D. (2009). Acculturation and sun-safe behaviors among US Latinos:
findings from the 2005 Health Information National Trends Survey. Am J Public Health,
99(4), 734-741. doi: 10.2105/AJPH.2007.122796
Angelopoulou, M. V., Kavvadia, K., Taoufik, K., & Oulis, C. J. (2015). Comparative clinical
study testing the effectiveness of school based oral health education using experiential
learning or traditional lecturing in 10 year-old children. BMC Oral Health, 15(1), 51. doi:
10.1186/s12903-015-0036-4
Armstrong, B. K., & Kricker, A. (2001). The epidemiology of UV induced skin cancer. J
Photochem Photobiol B, 63(1-3), 8-18.
Arthey, S., & Clarke, V. A. (1995). Suntanning and sun protection: a review of the psychological
literature. Soc Sci Med, 40(2), 265-274.
Bandura, A. (1986). Social foundations of thought and action : a social cognitive theory.
Englewood Cliffs, N.J.: Prentice-Hall.
Beames, S., Higgins, P. J., & Nicol, R. (2012). Learning outside the classroom : theory and
guidelines for practice. New York, NY: Routledge.

107
Behrens, C. L., Thorgaard, C., Philip, A., & Bentzen, J. (2013). Sunburn in children and
adolescents: associations with parents' behaviour and attitudes. Scand J Public Health,
41(3), 302-310. doi: 10.1177/1403494813476158
Berry, J. W. (1990). Acculturation and adaptation: health consequences of culture contact among
circumpolar peoples. Arctic Med Res, 49(3), 142-150.
Birch, L. L., & Fisher, J. O. (1998). Development of eating behaviors among children and
adolescents. Pediatrics, 101(3 Pt 2), 539-549.
Buller, D. B., & Borland, R. (1998). Public education projects in skin cancer prevention: child
care, school, and college-based. Clin Dermatol, 16(4), 447-459.
Buller, D. B., & Borland, R. (1999). Skin cancer prevention for children: a critical review.
Health Educ Behav, 26(3), 317-343.
Buller, D. B., Cokkinides, V., Hall, H. I., Hartman, A. M., Saraiya, M., Miller, E., . . . Glanz, K.
(2011). Prevalence of sunburn, sun protection, and indoor tanning behaviors among
Americans: review from national surveys and case studies of 3 states. J Am Acad
Dermatol, 65(5 Suppl 1), S114-123. doi: 10.1016/j.jaad.2011.05.033
Buller, D. B., Geller, A. C., Cantor, M., Buller, M. K., Rosseel, K., Hufford, D., . . . Lew, R. A.
(2002). Sun protection policies and environmental features in US elementary schools.
Arch Dermatol, 138(6), 771-774.
Buller, D. B., Reynolds, K. D., Ashley, J. L., Buller, M. K., Kane, I. L., Stabell, C. L., . . . Cutter,
G. R. (2011). Motivating public school districts to adopt sun protection policies: a
randomized controlled trial. Am J Prev Med, 41(3), 309-316. doi:
10.1016/j.amepre.2011.04.019
Buller, D. B., Reynolds, K. D., Yaroch, A., Cutter, G. R., Hines, J. M., Geno, C. R., . . .
Grandpre, J. (2006). Effects of the Sunny Days, Healthy Ways curriculum on students in
grades 6 to 8. Am J Prev Med, 30(1), 13-22. doi: 10.1016/j.amepre.2005.08.046
Buller, D. B., Taylor, A. M., Buller, M. K., Powers, P. J., Maloy, J. A., & Beach, B. H. (2006).
Evaluation of the Sunny Days, Healthy Ways sun safety curriculum for children in
kindergarten through fifth grade. Pediatr Dermatol, 23(4), 321-329. doi: 10.1111/j.1525-
1470.2006.00270.x
CDC. (2007). Sunburn prevalence among adults – United States, 1999, 2003, and 2004. MMWR,
56(21), 524-528.
CDC. (2011). The Hispanic Population: 2010. Retrieved January 5, 2015

