Incorrect text: In “Frequency of sunburn in Queensland adults: still a burning issue” published in the 6 May 2013 issue of the Journal (Med J Aust 2013; 198: 431-434), the third sentence on page 431 should read: “In Australia, skin cancers cost more than other major cancers.2”

The impact of skin cancer is disproportionately high due to its extraordinarily high incidence in white populations compared with other cancers. In the United States, skin cancer costs an estimated US$2 billion annually.1 In Australia, skin cancers cost more than other major cancers.2 Sunburn, the acute inflammation caused by excessive exposure to solar ultraviolet radiation, is a determinant of all major skin cancers.3 An estimated 1.3 million skin cancer cases were due to excessive sun exposure4 in the US in 2003, and a regional study in Texas calculated the economic impact of sunburn as over US$10 million annually through lost work and treatment costs.5 Sunburn prevalence (at least one sunburn in the past year) in white US adults aged 18–29 years has been constant for a decade at 66%.6 Men, the young and high-income groups appear susceptible.7 In Australia, weekend sunburn prevalence was 9% overall in Victoria in 20028 but in Queensland in 2004, 70% of surveyed residents aged 20–75 years reported sunburn in the past year.9

Sunburn is of crucial public health importance as a key preventable and common risk factor for skin cancer. We assessed frequency of sunburn and associated factors in Queensland in two surveys in 2009 and 2010.

The Queensland Health Population Epidemiology Unit conducted self-reported health status surveys from January to March 2009 (SRHS 2009) and from October 2009 to February 2010 (SRHS 2010) using computer-assisted telephone interviews10 with approval from the Queensland Health Central Office Human Research Ethics Committee. The target population sourced by random-digit-dialling was Queensland households with a person aged ≥ 18 years (SRHS 2009) or ≥ 16 years (SRHS 2010). From each selected household, one eligible resident was asked to participate. Response rates were calculated as the number of completed interviews expressed as a percentage of the number of eligible persons contacted.

A structured interview using a scripted questionnaire covered demographic and risk factors including sunburn on the previous weekend. Sunburn was defined as any reddening of the skin lasting longer than 12 hours after sun exposure. Other information collected included age, sex, height and weight, education, marital and employment status, fruit and vegetable intake, smoking, general use of sun protection and physical activity based on the Active Australia instrument.11 For analysis, employment status was categorised as employed, retired or student/carer/unemployed; fruit/vegetable consumption as meeting daily recommended levels (≥ 2 serves of fruit and ≥ 5 serves of vegetables);12 smoking as daily or not; and physical activity as none, insufficient (1–4 sessions of walking/moderate/vigorous activity and/or < 150 minutes total, weekly), or sufficient to meet national guidelines13 (≥ 5 sessions for ≥ 150 minutes total, weekly). Statistical Local Areas of residence were classified by the Accessibility/Remoteness Index of Australia14 (ARIA+), the Socio-Economic Indexes For Areas15 (SEIFA) and geographically (North, Central, South and West Queensland).

Identical variables from the 2009 and 2010 surveys were pooled for adults aged ≥ 18 and weighted by age, sex and Queensland Health Service District distribution using 2008 estimated resident population data for Queensland (Australian Bureau of Statistics, http://www. abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/3235.02008?OpenDocument) to minimise bias due to over- or underrepresentation of any demographic group, and by the number of in-scope people and fixed telephone lines per household to minimise selection bias. Weighted proportions (with 95% confidence intervals) of the population who reported being sunburnt the previous weekend were calculated, stratified by age and sex. Univariate and multivariate logistic regressions — adjusting for age (18–24, 25–34, 35–44, 45–54, 55–64, ≥ 65 years), sex, education (bachelor degree or higher, diplomas/certificates/trade, no qualifications) and physical activity — were undertaken to estimate the association between sunburn and risk factors, using Stata SE, version 11.0 (StataCorp).

