Characteristics and risk factors of bike related accidents : Preliminary analysis

Material and methods. The survey was based on a 39-item questionnaire created for this study. We collected data concerning the demographical status of respondents, their cycling preferences, attitude toward safety behaviors during cycling, and the characteristics of bike-related injuries. Our research covered 729 people who declared themselves as cyclists (302 women – average age: 31.5 ±10.3 years; 427 men – average age: 32.6 ±10.7 years).


Introduction
The growing popularity of cycling as a means of transport, a part of a certain lifestyle and a simple and healthy way of spending time creates the possibility not only to reduce air pollution and traffic, but also to improve public health.2][3] The benefits of cycling are truly impressive and include physical and psychosocial aspects.
It has been proven that cycling enhances self-efficacy, physical well-being and improves quality of life. 1,4,5By following the recommended 30 min of daily physical activity, the reduction of risk for cardiovascular diseases can be obtained.The researchers also hypothesized that it could have a positive, preventive effect against certain types of cancer. 1,67][8] This beneficial influence on health makes this kind of sport especially advisable for patients with type 2 diabetes. 8It has been shown that cycling positively affects the musculoskeletal system, also in terms of rehabilitation. 9,10Application of cycling leg exercise has shown beneficial effects on motor abilities in patients suffering from subacute and chronic diseases. 10,11Moreover, cycling has a positive influence on instrumental activities of daily living and social function in elderly people with mobility limitations.Unfortunately, the elderly are the group which incur more accidents than adults. 12,13New research reports that the loss of the ability to cycle can be a new warning sign for atypical parkinsonism. 14Some authors indicated that cycling in big cities might be erroneously assessed as unfavorable for health, especially for the respiratory system because of exposure to air pollution.However, it has turned out that the benefits of active travel outweigh the harm caused by air pollution in all regions except for the most extreme air pollution concentrations. 2,15onsidering the safety aspects during cycling, several factors of bike-related accidents should be mentioned.First of all, the environment and the infrastructure built may be correlated with bicycle injuries. 1,3Cyclists are often forced to share the same road with cars, buses and trucks, and they are more vulnerable to accidents than motorized drivers. 2 Modification of the existing infrastructure is also important, e.g., replacing signal-controlled intersections with roundabouts. 4Also demographic factors, and traffic speed and density influence the bike-related accident rate. 3,5Different behaviors among cyclists, such as wearing helmets and reflective elements, using headphones, cell-phones or other devices, form a very differential group of factors.
The aim of this study was to characterize the population of cyclists in Poland, assess the occurrence of safety behaviors among them and to precise the risk factors for bike accidents.Additionally, we analyzed bike-related injuries.

Material and methods
This survey was based on a 39-item questionnaire created for this study, covering a wide spectrum of issues related to cycling.It included 2 parts.In most questions, the participants were allowed to choose 1 answer, unless stated otherwise.A few questions additionally had an option for including comments.The 1 st part consisted of 22 questions related to the demographical status of respondents (age, sex, professional activity, education, marital status, having children), their cycling preferences and attitude toward safety behaviors during cycling (e.g.type of bike, frequency of riding, season of cycling, wearing helmets).The 2 nd part concerned only people who had at least 1 cycling injury.It included 17 detailed questions about the circumstances of the accident.
The survey was prepared in a digital version, using online service for creating surveys and tests -www.ankietka.pl.It was available on the website http://www.ankietka.pl/ankieta/206347/ankieta-dla-rowerzystow.html between November 20, 2015 and of February 10, 2016.Our research covered 729 people who declared themselves as cyclists.To distribute the survey to the wide spectrum of respondents, link to the survey was published on Facebook fan pages devoted to cycling, active lifestyle or traveling, and on bicycling websites.

Statistical analysis
The Kolmogorov-Smirnov test was used to assess the normality of data distribution.The χ 2 and Mann-Whitney U tests were done to determine the relationship among variables.Univariate and multivariate logistic regression analyses were performed.The p-value <0.05 was considered statistically significant.

