A comparison of cognitive-motor integration performance and resting state functional brain network connectivity in female athletes suggests intact motor network and visuomotor skill in those with a concussion history

Structural neural changes following concussion are often not captured by standard imaging techniques. However, there is growing evidence that damage to white matter tracts and change in functional network connectivity may be observed following concussive injury. We investigated behavioural performance on a cognitive-motor integration (CMI) task in conjunction with alterations in resting state functional connectivity (rs-FC) in brain networks in a population of 30 female varsity athletes, with 16 having a previous history of concussion. Behavioural performance on accuracy, timing, and trajectory measures of a CMI task were assessed between the concussion history group and the control group. Rs-FC within the nodes of the Default Mode Network (DMN), Dorsal Attention Network (DAN), the Frontoparietal Network, and the Anterior Cerebellar Lobule Network, was assessed against performance scores on accuracy, timing, and trajectory measures. Main findings indicate no difference in behavioural performance between those with concussion history and those without, in contrast to previous findings in a group of primarily male varsity athletes. In addition, no difference in rs-FC was noted in correlation with behavioural performance scores on either accuracy, timing, or trajectory. These findings may suggest sex-related differences in performance on a CMI task, and a resiliency in both functional network connectivity and visuomotor skilled performance in varsity female athletes.

a process known as cognitive-motor integration (CMI). CMI is essential for the completion of 48 skilled activities that involve non-standard mappings, which are relationships used when the visual 49 information guiding the motor task does not come from the location of object one is interacting 50 with (12,13). These mappings rely on different neural computations compared to standard, direct 51 mapping (the spatial location of the viewed object is the same as the motor goal) that must 52 incorporate the spatial dissociation of gaze, attention, and overt motor output. CMI is often 53 required for complex skills, making it crucial for work, duty, sport, and daily life. Importantly, we 54 have observed impaired CMI in youth and adolescent athletes who had experienced a concussion 55 but were asymptomatic (using self-report) at the time of testing (14-17). These individuals 56 included children and adolescents who played sport at the recreational and select level (14-16), as 57 well as adolescents who played at an elite level (17). Notably, in a previous behavioural study of 58 varsity athletes, we observed impaired CMI performance in a sample of primarily males (16/18 59 participants) with a history of concussion who were asymptomatic at testing, relative to peers with 60 no history (16). In all cases, when having to incorporate some form of cognition and non-standard 61 mapping into their movement control during an eye-hand coordination task, there were significant 62 behavioural performance deficits in either the timing, trajectory formation, or both relative to peers 63 with no concussion history. These data suggest that the brain networks used for cognition and 64 motor control are affected by concussion along a timeline longer than for tasks typically used to 89 the brain activity of elite performers -in this case video gamers -during an integrated cognitive-90 motor task, Granek et al. (20) found additional prefrontal cortex activity during the planning stages 91 of movement, as well as a later onset of preparatory activity (19). In the context of concussion 92 recovery in experienced performers, our group has observed that number of years of sport 93 experience is correlated with recovery, such that those who have been playing their sport longer 94 demonstrated faster behavioural recovery relative to less-experience athletes (26). These data 95 suggest that sport experience mitigates concussion-related skilled behavioural impairment, 96 putatively related to 'neural efficiency'. The underlying brain activity controlling cognitive-motor 97 integration in experienced athletes with a concussion history has not been studied, however, in 98 order to directly explore this idea. 99 The purpose of the present study was to investigate whether there was a correlation between 100 cognitive-motor integration performance and functional network connectivity in select athletes as 101 a function of concussion history. Due to the sex-related differences in brain network activation 102 during equivalent cognitive-motor skill performance in healthy individuals (27), this initial study 103 focused on data collected from female athletes only. We hypothesize that those athletes with a 104 history of concussion (Hx -Concussion Group) will show decreased functional network 105 connectivity in networks required to successfully execute cognitive-motor integration compared 106 to those with no history of concussion (NoHx -Control group). Additionally, we expect the 107 concussion history group to show behavioural performance deficits in cognitive-motor integration 108 when compared to controls.

