Childrens sun exposure and sun protection: Prevalence in Australia and related parental factors

Suzanne Dobbinson, PhD,a Melanie Wakefield, PhD,a David Hill, PhD,a Afaf Girgis, PhD,b Joanne F. Aitken, PhD,c Kerri Beckmann, MPH,d Anthony I. Reeder, PhD,e Natalie Herd, BA & BSc (Hons),a Matthew J. Spittal, PhD,a Andrew Fairthorne, LLB, BSc,a and Kelly-Ann Bowles, BSca Melbourne, Newcastle, Brisbane, and Adelaide, Australia; and Dunedin, New Zealand
Background: Reducing childhood exposure to ultraviolet radiation is important to minimize lifetime skin cancer risk. Objectives: We sought to describe the prevalence of childrens sun-related behaviors and associated parental and other factors. Methods: In weekly cross-sectional telephone interviews during summer, 1140 parents/guardians of children aged 0 to 11 years were recruited. Parents provided proxy reports for one of their children. Key questions related to weekend sun protection and sunburn, parents sun-related attitudes, and demographic characteristics. Potential predictors of childrens sun protection and sunburn were analyzed adjusting for covariates including weather conditions on the previous weekend. Results: On summer weekends, 73% of children spent longer than 15 minutes outdoors in peak ultraviolet radiation periods. Of these, 64% were protected by a hat and 58% by sun-protection factor 15 or higher sunscreen, 32% stayed under shade, and 18% wore three-quarter or long-sleeved tops. Overall, 8% of children had sunburn. Parental attitudes were typically supportive of childrens sun protection. Parental use of hats (odds ratio [OR] 3.1; 95% confidence interval [CI] 1.6-6.2), shade (OR 9.6; 95% CI 4.4-20.8), sunscreen (OR 12.6; 95% CI 5.2-30.4), longer leg cover (OR 10.3; 95% CI 4.4-24.0), and two or more protective behaviors (OR 5.7; 95% CI 2.8-11.9) increased the odds of their children practicing these behaviors, as did some parental attitudes. Limitations: We relied on cross-sectional parent reports. Conclusion: Although childrens sun protection was favorable, there was room for improvement. Health promotion to improve sun-protection practices in adults may benefit childrens sun-safe behaviors. ( J Am Acad Dermatol 2012;66:938-47.) Key words: skin cancer prevention; sun protection; sunburn.

kin cancer is a signicant global health problem causing an estimated 60,000 deaths in 2002.1 Reducing childhood exposure to ultraviolet (UV) radiation (UVR) is crucial to preventing skin cancer in later life.2,3 Two thirds of Australians will be given the diagnosis of the disease in their
From the Cancer Council Victoria, Melbournea; Center for Health Research and Psycho-Oncology, Cancer Council New South Wales, University of Newcastle and Hunter Medical Research Instituteb; Cancer Council Queensland, Brisbanec; Cancer Council South Australia, Adelaided; and Cancer Society of New Zealand Social and Behavioral Research Unit, Department of Preventive and Social Medicine, Dunedin School of Medicine, University of Otago.e Supported by state cancer councils of Australia and the Australian Government Department of Health and Aging.

Abbreviations used: CI: OR: SPF: UV: UVR: condence interval odds ratio sun protection factor ultraviolet ultraviolet radiation

Conflicts of interest: None declared. Accepted for publication June 6, 2011. Reprint requests: Suzanne Dobbinson, PhD, Center for Behavioral Research in Cancer, Cancer Council Victoria, 1 Rathdowne St, Carlton, Victoria 3053, Australia. E-mail: Suzanne.Dobbinson@ cancervic.org.au. Published online September 5, 2011. 0190-9622/$36.00 2011 by the American Academy of Dermatology, Inc. doi:10.1016/j.jaad.2011.06.015

