Design

Our study was cross-sectional in design and followed STROBE guidelines [18]. A self-administered questionnaire was made accessible online worldwide from February 2016 to October 2016. The study complied with the Declaration of Helsinki and was approved by the ethics committee from the Universitat de València (approval number H1449762164108). Participants approved the informed consent in the first page of the online questionnaire.

Participants

The questionnaire was sent to several cycling organizations, cycling clubs, Internet blogs, associations, and posted on several internet forums intended for cyclists. These organizations and websites were based in numerous countries, with variations as to whether their audience was national or global.

To be eligible for this study, the respondent had to be: literate in English, Spanish, French, or Portuguese; aged 18 years or older; and be involved in non-sporadic cycling (≥2 rides/week, ≥50 km/week and ≥3 hours of cycling/week). We excluded those respondents submitting data that were incomplete or containing questionable values (e.g., body mass reported as 5 kg). Cycling modalities with less than 50 respondents were also excluded (e.g., BMX modality).

Questionnaire

Table 1 highlights the variables captured by the survey as well as the response options for each variable. To improve the comprehension of some of the questions, images were provided. Although we made a priori decisions on how to recode certain variables for analysis, such recoding was further modified in those cases in which resulting category counts were low in number. In this study, injury was defined as any physical complaint sustained by the participant, irrespective of the need for medical attention or time loss from cycling activities [19,20].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Information requested on the questionnaire.

https://doi.org/10.1371/journal.pone.0211197.t001

To obtain a large and diverse sample of cyclists, we created the questionnaire in four languages (English, Spanish, French, and Portuguese). All four authors assisted in the translations to ensure consistency across the versions. The survey was hosted in Google Forms.

Statistical analysis

Statistical analyses were performed using RStudio [25]. Descriptive analyses were performed to acquire averages and 95% confidence intervals (95%CI). Logistic regression analyses were performed to estimate odds ratios (OR) and 95%CI. A total of seven models were run. Outcome variables of interest were: reporting any injury within the last 12 months (recoded as injured within the last 12 months: yes/no); reporting an overuse/fatigue injury within the last 12 months (injuries in which participants perceived overuse/fatigue as the cause); reporting an acute injury within the last 12 months (injuries that participants perceived that the cause was fall, contact with vehicle, other bicyclist, pedestrian or with stand-still structure); reporting body comfort (recoded as body discomfort reported: yes/no); reporting saddle comfort (recoded as saddle discomfort reported: yes/no); reporting pain during cycling (recoded as yes/no); and reporting pain while not cycling (recoded as yes/no). The exposure variables included the variables captured from: demographics; characteristics of cycling profile within the last 12 months; characteristics of fitness training within the last 12 months; and bike characteristics. Stepwise multiple regressions in both directions were performed to find the model with the best AIC (Akaike Information Criterion) [26]. Final models were then adjusted to retain only variables yielding p-values <0.05.

Injuries reported in last 12 months

Most cyclists reported no injuries in the past 12 months (63.2%; 95%CI [59.7–66.7%]), 25.3% reported one injury (95%CI [22.2–28.5%]), and 11.5% reported two or more injuries (95%CI [9.2–13.8%]). Injury characteristics are presented in the Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Characteristics of injuries reported by participants within the last 12 months.

Percentages were obtained from the total number of cyclists injured (n = 275).

https://doi.org/10.1371/journal.pone.0211197.t002

Table 3 presents the logistic regression models for being injured in the past 12 months. The odds of reporting an injury within the past 12 months increased when cyclists also engaged in running, rode on trail terrain (compared to road terrain), biked to work, and reported pain during cycling and while not cycling. The odds of reporting an overuse/fatigue injury increased when the cyclist was also engaged in running and swimming, and reported pain during cycling and while not cycling. The odds of reporting an acute injury increased when cyclists biked to work, rode at lower speed, and had a low-end bike (compared to mid-end and high-end bikes).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Logistic regression model to assess the odds of reporting any injury, overuse/fatigue injury, or acute injury within the past 12 months.

Referent category was not reporting an injury.

https://doi.org/10.1371/journal.pone.0211197.t003

Comfort

Most cyclists reported their body posture during cycling to be comfortable (95.8%; 95%CI [94.4–97.3%]); 4.2% reported discomfort (95%CI [2.8–6.0%]). Most cyclists reported their saddle to be comfortable (89.3%; 95%CI [87.1–91.5%]); 10.7% reported discomfort with their saddle (95%CI [8.5–12.9%]).

