Seven-year trends in sun protection and sunburn among Australian adolescents and adults
Correspondence to: Suzanne Dobbinson, Centre for Behavioural Research in Cancer, Cancer Council Victoria, 1 Rathdowne Street, Carlton, Vic 3053; e-mail: Suzanne.Dobbinson@cancervic.org.au
Aims : To examine the change in sun protective behaviours and sunburn of Australians over a seven-year period, in the context of sustained skin cancer prevention campaigns and programs.
Methods : Weekly cross-sectional telephone interviews of Australians were conducted throughout summer in 2010/11 for comparison with 2003/04 and 2006/07. In 2010/11, n=1,367 adolescents (12–17 years) and n=5,412 adults (18–69 years) were interviewed about their sun-related attitudes, weekend sun protection and sunburn. Multivariate analyses adjusted for key demographics, temperature, cloud, wind and ultraviolet radiation (UVR) to assess change in outcomes over time.
Results : There were consistent improvements in adolescents’ and adults’ attitudes, intentional tanning and incidence of sunburn over time. Behavioural changes were variable. Adults spent less time outdoors during peak UVR compared to past surveys, while adolescents were less likely to be outdoors compared with 2006/07. Sunscreen use and wearing of long sleeves increased among adults, but hat wearing decreased for both age groups, as did leg cover by adolescents since 2003/04. There has been a sustained decrease in weekend sunburn among adolescents and adults.
Conclusions : The findings suggest improvements in skin cancer prevention attitudes of Australians over time. Australians’ compliance with sun protection during summer has improved in some areas, but is still far from ideal. The sustained decrease in weekend sunburn among adolescents and adults is encouraging, but further improvements are required.
Australia has the unenviable claim of having one of the highest skin cancer rates in the world.1,2 Treatment of non-melanocytic skin cancers poses a significant public health burden, resulting in more hospital admissions than any other cancer type.3 Skin cancers caused 1,897 deaths in 2010,4 with the majority attributed to melanoma.5 Moreover, malignant melanoma can occur at a younger age than many other cancers,6 and is the most common cancer in Australians aged 12–24 years.7
Raising public awareness about skin cancer, the risks of exposure to ultraviolet radiation, and the need for sun protection and early detection of skin cancers has been an essential start to abating this epidemic.8–10 Moreover, reductions in skin cancer rates have been seen among younger cohorts in line with the history of public education implemented in Australia.1,11,12 State health departments and cancer councils in most States and Territories have broadcast summer media campaigns targeting prevention messages, and have advocated for sun protective policies and environmental change in a range of organisational settings across the community. Although comprehensive and community-wide strategies for skin cancer prevention have been implemented over several decades, budgets and delivery of media campaigns have been variable. The intensity and reach of the media campaigns declined in Victoria in the late 1990s and early 2000s,13 with at least three summers with no paid SunSmart campaigns since 1996. Similarly, other States had varied levels of exposure to campaigns from year to year and variable funding support,14,15 with no paid campaigns in Tasmania, the Australian Capital Territory (ACT) and the Northern Territory (NT).
NSW Health broadcast paid campaign advertisements from 1991 to 2000 targeting adolescents and, later, parents of young children.16,17 Later campaigns were funded by the Cancer Institute NSW and were youth focused. Cancer Council Western Australia broadcast paid campaigns targeting youth in most summers since 1994. In contrast, South Australia (SA) relied on free community service announcements (CSAs), with their first paid campaign launched in 2000.18 Queensland Health, the Queensland Cancer Fund, and Sun-Corp developed sun protection ads relying on CSAs for broadcast up until 2008 when the State Government launched a paid TV campaign.19
In 2003/04 a national survey was conducted to provide a more detailed assessment of Australians’ compliance with sun protection messages. The baseline study revealed that children's sun protection was relatively good, while adolescents’ and adults’ levels of sun protection compliance were less than satisfactory.20,21 Following this study funded by the Australian government, the first national skin cancer awareness campaign was broadcast on television in all States and Territories over spring and summer months (except in spring 2007) from November 2006 to January 2010. The campaign advertisement featured a surgeon talking about the risks of skin cancer and melanoma and what can be done to prevent it, featuring scenes from an operation to remove a melanoma on a young woman.22 In 2006/07 four states (NSW, SA, WA and Qld) added paid TV campaigns of their own to maximise the likely impact. State broadcast campaigns continued with broadcast of the ‘Dark Side of Tanning’ advertisement in all states except Tasmania and NT in 2010/11.23
The current study examines the trends in the population's sun-related attitudes, behaviours and sunburn incidence in the context of increasing and relatively sustained campaign activity during a seven-year period.
