Sun Protection Use at Work and Leisure by Outdoor Workers in Alberta, Canada

Abstract

Objective: 

To assess the prevalence and determinants of sun protection behaviours among outdoors workers at work and leisure in Alberta, Canada.

Methods: 

We collected outdoor workers’ demographics, skin cancer risk factors, job information, and sun habits at work and leisure using self-completed questionnaires. For both settings, we compared use of specific behaviours and modelled determinants of sun protection behaviour scores.

Results: 

At work, wearing a sleeved shirt (81% often/always) and hat (73%) were most prevalent, while seeking shade (12%) and applying sunscreen (36%) were least prevalent. Workers had higher sun protection scores at work than leisure. Hours spent outdoors was a strong determinant for both models. Additional leisure model predictors were eye colour, sex, skin type, and job group.

Conclusions: 

Differences in behaviours across settings were observed and should be considered when developing solar UVR exposure reduction initiatives.

Non-melanoma skin cancer is the most commonly diagnosed malignancy in countries with large proportions of light-skinned people.¹ Non-melanoma skin cancer comprises multiple types of skin cancers but is dominated by and is often used to refer specifically to keratinocyte carcinoma (KC), which includes squamous cell and basal cell carcinomas. The impacts of KC are substantial due to its high incidence; in 2016, an estimated 1.5 million cases of KC contributed to 53,000 deaths and 1.0 million disability adjusted life years across the world.² In Canada, KC is the most commonly diagnosed cancer,³ and its incidence is increasing.⁴

Solar ultraviolet radiation (UVR) is a known human carcinogen⁵ and the main cause of skin cancer.⁶ Outdoor workers are at particular risk of KC, as well as solar keratosis, cataracts, ocular melanoma, pterygium, and macular degeneration, due to long hours spent outdoors.⁷ Chronic sun exposure has also been associated with some melanomas (eg, lentigo maligna melanoma).⁸ Despite the well-known associations between solar UVR exposure and these adverse health outcomes, the adoption of sun protective behaviours has been variable and often low among outdoor workers across nations and industries.^9,10 For example, among agricultural and construction workers in the United States, sunscreen use varied from 15% to 20%, shade seeking ranged from 15% to 29%, and the prevalence of wearing protective clothing was much higher among agricultural workers (70%) compared to construction workers (approximately 50%).¹¹ On the other hand, for construction workers in Vancouver, Canada, only 8% often or always sought shade, 29% often or always wore sunscreen, 79% often or always wore hats, and 82% often or always wore sleeved shirts.¹² In Australia, despite 95% of workers using some form of sun protection, only 8.7% were considered fully protected.¹³

Inadequate sun protection among outdoor workers in Canada has significant public health and economic impacts. An estimated 1.5 million workers (9% of the workforce) are exposed to solar UVR,¹⁴ contributing to an estimated 4600 diagnosed KC each year.¹⁵ The total combined direct and indirect costs (eg, healthcare, out-of-pocket, and productivity costs) of these KC cases amount to approximately $28.9 million each year.¹⁶

An increased understanding of the prevalence and determinants of sun protection behaviours can lead to better-informed policies on solar UVR exposure reduction efforts in the workplace. This study builds on previous efforts in Canada¹⁷ to assess the prevalence and determinants of sun protective behaviours among outdoors workers at work and at leisure in Alberta, Canada's sunniest province, which receives an average of 312 sunny days each year.¹⁸

METHODS

Study Design

This cross-sectional study primarily took place in the province of Alberta, Canada, in the summer of 2019. Workers completed questionnaires to capture information on their demographics, jobs, skin cancer risk factors, and sun protection behaviours during work and during leisure time.

Study Sample

Participants were recruited by management or health and safety teams from building trade unions and employers in Alberta. Companies from other provinces (ie, British Columbia and Saskatchewan) who learned about the study via word of mouth were also included. In total, 13 companies were invited and participated in the study. The total number of workers invited to participate in the study is not known since the managers and health and safety representatives were responsible for identifying and inviting their outdoor workers. However, general reports from our contacts indicated that workers were happy to participate and that refusals were low. Upon company consent, supervisors identified and invited all outdoor workers (ie, those who regularly spend at least 2 h outdoors per workday) to participate. Workers of all ages were eligible to participate; however, signed parental or guardian consent was required for workers who were 17 years of age or younger. The study received ethical approval from the Health Research Ethics Board of Alberta—Cancer Committee (certificate HREBA.CC-18-0615).

