Estimates of familiality using sibling correlations
Sibling and parent-offspring correlations for EEG spectral power in frontal, central and parietal regions in each of the three conditions, EO, EC, and CPT are presented . Overall, there is a high degree of sibling similarity in EEG power, however, these data indicate differential levels of sibling correlation according to region and frequency band. Siblings are most similar at rest (EC) in the lower frequencies (theta power r2 range: .36–59, p < .001) and (alpha power range: .42–49, p < .001) and most similar under cognitive activation conditions (EO & CPT) in the beta 1 frequency band (range: .45–61, p < .001). Parent-offspring correlations were more variable, but strongest and significantly correlated in the alpha frequency band during EO and CPT conditions (range: .39–52, p < .001).
Parent-Offspring and Sibling Correlations for EEG frequency bands during baseline and cognitive activation conditions.
Association of EEG measures with ADHD affection status and subtype within families
Repeated-measures ANOVAs using ADHD affection status and subtype (Non-ADHD, ADHD-Inattentive Type or ADHD-Combined Type) were run separately within each brain region for children and parents. Among the children, significant between-subjects effects on theta power emerged in the anterior frontal [F(2, 169)=3.49, p=.03, FDR p=.12], frontal F(2,169)=4.34, p=.01, FDR p=.04], central [F(2, 169)=4.38, p=.01, FDR p=.04], and parietal [F(2, 169)=3.51, p=.03, FDR p=.12] regions, with non-ADHD children having lower theta power than children with ADHD-PI or ADHD-CT. There were no significant differences between the two ADHD subtypes as shown in . Findings for anterior frontal and parietal theta power become non-significant trends with FDR adjustment. None of WS or BS effects for EEG power in other frequency bands reached statistical significance. These analyses were re-run using ADHD status (since there are no differences by ADHD subtype) and only one child per family (either the eldest affected sib or the unaffected sibling when available) to ensure these effects were not the result of multiple children per family (data non-independence). Results were in the same direction with the ADHD group (N=75) exhibiting increased theta power when compared to the non-ADHD group (N=23) (Frontal: F(1, 95)=12.4, p=.001, FDR p=.02; Central: F(1, 95)=7.94, p=.006, FDR p=.03; Parietal F(1,95)=5.07, p=.03, FDR p=.09).
Familial clustering of EEG measures with ADHD affection status and subtype
Among adults, significant differences emerged in the theta, alpha and beta2 frequency bands. Similar to children, significant between-subjects effects emerged in frontal and central theta power among the parents. Theta power was significantly elevated for parents with ADHD-PI compared to the ADHD-CT and non-ADHD groups in the anterior frontal (F (2, 126)=5.16, p=.01, FDR p=.04) and frontal (F (2, 126)=5.16, p=.01, FDR p=.04) region, with a trend toward significance in the central (F (2, 126)=2.8, p=.07, FDR p=.10) region (see ). In addition, significant between-subjects effects in the beta2 frequency band emerged with ADHD-CT parents exhibiting significantly increased beta2 power when compared to the non-ADHD group in the anterior frontal (F(2,125)=3.64,p=.02, FDR p=.05), frontal (F(2,125)= 4.91, p=.01, FDR p=.04), central (F(2,124)=4.33, p=.02, FDR p=.04), and parietal (F(2, 125)= 3.90, p=.02, FDR p=.04). The ADHD-PI group did not differ significantly from either the non-ADHD or ADHD-CT groups. Finally, significant within-subjects and between-subjects effects emerged in the alpha band suggesting that there were different profiles of EEG activation across recording conditions by ADHD group as well as different levels of alpha power by ADHD group status. Within-subjects effects were primarily in the frontal regions (anterior frontal: F(4, 338)=4.13, p=.01, FDR p=.04; frontal F(4, 338)=3.55, p=.01, FDR p=.04) and indicate that the ADHD-PI group exhibited a different pattern of EEG activation where alpha power does not attenuate when going from EC to EO as it does for the ADHD-CT and non-ADHD groups. Between-subjects effects emerged globally for alpha power (anterior frontal: F(2, 126)=4.08, p=.02, FDR p=.04; frontal F(2, 126)=4.27, p=.02, FDR p=.04; central F(2, 126)=4.21, p=.02, FDR p=.04; parietal F(2, 126)=3.63, p=.03, FDR p =.05), indicating that adults with ADHD-CT exhibit reduced levels of alpha power compared to the non-ADHD and ADHD-PI groups.
Effects of the Dopamine receptor D4 gene (DRD4) on EEG
The DRD4 gene was significantly associated with cortical activity in the beta2 frequency band among both parent and child samples (see ) and the theta band within the child sample only. DRD4 effects on all other frequency bands were not significant. Within the child sample, the within-subjects analyses were not significant but two significant between-subjects effects emerged, suggesting DRD4-7-repeat allele (7R) carriers had significantly increased frontal theta power (frontal: F(1,206)=5.77, p=.02, FDR p= .05; central: F(1,206)=3.83, p=.05, FDR=.10; parietal F(1,206)=1.83, p=.18, FDR p= .45) and reduced global beta2 power (frontal: F(1,206)=5.79, p=.02, FDR p= .05; central: F(1,206)=5.82, p=.02, FDR=.05; parietal F(1,206)=6.35, p=.01, FDR p= .04) when compared to those children who do not have at least one copy of the 7-allele (not 7). The significant between group effect suggests that the DRD4-7 repeat allele is associated with EEG variability among children with ADHD. Among parents, DRD4 also had significant global (i.e., over the whole head) effects in the beta2 frequency band, however, it was the within-subjects and not the between-subjects effect that was significant (frontal: F(2, 312)=4.46, p=.01, FDR p=.04; central: F(2, 312)=6.03, p=.003, FDR p=.02; parietal: F(2, 304)=3.77, p=.03, FDR p=.07). The significant within-subjects effect suggests that DRD4-7R carriers had similar beta2 power in the eyes closed condition, but exhibited reduced beta2 power in the eyes open and CPT conditions compared to the DRD4 not-7 group. DRD4 appeared to have similar effects on theta power among adults, however, the results for the between-subjects effect were non-significant trends (frontal: F(1, 156)=2.6, p=.10, FDR p=.11; central: F(1, 156)=1.9, p=.16, FDR p=.16; parietal: F(1, 152) <1, p=.49, FDR p=.49).
DRD4 effects on EEG power among children and parents