In three summers (2003/04, 2006/07 and 2010/11) a national cross-sectional survey was conducted of the sun-related attitudes, weekend sun protection behaviour, and incidence of sunburn of Australian adolescents (12–17 years) and adults (18–69 years). Respondents were interviewed weekly from late November to the end of January on Mondays and Tuesdays, with a break over the public holiday period, to maximise the accuracy of self-reported behaviour.24
Sample quotas stratified by age, gender, and area of residence (metropolitan/regional location) ensured that the achieved sample was representative of the Australian population. Oversampling of respondents in smaller States and Territories required population weights stratified by age and gender to retain proportional prevalence figures.25–27 Due to severe flooding in Brisbane and nearby areas in south-east Queensland,28 quotas for Queensland were not fulfilled and interviews halted after week five of the survey. However, an analogous exclusion of respondents from these weeks was shown not to have a bearing on the results of trends over time for 2003–2006. An experienced market research company conducted the interviews. One person was recruited from each household, with households contacted via random selection of active telephone numbers drawn from the Australian Communications and Media Authority (ACMA) Numbering Plan. This sample frame was advised to provide 79% coverage of households in 2010. The response rate in 2003/04 was 24%, 16% in 2006/07, and 21% in 2010/11. Data collection being restricted to two days without possibility of follow-up in later survey weeks to avoid priming bias contributes substantially to the low recruitment rates.
Respondents took 15 minutes on average to complete a telephone interview which assessed respondents’ tanning attitudes and skin cancer related beliefs, their incidence of sunburn and sun protective behaviours on the previous weekend. The face validity and internal validity of these items has been confirmed.29,30 Adults and adolescents were asked identical questions, except for measures of employment and education.
A number of key interview questions have been previously reported.13,21,29 Dichotomous outcome variables were produced from respondents’ detailed reports of their tanning attitudes and behaviours. Outcome variables were produced for weekend behaviours to indicate whether or not during their main activity outdoors (on Sunday primarily, except if outdoors on Saturday only) they wore a hat (hat, cap or wide-brimmed hat), covering clothing (at least three-quarter length top/shirt/dress, or three-quarter length leg-cover), sunscreen (SPF 15+), and sunglasses, and/or stayed primarily under shade. The amount time spent outdoors in peak UVR (in minutes) was analysed as a continuous variable.
A key outcome question was “The next questions are about sunburn. By sunburn we mean any amount of reddening of the skin after being in the sun. Did you get at all sunburnt yesterday? What about on Saturday?” Occurrence of sunburn on either day was coded ‘sunburnt’ for the purpose of incidence figures.
Frequencies and means were used to describe the prevalence and determinants of adolescents’ and adults’ tanning attitudes, skin cancer related beliefs, weekend sun protection and sunburn. Multivariate analyses (logistic regression and multiple regression) examined trends for these outcomes, controlling for a number of demographic and weather-related covariates where appropriate. We were unable to model the proportion of respondents outdoors on the weekend adjusting for weather, as it was not possible to assign weather data preferentially for just one day of the weekend to these respondents. All these models treated survey year as the main independent variable and provided adjusted odds of adolescents’ and adults’ sun protective behaviours and sunburn in 2010/11. Odds ratios and 95% confidence intervals (or regression coefficients and standard errors for continuous variables) were used to describe the strength of the association between being surveyed in 2010/11 relative to 2003/04 and relative to 2006/07 with the outcome variable, when controlling for the other covariates.
