Assessing the Phenotypic and Genetic Structure of Psychopathology Across the Life Course

• In Event: 1-180 - Genetic Effects on Health and Education Across the Life Course

Integrative Statement

Major depressive disorder, anxiety disorders, and opioid use disorders rank among the top 10 diseases with the largest impact on disability-adjusted life years (US Burden of Disease Collaborators, 1990-2016). Such mental health disorders are potent predictors of health, wealth, and wellbeing across the life course (Kate et al., 2015; Schaefer et al., 2017), making our understanding of their phenotypic presentation and etiology critical for treatment and prevention efforts. A burgeoning literature suggests that there is an organized phenotypic structure of psychopathology, in which internalizing behaviors (e.g., anxiety, depression) load onto one factor and externalizing behaviors (e.g., attention problems, antisocial behavior) load onto another; thus creating a “two-factor” meta-structure. (Lahey, 2017). Behavioral genetic studies also report a two-factor model of additive genetic risk for psychopathology (e.g., Kendler et al., 2003). The purpose of the current paper is to extend this work by investigating whether a two-factor phenotypic and genotypic model (i.e., as compared to a one-factor model) of psychopathology fits similarly across the life course using two nationally-representative samples. Consistent with previous research in community-based samples, we expect a two-factor model of psychopathology to fit similarly across developmental stages. However, given several recent empirical reports suggesting that polygenic risk for psychopathology is highly comorbid across domains (e.g., Brikell et al., 2018), we expect that a one-factor genotypic model of psychopathology will fit better than a two-factor, and similarly across developmental periods.

Data were drawn from the Fragile Families and Child Wellbeing Study (FFCWS) and the Health and Retirement Study (HRS). The FFCWS is a nationally-representative sample of 4,898 children and their families followed from birth through age 15, with an oversample of non-marital births. Phenotypic data measuring internalizing and externalizing behaviors was drawn from multi-informant reports at age 15. Genetic data was available for N = 2,655 youth. The HRS is a nationally-representative sample of U.S. adults over age 50. Phenotypic data was drawn from the 2010/2012 data collection, which included self-report measures of internalizing and externalizing behaviors. Genetic data was available for N = 10,342. Genome-wide polygenic genic scores were derived using weights from large GWAS of psychiatric outcomes (Table 1). Structural equation modeling was used to evaluate whether one-factor or two-factor phenotypic and genotypic models of psychopathology fit the data within each cohort. To ensure replicability, all models were fit on a random half of the data and replicated in the second half of the data.

Preliminary results using the HRS indicated that a two-factor phenotypic model fit the data better than a one-factor model, as indicated by a chi-square difference test (Figure 2a). There was no difference in model fit of the two-factor model when HRS participants were divided by age into young old (aged 51 to 64) and older adult (aged 65 to 82) groups. By contrast, the one-factor genotypic model fit the data better than a two-factor model in both the cohorts (Figure 2b). The same analysis will be completed in the FFCWS data to evaluate whether these phenotypic and genotypic structures fit similarly during adolescence.

Authors

• Arianna Morgan Gard, University of Michigan, Ann Arbor
  Presenting Author

• Erin B Ware, University of Michigan, Ann Arbor
  Non-Presenting Author

• Luke W. Hyde, University of Michigan
  Non-Presenting Author

• Lauren Schmitz, University of Michigan, Ann Arbor
  Non-Presenting Author

• Jessica Faul, University of Michigan, Ann Arbor
  Non-Presenting Author

• Daniel Notterman, Princeton University
  Non-Presenting Author

• Colter Mitchell, University of Michigan
  Non-Presenting Author