108
Centers for Disease, C., & Prevention. (2002). Trends in cigarette smoking among high school
students--United States, 1991-2001. MMWR Morb Mortal Wkly Rep, 51(19), 409-412.
Centers for Disease Control and Prevention (CDC). (1998). Sun-protection behaviors used by
adults for their children--United States, 1997. MMWR. Morbidity and Mortality Weekly
Report, 47(23), 480–482.
Cialdini, R., Reno, R., & Kallgren, C. (1990). A Focus Theory of Normative Conduct: Recycling
the Concept of Norms to Reduce Littering in Public Places. J Pers Soc Psychol, 58(6),
1015-1026.
Clarke, V. A., Williams, T., & Arthey, S. (1997). Skin type and optimistic bias in relation to the
sun protection and suntanning behaviors of young adults. J Behav Med, 20(2), 207-222.
Cockburn, M. G., Zadnick, J., & Deapen, D. (2006). Developing epidemic of melanoma in the
Hispanic population of California. Cancer, 106(5), 1162-1168. doi: 10.1002/cncr.21654
Cody, R., & Lee, C. (1990). Behaviors, beliefs, and intentions in skin cancer prevention. J Behav
Med, 13(4), 373-389.
Cokkinides, V. E., Johnston-Davis, K., Weinstock, M., O'Connell, M. C., Kalsbeek, W., Thun,
M. J., & Wingo, P. A. (2001). Sun exposure and sun-protection behaviors and attitudes
among U.S. youth, 11 to 18 years of age. Prev Med, 33(3), 141-151. doi:
10.1006/pmed.2001.0877
Cokkinides, V. E., Weinstock, M. A., Cardinez, C. J., & O'Connel, M. A. (2004). Sun-safe
practices in U.S. youth and their parents: role of caregiver on youth sunscreen use. Am J
Prev Med, 26(2), 147-151.
Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data:
questions and tips in the use of structural equation modeling. J Abnorm Psychol, 112(4),
558-577. doi: 10.1037/0021-843X.112.4.558
Collins, L. M., & Lanza, S. T. (2010). Latent class and latent transition analysis : with
applications in the social behavioral, and health sciences. Hoboken, N.J.: Wiley.
Coogan, P. F., Geller, A., Adams, M., Benjes, L. S., & Koh, H. K. (2001). Sun protection
practices in preadolescents and adolescents: a school-based survey of almost 25,000
Connecticut schoolchildren. J Am Acad Dermatol, 44(3), 512-519. doi:
10.1067/mjd.2001.111621

109
Cormier, J. N., Xing, Y., Ding, M., Lee, J. E., Mansfield, P. F., Gershenwald, J. E., . . . Du, X. L.
(2006). Ethnic differences among patients with cutaneous melanoma. Arch Intern Med,
166(17), 1907-1914. doi: 10.1001/archinte.166.17.1907
Coups E.J., S. J. L., Hudson S.V., et al. . (2013). Linguistic acculturation and skin cancer-related
behaviors among Hispanics in the southern and western United States. JAMA Dermatol,
149(6), 679–686. doi: 10.1001/jamadermatol.2013.745
Coups, E. J., Manne, S. L., & Heckman, C. J. (2008). Multiple skin cancer risk behaviors in the
U.S. population. Am J Prev Med, 34(2), 87-93. doi: 10.1016/j.amepre.2007.09.032
Coups, E. J., Stapleton, J. L., Hudson, S. V., Medina-Forrester, A., Natale-Pereira, A., &
Goydos, J. S. (2012). Sun protection and exposure behaviors among Hispanic adults in
the United States: differences according to acculturation and among Hispanic subgroups.
BMC Public Health, 12, 985. doi: 10.1186/1471-2458-12-985
Coups, E. J., Stapleton, J. L., Manne, S. L., Hudson, S. V., Medina-Forrester, A., Rosenberg, S.
A., . . . Goydos, J. S. (2014). Psychosocial correlates of sun protection behaviors among
U.S. Hispanic adults. J Behav Med, 37(6), 1082-1090. doi: 10.1007/s10865-014-9558-5
Craciun, C., Schuz, N., Lippke, S., & Schwarzer, R. (2012). A mediator model of sunscreen use:
a longitudinal analysis of social-cognitive predictors and mediators. Int J Behav Med,
19(1), 65-72. doi: 10.1007/s12529-011-9153-x
Cress, R. D., & Holly, E. A. (1997). Incidence of cutaneous melanoma among non-Hispanic
whites, Hispanics, Asians, and blacks: an analysis of california cancer registry data,
1988-93. Cancer Causes Control, 8(2), 246-252.
Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. . Psychometrika,
16(3 ), 297–334. doi: doi:10.1007/BF02310555
Cust, A. E., Jenkins, M. A., Goumas, C., Armstrong, B. K., Schmid, H., Aitken, J. F., . . . Mann,
G. J. (2011). Early-life sun exposure and risk of melanoma before age 40 years. Cancer
Causes Control, 22(6), 885-897. doi: 10.1007/s10552-011-9762-3
Davies, H., Bignell, G. R., Cox, C., Stephens, P., Edkins, S., Clegg, S., . . . Futreal, P. A. (2002).
Mutations of the BRAF gene in human cancer. Nature, 417(6892), 949-954. doi:
10.1038/nature00766
Davison, K. K., & Birch, L. L. (2002). Obesigenic families: parents' physical activity and dietary
intake patterns predict girls' risk of overweight. Int J Obes Relat Metab Disord, 26(9),
1186-1193. doi: 10.1038/sj.ijo.0802071