Trends in sunburn were examined in a multivariate analysis for the full sample of metropolitan residents. These analyses included respondents who were indoors during peak UVR times, given staying indoors is an effective means of preventing sunburn. A separate model also examined only respondents who were outdoors in a metropolitan area. As with the sun protection behaviours, this focus on sunburn within the metropolitan area was due to the limited weather and UVR data available by region. Sunburnt respondents who were outdoors at times other than peak UVR (not a focus of the survey) were excluded from the models given the unknown location of their activities. For both models, the effects of temperature and cloud-cover on sunburn were modelled separately to the effects of UVR levels, due to co-linearity between these variables.
The weather covariates were obtained from the official climate records for the cities and dates relating to respondents’ weekend activities. Sun protective behaviours and sunburn models were adjusted by 3.00 pm temperature (degrees Celsius) and cloud cover (oktas). Wind speed (km/hr) was additionally adjusted for in models predicting hat wearing. Separate models predicting sunburn adjusting for UVR, the known physical cause of sunburn, used the mean UV Index between 10.00 am and 2.00 pm (Standard Time) on the Sunday/Saturday the respondent was outdoors.
PASW Statistics 18 for Windows was used for all statistical analyses. Population weights were applied to all the data analyses.
Table 1 describes the characteristics of the 5,412 adults and 1,367 adolescents who were interviewed. As with the previous surveys in 2003/04 and 2006/07, the weighted distribution of 2010/11 respondents across age group and gender was proportional to the population of Australians aged 12- 69 years in 2010.27 The majority of respondents had moderately sensitive skin that would ‘burn then tan’ following exposure to sunlight for more than 30 minutes at the start of summer.
Table 1. 2010–11 survey sample characteristics.
|Age|| || |
| Gender || || |
| Skin type || || |
|Not sensitive (tan only)||31||25|
|Moderately sensitive (burn then tan)||42||46|
|Highly sensitive (burn only)||24||27|
|Nothing would happen or can't say||3||2|
Adolescents’ beliefs and social norms, compared with adults, more commonly supported intentional tanning. Forty-five per cent of adolescents reported that they like to get a suntan compared with only 27% of adults (Table 2). Thirty-six per cent of adolescents and 42% of adults reported they believed a suntan was beneficial to appearance, while 12% and 10%, respectively, believed a suntan had a health benefit. The majority (66%) of adolescents perceived their friends as having a pro-tanning attitude, while 36% of adults had this perception. Despite adolescents’ strong desire for a suntan, approximately one-fifth reported they attempted a suntan in the current season. Similarly, the prevalence of recent tanning attempts among adults was rare (9%).
Table 2. Trends in adolescents’ and adults’ intentional tanning attitudes and behaviour.
|Prefer a suntan||60||51||45||0.52***|
|Believe suntanned person looks more healthy||41||43||36||0.86|
|Believe suntanned person is more healthy||19||17||12||0.59***|
|Perceive friends think a suntan is a good thing||77||71||66||0.56***|
|Attempted a suntan this season||32||22||22||0.57***|
The majority of adolescents and adults were outdoors for more than 15 minutes in peak UVR times on summer weekends, spending on average 112 and 111 minutes outdoors, respectively. Table 3 shows there was a low prevalence of the use of specific sun protective behaviours outdoors. Among adolescents, use of sunscreen (37%) and clothing with longer leg-cover (28%) were the most common forms of protection. The most common sun protective behaviours among adults were wearing sunglasses (57%), hats (45%) and clothing with longer leg-cover (44%). The prevalence of sunscreen use by adults was similar to that of adolescents. Overall, adults displayed more frequent use of sun protection than did adolescents in 2010/11.
