By Kristina Aagaard
Principal supervisor: Professor Klaus Bønnelykke, COPSAC
Principal co-supervisor: Professor Bo Chawes, COPSAC
Co-supervisors: PhD Jens Richard Møller-Jepsen (CSNR), PhD Rebecca Vinding and PhD Niklas Brustad (COPSAC)
Chairperson: Christina Engel Høi-Hansen, MD, PhD, Professor, Department of Clinical Medicine, University of Copenhagen, Denmark
Opponent:Monica Guxens, MD, PhD, Professor, Barcelona Institute for Global Health, Campus MAR, Spain
Opponent: Kerstin Von Plessen, MD, PhD, Professor, Department of Psychiatry, University of Lausanne, Switzerland
Summary
This thesis investigates prenatal risk factors for neurodevelopmental disorders, with a particular focus on Attention-Deficit/Hyperactivity Disorder (ADHD) and autism. The identification of early life exposures associated with later presentation of these disorders may provide valuable insights for future prevention strategies. ADHD and autism are known to be highly heritable, with genetic predisposition representing the strongest risk factor. This strong genetic component to neurodevelopmental disorders may introduce bias in observational studies investigating early environmental influences. In the attempt to address this, this thesis investigates the causality of proposed early life risk factors using genetically informed study designs that have the ability to account for potential confounding due to genetics.
The studies included in this thesis focused on the effect of a high-dose vitamin D3 supplementation throughout third trimester of pregnancy on risk of ADHD and autism as well as impaired cognitive functioning evaluated at the age of 10 at the Copenhagen Prospective Study on Neuro-PSYCHiatric Development (COPSYCH) visit. The COPSYCH 10-year visit consisted of an extensive psychopathological and neurocognitive evaluation of the Copenhagen Prospective Studies on Asthma in Childhood (COPSAC) 2010 cohort constituting 700 mother- child pairs followed prospectively since week 24 of pregnancy. The thesis further investigated the potential role of genetic confounding of the observed influence of a maternal unhealthy dietary pattern in pregnancy and maternal inflammation in pregnancy on neurodevelopment utilizing both the COPSAC cohort and two larger external mother-child cohorts: the Avon longitudinal study of parents and children (ALSPAC) including 1,199 trios with information on maternal, paternal and child genetics (trios) and the Norwegian mother, father, and child cohort study (MoBa) including 41,580 trios.
As a post hoc analysis of the high-dose vitamin D3 randomized clinical trial (RCT) conducted in the COPSAC2010 cohort throughout third trimester of pregnancy, we did not show any effect of this intervention on offspring risk of ADHD or autism at the age of 10 in Paper I. The paper further included observational analyses of the association between maternal pre-intervention blood levels of vitamin D and offspring risk of neurodevelopmental disorders. Adjusted for potential confounders, higher serum vitamin D in pregnancy was associated with a decreased risk of autism, fewer autistic traits, and a decreased risk of ADHD. 11As existing observational studies also have pointed to an association between pregnancy vitamin D and offspring cognition, we in Paper II conducted an additional post hoc analysis of the COPSAC2010 vitamin D3 intervention investigating the effect hereof on 11 cognitive functions likewise tested at the COPSYCH visit at age 10. The results were suggestive of a positive effect of the high-dose supplementation on verbal memory, visual memory and flexibility/set shift. However, these effects did not pass multiple test correction. In observational analyses, we only showed a positive association between higher serum vitamin D in pregnancy and offspring improved flexibility/set shift.
In Paper III we conducted a genetic trio analysis of the suggested association between an unhealthy maternal dietary pattern in pregnancy and offspring risk of ADHD. These analyses were performed using genetic risk scores for a healthy/unhealthy diet. By including maternal, paternal and child genetic risk scores in the analyses, we were able to investigate potential causal effects of maternal diet in pregnancy on ADHD adjusting out potential bias from overlapping genetic predisposition to both eating pattern and ADHD risk. The analyses in the COPSAC2010 cohort suggested a causal negative effect of an unhealthy diet in pregnancy on severity of ADHD traits, however this finding was not replicated in the larger ALSPAC and MoBa cohorts. Inflammation in pregnancy has been associated with risk of offspring neurodevelopmental disorders. Therefore, we in Paper IV investigated potential causal effects of pregnancy inflammation on neurodevelopmental traits and diagnoses using genetic risk scores for C- reactive protein (CRP), interleukin-6 (IL-6) and glycoprotein acetyls (GlycA) in pregnancy. The analyses were conducted in the MoBa cohort using the same genetic trio approach as described for Paper III. We included two additional methods to investigate the association: intergenerational mendelian randomization (MR) to test for causal effects of inflammation and classic two-sample MR to investigate potential confounding of observational associations. Neither the trio analyses nor the intergenerational MR analyses suggested causal effects. Two- sample MR suggested risk of genetic confounding of previously reported observational associations.
In conclusion, this thesis provides a thorough investigation of the influence of vitamin D in pregnancy on offspring neurodevelopment at the age of 10. A high dose vitamin D3 intervention in third trimester of pregnancy did not affect the risk of neurodevelopmental disorders at age 10. The supplementation however had a suggestive positive effect on verbal memory, visual memory 12and flexibility/set shift; however, these results did not pass multiple test correction. Finally, using genetically informed study designs, we investigated the causal influences of maternal dietary pattern and maternal inflammation in pregnancy on offspring risk of neurodevelopmental disorders. Our findings were not supportive of causal effects of either diet or inflammation in pregnancy on offspring neurodevelopment, but suggested risk of genetic confounding of observational associations.
