An Integrative Approach to the Etiology of Internalizing and Functional Disorders in the Lifelines Cohort

Although internalizing Disorders (IDs) together account for a substantial proportion of global disease burden, we still have a quite limited understanding of their causes. The overarching aim of this project is to clarify the etiology, course and comorbidities of five major internalizing disorders: Major Depression [MD], Dysthymia [DYS], Generalized Anxiety disorder [GAD], Panic Disorder [PD], and Social Phobia [SPH]. To aid this effort, we also study the three major Functional Disorders (FDs): Chronic Fatigue Syndrome [CFS], Fibromyalgia [FM] and Irritable Bowel Syndrome [IBS]. We do so because i) IDs and FDs are highly comorbid and share a number of etiologic pathways, clarification of which will further improve our understanding of IDs, and ii) FDs are markedly under-researched given their considerable public health burden4. This project will take place in Lifelines, a prospective population-based cohort in North-Eastern Netherlands following 167,729 subjects since 2006 with assessments of psychiatric and general medical health every five years. Lifelines includes high quality data on these phenotypes, relevant environments, and genotypes of multiple generations of family members. These rich data allow for an in-depth analysis of disease mechanisms for IDs and FDs. We will build multivariate models based on phenotypic, environmental, familial, and molecular genetic levels to understand which mechanisms are shared among versus specific to these eight related conditions. To achieve this goal, we propose three specific aims:

13.    Research questions 
Please be specific regarding the research questions.
Aim 1 – To investigate the associations between IDs and between IDs and FDs at the phenotypic level and clarify their causal inter-relationships, we will leverage the longitudinal and family structure of Lifelines using standard and newly developed methods:

1a: Examine the impact of age and sex with generalized additive models and age-period-cohort analyses to inform hypotheses about shared versus unique etiologies. 
1b: Explore the structure of comorbidity patterns within and over time using factor, network, and latent transition analyses on both disorder and symptom levels to obtain an empirical typology of IDs and IDs/FDs. 
1c: Clarify the degree to which the observed associations within IDs, between IDs, and between IDs and FDs are likely to be causal using cross-lagged, co-sibling, and propensity matching designs. 

Aim 2 – To identify shared and unique genetic risk factors for IDs/FDs, we will utilize family and molecular genetic data: 

2a: Clarify familial transmission and genetic architecture using parent-offspring, sibling, and molecular genetic designs and calculate familial and single nucleotide polymorphism (SNP) based genetic correlations for IDs/FDs. Because of the heterogeneity of IDs/FDs we will extend our analyses to more homogeneous empirical subtypes of IDs/FDs, identified by factor analyses, latent class analyses and factor mixture models.
2b: Identify shared and unique SNP and polygenic risk score (PRS) effects for IDs/FDs using genome-wide structural equation models (‘GW-SEM’ or “Genomic SEM”).
2c: Further investigate genetic architecture of IDs/FDs by analyses of blood and brain expression quantitative trait loci (eQTLs). To evaluate the accuracy of our eQTL-based imputation of gene expression levels, we will expand our analyses by sensitivity analyses on experimentally measured whole blood gene expression in the Lifelines Deep cohort.  
2d: Utilize within-family Mendelian Randomization (MR) to clarify causal relations between IDs and FDs.

Aim 3 –To investigate the joint action and interaction of genetic and environmental risk factors on the development(s) of IDs/FDs, we will:

3a: Use regularized regression methods to identify common and unique early and recent environmental risk factors for the onset of IDs/FDs.  
3b: Investigate potential gene-environment interactions using family history x environmental risk and PRS x environmental risk designs to predict onset of IDs/FDs.
3c: Elucidate how genetic nurture effects –the impact of parental genes on offspring IDs/FDs through the rearing environment– contribute to risk of IDs/FDs.