Table 3 Results of statistical models with a significant overall effect of group (NSSI+, NSSI−, CTL) on network coherence with planned linear contrast tests (NSSI+ versus NSSI− and NSSI+ versus CTL).

Model	Statistic	p (corrected)	p (uncorrected or posthoc)	B±SE	Δ R²
DMN-A Coherence
Group	F(2,63) = 6.835	0.007	0.002**		0.178
NSSI+ versus NSSI−	t(63) = −2.223		0.030*	−0.348 ± 0.157
NSSI+ versus CTL	t(63) = −3.568		0.0007***	−0.622 ± 0.174
DMN-P Coherence
Group	F(2,63) = 5.333	0.016	0.007**		0.145
NSSI+ versus NSSI−	t(63) = −1.052		0.297	−0.236 ± 0.225
NSSI+ versus CTL	t(63) = −3.261		0.002**	−0.814 ± 0.250
NSSI− versus CTL	t(63) = −2.165		0.034*	−0.578 ± 0.267
DMN-V Coherence
Group	F(2,63) = 4.102	0.037	0.021*		0.115
NSSI+ versus NSSI−	t(63) = −2.456		0.017*	−0.459 ± 0.187
NSSI+ versus CTL	t(63) = −2.279		0.026*	−0.473 ± 0.207
SN-Ins Coherence
Group	F(2,63) = 8.833	0.003	0.0004***		0.219
NSSI+ versus NSSI−	t(63) = −3.329		0.001**	−0.731 ± 0.220
NSSI+ versus CTL	t(63) = −3.618		0.0006***	−0.883 ± 0.244

Posthoc linear contrast tests (NSSI- versus CTL) with significant results are also reported. All results shown include age, sex, motion, and medication use as covariates (see Table S2 for more details). See Fig. 3 for more details. FDR-corrected p-values are reported for the main effect of group only.
NSSI non-suicidal self-injury, CTL healthy control, DMN-A Anterior Default Mode Network, DMN-P Posterior Default Mode Network, DMN-V Ventral Default Mode Network, SN-Ins Salience Network-Insula.
*p < 0.05, **p < 0.01, ***p < 0.001.

Quick links

Search