The Role of Inflammatory Cytokines as Intermediates in the Pathway from Increased Adiposity to Disease

Abstract

Objective: This study aimed to investigate the role of cytokines as intermediates in the pathway from increased adiposity to disease.

Methods: BMI and circulating levels of up to 41 cytokines were measured in individuals from three Finnish cohort studies (n = 8,293). Mendelian randomization (MR) was used to assess the impact of BMI on circulating cytokines and the impact of BMI-driven cytokines on risk of obesity-related diseases.

Results: Observationally, BMI was associated with 19 cytokines. For every SD increase in BMI, causal effect estimates were strongest for hepatocyte growth factor, monocyte chemotactic protein-1 (MCP-1), and tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and were as ratios of geometric means 1.13 (95% CI: 1.08-1.19), 1.08 (95% CI: 1.04-1.14), and 1.13 (95% CI: 1.04-1.21), respectively. TRAIL was associated with a small increase in the odds of coronary artery disease (odds ratio: 1.03; 95% CI: 1.00-1.06). There was inconsistent evidence for a protective role of MCP-1 against inflammatory bowel diseases.

Conclusions: Observational and MR estimates of the effect of BMI on cytokine levels were generally concordant. There was little evidence for an effect of raised levels of BMI-driven cytokines on disease. These findings illustrate the challenges of MR when applied in the context of molecular mediation.

Figure 1

Study overview. (1A) represents the analysis of the observational association between BMI and the inflammation‐related variables, whereas (1B) represents the one‐sample Mendelian randomization (MR) analysis of the effect of BMI on the inflammation‐related variables. (2) Represents the two‐sample MR analysis of the effect of BMI‐driven inflammation‐related variables on relevant disease outcomes. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 2

Observational (Observ.) and Mendelian randomization (MR) associations of BMI and inflammation‐related variables. Effect estimates are given in normalized SD units per 1‐SD‐higher BMI. Estimates represent meta‐analyzed results. (A) CRP result (YFS, FINRISK 1997, FINRISK 2002). (B) Results for cytokines available in all three cohorts (YFS, FINRISK 1997, FINRISK 2002). (C) Results for cytokines available only in two cohorts (YFS, FINRISK 2002).

Figure 3

Mendelian randomization associations of BMI‐driven inflammation‐related variables and disease outcomes. Estimates correspond to the odds ratio (OR) per unit increase in natural log‐transformed CRP or per normalized SD increase in HGF, MCP‐1, and TRAIL. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 4

SNP‐specific Mendelian randomization associations for MCP‐1 and inflammatory bowel disease (IBD), ulcerative colitis (UC), and Crohn disease (CD). (A) SNP‐specific associations between circulating MCP‐1 and odds of IBD. Estimates correspond to the odds ratio for IBD using either Immunochip data (N cases = 31,665, N controls = 33,977) or GWAS data (N cases = 12,882, N controls = 21,770) per normalized SD increase in circulating MCP‐1 (N = 8,337) levels. (B) SNP‐specific associations between circulating MCP‐1 and odds of UC. Estimates correspond to the odds ratio for UC using either Immunochip data (N cases = 13,768, N controls = 33,977) or GWAS data (N cases = 6,968, N controls = 20,464) per normalized SD increase in circulating MCP‐1 (N = 8,337) levels. (C) SNP‐specific associations between circulating MCP‐1 and odds of CD. Estimates correspond to the odds ratio for CD using either Immunochip data (N cases = 17,897, N controls = 33,977) or GWAS data (N cases = 5,956, N controls = 14,927) per normalized SD increase in circulating MCP‐1 (N = 8,337) levels.