None of the authors report any conflict of interest.
How to Cite this Article: Jacobsen KK, Halmøy A, Sánchez-Mora C, Ramos-Quiroga JA, Cormand B, Haavik J, Johansson S. 2013. DISC1 in Adult ADHD Patients: An Association Study in Two European Samples. Am J Med Genet Part B 162B:227–234.
In 1990, St Clair et al. reported a large Scottish family with mental and/or behavioral disorders, which co-segregated with a t(1;11) balanced translocation leading to the subsequent discovery of the Disrupted in Schizophrenia (SCZ) 1 (DISC1) gene, located in a 415 kb region on chromosome 1q42.1 [St Clair et al., 1990; Millar et al., 2000]. Although an emphasis was made on its connection with SCZ, bipolar disorder (BPD), and major depression, the strongest association between phenotype and translocation was found when five individuals with adolescent conduct and emotional disorder were also included in the analysis. Furthermore, the original proband was diagnosed with an adolescent conduct and emotional disorder, but not with SCZ [St Clair et al., 1990]. DISC1 has since been implicated in several psychiatric disorders, including autism spectrum disorders, and cognitive functions such as sustained attention and visual working memory [Salisbury et al., 1999; Hennah et al., 2005; Chubb et al., 2008; Kilpinen et al., 2008]. The DISC1 protein is known to interact with proteins encoded by a number of other neuropsychiatric candidate genes, many of which have been implicated in Attention-deficit/hyperactivity disorder (ADHD) [Williams et al., 2010; Won et al., 2011; Bradshaw and Porteous, 2012]. Furthermore, it has been shown that DISC1 is important for neuronal development and migration [Brandon and Sawa, 2011; Soares et al., 2011], thus DISC1 could be hypothesized to be a general liability-gene for neuropsychiatric traits.
ADHD is a common, complex and heterogeneous disorder characterized by symptoms and problems related to inattention, hyperactivity, and impulsiveness [Haavik et al., 2010]. The disorder has predominantly been studied in children, but at least one-third of patients continue to have impairing symptoms into adulthood [Faraone, 2006]. The clinical presentation of ADHD in adults typically includes comorbidity with other psychiatric disorders [Kessler et al., 2006; Sobanski et al., 2007; McGough et al., 2008; Haavik et al., 2010]. We have recently shown that 12% of our clinically diagnosed adult ADHD patients report a lifetime history of BPD and that 51% of the ADHD patients screened positive for a bipolar spectrum disorder, as defined by the Mood Disorder Questionnaire (MDQ), compared to 1.7% and 8.3% of population derived controls, respectively [Halmøy et al., 2010]. Importantly, ADHD patients with bipolar symptoms had poorer functioning than ADHD patients without bipolar symptoms [Halmøy et al., 2010].
As DISC1 is emerging as a susceptibility gene involved in several psychiatric disorders, we wanted to investigate whether it is also associated with a diagnosis of ADHD and with symptoms of mood disorders in a sample of Norwegian adult ADHD patients. The specific aims were to genotype the most studied candidate single nucleotide polymorphisms (SNPs) from previous work on BPD and SCZ and a putatively functional CNV at the start of DISC1 locus. Here we report a significant association between a DISC1 SNP and a diagnosis of adult ADHD in a meta-analysis of a Norwegian and a Spanish cohort.
MATERIALS AND METHODS
Subjects and Measures in the Norwegian Sample
The Norwegian sample consists of 561 ADHD patients and 713 controls above 18 years of age, after excluding eight patients with self reported mental retardation. Males were more common among cases than in controls, see Table I. Patients were mainly recruited from a Norwegian national ADHD registry, but also by psychiatrists and psychologists working in outpatient clinics nation-wide. A clinical diagnosis of ADHD/hyperkinetic disorder was made according to ICD-10 or DSM-IV criteria. Controls were randomly recruited from the Norwegian population through the Medical Birth Registry of Norway [Johansson et al., 2008; Halmøy et al., 2009].
Table I. Sociodemographic Data and Average Scores on the Mood Disorder Questionnaire (MDQ) and the ADHD Self Report Scale (ASRS) in Patients and Controls in the Norwegian Sample
ADHD patients (n = 561)
Controls (n = 710)
The three individuals excluded from genotype analysis are not included in this table. For further descriptive information of the sample, see Halmøy et al. .