Table 4 presents the logistic regression models predicting the odds of self-reported body posture discomfort and saddle discomfort. The odds of reported body posture discomfort increased with reporting saddle discomfort and pain while cycling. The odds of self-reported saddle discomfort increased with reporting body posture discomfort and perceived pain during cycling, but decreased when cyclists reporting biking to work.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Logistic regression models assessing the odds of reporting body posture discomfort and saddle discomfort while cycling, respectively.

Referent categories for models were not reporting body discomfort and saddle discomfort while cycling, respectively.

https://doi.org/10.1371/journal.pone.0211197.t004

Pain

While 36.8% of cyclists did not report perceived pain while cycling (95%CI [33.3–40.3%]), 63.2% reported perceived pain (95%CI [59.7–66.7%]). The body regions where pain was most often reported were: neck (23.1%; 95%CI [20.1–26.2%]), lower back (22.2%; 95%CI [19.2–25.2%]), knee (15.6%; 95%CI [12.9–18.2%]), hand (13.3%; 95%CI [10.8–15.7%]), genital area (11.6%; 95%CI [9.3–14.0%]), and shoulder (11.1%; 95%CI [8.8–13.4%]).

Overall, 58.2% of cyclists noted no pain when they were not cycling (95%CI [54.6–61.8%]); 42.0% reported pain while not cycling (95%CI [38.3–45.4%]). The body regions where pain was most often perceived were: lower back (19.4%; 95%CI [16.5–22.2%]), knee (11.6%; 95%CI [9.3–14.0%]), and neck (8.8%; 95%CI [6.8–10.8%]).

Table 5 presents the logistic regression models predicting the odds of reporting perceived pain during cycling and while not cycling. The odds of reporting perceived pain during cycling increased with reporting saddle discomfort, perceived pain while not cycling, and injury within the past 12 months; however, biking to work, having more experience, and training the core musculature were associated with decreased odds for perceived pain. The odds of reporting perceived pain while not cycling increased with reporting perceived pain while cycling and injury within the past 12 months; however, odds decreased when also engaged in running. In addition, the odds of reporting perceived pain while not cycling also varied by cleats use/adjustment.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Logistic regression models assessing the odds of reporting pain while cycling and pain while not cycling, respectively.

Referent categories for models were not reporting pain.

https://doi.org/10.1371/journal.pone.0211197.t005

Factors associated with cycling-related injury and resulting prevention

As found in previous research, the knee, lower back, and shoulder were commonly injured body regions among cyclists [5,6,27,28]. Common diagnoses include inflammatory conditions, sprain/strain, and contusion [5,28,29]. Fatigue and falls are the main causes for these injuries [28,30,31]. The need for continued examination of strategies to prevent cycling-related injury is highlighted by our finding that 28.7% of injured respondents required medical leave; most injuries resulted in recovery time durations of over 2 weeks. Thus, effective prevention strategies may benefit from focusing on these common injury types. In additional, clinical research should examine management and care guidelines that can help cyclists safely return to participation and avoid risk of further complications from injury.

Reporting overuse/fatigue injury within the past 12 months was associated with engaging in running or swimming training. This finding may be plausible since it has been found that cyclists who combine more than one discipline tend to have poor technique, which has been suggested with increased risk of injury [32]. Further, previous studies have found that triathletes produce less effective force and have a greater variation in muscle recruitment than cyclists [33,34]. Additionally, combining running, swimming and cycling without a proper periodization of the training can lead to overload and fatigue, both of which are previously-identified risk factors (also were the main perceived reason of injury by cyclists in our study) [32,35]. However, as a cross-sectional study, we are unable to assess causality. It is possible that running or swimming training may have also been undertaken while recovering from the cycling injury. Future research is needed to further examine the relationship between engaging in multiple forms of physical activity and an increased injury risk.