Data Collection and Analytic Variables

Data were collected using self-completed questionnaires. The questions were based on those used in the study of construction workers in Vancouver,¹² which were adapted from a standardized set developed to assess sun protection behaviours.¹⁹ The questionnaire included four sections to collect information on workers’ demographics, skin cancer risk factors, jobs, and sun habits. Demographic variables included: sex, date of birth, racial/ethnic background, and level of education (high school or less; some college; or, completed college). See the Supplemental Content for a copy of the questionnaire, https://links.lww.com/JOM/A868.

Skin cancer risk factor variables assessed were hair colour (red or blonde; dark blonde or light brown; or, dark brown or black), eye colour (blue, green, or grey; hazel or light brown; or, dark brown or black), childhood sunburn history (yes or no to a painful or blistering sunburn), and skin colour. Workers’ skin types were categorized by the researchers according to the Fitzpatrick skin types²⁰; light to fair skin types were classified as Fitzpatrick types I and II, white to olive skin types were classified as Fitzpatrick III and IV, while olive to dark brown skin types were classified as Fitzpatrick V-VI. Information on workers and their family history of skin cancer history was collected (yes or no, and type of skin cancer).

Job variables included current job title or trade, main job tasks, length of time at current job or trade, and current job schedule (<30, 30 to 40, 40 to 50, 50+ h). In addition, for the longest held job, information on tasks completed, duration in years, and number of hours spent outdoors in summer between 10:00 and 16:00 was collected. An occupational hygienist categorized workers’ jobs into three broad groups based on their self-reported job titles: (1) trades (eg, carpenters, bricklayers, concrete labourers, foremen, electricians, plumbers); (2) recreational, landscaping, or professional services (eg, coaches, dog walkers, groundskeepers, park labourers, mail delivery agents, fish and wildlife specialists); and, (3) security (eg, campus security, equipment technicians, parking enforcement).

Finally, self-reported sun-related variables included the average number of hours spent outside per day while at work and during leisure, number of sunburns during the last summer, and frequency of time spent in the sun to get a tan (never, rarely, sometimes, often, always). Sun protection behaviours (wearing sunscreen, a sleeved shirt, a hat, sunglasses, and shade seeking) were assessed using a five-point Likert scale (see Table 1), separately for work and leisure. Similar to the Vancouver study,¹² leisure and work sun protection behaviours were translated to a five-point ordinal scale, from 0 (never using the behaviour) to 4 (always using the behaviour). A composite score (ranging from 0 to 4) was then calculated by averaging the scores of all sun protection behaviours, separately for work and leisure. This score captures how many of the sun protection behaviours were practiced, and to what extent, by each worker.

TABLE 1 - Sun Protection Behaviour Questions from the Outdoor Workers Project Questionnaire¹²

For the Following Questions, Think About What You Do When You Are Outside AT WORK During the Summer on a Warm Sunny Day∗
1. How often do you wear SUNSCREEN?
2. How often do you wear a SHIRT WITH SLEEVES that cover your shoulders?
3. How often do you wear a HAT?
4. How often do you stay in the SHADE or UNDER AN UMBRELLA?
5. How often do you wear SUNGLASSES?

^∗Answers were on a five-point scale: never, rarely, sometimes, often, or always.

Statistical Analysis

Basic descriptive statistics were calculated for all variables. Mean sun protection behaviours scores and standard errors were generated. For each covariate, bivariate analyses were conducted, separately for work and leisure, by fitting a simple linear regression model, using the behaviour score as the response and the covariate as the lone predictor. For age and job tenure, the reported data is the estimated coefficient in the linear regression with its standard error and P value. For all other covariates (the categorical variables), the reported data is the least-squares mean score for each level of the variable (and associated standard error), and P value.