938

J AM ACAD DERMATOL
VOLUME 66, NUMBER 6

Dobbinson et al 939

lifetime.4 Without the major investments in prevenand sun-protection levels and to examine parental tion in Australia (outlined below),5,6 skin cancer and other factors. incidence and mortality would likely be much worse.4,7,8 For younger cohorts born after 1960 METHODS nonmelanoma skin cancer incidence has stabilized4 Study design and participants and melanoma mortality reduced.7 Parents and guardians (parents) were recruited State cancer councils and health departments from a subset of 5073 adults (18-69 years) participathave implemented largeing in a nationwide survey of scale public education on sun protection commisCAPSULE SUMMARY skin cancer in Australia for sioned by the state cancer over 30 years. The rst public councils and Australian govChildhood is a critical period for education included distribuernment.20 During the sumreducing ultraviolet exposure and tion of information in lecmer from November 2003 to lifetime skin cancer risk. tures, pamphlets, iers, and January 2004, weekly crossThis study found Australian children newsletters in the 1960s.7 By sectional telephone surveys were generally well protected from sun the early 1980s the first meswere conducted except durexposure on summer weekends sages appeared on televiing public holidays. Parents although 8% had sunburn. sion, and by the 1990s paid residing with children aged television advertisements 11 years or younger were Parents personal behaviors were were regularly broadcast interviewed about one childs strongly associated with their childrens during summer.9 Campaign sun protection and sunburn sun protection, suggesting that messages and images aimed on the previous weekend. promoting sun-safe behavior in adults to encourage parents to proSun-protection behaviors of may benefit children. tect young children from the parent and child were assun included the Kids Cook sessed by similar measures, Quick billboard, and Sid Seagull9 and Seymour the with parents assessed first. The initial survey sample Snowman.10 Most television campaigns targeted inincluded adults from all states and territories, with tentional tanning and sun-protection behaviors in some oversampling to improve precision. The readolescents, young adults, and the general populasponse rate for the adult survey was low (24%), but tion. These campaigns were supported by policy and limited by a short 2-day window for weekly interenvironmental changes in a range of settings, with viewing. The parents were recruited using quotas for organizations accountable for childrens safety rechilds age (0-2, 3-4, 5-6, and 7-11 years) and gender ceiving particular attention.9 to provide proxy reports of children aged 0 to 11 Few studies have examined the outcomes of years. In households with two or more children, specic interventions for Australian childrens sun when multiple quotas were open, the interviewer protection.11-15 Two specific initiatives for children randomly selected which child would be the interare viewed as successful. Kidskin implemented a view subject. Cancer Council Victoria Human comprehensive educational and environmental inResearch Ethics Committee approved the study. tervention in primary schools, which reduced childrens tanning and sun exposure during the Measures intervention.16,17 The SunSmart Schools program Demographic questions included parent educapromotes development of comprehensive suntion, area of residence, and the age and gender of the protection policies. Program evaluation showed child. The childs skin type was assessed by asking school policy implementation increased from 17% how the childs skin would react after exposure for to 76% in 4 years in one state18 and reached 80% 30 minutes to strong sunshine at the beginning of Australia-wide in 2005.13 Anecdotally, compliance summer with no protection at all. with sun-protection recommendations is supported The childs experience of sunburn on the weekby low pricing and new application options for end was assessed by asking whether the child got at sunscreen, availability of long-sleeved swimwear,9 all sunburnt on either Saturday or Sunday. Parents and shade provided in child-focused and other were then asked which body part or parts of their outdoor facilities.19 child got sunburn on the weekend, and on which In this context we expected Australian children to body part the burn was worst. Severity of the burn/ be well protected when outdoors in summer, and worst burn was assessed by asking parents if the parents to play an important role. Our study aimed to childs burn was red without being tender, red provide national data on childrens sun exposure and tender, or red, tender, and blistered.
d d d

940 Dobbinson et al

J AM ACAD DERMATOL

JUNE 2012

Subsequent questions determined if the redness was gone the next day and if the blisters were weeping. The childs exposure and sun-protection behaviors were assessed by separate questions on Saturdays and Sundays activities. Parents were asked if their child was outdoors on Sunday for longer than 15 minutes between (11 AM and 3 PM). They were asked about the amount of time the child spent outdoors, and the main activity and location during this period. Subsequent questions examined the childs use of clothing, a hat, sunglasses, and sunscreen during the main activity. Parents were told we need to ask you some questions in detail about what your child was wearing (on Sunday) to find out how much (his/her) skin was exposed to direct sunlight. Would you mind telling me what your child was wearing while he/she was (doing their activity outdoors)? Further questions determined the type of top worn (including sleeve length); whether a cap, hat, or sun visor was worn; and if the headwear had a wide brim or a flap covering the back of the neck. Sunscreen use was assessed by asking Was your child wearing a sunscreen between (11 AM and 3 PM) on Sunday? and, What was the sun-protection factor (SPF) of the sunscreen your child was using? The responses were reduced to dichotomous outcome variables to describe whether or not the child had sunburn; wore a hat, a three-quarter or long-sleeved top, three-quarter or long leg cover; used SPF 15 or higher sunscreen; stayed mainly under shade; or wore sunglasses. Time spent outdoors was assessed in minutes and analyzed as a continuous variable. Dichotomous summary outcome variables described the childs use (or not) of two or more, and 3 or more sun-protection behaviors during their outdoor activity. For two or more behaviors a value of 1 was assigned if the child wore either a hat or sunscreen, and complied with one other sun-protection behavior. The 3 or more summary variable was constructed similarly with two other sun-protection behaviors needed. This format avoided the context where clothing covering the legs and arms did not reflect sun-protection considerations given that the head was unprotected. The parents sunburn and sun-protection behaviors were described by an equivalent set of outcome variables. These weekend variables considered the activity data for the most recent day (Sunday) unless the respondent/child was only outdoors (or for sunburn outcomes only had sunburn) on Saturday. This enabled merging with the relevant meteorological records (temperature, cloud cover, wind speed, and UV index in the state/territory capital city) during 10 AM to 2 PM/11 AM to 3 PM on the specic date and

