You are here

Article Text

Article menu
Original articles
Family-based association study of ADHD and genes increasing the risk for smoking behaviours
  1. Geeta A Thakur1,4,
  2. Sarojini M Sengupta4,
  3. Natalie Grizenko2,4,
  4. Zia Choudhry3,4,
  5. Ridha Joober1–,4
  1. 1Integrated Program in Neuroscience, McGill University, Montreal, Quebec, Canada
  2. 2Department of Psychiatry, McGill University, Montreal, Quebec, Canada
  3. 3Department of Human Genetics, McGill University, Montreal, Quebec, Canada
  4. 4Douglas Mental Health University Institute, Montreal, Quebec, Canada
  1. Correspondence to Dr Ridha Joober, Douglas Mental Health University Institute, 6875 LaSalle Blvd, Montreal, Quebec H4H 1R3, Canada; ridha.joober{at}douglas.mcgill.ca

Abstract

Objective To investigate five top single nucleotide polymorphisms (SNPs) located in different genes and loci (CHRNA3, BDNF, DBH and LOC100188947) that were highly associated with different dimensions of smoking behaviour, in relation to attention-deficit hyperactivity disorder (ADHD).

Design Cohort study consisting of a clinical sample of children with ADHD.

Setting Douglas Institute ADHD Clinic, Montreal, Canada.

Patients Families of 454 children with ADHD aged 6–12 years old.

Interventions Family-based association tests used to study the transmission of risk alleles within these five genetic markers.

Main outcome measures Clinical diagnosis of ADHD, and a number of behavioural and neurocognitive phenotypes relevant to the disorder.

Results One SNP (rs1329650) from a non-coding RNA (LOC100188947) was significantly associated with overall ADHD diagnosis with the C* risk allele being over-transmitted from parents to children with ADHD (p=0.02). It was also over-transmitted to children with higher scores on Conners’ Parents (p=0.01) and Conners’ Teacher (p=0.002) index scores, and Child Behaviour Checklist withdrawn (p=0.001) and aggressive (p=0.007) behaviours. Children with poorer performances on executive and attention tasks were more likely to inherit the risk allele.

Conclusions The C* allele of rs1329650 may be increasing the risk for ADHD and smoking behaviour through a common mechanism, possibly externalising behaviours and specific cognitive deficits that manifest as ADHD in childhood and are the gateway to smoking behaviour later in life. This exploratory study illustrates the use of comorbid disorders to investigate ADHD genetics. In spite of its relatively large sample size, replication in future studies is warranted.

Trial Registration Number NCT00483106.

  • Child Psychiatry
  • Genetics

https://doi.org/10.1136/archdischild-2012-301882

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What is already known on this topic

    • Attention-deficit hyperactivity disorder (ADHD) and smoking behaviour are two highly comorbid phenotypes.

    • ADHD has been associated with multiple genes of small effect and several genes, reliably identified by genome-wide association studies (GWAS), are involved in smoking behaviour.

    • Comorbidity among these two phenotypes could be due to shared genetic factors.

    What this study adds

    • To the best of our knowledge, smoking genes have not previously been tested in ADHD.

    • This is the first report of single nucleotide polymorphisms identified through GWAS of smoking behaviour shown to be tentatively associated with ADHD.

    • This study illustrates the use of comorbid disorders to investigate ADHD genetics.

    Introduction

    Attention-deficit hyperactivity disorder (ADHD) is an aetiologically complex, heterogeneous, highly heritable, neurobehavioural childhood disorder with 8–12% prevalence in the general population.1 In spite of this high heritability, identifying genes in ADHD has been a challenging task. Five genome-wide association studies (GWAS) of ADHD have been conducted2–6 and identified 85 top-ranked ADHD candidate genes (p<0.0001). However, none of the findings passed the GWAS significance threshold.

