Article Text
Abstract
Background Female recreational skiers have twice the rate of knee injuries and three time the rate of ACL injuries compared with their male counterparts. Female skiers suffering from a knee injury reported a significantly higher proportion of failure of binding release than knee injured male skiers.
Purpose To evaluate factors associated with failure of binding release among ACL injured male and female recreational skiers.
Study design Cohort study.
Methods Among a cohort of 498 recreational skiers (68% females) suffering from an ACL injury (complete rupture or partial rupture), age, sex, height, weight, self-reported skill level and self-reported risk taking behaviour, gear origin, ski length, date of last binding adjustment, perceived speed at the moment of injury, type of fall, and failure of binding release of the ski of the injured knee, were collected by questionnaire.
Results Failure of binding release was reported within 78% of cases and was significantly higher for females compared to males (83 vs 66%, p<0.001) with an adjusted OR of 2.7 (95% CI 1.7 to 4.4). A higher perceived speed at the moment of injury was significantly associated with a decreasing proportion of failure of binding release. A slow perceived speed was independently associated with failure of binding release (adjusted OR 2.0; 95% CI 1.2 to 3.5). There was a significantly higher proportion of failure of binding release during backward falling compared to forward falling (87 vs 72%, p=0.002); similarly, a higher proportion of failure of binding release occurred in cases of complete rupture compared with a partial tear of the ACL (81 vs 64%, p=0.001), respectively.
Conclusions Among this cohort of ACL-injured skiers, failure of binding release was significantly associated with female sex, a slow perceived speed at the moment of injury and complete rupture of the ACL.
- ACL
- Alpine skiing
- Skiing
- Speed
Statistics from Altmetric.com
Introduction
In recreational alpine skiing, the knee joint is the most common anatomical location of an injury with about one-third of all injuries sustained by male and female skiers.1–3 The most common diagnosis in injured recreational alpine skiers is a tear of the ACL with 15–21% of all injuries in adult skiers of both sexes.4 ,5 However, female recreational skiers have twice the knee injury incidence of male skiers and the ACL injury risk is three times greater in female skiers.1 ,2 ,6 This decisive sex difference in risk of ACL injuries is also seen in other sports and may be related to different hormonal, anatomical and neuromuscular risk factors.7
Menstrual cycle phase and leg dominance are potential risk factors for non-contact ACL injuries among female recreational skiers.8–11 ACL injury risk seems to be significantly higher in the preovulatory phase of menstrual cycle8–10 and female skiers show a twofold higher risk of sustaining an ACL rupture on their left leg as well as on their non-dominant leg.11
Another decisive sex difference among knee injured male and female skiers concerns the timing of release of ski bindings.12–14 In 55–67% of cases, male skiers reported a failure of binding release at the moment of injury, compared to 74–88% among female skiers.12–14 However, despite sex differences the large amount of failure of binding release causing a knee injury seems to be a major problem among alpine skiing. Therefore, we evaluated factors associated with failure of binding release among ACL injured male and female recreational skiers.
Methods
This prospective study was conducted during four winter seasons from 2009/2010 to 2012/2013 in two Austrian ski injury clinics. Inclusion criterion was a non-contact fall during skiing leading to an injury of the ACL. MRI was used for the diagnosis of partial or total tear of the ACL. Recruitment of patients aged 17 years and above was performed with a random selection. This was done independently of injury severity at all days of the week throughout the overall winter seasons. The recruitment was sometimes dependent on organisational aspects at the clinics (availability of rooms and personnel) and also the willingness of patients to volunteer. However, more than 90% of invited participants agreed to participate. We abided by the ethical standards of the 2008 Declaration of Helsinki. Informed consent was obtained from all participants prior to beginning research. In addition, the study was approved by the Institutional Review Board.
