Prevalence and Correlates of Observed Sun Protection Behaviors across Different Public Outdoor Settings in Melbourne, Australia

Skin cancer prevention efforts in Australia have increasingly incorporated a focus on protection during incidental sun exposure. This complements the long-present messages promoting protection in high-risk settings and avoidance of acute intense bouts of sun exposure. Data from two waves of a cross-sectional direct observational survey was used to assess the prevalence and correlates of N = 12,083 adolescents' and adults' sun protection behavior (arm and leg cover, hat, sunglasses, and shade cover). Individuals were observed in public outdoor settings in Melbourne, Australia during peak ultraviolet (UV) times (11 a.m.–3 p.m.) on summer weekends. Settings included pools and beaches, parks and gardens, and for the first time in 2018, outdoor streets and cafés which may capture more incidental forms of sun exposure and represent another public setting where Australians commonly spend time outdoors. Females and older adults were consistently better protected than males and adolescents. Physical activity was strongly associated with low shade cover across settings. Weather was more strongly associated with sun protection at outdoor streets/cafes and parks/gardens than at pools/beaches but use of observed sun protection (particularly arm cover and covering hat) was low across settings. Continued public education about UV risk and its relation to weather and the seasons is needed to promote the routine use of multiple forms of sun protection during outdoor activities in peak UV times, especially among males and adolescents. Findings also highlight the importance of considering activity demands of public spaces in shade planning to optimize sun protection during outdoor activities in public spaces.

Keywords: cancer prevention and screening; health behavior; health promotion; multilevel modeling; neighborhood; quantitative methods

Exposure to ultraviolet (UV) radiation causes skin cancer. While intermittent bouts of acute UV exposure are associated with increased risk of malignant melanoma (the more fatal form of skin cancer), cumulative low doses are associated with both melanoma and squamous cell carcinoma ([2]; [38]). Australia's comprehensive skin cancer prevention program SunSmart was established in 1988. It encompasses public education, advocacy, and partnerships with workplaces, schools, and local governments to promote both personal protective behaviors and environmental change such as shade provision in public places ([43]; [46]). Early SunSmart advertising campaigns primarily featured imagery of individuals modelling protective behaviors during outdoor leisure activities in water-based settings such as beaches and later campaigns focused on raising awareness of the dangers of tanning, both of which are characterized by intense acute bouts of UV exposure ([26]; [61].). A substantial decrease in protanning attitudes among Australian adolescents and adults occurred between 2003 and 2011 ([65]). Given the unique risks of cumulative doses of UV, recent SunSmart public education efforts have focused on protection during everyday activities in settings where incidental exposure may occur, such as while doing household chores, shopping, or socializing outdoors ([26]; [50]; [67]). To measure the impact of these efforts and to inform ongoing prevention initiatives, methods of monitoring protective behavior during incidental sun exposure are needed.

People's use of protective clothing and shade are influenced by a range of factors beyond UV protection, including thermal comfort, social norms, fashion, activity demands, and the physical environment ([9]; [17]; [18]; [19]; [39]; [45]; [54]; [56]; [64]; [67]). Consideration of the interplay of environmental and personal factors in different settings is therefore useful to inform ongoing skin cancer prevention on both environmental and personal behavior fronts. A long-running observational survey of sun protection behaviors in public outdoor settings in Melbourne Australia demonstrated improved protective clothing use at pools and beaches, parks and gardens, and other public settings in the years following SunSmart's initiation ([19]; [31]). In 2018, the survey was expanded to include outdoor streets and cafés. These new settings were included to capture more incidental forms of UV exposure (relative to pools/beaches where people may intend to spend extended time in the sun). They also represent the public settings where Australians commonly spend time outdoors on summer weekends ([47]). The aim of this study is to examine the prevalence and predictors of adolescents' and adults' sun protection behaviors in newly observed outdoor street/café settings for comparison with other public settings (pools/beaches, parks/gardens).