location of the childs activity and sunburn, and was used in multivariate analyses. Parents were asked their opinion on statements relating to protecting the child from the sun. Responses, recorded on a 5-point scale from strongly agree to strongly disagree, were recoded into dichotomous variables for multivariate analysis with positive sun-protection attitudes in a consistent direction. Data were also reduced for the independent variables used in the multivariate models. Childs age recorded as: younger than 2, 3 to 4, 5 to 6, or 7 to 11 years was reduced to 0 to 4 and 5 to 11 years for analysis. Childs skin type considered the categories of burn only, burn first and then tan, or tan only. Cases where the parent could not say what would happen to the childs skin (n = 90) or said nothing would happen (n = 9) were excluded from the analyses. Statistical analyses The sample characteristics, and the prevalence of childrens sun-protection behaviors and sunburn were described using weighted frequencies and means. Bivariate relationships with demographic variables were analyzed using x2 and t tests. Weights were standardized to the Australian population of children 0 to 11 years in 2003, and adjusted for minor variations by age, sex, state, and region.21 Potential predictors of child sun-protection outcomes were examined in 3 stages using multivariate logistic and linear regression models. First we examined potential predictors of specic behaviors in relation to demographic and meteorological factors (meteorological records for relevant dates) in a weighted analysis. Next, the independent relationships between childrens sun protection and parents attitudes and behaviors were examined using unweighted analyses adjusting for child and parent demographics and weather conditions. For each outcome variable we examined the relationship between parent behavior and child sun protection, first with parent behavior, alone, as the primary independent variable. Subsequently a full model was analyzed that included the independent variables for parent attitudes and behaviors. Variables were considered to have an independent effect on childrens sun-protection behaviors at P less than or equal to .05. Software (SPSS 14.0, SPSS Inc, Chicago, IL) was used for all statistical analyses.

RESULTS
Parents (N = 1140) provided proxy reports for one of their children. Of respondents, 58% were women; mean age was 37 years (SD = 6.9), 24% held a

J AM ACAD DERMATOL
VOLUME 66, NUMBER 6

Dobbinson et al 941

Table I. Prevalence of childrens sun protection and sunburn on summer weekends in 2003 to 2004 by sex and age, as reported by parent proxy
Boys (N = 585) Girls (N = 555) Age 0-4 y (N = 457) Age 5-11 y (N = 683) All children (N = 1140)

Sun exposure Outdoors, %* 76 69 65 78 73 Mean time outdoors, min 115 (SD = 71) 103 (SD = 67) 107 (SD = 70) 112 (SD = 70) 110 (SD = 70) Sunburnt on weekend, % 10 7 5 11 8 Sun protection when outdoors, % N = 430 N = 364 N = 293 N = 501 N = 794 Wore a hat 69 57 71 59 64 Used SPF $ 15 sunscreen 57 59 59 58 58 Wore three-quarter or long-sleeved top 20 15 22 15 18 Wore three-quarter or long leg cover 31 26 32 27 29 Stayed mostly under shade 31 32 43 25 32 Wore sunglasses 11 10 10 11 11 Used $ 2 sun protection 66 54 70 56 61 behaviors
Sun-protection prevalence is reported for children outdoors during peak ultraviolet times (10 am-2 pm/11 am-3 pm Daylight Saving Time) in their main activity on previous Saturday or Sunday. Prevalence estimates were weighted to 2003 population of Australian children aged 0-11 years.21 SPF, Sun-protection factor. *Prevalence figures for children outdoors exclude n = 46 (4.1%) parents reporting they could not say if their child was outdoors or not on previous Saturday or Sunday.

university degree, 64% lived in metropolitan areas, and 18% were from the highest and 14% the lowest socioeconomic quintile areas.22 The majority of parents (72%) had spent both weekend days with their child. Children surveyed by proxy were 51% boys and 49% girls; 60% of children were aged 5 to 11 years and 40% aged 0 to 4 years. Most parents and children had highly (29%) or moderately (50%) sunburn-sensitive skin. Sun protection during activities outdoors Most children (73%) were outdoors during peak UVR on either Saturday or Sunday on the previous weekend (Table I). They spent on average of 110 minutes outdoors, and were mostly engaged in activities in the backyard at home, or at a friends or relatives home (42%). Water-related activities at pools, rivers, and beaches (29%) were also common. Sunscreen (58%) and hats (64%) were the forms of sun protection most often used by children; clothing covering most of the arms and/or legs, sunglasses, or staying under shade were less common. Of children, 42% wore a protective hat including a hat/cap with a wide brim (33%) and/or a ap covering the neck (20%). The majority of children used two or more sun-protection behaviors (61%), whereas only 29% used 3 or more. Overall, 8% of children (n = 94) had sunburn on the weekend before interview. Few experienced severe burns, with 87% described as red without being tender, 13% red and tender, and 0.1% blistered.