Questionnaire
Age, sex, height, weight, self-reported skill level according to Sulheim et al15 and self-reported risk taking behaviour (more risky vs more cautious) according to Ruedl et al,16 gear origin (own, ski shop rental, borrowed from family or friends), ski length, date of last professional binding adjustment in a ski shop (this season, last season and older), the own perceived speed at the moment of injury (very fast, fast, moderate, slow, very slow), type of fall, and failure of binding release of the ski of the injured knee were collected by a paper and pencil questionnaire in the ski clinic within 2 days after injury with a physician present.
Regarding self-reported skill level (expert, advanced, intermediated, beginner), we divided participants into more skilled (advanced and experts) and into less skilled (beginners and intermediates) persons as a tendency was observed to underestimate the individual skill level.15 According to a previous study14 patients recalled their type of fall by observing pictures of the most typical ACL injury mechanisms in skiing. There were four categories for the direction of the fall: (1) forward fall with body rotation, (2) forward fall without body rotation, (3) backward fall with body rotation and (4) backward fall without body rotation.14
Statistical analysis
Data are presented as means and absolute and relative frequencies. Differences in mean age, body mass index (BMI) and ski length to height ratio with regard to binding release and sex were evaluated by Mann-Whitney-U tests. Differences in frequencies were evaluated by χ2 tests. To control for confounding, variables with p values <0.1 were included in the multivariate logistic regression model (forward stepwise selection). Adjusted ORs and their 95% CIs for failure of binding release among ACL-injured skiers are reported.
Regarding variables with more than two categories (perceived speed and type of fall), each type of speed (very fast, fast, moderate, slow, very slow) as well as each type of fall (forward fall with body rotation, forward fall without body rotation, backward fall with body rotation, backward fall without body rotation) were separately coded by making dummy variables. All p values were two-tailed and values less than 0.05 were considered to indicate statistical significance.
Results
A total of 498 skiers (32.2% males, 67.8% females) suffering from a non-contact ACL injury with a mean age of 41.3±11.2 years volunteered for this study. Failure of binding to release was reported within 77.9% of cases at the moment of the injury.
In table 1, characteristics of factors associated with self-reported failure of binding release among ACL-injured skiers were shown. No differences were found with regard to age and BMI. Failure of binding release was significantly higher for females compared to males (83.4 vs 66.3%). A higher perceived speed at the moment of injury was significantly associated with a decreasing proportion of failure of binding release. In addition, the type of fall and the severity of ACL injury were significantly associated with failure of binding release, that is, higher proportions of failure of binding release during backward falls compared to forward falls and with a complete rupture compared to a partial tear of the ACL, respectively.
In table 2, sex-specific factors associated with self-reported failure of binding release are presented. Males with an ACL injury showed a significantly higher BMI and higher proportions of more skilled (72 vs 53%) and more risky (42 vs 15%) skiers compared to females. Significantly more females than males reported that their bindings were professional adjusted in a ski shop in the actual season (75 vs 65%). Regarding injury-related factors, the distribution of perceived speed showed a significant difference between sexes while the distributions of types of fall and of severity of ACL injury did not, respectively (table 2).
Results of the multivariate regression analysis including factors sex, perceived speed at the time of injury, type of fall and severity of ACL injury are shown in table 3. Factors female sex (OR 2.7) and slow speed (OR 2.0) were associated with increasing the odds of failure of binding release, while factors forward fall with rotation (OR 0.4), forward fall without rotation (OR 0.3), and a partial tear of the ACL (OR 0.4) were associated with decreasing the odds of failure of binding release.
Discussion
We evaluated factors associated with failure of bindings to release among ACL injured male and female recreational skiers. Female sex and a slow perceived speed at the moment of injury were independent risk factors associated with failure of bindings to release while forward falling with and without rotation and partial tears of the ACL were independently associated with binding release.
Why were there sex differences in failure of binding release?
