Sonuga-Barke, E. J. S., Lasky-Su, J., Neale, B. M., Oades, R. D., Chen, W., Franke, B., Buitelaar, J. K.., Banaschewski, T., Ebstein, R. P., Gill, M., Anney, R. J. L., Miranda, A., Mulas, F., Roeyers, H., Rothenberger, A., Sergeant, J. A., Steinhausen, H-C., Thompson, M., Asherson, P. & Faraone, S. V.

Does parental expressed emotion moderate genetic effects in ADHD? An exploration using a genome wide association scan.
(2008) American Journal of Medical Genetics Part B, 147B, 1359-1368. DOI: 10.1002/ajmg.b.30860.
[Request a copy] View Article

Introduction: Studies of gene x environment (G x E) interaction in ADHD have previously focused on known risk genes for ADHD & environmentally mediated biological risk. Here we use a G x E analysis in the context of a genome-wide association-scan to identify novel genes whose effects on ADHD symptoms & comorbid conduct disorder are moderated by high maternal expressed emotion (EE).

Methods: 600,000 SNPs were genotyped in 958 ADHD proband-parent trios. After applying data cleaning procedures we examined 429,981 autosomal SNPs in 909 family trios. Pre-specified ADHD candidate genes that was generated by the IMAGE study investigators [Brookes et al., 2006a]: NR4A2, PER2, SLC6A1, DRD3, SLC9A9, HES1, ADRA2C, ADRB2, ADRA1B, DRD1, HTR1E, DDC, STX1A, ADRA1A, NFIL3, ADRA2A, ADRB1, SLC18A2, TPH1, BDNF, FADS1, FADS2, ADRBK1, ARRB1, DRD2, HTR3B, TPH2, SYT1, HTR2A, SLC6A2, ARRB2, PER1, PNMT, CHRNA4, COMT, ADRBK2, and CSNK1E.

ADHD symptom severity and comorbid conduct disorder was measured using the PACS. Maternal criticism & warmth (i.e., EE) were coded by independent observers on comments made during the Camberwell family interview (warmth, from much to little was rated from 0-3). Criticism from much (4) to little (0) [After training in London - A mean Kappa coefficient across all sites of 0.88 (range 0.71-1.00) was achieved and an average agreement percentage of 96.6% (range 78.6-1.00) was obtained indicating a substantial level of inter-rater agreement. FBAT1-Interaction methodology was used (Vansteelandt et al. 2008) for nuclear families where gene-by-environment interactions can be modeled using SNPs, quantitative phenotypes, & complex exposure variables.

1 - No G x E interactions reached genome-wide significance (p<10-8). The 37 interaction p-values < 10-5 are summarized in tables 2 & 3.

2 - Nominal effects were found both with & without genetic main effects.

3 - For those with genetic main effects 36 uncorrected interaction p-values were less than 10-5 implicating both novel genes as well as some previously supported candidates.

4 - These were found equally often for all of the interactions being investigated.

5 - a) nominally significant G x Es found [without main genetic effect (p > 0.05)] with an interaction p-value < 0.001 (table 5). 19 of these related to conduct disorder (CD) while only 6 related to ADHD severity. A number of these were in genes previously associated with psychiatric disorders thought to be unrelated to ADHD.

.....b) The observed interax in SLC1A1 (GABA transporter rs10974610 located in 9p24) & neuroregulin NRG3 SNPs (rs17746658: intronic region) implicated in G x E interactions for CD (as in other disorders) represent reasonable candidate genes for further investigation given their previous association with several psychiatric illnesses.

In detail -- For the CD symptom count, strong interactions between mother's warmth and SNPs in RIT1, ADH1C, SLC6A1, A2BP1, & MFHAS1 were observed. One of these findings is in one of the ADHD candidate genes that had been selected a priori (SLC6A1, GABA transporter). Finally, interactions between SNPs in PPM1K & ZBTB16 & mother's criticism were observed when the CD symptom count was used as the phenotype. For all of these interaction associations, the association p-value for the main genetic effect was less than 0.01 -- p<0.01, we found associations at the following ADHD candidate genes: ADRA1A, ADRA1B, ADRA2C, ARRB1, DBH, DDC, DRD3, FADS2, HTR2A, HTR3B, SLC6A1, SLC6A2, SLC6A3, SLC6A4, SLC9A9, SYT1, TPH1, & TPH2.

One association in SLC6A1, rs9990174, was significant after adjusting for the comparisons among the 2 phenotypes, 4 environmental variables, and all of the candidate gene SNPs (p-value = 5.91x10-6).

Discussion: We find evidence for the role of EE in moderating the effects of genes on ADHD severity & comorbid conduct disorder, implicating both novel & established candidates. These findings need replicating in larger independent samples.

support NIH