Role of inflammation in depression and anxiety: Tests for disorder specificity, linearity and potential causality of association in the UK Biobank.

Abstract

BackgroundConcentrations of C-reactive protein (CRP), interleukin 6 (IL-6) and other inflammatory markers are elevated in people with depression and anxiety compared to controls, but evidence for disorder-specificity, linearity and potential causality is sparse.MethodsUsing population-based data from up to 144,890 UK Biobank cohort participants, we tested associations of circulating CRP concentrations with depression and anxiety symptom scores and probable diagnosis, including tests for linearity, disorder-specificity and sex difference. We examined potential causality using 1-sample and 2-sample Mendelian randomisation (MR) analyses testing associations of genetically-predicted CRP concentration and IL-6 activity with depression and anxiety. The study was conducted from June 2019 to February 2021.FindingsCRP concentration was associated with depressive and anxiety symptom scores and with probable diagnoses of depression and generalised anxiety disorder (GAD) in a dose-response fashion. These associations were stronger for depression than for anxiety, and for women than for men although less consistently. MR analyses provided consistent results suggesting that genetically predicted higher IL-6 activity was associated with increased risk for depressive symptoms, while genetically-predicted higher CRP concentration was associated with decreased risks of depressive and anxiety symptoms.InterpretationAltered activity of the IL-6/IL-6R pathway could be a risk factor for depression. The field now requires experimental studies of IL-6 modulation in humans and animal models to further examine causality, mechanisms and treatment potential. Such studies are also needed to elucidate mechanisms for divergent associations of genetically-predicted higher IL-6 activity (risk increasing) and higher CRP concentrations (protective) with depression/anxiety.FundingThis research was funded in whole, or in part, by the Wellcome Trust (grant code: 201486/Z/16/Z). For the purpose of open access, the author has applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission. This work was supported by a Data Science Award from the MQ: Transforming Mental Health (grant code: MQDS17/40) to GMK and PBJ, which also supported ZY. GMK also acknowledges funding support from the Wellcome Trust (grant code: 201486/Z/16/Z), the Medical Research Council UK (grant code: MC_PC_17213 and MR/S037675/1), and the BMA Foundation (J Moulton grant 2019). NK and SM are supported by the International Max Planck Research School of Translational Psychiatry (IMPRS-TP). GDS works in the Medical Research Council Integrative Epidemiology Unit at the University of Bristol, which is supported by the Medical Research Council (MC_UU_00011/1).