The McNemar test was used to compare differences in the proportion of workers who often or always practiced a sun protection behaviour, by setting (ie, at work or during leisure). Paired t-tests were used to compare hours spent outdoors and composite sun protection behaviour scores, by setting. Spearman correlation coefficient comparing the overall sun protection score at work to leisure was calculated.

The determinants of the composite sun protection behaviour scores and individual sun protection behaviour results were modelled separately for work and leisure, using multiple linear regression (SAS PROC GLM). All variables summarized in Table 2 were offered to the model. A manual backwards stepwise regression method was used to restrict variables in the final model; variables with P > 0.20 were sequentially removed. For each reported model, only participants with non-missing values for the response and for all significant predictors were included in the modelling population. In the final, most parsimonious model after the backwards selection process, participants with non-missing values for the final predictor variables were considered. Because the final model included fewer predictor variables, it had a greater sample size than the full model. Results are presented as the least-squares means (using Tukey's multiple-comparisons adjustment), which estimate the marginal means over a balanced population, to facilitate interpretation.

TABLE 2 - Descriptive Statistics and Sun Protection Behaviours Scores at Work and Leisure (N = 192)

		Sun Protection Scores (SE)
Variable	N (%)∗	Work	Leisure
Sex	189 (98)	P = 0.52	P = 0.045
Male	142 (75)	2.46 (0.05)	2.27 (0.05)
Female	47 (25)	2.40 (0.08)	2.06 (0.09)
Age (y)	183 (95)	P = 0.82	P = 0.65
Mean (range)	38.07 (14–70)	−0.00 (0.00)§	−0.00 (0.00)§
Race	189 (98)	P = 0.22	P = 0.67
Caucasian or White	141 (75)	2.48 (0.05)	2.23 (0.05)
Other	48 (25)	2.36 (0.08)	2.19 (0.09)
Skin type†	180 (94)	P = 0.10	P = 0.07
Light to fair (I–II)	55 (32)	2.54 (0.08)	2.37 (0.08)
White to olive (III–IV)	74 (41)	2.46 (0.07)	2.19 (0.07)
Olive to dark brown (V–VI)	51 (28)	2.31 (0.08)	2.09 (0.09)
Hair colour	191 (99)	P = 0.63	P = 0.92
Red or blonde	35 (18)	2.52 (0.10)	2.26 (0.10)
Dark blonde or light brown	65 (34)	2.44 (0.07)	2.21 (0.08)
Dark brown or black	91 (48)	2.41 (0.06)	2.22 (0.06)
Eye colour	191 (99)	P = 0.24	P = 0.14
Blue, green, or grey	95 (50)	2.51 (0.06)	2.29 (0.06)
Hazel or light brown	32 (17)	2.35 (0.10)	2.04 (0.11)
Dark brown or black	64 (34)	2.39 (0.07)	2.22 (0.08)
Family history of skin cancer	190 (99)	P = 0.26	P = 0.76
Yes	25 (13)	2.56 (0.11)	2.26 (0.12)
No or don’t know	165 (87)	2.42 (0.04)	2.22 (0.05)
More than 1 severe childhood sunburn?	188 (98)	P = 0.26	P = 0.69
Yes	106 (56)	2.48 (0.06)	2.24 (0.06)
No	82 (44)	2.39 (0.06)	2.20 (0.07)
Number of sunburns the previous summer	176 (92)	P = 0.90	P = 0.79
Zero	59 (34)	2.44 (0.07)	2.23 (0.08)
One or more	117 (66)	2.42 (0.05)	2.21 (0.06)
Education	183 (95)	P = 0.43	P = 0.62
Completed college	68 (37)	2.45 (0.07)	2.27 (0.08)
Some college	64 (35)	2.51 (0.07)	2.22 (0.08)
High school or less	51 (28)	2.37 (0.08)	2.16 (0.09)
Job location	192 (100)	P = 0.60	P = 0.99
Calgary	83 (43)	2.40 (0.06)	2.23 (0.07)
Edmonton	86 (44)	2.47 (0.06)	2.21 (0.07)
Other	23 (12)	2.51 (0.12)	2.23 (0.13)
Job group	192 (100)	P = 0.98	P = 0.57
Trade worker	88 (46)	2.43 (0.06)	2.20 (0.07)
Recreational, landscaping, or professional services	84 (44)	2.44 (0.06)	2.27 (0.07)
Security worker	20 (10)	2.49 (0.13)	2.13 (0.14)
Job tenure (y)	189 (98)	P = 0.86	P = 0.55
Mean (range)	8.5 (0–45)	−0.00 (0.01)§	0.00 (0.01)§
Hours spent outside on workdays‡	188 (98)	P = 0.01	NA
0–2 h	21 (11)	2.45 (0.12)	—
3–5 h	86 (46)	2.31 (0.06)	—
6 or more hours	81 (43)	2.57 (0.06)	—
Hours spent outside at leisure‡	188 (98)	NA	P = 0.08
0–2 h	64 (34)	—	2.10 (0.08)
3–5 h	102 (54)	—	2.31 (0.06)
6 or more hours	22 (12)	—	2.19 (0.13)
Frequency of sun tanning	184 (96)	P = 0.36	P = 0.10
Never or rarely	117 (64)	2.47 (0.05)	2.29 (0.06)
Sometimes	45 (25)	2.37 (0.08)	2.16 (0.09)
Often or always	22 (12)	2.33 (0.12)	2.00 (0.13)