Data were derived from years 2018 and 2019 of the Sun Observation Study, a serial cross-sectional observational field survey of sun protection behavior in selected public outdoor settings (pools/beaches, parks/gardens, and outdoor streets/cafés) in Melbourne. Trained field-workers were assigned postcodes and identified eligible venues for observation from a Melbourne electronic street directory. Pools without an outdoor pool area, small corner parks, ovals, and wetlands, were excluded; and for outdoor streets/cafés, only main shopping strips with a mixture of shops and cafés with outdoor seating areas were included. All observations were conducted on summer weekends in January and February between 11 a.m. and 3 p.m. (coinciding with peak UV times). Field-workers used a mobile phone application to record observed cloud cover (dichotomized for analysis: sunshine vs. mixed thin cloud, mixed thick cloud, and overcast) per venue; and individuals' estimated age (14–19, 20–49, 50+ years), sex, and physical activity (dichotomized for analysis: stationary [lying, sitting, and standing] vs. moderate/vigorous physical activity [walking and exercising]). Due to low numbers of individuals in some physical activity categories (see Supplemental Material), this variable was dichotomized (stationary vs. moderate/vigorous) to allow comparability of models across settings. Whether individuals were part of a social group with other individuals observed was also recorded. Observations were assigned temperature, wind speed, and seasonal rainfall values from the Bureau of Meteorology (mean-centered for analysis). Individuals who were swimming, working, wearing uniforms, or playing organized sport were not observed, and those wearing religious head coverings were excluded from analysis because clothing selection was restricted. See Supplemental Materials for detail on weather variables and selection of venues and individuals.

Outcomes were five binary sun protection variables observed by field-workers: arm cover and leg cover (clothing covering ≥3/4 vs. <3/4 of the limb), covering hat (broad-brimmed, legionnaire's vs. narrow-brimmed, peaked cap/visor, or no hat), sunglasses (wearing vs. not wearing), and shade cover (total vs. partial or no shade cover). In contrast to previous work describing trends over time ([31]), an index of overall body coverage was not included in the present study because the analysis of prevalence and predictors of discrete sun protection behaviors was more appropriate to inform practice. Field-workers also recorded the type of shade used: purpose-built (gazebo, shade sail), natural (tree), portable (umbrella, shade tent), or "other" (structure with purpose other than shade provision: building wall, vehicle). Sunscreen use was not observed.

Field-workers completed a 1-day training session prior to beginning data collection. Project staff outlined the standardized protocols for selecting venues and people for observation, and for coding sun protection behaviors, individual, and venue characteristics using descriptions, diagrams, and photographs. Field-workers carried out practice observations based on stock photographs and YouTube videos, or outdoors in the park and street (weather permitting), to test coding proficiency. Field-workers were provided with a training manual and on-call support for reference during data collection. Interrater reliability assessed in 2019 for all measures in the present analyses was good to excellent (Cohen's kappa = 0.70–1.00 for nominal variables, intraclass correlation coefficient = 0.73–1.00 for ordinal variables; [47]). The study was approved by the Institutional Research Review Committee at Cancer Council Victoria and was conducted in accordance with the Declaration of Helsinki.

Separate multilevel logistic regression analyses were conducted for each outcome in each setting type in Stata 16.1 ([57]). Models included observation year, individual (sex, age, physical activity) and weather variables as covariates, and random intercepts for social group and venue to account for clustering.

Individual and weather characteristics of 2018 and 2019 observations (N = 12,083) in each setting are reported in Table 1.

Graph

Table 1. Characteristics of Included Observations by Setting (N = 12,083).

1 Percentage of those in total shade per setting type. Values are n (%), except bM(SD) before values were centered at mean for inclusion in regression models.

At pools/beaches, nearly half of observed individuals were wearing sunglasses, but arm cover, leg cover, covering hat, and shade were used by relatively few (<12%, Table 2). Purpose-built structures were the most commonly used shade type (Table 1). Females had higher odds of arm and leg cover, covering hat, and sunglasses use than males, but were no more likely to use shade. Adults had higher odds of covering hat and sunglasses use than adolescents, and those aged 50+ years had higher odds of leg cover than adolescents. Individuals who were physically active had higher odds of arm or leg cover but lower odds of shade cover than those who were stationary. Higher temperature was associated with lower odds of arm and leg cover, while cloud cover was associated with lower odds of covering hat use. Higher temperature, and higher seasonal rainfall were each associated with greater odds of total shade cover, while higher wind speed was associated with lower odds of total shade cover.

Graph

Table 2. Predictors of Sun Protection Behaviors at Pools/Beaches (Adjusted Odds Ratios [95% CI]).