    ADHD and cigarette smoking are two highly comorbid phenotypes with rates of comorbidity varying between 35% and 45%.7 ADHD subjects begin smoking at an earlier age and are likely to smoke twice as much as controls. Evidence suggests that a history of childhood ADHD may predict worse smoking cessation outcomes.8 Although the underlying mechanisms of ADHD and smoking have yet to be fully understood, an integrated model of their comorbidity has been proposed9 stating that both neurobiological and psychosocial factors may contribute to an increased risk of nicotine use and dependence in ADHD.9 Neuropsychological theories suggest that children with ADHD may have deregulations of executive functions (EFs), including inhibitory control,10 ,11 resulting in increased impulsivity, risk taking, and novelty seeking behaviour, thereby increasing the risk for later substance/alcohol abuse and cigarette smoking.12 ADHD may itself be a risk for smoking later in life, and early treatment with stimulant medication seems to have a protective effect in adolescence.13 Converging data suggest that these two phenotypes may share underlying neurobiological mechanisms,9 related to monoaminergic transmission, specifically an altered dopamine/norepinephrine as well as cholinergic transmission. Thus, it is highly likely that genes implicated in smoking behaviour may also increase the risk for ADHD and vice versa.

    Smoking behaviour is a complex phenotype with several genetic factors involved. Interestingly, genetic studies of smoking behaviour have witnessed important advances in recent years due to the power of large scale GWAS. Meta-analytical results were reported from three GWAS smoking consortia: the Tobacco and Genetics (TAG) Consortium (16 studies, population-based and case–control), the European Network of Genetic and Genomic Epidemiology (ENGAGE), and the Oxford-GlaxoSmithKline (Ox-GSK) consortia, where a number of loci were identified and associated with different dimensions of smoking behaviour, such as number of cigarettes smoked per day, smoking initiation and cessation.14 Among the five top associated markers, a synonymous single nucleotide polymorphism (SNP) in the nicotinic receptor gene CHRNA3 (rs1051730) was associated with number of cigarettes smoked per day (p<3×10−70), several SNPs in the BDNF gene were associated with smoking initiation (p<5×10−8) and one SNP near the DBH gene (rs3025343) was associated with smoking cessation (p<4×10−8). In addition, two 10q25 SNPs (rs1028936 and rs1329650), located in a non-coding RNA (LOC100188947), were very highly associated with number of cigarettes smoked per day (p<2×10−9 and p<6×10−10, respectively).14

    Interestingly, two of these SNPs have previously been implicated in externalising behaviours often seen in ADHD. More specifically, in an adolescent sample, an association with the CHRNA5/CHRNA3/CHRNB4 locus and externalising behaviours was reported.12 Additionally, in a community-based cohort of 1236 Swedish individuals, multivariate regression analysis showed that the Met allele of the Val66Met polymorphism in the BDNF gene was associated with ADHD, where the association was primarily driven by persistent hyperactivity-impulsivity symptoms.15 Various polymorphisms within the DBH gene were linked with poorer cognitive performance in ADHD children,16 ,17 especially on tasks indexing cognitive impulsiveness. However, the two 10q25 SNPs have not yet been studied in ADHD or related phenotypes. In this study, we investigated transmission of risk alleles within these five SNPs in 454 ADHD families with respect to clinical diagnosis of ADHD and quantitative (behavioural and neurocognitive) phenotypes relevant to ADHD.

    Methods

    Subjects

    Four hundred and fifty-four ADHD subjects were sequentially recruited from the Disruptive Behaviour Disorders Programme and the child psychiatry outpatient clinics at the Douglas Mental Health University Institute (DMHUI) in Montreal. They were referred to these specialised care facilities by schools, community social workers, family doctors and paediatricians.

    Children were diagnosed with ADHD using DSM-IV criteria18 and based on clinical interviews of the child and at least one parent by a child psychiatrist (RJ or NG). A structured clinical interview of parents using the Diagnostic Interview Schedule for Children-IV (DISC-IV)19 and school reports were used to corroborate the diagnoses. Mothers were primary informants in most cases. Details about diagnostic procedures have been described elsewhere.20

    Children with a history of Tourette's syndrome, pervasive developmental disorder, or psychosis were excluded. The research protocol was approved by the Research Ethics Board of the DMHUI. Parents provided written informed consent while children gave their verbal assent.

