Table 4.

longitudinal effects of concussion on c₁, averaged over all significant ROIs in Table 3. (left) longitudinal analysis using an LMM to compare postacute imaging sessions (SUB, RTS, 1MO, 1YR) to ACU. (Right) Cross‐sectional analysis using a GLM to compare concussed imaging sessions (ACU, SUB, RTS, 1MO, 1YR) to uninjured controls. Statistics include coefficients of effect b, 95% confidence intervals (95% CIs), BSRs, and p‐values. Significant longitudinal change relative to ACU is noted with “*” and significant cross‐sectional difference relative to controls is noted with “**” (FDR = 0.05 threshold)

	Longitudinal (LMM)				Cross sectional (GLM)
	b	95%CI	BSR	p	b	95%CI	BSR	p
ACU	—	—	—	—	−0.030	−0.077, 0.016	−1.24	.206
SUB	0.038	−0.014, 0.102	1.43	.152	−0.006	−0.047, 0.037	−0.29	.782
RTS	0.118	0.077, 0.163	5.38	<.001*	0.083	0.037, 0.129	3.58	<.001**
1MO	0.079	0.024, 0.135	2.86	.008*	0.047	0.003, 0.090	2.06	.032
1YR	0.071	0.005, 0.137	2.03	.046	0.039	−0.011, 0.086	1.58	.124

Abbreviations: BNA, Brainnetome Atlas; BSR, bootstrap ratio; FDR, false discovery rate; GLM, general linear model; LMM, linear mixed model; ROI, region of interest.