110
Dennis, L. K., Vanbeek, M. J., Beane Freeman, L. E., Smith, B. J., Dawson, D. V., & Coughlin,
J. A. (2008). Sunburns and risk of cutaneous melanoma: does age matter? A
comprehensive meta-analysis. Ann Epidemiol, 18(8), 614-627. doi:
10.1016/j.annepidem.2008.04.006
Dixon, H., Borland, R., & Hill, D. (1999). Sun protection and sunburn in primary school
children: the influence of age, gender, and coloring. Prev Med, 28(2), 119-130. doi:
10.1006/pmed.1998.0392
Dobbinson, S., Wakefield, M., Hill, D., Girgis, A., Aitken, J. F., Beckmann, K., . . . Bowles, K.
A. (2012). Children's sun exposure and sun protection: prevalence in Australia and
related parental factors. J Am Acad Dermatol, 66(6), 938-947. doi:
10.1016/j.jaad.2011.06.015
Donavan, D. T., & Singh, S. N. (1999). Sun-safety behavior among elementary school children:
the role of knowledge, social norms, and parental involvement. Psychol Rep, 84(3 Pt 1),
831-836. doi: 10.2466/pr0.1999.84.3.831
Dubas, L. E., & Adams, B. B. (2012). Sunscreen use and availability among female collegiate
athletes. J Am Acad Dermatol, 67(5), 876 e871-876. doi: 10.1016/j.jaad.2011.11.962
Dudley, D. A., Cotton, W. G., & Peralta, L. R. (2015). Teaching approaches and strategies that
promote healthy eating in primary school children: a systematic review and meta-
analysis. Int J Behav Nutr Phys Act, 12(1), 28. doi: 10.1186/s12966-015-0182-8
Eide, M. J., & Weinstock, M. A. (2005). Association of UV index, latitude, and melanoma
incidence in nonwhite populations--US Surveillance, Epidemiology, and End Results
(SEER) Program, 1992 to 2001. Arch Dermatol, 141(4), 477-481. doi:
10.1001/archderm.141.4.477
Elwood, J. M., & Jopson, J. (1997). Melanoma and sun exposure: an overview of published
studies. Int J Cancer, 73(2), 198-203.
Emmons, K. M., Geller, A. C., Puleo, E., Savadatti, S. S., Hu, S. W., Gorham, S., . . . Dana-
Farber Skin Cancer Screening, G. (2011). Skin cancer education and early detection at
the beach: a randomized trial of dermatologist examination and biometric feedback. J Am
Acad Dermatol, 64(2), 282-289. doi: 10.1016/j.jaad.2010.01.040
Ennis, S. R., Ríos-Vargas, M., & Albert, N. G. (2011). The Hispanic Population: 2010. U.S.
Census Bureau. Retrieved from http://www.census.gov/prod/cen2010/briefs/c2010br-‐
04.pdf.

111
Fitzpatrick, T. B. (1988). The validity and practicality of sun-reactive skin types I through VI.
Arch Dermatol, 124(6), 869-871.
Gandini, S., Sera, F., Cattaruzza, M. S., Pasquini, P., Picconi, O., Boyle, P., & Melchi, C. F.
(2005). Meta-analysis of risk factors for cutaneous melanoma: II. Sun exposure. Eur J
Cancer, 41(1), 45-60. doi: 10.1016/j.ejca.2004.10.016
Geller, A., Rutsch, L., Kenausis, K., & Zhang, Z. (2003). Evaluation of the SunWise School
Program. J Sch Nurs, 19(2), 93-99.
Geller, A. C., Colditz, G., Oliveria, S., Emmons, K., Jorgensen, C., Aweh, G. N., & Frazier, A.
L. (2002). Use of sunscreen, sunburning rates, and tanning bed use among more than 10
000 US children and adolescents. Pediatrics, 109(6), 1009-1014.
Glanz, K., Chang, L., Song, V., Silverio, R., & Muneoka, L. (1998). Skin cancer prevention for
children, parents, and caregivers: a field test of Hawaii's SunSmart program. J Am Acad
Dermatol, 38(3), 413-417.
Glanz, K., Lew, R. A., Song, V., & Cook, V. A. (1999). Factors associated with skin cancer
prevention practices in a multiethnic population. Health Educ Behav, 26(3), 344-359.
Glanz, K., & Mayer, J. A. (2005). Reducing ultraviolet radiation exposure to prevent skin cancer
methodology and measurement. Am J Prev Med, 29(2), 131-142. doi:
10.1016/j.amepre.2005.04.007
Glanz, K., Yaroch, A. L., Dancel, M., Saraiya, M., Crane, L. A., Buller, D. B., . . . Robinson, J.
K. (2008). Measures of sun exposure and sun protection practices for behavioral and
epidemiologic research. Arch Dermatol, 144(2), 217-222. doi:
10.1001/archdermatol.2007.46
Gordon, R. (2013). Skin cancer: an overview of epidemiology and risk factors. Semin Oncol
Nurs, 29(3), 160-169. doi: 10.1016/j.soncn.2013.06.002
Green, A. C., Wallingford, S. C., & McBride, P. (2011). Childhood exposure to ultraviolet
radiation and harmful skin effects: epidemiological evidence. Prog Biophys Mol Biol,
107(3), 349-355. doi: 10.1016/j.pbiomolbio.2011.08.010
Guy, G. P., Jr., Berkowitz, Z., Tai, E., Holman, D. M., Everett Jones, S., & Richardson, L. C.
(2014). Indoor tanning among high school students in the United States, 2009 and 2011.
JAMA Dermatol, 150(5), 501-511. doi: 10.1001/jamadermatol.2013.7124
Hall, H. I., Jones, S. E., & Saraiya, M. (2001). Prevalence and correlates of sunscreen use among
US high school students. J Sch Health, 71(9), 453-457.

