The polygenic and brain correlates of attention deficit hyperactivity symptoms and overweight and obesity

• In Event: 1-051 - Use of polygenic scores in developmental psychology

Integrative Statement

Importance: There is an extensive body of literature linking attention deficit hyperactivity (ADHD) symptoms to overweight and obesity. In brief, during the late teens-to-adulthood individuals showing symptoms of ADHD tend to also gain weight. Traditional explanations have focused on impulsivity and/or inattention symptoms as a gateway to unhealthy snacking and weight gain. However, given that both ADHD and overweight/obesity show moderate heritability, a common genetic basis is also a possibility. However, despite the established link between ADHD and overweight and obesity, the biological basis of this association is poorly understood.

Objective: There are two primary objectives for this study: (i) Examine if polygenic scores scores (PGS) of ADHD and body mass index (BMI), respectively, are associated with one another, thus suggesting a shared genetic basis for these conditions; (ii) To determine whether these two phenotypes, and their polygenic risk scores, are associated with a shared neural substrate.

Design, Setting, and Participants: The sample for this study comprised participants in IMAGEN, a general population sample of Caucasian adolescents recruited from eight sites across France, Ireland, England and Germany. Participants were assessed at ages 14 and 19. In the present study, we focus on individuals at age 19.

Main Outcome and Measures: ADHD symptoms (from the Barratt Impulsivity Scale) and BMI were measured in 920 IMAGEN participants at age 19. T1 and DTI weighted MRI, and task and resting-state fMRI scans were also taken at this age. Genome-wide loci were available from blood collected at age 14.

Analysis: We used a novel, two-stage, sparse multiple canonical correlation analysis framework to find neuroimaging measures that are simultaneously associated with ADHD, BMI and their respective PGS.

Results: We replicated the results of previous investigations by finding an association between measurements of ADHD symptoms and BMI (r=0.11,n = 920, p =4x10-4). Furthermore, we found that the PGS of ADHD and BMI were also correlated (r=0.19, n=920, p<1x10-5). Using a whole brain analysis approach, the PGS of ADHD (r = 0.39, n = 920, p =0.019) and the PGS of BMI (r = 0.37, n = 920, p = 0.037) jointly associated with a single, multimodal brain phenotype (i.e. orbitofrontal, bi-lateral amygdala and other reward award related areas). This multimodal brain phenotype, in turn, associated with the measurements of ADHD (r = 0.38, n = 920, p = 0.023) and BMI (r = 0.41, n = 920,p = 0.012).

Conclusions and Relevance: Although a large body of literature linking ADHD and BMI, investigations into the biological underpinnings of this comorbidity have been limited. In the present investigation, we found ADHD and BMI shared genetic and neuronal liability. These results are important as they suggest (a) potential developmental mechanisms, and (b) targets for multimodal interventions. We also introduce a novel, two-stage, sparse multiple canonical correlation analysis framework, which is generally applicable to research aimed at associating multimodal multivariate measures (e.g. biological with several behavioural and psychological trajectories).

Authors

• Alex Ing, King's College London
  Presenting Author

• Edward Dylan Barker, King's College London
  Non-Presenting Author

• Francesca Biondo, King's College London
  Non-Presenting Author

• Tianye Jia, Fundan University
  Non-Presenting Author

• JB Pingault, University College London
  Non-Presenting Author

• Ebba Du Rietz, King's College London
  Non-Presenting Author

• Gunter Schumann, King's College London
  Non-Presenting Author