Prediction and injury risk based on movement patterns and flexibility in a 6-month prospective study among physically active adults

Introduction

Physical activity has many health benefits but can lead to injury (Meeuwisse et al., 2007). Many studies confirm the prevalence of injuries that occur during physical activity among ordinary people around the world (Złotkowska et al., 2015; Marar et al., 2012; Cai et al., 2018). Injuries may be caused by incorrect warm-up, excessive physical load (Ristolainen et al., 2014), previous injuries (Giroto et al., 2017), poor quality movement patterns (Kiesel, Plisky & Voight, 2007; Garrison et al., 2015), and a low level of motor fitness, especially flexibility (De la Motte et al., 2019a; De la Motte et al., 2019b). Injuries may be predicted by identifying the determinants of injuries and preventive measures may reduce the risk of injury (Emery & Pasanen, 2019).

Cook et al. (2010) indicated that is possible to predict injury based on the quality of movement patterns assessed using the Functional Movement Screen (FMS) test (Cook et al., 2010). Numerous studies have indicated that low FMS scores (14 ≥ FMS) are associated with a higher number of injuries (Kiesel, PJ & Voight, 2007; Chorba et al., 2010; Chimera et al., 2015). An elevenfold increase in the risk of injury was found in people with low-quality movement patterns (14 ≥ FMS) (Kiesel, PJ & Voight, 2007). Bunn, Rodrigues & BezerradaSilva (2019) conducted a systematic review with meta-analysis to show that subjects with low FMS scores are more likely to suffer injury.

However, observations of various durations were conducted, providing ambiguous results. Kiesel, PJ & Voight (2007) and Duke, Martin & Gaul (2017) indicated that it is possible to accurately predict injuries when assessing the quality of movement patterns in studies spanning 4.5 months and 8 months, respectively. However, studies conducted by Słodownik, Ogonowska-Słodownik & Morgulec-Adamowicz (2018) and De la Motte et al. (2019a); De la Motte et al. (2019b) suggest that it is not possible to predict an injury within 6 months of taking an FMS test. Most studies were conducted in groups of athletes and soldiers. There are no similar studies on the average physically-active population.

Existing discrepancies indicate the need to consider other comorbid factors associated with injuries. They include the level of motor skills, especially flexibility (De la Motte et al., 2019a; De la Motte et al., 2019b). Appropriate flexibility of soft tissues allows for better absorption of forces acting on them during physical activity (Safran, Seaber & Garrett, 1989). Lower levels of flexiblity put tissues at risk for damage by these forces (McHugh et al., 1999). This is evidenced by studies conducted by Bradley & Portas (2007) and Gabbe et al. (2005) among football players, which indicated that a low level of flexibility is related to an increase in injuries.

There are no studies on the determinants of injuries that have simultaneously captured the quality of the movement patterns and flexibility, although it has been suggested that higher quality movement patterns are associated with a higher level of flexibility (Chimera et al., 2017; Yildiz, Pinar & Gelen, 2019). However, both, movement patterns and flixibility are factors predispose to injury (Kiesel, PJ & Voight, 2007; Garrison et al., 2015; De la Motte et al., 2019a; De la Motte et al., 2019b). Further study of simultaneous use movement patterns quality and flexibility should allow for a more accurate prediction of injury. Few authors have taken this approach in the literature on FMS and motor fitness assessment. Lisman et al. (2013) indicated that combining the FMS score with a 3-mile run time allowed a more accurate prediction of an injury’s occurrence. Other observations provided by Kodesh et al. (2015) indicate that there is no relationship between FMS assessment, running time at various distances, and injury.

The literature lacks data for estimating the risk of injury using the two factors as a movement patterns quality and flexibility. Available studies show inconsistent results and cannot be not applied to a large population of average people undertaking various physical activities. Lehr et al. (2013) indicated that a more comprehensive assessment may be effective in predicting injury.

Therefore, the aim of this study is to determine the level of injury risk and the accuracy of injury prediction during a prospective 6-month period based on the quality of movement patterns and level of flexibility among average physically active young adults. Obtained results should identified factors increase the risk of injury and predictive possibilities, allowing to select appropriate methods to reduce the risk of injury during physical activity.