The clearly higher frequency of failure of binding release among female skiers with an ACL injury (83 vs 66%) is well known from other studies investigating potential risk factors of knee injuries in recreational alpine skiing.12–14 It has to be mentioned, that to date release values of the bindings according to the ISO 11 088 standard17 are determined using individual age, height, weight, self-estimated skiing style and sole length of ski boots of the skiers without considering any sex-specific differences. Therefore, one could speculate that sexes differ with regard to the date of the last binding adjustment; Burtscher et al1 found that newly adjusted bindings clearly reduced knee injury risk among female skiers. However, among this cohort of ACL injured skiers we found no association between the last binding adjustment and binding release. We even demonstrated a significantly higher proportion of actual binding adjustment among female skiers with an ACL injury.
Another explanation for the observed sex difference might be due to different types of falls as we report a significant association between binding release and type of falling. In accordance with other studies, the most common type of self-reported fall among knee injured skiers seems forward falling with body rotation (‘valgus external rotation’).14 ,18 ,19 The larger amount of reported binding release among forward falls with and without body rotation becomes plausible because current bindings are mostly designed to release during a simple forward fall (upward release function of the heel piece) or during a forward fall with body rotation (side release function of the toe piece of the ski binding).20 However, when comparing sexes, types of fall did not differ; again in line with an earlier study by Ruedl et al.14 Therefore, it seems unlikely that the higher proportion of female failure of bindings to release is due to a different distribution of forward and backward falls among male and female skiers.
Slow perceived speed and failure of binding release
Failure of binding release was independently associated with a slow perceived speed at the moment of injury. The potential problem of ski binding release at low speed is well known.13 ,21 LaPorte et al13 found that 44% of lower leg injuries (50% of all tibia fractures, 44% of all MCL injuries, 47% of all complex knee sprains, and 43% of all ACL injuries) occurred at low speed or in a stationary position. Senner et al21 refer to a ‘landmark study’ by Merino et al22 “which confirmed the expectation that, the risk is slightly higher for false releases at higher speeds while at low speeds, however, cruciate ligament ruptures are likelier”.
Interestingly, in our study males and females differed significantly with regard to perceived speed at the moment of injury. The proportions of moderate (41 vs 36%), slow (34 vs 26%) and very slow (12 vs 6%) speed were clearly higher among females compared to males. In comparison, studies measuring skiing speed with a radar gun found an about 7 km/h lower mean speed for female compared to male skiers.23 ,24
According to the ISO 11 088 standard17 for binding values, skiers have to differentiate between skiing speed (slow to moderate vs fast), terrain (gentle to moderate vs steep) and skiing style (cautious vs aggressive) to classify themselves into one of three skiing types.17 Let's assume a male and a female of equal age, height, and weight and of equal ski boot sole length also both classified themselves as type-3 skier (fast speed, steep terrain, aggressive style). They both would get the same binding setting values without considering sex.
At least two potential sources of error could explain the higher rate of failure of binding release among female skiers. First, a recent study by Brunner et al25 found that males, more skilled skiers and risky skiers perceived their actual speed as fast, moderate and slow when skiing up to 10 km/h faster than females, less skilled and cautious skiers. Therefore, one might suspect that compared to a self-reported ‘slow to moderate’ or ‘fast’ skiing male, the binding setting for a self-reported ‘slow to moderate’ or ‘fast’ skiing female is too high. This would result in a higher rate of failure of binding release as sexes seem not to differ with regard to date of last binding adjustment,14 incorrectly adjusted bindings,26 or self-reported types of falling.14
Second, a study by Werner and Willis27 found that muscle strength is highly correlated with the ability to release the binding in a self-release test. With regard to the equal weight of the male and female skier in the aforementioned example it has to be considered that the weight-to-strength ratio is negatively influenced by the higher fat mass in females28 maybe partly explaining the sex difference in the lack of binding release due to less muscular strength among females.
In addition, in a pilot study (8 males, 9 females) presented at the 21st International Congress on ski trauma and skiing safety, Posch et al29 found that about one-third of females compared to about 88% of males released their bindings at least once in a self-release test. They found significant correlations between body weight and isometric strength of the dominant leg (r=0.90, p=0.003) and between body cell mass (measured via bioimpedance analysis) and isometric strength of the dominant leg (r=0.78, p=0.021) among males while no such significant correlations were found for the females.29
Time to include sex in the ISO binding setting values?