    Maternal smoking during pregnancy

    The Kinney Medical Gynecological Questionnaire21 was used to systematically evaluate pregnancy, delivery and perinatal complications. Mothers retrospectively reported maternal smoking (yes/no) during pregnancy (MSDP). We had information pertaining to MSDP exposure for 394 of the families in the study: 171 were categorised as smoking and 223 as non-smoking families. In the families where mothers smoked during pregnancy, 76.6% of affected children were boys, whereas 81.6% were boys in the unexposed group. The mean age of exposed children was 9.1 years (SD 1.7) and 8.9 years (SD 1.8) in unexposed children. Neither gender (χ2=1.49, df=1, p=0.22) nor age (F1,393=0.38, p=0.54) differed among the two groups of children.

    Behavioural evaluations

    The Child Behaviour Checklist (CBCL),22 which assesses children's behavioural and emotional problems, was completed by the parents and the child's overall behaviour (without a specific timeframe) was evaluated. The Conners’ Global Index for parents (CGI-P) and teachers (CGI-T)23 were used to assess behaviours relevant to ADHD in home and school settings, respectively. The CGI-P and CGI-T are subsets of the original Conners’ Rating Scales, which are widely used to assess ADHD symptoms and other psychopathology in children between 3 and 17 years of age. The raw total scores are transformed into normalised T scores. All assessments were completed while children were not taking any medication.

    Neurocognitive assessment

    A neuropsychological battery of tests was used to study attention and EFs in these children. Assessments were carried out at the end of a 1-week washout period if children were previously medicated. Full scale, verbal and performance IQ were evaluated using the Wechsler Intelligence Scale (WISC).24 Children with IQ <70 were excluded from the study. The Wisconsin Card Sorting Test (WCST),25 the Wide Range Assessment of Memory and Learning Finger Windows (FW) subtest,26 the Tower of London test (TOL),27 the Self-Ordered Pointing Task (SOPT)28 and the Conners’ Continuous Performance Test (CPT)29 were carried out. These tasks are conceptually used to assess mental flexibility, visual working memory, planning capacities, working memory and response inhibition and attention profile, respectively.

    Genotyping and marker selection

    DNA was extracted from a blood, buccal swab or saliva sample from each affected child, parents and unaffected siblings, whenever possible. Based on findings from the TAG study, five markers associated with different dimensions of smoking behaviour were selected. The panel of SNPs was genotyped using Sequenom iPlex Gold Technology.30 Every plate included duplicates of two reference samples to estimate genotyping error and genotypes were read with 100% accuracy on each of the plates. The genotype distribution of all five markers did not depart from Hardy–Weinberg equilibrium: rs1051730 (p=0.96), rs1028936 (p=0.11), rs1329650 (p=0.12), rs3025343 (p=0.49) and rs6265 (p=0.23).

    Family-based association tests analysis

    Single SNP tests of association were performed using family-based association tests (FBAT) to investigate the association between selected markers with ADHD diagnosis and quantitative phenotypes relevant to ADHD (V2.0.3 Harvard School of Public Health, Departments of Biostatistics and Environmental Health, Programme for Population Genetics, Boston, Massachusetts, USA).31 First, the overall association between each of these markers and ADHD diagnosis was investigated. Second, association between a number of behavioural and cognitive quantitative traits was tested. Offsets used in the FBAT analysis were based on average scores found in the population (50 in the case of T scores). All tests were performed under the assumption of an additive model, with a null hypothesis of no linkage and no association. Further details pertaining to the principle of FBAT have been described elsewhere.32 Haploview V4.0 was used to determine linkage disequilibrium between the two 10q25 SNPs and haplotype analysis was then conducted in FBAT for all the ADHD phenotypes. Given the exploratory nature of this study, the significance level was set at p=0.05.

    To obtain a rough estimate of effect size, we calculated the effect size F as for a χ2 test, using the following formula F=square root [χ2/N (k−1)], where N indicates sample number, and k indicates the number of rows or columns or 2 in the McNemar's test. This was based on the assumption that FBAT is an extension of McNemar's test used to calculate transmission disequilibrium in a pedigree, where χ2TDT=(T−NT)2/(T+NT). T and NT denote the number of transmissions and non-transmissions of a specific allele from the parent to the affected offspring. In a specific case of the FBAT (where both parents are known, and when the additive model is used), the Z2 statistic can be considered equivalent to a χ2TDT statistic (N Laird, personal communication, 5 January 2012).

