A network analysis of depressive symptoms and metabolomics

• DOI: 10.1017/s0033291723001009
• Publication Date: 2023-04-24T13:06:46Z
• Authors: Arja O. Rydin; Yuri Milaneschi; Rick Quax; Jie Li; Jos A. Bosch; Robert A. Schoevers; Erik J. Giltay; Brenda W. J. H. P...; Femke Lamers
• Keywords: Depression

Depression is associated with metabolic alterations including lipid dysregulation, whereby associations may vary across individual symptoms. Evaluating these using a network perspective yields more complete insight than single outcome-single predictor models.We used data from the Netherlands Study of and Anxiety (N = 2498) leveraged networks capturing between 30 depressive symptoms (Inventory Depressive Symptomatology) 46 metabolites. Analyses involved 4 steps: creating Mixed Graphical Models; calculating centrality measures; bootstrapping for stability testing; validating central, stable by extra covariate-adjustment; validation another wave collected 6 years later.The yielded 28 symptom-metabolite associations. There were 15 highly-central variables (8 symptoms, 7 metabolites), 3 links involving Low energy (fatigue), Hypersomnia. Specifically, fatigue showed consistent higher mean diameter VLDL particles lower estimated degree (fatty acid) unsaturation. These remained present after adjustment lifestyle health-related factors wave.The somatic Fatigue Hypersomnia cholesterol fatty acid measures stable, relationships in our network. The analyses how are consistently linked to specific symptom profiles.