Results
Most respondents were male (58.6%).As shown in Fig. 1, almost half of the group were between 19 and 29 years old (43.8%), and around 1/3 was between 30 and 29 years old (31.3%).Cyclists younger than 19 years and older than 39 years accounted for <25% of the studied sample.
Important differences between men and women were observed during data analysis.The following graphs show these diverse patterns of using a bike by both sexes.Figure 2 illustrates that despite the fact that both men and women prefer spring and summer as the most suitable seasons for cycling, men significantly more often use bikes in spring (by 6.4 percentage points; χ 2 = 10.170;p = 0.002), autumn (by 19.4 percentage points; χ 2 = 46.469;p < 0.001) and winter (by 25.4 percentage points; χ 2 = 46.0234;p < 0.001).No statistically significant difference among male and female groups in the frequency of riding a bike in summer was observed.
According to the survey results, 3 types of bikes were the most popular -mountain bikes (40.9%), touring bikes (25.4%) and city bikes (24.3%).But here significant gender differences were also identified.For women, the most popular type of bike was a city bike (39.7%), while men more often preferred mountain (45%) and touring (28.6%) bikes.The popularity of city bikes among men (13.4%) was close to the popularity of road bikes (11.5%) (χ 2 = 51.082;p < 0.001) (Fig. 4).
About 71.3% of the interviewees declared that they did not always wear helmets.Despite stereotypes about attitudes of men and women toward risk, women more often than men claimed that they did not always use helmets (19 percentage points more; χ 2 = 31.175;p < 0.001) (Fig. 5).This can be explained by the fact that women usually used different type of bikes and also they used them less often than men.
To determine the risk factors, the logistic regression analysis was conducted.In the first step, 19 variables with possible impact on the probability of having an accident were identified and then used in binary logistic regressions with single categorical predictors.These variables were: gender, age group, professional activity, educational level, marital status, number of children and their age, time of the year when a bike is used, the frequency of using a bike, using reflectors, reason for cycling, listening to music, opinions about the obligatory use of helmets and about the usefulness of helmets, frequency of using a helmet, having a driving license and a cycling license, type of bike, and alcohol consumption.It turned out that 10 of these variables were statistically significant in these models (p < 0.05).
In the next step, these 10 variables were used as independent variables in a multivariable logistic regression model.The outcome of this model is presented in Table 1.
Out of 10 variables used in a multivariable logistic regression, 5 remained statistically significant (p < 0.05): educational level, number of children, frequency of using a bike, type of bike used, and frequency of using a helmet.People with secondary education (p < 0.01) or with an incomplete higher education (p < 0.05) were almost 2 times less likely to have an accident than people with a university degree.Another significant predictor was the number of children.Having at least 3 children increased the probability of accidents by 3 times (p < 0.01).What is quite intuitive, cyclists using a bike every day were more exposed to accidents than others.The probability of an accident for them was 3 times higher than the probability of an accident for people who used a bike a few times a month or only during holidays (p < 0.01).The type of bike used also strongly affected the probability of an accident.In comparison with people using city bikes, those riding mountain bikes and fold-up bikes were more prone to accidents -2 times (p < 0.01) and 8 times (p < 0.05), respectively.The last significant variable was the frequency of using a helmet.People who did not always use head protection had less probability of having an accident than people always using helmets during their bike rides.Generally, the respondents declared that they rode a bike a few times a week (45.5%) or every day (26.9%).But data presented in Fig. 3 confirms the fact that women used bikes less frequently than men (χ 2 = 51.082;p<0.001).Female respondents usually declared that they rode a bike a few times a week (36.8%) or from time to time (27.5%), while more than half of male respondents stated that they used a bike a few times a week (51.8%,i.e., 15 percentage points more than women) and around 1/3 claimed that they used a bike every day (30.4%).In our cohort, 277 out of 729 respondents suffered from bike-related injuries.In this group, we noted 870 accidents that resulted in 1,671 different injuries.The most common injuries were related to the upper and lower extremities (48.5% and 32.4% of all injuries, respectively).Injuries of the upper extremities were more often observed in the male group and they constituted 50.1% of all injuries in men (p < 0.025).Wounds of the upper extremities also appeared more frequently in males (17.3% of injuries; p < 0.005).Women more often suffered from injuries of the lower extremities in comparison to men (37.7% vs 30.6%; p < 0.01).This observation was similar for abrasions of lower limbs (24.6% vs 18.8%; p < 0.02).The female group was also at a higher risk of facial skeleton fractures in comparison to the male group (1.9% vs 0.2%; p < 0.001) (Table 2).
In our study, wearing helmets did not influence the region of trauma nor its type (p < 0.05).Detailed data regarding injuries in the helmeted and non-helmeted groups of cyclists was presented in Table 3.