112
Hall, H. I., Jorgensen, C. M., McDavid, K., Kraft, J. M., & Breslow, R. (2001). Protection from
sun exposure in US white children ages 6 months to 11 years. Public Health Rep, 116(4),
353-361.
Hall, H. I., May, D. S., Lew, R. A., Koh, H. K., & Nadel, M. (1997). Sun protection behaviors of
the U.S. white population. Prev Med, 26(4), 401-407. doi: 10.1006/pmed.1997.0168
Hart, K. M., & Demarco, R. F. (2008). Primary prevention of skin cancer in children and
adolescents: a review of the literature. J Pediatr Oncol Nurs, 25(2), 67-78. doi:
10.1177/1043454208314499
Hay, J., Coups, E. J., Ford, J., & DiBonaventura, M. (2009). Exposure to mass media health
information, skin cancer beliefs, and sun protection behaviors in a United States
probability sample. J Am Acad Dermatol, 61(5), 783-792. doi:
10.1016/j.jaad.2009.04.023
Heckman, C. J., & Cohen-Filipic, J. (2012). Brief report: ultraviolet radiation exposure,
considering acculturation among Hispanics (project URECAH). J Cancer Educ, 27(2),
342-346. doi: 10.1007/s13187-011-0308-x
Heckman, C. J., & Coups, E. J. (2011). Correlates of sunscreen use among high school students:
a cross-sectional survey. BMC Public Health, 11, 679. doi: 10.1186/1471-2458-11-679
Hernandez, C., Calero, D., Robinson, G., Mermelstein, R., & Robinson, J. K. (2012).
Comparison of sunscreen availability in Chicago Hispanic and non-Hispanic
neighborhoods. Photodermatol Photoimmunol Photomed, 28(5), 244-249. doi:
10.1111/j.1600-0781.2012.00688.x
Hernandez, L., Rodriguez, A. M., & Spirito, A. (2015). Brief Family-Based Intervention for
Substance Abusing Adolescents. Child Adolesc Psychiatr Clin N Am, 24(3), 585-599.
doi: 10.1016/j.chc.2015.02.010
Hillhouse, J., Turrisi, R., Jaccard, J., & Robinson, J. (2012). Accuracy of self-reported sun
exposure and sun protection behavior. Prev Sci, 13(5), 519-531. doi: 10.1007/s11121-
012-0278-1
Holman, D. M., & Watson, M. (2013). Correlates of intentional tanning among adolescents in
the United States: a systematic review of the literature. J Adolesc Health, 52(5 Suppl),
S52-59. doi: 10.1016/j.jadohealth.2012.09.021
Hu, S., Ma, F., Collado-Mesa, F., & Kirsner, R. S. (2004). UV radiation, latitude, and melanoma
in US Hispanics and blacks. Arch Dermatol, 140(7), 819-824. doi:
10.1001/archderm.140.7.819

113
Hu, S., Parmet, Y., Allen, G., Parker, D. F., Ma, F., Rouhani, P., & Kirsner, R. S. (2009).
Disparity in melanoma: a trend analysis of melanoma incidence and stage at diagnosis
among whites, Hispanics, and blacks in Florida. Arch Dermatol, 145(12), 1369-1374.
doi: 10.1001/archdermatol.2009.302
Hunter, S., Love-Jackson, K., Abdulla, R., Zhu, W., Lee, J. H., Wells, K. J., & Roetzheim, R.
(2010). Sun protection at elementary schools: a cluster randomized trial. J Natl Cancer
Inst, 102(7), 484-492. doi: 10.1093/jnci/djq010
International Agency for Research on Cancer, I. (2010). Cancer incidence in five continents :
volumes I to IX IARC CancerBase No. 9. (pp. 1 online resource.). Lyon, France: IARC.
Jemal, A., Saraiya, M., Patel, P., Cherala, S. S., Barnholtz-Sloan, J., Kim, J., . . . Wingo, P. A.
(2011). Recent trends in cutaneous melanoma incidence and death rates in the United
States, 1992-2006. J Am Acad Dermatol, 65(5 Suppl 1), S17-25 e11-13. doi:
10.1016/j.jaad.2011.04.032
Juckett, G. (2013). Caring for Latino patients. Am Fam Physician, 87(1), 48-54.
Kanetsky, P. A., Rebbeck, T. R., Hummer, A. J., Panossian, S., Armstrong, B. K., Kricker, A., . .
. Berwick, M. (2006). Population-based study of natural variation in the melanocortin-1
receptor gene and melanoma. Cancer Res, 66(18), 9330-9337. doi: 10.1158/0008-
5472.CAN-06-1634
Katiria Perez, G., & Cruess, D. (2014). The impact of familism on physical and mental health
among Hispanics in the United States. Health Psychol Rev, 8(1), 95-127. doi:
10.1080/17437199.2011.569936
Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New
York: Guilford Press.
Koh, H. K., Bak, S. M., Geller, A. C., Mangione, T. W., Hingson, R. W., Levenson, S. M., . . .
Howland, J. (1997). Sunbathing habits and sunscreen use among white adults: results of a
national survey. Am J Public Health, 87(7), 1214-1217.
Kristjansson, S., Branstrom, R., Ullen, H., & Helgason, A. R. (2003). Transtheoretical model:
investigation of adolescents' sunbathing behaviour. Eur J Cancer Prev, 12(6), 501-508.
doi: 10.1097/01.cej.0000102805.33147.85
Lanza, S. T., Collins, L. M., Lemmon, D. R., & Schafer, J. L. (2007). PROC LCA: A SAS
Procedure for Latent Class Analysis. Struct Equ Modeling, 14(4), 671-694.