^∗Number of subjects with complete data.

^†Skin type categories and values correspond to the Fitzpatrick scale.²⁰

^‡Between 10:00 am and 4:00 pm (ie, max value = 6).

^§Mean change per year.

RESULTS

In total, 192 participants were recruited for the study, of whom 140 (73%) had complete information on all variables. Most participants identified as male (75%) and Caucasian (75%), with a mean age of 38 years (Table 2). Only 30% of workers reported light to fair skin types (Fitzpatrick skin types I and II), but 56% reported experiencing at least one painful or blistering childhood sunburn and 66% reported one or more sunburns in the previous summer.

The majority of workers were located in Calgary (43%) and Edmonton (45%). Workers were most often employed in the trades (46%), but also in recreational, landscaping, or professional services (44%) and security work (10%). Job tenure was high (8.5 y) and 43% of workers spent six hours or more outdoors on workdays. During leisure time, 12% of workers spent six hour or more outdoors, per day.

Results from the bivariate analyses are also presented in Table 2. Sun protection scores differed by sex when at leisure (P < 0.05), with men engaging in more sun protection behaviours than women. At work, overall sun protection score varied by time spent outdoors (P = 0.01).

Sun-Related Behaviours at Work and Leisure

Sun protection behaviours used by participants at work and leisure are summarized in Table 3. The majority of workers often or always wore shirts with sleeves (81%), a hat (73%), or sunglasses (59%) at work. Relatively fewer workers wore sunscreen (36%) and sought shade (12%).

TABLE 3 - Distribution of Sun Protection Behaviours Among Outdoor Workers (N = 192)

	Never (0), Rarely (1), or Sometimes (2) (%)		Often (3) or Always (4) (%)
Behaviour	Work	Leisure	Work	Leisure	P Value∗
Wear sunscreen	121 (64)	140 (74)	69 (36)	50 (26)	<0.001
Wear a shirt with sleeves	36 (19)	80 (43)	151 (81)	107 (57)	<0.001
Wear a hat	52 (27)	101 (53)	138 (73)	89 (47)	<0.001
Stay in the shade or under umbrella	166 (88)	138 (73)	23 (12)	51 (27)	<0.001
Wear sunglasses	79 (41)	76 (40)	112 (59)	115 (60)	0.68
Spend time in the sun to tan	—	162 (88)	—	22 (12)	—

^∗P values of the McNemar analyses assess the significance of differences in marginal frequencies of work- and leisure-time sun protection behaviours among the dichotomized variables (ie, Never, rarely, or sometimes; and, often or always).

Use of sunglasses at work and leisure was similar (59% and 60%, respectively; P = 0.68). However, participants were more likely to protect themselves from the sun at work than leisure by wearing sunscreen (36% vs 16%, respectively, P < 0.001), a shirt with sleeves (81% vs 57%, P < 0.001), and a hat (73% vs 47%, P < 0.001). On the other hand, participants were more likely to seek shade during leisure time (27% vs 12% at work, P < 0.001). The majority of workers did not spend time outdoors to obtain a tan (88%) (Table 3). The overall sun protection scores at work and at leisure were moderately correlated (Spearman correlation coefficient = 0.64).

