Supplement
Non-participating Children
Eight anxious families who participated in the therapy program declined to participate in the research study. Table 1 presents the child characteristics for these clinical participants.
Excluded Children
Twenty-three anxious families who participated in the therapy program and research study, but did not meet the study inclusion criteria, were excluded from all subsequent data analyses: seven clinical children were rated below the clinical range of the CBCL Internalizing Scale (T-score < 63) at pre-treatment, one clinical child’s pre-treatment assessment occurred after the third therapy session, seven clinical children attended less than 75% of the therapy sessions or dropped out of the therapy group, one child completed the therapy group but did not return to the laboratory for a post-treatment assessment, five children had unusable EEG data due to technical difficulties, and two clinical children had fewer than ten usable correct no-go trials per emotion condition. Tables 2 and 3 present the child characteristics and questionnaire data for these clinical participants. Four comparison children were excluded from the study as well: three children were rated above the normal range of the CBCL Internalizing Scale (T-score ≥ 60), while one child had unusable EEG data due to technical difficulties.
EEG Net Application
EEG was recorded using a 128-channel Geodesic Sensor Net (electrode map presented in Figure 1). The circumference of each participant's head was measured, and using a sizing chart, the appropriate net was selected. The intersection between the midpoint of the nasion/inion line and the midpoint of the preauricular line was used to mark the vertex. The net was then applied, ensuring that electrode Cz was aligned over the vertex, and the nasion electrode was aligned over the nasion. The chin straps were then tightened and adjusted to secure the net. Each individual electrode, interconnected with lines of elastomer, was then moved side to side to ensure direct contact with the scalp.
Comorbidity
To examine whether the finding that higher P1 amplitudes predicted non-improvement following therapy was influenced by comorbid disorders such as ADHD/ ODD, a supplementary set of analyses was conducted with the ADHD/ODD participants removed from the regression analysis (n=19). The Wald chi-square for the coefficient associated with pre-treatment P1 amplitudes was statistically significant, B = .201, c2 (1) = 4.36, p = .037. Exp(B) for pre-treatment P1 amplitudes was 1.22, indicating that for each microvolt increase in pre-treatment P1 amplitudes, the predicted odds of non-improvement following treatment was approximately 22% higher. The 95% CI for exp(B) ranged from 1.01 to 1.48. In summary, even with these participants removed, high P1 amplitudes predicted non-improvement.
Table 1
Characteristics for non-participating clinical children
Clinical Children
(n=8)
Age (years) 10.34 (1.47)
Gender (M:F) 3:5
Anxiety Disorder Diagnosis (n)
GAD only 3
SEP only 2
GAD & SOC 1
GAD & SEP 1
GAD, SOC & SEP 1
Table 2
Child characteristics for excluded clinical participants
Clinical Children
(n=23)
Age (years) 10.03 (1.10)
Gender (M:F) 11:12
Anxiety Disorder Diagnosis (n)
GAD only 7
SOC only 2
SEP only 2
SOC & SEP 1
GAD & SOC 2
GAD & SEP 7
GAD, SOC & SEP 1
Unknown 1
Table 3
Questionnaire measures for excluded clinical participants
Clinical Children
Pre-Treatment Post-Treatment
(n=23) (n=19)
CBCL Internalizing Score 63.48 (12.09) 57.42 (9.81)
MASC Total Score 54.61 (11.07) 52.11 (10.45)
STAIC S-Anxiety Score 48.74 (11.91) 42.00 (10.17)
GIS Score n/a 3.26 (1.41)
Note: CBCL: Child Behavior Checklist; MASC: Multidimensional Anxiety Scale for Children; STAIC-S: State-Trait Anxiety Inventory for Children; GIS: Global Improvement Scale.
Figure S1. 128-channel map used for EEG recordings.
1