Disentangling shared and unique effects of parenting on psychopathology: Evidence from two prospective, genetically sensitive cohort studies

Background Multiple dimensions of parenting–including hostility, conflict, and warmth–are associated with an array of mental disorders in youth. We tested two possible explanations for these associations: Parenting may shape a shared liability to experience many forms of psychopathology (“shared‐liability” hypothesis), or parenting may shape unique liabilities to externalizing and internalizing psychopathology (“unique‐liability” hypothesis). Methods Participants were from the Early Growth and Development Study (EGDS; N = 561 adoptive families) and the Michigan Twin Neurogenetics Study (MTwiNS; N = 354 twin families). In EGDS, we measured parental hostility and warmth at child age 27 months. In MTwiNS, we measured parent‐child conflict and parental involvement at child mean age 8.1 years. We extracted shared and unique liabilities from child‐psychopathology data using structural equation models. We tested whether parenting longitudinally predicted psychopathology, and to what extent associations between parenting and children's externalizing and internalizing psychopathology liabilities were explained by associations with a general psychopathology liability (“p”). We addressed genetic confounding by testing associations of adoptive parents' parenting with psychopathology in their genetically unrelated children, and by using a twin‐difference design. Results Across ages and cohorts, parenting characterized by hostility and conflict was associated with greater p, consistent with the shared‐liability hypothesis. Also consistent, parental warmth was associated with lower p in middle and late childhood in EGDS participants, though parental involvement was not associated with p in adolescent MTwiNS participants. There was limited support for the unique‐liability hypothesis; only one association (with externalizing psychopathology) remained after accounting for p. In childhood, but not adolescence, findings remained after accounting for genetic confounding. Conclusion Parenting behaviors, particularly hostility and conflict, are associated with children's broad vulnerability to psychopathology, and these associations are evident in childhood even after controlling for genetic influences. These findings support parenting interventions as a means to reduce children's liability to diverse mental‐health problems.