Common mental distress, neuroticism, and markers of inflammation : network and latent variable analyses

Abstract

Aims: Inflammatory processes have been linked to depression as well as common mental distress (CMD). Similarly, the personality dimension ‘neuroticism’ has been shown to predict CMD and its health-related costs. But questionnaires to assess CMD and neuroti- cism often cover related content. The goal of this presentation is to investigate whether patient-reported outcome dimensions of CMD can be differentiated from neuroticism and the degree to which inflammatory markers are related to these assessments. Methods: Blood samples of N = 2835 adults (British Household Panel Survey 2009) were analysed for levels of C-Reactive Protein (CRP) and Fibrinogen (FIB). Health-related Quality of Life (HRQoL) and CMD were assessed with the General Health Questionnaire-12 and the Short Form-12. Neuroticism was assessed with the Big Five Inventory (short-form). Network analysis was used to explore all bivariate relationships and common sources of variance. Exploratory structural equation modelling (ESEM) for categorical data was employed to explore the relationships between inflammatory markers and latent variables. Model results were cross-validated in a randomly selected hold-out sample. Results: Network analysis showed (1) a gradated relationship from inflammatory markers, over ratings of physical health, to CMD and finally to neuroticism; and (2) that the variables shared large amounts of variance. ESEM showed good fit for a bifactor model with one general and three specific factors [RMSEA = .038; CFI/TLI = .98/.98] and a model with five corre- lated factors [RMSEA = .033; CFI/TLI = .99/.98]. In both models, CRP was significantly related to CMD factors, while FIB was only related to ratings of physical health (controlling for gender, age, chronic conditions, medication). Higher levels of these markers cor- related with lower HRQoL/higher CMD. Effect sizes were small (CRP: r = .05–.14; FIB: r = .24–.25), but larger in the bifactor model. Tests for moderation effects showed that the relationship was more pronounced in females. Conclusions: In the general population, inflammatory markers are differentially linked with subjective ratings of physical and psychological health. Neuroticism is only distally connected to such markers. Although these findings are currently only cross-sectional and need to be replicated, they are in line with theories of PRO assessment (e.g., Wilson & Cleary, 1995, JAMA, 59–65) and speak to neuroticism as part of a potential CMD endophenotype.