Gender (% men)
Age (range of years)
Average age (years)
Average MDQ score
Average ASRS score
DNA from patients and controls were extracted from EDTA blood using standard methods or from saliva using the Oragene™ DNA Self Collection Kit (DNA Genotek, Inc., Ontario, Canada). All participants filled in questionnaires including the ASRS (adult ADHD self report scale, 18 item version) measuring the presence and severity of current ADHD symptoms, and the MDQ, a screening instrument for BPD [Hirschfeld et al., 2000; Kessler et al., 2005; Adler et al., 2006]. In addition, they answered questions concerning sociodemographic and clinical factors including comorbid symptoms and disorders.
The standard scoring of MDQ was used to create a dichotomous variable of MDQ positive and negative individuals [Halmøy et al., 2010]. Patients missing one or more item on the MDQ were excluded from the analyses. These patients (n = 63) did not differ significantly from the rest of the patients regarding age, gender, occupational/educational level, or ASRS score.
Written consent from all participants was obtained at the time of inclusion. The study was approved by the Regional Ethics Committee of Western Norway (IRB 00001872).
Subjects and Measures in the Spanish Sample
The Spanish sample consisted of 694 patients and 735 controls that were matched for age and sex, collected and diagnosed at the Department of Psychiatry at the Hospital Universitari Vall d'Hebron [Ribasés et al., 2009]. The diagnosis of ADHD in adulthood was evaluated with the Structured Clinical Interview for DSM-IV Axis I and II Disorders (SCID-I and SCID-II) and the Conners' Adult ADHD Diagnostic Interview for DSM-IV (CAADID Parts I and II). The diagnosis of mood disorders was based on the Structured Clinical Interview of DSM-IV for axis-I disorders (SCID-I). Exclusion criteria included IQ <70, SCZ or other psychotic disorders, ADHD symptoms due to mood, anxiety, dissociative or personality disorders, adoption, sexual or physical abuse, birth weight <1.5 kg, and other neurological or systemic disorders that might explain ADHD symptoms [Ribasés et al., 2009].
The control sample consisted of Spanish Caucasoid-unrelated adult subjects in whom DSM-IV ADHD symptoms have been excluded retrospectively.
The study was approved by the ethics committee of each participating institution, and written informed consent was obtained from all adult subjects.
The 101 kb DISC1 promotor/exon 1 copy number polymorphism variation 2350 was assessed using the TaqMan CNV assay Hs04200080_cn from Applied Biosystems (Life Technologies, Carlsbad, CA) [Redon et al., 2006]. All assays were performed in 384-well format, 10 µl total volume, with three parallels per sample and one positive control (CN-state = 3, verified on microarray) per plate. The Copy Caller-software from Applied Biosystems was used for copy number calling. All samples with a calculated CN of ≥2.3 or a confidence value of <0.95 were re-run a second time (using the same protocol). All samples with a confidence score of less than 0.95 on both assays were removed from further analysis (n = 4).
DISC1 SNPs were prioritized based on previous findings, mainly by Schumacher et al.  and Hennah et al. [2009, 2003], as well as those studies summarized by Chubb et al. , concerning DISC1 and SCZ, major depression disorder (MDD) or BPD. The two non-synonymous SNPs, rs6675281 (Leu607Phe) and rs821616 (Ser704Cys) have been tested in several studies, with varying results [Devon et al., 2001; Chubb et al., 2008]. The former is also part of the HEP1 haplotype described by Hennah et al. . These markers, as well as the HEP3 haplotype, are those most frequently studied previously. One SNP had to be excluded because of assay failure, leaving a total of 11 SNPs. A full list of SNPs and previous findings can be found in Table II. Figure 1 shows the gene with exons and markers indicated, and an LD-plot of our data.
Table II. An Overview of Markers Selected for Analysis
All but one SNP were genotyped at the CIGENE platform at the Norwegian University of Life Sciences at Ås, using the MassARRAY iPLEX system (Sequenom, San Diego, CA). SNP rs821616, which did not fit with the multiplex assay design, was genotyped using TaqMan genotyping assays (Life Technologies) in a 384-well format, using the standard protocol.