Among these children the most common sunburn sites were the head and face (56%), the shoulders (27%), and the arms and hands (22%). Less commonly burned were the neck (11%), legs and feet (13%), and the chest, stomach, and back (12%). Of these children, most (47%) were burned while close to water, followed by those playing (15%) or engaging in active recreation (9%). The explanations given for the sunburn included that the sunscreen had worn off (26%), they had stayed in the sun too long (18%), or had forgotten to protect the burnt area (13%). Factors associated with childrens sun protection In multivariable analyses, weather conditions and gender, age, and skin type were strongly associated with the pattern and level of child sun protection (Table II). On average, girls spent less time outdoors than boys. However, except for shade use, boys were generally better protected when outdoors, with decreased odds of girls wearing a hat or using two or more sun-protection behaviors. The older children were less well protected than the younger children: those aged 5 to 11 years had decreased odds of wearing a hat or clothing covering their legs ( $ 3/4), or staying under shade. Children with skin that was not sensitive to sunburn had decreased odds of staying under shade when compared with children with sunburn-sensitive skin. Warmer temperatures predicted increased time spent outside and childrens increased use of hats, sunscreen, and shade,

942 Dobbinson et al

Table II. Factors associated with childrens sun protection on previous weekend, weighted analyses
Time outside Regression coefficient (SE) Hat OR (95% CI) Sunscreen (SPF $ 15) OR (95% CI) Three-quarter or long-sleeved top OR (95% CI) Three-quarter or long leg cover OR (95% CI) Mostly in shade OR (95% CI) $ 2 Sun-protection behaviors OR (95% CI) Sunburn OR (95% CI)

Sex Girls (reference: boys) Age, y 5-11 (reference: 0-4) Childs skin type Moderately sensitive (reference: highly sensitive) Not sensitive (reference: highly sensitive) Temperature 22-278C (reference: cold days \228C) $ 288C (reference: cold days \228C) Cloud cover Wind speed UV*

e13.4y (6.7) 3.8 (7.0) 2.6 (7.8) 16.0 (9.7) 30.6z (8.8) 39.7x (10.5) 0.02 (1.4) N/A N/A

0.6y (0.4-0.9) 0.4z (0.3-0.7) 0.8 (0.5-1.5) 0.8 (0.4-1.6) 1.8y (1.0-3.2) 1.0 (0.5-2.0)

0.8 (0.5-1.2) 0.6 (0.4-1.0) 0.7 (0.4-1.2) 0.6 (0.3-1.1) 2.3z (1.3-4.1) 3.4x (1.7-6.8)

1.0 (0.6-1.7) 0.6 (0.4-1.0) 1.1 (0.6-2.1) 0.8 (0.4-1.7) 0.3x (0.2-0.6) 0.5 (0.2-1.00) 1.05 (0.9-1.2) N/A N/A

1.1 (0.7-1.7) 0.6y (0.4-0.98) 1.0 (0.5-1.7) 0.9 (0.5-1.8) 0.2x (0.1-0.4) 0.1x (0.1-0.3) 1.05 (0.95-1.2) N/A N/A

1.7y (1.1-2.7) 0.5z (0.3-0.8) 0.8 (0.5-1.4) 0.3z (0.1-0.6) 2.7z (1.4-5.2) 2.8z (1.3-6.1) 1.07 (0.97-1.18) N/A N/A

0.6y (0.4-0.9) 0.4x (0.2-0.6) 0.6 (0.3-1.0) 0.4y (0.2-0.8)

0.2x (0.1-0.5) 2.2 (1.0-4.8) 1.9 (0.8-4.5) 0.7 (0.2-2.2) 12.4x (3.1-49.3) 5.5y (1.2-25.6) 1.09 (0.95-1.25) N/A 1.1 (0.98-1.27)

1.6 (0.9-2.8) 1.1 (0.6-2.3) 0.87z (0.79-0.96) 1.00 (0.96-1.03) N/A

0.9y (0.82-0.99) 0.85z (0.8-0.9) 1.0 (0.98-1.04) N/A N/A N/A

Results of multivariate logistic and linear regression models among 371 children outdoors in metropolitan areas on previous weekend. Adjusted OR with 95% CI; summarizing odds of performing specified behaviors compared with referent group (referenceeOR = 1.0) adjusting for covariates. Linear regression model R2 = 0.067, P \ .001, and model constant regression coefficient = 95.2 minutes (SE = 15.5). CI, Confidence interval; N/A, not applicable; OR, odds ratios; SPF, sun protection factor; UV, ultraviolet. *Results for independent variable UV were from second logistic regression model of sunburn predictors (N = 524) considering sunburn among all children whether indoors or outdoors, excluding children residing in nonmetropolitan areas where UV data were not available. Several sunburn models were examined given colinearity between temperature and UV radiation values and to compare potential predictors of sunburn considering protection used by children when outdoors and separately considering protection used when either outdoors or protected indoors. y P \.05. z P \.01. x P \.001.

J AM ACAD DERMATOL

JUNE 2012

J AM ACAD DERMATOL
VOLUME 66, NUMBER 6

Dobbinson et al 943

Table III. Parents opinions on protecting their children from sun, N = 1140, percentages
Strongly agreed Mildly agreed Neither agreed nor disagreed Mildly disagreed Strongly disagreed Could not say

Provided opinions on statements: Sunburn during childhood is dangerous Children enjoy summer more when protected from sun I try to set a good example with my own sun protection My children look attractive with a suntan Protecting my child from the sun takes too much time, money, and trouble I have great difficulty getting my child to wear sunscreen I have great difficulty getting my child to wear a hat Children need to get some direct sunlight on their skin to be healthy