Compared to road cycling, cycling on trail terrain increased the odds of injury. It is important to note that our terrain categories could not account for variations within each terrain, such as changes in traffic presence in road and trail courses. However, our findings may be attributable to the fact that whereas road cycling has a more level terrain, trail terrain is more varied and thus, results in more vibrations on the bicycle and the cyclist. Previous research has suggested that high levels of vibration on the bicycle may be associated with musculoskeletal disorders, particularly related to the upper extremity, the knee, and the back [36,37]. This additional mechanical load could also increase muscular activity, with such articular strain contributing to injury risk [36,37]. Cyclists engaged in trail terrain cycling may benefit from the use of recovery strategies such as cryotherapy, stretching compression garments, nutrition, or massage therapy under medical supervision [38,39].

Cycling to work was associated with increased odds of reporting an acute injury in the past 12 months. This result may highlight the need for safety-related infrastructure for cyclists as they commute [40,41]. The inverse relationship found between the cyclist’s average ride speed and the probability of having an acute injury may reflect the greater acute injury risk within a city setting, where cycling speed is typically slower. Last, decreased odds of acute injury were found when riding a high-end or mid-range quality bike, compared to a low-end quality bike. Although not measured specifically in our study, such results may highlight the higher quality of components on these bikes (e.g., brakes). Future research is needed to better understand the specific mechanisms by which these associations were found.

Factors associated with reported discomfort and pain

Building upon etiological research focused on injury incidence in cycling, our study also aimed to identify factors associated with perceived discomfort and pain. We hypothesized that discomfort and pain related to cycling may be associated with injury risk. Although our logistic models only found an association between perceived pain and injury, we believe continued exploration of these associations is required with additional samples from the cycling population to further validate or refute our findings. In addition, we believe that examining perceived discomfort and pain are nonetheless important as they may provide valuable information to drive the development of cycling-related injury prevention strategies. Such findings may be of interest to the many stakeholders within the cycling population, including coaches, bike fitting technicians, and medical staff treating such injuries.

Different studies observed how the modification of saddle discomfort was related with factors such as the variability of the sitting postural control [42,43], trunk flexion [44], forward-backward sitting position, and neuromuscular activation of gastrocnemius [43]. A previous study observed that saddle discomfort increases with cycling time [43]. Riding duration can explain why in our study the cyclists that bike to work presented lower odds of saddle discomfort. Although discomfort increases with cycling time [43], it is unknown how this discomfort can alter cycling posture, neuromuscular activation, or pedal forces, thus warranting future examination. In our study, approximately 1 in 10 of the cyclists reported saddle discomfort. Saddle discomfort should not be ignored since this can be the result of a high compression in the gluteal area that is often accompanied by skin discomfort, syndromes such as urinary pain, and numbness in the perineal region [45–47].

Interestingly, we found that the odds of reporting pain while cycling decreased with more cycling experience and when the cyclist reported cycling to work. It is likely that respondents with a higher exposure to cycling may have acclimated themselves to a higher pain threshold [29,48]. This is in agreement with previous studies that found that aerobic exercise was associated with higher pain tolerance [49,50]. The increase in pain tolerance in cycling is important because it was suggested to be an important factor for endurance performance [48]. Also, cyclists that trained their core musculature had decreased odds of reporting pain while cycling. This parallels previous findings suggesting that lumbar pain was reduced by lumbopelvic stabilization training [51]; more specifically in cycling, such pain was also associated with lower control of the trunk musculature [52,53]. Finally, cyclists without adjustments made to their cleats had increased odds of reporting pain while not cycling. A wrong adjustment of the cleats can lead to an excessive Q-angle, an important factor associated with chondromalacia [54]. It is important to note that we measured perceived pain through self-report, which may result in limitations. However, our findings highlight the need to further examine perceived pain in a variety of manners in the context of injury risk to better gauge how training load, pain threshold, and willingness to participate while in pain are related.

Limitations

As in other studies [5,27], a prospective design and a data collection by interviews was considered. Although the use of web-based questionnaires has been supported by the literature [55,56], it is important to be aware of their limitations. One limitation is combining data from different methods of survey, which was avoided here by using the same instrument for all participants, being the language the only difference [57]. The range of languages used and the countries targeted must be considered. Although we included four different languages, it is possible that participation of some countries could be higher if providing the questionnaire in additional languages. Definitions for each injury diagnosis were not included in the questionnaire, which could have potentially lead to respondents incorrectly stating their respective injury diagnoses. Finally, despite the large sample size, we had a larger proportion of respondents that were males than females. Future research focused on females and the potential risk factors associated with cycling is warranted.