Postural Control in Children and Adolescents After Suffering From a Concussion

Overview

Sport-related concussions occur during different types of sport and are still an underestimated brain injury. Especially children are affected due to their lacking movement control and thereby at higher risk of situations leading to concussion. However, research about the rehabilitation of balance and coordination in children after sustaining a concussion is lacking. Therefore, the return-to-sport question cannot be answered reliable due to the missing understanding of the underlying mechanisms disturbing coordination, yet. Analyzing postural control, meaning the ability of the body controlled by the brain to maintain balanced, is suggested to be a valid method to investigate movement coordination. A newer method to analyze postural control using reflective marker data will be used to study the rehabilitation process. The findings may help to improve concussion treatment and give implication to the return-to-sport decision. The investigators expect to see an altered postural control after sustaining a concussion visible in the movement amplitude especially short after the injury. Moreover, the researchers assume coordination patterns which are not visible to be altered for an extended time period of up to 30 days as well. Participants will be children aged 10 to 16 years and the aim is to recruit 30 children and adolescents who suffer from a concussion. The data of the concussed participants will be compared with data of healthy volunteers.

Concussions are a major public health issue and occur in many sports, for example soccer, basketball, or American football. After sustaining a concussion balance is affected, often for weeks. The return to motor and cognitive tasks frequently occurs as soon as postural stability seems to be recovered. However, a research group around Cavanaugh investigated already in 2005 changes in postural control by analyzing the center of pressure (COP) movement. Their findings suggest that postural sway variables cannot be interpreted as the only parameters regarding rehabilitation of balance, since postural control was still noticeably affected after the sway variables returned to normal levels. They further state that athletes who return too early back to sport may still be "unstable" and are likely to be more vulnerable than healthy athletes. Furthermore, a compromised balance may lead to falls, shown in several studies, which found correlations between weak balance and the risk to fall. An additional and very important issue is that a second impact to the head of a not fully recovered person can lead to cumulative effects. Athletes with successive concussions showed significantly lower performance on memory testing than athletes with only one concussion. Consequently, it was stated that the "return-to-field" or "return-to-play" decision, for instance in rugby, is one of the most important questions to solve in the context of concussions in sport. However, despite the increasing research interest in the last decade, a very recent systematic review conducted by McLeod still highlights the difficulty of diagnosing a concussion and of making "return-to-action" decisions. Due to their natural need of movement, adolescents and children are specifically at risk of returning to unsafe situations too early after sustaining a mild traumatic brain injury (mTBI). Hugentobler pointed out that age seems to have a major effect on common used post-concussion postural control assessments. During the ten competition days of the Youth Olympic Games in Innsbruck 2012, 7,2% of the injured athletes suffered from a concussion. Research into concussions often focuses on collegiate athletes; fewer studies are available on adolescents or children. Researchers observed that high school children had prolonged memory dysfunction after a concussion compared to college athletes. The reason for this age difference might be the different brain structure of children compared to that of adults, observed as overall increase in white matter volumes, regional differences in gray matter mass and also volumetric differences. These age-related distinctions may lead to different alterations in postural control in children, maybe due to not fully automated motion patterns, as shown for hand movements during writing. Hence, while the study of Cavanaugh already identified limitations in the standard assessment of concussion-related deficits in the postural control of healthy adults, it remains an open research question, how the postural control system of children or adolescents is affected and how the recovery process develops in this age group in which postural control is less automated. Furthermore, assessments of the COP, as in the study of Cavanaugh, give an indication that changes exist, but cannot answer the question which control mechanisms are affected. For example, concussions might disturb the coordination of postural control movements, e.g. change the coordination between hip and ankle strategies, or concussions could affect how the postural movements are controlled, e.g. slower reaction mechanisms or slower or false anticipatory mechanisms. None of these mechanisms can be distinguished in an assessment of the COP excursion. However, a whole body