88 55 47 4 2 5 9 20

8 26 30 15 3 10 12 40

\1 7 4 7 \1 1 3 3

\1 8 13 23 7 18 21 16

2 3 6 50 88 64 55 21

\1 1 1 1 \1 2 1 \1

but reduced use of covering clothes. On days with thicker cloud cover the odds of children wearing a hat and using sunscreen decreased, as did use of two or more sun-protection behaviors. Adjusting for other covariates, gender and the ambient temperature were strongly associated with childrens sunburn when outdoors (Table II, column 8). Girls had lower odds of being sunburnt than boys, and children had higher odds of being sunburnt on warmer days above 228C. A separate model adjusting for ambient UVR levels predicted sunburn among all children on the weekend, revealing factors strongly associated with sunburn and considering the protection afforded to children by staying indoors during peak UVR times. In this model age was also related to sunburn, with increased odds of sunburn for older children aged 5 to 11 years (odds ratio [OR] 2.4; 95% confidence interval [CI] 1.1-5.4) compared with children aged 0 to 4 years. This suggests younger children were more likely to be protected by staying indoors. Parental attitudes The majority of parents held beliefs supportive of their childs sun protection (Table III). Almost all parents (88%) indicated they strongly believed that sunburn during childhood is dangerous, and that it was not too much trouble to protect their child from the sun. However, up to 43% of parents agreed with at least one of the less supportive statements: namely protecting their child was too much trouble, suntans were attractive, or that their child was resistant to wearing hats or sunscreen. Parental factors associated with childrens behaviors Table IV shows that with the exception of wearing long-sleeved tops, the parents use of

specific sun-protection behaviors was associated with the childs use of the same protective behavior. The odds of a child using the specific behavior if the parent had used that behavior on the weekend (compared with parents not using that behavior) ranged from 3 times greater for wearing a hat (OR 3.1; 95% CI 1.6-6.2) to 12 times greater for using sunscreen (OR 12.6; 95% CI 5.2-30.4). Similarly the odds of a child using at least two sun-protection behaviors were higher if the parent had also used two or more sun-protection behaviors. Few parental attitudes were independently associated with childrens sun-protection behaviors after adjusting for parents use of the same sun-protection behavior (full model Table IV). Childrens hat and sunscreen use were predicted by attitudes about the childs resistance to using these behaviors. Childrens shade use was related to a broader belief about the perceived benefits of protecting their child from the sun. Perceived resistance to hat use was also associated with increased odds of children using at least two sun-protection behaviors.

DISCUSSION
This study found high levels of sun protection among Australian children aged 0 to 11 years as part of a larger concurrent study of sun protection in Australian adults and adolescents.20 These directly comparable data reveal that children were better protected during their main activity on summer weekends compared with adults and adolescents. For example, 64% of children, 48% of adults, and 38% of adolescents wore hats; similarly, 58%, 33%, and 37%, respectively, used SPF 15 or higher sunscreen. Consequently fewer children (8%) were burned on these weekends compared with adults (18%) and adolescents (25%).

944 Dobbinson et al

J AM ACAD DERMATOL

JUNE 2012

Table IV. Parental factors (behaviors and attitudes) independently associated with childs use of specific behaviors outdoors on previous summer weekend
Initial model with parent behaviors and key covariates N = 229 Dependent variables Independent variables OR (95% CI) Full model with parent behaviors and parent attitudes N = 225 OR (95% CI)

Childs sun protection predicted SPF $ 15 sunscreen

Three-quarter or long leg cover Stayed mostly in shade

Parent modeled same behavior/and significant parent attitudes Parent used sunscreen (yes) Have difficulty getting child to wear sunscreen (disagreed) Parent wore longer leg cover (yes) Parent used shade (yes) Children enjoy summer more when protected (agreed) Parent wore sunglasses (yes) Parent wore a hat (yes) Have difficulty getting child to wear a hat (disagreed) Parent wore three-quarter or long-sleeved top (yes) Parent used $ 2 Sun-protection behaviors (yes) Have difficulty getting child to wear a hat (disagreed)

12.6* e 10.3* 9.6* e 4.0y 3.1x e 2.0 5.7 e

(5.2-30.4) e (4.4-24.0) (4.4-20.8) e (1.4-11.6) (1.6-6.2) e (0.9-4.7) (2.8-11.9) e

16.3* 3.8y 9.4* 10.7* 4.7y

z

(5.9-45.2) (1.3-11.0) (4.0-22.2) (4.7-24.4) (1.4-15.7)

Sunglasses Any hat

2.7x 4.3x 1.9 6.4* 3.3y

(1.3-5.8) (1.7-11.1) (0.8-4.7) (2.8-14.3) (1.2-9.0)

Three-quarter or long-sleeved top $ 2 Sun-protection behaviors

Results of logistic regression models of childs use of specific sun-protection behaviors adjusting for covariates: parents gender, education, and socioeconomic status; weather conditions (temperature, cloud, and wind speed for analyses of hat use); childs age, gender, and skin type. Note in these models dichotomous variables were used to describe parents attitudes, with responses on 5-point scale recoded according to statement viewpoint. For positive statement parents who agreed mildly or strongly were coded 1, whereas for negative statement parents who disagreed mildly or strongly were coded 1, versus referent response categories coded 0. OR for parent attitudes were reported only if significantly associated with childs sun-protection outcomes. OR for independent variables used referent categories (OR = 1.0) representing: (1) for parents behaviorseparents responding they did not perform given behavior; and (2) for parent attitudeseparents responding with less positive opinion toward sun-protection statements. N = complete data for children spending most of the time under parents supervision while outdoors in metropolitan area on previous weekend. CI, Confidence interval; OR, odds ratios; SPF, sun-protection factor. *P \.001. y P \.05. z Model did not converge. x P \.01.