    The number of informative families was used to calculate N. Effect sizes of 0.1, 0.3 and 0.5 are considered small, medium and large, respectively.

    Results

    ADHD clinical diagnosis

    Only one SNP (rs1329650) was nominally significantly associated with overall ADHD diagnosis (p=0.02) with a small effect size (ES) of 0.19, in the total sample with the C* allele over-transmitted from parents to children with ADHD (table 1). Haplotype analysis of the two 10q25 SNPs showed they are in strong linkage disequilibrium (D´=0.98, r2=0.59). FBAT analysis of the four haplotypes derived from these two SNPs showed that the A-C haplotype containing the A* allele from rs1028936 and the C* allele from rs1329650, was significantly over-transmitted in children (p=0.02, ES=0.2), whereas none of the other haplotypes showed a significant association with ADHD, suggesting that the association is mainly driven by the C* allele in rs1329650. When the total sample was stratified by MSDP, a marginally significant association between ADHD diagnosis and the risk variant in rs1329650 was distributed evenly between the two exposure subgroups (p=0.08 for both groups). Since none of the other genes were associated with ADHD diagnosis (before or after stratification with regard to MSDP, supplementary table S1, we further investigated the two 10q25 SNPs and their derived haplotypes with respect to a number of quantitative behavioural and cognitive traits in ADHD children.

    View this table:
    Table 1

    FBAT analysis of genes increasing the risk for smoking behaviours in children with ADHD

    Quantitative behavioural traits

    A pattern of associations was formed with the two 10q25 SNPs and their derived A-C haplotype with several behavioural traits (table 2). For rs1028936, the A* allele was significantly over-transmitted to children with higher T scores on the CGI-P (p=0.03, ES=0.22) and CGI-T (p=0.02, ES=0.24), as well as the withdrawn (p=0.01, ES=0.26) and aggressive (p=0.008, ES=0.26) dimensions of the CBCL. For rs1329650, the C* allele was significantly over-transmitted to children with a higher total number of ADHD items (p=0.04, ES=0.18), T scores on the CGI-P (p=0.01, ES=0.21) and CGI-T (p=0.002, ES=0.27) and several dimensions of the CBCL behaviours T scores: externalising (p=0.02, ES=0.21), withdrawn (p=0.001, ES=0.3), attention (p=0.01, ES=0.21) and aggressive (p=0.007, ES=0.24). The A-C haplotype results mimicked those of the C* allele of rs1329650 with similar effect sizes (table 2).

    View this table:
    Table 2

    Association between two LOC100188947 SNPs and derived haplotypes with behavioural traits in a sample of children with ADHD

    View this table:
    Table 3

    Association between two LOC100188947 SNPs and derived haplotypes with cognitive traits in a sample of children with ADHD

    Cognitive traits

    Several cognitive traits were significantly associated with both of the 10q25 SNPs and their derived A-C haplotype (table 3). For rs1028936, the A* allele was associated with poorer performance on the WCST, more specifically, number of total (p=0.02, ES=0.24) and non-perseverative (p=0.01, ES=0.26) errors. For rs1329650, the C* allele was also associated with poorer performance on the WCST, number of total and non-perseverative errors (p=0.03, ES=0.2), as well as SOPT total score (p=0.02, ES=0.2) and the CPT omissions T score (p=0.04, ES=0.19). Again, the results and effect sizes obtained with the A-C haplotype mirrored those of the C* allele of rs1329650 (table 3).

    Discussion

    To the best of our knowledge, this is the first report of SNPs identified through GWAS of smoking behaviour shown to be tentatively associated with ADHD. Two 10q25 SNPs (rs1028936 and rs1329650) from a non-coding RNA (LOC100188947) show a distinct pattern of association with respect to several behavioural and neurocognitive traits characteristic of ADHD. Risk variants of both SNPs, and their derived haplotype, were significantly over-transmitted from parents to affected offspring, and associated with a more severe ADHD phenotype, where children who inherited the risk variant from their parents had more externalising symptoms and poorer performance on several neurocognitive tasks.

