Depression is the leading cause of mental health-related global disease burden 1, 2. Persistent cognitive problems, such as poor memory and concentration, are reported in 11% of adults aged ≥ 45 years 3 and are frequently observed across physical [cancer (35%); COVID-19 (22%); HIV (43%); hepatitis C (50%)] 4–7 and mental health conditions [depression (30%); schizophrenia (50%)] 8, 9. Existing treatments for depression are only modestly effective 10 and almost inexistent for cognitive dysfunction 11. A mechanistic understanding of depression and cognitive dysfunction is urgently needed to inform the development of effective treatments and prevention approaches.
Chronic, low-grade systemic inflammation may represent one such mechanism. Indices of inflammation [e.g., circulating levels of cytokines (e.g., interleukin-6 (IL-6) and acute phase proteins (e.g., C-reactive protein (CRP)] are elevated in individuals with depression compared to controls 12 and inflammatory biomarkers have been linked to specific aspects of depression, such as anhedonia and negative affect 13, 14. Further, longitudinal observational studies have found that higher levels of inflammatory biomarkers (e.g., IL-6, CRP) are prospectively associated with higher depressive symptoms 15. Observational studies have linked inflammation with impaired cognition in population-based 16–19 and in physical 20–22 and mental health conditions 23–26. Inflammation also impacts neural circuitry relevant to affective disorder and cognitive task performance 27, 28, particularly the hippocampus 29 and striatum 30–32. To date, inflammation-cognition research has primarily relied upon observational data.
Inferring causality from observational studies is a challenge due to confounding (e.g., stress, poor sleep 22) and reverse causality (i.e., whether inflammation impacts depression/cognition, or vice versa). Mendelian randomization (MR) is a genetic epidemiological method that can test causal relationships by using genetic variants associated with an exposure (e.g., inflammation) as proxies for the exposure 33. As genetic variants are randomly inherited from parents to offspring and are fixed at conception, they are less likely to be associated with confounders and overcome issues of reverse causation 33. Preliminary evidence, using MR, implicate IL-6 and its soluble IL-6 receptor (sIL-6R) in depression 34–37. To date, most MR studies examining the effect of IL-6 on health have focused on circulating IL-6 levels. However, IL-6 signals via multiple pathways (trans-signaling, classical-signaling, and trans-presentation) and there is growing evidence that IL-6 trans-signaling is primarily responsible for the pathogenic inflammatory effects of IL-6 38. Here, we include variants related to circulating IL-6 levels, and sIL-6R levels (relevant for IL-6 trans-signaling). Causal evidence for CRP and other proinflammatory markers [i.e., Glycoprotein Acetyls (GlycA) a composite biomarker thought to provide a more stable marker of inflammation which reflects the glycosylation of multiple acute-phase proteins 39] on depression are mixed 34, 37, 40–43. Regarding cognition, few studies have examined potential causal relationships with inflammation. MR analyses using available genome-wide association studies (GWAS) report both null results of inflammatory biomarkers on emotion recognition, working memory, response inhibition 44 as well as effects of specific cytokines/chemokines (i.e., Eotaxin, IL-8) on fluid intelligence 45.
The current study used data from the Lifelines Cohort Study – a large population-based cohort in the Netherlands – to conduct complementary non-genetic and genetic analysis to investigative the causal relationship between inflammation and negative affect, depressive disorders, and cognitive task performance. First, we used cross-sectional and longitudinal non-genetic analysis examine the association between circulating levels of CRP and depression/cognitive performance. Second, we conducted genetic risk score (GRS) and MR analysis to test whether genetic variants regulating levels and activity of CRP, IL-6, and GlycA were causally related with depression/cognitive performance. We also conducted the above analyses on closely related constructs (e.g., anxiety, negative/positive affect), for which associations with inflammation have previously been observed 46–51 but for which considerably less empirical data has been published. We hypothesized that both circulating CRP levels and genetically predicted inflammatory biomarkers (i.e., CRP, IL-6, sIL-6R, and GlycA) would be associated with depression, cognitive task performance, affect, and anxiety.