Importance Depression is associated with alterations in blood immuno-metabolic biomarkers, but it remains unclear whether these alterations are limited to select measures or represent broader patterns and can predict depression diagnosis. Objective To examine immuno-metabolic biomarker changes in depression, pattern of effect at the symptom and symptom-dimension level, and prediction of depression diagnosis. Design, Setting, and Participants Case-control and cohort-wide analyses of ICD-10 depression, depressive and anxiety symptoms based on up to N=4161 participants (2363 female) aged 24 years from the Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort. Exposures Blood-based immunological and metabolic biomarkers (n=93) comprising inflammatory proteins, cell count, lipids, hormones, and metabolites. Main Outcomes and Measures ICD-10 diagnosis of depression, 11 individual depressive and anxiety symptoms, and four domain scores were used as outcomes. Confounders included sex assigned at birth, body mass index, smoking, and alcohol use. Results After adjusting for potential confounders and multiple testing, depression was associated with changes in concentrations of specific immuno-metabolic markers (IL-6, CDCP1, neutrophil count, and insulin), and greater number of extreme-valued inflammatory markers. We identified three distinct affective symptom-related biomarker clusters, including one comprising inflammatory cytokines, chemokines and cells which was positively associated with somatic and mood symptoms, and one comprising liver-related biomarkers which was negatively associated with anxiety symptoms. Then using Partial Least Squares regression we identified two latent variables that capture the biomarker-symptom associations (Component 1: Somatic-Depressive-Inflammation and Component 2: Anxiety-Hepatic). Higher Component 1 score was associated with higher depressive symptom severity consistently over subsequent five years. Immuno-metabolic biomarkers performed poorly in predicting ICD-10 depression (0.569 Balanced Accuracy). However, within depression cases the addition of immuno-metabolic biomarkers improved the prediction of depression with high levels of mood (0.720 Balanced Accuracy) or anxiety symptoms (0.636 Balanced Accuracy). Conclusion and Relevance Depression is associated with disruption in immuno-metabolic homeostasis. Specific patterns of immuno-metabolic biomarkers are associated with differing subsets of affective symptoms, which are potentially relevant for poor depression prognosis. Immuno-metabolic biomarkers improve predictions of high levels of mood symptoms within people with depression, highlighting the symptom-level heterogeneity of depression and opportunities for immuno-metabolic biomarker-based subtyping, prediction, and targeted intervention. Key Points Question Depression is linked to immuno-metabolic dysfunction, but what is the precise nature of these associations at biomarker and symptom level, and can we predict depression using immuno-metabolic biomarkers? Findings Using 93 blood immuno-metabolic biomarkers and depression measures from up-to 4196 participants, we identified distinct clusters/groupings of immuno-metabolic biomarkers and depressive and anxiety symptoms which are differently associated with subsequent symptom persistence. These biomarkers predict specific symptom profiles better than others within people with depression. Meaning Heterogeneity in the associations of immuno-metabolic biomarkers with mood and anxiety symptoms is relevant for prognosis and could aid better stratification/prediction of depression.
No takes yet. Share an insight, caveat, or question.
Donnelly et al. (2024) studied this question.