114
Lin, J. S., Eder, M., & Weinmann, S. (2011). Behavioral counseling to prevent skin cancer: a
systematic review for the U.S. Preventive Services Task Force. Ann Intern Med, 154(3),
190-201. doi: 10.7326/0003-4819-154-3-201102010-00009
Linos, E., Swetter, S. M., Cockburn, M. G., Colditz, G. A., & Clarke, C. A. (2009). Increasing
burden of melanoma in the United States. J Invest Dermatol, 129(7), 1666-1674. doi:
10.1038/jid.2008.423
Little, R. J. A., & Rubin, D. B. (1987). Statistical analysis with missing data. New York: Wiley.
Lomas, A., Leonardi-Bee, J., & Bath-Hextall, F. (2012). A systematic review of worldwide
incidence of nonmelanoma skin cancer. Br J Dermatol, 166(5), 1069-1080. doi:
10.1111/j.1365-2133.2012.10830.x
Lowe, J. B., Balanda, K. P., Gillespie, A. M., Del Mar, C. B., & Gentle, A. F. (1993). Sun-
related attitudes and beliefs among Queensland school children: the role of gender and
age. Aust J Public Health, 17(3), 202-208.
Lower, T., Girgis, A., & Sanson-Fisher, R. (1998a). How valid is adolescents' self-report as a
way of assessing sun protection practices? Prev Med, 27(3), 385-390.
Lower, T., Girgis, A., & Sanson-Fisher, R. (1998b). The prevalence and predictors of solar
protection use among adolescents. Prev Med, 27(3), 391-399. doi:
10.1006/pmed.1998.0309
Ma, F., Collado-Mesa, F., Hu, S., & Kirsner, R. S. (2007). Skin cancer awareness and sun
protection behaviors in white Hispanic and white non-Hispanic high school students in
Miami, Florida. Arch Dermatol, 143(8), 983-988. doi: 10.1001/archderm.143.8.983
MacKinnon, D. (2008). Introduction to statistical mediation analysis. New York: Lawrence
Erlbaum Associates.
Mackinnon, D. P., & Dwyer, J. H. (1993). Estimating Mediated Effects in Prevention Studies.
Evaluation Review, 17(2), 144-158.
Maldonado, J. L., Fridlyand, J., Patel, H., Jain, A. N., Busam, K., Kageshita, T., . . . Bastian, B.
C. (2003). Determinants of BRAF mutations in primary melanomas. J Natl Cancer Inst,
95(24), 1878-1890.
Malotte, C. K., Jarvis, B., Fishbein, M., Kamb, M., Iatesta, M., Hoxworth, T., . . . Bolan, G.
(2000). Stage of change versus an integrated psychosocial theory as a basis for
developing effective behaviour change interventions. The Project RESPECT Study
Group. AIDS Care, 12(3), 357-364. doi: 10.1080/09540120050043016

115
Marks, R., & Hill, D. (1988). Behavioural change in adolescence: a major challenge for skin-
cancer control in Australia. Med J Aust, 149(10), 514-515.
Mermelstein, R. J., & Riesenberg, L. A. (1992). Changing knowledge and attitudes about skin
cancer risk factors in adolescents. Health Psychol, 11(6), 371-376.
Muthén, B., Brown, C. H., Masyn, K., Jo, B., Khoo, S. T., Yang, C. C., . . . Liao, J. (2002).
General growth mixture modeling for randomized preventive interventions. Biostatistics,
3(4), 459-475. doi: 10.1093/biostatistics/3.4.459
Muthén, B., & Muthén, L. K. (2000). Integrating person-centered and variable-centered analyses:
growth mixture modeling with latent trajectory classes. Alcohol Clin Exp Res, 24(6), 882-
891.
Muthén, L. K., & Muthén, B. O. (1998-2011). Mplus User's Guide. Los Angeles, CA: Muthén &
Muthén.
National Cancer Institute. SEER Stat Fact Sheets: Melanoma of the Skin. Retrieved 10-22-
2014, from http://seer.cancer.gov/statfacts/html/melan.html
Negy, C., & Woods, D. J. . (1992). The Importance of Acculturation in Understanding Research
with Hispanic-Americans. Hispanic Journal of Behavioral Sciences,, 14(2), 224–247.
doi: 10.1177/07399863920142003
Nelson, A. A., & Tsao, H. (2009). Melanoma and genetics. Clin Dermatol, 27(1), 46-52. doi:
10.1016/j.clindermatol.2008.09.005
O'Riordan, D. L., Geller, A. C., Brooks, D. R., Zhang, Z., & Miller, D. R. (2003). Sunburn
reduction through parental role modeling and sunscreen vigilance. J Pediatr, 142(1), 67-
72. doi: 10.1067/mpd.2003.mpd039
O'Riordan, D. L., Nehl, E., Gies, P., Bundy, L., Burgess, K., Davis, E., & Glanz, K. (2009).
Validity of covering-up sun-protection habits: Association of observations and self-
report. J Am Acad Dermatol, 60(5), 739-744. doi: 10.1016/j.jaad.2008.12.015
Olson, A. L., Gaffney, C. A., Starr, P., & Dietrich, A. J. (2008). The impact of an appearance-
based educational intervention on adolescent intention to use sunscreen. Health Educ
Res, 23(5), 763-769. doi: 10.1093/her/cym005
Ortiz, A. (March 2015). Paper presented at the American Academy of Dermatology 73rd Annual
Meeting, San Francisco.