Some population-based studies indicated Australian childrens sun protection to be considerably higher than in other countries. One study of US children in 1999 found only 6.6% always used a hat, 2.5% used shade, and 33% used SPF 15 or higher sunscreen in summer sun.23 A few other studies showed relatively good sun-protection levels among children in the United States,24 New Zealand,25 and Canada,26 but comparability is limited by different timing and methods.27,28 Most previous research on childrens sun protection in Australia has been of local and regional sample frames, in schools,29,30 childcare centers,12,31,32 and beaches,33-35 with limited comparability with our

measures. Our baseline findings are consistent with positive impacts of the skin cancer prevention strategies implemented for children and parents.11,13,36,37 Despite relatively high protection levels, 8% of children had sunburn, indicating a need for more thorough sun protection, particularly on the head, shoulders, and arms. Although campaigns have promoted use of multiple sun-protection behaviors,5,9 only 29% of children reportedly used 3 or more sun-protection behaviors. Male gender and higher atmospheric temperature were the strongest factors associated with childrens sun-protection and sunburn outcomes. Participation in water-related activities, and forgetting to protect or reapply

J AM ACAD DERMATOL
VOLUME 66, NUMBER 6

Dobbinson et al 945

sunscreen, were common explanations for the childrens sunburn. Previously among adolescents and adults some sun-protection behaviors (using shade and spending less time outdoors) were associated with less sunburn.20 Low sunburn numbers prohibited our analysis of the full range of explanatory variables in predicting sunburn. Nonetheless, the probable explanation for boys having more sunburn than girls may be that boys spent considerably more time outdoors and less commonly used shade. Another key nding was that, except for wearing long-sleeved tops, parental behaviors were strongly associated with childrens sun protection. The parents use of shade, sunscreen, and longer leg cover were highly predictive, with the adjusted odds of children practicing these specific behaviors 9 to 12 times greater when the parent had used the behavior. In contrast few parental attitudes were independently associated with childrens sun protection. Theoretical models recognize role modeling as an important cue for (health) behaviors in general.38,39 Moreover, parents often have some responsibility for and control over childrens sun-protection behaviors.40 Few studies have closely examined the relationships between parental behaviors and sun protection in children23,25,33 or adolescent.41 Three studies showed strong relationships between the parents specific sun-protection behaviors and the childs behavior.25,33,41 Another study showed childrens behaviors were more strongly associated with the parents tan preferences but few had modeled good sun protection.23 These ndings suggest campaigns targeting adults (and hence parents) may have a wider benet in improving childrens sun protection. The intensity of exposure to these campaigns may be a key factor in behavior change.37 Since the preliminary study findings, a national campaign was launched (http:// www.skincancer.gov.au/) and state-based campaigns increased to address the broad gaps in compliance past childhood. Involving parents and caregivers in interventions targeting young children may enhance outcomes.12,29,42-46 Clinicians might counsel parents about providing good sun-protection role models for their children. Further evidence of the efficacy of these and other strategies involving parent role models is needed.47 We acknowledge that our measures did not account for parental enforcement of SunSmart behaviors, which likely correlates with parental example. Limited inference about the effects of public education on skin cancer in Australia can be made, given the cross-sectional study design with no

baseline or control group. No directly comparable census data on the demographics of parents of children aged 0 to 11 years are available to fully determine the representativeness of our sample,22,48 but educational attainment is similar to those aged 15 to 64 years.49 The sample was limited to households with a landline telephone and initial response rates were low.20 One child was systematically selected as the survey subject per household, and results may overrepresent households with smaller families. Clustering effects were not examined, as the presence of other children in the household was not recorded. Parenting styles and advice given to their children about sun protection might partly explain the parental role model effects and future research might examine this. In line with campaign messages over the years, our study focused on use of multiple sun-protection behaviors. An index representing the proportion of the body protected might also be used in future research to evaluate trends in childrens sun-protection compliance. A study strength was that parental sun-protection role modeling was assessed using measures for the same time period for each parent-child pair. We relied on parent proxy reports, which are considered valid30,50 and have been widely used.23-25 Social desirability bias is possible, as are other threats to validity. Childrens activity levels may influence the accuracy of parents reports of childrens sun protection.15 The sun-protection measures may not fully determine total exposure if, for example, shirtsleeves are rolled up, hats taken off frequently, or a cap worn backwards. Our nding that Australian children were generally well protected from the summer sun is encouraging given considerable investment in public education on skin cancer in past decades. A small but not insignicant percentage of children each weekend were sunburnt, highlighting the need for further efforts. The strong association between parental and childrens sun-protection behaviors suggests that targeting parental behaviors is potentially an important means of improving childrens sun protection. Climate and UV data were provided by the Australian Radiation Protection and Nuclear Safety Agency, the Bureau of Meteorology, and University of Tasmania. We thank the national research group who contributed to the survey development.
REFERENCES 1. Lucas R, McMichael T, Smith W, Armstrong B. Solar ultraviolet radiation: global burden of disease from solar ultraviolet radiation. Environmental burden of disease series, No. 13. Geneva: World Health Organization, Public Health and the Environment; 2006.