Differences in the composite sun protection behaviour scores and hours spent outside by setting are summarized in Table 4. Participants were more likely to spend time outdoors between 10:00 and 16:00 at work than during leisure time (4.8 compared to 3.3 h, P < 0.001). In addition, participants were more likely to adopt sun protection behaviours at work (score: 2.44, compared to 2.22 during leisure time, P < 0.001).

TABLE 4 - Hours Spent Outside Per Day and Sun Protection Behaviour Scores Among Outdoor Workers, During Work and Leisure

Variable	Setting	Mean	Median	Range	P Value†
Hours spent outside∗ at…	Work	4.76	5.00	≤1–6	<0.001
	Leisure	3.28	3.00	≤1–6
Sun protection behaviour score (range 0–4) at…	Work	2.44	2.40	0.80–3.80	<0.001
	Leisure	2.22	2.20	0.00–3.60

^∗Between 10:00 am and 4:00 pm (ie, max value = 6).

^†P values are for the difference between work and leisure, first hours spent outdoors (calculated using the Wilcoxon signed-rank test), and then the sun protection behaviour score (calculated using the paired t test).

Determinants of Sun Protection Behaviour Scores

Modelling results for the determinants of sun protection behaviour scores at work and leisure are presented in Table 5. Only three variables remained in the final work model. Workers that spent 6 or more hours outdoors had higher sun protection behaviour scores (2.58) than those who spent 0 to 2 and 3 to 5 hours outdoors (2.49 and 2.32, respectively; P = 0.01). In addition, workers with a family history of skin cancer had higher sun protection behaviour scores (2.56 vs 2.37), as did workers that never or rarely sun tanned (2.57 vs 2.41), although neither of these results were statistically significant.

TABLE 5 - Model Results: Determinants of Sun Protection Behaviour Scores^∗,†

	Determinants of Work Protection Behaviour Score||		Determinants of Leisure Protection Behaviour Score¶
Effect	Mean Score (SE)	P Value	Mean Score (SE)	P Value
Sex
Male	—	—	2.16 (0.09)	0.04
Female	—		1.92 (0.12)
Skin type‡
Light to fair	—	—	2.23 (0.13)	0.05
White to olive	—		1.96 (0.11)
Olive to dark brown	—		1.93 (0.11)
Eye colour
Blue, green, or grey	—	—	2.08 (0.10)	0.01
Hazel or light brown	—		1.81 (0.13)
Dark brown or black	—		2.23 (0.12)
Numbers of sunburns the previous summer
Zero	—	—	2.11 (0.11)	0.19
One or more	—		1.97 (0.09)
Family history of skin cancer
Yes	2.56 (0.12)	0.11	—	—
No or don’t know	2.37 (0.06)		—
Job group
Trade worker	—	—	1.95 (0.11)	0.04
Recreational, landscaping, or professional services	—		2.22 (0.09)
Security worker	—		1.96 (0.15)
Number of hours spent outdoors§
0–2 h	2.49 (0.14)	0.01	1.86 (0.10)	0.03
3–5 h	2.32 (0.08)		2.13 (0.09)
6 or more hours	2.58 (0.08)		2.13 (0.18)
Frequency of sun tanning
Never or rarely	2.57 (0.08)	0.18	2.17 (0.09)	0.19
Sometimes	2.41 (0.09)		2.05 (0.12)
Often or always	2.41 (0.14)		1.91 (0.15)

^∗Variables initially offered to the work and leisure models include those noted plus age, race, hair colour, childhood sunburn, education, job location, and job tenure. None of these variables remained in either of the final models.

^†SE, standard error.

^‡Skin type: Light to fair = Fitzpatrick Type I or II; white to olive = Fitzpatrick Type III or IV; olive to dark brown = Fitzpatrick Type V or VI.

^§Refers to hours spent outdoors at work when modelling work protection behaviour scores, and hours spent outdoors at leisure when modelling leisure protection behaviour scores.