110
Participants fMRI imaging female athletes motor reserve  including the varsity sport they play at York University (sport), their age, the number of years they 120 have been playing said sport (years of experience), whether they have sustained a concussion 121 previously (Concussion History -n = no, y = yes), and how many previous concussions they have 122 sustained throughout their life obtained through self-report (Number Previous Concussions).

Years of Experience Concussion History Number Previous Concussions
All participants included in the study had no history of neurological problems, and no other 125 injury preventing them from practicing and playing their sport. Additionally, all participants were 126 screened using MRI safety protocols to ensure they were cleared to enter the MRI scanner.     Multi-echo rs-fMRI data preprocessing was performed using the Multi-Echo Independent

190
Components Analysis (ME-ICA) pipeline in AFNI. In order to preprocess the data using ME-ICA,  In the case of the Trajectory score, PV raw output was multiplied by negative one before adding 237 to ensure that a higher score indicated a worse performance, in line with PLf. The number of 238 direction reversal errors (DR) was analyzed separately. To include a comparison of the 239 performance as a function of their change from their own standard behaviour, composite scores 240 will be composed of delta scores (standard condition subtracted from the non-standard condition).

241
Imaging: 242 The ME-ICA denoised functional data were used in the group level analyses using  We used these subject-specific spatial maps to investigate the differences in functional

274
Since the statistical models described above did not reveal statistically significant 275 differences between groups (see Results section, below), we also tested a model with all participants collapsed into a single group to investigate whether there was an interaction between 277 our behavioural performance scores (accuracy, timing, and trajectory composite scores modeled 278 as covariates) and functional network connectivity across all participants.    No statistically significant differences were found in mean functional connectivity between 311 the Concussion history group and the control group in the two-group unpaired t-test model for any 312 of the resting state networks of interest (p>0.05 for all).

344
As noted previously, the CMI task used in this experiment has shown sex-related differences in 345 activation in the healthy brain; therefore, there is a possibility that the functional network activity 346 following concussion is impacted differently in females than in males. Interestingly, a decrease in 347 these differences of network activation in those with concussion history and those without has 348 been correlated with a decrease in symptoms (38), and with time since injury (39.40). Given that 349 all of our participants were >6 months post-injury, and deemed asymptomatic and cleared to return 350 to play, this may also provide us with a possible explanation as to why there are no differences in 351 functional network connectivity between the two groups.

352
Previous studies investigating CMI performance with and without concussion history have 353 shown an impact of concussion on CMI performance (16,17) lasting up to two years (14).

354
However, these previous studies were done with youth (14), included both males and females in a 355 single group (16), or on males only (17). Given the nature of sex-related differences in cognitive performance, and brain network activation patterns in CMI tasks, there is a possibility that 357 previously observed differences in CMI performance in those with a concussion were driven by 358 males. Indeed, in terms of recovery from concussion, there is evidence of sex-related differences 359 in both the rate of recovery and the form of recovery, although these findings are mixed (41,42).

360
In humans it has been proposed that these differences may be related to neck girth (43), hormonal (but not visuomotor skills). To this end, further research is needed to determine whether analogous 371 behavioural differences in CMI performance exist in male varsity athletes, and whether they 372 exhibit a correlation between resting state functional network connectivity and CMI performance.

373
The capacity of the brain to continue to perform without noticeable impairment until in their sport at the time of testing. Therefore, it is possible that the lack of behavioural differences 395 in performance on a CMI task between females with and without a history of concussion is derived 396 from increased strengthening of the brain networks required to successfully perform a CMI task  reserve, will this damage be increasingly evident in the long-term, or will said resiliency provide 424 a protective effect leading to decreased or slowed cognitive decline down the road.

425
Overall, this study provides a starting point for many more interesting questions 426 surrounding the sex-related and long-term effects of concussion.