    Using SNPs reliably associated through GWAS studies with somatic or behavioural disorders comorbid to ADHD may be an interesting approach to decipher the genetics of this complex psychiatric disorder. This strategy is now used by some other investigators in other complex disorders. For example, Hansen and colleagues investigated a number of SNPs associated with type II diabetes in patients with schizophrenia to identify risk variants for this disorder.33 Here, we have used a similar approach relying on the observation that there is a strong link between ADHD and smoking. It has been suggested that this association may be due, at least in part, to shared genetic factors between these two phenotypes.34

    Behavioural disinhibition/externalising disorders are often comorbid with ADHD and smoking.35 Thus, genetic factors predisposing to ADHD and smoking may act through an increased level of behavioural disinhibition.35 Consistent with this hypothesis, we have observed in our sample that the risk variant associated with smoking in rs1329650 is also related to an increased risk of externalising behaviours as seen on the CBCL.

    In young adults, two theoretical models of tobacco use have been proposed: the self-medication and orbitofrontal/disinhibition model.36 The latter predicts smokers will perform worse on neurocognitive tasks related to orbitofrontal dysfunction as compared to non-smokers. In the present study, poorer performance on neurocognitive tasks such as the WCST, SOPT and CPT, was associated with the risk alleles of both 10q25 SNPs. This is in line with the disinhibition model, which also proposes that smokers obtain elevated scores on tasks that measure behavioural disinhibition and mirrors what we have seen. In our sample, children with the smoking risk variant seem to perform worse on tasks that measure response inhibition, such as the SOPT and CPT.

    Given that smoking is a preventable behaviour, studies identifying common genetic factors for ADHD and smoking may help in the earlier identification of subjects who are more prone to develop dependence to cigarette smoking. This genetic information would be crucial, once confirmed and furthered, to develop preventive strategies, especially since smoking in ADHD patients tends to start earlier in life and once initiated, is much more severe and harder to curve down than in the general public.8

    There is an extensive body of literature associating MSDP with increased risk for ADHD. Although exposure to smoking during pregnancy could be increasing the risk for ADHD through direct neurobiological effects of cigarette smoke content on the developing brain, there is mounting evidence that the association between MSDP and ADHD might be due to factors shared by both conditions, including genetic variants shared by mothers and children, putting them at higher risk for developing smoking behaviour and ADHD, respectively. In order to investigate this possibility, we stratified ADHD children according to whether or not their mothers had smoked during pregnancy (supplementary table S1). Our stratification revealed a marginally and equally significant pattern of transmission of the risk allele in both exposure groups, suggesting that environmental exposure to smoking during pregnancy is not interacting with rs1329650 allelic variants to increase the risk for ADHD or to modify its clinical and cognitive dimensions. However, since this stratification reduced the sample size of each group, we cannot disregard that the absence of significant association, particularly in the group exposed to maternal smoking, is not due to a lack of statistical power.

    Certain limitations should be kept in mind when interpreting these results. First, a larger sample size may have enabled us to have more statistical power to detect significant associations with other loci tested, as previous studies that investigated markers such as CHRNA3, had larger sample sizes12 and identified significant associations with ADHD. It is also possible that we did not identify an association with these markers since our study was conducted with a clinical sample of ADHD children, whereas previous studies investigated adolescents, adults and/or community-based samples.37 However, given the heterogeneity among these associations, it is important to note that our negative results with respect to certain genes, such as BDNF, have also been reported by recent meta-analyses which do not support an association between BDNF and ADHD.38 ,39 Also related to the problem of limited statistical power, the reported association with rs1329650 was only marginal and would not be significant had we applied the stringent Bonferroni correction (we tested five loci, thus corrected p=0.01). However, the a priori plausibility of the involvement of these loci, highly associated with smoking, justifies reporting these tentative results which warrant replication in larger samples.

