“A Genetically Informed Registered Report on Adverse Childhood Experiences and Mental Health”, 2022-12-08 (; backlinks):
Children who experience adversities have an elevated risk of mental health problems. However, the extent to which adverse childhood experiences (ACEs) cause mental health problems remains unclear, as previous associations may partly reflect genetic confounding.
In this Registered Report, we used DNA from 11,407 children from the United Kingdom and the United States to investigate gene-environment correlations and genetic confounding of the associations between ACEs and mental health.
Regarding gene-environment correlations, children with higher polygenic scores for mental health problems had a small increase in odds of ACEs. Regarding genetic confounding, elevated risk of mental health problems in children exposed to ACEs was at least partially due to pre-existing genetic risk. However, some ACEs (such as childhood maltreatment and parental mental illness) remained associated with mental health problems independent of genetic confounding.
These findings suggest that interventions addressing heritable psychiatric vulnerabilities in children exposed to ACEs may help reduce their risk of mental health problems.
…Controlling for polygenic scores for mental health problems in this manner can indicate whether there is likely to be a genetic contribution to the association between ACEs and mental health. However, one limitation of this methodological approach is that polygenic scores capture only a small proportion of heritability and thus do not fully account for genetic confounding. This can be addressed by a newly developed genetic sensitivity analysis that estimates shared genetic effects under scenarios in which the polygenic score captures additional genetic variance in the outcome (that is, SNP-based and/or twin-based heritability; see ‘Analysis plan’ in the Method for a detailed description of this method). A recent application of this genetic sensitivity analysis found that the associations between maternal education and offspring ADHD, educational achievement and body mass index were moderately explained by shared genetic effects27, consistent with findings from children-of-twins studies and adoption designs. For example, a latent polygenic score that captured SNP-based heritability in educational achievement (that is, 31%; ref. 32) explained 50% of the association between maternal education and child educational achievement27. However, this approach has never been applied to assess the extent to which genetic influences contribute to the associations between ACEs and mental health.
…Despite our cautious interpretation surrounding specific estimates of genetic confounding, the overall pattern of results supports findings from other genetically informed designs with different assumptions and sources of bias. For example, we found that child maltreatment was largely associated with internalizing and externalizing problems, independent of genetic confounding. This is consistent with evidence of causal effects of maltreatment on psychopathology from Mendelian Randomization42, co-twin control43 and other quasi-experimental studies44. We also found that parental mental illness was associated with internalizing and externalizing problems independent of genetic confounding, which supports evidence from children-of-twins and adoption studies45,46,47.
In contrast, we found that parental substance abuse, parental criminality and parental separation were predominantly associated with internalizing and externalizing problems via genetic confounding. Notably, similar genetically confounded associations with psychopathology have also been reported for parental substance abuse in children-of-twins48,49 and adoption studies50, for parental criminality in an adoption study51, and for parental separation in some52 (though not all53) children-of-twins studies.