Table 3. Trends in adolescents’ and adults’ weekend sun-protection behaviour and sunburn.
|Respondents outdoors >15 minutesb||80||82||77||0.81|
|Time spent outdoors in minutesb (B-value)||110||111||112||3.28|
|Sunscreen used (> SPF 15+)b||37||37||37||1.05|
|¾ or long sleeved top wornb||11||9||11||1.34|
|¾ or long leg cover wornb||37||30||28||0.59**|
|Stayed mostly in the shadeb||19||20||21||1.29|
|Two or more sun protective behavioursc||29||22||24||0.70*|
|Sunburn (almost all respondents)d||25||24||21||0.62**|
Twenty-one per cent of adolescents and 13% of adults reported being sunburnt on the weekend (Table 3). Most sunburnt respondents typically reported their sunburn as “red but not tender” (adolescents: 62% c.f. adults: 71%). Fewer respondents reported “red and tender” (adolescents 34% cf. 27%), or “red, tender and blistered” sunburn (adolescents: 4% c.f. adults: 2%). “Sunscreen wearing off” (23%) and “stayed in the sun too long” (28%) were the most frequently reported reason for sunburn among adolescents, while among adults these were “stayed in the sun too long” (22%) and “forgot to protect” (21%).
Changes over time
A pattern of improvement was evident in adolescents’ and adults’ tan preference since 2003/04 and within the recent period as compared with 2006/07 (Table 2). Preference for a suntan decreased among both adolescents and adults for both periods examined. Perceived attractiveness of a tan decreased within the recent period, among both adolescents and adults, as did misconception about the health benefits of tanning. Perceived tanning norms of both groups improved over time, although for adults there was only improvement relative to 2003/04. The adjusted odds for both groups attempting a suntan decreased relative to the baseline 2003/04 survey, as well as for adults relative to 2006/07.
There was some evidence respondents’ sun exposure decreased in 2010/11 compared with the 2003/04 and 2006/07 surveys. Table 3 indicates that when controlling for demographic factors, there was a decrease in the odds of adolescents reporting they were outdoors during peak UVR times relative to 2006/07. Among adults, the proportion outdoors since 2003/04 decreased with no change observed since 2006/07. Moreover adults’ time spent outdoors in minutes decreased over both time periods.
Respondents’ compliance with sun protective behaviours over time was variable with improvements in 2010/11 noted only among adults and not for adolescents. Adolescents showed no improvement in sun protection compliance since the more recent period (2006/07). Relative to 2003/04, adolescents’ use of hats, clothing with longer leg-cover and use of at least two sun protective behaviours decreased in 2010/11. For adults, use of hats decreased since 2006/07, but use of long-sleeved tops increased. Adults’ sunscreen (SPF 15+) use remained unchanged since 2006/07, but improved over the longer period (2003/04). Use of two or more sun protective behaviours did not change over either time period for adults.
Table 3 shows the adjusted odds ratio for adolescents’ sunburn (controlling for UVR) decreased since 2003/04 (OR=0.62; 95% CI=0.43–0.89) and 2006/07 (OR=0.58; 95% CI=0.40–0.84). Adults showed a decrease in odds of sunburn relative to 2003/04 (OR=0.49; 95% CI=0.41–0.59), and 2006/07 (OR=0.72; 95% CI=0.59–0.87). Similar results were found, in terms of a decreased odds ratio for adolescents and adults, in other sunburn models that substituted UVR with temperature and cloud cover, and considered only respondents outdoors.
The findings from three national surveys during summers between 2003/04 and 2010/11 suggest there has been a reduction in Australians’ pro-tanning attitudes. In 2010/11, fewer adolescents and adults desired a suntan, perceived social norms supported tanning, and held beliefs that a suntan is attractive or healthy. These improvements were more marked over the longer period than since 2006/07.
