A meta-analysis of differences in IL-6 and IL-10 between people with and without depression: Exploring the causes of heterogeneity

Abstract

Epidemiological evidence for the inflammatory hypothesis of depression is largely cross-sectional; people with depression have elevated levels of circulating pro-inflammatory markers compared to people without depression. The limitation of cross sectional research is the potential for extraneous factors to influence observed effects. The purpose of this meta-analysis of cross-sectional studies of interleukin(IL)-6 and IL-10 in people with and without depression is to provide a targeted analysis of potential moderator factors relating to the diagnosis of depression and to physical and psychiatric comorbidity. Electronic searches of Embase and Medline databases were conducted using subject headings “interleukin-6” or “interleukin-10” and those relating to depression. Studies were included if they measured circulating marker levels in serum or plasma in a group of people with and without depressive symptoms (99 studies for IL-6, 19 studies for IL-10). IL-6 was elevated in depressed compared to non-depressed groups (d = 0.46, 99% CI 0.34 to 0.58, I² = 85.9%). This effect was larger in subgroups where depressive disorders were diagnosed compared to those with only depressive symptoms via standardized inventory, and subgroups where participants were recruited from inpatient or outpatient settings compared to the general community. The effect was also larger in those who were not selected for a particular comorbidity compared to those selected for cardiovascular disease. IL-10 effect size was not significant (d = −0.31, 99% CI −0.95 to 0.32, I² = 94.1%) which was not accounted for in subgroup analyses or meta-regression, indicating there is not a global elevation in cytokines. These data highlight that comorbidity and behavioral aspects of depression need to be measured and controlled in future prospective and experimental research testing the inflammatory hypothesis of depression.

Introduction

Depressive disorders are common and contribute substantially to disease burden (Kessler et al., 2003, Üstün et al., 2004). The key to understanding their origins may be inflammation. The theory that inflammation causes depression relies on the idea that cytokines exert central and peripheral effects which cause the psychological and physiological experience of depression (Miller et al., 2009). Inflammation is thought to cause clinical depressive disorders by inducing sickness behaviors which are akin to neurovegetative symptoms of depression (e.g., lethargy, changed appetite, changed sleep), activating the hypothalamic–pituitary–adrenal axis and sympathetic nervous system, altering serotonin and dopamine synthesis and reuptake and causing neurodegenerative processes which lead to the phenomenology of depression (for reviews, see Dantzer et al., 2008, Irwin and Miller, 2007, Maes et al., 2009, Miller et al., 2009). Likely sources of depression-inducing cytokines include internal stressors (e.g., organic disease, adipose tissue) and external stressors (e.g., psychological distress, diet).

Animal models and some experimental human studies support the causal nature of this theory, where administration of cytokines or immune stimulants cause depression-like behavior and symptoms, and also evidence that medical illness and inflammation-based treatments [i.e., interleukin(IL)-2 or interferon alpha treatments for hepatitis C and cancer] are linked to incidence of depressed mood or depressive disorders (Capuron et al., 2009, Dantzer et al., 2008). However, the bulk of human literature is cross sectional; people with depression or depressive symptoms have elevated levels of circulating inflammatory markers compared to those without (Dowlati et al., 2010, Howren et al., 2009). It is impossible to infer causality in this context, but these studies do permit examination of the strength of association between inflammatory markers and depressive symptoms/disorder.

Several recent meta-analyses have verified that people with depression show elevated levels of the cytokine IL-6 compared to people without depression in circulating serum or plasma (Dowlati et al., 2010, Howren et al., 2009, Liu et al., 2012). IL-6 is a pleiotropic cytokine associated with inflammation, both acute and chronic, and also produced by adipose cells. Elevations in IL-6 have been associated with increases in several “trait” factors such as advancing age, high body mass index (BMI), smoking and physical comorbidity as well as “state” factors such as fat consumption, acute physical activity, and psychological stress (see O’Connor et al., 2009 for review). As people with depressive disorders frequently exhibit these lifestyle factors, there is a possibility that one or more of these is driving the observed elevations in circulating levels of inflammatory markers. This is supported in prospective studies which show that once BMI, smoking and physical activity are controlled for, the relationship between depression and IL-6 is no longer significant (Duivis et al., 2011). It is important to assess the contribution of lifestyle factors to the observed associations between depression and IL-6 to more accurately define the role of peripheral, circulating inflammatory markers as markers of an inflammatory depressive state. A second interleukin of potential interest is IL-10, a cytokine with anti-inflammatory effects which is in part stimulated by IL-6 among other factors. IL-10 is also elevated in response to acute immune challenge (Henry et al., 2009) and may be implicated in depression (Mesquita et al., 2008, Roque et al., 2009).

The current meta-analysis compares IL-6 and IL-10 in people with and without depression and is the most extensive to date. It includes a broad range of studies with different degrees of “depression” (diagnosed disorder vs. depressive symptoms), different recruitment sources and instances where depressed and control groups are matched for particular physical disease. The aim was to pool the evidence of a cross sectional association between depression and IL-6 and IL-10 in light of moderating factors to investigate whether the elevations in inflammatory markers are attributable to the physical and psychiatric comorbidities of depressive disorders.