This indicator shows the proportion of adult and adolescent Australians reporting that they had sunburn (i.e. any amount of reddening of the skin after being in the sun) and the proportion reporting that they used a combination of two or more sun protection behaviours on the previous weekend (in summer).
Numerator:
1) Adults (aged 18 to 69 years) and adolescents (aged 12 to 17 years) who reported getting sunburnt on the previous weekend (in summer).
2) Adults (aged 18 to 69 years) and adolescents (aged 12 to 17 years) who reported reporting that they used a combination of two or more sun protection behaviours on the previous weekend (in summer). Sun protection behaviours include wearing a hat or applying sunscreen, in combination with one other sun protection behaviour defined as wearing a hat, applying sunscreen, wearing at least ¾-length clothing on arms or legs, or staying mostly in the shade.
Denominator:
Adults (aged 18 to 69 years) and adolescents (aged 12 to 17 years) resident in Australia.
For the purpose of this analysis, ‘city residents’ are defined as residents of the capital cities of Australia’s states and territories. SEIFA refers to Socio-Economic Indexes for Areas. The SEIFA Index of Relative Socioeconomic Disadvantage was used which was derived from residential postcode characteristics. ‘Lowest decile’ refers to areas of greatest disadvantage.
Data sources:
National Sun Protection Surveys– 2003-04, 2006-07, 2010-11, 2013-14. Data supplied by the Centre for Behavioural Research in Cancer, Cancer Council Victoria, on behalf of the National Skin Cancer Committee, Cancer Council Australia.
The National Sun Protection Surveys assess Australian residents’ sun-related knowledge, attitudes and behaviours, and sunburn experiences in weekly cross-sectional telephone interviews over eight weeks in summer. The survey sample frame is selected to obtain a representative sample of Australian residents aged 12 to 17 years and 18 to 69 years in households with a landline telephone. Additional interviews were undertaken in smaller states and the Australian Capital Territory to improve precision of estimates. In 2013-14, a dual-frame landline/mobile telephone sample was used.
Methodology:
Prevalence figures were weighted to be representative of the Australian population with telephones by age, sex and state. For sunburn analyses in 2013-14, data was derived from representative samples of the adult (18-69 years, n=5,288), and adolescent (12-17 years, n=1,061) residents in Australia 2013-14. For sun protection behaviours, data was derived from representative samples of the adults (18-69 years, n=3,683), and adolescents (12-17 years, n=830) resident in Australia in 2013-14, and who were outdoors for at least 15 minutes during peak UVR hours on the designated weekend.
Comparisons of data from earlier National Sun Protection Surveys involved adjusting for theoretically important predictors influencing sunburn and sun protection behaviours that otherwise might have biased reported changes over time in response to preventive initiatives. Models were adjusted for age, sex, skin type, and the weather conditions or ambient ultraviolet radiation (UVR) levels for the date and city relevant to the location and time of respondents’ activities on the preceding weekend.
Due to correlations between weather variables and UVR, temperature, cloud cover and wind speed adjusted models were used for sun protection behaviours and UVR adjusted models were used for sunburn. The adjustment for UVR was considered to provide a more rigorous analysis of changes in sunburn over time, given that exposure to different UVR levels can directly determine sunburn outcomes for unprotected skin. Limited UVR and temperature records were available regionally. Therefore, the trend analyses included a subset of respondents for whom the relevant UVR and temperature data were available. These respondents were located within metropolitan areas and were outdoors during peak UVR hours on at least one day of the weekend preceding their interview. For the sunburn analyses, metropolitan residents who were indoors during peak UVR hours on the weekend were also included. Additionally, respondents who were sunburnt at times other than during peak UVR hours on the weekend were excluded from these sunburn analyses.