• 2 Note. Models include all listed covariates plus year of survey (2018, 2019), and include random intercepts for social group and venue. Adjusted odds ratios >1 for temperature, wind speed, and seasonal rainfall indicate higher odds of the outcome at higher levels of each predictor.
• 3 p <.05. **p <.01. ***p <.001.

At parks/gardens, 14% of individuals wore arm cover, and fewer than one in ten wore covering hats and/or used total shade (Table 3). Approximately 40% of individuals wore leg cover and/or sunglasses. The majority of individuals in total shade used natural or purpose-built shade (Table 1). Females and adults had higher odds of each form of sun protection than males and adolescents respectively, except for shade cover. Physically active individuals had higher odds of sunglasses use and lower shade cover than those who were stationary. Cloud cover was associated with lower hat and sunglasses use, and higher temperature was associated with lower odds of arm and leg cover and higher shade cover.

Graph

Table 3. Predictors of Sun Protection Behaviors at Parks/Gardens (Adjusted Odds Ratios [95% CI]).

• 4 Note. Models include all listed covariates plus year of survey (2018, 2019), and include random intercepts for social group and venue. Adjusted odds ratios >1 for temperature, wind speed, and seasonal rainfall indicate higher odds of the outcome at higher levels of each predictor.
• 5 p <.05. **p <.01. ***p <.001.

At outdoor streets/cafes, approximately half of individuals observed wore leg cover, while 38% used sunglasses, 28% used shade, and 17% arm cover (Table 4). Only 3% wore a covering hat. The majority of individuals in total shade used purpose-built shade structures. Females had higher odds of wearing sun protective garments than males, but shade cover did not differ. Adults aged 20 to 49 years had higher odds of leg cover and sunglasses use than adolescents, while older adults had higher odds of each sun protection behavior than adolescents. Individuals who were walking or exercising had higher odds of leg cover and sunglasses use, but lower odds of shade cover than those who were stationary. Cloud cover was associated with higher odds of arm, leg, and shade cover, but lower covering hat and sunglasses use. Higher temperature was associated with lower odds of arm, leg, and shade cover, but higher sunglasses and covering hat use. Higher wind speed was associated with lower shade and sunglasses use.

Graph

Table 4. Predictors of Sun Protection Behaviors at Outdoor Streets/Cafes (Adjusted Odds Ratios [95% CI]).

• 6 Note. Models include all listed covariates plus year of survey (2018, 2019), and include random intercepts for social group and venue. Adjusted odds ratios >1 for temperature, wind speed, and seasonal rainfall indicate higher odds of the outcome at higher levels of each predictor.
• 7 p <.05. **p <.01. ***p <.001.

This study demonstrates differences in the prevalence and observed forms of sun protection used in public outdoor settings and variation by demographic and activity characteristics and weather conditions. Clothing covering at least 3/4 of the arms and/or legs was less commonly worn at pools/beaches relative to streets/cafes and parks/gardens, while shade was more commonly used at streets/cafes, in line with previous observations in outdoor dining areas, pools and beaches at North American holiday resorts ([66]). In the present study, there were some commonalities across settings, including that females and older adults were better protected than males and adolescents respectively, but sun protection (especially covering hat use and arm cover) was largely inadequate.

The lower sun protection observed among males and adolescents in the present study is consistent with findings from self-report and direct observation surveys conducted both in Australia ([17]; [25]; [42]; [65]) and internationally ([4]; [36]; [53]; [66]). It is unlikely that males and adolescents' inferior use of physical sun protection would be offset by extensive use of sunscreen. The National Sun Protection Survey shows similar prevalence of sunscreen use among Australian adolescents and adults, and lower prevalence in men than women ([62]). Variation in personal sun protection between males and females in the present study may be primarily attributable to clothing norms and trends, but risk taking, resistance to protective behaviors, and lower health concerns among men may also play a role ([11]; [12]; [14]). These factors may also contribute to the lower sun protection observed among adolescents relative to adults. Ethnographic research revealed that adolescents perceived sun protection to be incompatible with their desire for autonomy, spontaneity, risk taking, and maintaining a social image ([54]). Evidence for patterning of sun protection behaviors by age and gender support the targeting of promotion efforts toward males and adolescents who may not routinely consider sun protection during peak UV times.