Materials & Methods

Results

The Cronbach’s alpha coefficient for IHQ was 0.836. This indicates very good reliability and repeatability of the survey (Aron, Aron & Coups, 2012). Functional Movement Screen - Compliance of self-assessment (video recordings) ICC (2.1): FMS overall assessment - 0.96; Repeatability of self-assessment (re-assessment) ICC (3.1): FMS overall assessment - 0.95.

Table 1 shows descriptive statistics of the studied group.

Table 2 presents the incidence of injuries in the retrospective and prospective periods in the research group.

The FMS score cut-off value was 14 points (Fig. 1). An increased frequency of injuries in the retrospective period, 12 months prior to the study was observed in subjects with a result equal to or lower than the FMS test cut-off point. On this basis, the participants were divided into groups of high-quality movement patterns (14 <FMS) in whom injuries occurred less frequently and people with low-quality movement patterns (14 ≥FMS) in whom injuries occurred more often.

The cut-off point value for the level of flexibility lower back and hamstrings was 15 cm (Fig. 2). An increased frequency of injuries in the retrospective period 12 months prior to the start of the study was observed in subjects with a score equal to or below the flexibility cut-off point. On this basis, the participants were divided into groups with a high (15 cm <Sit-and-reach score) and low (15 cm ≥Sit-and-reach score) level of flexibility.

The quality of movement patterns and quantitative and qualitative (dichotomous) flexibility and previous injury state (injury retrospectively) were included in analyzing the risk of injury. Table 3 shows the logistic regression models. Subjects with low-quality movement patterns (14 ≥FMS) were nearly seven times more likely to experience an injury than people with high-quality patterns. This model is statistically significant. A one unit (1 cm) decrease in flexibility may increase the risk of injury by 6%. Additionaly, subjects who suffered injuries in the retrospective period were 6.4 times more likely to suffer injury again prospectively. However, there was no reliability to the interactions between the above-mentioned factors. The quantified flexibility factor was used to show that the logistic regression model is statistically significant and indicates an injury’s predictive potential. The two-factor predictive models are not statistically significant. Neither model demonstrates predictive reliability.

The prediction of an injury’s occurrence was verified with the quality factors of movement patterns and flexibility of lower back and hamstrings. The division of the subjects was determined based on ROC curve models for the quality of movement patterns (Fig. 1) and flexibility (qualitative approach) (Fig. 2). Injury was expected for low-quality movement patterns (14 ≥FMS) and low levels of flexibility quantitative (15 cm ≥Sit-and-reach score). However, no injuries were expected with high-quality movement patterns and a high level of flexibility. The accuracy of the predictions was verified with the actual state (occurrence of an injury or its absence) and the predicted state (the expected injury or its absence) using the data on injuries sustained during the observation.

Table 4 compares the groups for injury occurence in terms of the predicted state and the actual state using the ROC curve model for the quality of movement patterns from data prospectively collected during the 6-month observation period.

The accuracy of predictions based on the quality of movement patterns was 73.18%. The χ2 statistic value indicates that at the significance level of 0.0001, we can reject the hypothesis about the independence of the predicted and actual state with the occurrence of an injury. However, in the case of lower back and hamstrings flexibillity, the accuracy of predictions was 41.14%. The χ2 statistic value indicates that at the significance level of 0.6708, we cannot reject the hypothesis about the predicted state’s independence and the actual state with the injury occurrence.

