Sunscreen use among recreational cyclists: How intentions predict reported behavior

Abstract

A nationwide survey measured 927 recreational cyclists’ cognitions and perceptions about skin cancer risks, along with sun protection practices and predictors of sunscreen use while cycling. Multiple regressions evaluated associations between perceived costs, rewards, photoaging, self-efficacy and sunscreen use, and potential moderators of the associations between intentions and sunscreen use were examined. Results suggest that when cyclists see the advantages of using sunscreen, are worried about photoaging, and feel efficacious, they have stronger intentions to apply sunscreen before riding. Intentions to use sunscreen while cycling predict reported use of sunscreen, particularly when cyclists perceive sunscreen application as easy and viable.

Keywords

cancer health behavior lifestyle protective factors risk reduction self-efficacy social cognitions

Skin cancer is the most common form of cancer in the United States (American Cancer Society, 2009). Although skin cancers are ‘the most easily and successfully treated human cancers’ (Amir et al., 2000), skin cancer accounts for nearly 50% of new cancer cases, and it is estimated that more than one million new cases will be diagnosed in the USA each year (American Academy of Dermatology, 2010; Rogers et al., 2006). One major factor that plays a role in the development of skin cancer is excessive exposure to ultraviolet radiation (UVR) placing outdoor enthusiasts at particular risk for extensive sun exposure (Liffrig, 2001). Many individuals appear to rationalize sun exposure, claiming it is an incidental but necessary effect of their lifestyle (Paul et al., 2011). Unfortunately, little effort is being made to target outdoor recreational groups regarding the practice of sun safety with the exception of sunbathing habits (Newman et al., 1996).

Recently, attention has been turned to recreational athletes who often spend many hours and weeks training and competing outdoors. Prior epidemiological research has found that outdoor recreational activities such as sun-bathing, water-skiing, mountain climbing and outdoor endurance sports increase the risk for skin cancer (Moerhle, 2008). Recreational cyclists appear to be an understudied population that may be at particular risk. Recent studies have tested levels of UVR by monitoring cyclists with dosimeters during their rides. Results indicate that cyclists receive significant doses of UVR on various sites of the body, often exceeding the limits recommended by the International Commission for Non-ionizing Radiation Protection (Kimlin et al., 2006). Therefore, sun protection behaviors should be a priority while engaging in this activity.

Use of sunscreen is a major cancer-preventive behavior for those often exposed to the sun. Recreational cyclists are a unique group who spend prolonged time in the sun but who otherwise tend to be motivated for improved health and fitness. One study found that cyclists participating in the Tour de Suisse were exposed to more than eight times the minimal dose of UV needed to produce sunburn. The personal exposure levels determined during these races were 30 times higher than recommended levels (Moehrle et al., 2000). Many cyclists appear to be unfamiliar with the risks of sun exposure. Protective means such as wearing adequate clothing, training and competing during certain hours of the day, and applying sunscreen still need to be propagated in the cycling community.

The present study sought to analyze cyclists’ perceptions about skin cancer risks and the actions they currently take to protect themselves from the sun while riding. Prior research has examined social cognitive predictors of sun protection intentions and behaviors with regard to tanning, and several other forms of outdoor recreation (Myers and Horswill, 2006). These researchers suggest that one’s intentions to perform a behavior, along with perceived behavioral control (an evaluation of whether or not the individual can perform the behavior based on internal and external factors), can substantially predict the behavioral outcome. That work focused on the theory of planned behavior, and emphasized intentions as predictors of behavior over time (Ajzen, 1991). Other work has emphasized components from the health belief model such as perceived rewards and costs of using sun protection. Perceived susceptibility to photoaging is another factor that has been shown to positively influence one’s intentions to use sunscreen, also lowering reported sunbathing behavior (Mahler et al., 2003). Finally, self-efficacy has been shown to play an important role in sunscreen use as well (Myers and Horswill, 2006). In the current study, we tested several of these social cognitive factors as predictors of intentions to use sunscreen and reported sunscreen use in a national sample of recreational cyclists. As found in prior research on sun protection in the general population, we hypothesized that social cognitive predictors such as perceived rewards of being tan, perceived costs of using sunscreen, perceived susceptibility to photoaging, and self-efficacy for using sunscreen would predict intentions to use sunscreen. Intentions would, in turn, predict reported sunscreen use. Self-efficacy has been shown to moderate the association between intentions and behavior (Jones et al., 2001). We reasoned that a similar role might emerge in our data. Thus, we explored whether each social cognitive predictor moderated the intention–behavior association in our sample as well.

