Unhealthy lifestyle may increase later depression via inflammation in older women but not men

Highlights

• We note a gender difference in depression-inflammation relationships.

• Baseline IL-6 predicts future depressive symptoms in women.

• IL-6 mediates the prospective relationship between adiposity and depression in women.

• Unhealthy lifestyle may be a source of inflammatory markers in depressed women.

Abstract

Depression and inflammatory markers have a reliable cross-sectional association although less is known about the prospective relationship. The current study investigated whether pro-inflammatory markers are prospectively associated with depression, and whether indicators of unhealthy lifestyle, physical health and psychosocial functioning may drive this association. Participants were drawn from the Hunter Community Study, a community-dwelling cohort of individuals aged 55–85 years (N = 1410). Participants completed baseline physiological assessment, health-related questionnaires, and blood sampling for the analysis of inflammatory markers, C-reactive protein (CRP) and interleukin (IL)-6. Participants completed the same depressive symptom questionnaire again after 3.5–5.5 years. Depression outcomes at follow-up were analysed dichotomously using established scale cut-off scores and continuously as a “residual score”, representing the variation in follow-up depressive symptoms not explained by baseline symptoms and age. Analyses were conducted on males and females separately. At baseline, indicators of unhealthy lifestyle, physical health and psychosocial functioning were associated with depressive symptoms and inflammatory markers. For males, there were no relationships between inflammatory markers and follow-up depression outcomes. In females, IL-6 was significantly associated with depression outcomes in univariate, but not multivariate analyses. However, IL-6 significantly mediated the association between the predictors of waist-to-hip ratio, smoking and psychological coping at baseline, and follow-up depression outcomes. The results support the inflammatory hypothesis of depression, although females may be more vulnerable to effects. The findings raise the possibility that unhealthy lifestyle and psychosocial stress may drive inflammation and subsequent depressive symptoms.

Introduction

The inflammatory hypothesis of depression posits that inflammation may have a causative role in depression. It is supported by observations of depressive-like behaviour following cytokine administration in animals and humans, and idiopathic major depressive disorder in patients treated with cytokines such as interferon-alpha or interleukin (IL)-2 (Anisman et al., 2005, Capuron et al., 2009, Dantzer et al., 2011, Miller et al., 2009, Myint et al., 2009, Reichenberg et al., 2001). Furthermore, inflammatory mediators, including IL-6, C-reactive protein (CRP), and tumor necrosis factor are consistently elevated in depression (Dowlati et al., 2010, Hiles et al., 2012, Howren et al., 2009). Emerging evidence from randomised controlled trials suggests that anti-inflammatory medications may improve depression outcomes (Akhondzadeh et al., 2009, Raison et al., 2013). Inflammatory mediators interact with key biological systems implicated in depression, including altering neuroendocrine stress activity, neural plasticity, cognitive functioning, reactive oxygen species, and neurotransmitter metabolism and activity (Irwin and Miller, 2007, Miller et al., 2009). Thus, a causal relationship is biologically plausible.

The source of the elevated inflammatory markers in depression remains unclear. Recent theories, such as the social signal transduction theory (Slavich and Irwin, 2014) and PATHOS-D (Raison and Miller, 2013), propose that real or imagined psychosocial stressors, represented cortically, activate autonomic and hormonal inflammatory pathways and upregulate inflammatory gene expression. This upregulation produces the elevated circulating inflammatory mediators that cause cognitive, emotional and behavioural symptoms of depression. However, the source of inflammation in people with depression may be broader than this, involving factors such as nascent or apparent physical illness, including obesity, and/or aspects of lifestyle (Berk et al., 2013). Aspects of physical illness and unhealthy lifestyle, including central adiposity, low physical activity, poor diet quality, smoking and alcohol use, are frequently observed in people with depression and also have inflammatory consequences (Hamer et al., 2009a, Lopresti et al., 2013, Milaneschi et al., 2009, O'Connor et al., 2009). For instance, adipose tissue, particularly hypertrophic abdominal fat, produces inflammatory cytokines and mediators (Bastard et al., 2006, Maury and Brichard, 2010, Odegaard and Chawla, 2013) and it may be this abdominal, and not subcutaneous, fat that is associated with depression (Everson-Rose et al., 2009).

Little attention has been given to examining potential sources of inflammation in depression within longitudinal contexts. Indeed, few published studies address longitudinal evidence of whether elevations in inflammatory markers precede or follow depressive symptoms, and the evidence that is published is mixed. Meta-analysis of longitudinal studies indicate a small significant association between elevated CRP (eight studies) or IL-6 (three studies) and subsequent depressive symptoms, with moderate heterogeneity (Valkanova et al., 2013). There is also support for a bi-directional prospective relationship between inflammatory markers and depressive symptoms (Hamer et al., 2009a, Hamer et al., 2009b, Matthews et al., 2010). Given the limited and mixed evidence, further exploration of the prospective relationship is warranted, with close consideration of the influence of effect modifiers. For instance, although previous prospective studies have selectively examined women (Matthews et al., 2007, Matthews et al., 2010) or men (Boyle et al., 2007), typically gender is considered as a control variable, rather than an effect modifier. There are well-established differences in the clinical presentation of depression in men and women (Marcus et al., 2005), likely due to both social factors and biological factors, including inflammatory markers and neuroendocrine stress hormones (Edwards et al., 2006, Kudielka and Kirschbaum, 2005, Larsson et al., 2009, Marriott and Huet-Hudson, 2006, McConnell et al., 2005). Therefore, examining the prospective relationship between depression and inflammatory markers by gender is pertinent.

To our knowledge, mediation analyses have not been completed to examine whether inflammatory markers mediate the relationship between baseline health and lifestyle factors, and later depression. This approach may highlight whether physical health and lifestyle could be a source of elevated inflammatory markers observed in people with depression. The current study explores the relationship between inflammatory markers, depressive symptoms and indicators of psychosocial functioning, physical health, and unhealthy lifestyle (central adiposity, low physical activity, poor diet quality, smoking and alcohol use). There are two discrete aims. The first aim is to explore a practical question from a biomarker perspective: whether baseline levels of inflammatory markers – IL-6 and CRP – are associated with levels of depressive symptoms at follow-up, and whether the effects remain after adjusting for confounding. The second aim is to examine lifestyle, physical health or psychosocial functioning as predictors of depressive symptom outcomes at follow-up, and explore whether inflammatory markers mediate this relationship, thereby providing evidence regarding potential sources of inflammatory markers in depression.