analysis of kinematic marker data using a principal component analysis (PCA) is a more detailed approach and proved to be reliable for investigating the mechanisms that play a role in human postural control. The method offers two new variables: 1) Analysis of movement components and their relative contribution to the overall motions needed to maintain balance (PC-eigenvalues) and 2) principal accelerations (PA), which can be interpreted as motor control actions and therefore facilitate a deeper insight into what aspects of the postural control processes may be compromised. The proposed study has two main goals. The first goal is to better understand the effects of concussions on balance and postural control in adolescents and children. Since postural control in this age group is less automatized and due to several other structural differences in the nervous system between adolescents and adults the investigators expect that concussions might affect their postural control differently. It is also expected that the recovery process may differ. The second goal is to develop a better understanding of how concussions compromise postural control by applying a novel analysis technique based on principal component analysis (PCA). The researchers expect that the PCA will be better able to distinguish effects on the coordination of segment movements from effects on how movement components are controlled. Specific hypotheses that will be tested are, for example, 1. Existing results suggest that the recovery of postural control after a concussion occurs in different phases. Early in the recovery process (three days to two weeks), postural sway amplitudes, which are substantially increased after a mTBI, return to normal levels. Yet, how postural movements are controlled (measured through the entropy of the COP movements) shows abnormalities for several weeks. The investigators hypothesize that the early phase will be characterized by disturbed coordination of movement components (quantifiable through PC-eigenvalues). The late recovery phases will be characterized by timing issues in the control of individual movement components, which can be detected in the movement component accelerations (PAs). 2. The PAs might also reveal underlying mechanisms of how concussion affects the neural postural control system. For example, if concussion prolongs sensorimotor delays, then one might expect less frequent activity in the PAs. 3. The opposite behavior might also occur: if the concussion disturbs automatized control processes in the brain, then one might hypothesize that more cognitive processes need to be involved in postural control after a concussion. This might manifest in more frequent changes of the PAs. To analyze the data in the prescribed way a principal component analysis (PCA) will be conducted with the kinematic data of the participants, representing the whole kinematics of the movement. The idea is to divide the movement into many one-dimensional principal components (PCs) with different impact on the whole movement. This method is valid for comparing the measured variance of center of pressure (COP) data of participants standing in quiet stance, with the variance given by calculated PCs as resulting from a PCA. The COP variance was explained by the resultant principal movements (PMs) better than other methods did so far. The first 15 PMs explained 99.3% of the postural variance during the quiet stand. Every PC described a little part of the whole movement and had its own impact on the overall motion called "eigenvalue". The method is therefore proven to be able to detect even small adjustment movements performed by the motor control system during quite stance. This ability of revealing very small adjustments enables it to be used as method investigating movement patterns. All statistical analyzes will be performed using the Statistical Package for the Social Sciences (SPSS) with the alpha-level set to 0.05. A Shapiro-Wilk test will be used to ensure the normal distribution of the data. The comparison of the PC-values as main output will be a within-subject analysis and therefore the used method is going to be a one-way repeated analysis of variance (rANOVA) to evaluate the data provided in case of normal distribution. If normal distribution is not given, a Friedmann test will be utilized. Applied Post-Hoc tests will be conducted using a Sidak correction. The sample rate is dependent on the amount of children and adolescents suffering from a concussion in the planned time period of the study that moreover want to participate. However, based on the study of Cavanaugh a power-analysis was conducted for the planned repeated measures ANOVA, within factors. The α-error was set to 0.05 and the expected power was set with 0.95. The effect size of 0.73 was calculated using the mean of difference in the concussed group (0.19) and the standard deviation of difference (SD = 0.26) calculated by multiplying the given standard error of the mean (SEM = 0.05) with the root out of the sample size (n = 27). Power analysis stated a required total sample size of 27 participants based on the calculations. In case a dropout-quote of 50% would occur and would slim the sample size to only 15 participants, this would yield to a power of 75% based on our calculations.