SNP Genotyping in the Spanish Replication Sample
Markers rs1538979 and rs6675281 were selected for follow up in the Spanish sample and genotyped using the TaqMan system. However, rs1538979 showed poor clustering in the TaqMan assay and was replaced by the tag SNP rs11122330 and genotyped in both samples (r2 = 0.93 between markers in the HapMap CEU sample and r2 = 0.89 in the Norwegian sample).
All genotype data were analyzed using the PLINK software [Purcell et al., 2007]. The promoter/exon 1 CNV was treated as a SNP since only two copy number states were detected (CN = 2, CN = 3). Three individuals were removed because of low genotyping efficiency (>30% missing genotype data) in the Norwegian sample, leaving 561 cases and 710 controls for the final analysis, of which 575 were male and 696 were female. All markers passed Hardy–Weinberg testing with P > 0.01. Concordance was 100% for all markers (n = 54 replicate samples). Total genotyping rate (including the CNV-assay) in remaining individuals was >99%. Allele and genotype based association analyses were performed using chi square tests and/or logistic regression, with ADHD status or a positive score on MDQ as outcome measures. Due to gender distribution differences between the case and control groups we controlled for gender in the ADHD case/control analysis, and both gender and ADHD status in the MDQ analysis. A 3-marker window sliding haplotype analysis was performed using PLINK, including analysis of the HEP1 haplotype. Analysis of the CNV data was restricted only to tests of allelic association with ADHD, due to the very low minor allele frequency. Based on previous findings we also performed a test of rs821633 conditioning on rs1538979 [Hennah et al., 2009].
Replication and meta-analysis
Rs11122330 was analyzed with ADHD diagnosis as an outcome measure. Both marker frequencies were in accordance with HWE with genotyping rates of 98% and 94%, respectively. Meta-analysis was performed using the metan command in STATA (StataCorp. 2011. Stata Statistical Software: Release 12; StataCorp LP, College Station, TX) and similar results were obtained using PLINK.
Tests of association with an ADHD diagnosis in the Norwegian sample revealed one nominally significant SNP [rs1538979: odds ratio (OR): 1.33, 95% confidence interval (CI) 1.03–1.73, P = 0.03], see Table III. We also observed a nominally significant association between rs6675281 and a positive MDQ-score after adjusting for gender and ADHD-status (OR = 1.44, 95% CI 1.08–1.93, P = 0.01), see Supplementary Table I. The two associated markers showed no evidence of inter-marker LD (r2 = 0.002), indicating that the two findings are independent. In view of earlier findings we also tested for putative haplotype effects but found none stronger than the single marker results. Likewise, we did not find any evidence of additional associations when conditioning upon rs1538979 (data not shown).
Table III. Allelic Association Between DISC1 and ADHD, Cases Versus Controls, in the Norwegian Sample
Nominal P-value is listed. DISC1, disrupted in schizophrenia 1; ADHD, attention-deficit/hyperactivity disorder; Freq, frequency; OR, odds ratio; CI, confidence interval; CNV, copy number variation.
Attempts of replication for the two nominally significant findings were performed in a Spanish sample of 694 adult ADHD patients and 735 controls. Replication of the association between adult ADHD and marker rs1538979 was performed using the near perfect tag SNP rs11122330 (r2 = 0.93). Similar results and direction of effects were obtained in both samples, with an overall meta-analysis association with ADHD status of P = 0.008 (OR 1.25, 95% CI 1.06–1.47), see Table IV and Figure 2.
Table IV. Meta-Analysis of rs11122330 in the Norwegian and Spanish Samples
OR and P-value for meta-analysis was the same for random and fixed effects models.
To follow up the putative association between MDQ scores and rs667528, we contacted other investigators, but as MDQ data were not available for other adult ADHD cohorts than the Norwegian sample, it was impossible to perform a formal replication study of this finding. In an attempt to explore the relationship between DISC1 variants and mood disorders, we obtained three measures of mood disorder diagnoses available in the Spanish sample; total, present, and lifelong mood disorder.
However, we observed no significant association between these diagnoses and the Phe607 variant of rs6675281 (P > 0.3).