^||The final, parsimonious model determined by backward selection had a sample size of N = 183.

^¶The final, parsimonious model determined by backward selection had a sample size of N = 158.

In the leisure model, sex, skin type, eye colour, job group, and number of hours spent outdoors were significant determinants of sun protection behaviour scores. Male workers had higher overall mean behaviour scores than women (2.16 and 1.92, respectively). Sex was also a significant determinant in all the individual sun protection behaviour models, except for shade use (results not shown); male workers were more likely to wear shirts (2.65 vs 1.87, P < 0.001), wear a hat (2.35 vs 1.36, P < 0.001), and sunglasses (2.62 vs 2.11, P = 0.02). However, female workers were more likely to use sunscreen (2.28 vs 1.55, P < 0.001).

Participants had higher leisure sun protection scores if they had light to fair skin (2.23 vs 1.96, P = 0.05), dark brown or black eyes (2.23 vs 2.08 for blue, green, or grey eyes and 1.81 for light brown eyes, P = 0.01), and spent greater than three hours outdoors between 10:00 and 16:00 during their leisure time (2.13 vs 1.86, P = 0.03). Participants whose jobs were related to recreation, landscaping, or professional services had higher sun protection scores during their leisure time (2.22) compared to the other job groups (1.95, P = 0.04). In addition, participants with zero sun burns in the previous summer had higher mean scores (2.11 vs 1.97), as did those who never or rarely tanned (2.17 vs 2.05 for sometimes and 1.91 for often or always), but these results were not significant.

DISCUSSION

This study builds on previous efforts to assess the prevalence and determinants of sun protection behaviours among outdoor workers in Canada. The prevalence of certain sun protection behaviours in this study was relatively high. Findings were similar to a study of construction workers conducted in Vancouver, Canada, in 2013 where 79% to 82% of workers wore a shirt with sleeves or a hat (compared to 73% to 81% in this study), 74% wore sunglasses, 29% wore sunscreen, and 8% sought shade (vs 59%, 36%, and 12% in this study, respectively).¹² The prevalence is much higher than the results of a national, Canadian sun survey conducted in 2006, which found that 54% to 58% of outdoor workers wore sunglasses, protective clothing, or hats,²¹ as well as findings from the United States, Denmark, and Germany.^9,11,22,23 Increased sun protection use among workers in Canada may be due to increased public awareness and prevention campaigns over the past years. Furthermore, variations in sun protection behaviours at work have been observed by occupation, sex, and jurisdiction^9,10 and may further explain observed differences, in addition to differences in the wording of questions across surveys.

Overall, participants were more likely to protect themselves at work than leisure, consistent with the Vancouver study.¹² Similarly, past studies have found that individuals from workplaces with solar UVR safety training programs use fewer sun safety behaviours when away from work.^24–26 Our findings can be explained in part by the fact that study participants spent more time outdoors at work, which is an important determinant of sun protection behaviour score in both models. Decreased sun protection during leisure could be due to a lower perceived risk since relatively less time is spent outdoors during leisure time than at work.²⁷ The overall sun protection scores for work and leisure were moderately correlated, indicating that individuals who protect themselves at work are more likely to protect themselves during leisure. However, a study of outdoor workers in Denmark found the opposite effect, with workers using more sun protective behaviours at leisure.²² This difference may be partly due to differences in occupations included, the definition of outdoor workers, workplace training, and workplace personal protective equipment requirements.

At work, participants were more likely to wear protective clothing, hats, and sunscreen, less likely to seek shade, and equally likely to use sunglasses than at leisure. The equal use of sunglasses was unsurprising given that sunglasses have been found to be primarily worn to control glare and not to reduce ocular solar UVR exposure.^28,29 The differential use of shade at work and leisure suggests that if more shade structures were available at work, uptake could increase. Indeed, shade structures specifically designed for workers to protect themselves from the sun while they were working were not observed during site visits. However, the introduction of shade structures must be paired with additional changes at the workplace; providing shade structures, where feasible, will not necessarily lead to change, as there are many additional barriers to uptake (eg, workers’ knowledge or attitudes, workplace culture).³⁰ For example, a study of sun protection behaviours in Australia found that while half of workplaces provided shade structures, only 20% of workers usually or always sought shade during the workday.³⁰

