63 A Multimodal Investigation of Attention in Pediatric Concussion

Abstract

Objective:

Concussion patients frequently report increased distractibility, with more than half endorsing “concentration difficulty”. Previous studies have demonstrated impairments in maintaining attention and voluntary attentional allocation in concussion patients. However, involuntary attentional allocation (distraction) is not well understood in the context of concussion. The goal of this study was to examine distraction in acute pediatric concussion patients, monitoring frontal lobe activity using functional near-infrared spectroscopy (fNIRS) - a noninvasive measure of local hemodynamic activity - to elucidate whether post-concussion distractibility is associated with the availability of attentional control resources.

Participants and Methods:

Participants included concussion patients (cases; n=19) presenting to specialty care within 28 days of injury (M=8.05, SD=5.55) and controls (n=16) presenting for reasons other than concussion. Participants were 13-17 years old (M=14.83, SD=1.10) and 57.1% female. Participants completed a computerized measure of behavioral distraction (the additional singleton paradigm) while frontal lobe activity was recorded using fNIRS 4-channel split sensor. On each trial, an array of shapes (five squares and one circle) was presented, and participants reported the orientation of a line segment inside a target shape (circle). The search array included a distractor (a square that differed in color) on 50% of trials. For each participant, the fNIR signal for epochs of each trial type (distractor present/absent) were averaged and subjected to a linear regression in which the data were fitted to a hemodynamic response function (HRF).

Results:

34 participants (19 cases, 15 controls) were included in our behavioral analysis. Reaction time (RT) was significantly slower on distractor present compared to distractor absent trials; F(1,32)=17.151, p<.001. There was no significant effect of group (case/control) on RT (F(1,32)=1.24, p=.273) or interaction between group and trial type (F(1,32)=1.05, p=.313). 29 participants (15 cases, 14 controls) were included in fNIRS analyses. The effect of group and distractor presence/absence on oxygenated hemoglobin (HbO2) was examined for each channel. A significant effect of distractor presence/absence was observed in channel 3; F(1,27)=8.510, p=.007. There were no significant effects of group or interactions between group and distractor presence/absence.

Lastly, a capture index was calculated for each participant by subtracting average RT on distractor absent trials from distractor present trials and correlated with HbO2 (beta weights averaged across trial type) for each group at each channel. No significant correlations were observed. There was a trend towards a negative correlation for case participants, particularly in channel 1, which strengthened when an outlier was removed (r=-.407, p=.149).

Conclusions:

Reaction time and frontal lobe activity - which serves as a proxy for attentional control resources - were significantly higher when a distractor was present. Although there were no significant differences in behavioral distraction between groups, concussion patients trended towards higher levels of frontal lobe activity. Likewise, although not statistically significant, there was a trend towards a negative correlation for cases such that more attentional control resources (i.e., higher frontal lobe activity) was associated with less behavioral distraction (i.e., smaller capture index). This suggests that concussion patients may recruit more neural resources to produce comparable behavioral responses to healthy controls.