Methods

Participants

Participants were 927 (696 male) recreational cyclists from all parts of the USA who participated in an anonymous online survey in exchange for entry into a lottery drawing for a gift certificate to a popular online bicycle shop. Participants were recruited from various recreational cycling listservs, as well as advertisements that were placed in several widely circulated newsletters intended for cyclists training for organized rides and charity events. During recruitment, recreational cyclists were invited to participate in a 10–15 minute online survey investigating cyclists’ perceptions about sun exposure risks and current sun protection strategies. Cyclists were required to be at least 18 years of age and actively involved in recreational cycling. Upon completing the survey, participants had the option of entering their name in a drawing for one of several $100 gift cards to Performance Bike. Participants were asked to complete the study only once and the data were scanned for duplicate entries.

Participants emerged from nearly every state in the USA. The sample was largely Caucasian (89.3%), with some Hispanic (5.6%), Asian (2.5%), and African American participants (.7%), and with 1.7% choosing ‘other’. Age ranged from 18–77 years old (Mean = 48, SD = 11). A total of 16.5% of participants reported a personal history of skin cancer, and 43.7% of participants reported a familial history of the disease. On average, participants reported spending 6.54 hours riding their bicycle during the previous week (SD = 5.97).

Procedure

The survey was conducted during the late spring and summer seasons of 2009, so that all parts of the USA would be in prime outdoor recreational season. Participants provided demographic information and indicated the number of hours during the previous week they had ridden their bicycle. They were also asked about their Fitzpatrick skin-type (Fitzpatrick, 1988), personal and family history of skin cancer, various sun-protective behaviors (e.g. wearing protective clothing), how often they use sunscreen while on their bike, and the preferred SPF. Participants rated several reasons for why they do or do not use sunscreen on a regular basis. Predictors surveyed were perceived rewards of not using sunscreen (e.g. getting a nice tan), perceived costs of using sunscreen (e.g. burdensome), self-efficacy with regard to sunscreen use, and perceived susceptibility to photoaging. Criteria assessed were intentions to use sunscreen and reported sunscreen use.

Measures

Skin type

Skin type was assessed using the methods illustrated by Fitzpatrick (1988), which consists of asking respondents about their tendencies to tan and burn. For example, a response of 1 = burn easily, and the burn would not turn into a tan; 2 = burn easily, and then the burn might turn a light tan; 3 = burn moderately, and then turn a light tan; 4 = burn minimally, and then turn a moderate brown tan; 5 = probably not burn, but rather develop a dark brown tan; and 6 = not burn, I am dark skinned naturally. This scale is considered a valid indicator of one’s tendency to burn.

Social cognitive predictors

Three social cognitive predictors were assessed using previous measures adapted from Mahler and colleagues (2007). Six items measured perceived rewards of sunbathing/tanning (Cronbach’s alpha = .87; e.g. ‘I think I look healthier when I have a tan’), nine items measured costs of using sunscreen (Cronbach’s alpha = .86; e.g. ‘Using sunscreen regularly is just too much trouble’), and seven items measured perceived susceptibility to photoaging (Cronbach’s alpha = .75; e.g. ‘I don’t need to worry about getting wrinkles or age spots until I am much older’). Participants rated each item on a five-point scale from 1 (strongly disagree) to 5 (strongly agree) and scores were averaged for each predictor.

Self-efficacy

Self-efficacy for regular sunscreen use was assessed with six items (Mahler et al., 2003). Participants indicated on a scale from 1 (extremely difficult) to 10 (extremely easy) how easy or difficult it would be to use sunscreen despite various obstacles (e.g. ‘How easy or difficult would it be for you to use sunscreen while cycling even if you have a base tan?’). Cronbach’s alpha was .95.

Intentions

Intentions to use sunscreen were assessed with items adapted from Mahler and colleagues (2003). Participants rated seven items on a scale from 1 (strongly disagree) to 5 (strongly agree) on the extent to which they intend to use sunscreen (Cronbach’s alpha = .92; e.g. ‘I plan to always use sunscreen when I am outside on my bike’). Higher scores indicated greater intentions to use sunscreen.

