White matter abnormalities are associated with chronic postconcussion symptoms in blast-related mild traumatic brain injury

Abstract

Blast-related mild traumatic brain injury (mTBI) is a common injury among Iraq and Afghanistan military veterans due to the frequent use of improvised explosive devices. A significant minority of individuals with mTBI report chronic postconcussion symptoms (PCS), which include physical, emotional, and cognitive complaints. However, chronic PCS are nonspecific and are also associated with mental health disorders such as posttraumatic stress disorder (PTSD). Identifying the mechanisms that contribute to chronic PCS is particularly challenging in blast-related mTBI, where the incidence of comorbid PTSD is high. In this study, we examined whether blast-related mTBI is associated with diffuse white matter changes, and whether these neural changes are associated with chronic PCS. Ninety Operation Enduring Freedom/Operation Iraqi Freedom (OEF/OIF) veterans were assigned to one of three groups including a blast-exposed no--TBI group, a blast-related mTBI without loss of consciousness (LOC) group (mTBI--LOC), and a blast-related mTBI with LOC group (mTBI + LOC). PCS were measured with the Rivermead Postconcussion Questionnaire. Results showed that participants in the mTBI + LOC group had more spatially heterogeneous white matter abnormalities than those in the no--TBI group. These white matter abnormalities were significantly associated with physical PCS severity even after accounting for PTSD symptoms, but not with cognitive or emotional PCS severity. A mediation analysis revealed that mTBI + LOC significantly influenced physical PCS severity through its effect on white matter integrity. These results suggest that white matter abnormalities are associated with chronic PCS independent of PTSD symptom severity and that these abnormalities are an important mechanism explaining the relationship between mTBI and chronic physical PCS.

Keywords: OEF/OIF; PTSD; diffusion tensor imaging; loss of consciousness.

Figure 1

Mean number of clusters with reduced FA for each group. Note: figure displays number of clusters with reduced FA that have been square root transformed. * = significant group difference from the no ‐ TBI group, P = 0.011. Error bars refer to standard error of the mean. FA = fractional anisotropy; mTBI = mild traumatic brain injury; LOC = loss of consciousness.

Figure 2

Distribution of clusters with reduced FA within groups. Scale represents the proportion of individuals within each group that have a cluster of reduced FA in the area marked. FA = fractional anisotropy; mTBI = mild traumatic brain injury; LOC = loss of consciousness.

Figure 3

Physical postconcussion symptoms are associated with the number of clusters with reduced FA. Individuals with a greater number of clusters with reduced FA had more physical PCS. Note: PCS represents mean score (range 0–4) and number of clusters with reduced FA is square root transformed. FA = fractional anisotropy; PCS = postconcussion symptoms.

Figure 4

The number of clusters with reduced FA mediates the relationship between mTBI + LOC and physical PCS. Individuals in the mTBI + LOC group had a greater number of clusters with reduced FA, which in turn was associated with greater physical PCS severity. Numbers shown represent unstandardized coefficients. Solid lines indicate significance (P < 0.05). Gray indicates covariates in the model. FA = fractional anisotropy; PCS = postconcussion symptoms; mTBI = mild traumatic brain injury; LOC = loss of consciousness; CAPS = Clinician‐Administered PTSD Scale.