    Given the age range of children included in this study, none has (or declared) smoking behaviour. Thus, it would be interesting to obtain future information about smoking behaviour in these subjects, to investigate whether risk variant carriers are more likely to initiate smoking behaviour in adolescence or later in life. It is also important to note that rs1329650 found to be associated with ADHD and its relevant behavioural dimensions is not located in any known gene.

    Here, we have found an association between the LOC100188947 marker and ADHD. A possible explanation may be that this variant influences regulatory elements or tags a variant in a nearby gene. The HECTD2 gene, located close to this marker, has been associated with prion disease40 and Alzheimer's disease,41 but has not yet been investigated in ADHD. Thus, it may be worthwhile to study the HECTD2 gene in the aetiology of ADHD.

    Future research will need to focus on understanding the functional consequences of these loci42 using fine mapping around these regions of interest to find out which gene(s) are close by and further dissect their relationship with smoking and ADHD.

    In conclusion, we present preliminary evidence indicating that rs1329650 increases the risk for ADHD and we suggest a neuropsychological and behavioural mechanism that could underlie the link between this genetic risk locus, ADHD and smoking.

    Acknowledgments

    We would like to thank Johanne Bellingham, Rosherrie DeGuzman, Marie-Ève Fortier, Phuong Thao Nguyen, Anna Polotskaia, Jacqueline Richard, Sandra Robinson and Marina Ter-Stepanian for clinical and technical assistance. And we would like to express our sincere gratitude to all the families who participated in this study.