116
Park, S. L., Le Marchand, L., Wilkens, L. R., Kolonel, L. N., Henderson, B. E., Zhang, Z. F., &
Setiawan, V. W. (2012). Risk factors for malignant melanoma in white and non-
white/non-African American populations: the multiethnic cohort. Cancer Prev Res
(Phila), 5(3), 423-434. doi: 10.1158/1940-6207.CAPR-11-0460
Pennello, G. A., Devesa, S. S., & Gail, M. H. (1999). Using a mixed effects model to estimate
geographic variation in cancer rates. Biometrics, 55(3), 774-781.
Perez, A. D., & Hirschman, C. (2009). The Changing Racial and Ethnic Composition of the US
Population: Emerging American Identities. Popul Dev Rev, 35(1), 1-51. doi:
10.1111/j.1728-4457.2009.00260.x
Pew Forum on Religion & Public Life. (2007). Changing Faiths: Latinos and the Transformation
of American Religion Retrieved November 20, 2014, from
http://www.pewhispanic.org/2007/04/25/changing-‐faiths-‐latinos-‐and-‐the-‐transformation-‐of-‐
american-‐religion/
Phinney, J. S. (1990). Ethnic identity in adolescents and adults: review of research. Psychol Bull,
108(3), 499-514.
Pichon, L. C., Corral, I., Landrine, H., Mayer, J. A., & Adams-Simms, D. (2010). Perceived skin
cancer risk and sunscreen use among African American adults. J Health Psychol, 15(8),
1181-1189. doi: 10.1177/1359105310364177
Pichon, L. C., Corral, I., Landrine, H., Mayer, J. A., & Norman, G. J. (2010). Sun-protection
behaviors among African Americans. Am J Prev Med, 38(3), 288-295. doi:
10.1016/j.amepre.2009.10.041
Pipitone, M., Robinson, J. K., Camara, C., Chittineni, B., & Fisher, S. G. (2002). Skin cancer
awareness in suburban employees: a Hispanic perspective. J Am Acad Dermatol, 47(1),
118-123.
Pollitt, R. A., Clarke, C. A., Swetter, S. M., Peng, D. H., Zadnick, J., & Cockburn, M. (2011).
The expanding melanoma burden in California hispanics: Importance of socioeconomic
distribution, histologic subtype, and anatomic location. Cancer, 117(1), 152-161. doi:
10.1002/cncr.25355
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects
in simple mediation models. Behav Res Methods Instrum Comput, 36(4), 717-731.
Reynolds, K. D., Blaum, J. M., Jester, P. M., Weiss, H., Soong, S. J., & Diclemente, R. J. (1996).
Predictors of sun exposure in adolescents in a southeastern U.S. population. J Adolesc
Health, 19(6), 409-415.

117
Reynolds, K. D., Buller, D. B., Yaroch, A. L., Maloy, J. A., & Cutter, G. R. (2006). Mediation of
a middle school skin cancer prevention program. Health Psychol, 25(5), 616-625. doi:
10.1037/0278-6133.25.5.616
Rimal, R. N., Lapinski, M. K., Cook, R. J., & Real, K. (2005). Moving toward a theory of
normative influences: how perceived benefits and similarity moderate the impact of
descriptive norms on behaviors. J Health Commun, 10(5), 433-450. doi:
10.1080/10810730591009880
Robinson, J. K., Rademaker, A. W., Sylvester, J. A., & Cook, B. (1997). Summer sun exposure:
knowledge, attitudes, and behaviors of Midwest adolescents. Prev Med, 26(3), 364-372.
doi: 10.1006/pmed.1997.0156
Robinson, J. K., Rigel, D. S., & Amonette, R. A. (2000). Summertime sun protection used by
adults for their children. J Am Acad Dermatol, 42(5 Pt 1), 746-753.
Rogers, H. W., Weinstock, M. A., Feldman, S. R., & Coldiron, B. M. (2015). Incidence Estimate
of Nonmelanoma Skin Cancer (Keratinocyte Carcinomas) in the US Population, 2012.
JAMA Dermatol. doi: 10.1001/jamadermatol.2015.1187
Rouhani, P., Parmet, Y., Bessell, A. G., Peay, T., Weiss, A., & Kirsner, R. S. (2009).
Knowledge, attitudes, and behaviors of elementary school students regarding sun
exposure and skin cancer. Pediatr Dermatol, 26(5), 529-535. doi: 10.1111/j.1525-
1470.2009.00908.x
Rouhani, P., Pinheiro, P. S., Sherman, R., Arheart, K., Fleming, L. E., Mackinnon, J., & Kirsner,
R. S. (2010). Increasing rates of melanoma among nonwhites in Florida compared with
the United States. Arch Dermatol, 146(7), 741-746. doi: 10.1001/archdermatol.2010.133
Saraiya, M., Glanz, K., Briss, P. A., Nichols, P., White, C., Das, D., . . . Rochester, P. (2004).
Interventions to prevent skin cancer by reducing exposure to ultraviolet radiation: a
systematic review. Am J Prev Med, 27(5), 422-466. doi: 10.1016/j.amepre.2004.08.009
Schuz, N., Eid, M. (2012). Beyond the usual suspects: target group- and behavior-specific factors
add to a theory-based sun protection intervention for teenagers. J Behav Med, 36(5), 508-
519. doi: 10.1007/s10865-012-9445-x
Seidler, A. M., Pennie, M. L., Veledar, E., Culler, S. D., & Chen, S. C. (2010). Economic burden
of melanoma in the elderly population: population-based analysis of the Surveillance,
Epidemiology, and End Results (SEER)--Medicare data. Arch Dermatol, 146(3), 249-
256. doi: 10.1001/archdermatol.2009.389