Demographic factors were more strongly associated with use of sun protective clothing and accessories than with shade cover. This may be because social norms surrounding clothing and accessories relate to fashion and social identity ([54]), while there may be weaker social norms for avoidance of shade which relate to pursuit of a suntan ([23]; [48]). Shade use is also dependent on thermal comfort (as described in a socioecological model) and shade availability ([3]; [5]; [23]; [48]; [54]).

On the other hand, physical activity (walking or exercising) was associated with lower shade use across all settings. This may reflect the localized nature of available shade in public areas suitable for different types of activity (e.g., purpose-built awnings or shade umbrellas in café seating areas and gazebos over barbecue facilities vs. variable shade over walking paths and common thoroughfares). Audits in various regions of Australia and New Zealand have shown many outdoor recreational and street settings to have insufficient shade ([28]; [29]; [52]). There are numerous options for increasing shade ([52]). Leafy tree canopies are attractive on a number of levels and can provide shade over extensive open space, including walking and cycle paths in parks and gardens ([69]). In cooler temperate climates such as Melbourne, the use of deciduous trees for shade may be beneficial to make these areas attractive for use across the seasons ([52]). Purpose-built shade offers an alternative to natural shade which is particularly useful in regions with poor soils and/or drought-prone regions ([3]; [24]). SunSmart and state governments around Australia provide support for local governments to increase shade provision in public spaces ([15]; [52]; [58]; [60]). Ongoing support of this kind to provide shade that is practical, attractive, and activity-appropriate would be useful to maximize opportunities for sun protection.

Although physical activity was associated with lower shade use in the present study, it was associated with higher odds of some other sun protection behaviors. In particular, individuals who were engaging in physical activity had higher odds of arm and leg cover at pools/beaches, sunglasses at parks/gardens, and sunglasses and arm cover at outdoor streets/cafés than those who were stationary. In a previous study of Australian adults, those who met physical activity guidelines were more likely to apply sunscreen during outdoor activity, but those who did not meet guidelines were more likely to seek shade ([40]). Other studies of adults in the United States ([33]) and Queensland (most of which is classed as subtropical; [35]) have demonstrated that individuals with higher levels of physical activity are more likely to experience sunburn. These findings reflect the joint contribution of more time spent in the sun and the different activities undertaken while outdoors by physically active individuals which increases their susceptibility to UV damage.

As highlighted in previous work ([35]), integrating sun protection messaging in physical activity promotion would be useful given the strong association between physical activity and sunburn risk. One such approach is settings-based promotion of UV protection with implementation of either top–down (sporting association-level) or grassroots (club-level) engagement strategies. Examples include support for policy development at the state level and dissemination of educational messages in the form of posters and signage as prompts for sun protection at the local and club level ([21]; [27]; [46]; [59].). Additionally, partnering with other organizations delivering health promotion messages and/or engaging sports role models for social marketing campaigns may assist in leveraging better engagement with sun protection messages among grassroots sportspeople ([10]; [13]; [21]; [20]). Promotional materials might portray physically active individuals modelling optimal sun protection and include education on the appropriate application of sunscreen, use of covering styles of swimwear and sportswear with maximal UV protection.

Weather conditions were more strongly associated with sun protection at outdoor streets/cafes and parks/gardens than at pools/beaches, while the direction of associations with cloud and temperature were more variable for shade use at outdoor streets/cafes compared with other settings. The limited impact of weather conditions for pools/beaches may reflect social norms for briefer clothing and a more ingrained schema for attire at water-based settings ([9]). However, our findings suggest that sun-protective clothing is likely affected by thermal comfort and/or perceived UV risk across settings, despite the fact that all observations were conducted during peak UV times when five forms of sun protection (hat, shirt, sunglasses, sunscreen, and shade) are recommended ([6]). For example, lower odds of arm and leg cover were observed during hotter sunny weather which is likely a function of thermal comfort ([55]). In contrast, covering hat use was more common on sunny days across settings, and at higher temperatures at outdoor streets/cafes. Hats can keep the head cool and protect the eyes from the sun's glare. However, some people may misunderstand the level of UV risk on cooler and cloudy days and continued messaging to make UV risk top of mind during outdoor activities at peak UV times is needed ([7]; [68]).

