Maternal and child genetic liability for smoking and caffeine consumption and child mental health: an intergenerational genetic risk score analysis in the ALSPAC cohort

Abstract

Background and aims Previous studies suggest an association between maternal tobacco and caffeine consumption during and outside of pregnancy and offspring mental health. We aimed to separate effects of the maternal environment (intrauterine or postnatal) from pleiotropic genetic effects. Design Secondary analysis of a longitudinal study. We (i) validated smoking and caffeine genetic risk scores (GRS) derived from published genome‐wide association study (GWAS) for use during pregnancy, (ii) compared estimated effects of maternal and offspring GRS on childhood mental health outcomes and (iii) tested associations between maternal and offspring GRS on their respective outcomes. Setting We used data from a longitudinal birth cohort study from England, the Avon Longitudinal Study of Parents and Children (ALSPAC). Participants Our sample included 7921 mothers and 7964 offspring. Measurements Mental health and non‐mental health phenotypes were derived from questionnaires and clinical assessments: 79 maternal phenotypes assessed during and outside of pregnancy and 71 offspring phenotypes assessed in childhood (<10 years) and adolescence (11–18 years). Findings The maternal smoking and caffeine GRS were associated with maternal smoking and caffeine consumption during pregnancy (2nd trimester: Psmoking = 3.0 × 10−7, Pcaffeine = 3.28 × 10−5). Both the maternal and offspring smoking GRS showed evidence of association with reduced childhood anxiety symptoms (βmaternal = −0.033; βoffspring = −0.031) and increased conduct disorder symptoms (βmaternal = 0.024; βoffspring = 0.030), after correcting for multiple testing. Finally, the maternal and offspring smoking GRS were associated with phenotypes related to sensation seeking behaviours in mothers and adolescence (e.g. increased symptoms of externalising disorders, extraversion and monotony avoidance). The caffeine GRS showed weaker evidence for associations with mental health outcomes. Conclusions We did not find strong evidence that maternal smoking and caffeine genetic risk scores have a causal effect on offspring mental health outcomes. Our results confirm that the smoking genetic risk scores also captures liability for sensation seeking personality traits.

Prenatal smoking, alcohol and caffeine exposure and maternal‐reported attention deficit hyperactivity disorder symptoms in childhood: triangulation of evidence using negative control and polygenic risk score analyses

“…Although in GenR and MoBa we found suggestive evidence for a causal relationship of maternal prenatal smoking on high risk of maternal‐reported ADHD symptoms, when comparing the findings across the cohorts, reporters and questionnaires the evidence was weak and inconsistent. Additionally, our PRS analyses with lifetime smoking PRS in ALSPAC and MoBa indicated pleiotropic associations which are consistent with recent findings in ALSPAC [54]. There is also a large body of evidence showing pleiotropy between smoking, impulsivity and sensation‐seeking type of personality [55, 56] which could confound observed phenotype associations in the present study.…”

Exploring pleiotropy in Mendelian randomisation analyses: What are genetic variants associated with “cigarette smoking initiation” really capturing?

“…Some (although not all) of these results were replicated in ALSPAC. Our results are in line with recent studies suggesting associations of increased smoking initiation PRS with increased risk-taking behaviours and decreased age at first birth, amongst other phenotypes [12,13,22]. However, we also found novel associations with other phenotypes not plausibly caused by smoking, suggesting that these potential pleiotropic effects may occur via a range of pathways.…”

“…However, subsequent studies have suggested that some of these SNPs may also be influencing different phenotypes via independent pathways (i.e., not via smoking). For example, in studies by Khouja and colleagues [12] and by Schellhas and colleagues [13],…”

The effect of maternal BMI, smoking and alcohol on congenital heart diseases: a Mendelian randomisation study

“…The study was a GWAS of a lifetime smoking index (which combined smoking initiation, duration, heaviness and cessation), conducted in a sample of 462,690 current, former and never smokers in UK Biobank [ 30 ]. A GRS based on the lifetime smoking GWAS has previously been shown to be associated with smoking behaviours during pregnancy in the ALSPAC cohort [ 31 ]. For the alcohol weighted GRS, there were 99 conditionally independent SNPs ( P < 5 × 10 −8 ), measured as the number of alcoholic drinks per week [ 32 ].…”