Sunscreen use

Sunscreen use was measured by asking participants to indicate how frequently they use sunscreen as a percentage of time when riding their bicycle outside (0–100%). For presentation in the analysis, the scale was converted from a percentage to a number ranging from 0 (0% of the time) to 10 (100% of the time). Separate items assessed sunscreen use on the face and on the body. The two items were highly correlated (r = .77) and were averaged to form a sunscreen use index. Participants also indicated the reasons they use sunscreen (e.g. to avoid getting sunburned, to avoid getting wrinkles) and do not use sunscreen (e.g. it is too much trouble, it is too messy, it causes me to break out).

Skin cancer history

Participants were asked if they had ever been diagnosed with any form of skin cancer (e.g. squamous cell, melanoma). They were also asked to report the number of people in their family (both immediate and extended) who have had any type of skin cancer.

Data preparation and statistical analysis

Statistical analyses were performed using SAS (SAS Institute, Cary, NC). All available data were used for descriptive statistics and zero-order correlations. Twelve participants had missing data on one or more variables in the multiple regression analyses and thus were not included in those analyses. We computed descriptive statistics to describe sun protection practices and attitudes of the study cohort. Multiple regressions evaluated associations between perceived costs, rewards, photoaging, and self-efficacy and sunscreen use. We then explored potential moderators of the associations between intentions and sunscreen use. The hypothesized associations between each predictor and sunscreen use, and the mediating role of intentions to use sunscreen, were tested using multiple regression models in accordance with the procedures for mediation analysis (Kenny et al., 1998). First, the mediator (intentions) was regressed on the four predictors (rewards, costs, photoaging, and self- efficacy) in a simultaneous multiple regression analysis. Next, the criterion (sunscreen use) was regressed on the four predictors. Finally, the criterion was regressed onto the four predictors along with the mediator.

The statistical significance of intentions as a mediator of each path was tested with the modified Sobel test (Kenny et al., 1998; Sobel, 1982), which evaluates the degree of decline in the direct path (i.e. costs predicting sunscreen use) when the mediator is included in the regression (i.e. costs predicting sunscreen use controlling for intentions).

Results

Participant characteristics

Skin type distribution was consistent with previous results from a sample population (I = 8.71%; II = 21.72%; III = 25.38%; IV = 30.75%; V = 13.44%), where type I represents a lighter skin type that is likely to always burn while out in the sun (Fitzpatrick, 1988). Most participants reported occasionally using sunscreen while cycling (n = 703 [75%]) either to avoid getting burned (n = 598 [85%]) or to avoid increasing their risks for skin cancer (n = 548 [78%]). Commonly reported reasons for not using sunscreen while cycling included forgetfulness (n = 371 [40%]), too much trouble or time to apply (n = 176 [19%]), too greasy and messy (n = 181 [20%]), and causes breakouts in the skin (n = 53 [6%]). Only 6% of cyclists (n = 57) reported wanting a tan.

Preliminary analyses

Table 1 provides means, standard deviations, and correlations between variables. Because of the large number of correlations, we set significance levels at p < .001. Intentions to use sunscreen and reported sunscreen use were highly correlated, as well as rewards and costs, photoaging, and self-efficacy. Having a family history of skin cancer was not significantly correlated with rewards of having a tan or perceived costs of using sunscreen, and having a personal history of skin cancer was not significantly correlated with photoaging.

Table 1.

Means, standard deviations, and zero-order correlations among observed variables

	M	SD	2	3	4	5	6
1. Rewards of being tan	2.63	.72	.33*	–.17*	–.42*	–.29*	–.30*
2. Costs of use	2.42	.91	–	–.34*	–.51*	–.54*	–.58*
3. Photoaging	3.79	.63		–	.33*	.39*	.40*
4. Self-efficacy	8.05	2.04			–	.54*	.56*
5. Intentions	3.21	.96				–	.74*
6. Sunscreen use	5.84	3.46					–

Note: Sunscreen use was converted from a percentage to a number ranging from zero to 10.

p < .001.

The hypothesized model

To test whether the social cognitive predictors of perceived rewards, costs, susceptibility to photoaging, and self-efficacy significantly predict intentions to use sunscreen while cycling, we conducted a simultaneous multiple regression. Perceived rewards, costs, susceptibility to photoaging and self-efficacy were entered as predictors with intentions to use sunscreen as the criterion. The overall model predicted 41% of the variance in intentions to use sunscreen, R² = .41, F(4, 915) = 158.49, p < .001. Rewards of having a tan did not significantly predict intentions to use sunscreen, whereas perceived costs, β = –.31, p < .001, susceptibility to photoaging, β = .18, p < .001, and feeling efficacious about using sunscreen, β = .31, p < .001, each uniquely predicted intentions.

A second regression was conducted with perceived rewards, costs, susceptibility to photoaging and self-efficacy as predictors and sunscreen use as the criterion. The overall model accounted for 46% of the variance in sunscreen use, R² = .46, F(4, 915) = 191.21, p < .001. Rewards of tanning did not significantly predict sunscreen use whereas perceived costs, β = –.35, p < .001, photoaging, β = .17, p < .001, and self-efficacy, β = .32, p < .001 each uniquely predicted sunscreen use.

A third regression was conducted with sunscreen use as the criterion and all five predictors, including intentions, entered simultaneously. The overall model predicted 62% of the variance in sunscreen use, R² = .62, F(5, 914) = 293.63, p < .001. In this model, stronger intentions to use sunscreen significantly predicted higher frequency of sunscreen use while cycling, controlling for perceived rewards and costs, susceptibility to photoaging, and self-efficacy for using sunscreen, β =.52, p < .001. To formally test intentions as a significant mediator between the predictors and the criterion, Sobel tests (Kenny et al., 1998; Sobel, 1982) were conducted on the difference between the direct path and the direct path when controlling for the mediator (intentions). Intentions significantly mediated each of the associations between costs, photoaging, and self-efficacy in predicting sunscreen use (all Sobel Zs > 6.00, ps < .001). This can be considered partial mediation in that the direct effects (in parentheses in Fig. 1) remained statistically significant, partly because of the large sample size and statistical power.

Figure 1.

Path model and coefficients illustrating mediation. Numbers in parentheses are standardized regression coefficients for the direct path predicting sunscreen use with intentions in the regression equation.

Tests of moderation

We also tested each predictor as a potential moderator of the association between intentions and sunscreen use with a series of moderated multiple regression analyses. For example, with regard to self-efficacy as a moderating variable, frequency of sunscreen use was regressed on intentions, self-efficacy, and their product simultaneously. The product term was significant indicating that self-efficacy moderated the association between intentions and sunscreen use, F(1, 917) = 36.41, partial r = .20, p < .001. Figure 2 provides reports of sunscreen use as a function of intentions and self-efficacy (± 1 SD above and below the mean) derived from the parameter estimates with the product term included (Cohen et al., 2003). As shown, stronger intentions to use sunscreen were associated with more frequent sunscreen use, especially when participants felt more efficacious in using sunscreen. Simple slopes of intentions predicting sunscreen use were positive and significant for those higher (pr = .62, p < .001) and lower in self-efficacy (pr = .43, p < .001), with the significant interaction indicating a steeper slope for those higher (relative to lower) in self-efficacy. The other social cognitive variables were not significant moderators of the intentions and sunscreen use association.

Figure 2.

Reported sunscreen use as a function of intentions and self-efficacy.

Discussion

Research has previously shown that outdoor recreational activities place one at greater risk for skin cancer (Kimlin et al., 2006), thus recent attention has been given to outdoor athletes. Recreational cyclists appear to be one understudied population who receive significant doses of UV radiation, consequently putting them at greater risk for sun exposure and skin cancer (Moehrle et al., 2000). Therefore, sun protection behaviors and sunscreen use should be a priority while engaging in this activity. The present study used data from a large and representative sample of cyclists to examine social cognitive predictors of intentions and reported sunscreen use while cycling.

The proposed model successfully predicted 65% of the variance in self-reported sunscreen use while cycling. As hypothesized, intentions to use sunscreen while cycling were predicted by the perceived costs of sunscreen use (i.e. too much trouble, messy), perceived susceptibility to photoaging (i.e. wrinkled or damaged skin), and feeling efficacious about using it. However, appearance-related factors or perceived rewards for being tan did not predict intentions or reported sunscreen use, suggesting that developing a tan for the sake of enhancing one’s appearance was not of concern for these cyclists. These findings may be a reflection of the sample in that approximately 75% of the participants were male. If the sample had comprised more women, appearance-related factors may have played a more significant role in the overall model. What can be concluded is that when recreational cyclists, particularly men, see the advantages for using sunscreen, are worried about damaging their skin, and feel efficacious, they have more intentions of applying sunscreen prior to their ride.

Results also suggest that one’s intentions to use sunscreen while cycling predict the percent of the time that one reports using sunscreen when cycling. Furthermore, the association between intentions to use sunscreen while cycling and percentage of sunscreen use becomes stronger when cyclists feel efficacious about being able to use sunscreen effectively. That is, when cyclists perceive sunscreen application as easy and viable, regardless of obstacles, they are more likely to use it. It is also evident that recreational cyclists’ intentions are less a function of rewards of being tan rather than perceived costs and burdens of applying sunscreen. Rewards may not predict intentions because cyclists may not consider sun exposure for tanning or appearance reasons while riding (as opposed to going to the beach). Intentions to use sunscreen appear to be more motivated by the potential costs of using (e.g. it is messy) and effects of photoaging (e.g. worrying about wrinkled skin), rather than rewards. Because a majority of the cyclists in the current study were middle age, these individuals may not think about the rewards of a tan appearance as being a primary concern for use. For middle aged and older adults, the effects of photoaging and appearance of wrinkles may be more salient and important.

Interventions to increase sunscreen use among recreational cyclists may want to target perceived costs and self-efficacy in particular, by making sunscreen seem more convenient to access and apply, and make it seem worth using even if one has a base tan, for example. Findings from Craciun et al. (2012) indicate that using volitional interventions primarily based on planning can facilitate sunscreen use. Sutton (1999) has argued that for interventions to change behavior, the amount of actual control over the behavior is important. If one perceives a task or behavior to be easy, it does not mean he or she has control over that behavior. For example, a cyclist may feel that sunscreen application is a relatively easy task, but then frequently fails to apply it prior to a ride due to uncontrollable or unforeseen circumstances (e.g. time, skin allergies). Thus, the actual control one has may ultimately influence the behavioral outcome. Future studies should investigate actual control as well as (perceived) self-efficacy.

Limitations and future directions

This study has several limitations. People who have had experiences with skin cancer may have been more likely to take an interest and complete the survey. This study also relied solely on cyclists’ self-reports of their sunscreen use. As with much skin cancer prevention research, attitudes, intentions, and behavior were assessed via self-reports that may be affected by social desirability biases and demand characteristics (O’Riodan et al., 2006).

Many cyclists reported not using sunscreen for the sole reason that it gets messy and stings their eyes when mixed with sweat. In addition, several cyclists stated not using sunscreen because of the time of day they trained or cloudy weather conditions. Another limitation was not accounting for the time of day, on average, that individuals cycled. Those who report early morning or late evening rides may use less sunscreen due to the fact that the sun is not as strong as peak sun hours (between 10 am and 4 pm). Specific variations in weather and geographic location were also not examined. Further, planning to use sunscreen when cycling is a relatively specific and concrete implementation intention (Gollwitzer, 1999), which tend to be strong predictors of behavior. Using less specific measures of intention may not yield the same findings. Finally, self-efficacy traditionally involves perceived capability rather than actual capability (Bandura, 1997). The items we employed included references to barriers that might prevent sunscreen use and thus our definition might be broader than other uses. We reasoned that in this particular domain, a lack of efficacy belief may largely involve these kinds of perceived barriers.

Future research can develop interventions to facilitate acceptance of sunscreen use among recreational cyclists, who are already health-conscious, but likely to be at greater risk for skin cancer. When individuals are health-motivated, as opposed to appearance-motivated, they may be more motivated by long-term benefits of sunscreen use (Ingledew et al., 2010). Skin cancer is preventable through reducing exposure to UVR; and it is often the case that athletes know little about the risks of sun exposure (Darlington et al., 2003). Protective means such as avoiding cycling outside during the peak UVR times (between 10 am and 4 pm), wearing adequate clothing and sunglasses, and applying water-resistant sunscreen need to be strongly communicated in the cycling community. The use of dosimeters can quantify the sun exposure and risks in relation to the use of sunscreen (Thieden et al., 2005).

This study was an important preliminary step in delineating the factors related to cyclists’ sunscreen use and active behaviors taken before cycling. Analyzing cumulative exposure to specific body sites could provide useful information for athletic organizations and health educators to make suggestions to cyclists new to or actively involved in the sport. The current research contributes to understanding the general perspective of cyclists regarding UV protection and sun safety issues. Recognizing the factors that influence cyclists’ current behaviors may be helpful for designing educational messages and sun safety messages that could be tailored to influence change and a healthier lifestyle, especially because cyclists are already open to improving health and fitness.

Footnotes

Competing Interests

None declared.

References

Ajzen

(1991) The theory of planned behavior. Organ Behavior Human Decision Process 50: 179–211.

American Academy of Dermatology (2010) www.aad.org (accessed 9 January 2010).

American Cancer Society (2009) Skin cancer facts. Available at: http://www.cancer.org/docroot/CRI/content/CRI_2_2_1X_How_many_people_get_melanoma_skin_cancer_50.asp?sitearea=.

Amir

Wright

Kernohan

EEM

Hart

(2000) Attitudes, beliefs, and behaviour regarding the use of sunbeds among healthcare workers in Bradford. European Journal of Cancer Care 9: 76–79.

Bandura

(1997) Self-Efficacy: The Exercise of Control. New York: Freeman.

Cohen

West

Aiken

(2003) Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. Mahwah, NJ: Erlbaum.

Craciun

Schuez

Lippke

Schwarzer

(2012) Facilitating sunscreen use in women by a theory-based online intervention: A randomized controlled trial. Journal of Health Psychology 17(2): 207–216.

Darlington

Williams

Neale

Frost

Green

(2003) A randomized controlled trial to assess sunscreen application and beta carotene supplementation in the prevention of solar keratoses. Archives of Dermatology 139: 451–455.

Fitzpatrick

(1988) The validity and practicality of sun-reactive skin types I through IV. Archives of Dermatology 124: 869–871.

10.

Gollwitzer

(1999) Implementation intentions: Strong effects of simple plans. American Psychologist 54(7): 493–503.

11.

Ingledew

Ferguson

Markland

(2010) Motives and sun-related behaviour. Journal of Health Psychology 15(1): 8–20.

12.

Jones

Abraham

Harris

Sculz

Chrispen

(2001) From knowledge to action regulation: Modeling the cognitive prerequisites of sunscreen use in Australian and UK samples. Psychological Health 16: 191–206.

13.

Kenny

Kashy

Bolger

(1998) Data analysis in social psychology. The Handbook of Social Psychology, Vols. 1 and 2 [e-book]. New York: McGraw-Hill, pp. 233–265. Available at: PsycINFO, Ipswich, MA (accessed 1 March 2010).

14.

Kimlin

Martinez

Green

Whiteman

(2006) Anatomical distribution of solar ultraviolet exposures among cyclists. Journal of Photochemistry and Photobiology B 85: 23–27.

15.

Liffrig

(2001) Phototrauma prevention. Wilderness and Environmental Medicine 12(3): 195–200.

16.

Mahler

HIM

Kulik

Gerrard

Gibbons

(2007) Long-term effects of appearance-based interventions on sun protection behaviors. Health Psychology 26: 350–360.

17.

Mahler

HIM

Kulik

Gibbons

Gerrard

Harrell

(2003) Effects of appearance-based intervention on sun protection intentions and self-reported behaviors. Health Psychology 22: 199–209.

18.

Moerhle

(2008) Outdoor sports and skin cancer. Clinics in Dermatology 26: 12–15.

19.

Moehrle

Heinrich

Schmid

Garbe

(2000) Extreme UV exposure of professional cyclists. Dermatology 201: 44–45.

20.

Myers

Horswill

(2006) Social cognitive predictors of sun protection intention and behavior. Behavioral Medicine 32: 57–63.

21.

Newman

Woodruff

Agro

Mayer

(1996) A survey of recreational sun exposure of residents of San Diego, California. American Journal of Preventative Medicine 13(3): 186–194.

22.

O’Riodan

Lunde

Steffen

Maddock

(2006) Validity of beachgoers’ self-report of their sun habits. Archives of Dermatology 142: 1304–1311.

23.

Paul

Paras

Harper

Coppa

(2011) Harm minimization in tan seekers: An exploration of tanning behaviour and the potential for substitutional use of sunless tanning products. Journal of Health Psychology 16(6): 929–937.

24.

Rogers

Weinstock

Harris

. (2006) Incidence estimate of nonmelanoma skin cancer in the United States. Archives of Dermatology 146(3): 282–287.

25.

Sobel

(1982) Asymptotic confidence intervals for indirect effects in structural equation models. In: Leinhart

(ed.) Sociological Methodology. San Francisco, CA: Jossey-Bass, pp. 290–312.

26.

Sutton

(1999) How accurate are smokers’ perceptions of risk? Health, Risk and Society 1: 223–230.

27.

Thieden

Philipsen

Sandby-Møller

Wulf

(2005) Sunscreen use related to UV exposure, age, sex, and occupation based on personal dosimeter readings and sun exposure behavior diaries. Archives of Dermatology 141(8): 967–973.