118
Spencer, J. E., Cooper, H. C., & Milton, B. (2013). The lived experiences of young people (13-
16 years) with Type 1 diabetes mellitus and their parents--a qualitative phenomenological
study. Diabet Med, 30(1), e17-24. doi: 10.1111/dme.12021
StataCorp. (2011). Stata Statistical Software: Release 12 (Version Release 12). College Station,
TX: StataCorp LP.
Stern, R. S., Weinstein, M. C., & Baker, S. G. (1986). Risk reduction for nonmelanoma skin
cancer with childhood sunscreen use. Arch Dermatol, 122(5), 537-545.
Thieden, E., Philipsen, P. A., Sandby-Moller, J., & Wulf, H. C. (2005). Sunscreen use related to
UV exposure, age, sex, and occupation based on personal dosimeter readings and sun-
exposure behavior diaries. Arch Dermatol, 141(8), 967-973. doi:
10.1001/archderm.141.8.967
Thomson, C. E., White, K. M., & Hamilton, K. (2012). Investigating mothers' decisions about
their child's sun-protective behaviour using the Theory of Planned Behaviour. J Health
Psychol, 17(7), 1001-1010. doi: 10.1177/1359105311433905
Tripp, M. K., Carvajal, S. C., McCormick, L. K., Mueller, N. H., Hu, S. H., Parcel, G. S., &
Gritz, E. R. (2003). Validity and reliability of the parental sun protection scales. Health
Educ Res, 18(1), 58-73.
Tripp, M. K., Diamond, P. M., Vernon, S. W., Swank, P. R., Dolan Mullen, P., & Gritz, E. R.
(2013). Measures of parents' self-efficacy and perceived barriers to children's sun
protection: construct validity and reliability in melanoma survivors. Health Educ Res,
28(5), 828-842. doi: 10.1093/her/cys114
Turner, L. R., & Mermelstein, R. J. (2005). Psychosocial characteristics associated with sun
protection practices among parents of young children. J Behav Med, 28(1), 77-90.
U.S. Cancer Statistics Working Group. (2014). United States Cancer Statistics: 1999–2011
Incidence and Mortality Web-based Report. . Atlanta: Retrieved from
http://www.cdc.gov/uscs.
Unger, J. B., Gallaher, P., Shakib, S., Ritt-Olson, A., Palmer, P. H., & Johnson, C. A. . (2002).
The AHIMSA acculturation scale: A new measure of acculturation for adolescents in a
multicultural society. The Journal of Early Adolescence, 22(3), 225-251.
United States Census Bureau. (2013). Hispanic or Latino Origin by Specific Origin. from
http://factfinder2.census.gov

119
Viros, A., Sanchez-Laorden, B., Pedersen, M., Furney, S. J., Rae, J., Hogan, K., . . . Marais, R.
(2014). Ultraviolet radiation accelerates BRAF-driven melanomagenesis by targeting
TP53. Nature, 511(7510), 478-482. doi: 10.1038/nature13298
Whiteman, D., & Green, A. (1994). Melanoma and sunburn. Cancer Causes Control, 5(6), 564-
572.
Whiteman, D. C., Whiteman, C. A., & Green, A. C. (2001). Childhood sun exposure as a risk
factor for melanoma: a systematic review of epidemiologic studies. Cancer Causes
Control, 12(1), 69-82.
World Health Organization. (2002). Sun protection: an essential element of health-promoting
schools.: Retrieved from Retrieved from
http://cdrwww.who.int/entity/school_youth_health/media/en/456.pdf.
Wurdinger, S. D., & Carlson, J. (2010). Teaching for experiential learning : five approaches that
work. Lanham, Md.: Rowman & Littlefield Education.