Discussion
Bike-related accidents are becoming a high interest topic in the medical press, mainly due to the growing popularity of cycling and an observed peak of injuries among cyclist. 16,17This research article examined the population of cyclists -their habits, cycling behavior in relation to potential risk factors of bike-related accidents -and also described the injury pattern.
9][20][21][22][23][24][25][26] Most articles concerned the study population of adolescences with a small sample of elderly people.In the abovementioned articles, males were the main group of cyclists.However, bicycle users are not a homogenous group.8][29] One of the features being the basis for distinguishing bike users is the reason for cycling.For some people, this activity is only a form of recreation, for others, bicycle is a very effective and cheap means of transport (commuters), and lastly there is a group which treats it as a sport, also in its extreme version. 4,28,29Additionally, the differences among cyclists are correlated with the frequency of cycling and the level of their experience. 22Different classifications of cyclists have been developed, but none of them is widely accepted.The proportion of cycling population differs among different countries.In Australia and the United States, bicycles are used by a relatively small group of inhabitants; cycling is treated mostly as a form of recreation and rarely as a means of transport. 30,31In the European countries, such as the Netherlands, Denmark or Germany, cycling is much more popular and mainly used as a mainstream mode of transport. 31Researchers observed that cyclists who used bicycles as a means of transport tended to be younger and travel more frequently (more days per week), in the morning and evening peak hours, than those who rode for recreational purposes.Recreational cyclists treated cycling as physical exercise that helped them maintain a good physical condition; they more frequently used road bikes, while commuters spent more time on hybrid, city, cruiser and comfort bikes. 26 our article, we additionally included other sociological factors, such as professional activity, education, marital status, having children, cycling preferences, and attitude toward safety behaviors during cycling, as potential risk factors for bike-related injury.Our cohort included mostly adult cyclists at the age 19-29 years and only a small group of very young or old people, similarly to the major part of research considering cycling patterns.Also, in our study, * The percentage was calculated with regard to the total number of accidents (n = 870).For the rest of the factors, the percentage was calculated with regard to the total number of injuries in each group (female = 414, male = 1,257, total = 1,671); ** statistically significant.* The percentage was calculated with regard to the total number of accidents (n = 870).For the rest of the factors, the percentage was calculated with regard to the total number of injuries in each group (female = 691, male = 980, total = 1,671).
like in most publications, males constituted the main group of respondents.Our interviewees most frequently declared that they treated cycling as a form of recreation or a means of transport.Women declared riding a bike a few times a week or from time to time, and they more often chose city bikes.Men on the other hand stated that they used a bike more often, i.e., a few times a week or every day, and they usually preferred riding a mountain bike.Other studies also indicated that men had larger experience in cycling.When it comes to safety considerations, women more often than men claimed that they did not always wear helmets.This can be explained by the fact that women used bikes less often than men.However, more than half of the cyclists declared to have reflective elements on their clothes or bikes.Bike accidents are a growing public health problem worldwide.Risk factors of these incidents were discussed in many medical publications.Researchers around the world confirmed that the age of cyclist is correlated with the number of accidents -children, adolescents and people older than 65 years of age take part in more accidents than adults. 20,22,23,25Taking into consideration the number of accidents that take place, the exposure indicator should not be omitted.Subjectively, a higher number of bike crashes noted in the urban areas is strongly correlated with a greater number of bike users.Some authors compared the number of cyclists and the number of bikerelated injuries in and outside the city and it turned out that off-road cycling was much more unsafe. 18,20There is little information in the literature about the correlation of the bicycle type and bike-related injuries.It is probably related to the belief that many kinds of bicycles are used not necessarily in line with their main purpose, e.g., mountain bikes are often used as a means of transport in the city and city bikes can be seen on short off-road trips.
Our study recognized several groups of cyclists with a higher probability of undergoing bike-related accidents.They included people with secondary education and incomplete higher education, parents with at least 3 children, people who used a bike every day (higher exposure), people riding city bikes, mountain bikes and fold-up bikes.People who declared always wearing helmets during cycling proved to be more prone to accidents.This seems counterintuitive, but there can be 2 possible explanations.The first one is similar to the conclusion made by Peltzman, who suggested that the probability of risky behaviors increase along with the increase in perceived safety. 32Some studies also indicate that drivers can act less carefully toward cyclists wearing helmets than toward those without. 33t should be stated that both these hypotheses are controversial.Several studies have shown that head injuries are more common and more severe in cyclists that do not wear helmets.Some epidemiological data indicates that head injuries are the most common traumas among cyclists. 34,35It was estimated that among victims of fatal accidents, around 2/3 of them sustained head traumas. 36The most frequent head injuries in bike-related accidents are skull fractures and cerebral contusions. 35Cyclists are also at a higher risk of undergoing facial fractures. 379][40] Some authors emphasized that wearing bicycle helmets did not reduce the incidence of mid-facial fractures and was even associated with an increased risk of mandibular fractures. 37ontrary to the abovementioned articles, in our studies, the most common injuries were those of the upper extremities (48.53% of all injuries) and lower extremities (32.38%).Injuries of the upper extremities, including wounds, were more often observed in the male group.Women more often suffered from injuries of the lower extremities (37.68%), comprising abrasions of the lower limbs.The female group was also at a higher risk of facial skeleton fractures in comparison to the male group.In our study, wearing helmet did not influence the region of trauma nor its type.
In the cited articles, the injuries of the upper and lower extremities were the second most common type of traumas.[40]

Conclusions
Young men prevailed among cyclists in Poland.Gender-related dependencies can be observed in preferences for bike type, and cycling characteristics and pattern.For women, the most popular type of bike is a city bike, while men more often prefer mountain and touring bikes.The overwhelming majority of bike users admitted to not wearing protective helmets.In our study, we identified 5 risk factors for bike-related accidents.People with secondary education and incomplete higher education were at a higher risk of injuries as well as those who have at least 3 children.Frequency of using a bike (the more often, the more injuries), type of bike used (mountain bikes and fold-up bikes) and frequency of wearing helmets constituted the other group of factors correlated with a higher rate of injuries.The most common locations of injuries proved to be the upper and lower extremities.

Table 1 .
Coefficients of logistic regression model OR -odds ratio; CI -confidence interval; * statistically significant.

Table 2 .
Characteristic of injuries among the female and male groups of cyclists

Table 3 .
Characteristic of injuries among the helmeted and non-helmeted groups of cyclists