    References

    1. ↵
      1. Biederman J
      . Attention-deficit/hyperactivity disorder: a selective overview. Biol Psychiatry 2005;57:1215–20.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Lasky-Su J,
      2. Neale BM,
      3. Franke B,
      4. et al
      . Genome-wide association scan of quantitative traits for attention deficit hyperactivity disorder identifies novel associations and confirms candidate gene associations. Am J Med Genet B Neuropsychiatr Genet 2008;147B:1345–54.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Lesch KP,
      2. Timmesfeld N,
      3. Renner TJ,
      4. et al
      . Molecular genetics of adult ADHD: converging evidence from genome-wide association and extended pedigree linkage studies. J Neural Transm 2008;115:1573–85.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Mick E,
      2. Todorov A,
      3. Smalley S,
      4. et al
      . Family-based genome-wide association scan of attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry 2010;49:898–905 e3.
      OpenUrlCrossRefPubMed
    5. ↵
      1. Neale BM,
      2. Lasky-Su J,
      3. Anney R,
      4. et al
      . Genome-wide association scan of attention deficit hyperactivity disorder. Am J Med Genet B Neuropsychiatr Genet 2008;147B:1337–44.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Neale BM,
      2. Medland S,
      3. Ripke S,
      4. et al
      . Case-control genome-wide association study of attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry 2010;49:906–20.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Pomerleau OF,
      2. Downey KK,
      3. Stelson FW,
      4. et al
      . Cigarette smoking in adult patients diagnosed with attention deficit hyperactivity disorder. J Subst Abuse 1995;7:373–8.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Humfleet GL,
      2. Prochaska JJ,
      3. Mengis M,
      4. et al
      . Preliminary evidence of the association between the history of childhood attention-deficit/hyperactivity disorder and smoking treatment failure. Nicotine Tob Res 2005;7:453–60.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. McClernon FJ,
      2. Kollins SH
      . ADHD and smoking: from genes to brain to behavior. Ann N Y Acad Sci 2008;1141:131–47.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Barkley RA
      . Behavioral inhibition, sustained attention, and executive functions: constructing a unifying theory of ADHD. Psychol Bull 1997;121:65–94.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Quay HC
      . Inhibition and attention deficit hyperactivity disorder. J Abnorm Child Psychol 1997;25:7–13.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Stephens SH,
      2. Hoft NR,
      3. Schlaepfer IR,
      4. et al
      . Externalizing Behaviors are Associated with SNPs in the CHRNA5/CHRNA3/CHRNB4 Gene Cluster. Behav Genet 2012;42:402–14.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Wilens TE,
      2. Morrison NR
      . The intersection of attention-deficit/hyperactivity disorder and substance abuse. Curr Opin Psychiatry 2011;24:280–5.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      Tobacco and Genetics Consortium. Genome-wide meta-analyses identify multiple loci associated with smoking behavior. Nat Genet 2010;42:441–7.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Bergman O,
      2. Westberg L,
      3. Lichtenstein P,
      4. et al
      . Study on the possible association of brain-derived neurotrophic factor polymorphism with the developmental course of symptoms of attention deficit and hyperactivity. Int J Neuropsychopharmacol 2011;14:1367–76.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Barkley RA,
      2. Smith KM,
      3. Fischer M,
      4. et al
      . An examination of the behavioral and neuropsychological correlates of three ADHD candidate gene polymorphisms (DRD4 7+, DBH TaqI A2, and DAT1 40 bp VNTR) in hyperactive and normal children followed to adulthood. Am J Med Genet B Neuropsychiatr Genet 2006;141B:487–98.
      OpenUrlPubMed
    17. ↵
      1. Kieling C,
      2. Genro JP,
      3. Hutz MH,
      4. et al
      . The -1021 C/T DBH polymorphism is associated with neuropsychological performance among children and adolescents with ADHD. Am J Med Genet B Neuropsychiatr Genet 2008;147B:485–90.
      OpenUrlCrossRefPubMed
    18. ↵
      1. Lahey BB,
      2. Applegate B,
      3. McBurnett K,
      4. et al
      . DSM-IV field trials for attention deficit hyperactivity disorder in children and adolescents. Am J Psychiatry 1994;151:1673–85.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Shaffer D,
      2. Fisher P,
      3. Lucas CP,
      4. et al
      . NIMH Diagnostic Interview Schedule for Children Version IV (NIMH DISC-IV): description, differences from previous versions, and reliability of some common diagnoses. J Am Acad Child Adolesc Psychiatry 2000;39:28–38.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Grizenko N,
      2. Bhat M,
      3. Schwartz G,
      4. et al
      . Efficacy of methylphenidate in children with attention-deficit hyperactivity disorder and learning disabilities: a randomized crossover trial. J Psychiatry Neurosci 2006;31:46–51.
      OpenUrlPubMedWeb of Science
    21. ↵
      1. McNeil TF,
      2. Cantor-Graae E,
      3. Sjostrom K
      . Obstetric complications as antecedents of schizophrenia: empirical effects of using different obstetric complication scales. J Psychiatr Res 1994;28:519–30.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Achenbach TM