Statistical analysis
Cancer Council Victoria performed statistical analysis of the trend data on behalf of Cancer Australia. These multivariate analyses - logistic regression using Stata adjusting for age, sex, skin type, and weather/UV conditions - examined the change in the prevalence of sunburn prevalence and sun protection behaviours over time. The following tables show the outcomes of these analyses:
| Table 1: Australian adults’ sunburn prevalence on summer weekends 2003-2014 |
| |
Persons |
Males |
Females |
| |
% |
95% Cl |
Significance |
% |
95% Cl |
Significance |
% |
95% Cl |
Significance |
| 2003-04 |
13.5% |
12.0%-14.9% |
Ref. |
15.7% |
13.5%-17.9% |
Ref. |
11.3% |
9.4%-13.2% |
Ref. |
| 2006-07 |
9.8% |
8.5%-11.0% |
*** |
11.9% |
9.9%-13.9% |
* |
7.8% |
6.2%-9.4% |
** |
| 2010-11 |
7.3% |
6.2%-8.4% |
*** |
9.4% |
7.6%-11.2% |
*** |
5.2% |
3.9%-6.5% |
*** |
| 2013-14 |
9.6% |
8.4%-10.7% |
*** |
11.8% |
9.9%-13.6% |
** |
7.5% |
6.0%-8.9% |
** |
Adjusted percentages, 95% CI and significant change since 2003-04 modelled in multivariate analyses adjusting for age, sex, skin type and weekend UV Index levels.
Sign. For each survey year the statistical significance of the change in adjusted prevalence was compared with 2003-04. Statistical significance was set at p=0.05.
Ref. Referent survey year (2003-04)
*** p <0.001 ** p<0.01 *p<0.05
Weights were applied to retain a representative sample of the population within each survey year.
UVR data was provided by the Bureau of Meteorology and the Australian Radiation Protection and Nuclear Safety Agency.
| Table 2: Australian adolescents’ weekend sunburn prevalence on summer weekends 2003-2014 |
| |
Persons |
Males |
Females |
| |
% |
95% Cl |
Significance |
% |
95% Cl |
Significance |
% |
95% Cl |
Significance |
| 2003-04 |
19.4% |
14.7%-24.1% |
Ref. |
19.3% |
13.0%-25.7% |
Ref. |
19.3% |
12.4%-26.2% |
Ref. |
| 2006-07 |
20.3% |
15.1%-25.4% |
n.s. |
24.6% |
16.6%-32.6% |
n.s. |
16.4% |
9.7%-23.0% |
n.s. |
| 2010-11 |
13.1% |
9.8%-16.3% |
* |
14.6% |
9.3%-20.0% |
n.s. |
11.6% |
8.0%-15.1% |
* |
| 2013-14 |
17.1% |
13.6%-20.6% |
n.s. |
19.4% |
14.1%-24.7% |
n.s. |
14.8% |
10.3%-19.3% |
n.s. |
Adjusted percentages, 95% CI and significant change since 2003-04 modelled in multivariate analyses adjusting for age, sex, skin type and weekend UV Index levels.
Sign. For each survey year the statistical significance of the change in adjusted prevalence was compared with 2003-04. Statistical significance was set at p=0.05.
Ref. Referent survey year (2003-04)
*** p <0.001 ** p<0.01 *p<0.05
Weights were applied to retain a representative sample of the population within each survey year.
UVR data was provided by the Bureau of Meteorology and the Australian Radiation Protection and Nuclear Safety Agency.
International data:
New Zealand data were sourced from the Sun Exposure Survey 2013: Topline Time Series Report (see: http://www.hpa.org.nz/research-library/research-publications/sun-exposure-survey-2013-topline-time-series-report). Data were collected only in areas in which at least one weekend day received a high score using a fine weather criteria standard, for respondents who reported being outdoors for 15 minutes or more between 10am and 4pm. Data from this survey were weighted so that no specific population was over- or under-represented.
Caution should be applied in comparing the Australian and New Zealand data, due to the New Zealand survey being conducted when “fine weather” is present, while the Australian NSPS is conducted across multiple weekends regardless of the weather. This difference in methodology could therefore account for some of the apparent difference in sunburn incidence between the two countries.
