Systematic review and meta-analysis of the association between peripheral inflammatory cytokines and generalised anxiety disorder

Objective Inflammation has been implicated in the aetiology of mental illness. We conducted the first systematic review and meta-analysis of the association between peripheral markers of inflammation and generalised anxiety disorder (GAD). Design Systematic review and meta-analysis of studies measuring peripheral cytokine levels in people with GAD compared with controls. Data sources MEDLINE (1950-), EMBASE (1947-), PsycINFO (1872-) and Web of Science (1945-) databases up until January 2018. Eligibility criteria Primary, quantitative research studies of people with a diagnosis of GAD assessed using a standardised clinical interview that measured peripheral inflammatory markers. Data extraction and synthesis Two independent reviewers extracted data and assessed study quality. Meta-analysis using a random-effects model was conducted for individual cytokines where data from three or more studies were available. Results 14 of 1718 identified studies met the inclusion criteria, comprising 1188 patients with GAD and 10 623 controls. In total 16 cytokines were evaluated. Significantly raised levels of C reactive protein (CRP), interferon-gamma and tumour necrosis factor-alpha were reported in patients with GAD compared with controls in two or more studies. Ten further proinflammatory cytokines were reported to be significantly raised in GAD in at least one study. However, 5 of 14 studies found no difference in the levels of at least one cytokine. Only CRP studies reported sufficient data for meta-analysis. CRP was significantly higher in people with GAD compared with controls, with a small effect size (Cohens d=0.38, 0.06-0.69), comparable with that reported in schizophrenia. However, heterogeneity was high (I-2=75%), in keeping with meta-analyses of inflammation in other psychiatric conditions and reflecting differences in participant medication use, comorbid depression and cytokine sampling methodology. Conclusion There is preliminary evidence to suggest an inflammatory response in GAD, but it remains unclear whether inflammatory cytokines play a role in the aetiology. GAD remains a poorly studied area of neuroinflammation compared with other mental disorders, and further longitudinal studies are required.