      . The child behaviour checklist/4–18. Burlington: University of Vermont, 1991.
    23. ↵
      1. Conners CK
      . Rating scales for use in assessment and clinical trials with children. In: Greenhill LL, Osman BB, eds. Ritalin theory and practice. New York: Mary Ann Liebert, Inc, 1999:113–26.
    24. ↵
      1. Weschler D
      . Weschler intelligence scale for children-third edition: manual. San Antonio, TX: Psychological Corporation, 1991.
    25. ↵
      1. Heaton RK,
      2. Chelune GJ,
      3. Talley JL,
      4. et al
      . The Wisconsin card sorting test manual–revised and expanded. Odessa, FL: Psychological Assessment Resources, 1993.
    26. ↵
      1. Sheslow D,
      2. Adams W
      . Wide range assessment of memory and learning: administration manual. Wilmington, DE: Jastak, 1990.
    27. ↵
      1. Shallice T
      . Specific impairments of planning. Philos Trans R Soc Lond B Biol Sci 1982;298:199–209.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Petrides M,
      2. Milner B
      . Deficits on subject-ordered tasks after frontal- and temporal-lobe lesions in man. Neuropsychologia 1982;20:249–62.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Conners CK
      . Conners continuous performance test computer program. Toronto, Ontario: Multi-Health Systems, 1995.
    30. ↵
      1. Ehrich M,
      2. Bocker S,
      3. van den Boom D
      . Multiplexed discovery of sequence polymorphisms using base-specific cleavage and MALDI-TOF MS. Nucleic Acids Res 2005;33:e38.
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Laird NM,
      2. Horvath S,
      3. Xu X
      . Implementing a unified approach to family-based tests of association. Genet Epidemiol 2000;19(Suppl 1):S36–42.
      OpenUrlCrossRefPubMedWeb of Science
    32. ↵
      1. Choudhry Z,
      2. Sengupta SM,
      3. Grizenko N,
      4. et al
      . LPHN3 and attention-deficit/hyperactivity disorder: interaction with maternal stress during pregnancy. J Child Psychol Psychiatry 2012;53:892–902.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Hansen T,
      2. Ingason A,
      3. Djurovic S,
      4. et al
      . At-risk variant in TCF7L2 for type II diabetes increases risk of schizophrenia. Biol Psychiatry 2011;70:59–63.
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Laucht M,
      2. Hohm E,
      3. Esser G,
      4. et al
      . Association between ADHD and smoking in adolescence: shared genetic, environmental and psychopathological factors. J Neural Transm 2007;114:1097–104.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Sousa NO,
      2. Grevet EH,
      3. Salgado CA,
      4. et al
      . Smoking and ADHD: an evaluation of self medication and behavioral disinhibition models based on comorbidity and personality patterns. J Psychiatr Res 2011;45:829–34.
      OpenUrlCrossRefPubMed
    36. ↵
      1. Dinn WM,
      2. Aycicegi A,
      3. Harris CL
      . Cigarette smoking in a student sample: neurocognitive and clinical correlates. Addict Behav 2004;29:107–26.
      OpenUrlCrossRefPubMedWeb of Science
    37. ↵
      1. Winterer G,
      2. Mittelstrass K,
      3. Giegling I,
      4. et al
      . Risk gene variants for nicotine dependence in the CHRNA5-CHRNA3-CHRNB4 cluster are associated with cognitive performance. Am J Med Genet B Neuropsychiatr Genet 2010;153B:1448–58.
      OpenUrlCrossRefPubMed
    38. ↵
      1. Sanchez-Mora C,
      2. Ribases M,
      3. Ramos-Quiroga JA,
      4. et al
      . Meta-analysis of brain-derived neurotrophic factor p.Val66Met in adult ADHD in four European populations. Am J Med Genet B Neuropsychiatr Genet 2010;153B:512–23.
      OpenUrlPubMed
    39. ↵
      1. Gizer IR,
      2. Ficks C,
      3. Waldman ID
      . Candidate gene studies of ADHD: a meta-analytic review. Hum Genet 2009;126:51–90.
      OpenUrlCrossRefPubMedWeb of Science
    40. ↵
      1. Lloyd SE,
      2. Maytham EG,
      3. Pota H,
      4. et al
      . HECTD2 is associated with susceptibility to mouse and human prion disease. PLoS Genet 2009;5:e1000383.
      OpenUrlCrossRefPubMed
    41. ↵
      1. Lloyd SE,
      2. Rossor M,
      3. Fox N,
      4. et al
      . HECTD2, a candidate susceptibility gene for Alzheimer's disease on 10q. BMC Med Genet 2009;10:90.
      OpenUrlCrossRefPubMed
    42. ↵
      1. Freedman ML,
      2. Monteiro AN,
      3. Gayther SA,
      4. et al
      . Principles for the post-GWAS functional characterization of cancer risk loci. Nat Genet 2011;43:513–18.
      OpenUrlCrossRefPubMed

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • NG and RJ contributed equally to the collection of the research project.

    • Contributors RJ and NG designed the study and acquired the data; GAT, SMS, ZC and RJ analysed the data. GAT and RJ wrote the article; SMS, NG and ZC reviewed the article. All authors approved its publication.

    • Funding This work was supported by grants from the Fonds de la recherche en santé du Québec and the Canadian Institutes of Health Research to RJ and NG. GAT holds a doctoral award from the CIHR. SMS is a recipient of the 2008 NARSAD Young Investigator and 2009 Dr Mortimer D. Sackler Developmental Psychology Investigator Awards. RJ receives consultancy honorarium from Janssen Ortho and Pfizer Canada.

    • Competing interest None.

    • Ethics approval The Research Ethics Board of the Douglas Mental Health University Institute approved this study.

    • Provenance and peer review Not commissioned; externally peer reviewed.