Taken together, our data provide strong arguments for discussing the implementation of a sex factor within the ISO binding setting values. This idea, however, is not new.30 Opponents argue that reducing the binding release values would not result in fewer ACL injuries as the primary injury mechanism, the so-called ‘phantom foot’ (ie, flexion-internal rotation), is not related to binding function. Those opponents suggest low settings for binding release would increase the number of inadvertent releases.31
We disagree with that claim as several studies14 ,18 ,19 found that since the introduction of the short and shaped carving skis the primary self-reported ACL injury mechanism is forward falling with body rotation (‘valgus external rotation’) which is directly related to binding function. Also, Senner et al21 did not support the conclusion from a work by Shealy et al32 on using signal detection theory as a model to evaluate release/retention criteria in alpine skiing that the reduction of retention values would inevitably increase the overall risk of injury due to a higher rate of inadvertent releases. In addition, LaPorte et al30 reported no increase in injuries from inadvertent binding release among females with lower binding release values set 15% below those recommended by the ISO 11088 standard compared to males.
Practical steps—ISO binding settings
Interestingly, the ISO 11088 standard accepts a deviation of 15% between the measured release moment (‘reference moment’) determined according to the setting tables in ISO 8061 and ski binding settings may also lowered by the same magnitude on request of the skier.22 However, this is largely unknown among the overall skier population, but could represent a potential preventive measure, especially for female recreational skiers.
Failure of binding release in this study was also associated with higher proportion of a complete rupture of the ACL compared to a partial tear (81 vs 64%). The higher grade of injury severity seems plausible because during a non-contact fall, failure of ski binding release means the ski acts as a lever at the knee joint for a longer period of time than it would if the binding released.
Limitations
At least two limitations should be considered. First, a selection bias due to the specific type of patient recruitment cannot entirely be excluded. However, at least with regard to the age and sex distribution our sample seems to be representative for the total skier population aged over 17 years suffering from knee injuries in Austrian skiing regions.33
Second, we acknowledge the possibility of recall bias especially regarding the type of falling and failure of binding release may be taken into consideration. To control for putative confounding, multivariate logistic regression analysis has been performed. Although the authors have many years of experience in collecting ski data and performing ski studies, the existence of unknown confounders, for example, previous knee injuries, cannot be ruled out. Strengths of the study include the large number of skiers suffering from ACL injury.
In conclusion, among this cohort of skiers suffering from ACL injury (partial rupture and complete rupture combined), failure of binding release was significantly associated with female sex, a slow perceived speed at the moment of injury and the occurrence of complete rupture of the ACL. We recommend ski binding adjustment for female skiers should be revisited so that skis release at lower forces than they do if the current ISO settings are used.
What are the findings?
Failure of binding release among this cohort of skiers suffering from ACL injury was independently associated (ie, more likely to have been the case) with female sex, slow perceived speed at the moment of injury and when the individual suffered a complete rupture of the ACL. At present the ISO standards do not take sex into account.
Binding release among this cohort of skiers suffering from ACL injury is independently associated with forward falling with and without rotation.
How might it impact on clinical practice in the future?
The ISO 11 088 standard accepts a 15% lowering of binding setting values at the skier's request. This is little known among skiers, but could represent a potential preventive measure, especially for female recreational skiers.
Implementing earlier binding release values for female skiers likely represents a very important measure of injury prevention.
References
Footnotes
Contributors All listed authors have significantly contributed to this work to justify authorship. This study was conceived and designed by GR, AS, and CF. Literature search was done by GR, KH and KT. Acquisition of data and statistical analysis were done by GR, KH, KT, AS, CF and by MB, respectively. All listed authors (GR, KH, KT, AS, CF, MB) contributed to the interpretation and discussion of the findings and participated in editing or re-writing of the article lead by GR (the guarantor).
Competing interests None declared.
Ethics approval IRB.
Provenance and peer review Not commissioned; externally peer reviewed.