946 Dobbinson et al

J AM ACAD DERMATOL

JUNE 2012

2. Whiteman DC, Whiteman CA, Green AC. Childhood sun exposure as a risk factor for melanoma: a systematic review of epidemiologic studies. Cancer Causes Control 2001;12:69-82. 3. Armstrong BK. How sun exposure causes skin cancer: an epidemiological perspective. In: Hill D, Elwood JM, English DR, editors. Prevention of skin cancer. The Netherlands: Kluwer Academic Publishers; 2004. pp. 89-116. 4. Staples M, Elwood M, Burton R, Williams J, Marks R, Giles G. Non-melanoma skin cancer in Australia: the 2002 national survey and trends since 1985. Med J Aust 2006;184:6-10. 5. McCarthy WH. The Australian experience in sun protection and screening for melanoma. J Surg Oncol 2004;86:236-45. 6. Edlich RF, Winters KL, Cox MJ, Becker DG, Horowitz JH, Nichter LS, et al. National health strategies to reduce sun exposure in Australia and the United States. J Long Term Eff Med Implants 2004;14:215-24. 7. Giles GG, Armstrong BK, Burton RC, Staples MP, Thursfield VJ. Has mortality from melanoma stopped rising in Australia? Analysis of trends between 1931 and 1994. BMJ 1996;312: 1121-5. 8. Whiteman DC, Bray CA, Siskind V, Green AC, Hole DJ, Mackie RM. Changes in the incidence of cutaneous melanoma in the west of Scotland and Queensland, Australia: hope for health promotion? Eur J Cancer Prev 2008;17:243-50. 9. Montague M, Borland R, Sinclair C. Slip! Slop! Slap! and SunSmart, 1980-2000; skin cancer control and 20 years of population-based campaigning. Health Educ Behav 2001;28: 290-305. 10. Smith BJ, Ferguson C, McKenzie J, Bauman A, Vita P. Impacts from repeated mass media campaigns to promote sun protection in Australia. Health Promot Int 2002;17:51-60. 11. Giles-Corti B, English DR, Costa C, Milne E, Cross D, Johnston R. Creating SunSmart schools. Health Educ Res 2004;19:98-109. 12. Harrison SL, Buettner PG, Maclennan R. The North Queensland "Sun-Safe Clothing" study: design and baseline results of a randomized trial to determine the effectiveness of sun-protective clothing in preventing melanocytic nevi. Am J Epidemiol 2005;161:536-45. 13. Jones SB, Beckmann K, Rayner J. Australian primary schools sun protection policy and practice: evaluating the impact of the national SunSmart schools program. Health Promot J Austr 2008;19:86-90. 14. Schofield MJ, Edwards K, Pearce R. Effectiveness of two strategies for dissemination of sun-protection policy in New South Wales primary and secondary schools. Aust N Z J Public Health 1997;21:743-50. 15. Girgis A, Sanson-Fisher RW, Tripodi DA, Golding T. Evaluation of interventions to improve solar protection in primary schools. Health Educ Q 1993;20:275-87. 16. Milne ER, English DR, Johnston R, Cross D, Borland R, Giles-Corti B, et al. Reduced sun exposure and tanning in children after 2 years of a school-based intervention (Australia). Cancer Causes Control 2001;12:387-93. 17. Milne E, Jacoby P, Giles-Corti B, Cross D, Johnston R, English DR. The impact of the kidskin sun protection intervention on summer suntan and reported sun exposure: was it sustained? Prev Med 2006;42:14-20. 18. Dobbinson SJ, Peipers AM, Borland R, Nolan KM. Are Victorian primary schools SunSmart? The ongoing development of the SunSmart schools program. Health Promot J Aust 2000;10: 43-50. 19. Gartland D, Dobbinson S. A survey of the sun protection environment at swimming pools in Victoria: 2000/2001. SunSmart evaluation studies No. 7. Melbourne: Cancer Council Victoria; 2004.