Response rates were 57% (7537/13 289) and 65% (8938/13 857) in the 2009 and 2010 surveys, respectively. Pooling of the two surveys (results did not differ) provided a total of 16 473 Queensland adults. Of these, 15% lived in North, 20% in Central, 63% in South, and 2% in West Queensland, while 55% lived in major cities, 39% in regional areas and 6% in remote areas, based on census data of the estimated resident population of Queensland. Of the 8150 men, 901 (12.9%, weighted; 95% CI, 11.9%–14.0%) reported being sunburnt the previous weekend, compared with 596 of 8323 women (8.6%, weighted; 95% CI, 7.7%–9.6%) (Box 1). Peak weighted prevalence of 22% (102/459) was among young men aged 18–24 years. There was a clear trend of decreasing sunburn prevalence with increasing age in both sexes (P < 0.001), such that sunburn was least common among participants aged ≥ 65 years (men, 2.2%, weighted [54/1938]; women, 1.8%, weighted [35/1950]) (Box 1).

After adjustment, the strongest predictor of sunburn was youth: people aged 18–24 years were seven times more likely to report being sunburnt than those aged ≥ 65 years, although adults aged 45–64 years were also at significantly higher risk of being sunburnt than those ≥ 65 years (Box 2).

Other sunburn predictors were male sex, not having a tertiary education, and being in the workforce. People who had undertaken physical activity in the previous week, regardless of frequency or duration, were more likely to be sunburnt than those who did not. People who undertook the recommended level of physical activity had twice the odds of sunburn of inactive people, while those who undertook one to four sessions were also at risk (Box 2). Those who generally used sunscreen in summer tended to be less likely to be sunburnt than non-users (adjusted OR, 0.86) (Box 2). Although hat-wearing per se was not associated with sunburn after adjustment for confounding factors (Box 2), those who usually sought shade or wore protective clothing in summer also had significantly lower odds of sunburn than those who did not, and respondents’ most common explanation for sunburn was failure to use clothing or sunscreen protection (SRHS 2010 only; data not shown).

No other factors, including marital status, body weight, smoking, fruit/vegetable consumption, location of residence, socioeconomic status (Box 2) or skin colour (SRHS 2010 only; data not shown) were associated with sunburn.

If sunburn in Australia were reduced, decreased rates of skin cancer would follow.3 Queensland has the highest melanoma rates in Australia, and Australia and New Zealand have the highest rates in the world.16 We have shown that around one in 10 Queensland adults report being sunburnt on the weekend in summer. Men were about 50% more likely than women to experience sunburn and adults aged under 65 years were more likely to be sunburnt than those aged 65 years and over. Youngest adults (aged 18–24 years) were sunburnt most often. While people in the workforce and those without tertiary qualifications are more likely to report sunburn than others, this probably reflects sunburn in leisure more than work hours, since weekend sunburn prevalence was assessed. Physical activity was also associated with sunburn.

Our results are broadly consistent with a 2004 Queensland survey showing young age and male sex greatly increase odds of sunburn.9 An earlier analysis of SRHS 2010 data also showed that people engaging in physical activity were more likely to experience sunburn on the previous weekend and during the past year, especially those who undertook ≥ 7 hours of activity per week.17 Sun-safety and physical activity promotion messages must be integrated, acknowledging the importance of both for health and wellbeing.

This study was drawn from adults of all ages across Queensland but excluded households without fixed telephones, and restricted detail was available about timing of physical activity in relation to sunburn. However, we have shown that despite half a century of campaigns, sun protection in Queensland remains far from optimal. Vigilance to avoid sunburn is essential among active adults. Men still are more likely to be sunburnt and need targeted encouragement to practise prevention measures. By experiencing sunburn, often repeatedly,9 Queenslanders are driving their already high risk of skin cancer higher.3 In Victoria, where the SunSmart program originated, prevention programs have recently stalled.8 Redoubled efforts are required through television and other media campaigns to improve sun-protection and ultimately reduce skin cancer.