Here we investigated DISC1 SNPs previously associated with major psychiatric disorders such as SCZ, BPD, and major depression in a sample of clinically diagnosed ADHD patients. In the ADHD case/control analysis we found an association for the intronic DISC1 SNP, rs1538979 (OR: 1.33, 95% CI 1.03–1.73, P = 0.03), which was further strengthened using a Spanish cohort for replication (meta-analysis OR 1.25, P = 0.008 for the tested tag-SNP rs11122330). The rs11122330/rs1538979 markers have been studied by several different groups [Wood et al., 2007; Sullivan et al., 2008; Hennah et al., 2009; Schumacher et al., 2009]. Hennah et al.  found different trends in the different cohorts for SCZ and BPD, both regarding risk allele and gender specific associations. In contrast to our results, they also reported an interaction between rs1538979 and rs821633. This could indicate that the true risk locus resides elsewhere on a haplotype marked by these two SNPs, or alternatively, that there are several risk variants in DISC1, or that the signal is secondary to another gene in the region [Ram et al., 2012]. Furthermore, since DISC1 is involved in many cellular pathways and has several known binding sites spanning the protein [Bradshaw and Porteous, 2012], it has been speculated that different alterations of the gene could lead to increased susceptibility for different neuropsychiatric disorders. Palo et al.  suggested that the 5′ end of DISC1 was associated with psychotic disorder, while the 3′ end was associated with BPD. Thus, it was interesting to note that we found different SNPs nominally associated with susceptibility to ADHD, and a measure of bipolar spectrum disorder in the Norwegian sample, although it should be noted that our replication sample only supported the former association (see below).
The relationship between ADHD and mood disorders, in particular BPD has received much attention [Faraone et al., 1997; Singh et al., 2006; Wingo and Ghaemi, 2007]. We showed in our clinically recruited cohort of adult ADHD patients that many ADHD patients have a high load of affective symptoms as measured with the MDQ, a rating scale developed to screen for bipolar spectrum disorders [Halmøy et al., 2010]. It is therefore interesting that we observed an association between rs6675281 and a positive MDQ-score (OR = 1.44, 95% CI 1.08–1.93, P = 0.01). Rs6675281 (Leu607Phe) is one of the most studied polymorphisms in DISC1, and has been associated with changes in neuronal signaling [Eastwood et al., 2009; Nakata et al., 2009; Brandon and Sawa, 2011]. This finding was not replicated in the Spanish cohort using a diagnosis of mood disorder as an outcome measure. This could be due to a false positive finding, but also be a result of the different phenotypes that were compared.
Although DISC1 for a long time was considered a susceptibility gene for psychotic disorders, more recent findings have shown that it is involved in general neurodevelopment and signaling, and it is possible that unknown functional variants may predispose an individual for a range of different mental illnesses. To our knowledge, our study is the first systematic investigation of DISC1 in ADHD and it adds ADHD to the traits possibly associated with DISC1 variation. It should be noted that although we find a consistent effect in both tested cohorts, results are only nominally significant, and therefore need to be confirmed in larger samples. However, several indirect links between DISC1 and ADHD have been reported, including a recent genome-wide study that showed an increased number of CNVs involving one of DISC1s interaction proteins, NDE1 [Williams et al., 2010]. Furthermore, DISC1 and its interactome belong to a large group of proteins that are involved in early prenatal brain development and it could be speculated that genetic variants in these key genes might influence later liability for psychiatric disorders [Haavik et al., 2010]. Therefore, in conclusion, our results suggest that DISC1 should be considered a candidate gene in ADHD.
Norway: We would like to thank Paal Borge for his help with genotyping. The project was supported from the Research Council of Norway, the Regional Health Authority of Western Norway and the K.G. Jebsen Foundation.
Spain: Financial support was received from “Instituto de Salud Carlos III” (PI11/00571, PI11/01629), “Fundació La Marató de TV3” (092330/31), “Plan Nacional sobre Drogas” (2011/0080). “Ministerio de Economía y Competitividad, Spain” (SAF2012-33484), “Agència de Gestió d'Ajuts Universitaris i de Recerca-AGAUR” (2009SGR0971) and the Departament de Salut, Government of Catalonia, Spain.