Discussion

The risk of injury is ubiquitous, regardless of the type and level of physical activity undertaken (Złotkowska et al., 2015; Marar et al., 2012; Cai et al., 2018). Our study sought to determine the risk of injuries among young, physically active adults and the assessment of the accuracy of injury prediction was based on the quality of movement patterns and the level of flexibility. The ability to accurantely predict injury allows participants to take preventive measures to avoid injury (Emery & Pasanen, 2019). Garrison et al. (2015) demonstrated the accuracy of the prediction of injury based on the FMS assessment. A significant level for injury risk was observed among people with an overall FMS score below the cut-off point. The risk of injury was nearly six times higher in those with lower quality movement patterns than in those with a higher quality of movement patterns. In the case of people with previous injuries, their risk increased 15 times. This result is almost two times higher than our research, where the risk of injury in the case of low-quality movement patterns (14 ≥ FMS) was seven times higher. However, research by Garrison et al. (2015) was conducted on athletes. Similarly, in a study of rugby players Duke, Martin & Gaul (2017) indicated that those characterized by a lower quality of movement patterns were more than ten times more likely to be injured. The risk of injury was correlated with longer breaks from training due to injuries sustained during the season. A study by Chorba et al. (2010) conducted on 38 athletes showed a 4-fold increase in injury risk with poor movement patterns than in people with good quality movement patterns. This model accurantely predicted injuries among 69% of the examined women over a period of a year. Kiesel, PJ & Voight (2007) found an 11-fold risk of injury based on their assessment over 4.5 month, during which the accuracy of the prediction of injury was indicated at 51%. Our research indicated a 73% accuracy of prediction during the observation period. Some authors indicate lack of accuracy prediction based on the FMS assessment. De la Motte et al. (2019a); De la Motte et al. (2019b) used several functional tests, including FMS, but did not indicate the possibility of predicting injury in a large group of over 1,600 soldiers during the 180-day observation period. Słodownik, Ogonowska-Słodownik & Morgulec-Adamowicz (2018) examined injuries among handball players retrospectively (12 months) with the FMS assessment and 6 months after (observation period), and did not show simple relationships between the quality of movement patterns and injuries. This suggests that other factors related to the risk of injury should be considered. Literature reports indicate a strong relationship between flexibility and injury (De la Motte et al., 2019a; De la Motte et al., 2019b). Witvrouw et al. (2004) showed that footballers with lower muscle flexibility, manifested by weaker ranges of motion in the joints, were more likely to be injured than those with standard values. In the following years, these results were confirmed by Bradley & Portas (2007), also indicating that greater flexibility protects against injuries.

The predictive model in our research on the flexibility factor showed that a decrease in its value by one cm was associated with a 6% increase in injury risk. Witvrouw et al. (2003) indicated that preseason measurements of the level of flexibility among soccer players may be a valuable diagnostic tool in predicting injury. Prospective studies of young male and female runners showed associations between less flexibility and overload injuries (Yagi, Muneta & Sekiya, 2013).

However, as Arnason et al.(2008) showed that flexibility alone is not enough to prevent injury. A holistic approach is essential, including appropriate training of other motor skills. This explains the lack of accuracy of the predictive model in our research, where 41.14% of injuries were correctly predicted based on flexibility. Similarly, Van Doormaal et al. (2017) did not show the possibility of predicting the occurrence of an injury in the next 12 months on this basis when examining the flexibility of 450 amateur soccer players.

However knowing the determinants of injuries and the possibility of their diagnosis, there are many coexisting factors that often intertwine, which may ultimately determine the possibility of predicting an injury (Moore et al., 2019). Despite the reports (Chimera et al., 2017; Yildiz, Pinar & Gelen, 2019) indicating the relationship between the quality of movement patterns and flexibility, few researchers have considered the common influence of these factors on an injury.

We found that participants injured in retrospective period were 6.4 times more likely to suffer injury again in a prospective period. Unhealed injuries may cause constant pain, become reinjured and worsen, and may cause further injuries (Meeuwisse et al., 2007; Swenson et al., 2009).

The recognition of the connection between the quality of movement patterns and flexibility of the lower back and hamstrings could enable a more accurate prediction of an injury’s occurrence. However, until now, attempts have been made to predict injuries based on the quality of movement patterns using other motor skills. Lisman et al. (2013) indicated that a slower time in the 3-mile run combined with lower quality patterns was more strongly associated with injuries. People who had weaker scores in both areas were four times more likely to be injured. Kodesh et al. (2015) made a similar attempt to assess the determinants of injuries. However, their research showed the relationship between injuries and endurance without indicating the relationship between FMS scores and injuries in a group of 145 female soldiers. We attempted to define the determinants of injuries taking into account quality of movement patterns and flexibility of lower back and hamstrings. However, the simultaneous treatment of the quality of movement patterns and flexibility did not constitute a reliable multi-factor predictive model.

Conclusions

The risk of an injury increases with low-quality movement patterns and a low level of flexibility of the lower back and hamstrings. Prior injuries increase the risk of further injuries and full recovery is important. Exercises that improve the flexibility and quality of movement patterns and promote the full healing of injuries as preventive measures may effectively reduce the risk of injury. However, further investigation is nedeed. The quality of movement patterns is an accurate predictor for the risk of injury and is applicable here. Flexibility is a weak predictor after six months of observation and should not be used as a predictive tool.

Supplemental Information