The only statistically significant determinant in the work model was hours spent outside; longer hours outdoors contributed to the most protective behaviours, consistent with previous findings.³¹ In the leisure model, sex was a significant determinant of sun protection behaviours, with male participants having a higher sun protection score than the female counterparts. These findings were unexpected given that women have typically been associated with better sun protection behaviours.^21,22,31 However, men have been shown to be more likely to wear hats at work and leisure,^21,25 and we did see this in our study, with hat wearing having the largest difference in score by sex. The analysis of individual sun protection behaviours also showed that female participants still had higher sunscreen use compared to the male counterparts, in agreement with previous findings.

Other significant determinants in the leisure model include skin type, in which lighter skin types are associated with better sun protection,^31,32 and hours spent outdoors, with longer hours associated with higher scores. Job group was a significant determinant of protection behaviours at leisure, with recreational, landscaping, or professional services workers having overall higher scores compared to trade and security workers. We were unable to ascertain the reason for this finding with the data we collected, but future studies could consider asking outdoor workers about the sun safety culture of their workplaces to see if that may explain differential use by job title. Eye colour was also a significant determinant of protection behaviours at leisure, with dark brown/black-eyed workers having the highest sun protection behaviours cores. These findings were surprising because lighter eye colours are a well-known risk factor for KC.³³

Strengths and Limitations

This study used self-completed questionnaires and is thus susceptible to response bias, and in particular, social desirability and/or recall bias. Individuals have been shown to incorrectly assess their skin type;^22,34 similarly, we found that the percentage of people reporting light/fair skin types was incongruent with the percentage of people reporting sun burn in the previous summer. However, the potential response bias for sun protection behaviours is expected to be minimal, as studies have shown that these self-reports are relatively accurate in occupational and leisure settings.^35–37 In addition, selection bias, in which companies with sun safety policies may be more likely to participate, is possible. The potential impact of this bias is unknown because the sun safety policies of the participating workplaces were not assessed. However, where possible, selection bias was minimized; for example, employers were asked to invite all individuals working outdoors for two or more hours to participate. Another limitation is the lack of information collected on workplace training or culture, both of which could have influenced how sun protection behaviours were employed at work.

The use of composite scores may lead to some skewed results. Furthermore, the composite score assigns the same weight to each sun protection behaviour, suggesting that each is equally effective at protecting the individual, which is not the case. However, in our results, relative scores were mostly consistent across the different sun protection behaviours for the various variables, and thus the potential for skewed results is likely low.

This study builds on the existing Canadian literature to evaluate sun protection behaviours among outdoor workers. The study has a large sample size and includes a more heterogenous sample compared to previous Canadian studies, and is thus more generalizable.¹² In the future, the collection of more detailed data on the quality of the sun protection behaviours (eg, the adequacy of protection provided by hats, sunscreen re-application), as well as sun safety policies that are already in place in workplaces and/or legislated by the jurisdiction, would be beneficial.

CONCLUSION

Overall, the outdoor workers in this study, who were primarily employed in the trades and recreational, landscaping, or professional services, were more likely to practice sun protection behaviours at work than at leisure. Participants were more likely to wear sunscreen, hats, and long-sleeved shirts at work, but were more likely to seek shade at leisure. Number of hours spent outside was an important predictor of the sun protection score at work and at leisure. Being male, having fair skin and dark brown/black eyes, and having a job in recreation, landscaping, or professional services were also important predictors of better sun protection by workers at leisure. Greater sun protection at work indicates that more emphasis should be placed on sun protection at leisure in sun safety messaging and programs that target outdoor workers, as solar UVR exposure during leisure may be significant. In addition, sun safety programs may benefit from targeting participants who have lower sun protection behaviour scores at work and leisure. While the prevalence of sun protection behaviours in this sample of workers is higher than observed in the past,²¹ adoption of some measures remain low. To be effective, multiple sun protective behaviours must be used. Sun safety among outdoor workers should continue to be emphasized by employers, for example through implementing educational programs, mandating the use of specific sun protective behaviours, and providing sunscreen and shade structures.