Behaviour change in favour of sun protection was also noted, but mainly for adults. In 2010/11 the amount of time adults spent outdoors during peak UVR was significantly less than the baseline and the recent period, and the proportion who were outdoors decreased compared with 2003/04. For adolescents the main change was reduced odds of being outdoors during peak UVR hours in 2010/11 compared with 2006/07. There were some other improvements in adults’ sun protection in 2010/11 noted, with increased odds of sunscreen use since 2003/04, and increased odds of wearing long-sleeved clothing since 2006/07. Disappointingly, there was evidence that hat use among adults decreased in 2010/11 compared with the two previous surveys.
Despite the variable improvements in adolescents’ and adults’ sun protection, there was an overall reduction in odds of sunburn on the previous weekend for both age groups in 2010/11 compared with the other two survey periods. Although this reduction in sunburn incidence was a good outcome for skin cancer prevention, the 2010/11 summer was relatively wet for the east coast,28 and our analyses may not fully adjust for varying weather conditions and UVR levels across the years. Therefore, the reductions in sunburn incidence may not be totally related to behavioural changes, rather an artefact of unaccounted environmental UVR patterns. Nonetheless, there was sufficient evidence of change in time spent outdoors in 2010/11 together with improvements in adults’ increased use of long-sleeved clothing to suggest at least some behavioural contributions. Previous analyses from the 2003/04 survey have shown both time spent outdoors and use of long-sleeved clothing reduced the odds of sunburn among adults, but that use of sunscreen and hats did not.21 Thus, improvements in adults’ sunscreen use since 2003/04 is likely to have had only limited impact on the observed decrease in sunburn odds. However, together with the shift in attitudes, it seems feasible that at least some of this improvement might be attributed to the implementation of campaigns and prevention strategies.
Despite these reported attitudinal and behavioural changes, there were areas where respondents’ compliance with sun protection was less than optimal in summer 2010/11. Sunscreen was the most common form of sun protection used among adolescents. However, prevalence of use of other specific sun protection behaviours was typically reported by less than one-quarter of adolescents and was lowest for three-quarter length or long-sleeved clothing. Although adults were generally more compliant with sun protection than adolescents, fewer than half used sunscreen, wore a hat, cap or visor outdoors on the weekend, or had used two or more sun-protective behaviours. These low rates of compliance with sun protection might reflect the relatively cool and wet summer in 2010/11.28 However, improvements were noted across the seven-year period, and therefore other factors are maintaining this low compliance. Even the relatively low sunburn incidence across the eight weeks of the survey in 2010/11 translates to about 363,000 adolescents and two million adults being sunburnt on any given summer weekend,27 so there is plenty of cause for concern.
Previous evidence suggests intensity and reach of exposure to campaigns can influence sun-related attitudes and behaviours.13 The shift in attitudes observed is consistent with relatively sustained intensive broadcast of summer campaigns since 2006, compared with summer 2003/04. The target audience and content of campaign advertisements has been consistent across the survey period. Typically, advertisements targeted youth, while appealing to a broad audience. The messages aimed to raise awareness about severity and susceptibility to skin cancer, the dangers of tanning, and reminding people to use sun protection. The slowing of improvements in attitudes and adults’ sunscreen use in the recent period is consistent with a reduction in campaign broadcasting in 2010/11, following a peak of broadcast activity during the national campaign that extended from 2006/07 until January 2010. Moreover, since this time there has been limited capacity of smaller States and Territories to deliver campaign messages on television.
