New Zealand has among the highest age-standardised incidence and mortality rates for cutaneous malignant melanoma (CMM) in the world.[[1]] In 2018, New Zealand Cancer Registry age standardised registrations for melanoma were 42.1 and 31.8 per 100,000 for males and females respectively; 296 deaths from CMM were recorded.[[2]] The more numerous keratinocyte cancers are not required to be registered but accounted for an additional 204 deaths that year.[[2]]

The most effective way to prevent skin cancer is to reduce exposure to ultraviolet radiation (UVR) and sunburn with: body coverage (clothing, hat, sunglasses), using sunscreen and seeking shade or rescheduling activity when UVR is lower.[[3]] However, concurrent use of multiple sun-protection strategies is rarely reported by research participants, as many people use only one or two options when outdoors.[[4]] Previous research by Bleakley and colleagues[[4]] conceptualised three main forms of sun-protection (“Cover” using clothing or hat, “Protect” using sunscreen or sunglasses, and “Avoid” using shade or rescheduling outdoor activity) based on known mechanisms to reduce UVR exposure and sunburn. Conceptualisation of sun-protection in this manner is useful to show groups of similar strategies and their influence on sunburn. Their study showed that using one sun-protective strategy may preclude the use of other strategies, known as compensation behaviour.[[4]] However, their findings were limited by applying regression approaches to each sun-protection strategy individually. We hypothesise that it is possible to extend Bleakley and colleague’s study[[4]] by employing a more sophisticated Seemingly Unrelated Structural Equation Model (SEM) to analyse the data, which allows for simultaneous testing of all sun-protection strategies against the outcome sunburn.

With the Southern Hemisphere summer bringing extreme levels of UVR between September and April,[[3,5]] now is the time to encourage the use of multiple sun-protection strategies among New Zealanders, particularly those with sun-sensitive skin types who sunburn readily and tan rarely.[[6]] The aims of this paper are to broadly replicate the study by Bleakley and colleagues[[4]] to report: 1) frequencies of specific sun-protection strategies (“Cover”, “Protect”, and “Avoid”); 2) associations of these strategies with sunburn outcomes; and 3) evidence of compensation behaviours for sunburn prevention among the New Zealand population, using data from a cross-sectional nationally representative dataset.

Methods

Study design and participants

Te Hiringa Hauora/New Zealand Health Promotion Agency Sun Exposure Survey (SES) dataset is a cross-sectional, nationally representative sampling of New Zealanders aged 13+ years, conducted during the 2016 Southern Hemisphere summer, who were outdoors for at least 15 minutes between 10am to 4pm on the day selected as the reference interview day.[[7]] In brief, n=2,272 people were interviewed, distributed by geographic region according to quota targets based on known population distributions.[[7]] For consistency with previous studies, only data for participants aged 15+ years  who met the outdoor criteria were analysed (n=1,924).[[8,9]] Prior to weighting, sample sizes for age groups were 15–24 years 25.5%, n=491; 35–34 years 11.9%, n=229; 35–44 years 18.7%, n=359; 45–54 years 20.5%, n=395; and 55+ years 23.4%, n=450. Males 45.9%, n=884; females 54.1% n=1,040. Ethnicity sample sizes were Māori 10.6%, n=204; Pasifika 3.0%, n=58; Asian 6.0%, n=116; NZ European 79.9%, n=1520; and Other 1.4%, n=26. Due to missing data on sun protection strategies, some analyses are based on smaller samples. Ethical approval for analysing and reporting this audit-related data was obtained from the University of Otago Human Ethics Committee (HD17/039). Respondent participation was taken as informed consent.

Measures

Sunburn

Participants were asked “Did you get sunburnt? By sunburnt we mean any amount of reddening of the skin after being in the sun”.

Sun-protection behaviours

Participants were asked a series of dichotomous questionnaire items about sunburn prevention behaviours they used while outdoors, grouped into three strategies based on Bleakley and colleagues:[[4]] Cover (with t-shirt, shorts, wide-brimmed hat); Protect (with sunglasses, sunscreen); Avoid (using shade, rescheduling outdoor activity). These strategies, conceptualised by Bleakley and colleagues,[[4]] were based on known mechanisms to reduce UVR exposure and sunburn/skin cancer. Each strategy was classified as 1=“Yes”/0=“No” reflecting whether the respondent had reported any of the associated sunburn prevention behaviours.

Covariates included self-reported socio-demographic characteristics (respondent sex, age (years), skin-type using Fitzpatrick sun-sensitivity scale I–IV,[[6]] ethnicity, region of residence and of outdoor activity, highest educational qualification, household income, self-assessed skin cancer risk score), duration outdoors (minutes) and concurrent climatic conditions (UVR, air temperature).

Statistical methods

Sun-protection behaviours were concatenated to show strategies used by participants. A Seemingly Unrelated SEM[[10]] was applied to the data to simultaneously test all sun-protection strategies against the outcome sunburn/not sunburned. The parameters were estimated using the full information likelihood method to account for indigeneity problems. Data were analysed using SAS 9.4[[11]] and Mplus.[[12]]

Results

Sunburn was reported by 14.9% (279/1,877) of participants. Figure 1 shows the sun-protection strategies of those sunburned on the target day: 16.8% (19/113) did not use any sun-protection strategies and 13.2% (60/456) used all strategies. Most (69.7%) (1,308/1,877) behaviour was consistent with a compensation strategy of only one or two to sun-protection options. The protect strategy (sunglasses and sunscreen) yielded the most (23.4%) (52/222) sunburn cases. Examination of the SEM (see Table 1) shows no particular behavioural strategy statistically significantly prevented sunburn (p>0.07); sunglasses and sunscreen were positively associated with sunburn (p<0.001).

View Figure 1 & Table 1.

Discussion

The aims of this study were, firstly, to report use of specific sun-protection practices. We found that, during the Southern Hemisphere summer of 2016, 16.8% of sunburned participants did not use any sun-protection strategies. The highest proportion of sunburn cases were among those who used the Protect strategy (23.4%). Secondly, two of three SEM pathways between protective strategies and prevention of sunburn were statistically non-significant. Use of sunglasses and sunscreen was positively associated with sunburn (p<0.001). We concluded that these findings were consistent with the practice of compensation behaviours—in support of Bleakley and colleagues[[4]] in which the majority of sunburned participants used sub-optimal sun-protection. Caution is advised regarding interpretation of results as assessment of weekend sunburn prevalence may not accurately extend to findings on studies reporting annual sunburn prevalence. Further, our approach may potentially yield different estimates as our sun protection scores were calculated differently to Bleakley and colleagues (2018).[[4]] In addition, sun protection factor (SPF), brand or amount of sunscreen used by the participants was not recorded. Other limitations include the potential for bias from untested instrument validity among the New Zealand population. Nevertheless, the study sample was representative of the New Zealand population.

In conclusion, this study confirms that many sunburned New Zealanders did not use optimal sun-protection, consistent with the operation of compensation behaviours. Each protective strategy may be seen as an alternative rather than complementary pattern of behaviour, precluding the use of other strategies. Further investigation of compensation behaviours is warranted to provide insight into sun-protection barriers because ineffective sun-protection is problematic among populations with high skin cancer rates. Now that New Zealand is in a period of high UVR during months September to April,[[3,5]] there is an opportunity to reduce sunburn and consequent CMM among those with sun-sensitive skin types. This can be achieved by using multiple sun-protection strategies (body coverage, using sunscreen, seeking shade or rescheduling outdoor activity) during high UVR months.