20. Dobbinson S, Wakefield M, Hill D, Girgis A, Aitken JF, Beckmann K, et al. Prevalence and determinants of Australian adolescents and adults weekend sun protection and sunburn, summer 2003-2004. J Am Acad Dermatol 2008;59: 602-14. 21. Australian Bureau of Statistics. Population by age and sex, Australia 2003. Electronic delivery: Australian Bureau of Statistics; 2004. Available at: http://www.abs.gov.au/AUSSTATS/ abs@.nsf/Lookup/3235.0.55.001Main+Features1Jun%202003? OpenDocument. Accessed August 5, 2011. 22. Australian Bureau of Statistics. Census of population and housing: socio-economic indexes for areas (SEIFA). Canberra: Australian Bureau of Statistics, Australian Government Publishing Service of Australia; 2004. 23. ORiordan DL, Geller AC, Brooks DR, Zhang Z, Miller DR. Sunburn reduction through parental role modeling and sunscreen vigilance. J Pediatr 2003;142:67-72. 24. Robinson JK, Rigel DS, Amonette RA. Sun-protection behaviors used by adults for their childreneUnited States, 1997. MMWR Recomm Rep 1998;47:480-2. 25. McGee R, Williams S, Glasgow H. Sunburn and sun protection among young children. J Paediatr Child Health 1997;33: 234-7. 26. Lovato CY, Shoveller JA, Peters L, Rivers JK. Canadian national survey on sun exposure and protective behaviors: parents reports on children. Cancer Prev Control 1998;2:123-8. 27. Glanz K, Yaroch AL, Dancel M, Saraiya M, Crane LA, Buller DB, et al. Measures of sun exposure and sun protection practices for behavioral and epidemiologic research. Arch Dermatol 2008;144:217-22. 28. Dobbinson S, Hill DJ. Patterns and causes of sun exposing and sun protecting behavior. In: Hill D, Elwood JM, English DR, editors. Prevention of skin cancer. The Netherlands: Kluwer Academic Publishers; 2004. pp. 211-40. 29. Milne E, English DR, Johnston R, Cross D, Borland R, Costa C, et al. Improved sun protection behavior in children after two years of the Kidskin intervention. Aust N Z J Public Health 2000;24:481-7. 30. Dixon H, Borland R, Hill D. Sun protection and sunburn in primary school children: the influence of age, gender and coloring. Prev Med 1999;28:119-30. 31. Harrison SL, Saunders V, Nowak M. Baseline survey of sun-protection knowledge, practices and policy in early childhood settings in Queensland, Australia. Health Educ Res 2007;22:261-71. 32. Stanton WR, Saleheen HN, ORiordan D, Roy CR. Environmental conditions and variation in levels of sun exposure among children in child care. Int J Behav Med 2003;10:285-98. 33. Bennetts K, Borland R, Swerissen H. Sun protection behavior of children and their parents at the beach. Psychol Health 1991;5: 279-87. 34. Pincus MW, Rollings PK, Craft AB, Green A. Sunscreen use on Queensland beaches. Australas J Dermatol 1991;32:21-5. 35. Foot G, Girgis A, Boyle CA, Sanson-Fisher RW. Solar protection behaviors: a study of beachgoers. Aust J Public Health 1993; 17:209-14. 36. Dobbinson S, Peipers A, Reading D, Sinclair C. A national approach to skin cancer prevention: the national SunSmart schools program. Med J Aust 1998;169:513-4. 37. Dobbinson SJ, Wakefield MA, Jamsen KM, Herd NL, Spittal MJ, Lipscomb JE, et al. Weekend sun protection and sunburn in Australia: trends (1987-2002) and association with SunSmart television advertising. Am J Prev Med 2008;34:94-101. 38. Bandura A. Social foundations of thought and action: social cognitive theory. Englewood Cliffs (NJ): Prentice-Hall; 1986.

J AM ACAD DERMATOL
VOLUME 66, NUMBER 6

Dobbinson et al 947

39. Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the health belief model. Health Educ Q 1988;15:175-83. 40. Lowe JB, McDermott LJ, Stanton WR, Clavarino A, Balanda KP, McWhirter B. Behavior of caregivers to protect their infants from exposure to the sun in Queensland, Australia. Health Educ Res 2002;17:405-14. 41. Cokkinides VE, Weinstock MA, Cardinez CJ, OConnel MA. Sun-safe practices in US youth and their parents: role of caregiver on youth sunscreen use. Am J Prev Med 2004;26:147-51. 42. Glanz K, Carbone E, Song V. Formative research for developing targeted skin cancer prevention programs for children in multiethnic Hawaii. Health Educ Res 1999;14:155-66. 43. Grant-Petersson J, Dietrich AJ, Sox CH, Winchell CW, Stevens MM. Promoting sun protection in elementary schools and child care settings: the SunSafe project. J Sch Health 1999;69:100-6. 44. Mayer JA, Lewis EC, Eckhardt L, Slymen DJ, Belch GE, Elder JP, et al. Promoting sun safety among zoo visitors. Prev Med 2001; 33:162-9.

45. Mayer JA, Slymen DJ, Eckhardt L, Johnston MR, Elder JP, Sallis JF, et al. Reducing ultraviolet radiation exposure in children. Prev Med 1997;26:516-22. 46. McDermott L, Lowe J. Using popular childrens characters to help parents protect their young children from the sun. Health Promot J Aust 2003;14:66-9. 47. Saraiya M, Glanz K, Briss P, Nichols P, White C, Das D, et al. Interventions to prevent skin cancer by reducing exposure to ultraviolet radiation: a systematic review. Am J Prev Med 2004; 27:422-66. 48. Trewin D. Australian social trends 2004. Canberra: Australian Bureau of Statistics; 2004. 49. Australian Bureau of Statistics. Education and work: Australia. Canberra: Australian Bureau of Statistics; 2004. 50. Mayer JA, Sallis JF, Eckhardt L, Creech L, Johnston MR, Elder JP, et al. Assessing childrens ultraviolet radiation exposure: the use of parental recall via telephone interviews. Am J Public Health 1997;87:1046-9.