120
APPENDIX: Table of Intraclass Correlations

Item
Variable
(Pre to Post Change) ICC
Standard
Error CI 95%
Use of sunscreen at school a2appspf15 0.01 0.008 (0.005, 0.044)
Use of hats at school a3worehat 0.01 0.004 (0.003, 0.021)
Use of long sleeves at school a4lngsleves 0.03 0.012 (0.017, 0.069)
Use of long pants at school a5longpants 0.07 0.021 (0.036, 0.122)
Shade-seeking at school a7plyinshd 0.01 0.004 (0.003, 0.019)
Use of sunscreen at home b2appsnscrn 0.01 0.004 (0.002, 0.019)
Use of hats at home b3worehat 0.01 0.002 (0.001, 0.149)
Use of long sleeves at home b4lngsleves 0.03 0.012 (0.016, 0.067)
Use of long pants at home b5longpants 0.04 0.014 (0.021, 0.081)
Shade-seeking at home b7plyinshde 0.01 0.003 (0.001, 0.015)
Knowledge of sun protection knowledge 0.05 0.016 (0.026, 0.092)
Friends play in shade (norms) h1frndsplayinshde 0.02 0.007 (0.007, 0.038)
Friends wear hats (norms) h2frndswrhats 0.01 0.004 (0.003, 0.019)
Friends wear long-sleeves (norms) h3frndswrlngslves 0.02 0.008 (0.009, 0.043)
Friends wearing sunscreen (norms) h4frndswrsnscrn 0.01 0.004 (0.002, 0.022)
Self-efficacy to use sunscreen j2spf15 0.01 0.005 (0.004, 0.029)
Self-efficacy to use long sleeves j3lngsleeves 0.01 0.004 (0.005, 0.024)
Self-efficacy to use hat j4wearhat 0.01 0.003 (0.001, 0.015)
Self-efficacy to play in shade j5plyinshde 0.01 0.002 (0.001, 0.014)
Appendix: Table of Intra-Class Correlations

Abstract (if available)

Abstract Malignant melanoma is the most serious form of skin cancer, accounting for more than 75% of all skin cancer deaths. In contrast with decreasing trends for most major cancer sites, incidence rates of melanoma have been rising for the past several decades globally. Because childhood sun exposure and sunburn have been associated with significantly increased risk of melanoma in adulthood, children are particularly important targets for melanoma primary prevention, which includes the reduction of solar overexposure through methods such as limiting sun during peak hours, using sunscreen with adequate sun protection factor, using protective clothing including long sleeves, long pants, and hats, and seeking shade. ❧ This dissertation examines determinants of sun protection behaviors in a multiethnic sample of children participating in a school-based skin cancer prevention intervention. To date, interventions targeting children have been largely unsuccessful in improving UV-related behaviors. In addition, data regarding the sun protective behaviors of non-white children are scarce, and few skin cancer prevention interventions have been conducted amongst ethnically diverse samples despite rising rates of melanoma in non-white populations, particularly among U.S. Hispanics. ❧ Findings from our studies indicate that ethnic minority children experience similar sun protection correlates previously observed among non-Hispanic white (NHW) children. Strong positive associations were found between parental variables and child sun protection across studies. Mediation analysis between the intervention and sunscreen use found short-term effects for psychosocial variables

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The role of acculturation in the sun-safe behaviors of US Latino adults in the United States

PDF

Examination of sun awareness behavior, knowledge, and attitudes before and after a sun awareness program in 4th to 8th grade students

PDF

Carcinogenic exposures in racial/ethnic groups

PDF

Genetic variants and smoking progression in Chinese adolescents

PDF

Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California

PDF

Multilevel sociodemographic correlates of the health and healthcare utilization of childhood cancer survivors

PDF

The role of dyadic and triadic factors on psychosocial wellbeing and healthcare interactions among childhood cancer survivors, parents, and medical providers

PDF

Sociocultural stress, coping and substance use among Hispanic/Latino adolescents

PDF

Melanoma in children, adolescents, and young adults

PDF

A sociocultural and developmental approach to intimate partner violence among a sample of Hispanic emerging adults

PDF

Multilevel influences of care engagement and long-term survival among childhood, adolescent, and young adult cancer survivors

PDF

The effect of an educational fotonovela on the prevention of secondhand smoke

PDF

Investigating factors that influence peer relationships and obesity during middle childhood

PDF

Psychosocial and behavioral ractors associated with emotional eating in adolescents

PDF

Diffusion of a peer‐led suicide preventive intervention in secondary schools: strategies to increase effectiveness of peer‐led interventions

PDF

Childhood cancer survivorship: parental factors associated with survivor's follow-up care behavior and mental health

PDF

Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents

PDF

Effects of sugar and fiber consumption in minority adolescents and self-tracking as a potential dietary intervention tool

PDF

Competing risks: the role of the perceived consequences of refusing to share injection equipment among injection drug users

PDF

The role of physicians in the late-stage diagnosis of melanoma among: a population-based study in Los Angeles County

Asset Metadata

Creator Miller, Kimberly Ann (author)

Core Title Psychosocial and cultural factors in the primary prevention of melanoma targeted to multiethnic children

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior Research)

Publication Date 11/09/2015

Defense Date 09/18/2015

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag children and adolescents,Hispanics,Melanoma,OAI-PMH Harvest,skin cancer,sun protection

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cockburn, Myles (committee chair), Huh, Jimi (committee member), Peng, David (committee member), Richardson, Jean (committee member), Unger, Jennifer (committee member)

Creator Email kim.miller@med.usc.edu,kimmiller23@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-198540

Unique identifier UC11277470

Identifier etd-MillerKimb-4024.pdf (filename),usctheses-c40-198540 (legacy record id)

Legacy Identifier etd-MillerKimb-4024.pdf

Dmrecord 198540

Document Type Dissertation

Format application/pdf (imt)

Rights Miller, Kimberly Ann

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA