Abstract
Introduction:
The purpose of this study was to determine the proportion of skateboarders who owned and who wore a helmet and which constructs from the Health Belief Model predicted helmet ownership and helmet use among undergraduate skateboarders.
Methods:
From March 2013 through March 2014, 83 skateboarders completed a helmet attitude and use survey.
Results:
Among participants, 47% owned a helmet while 18% wore a helmet on their most recent ride. Skateboarders who did not own a helmet were compared to skateboarders who owned a helmet but did not wear it and to skateboarders who wore a helmet. Positive trends for emotional benefits, friends and family cues to action and parental rules cues to action were observed. A negative trend for vanity and discomfort barriers was observed. Friends and family cues to action, parental rules cues to action and lower perceived danger were associated with helmet ownership. Friends and family cues to action were associated with helmet use.
Conclusion:
The findings are consistent with the thrill seeking culture of skateboarding. Implications for interventions to increase helmet use among undergraduate skateboarders are discussed.
Introduction
Skateboarding is a relatively young sport, beginning in earnest less than six decades ago. Since its inception as an offshoot of surfing, skateboarders have been viewed as a rebellious, anti-authoritarian culture. The potential for harm and the severity of injuries has also increased as the tricks and speeds have increased. Skateboards are comprised of three main components: the deck, the trucks and the wheels. Variations in the three components impact performance of the skateboard. In the early history of skateboarding, homemade boards with metal or clay wheels were extremely dangerous and the sport was outlawed in many cities (Beal, 2013; Mattern, 2009; Owen, 2013). With the development of urethane wheels in the 1970s, skateboarding experienced a resurgence with smoother rides. Skateboarders invented new tricks as they used drained swimming pools and skateparks for what came to be known as vertical skateboarding, or vert skating. However, many skateparks struggled under high insurance rates and vert skating on ramps lost dominance to street skating, where skateboarders use handrails, sidewalk curbs, walls and other surfaces to perform tricks. The ESPN X games began in 1995 and helped to entrench the sport as a (relatively) mainstream sport (Pickert, 2009). Longboards, with a deck typically over 3 feet long and with larger wheels than other skateboards, are often used for downhill racing and transportation. In the USA, skateboarding is now around a 5-billion-dollar industry including DVDs, video games, fashion and so on (Beal, 2013). It is also a recreational activity and a means of transportation for many college students.
League sports and government funded park and recreation programmes, school-sponsored sporting events and professional sports associations provide a mechanism to regulate behaviours, such as helmet use, among participants. Even under these circumstances, helmet use is rarely required. However, skateboarding is more than a sport; it is also a social experience and culture with its own language, clothing and identity (Karsten and Pel, 2000; Woolley and Johns, 2001). Skateboarders are often observed as being supportive of one another, focusing on achieving high skill levels and learning from one another rather than competing (Beal, 1995; Haines et al., 2010). As skateboarders work to develop their skills, they accept that injuries may occur, but relish the freedom accompanying the sport (Haines et al., 2010). Verbal heckling or physical interference, known as ‘snaking’, may be used to indoctrinate new skaters; acceptable behaviours within the skateboarding culture are taught through interactions with more experienced skaters (Petrone, 2010). Taking risks is rewarded within this culture, particularly among male skateboarders (Atencio et al., 2009). Understanding the beliefs of skateboarders about helmet use may ultimately be used to increase use in sporting and non-sporting events and decrease the incidence and severity of injuries suffered by skateboarders.
The US Consumer Product Safety Commission (CPSC, 2013) estimates that there were 114,120 skateboard injuries requiring treatment in hospital emergency departments in 2012. While the majority of those injured skateboarders were treated and released, 3.1% required hospitalisation or were fatal. The overall incidence of skateboard injury was 36.4 injuries per 100,000 persons (69.2 per 100,000 for men and 10.7 per 100,000 for women). Including estimations of participant exposure, the incidence of skateboard injury may be 8.9 per 1,000 participants (Kyle et al., 2002).
Skateboarding injuries requiring treatment may occur in greater numbers and at higher rates among those in late adolescence and early adulthood, including college students. Over half of the skateboard injuries (58,315) reported by the CPSC occurred among adolescents and young adults aged 15–24. The incidence of skateboard injury in this age group, 132.7 per 100,000, was higher than in any other age group. Within this age group, skateboard injuries were ranked eighth among sports and recreational equipment (CPSC, 2013). Young adults, aged 18–24 years, make up 10.5% of the skateboarding population; however, the age-specific incidence rates of skateboard-associated injury may be higher in this age group than in any other age group (Kyle et al., 2002).
Upper extremity and lower extremity injuries may be the most common injuries among skateboarders (Hunter, 2012); however, it is important to understand the frequency of head injuries because of the potential severity of related consequences. The frequency of head injury among skateboarders reported from different surveillance systems may differ because of different levels of care or disease classification systems (Tominaga et al., 2013). Data from the US National Electronic Injury Surveillance System (NEISS), which tracks emergency department visits, suggest 5.9% of injuries sustained by skateboarders under the age of 20 involved a traumatic brain injury (TBI) (Gilchrist et al., 2011). Using data from the US National Trauma Data Bank (NTDB), 36.3% of skateboard injuries resulted in a TBI (Lustenberger et al., 2010).
Based on injury surveillance data, helmet use and injuries may differ significantly by age, sex and skateboard type (Fabian et al., 2014; Keays and Dumas, 2014). Injuries among skateboarders over the age of 16 were more likely to be severe, more likely to involve a TBI and more likely to result in death (Lustenberger et al., 2010). Being male, failure to use a helmet, riding in the street and longboarding have been frequently associated with injury and TBI among skateboarders (Fabian et al., 2014; Keays and Dumas, 2014; Konkin et al., 2006; Lustenberger et al., 2010).
Wearing a helmet serves to protect against TBI (Cripton et al., 2014; Lustenberger et al., 2010). The American Academy of Pediatrics recommends the use of a bicycle/multi-sport helmet (Bull et al., 2002). Yet, helmet use is uncommon among skateboarders and may be absent among older skateboarders (Page et al., 2012). Other reports of skateboard helmet use are derived from injury data surveillance systems. For injured skateboarders of all ages, Fabian et al. (2014) report helmet use to be relatively low among injured skateboarders (1.7%) and longboarders (4.0%) presenting at a trauma centre; however, helmet status for the majority of those injured was unknown. Helmet use appears to be more common among younger skateboarders (Lustenberger et al., 2010; Page et al., 2012). For injured skateboarders under the age of 18 included in a surveillance system databank of emergency room visits, Keays and Dumas (2014) report helmet use to be slightly higher among injured skateboarders (23.7%) and longboarders (13.9%); however, helmet status was unknown for nearly one-third of those injured. Lindsay and Brussoni (2014) report helmet use among injured skateboarders under the age of 17 to be 32.9%. Using data from injured skateboarders and record review may be problematic for at least two reasons: (1) skateboarders who experienced a less severe injury because of wearing a helmet may not be represented in trauma centres or in injury registries and (2) data on helmet use may be inaccurate or incomplete (Konkin et al., 2006).
Helmet use and interventions among cyclists
Because of the limited information available for skateboarding, it is important to consider helmet use in other wheeled activities and extreme sports. Among US adult cyclists who had ridden in the previous year, 28% reported wearing a helmet on all rides while 46% reported never wearing a helmet (Schroeder and Wilbur, 2013). Helmet use among cyclists may differ by the primary purpose of the ride (Kakefuda et al., 2009). Among injured cyclists, those who wore a helmet were less likely to have a head injury and when a head injury occurred, it was less severe than in cyclists who were not wearing a helmet (Attewell et al., 2001; Bambach et al., 2013; Thompson et al., 1989). Effective helmet interventions among cyclists have relied on several strategies: directly or indirectly providing helmets to cyclists for free or at reduced cost, education of young cyclists and or their parents in schools, medical facilities or community settings and the use of peer educators (Lohse, 2003; McPherson et al., 2009; Quine et al., 2001; Royal et al., 2007; Wu and Oakes, 2005). Ludwig et al. (2005) employed social marketing with peer agents, pledge cards and helmet vouchers to increase helmet use among college cyclists. Mandatory helmet legislation for youth appears to increase helmet use and decrease head injury rates compared to adults or to areas without legislation (Lindsay and Brussoni, 2014; Macpherson and Spinks, 2008). However, few interventions for cyclists were developed within a specific theoretical framework (Thompson et al., 2002). The Health Belief Model (HBM) has been successfully used to predict helmet use among undergraduate bicyclists (Ross et al., 2010). As one of the most commonly used theories in the development of public health interventions, the HBM is relevant to the development of injury prevention strategies (Glanz and Bishop, 2010). Understanding the most salient constructs may suggest how to structure future interventions.
Health Belief Model
The HBM includes the following constructs: perceived susceptibility, perceived severity, perceived benefits, perceived barriers and cues to action (Janz et al., 2002). Within the context of helmet use among skateboarders, the constructs may explain why some skaters choose to own and wear a helmet while others do not. Perceived susceptibility would be the skateboarder’s beliefs of the likelihood of sustaining a head injury as a result of a skateboard crash (i.e. risk of being involved in a crash in which he or she would hit his or her head), while perceived severity refers to the magnitude of consequences if such an injury occurred. The perceived benefit of helmet use is the belief that the helmet may reduce either the susceptibility or the severity of the injury and resultant consequences. Perceived barriers to helmet use may include financial and psychosocial costs of the behaviour. Cues to action are triggers, which serve to initiate helmet use. In the presence of cues to action, skateboarders who believe they are sufficiently likely to incur injuries and believe that helmets reduce the severity of the injury are more likely to wear helmets provided any barriers can be overcome. The purpose of this study was to determine the proportion of skateboarders who own and wear a helmet and which constructs of the HBM predict helmet ownership and helmet use among undergraduate skateboarders. Determining the most salient constructs of the HBM is an important step in developing targeted interventions to increase helmet use among undergraduate skateboarders. As skateboarders remain present on college campuses with potential for increases in popularity, it is important to consider negative outcomes of the behaviour and develop harm reduction strategies.
Methods
Participants
Individuals enrolled in a general education health and wellness course from March 2013 through March 2014 were invited to participate in the research study if they self-identified as a current or recent skateboarder and were at least 18 years old. The general education class from which participants were recruited was one of two courses which students choose to fulfil a graduation requirement for a three-credit course within the wellness domain. Each semester, between 1,150 and 1,500 students are enrolled in the course. As a course requirement, students must attend a minimum of four out-of-class university sponsored events related to different dimensions of health. Participating in the study and the discussion that followed qualified as an event related to either the environmental or intellectual dimensions of health. Students could withdraw from the study yet still receive credit for attending the event. The research project was approved by the Institutional Review Board at James Madison University.
The university is located within a municipality that has a bicycle helmet ordinance requiring cyclists under the age of 15 to wear a helmet. The university has written policy in the handbook, which strongly recommends helmet use for cyclists and for skateboarders.
Procedures
Participants were recruited from all sections of a health and wellness course; the course was one of two courses that fulfils a general education requirement. Announcements were made during class sessions by one of the investigators or by the instructor of record to invite individuals who self-identified as either a current or recent skateboarder to attend a discussion session related to the health of skateboarders. The session was advertised via a course website listing all qualifying events. Attending the session, led by one or more of the investigators, fulfilled a course requirement for the participants. Following the informed consent procedure, participants completed a 70-item closed-form questionnaire. After completing the survey, participants discussed their skateboarding experiences within the context of the HBM and the different dimensions of health. The qualitative data from the semi-structured questions used in the discussion are not included in the present analysis.
Measures
Demographic questions included current age, sex, race/ethnicity, location of residence and year in school. Participants were asked about the types of skateboards owned and age at which the participant began skateboarding. Participants reported helmet ownership (or lack of ownership) and helmet use (or lack of use) on the most recent skateboard trip as the dependent variables.
Following demographic information and skateboarding history questions, the questionnaire included 55 items adapted from the Bicycle Helmet Attitude Scale (BHAS) (Ross et al., 2010). The BHAS was developed within the framework of the HBM. Five constructs of the HBM were represented with 10 subscales: Vulnerability (Perceived Exemption from Harm and Perceived Dangers), Perceived Severity of Harm, Perceived Benefits (Emotional Benefits and Safety Benefits), Perceived Barriers (Personal Vanity and Discomfort Barriers and Cost Barriers) and Cues to Action (Friends and Family, Parent Rules in Childhood and Media). The questions from the BHAS were modified to refer to skateboarders rather than to cyclists. The responses to the modified BHAS were coded on a 5-point Likert-type scale from 1 = strongly disagree to 5 = strongly agree. Subscale scores represent the average response for all answered questions within the construct.
Comparing the modified survey used within this research to the original survey is important. With the exception of the exemption subscale (α = .57), the remaining nine subscales had acceptable to good levels of internal consistency in the skateboard helmet attitude survey (α = .66–.84). The internal reliability of the subscales of the survey was slightly lower among the skateboarders than among the original bicyclists (Ross et al., 2010). This difference may be influenced by the smaller sample size in the current study (Shevlin et al., 2000).
Analysis
Non-parametric statistical tests, including χ2 and Mann–Whitney U, were used to determine if helmet ownership and helmet use differed between demographic groups. Analysis of variance (ANOVA) with Bonferroni post hoc and linear trend analyses was performed to compare the scores of skateboarders who did not own a helmet to skateboarders who owned a helmet but did not wear it on their last ride to skateboarders who did wear a helmet on the last ride. The normality of the distribution was assessed using the Shapiro–Wilk test. Although the data violate the assumption of normality, the ANOVA is robust against non-normal distributions (Norman, 2010; Schmider et al., 2010). The homogeneity of variances was assessed by Levene’s test for equality of variances; no violations occurred. Since outliers were identified by inspection of the boxplots, the ANOVAs were repeated three times as follows: excluding outliers, replacing outliers with the next most extreme score, and using square root transformed scores. Results were consistent across the three alternatives with the following exception: exemption from harm scores between those who owned a helmet but did not wear it and those who wore a helmet was significantly different when outliers were removed or altered. Therefore, the original results are reported. In addition, results from the ANOVA were compared to the Kruskal–Wallis tests which possess fewer restrictions (Bruce et al., 2008).
Logistic regression was performed to determine the effects of the HBM constructs on helmet ownership among all participants. The omnibus test of model coefficients suggests that the model is statistically significant, χ2(3) = 27.44, p < .001; furthermore, the Hosmer and Lemeshow test is not statistically significant (p = .363). The model explained 37.9% of the variation in helmet ownership based upon the Nagelkerke R2 value. A logistic regression was performed to determine the effects of the HBM constructs on helmet use among helmet owners. The omnibus test of model coefficients suggests that the model is statistically significant, χ2(3) = 15.45, p < .001; furthermore, the Hosmer and Lemeshow test is not statistically significant (p = .954). The model explained 45.4% of the variation in helmet use among helmet owners based upon the Nagelkerke R2 value. The HBM constructs used in the model meet the assumption of linearity (Box and Tidwell, 1962) and do not possess multicollinearity (variance inflation factor [VIF] < 1.20).
Results
Selected characteristics of the 83 study participants are presented in Table 1. The majority of participants were Caucasian (84.3%), male (80.7%), lived in a dormitory (62.7%) and were first year students (50.4%). While 47% of the skateboarders owned a helmet, only 18% wore a helmet on the previous ride. Longboards were the most frequently owned skateboard. Helmet ownership and reported use were more common among female skateboarders (χ2 = 5.38, p = .02; χ2 = 5.79, p = .02). Neither helmet ownership nor helmet use differed between the remaining demographic characteristics.
Demographic characteristics of study participants (N = 83).
Includes three American Indian/Alaskan Native, two African American and three Hispanic participants.
Participants may own more than one type of skateboard.
χ2 = 5.38, p = .020.
χ2 = 5.79, p = .016.
Health Belief Model
The reliability of the subscales and mean scores for the three categories of participants are presented in Table 2. The internal reliability of the 10 subscales ranged from .57 for exemption to .84 for friends and family. ANOVA revealed differences in emotional benefits (F = 5.99, p = .004), vanity and discomfort barriers (F = 5.09, p = .008), friends and family cues to action (F = 17.48, p < .001) and parental rules cues to action (F = 5.23, p = .007) between the three categories of skateboarders: (1) those who did not own a helmet, (2) those that owned a helmet, but did not use it and (3) those that used a helmet. Perceived exemption from harm, danger, severity, safety benefits and cost barriers did not significantly differ between the three groups.
Internal reliability and comparisons of subscales between (1) those who did not own a helmet, (2) those who owned a helmet, but did not use it and (3) those who used a helmet.
K-W: Kruskal–Wallis test; ANOVA: analysis of variance; SD: standard deviation.
Group 3 > Group 1, p = .003.
Group 3 < Group 1, p = .008; Group 3 < Group 2, p = .028.
Group 3 > Group 1, p < .001; Group 3 > Group 2, p = .003; Group 2 > Group 1, p = .031.
Group 2 > Group 1, p = .015.
The following results describe the Bonferroni post hoc analysis describing group differences and tests for linearity. Helmet users reported significantly greater emotional benefits than skateboarders who did not own a helmet (p = .007); non-users were not different from non-owners or from users. A significant, positive linear trend was observed for emotional benefits (F = 11.74, p = .001). Skateboarders who wore a helmet reported lower vanity and discomfort barriers than non-users (p = .03) and non-owners (p = .008); non-users were not different from non-owners. A significant, negative linear trend was observed for vanity and discomfort barriers (F = 9.63, p = .003). Those who wore a helmet reported greater family and friend cues to action than non-users (p = .03) and non-owners (p < .001); non-users reported significantly greater friends and family cues to action than non-owners (p = .031). Parental rules cues to action were greater among non-users compared to non-owners (p = .015). Parental rules cues to action were not different between non-users and users; however, when users and non-users were combined and compared to non-owners, parental rules cues to action remained significantly different (t = –3.56, p = .001; data not shown in table). Positive linear trends were significant for all three cues to action subscales: friends and family (F = 34.29, p < .001), parental rules (F = 5.26, p = .024) and media (F = 4.47, p = .038). Results of the Kruskal–Wallis tests were consistent with the results from the ANOVAs.
Predicting helmet ownership and helmet use
The results of the multivariate logistic regression examining helmet ownership among all participants and helmet use among helmet owners are presented in Table 3. Friends and family cues to action (odds ratio [OR] = 3.20 [1.55, 6.62], p = .002) and parental rules cues to action (OR = 3.50 [1.42, 8.66], p = .007) were positively associated with helmet ownership while perceived danger (OR = 0.34 [0.13, 0.89], p = .027) was negatively associated with helmet ownership. The associations were not affected by demographic variables. Among helmet owners, only friends and family cues to action (OR = 10.41 [2.39, 45.26], p = .002) were positively associated with helmet use.
Multivariate logistic regression of helmet ownership and helmet use.
OR: odds ratio; CI: confidence interval.
Discussion
Helmet ownership and helmet use was reported by 47% and 18% of participants, respectively. Reported helmet use was significantly higher than reported by Page et al. (2012), where none of the adolescent or adult skateboarders were observed wearing helmets. The prevalence of helmet use from this study may be less directly comparable to retrospective studies that used injury surveillance data. Fabian et al. (2014) report a prevalence of helmet use of 3.0% among injured longboarders and skateboarders of all ages. Keays and Dumas (2014) reported a prevalence of 23.1% helmet use among injured longboarders and skateboarders under the age of 18. However, helmet use data was missing for between 30% and 85% of their records.
Female skateboarders were more likely to wear a helmet than male skateboarders. None of the American Indian/Alaskan Natives, African American and Hispanic skateboarders used a helmet. Although it appeared that longboarders were less likely to use a helmet than classic skateboarders, the difference was not significant. Other studies have found inconsistent differences in helmet use between longboarders and skateboarders (Fabian et al., 2014; Keays and Dumas, 2014).
Surprisingly, reported helmet use among the skateboarders (18%) was higher than among cyclists (12%) using similar survey procedures (Ross et al., 2010). Among cyclists, significant differences were found between helmet wearers and non-wearers for all 10 subscales while only 4 of the subscales were different among the skateboard groups. As with the cyclists, the friends and family cues to action was the most significant predictor of helmet use among skateboarders. The friends and family cues to action subscale included questions about whether friends wear helmets and whether family and friends believe the participant should wear a helmet. Parents seem to be important in the acquisition of the helmet, but less important in whether the helmet owner wears the helmet. The significant trend observed in media cues to action may indicate that helmet owners and wearers are more perceptive of messages or heavier consumers of skateboard related media. The lower scores reported by non-helmet owners suggest that media messages are either not being seen or are not being remembered by those who do not own a helmet.
Emotional benefits were greater among helmet wearers while safety benefits were not different between the three groups. The susceptibility to and severity of harm from a fall did not differ between groups. This finding is consistent with skateboarders’ acceptance of injury as part of the sport (Haines et al., 2010). Furthermore, the safety benefits of wearing a helmet did not differ between the groups. It appears that these three constructs already exist in favourable conditions for the behaviour.
Helmet wearers had lower perceived vanity and discomfort barriers than non-wearers. Discomfort is also a barrier to helmet use among cyclists (Finnoff et al., 2001). Emotional benefits correlated with friends and family cues to action. These findings are consistent with the sense of identity and belonging that accompanies being part of the skateboarding culture. A helmet may not be seen as part of the culturally appropriate attire and may go against the risk-taking, extreme bravado of the sport (Beal, 2013; Hunter, 2012). However, those who report greater friends and family cues to action may be more likely to report emotional benefits because they have accepted the messages sent by their friends. Three of the items within the friends and family cues included that the skateboarder had friends that wore helmets or believed that their friends would want them to wear a helmet. Ignoring these cues to action may elicit feeling of guilt or violations of peer expectations.
Overall, participants perceived skateboarding as a potentially dangerous activity. However, after controlling for cues to action, lower perceived danger was associated with helmet ownership. One explanation for this finding is that those who are likely to own a helmet may also be those who are more cautious and therefore less susceptible to falling. Another possible explanation is that those who own a helmet may believe that they are less susceptible to injury when they do fall. Risk compensation occurs among male cyclists: those who wear helmets travel faster (Messiah et al., 2012). A similar phenomenon may occur among skateboarders.
Limitations and future research
The findings reported derive from undergraduate skateboarders at one university. Two characteristics of the study site may influence the generalisability of the findings to skateboarders at other universities. The terrain of the 721-acre campus includes substantial hills and is bisected by both an interstate and railroad tracks. These physical characteristics may influence travel patterns and level of risk associated with skateboarding on campus because of potential increases in speed and limited number of alternative travel routes. Investigating the influence of natural and built environment variables on perceptions of risk and behaviours should be pursued. Second, prior to the initiation of this study, serious injuries (including TBI) and a death occurred among skateboarders on the university campus. These events serve as anecdotal evidence of the risks associated with skateboarding at the university and may influence perceptions and behaviours of current skateboarders included in the study. The skateboarders volunteered to participate in the study as part of one of many potential assignments in a general education class. Thus, not all skateboarders in the recruitment courses participated in the research, nor do they necessarily represent all skateboarders at the university.
Future research should seek to confirm these findings among skateboarders at other universities that have different environmental and historical contexts and further investigate differences in helmet use by activity (transportation vs recreation) and location (street vs sidewalk and on-campus vs off-campus). Additionally, the perceptions of friends and family of skateboarders should be measured and compared to the perceptions of the skateboarders. Such comparisons may clarify relationships between the cues to actions that friends and family believe they are conveying and how these cues are received by skateboarders. Additional theories should be investigated to determine which has the strongest predictive value. Among adolescent cyclists, the Theory of Planned Behaviour has provided better fit than the HBM (Lajunen and Räsänen, 2004; Quine et al., 1998).
Implications for increasing helmet ownership and helmet use among undergraduate skateboarders
Although epidemiological data suggest an increase in the incidence of injuries among young adult skateboarders and point to the severity of those injuries among those who do not wear a helmet, little is known about the determinants of helmet use among college students. Therefore, the findings of the current study represent an important contribution to informing initiatives to increase helmet use among young adult skateboarders. The low scores of media cues to action reported by skateboarders who do not own a helmet suggest that additional research needs to be conducted to determine preferred communication routes. The impact of skateboarding professionals and the skateboarding industry on helmet use remains unclear as an intervention strategy. Providing resources to friends and family to support and increase their efforts to provide meaningful cues to action may be the most efficient strategy to increase helmet use among skateboarders. Ludwig et al. (2005) successfully increased helmet use among college cyclists by using peer agents to distribute an informational brochure, helmet use pledge cards and vouchers for a free helmet. Such efforts may also decrease perceived vanity barriers of wearing a helmet. As more friends are observed wearing helmets, the embarrassment of being seen as different (in a culture where being seen as part of the group is important) may decrease (Schieber and Olson, 2002). Interventions focused on increasing perceived danger of skateboarding and safety benefits while decreasing cost barriers may be ineffective alone.
Footnotes
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.