Prompts designed to ensure individuals habitually consider sun protection and to enhance understanding of the risks of cumulative damage may be particularly useful for increasing protection during incidental or brief intermittent exposures. Such prompts may be delivered via smartphone apps or in outdoor settings themselves. However, a key challenge with promoting sun protection during incidental exposure away from the home is that many forms of sun protection require forward planning (e.g., dressing appropriately, taking a hat, sunglasses, and sunscreen). A recent systematic review concluded that SMS text messaging of reminders was ineffective at improving sun protection habits ([8]), but the authors noted a need for higher quality studies. Evaluation of a SunSmart smartphone app designed to provide advice and timely reminders about UV levels and required sun protection concluded that low understanding of UV was a major barrier to potential efficacy ([49]). The field would benefit from targeted research investigating ways to promote habit formation and UV understanding to promote optimal sun protection during incidental sun exposure.

Although less than half of individuals observed in the present study wore sunglasses, they were one of the most prevalent forms of sun protection across settings. Similarly, a U.S. study found that use of sunglasses was common across pools, beaches, and parks ([41]). Use of sunglasses in outdoor streets/cafés in the present study was higher (38%) than observed among pedestrians in subtropical Queensland (33%), despite observations being made in an unshaded thoroughfare ([16]). The eyes are rapidly subject to the effects of glare such that people may feel immediate relief from wearing sunglasses, whereas the skin may be exposed for longer before effects are noticed. The relatively high use of sunglasses may also reflect high perceived risk of UV-related eye damage (e.g., cataracts, skin cancer, and photoaging), social norms and fashion trends, and/or more established habits for many individuals when venturing outdoors in summer ([1]; [34]; [44]; [54]). Investigating ways of promoting their more widespread and habitual use during incidental UV exposure would be useful.

A strength of this study is the direct observational methods, circumventing biases that can affect self-report surveys ([22]; [30]). However, the results are specific to individuals who remained outdoors during peak UV times, and may not apply to individuals in rural or private settings, or venues where sun protection policies may be in place. Another limitation is that we were not able to account for sunscreen use. Daily sunscreen use can prevent melanoma ([51]). Despite sunscreen being the most frequently nominated form of sun protection by Australians ([37]), it is commonly applied inadequately ([62]) and use of five forms of sun protection and avoidance of exposure during peak UV times are therefore optimal. Due to small sample sizes, prevalence and predictors of sun protection behaviors could not be estimated separately for outdoor streets and cafés. Future studies would benefit from a more granular categorization of outdoor settings as activity patterns differ and may have implications for sun protection practices.

This study examined the prevalence and predictors of sun protection behaviors in newly observed outdoor street/café settings for comparison with other public settings (pools/beaches and parks/gardens). The findings suggest that skin cancer prevention efforts in Australia should continue to promote the routine use of five forms of sun protection while outdoors during peak UV times, particularly among males and adolescents. Continued public education to promote awareness and understanding of UV risk (including its relation to weather and the seasons, and the risk of cumulative damage) may prompt individuals to consider UV protection with higher priority during outdoor activities across a range of settings. The findings also highlight the importance of considering activity demands of public spaces in shade planning. Despite gains in population-level sun protection since the initiation of SunSmart in Australia ([63]), this and other studies evidence a need for continued promotion efforts to maintain improvements and to target population groups and activities where protection is suboptimal. Without ongoing commitment and investment in skin cancer prevention, other sociocultural drivers and personal motives affecting UV exposure such as fashion trends, thermal comfort, and activity demands may prevail, resulting in declines in sun protection among future cohorts ([32]).

sj-docx-1-heb-10.1177_10901981211026535.docx

sj-docx-1-heb-10.1177_10901981211026535 – Supplemental material for Prevalence and Correlates of Observed Sun Protection Behaviors Across Different Public Outdoor Settings in Melbourne, Australia

Supplemental material, sj-docx-1-heb-10.1177_10901981211026535 for Prevalence and Correlates of Observed Sun Protection Behaviors Across Different Public Outdoor Settings in Melbourne, Australia by Ashleigh Haynes, Andrea Nathan, Clover Maitland, Helen Dixon, Anna Nicholson, Melanie Wakefield and Suzanne Dobbinson in Health Education & Behavior

We gratefully acknowledge David Hill, Catherine Segan and colleagues for conceiving and establishing the Sun Observation Survey. We also thank the Bureau of Meteorology for providing the data on weather conditions in Melbourne, and the teams of field-workers who collected and entered data throughout the history of the survey.