Further prevention initiatives are needed. A greater government investment in skin cancer prevention televised campaigns across Australia during high UVR periods of the year is therefore warranted, and supported by cost effectiveness evidence.31,32 Campaign themes need to be expanded to address specific circumstances where over-exposure is more common, and this might be informed by future analyses from these national surveys to identify settings and activities where risk of sunburn is high. Given news stories on vitamin D have been at times less supportive of sun protection messages in recent years,33 and Queensland adults commonly believed using sun protection will reduce their vitamin D levels,34 campaign messages may also benefit from providing more information about seasonal and daily variations in UVR. A number of skin cancer control programs in the southern States have begun to address this issue by communicating about vitamin D as well as sun protection requirements in recent years. Moreover, future analyses are required to examine the level of advertising that is related to improved outcomes nationally over the study period. This will require collation of objective records of exposure to skin cancer prevention activity in each State and Territory for use in future studies.
Additional prevention strategies are needed to advance policy to ensure sun protection options are easily accessible. Ensuring the availability of low-cost, high-quality sun protection products remains a legitimate course of advocacy and action by skin cancer control programs. Culturally normative options need to be explored and advanced. Efforts to promote the use of broad-brimmed hats as being more the social norm (i.e. fashionable and acceptable), particularly by adolescents, are warranted. Further qualitative research to examine gender differences is needed, and could be used to expand our understanding of what gender-related cultural norms are salient (e.g. “blokes don't use cosmetics” being a barrier to sunscreen use). Moreover, future analyses from this national survey focusing on gender differences in UVR exposure would also be valuable, given the growing disparity in males’ and females’ skin cancer incidence and mortality.3,4 Nonetheless, efforts to promote use of sunscreen (anecdotally, particularly among men and boys) also deserve further attention. Research suggests sun protection policies in workplaces may potentially have a normative influence on sun protection compliance and should also be promoted.35 Larger workplaces with better occupational health and safety infrastructure may be more likely to promote good sun protection practice in the workplace.36,37 Strengthening of codes of practice in this sphere by working with relevant regulators such as Safe Work Australia has substantial potential.
Similarly, improvement in sun protection opportunity in the built environment will assist to embed systematic sun protection into Australian cultural practice. Evidence suggests increased shade in high schools will be well taken up,38 and programs to increase the provision of shade in school settings are worth promoting to reduce adolescent sun exposure. Increasing pressure on relevant State and local building codes to include shade provisions in design requirements, particularly in recreational settings, also deserves examination. Policy relevant research investment should be prioritised to drive these initiatives.
A strength of this national sun protection study is the assessment of respondents’ sun protection behaviours for a specific day and location during peak UVR hours, which enables adjustment for temperature and UVR in analysis of trends over time to separate out behavioural from environmental effects. We rely on self-reports of sun protection which may be subject to socially desirable responses.39 Moreover, response rates for the surveys were low but, given the short two-day window for recruiting respondents to weekly interviews (to minimise cueing and recall biases), this is expected.
The findings suggest improvements in skin cancer prevention attitudes of Australians during the past seven years. Moreover, Australian adults’ compliance with sun protection during summer has improved in some areas but is still far from ideal. There were no significant improvements in adolescents’ sun protection, and a reliance on sunscreen for protection. Nonetheless, their sun exposure during peak UVR hours was reduced in 2010/11. It is encouraging that there has been a sustained decrease in weekend sunburn among adolescents and adults.
We acknowledge the work of the research advisory committee chaired by Professor David Hill for establishing the survey protocols in 2003, and Craig Sinclair and Professor Mark Elwood in establishing the initial funding for the national sun protection survey. We thank members of Cancer Council Australia's national skin cancer committee for contributions and support of the study, and the Australian Radiation Protection and Nuclear Safety Agency and the Bureau of Meteorology for provision of UVR and climate data respectively, and Bhavani Sridharan, Kimberley Dunstone, Philippa Niven, and Poly Makris at Cancer Council Victoria, who assisted with data management and reporting support.
The State and Territory cancer councils, the Australian government Department of Health and Aging and the National Cancer Control Initiative funded the 2003/04 survey; Cancer Council Australia and Cancer Australia funded 2006/07 survey; and Cancer Council Australia funded the study in 2010/11.