Erik Pettersson

Purpose and aims: There is so much comorbidity among psychiatric conditions that one can combine them into a single general psychopathology index. Although general psychopathology predicts adverse outcomes, it remains unknown if it responds to intervention. Therefore, we will estimate the causal effect of psychosocial interventions and psychotropic medications on general psychopathology. Method: We will apply casual inference designs (children-of-siblings design

co-twin control/within-individual design

instrumental variable

and front-door criterion) to data from the Swedish Twin Register and Swedish population registers to estimate the effect of childrearing conditions (year 1), psychotherapy (years 2-3), and psychotropic medication (years 1-5) on general psychopathology (derived from self/parent-reported symptoms and psychiatric diagnoses). The principal investigator will devote 50% of his time to this project, and he is supported by experts in epidemiology, psychotherapy, and psychiatry. Importance: Even though comorbidity is the rule rather than exception in psychiatry, most randomized clinical trials exclude individuals with several disorders. Therefore, there is a lack of knowledge of the effect of treatment on comorbidity. By applying innovative causal inference designs to large observational data, we aim to study the effect of treatment on general psychopathology in the best conceivable way. This could open a new area of research into transdiagnostic treatments.

Research problem and specific questions: About one quarter of all children have parents with mental health problems. These children are at increased risk for a wide range of adverse outcomes. Nevertheless, despite the importance of reducing this form of inequality, research on the intergenerational transmission of mental health problems is plagued by three important limitations. These include an overreliance on small samples and retrospective (biased) reports

a failure to address psychiatric comorbidity

and inadequate control of unmeasured confounding. The aim of this project is to answer three research questions: 1) What are the associations between psychiatric diagnoses in parents and adverse outcomes in their children? 2) Can these be attributed to broad comorbidity? 3) Does treatment of the parents’ psychiatric disorders reduce the risk of adverse outcomes in children?Data and method: We will include all individuals born in Sweden between 1970 and 2000 (N = 2 797 086). The exposures are 6 psychiatric diagnoses in their parents, and the outcomes are 34 adverse events in the offspring recorded until the end of 2019 (e.g., psychiatric disorders, psychotropic medications, criminality, school and employment problems, etc.). We will estimate associations between the exposures and outcomes

adjust for comorbidity using multiple regression and a general factor model

and apply causal inference techniques to control for unmeasured confounding.Plan for project realisation: We will hire a PhD student and a postdoc to complete the project. The PI, along with experts in epidemiology and psychiatry, will provide support.Relevance: Our project will contribute to the study of intergenerational transmission of mental health problems in three important ways. First, because psychiatric treatment resources are becoming increasingly scarce as more individuals seek help, parental psychiatric history can help guide treatment allocation by identifying those at higher risk and by screening out low risk persons. Second, we will highlight the importance of focusing on parental comorbidity, which will guide clinicians to additionally focus on the number of parental diagnoses (rather than only type) when predicting patient prognosis. Third, if a parental psychiatric diagnosis appears causally related to an outcome, then it would be beneficial to develop a policy for child and adolescent mental health providers to recommend that the parent also seek treatment.