Background
The characterization of physiological immune signatures in a population-based cohort is a prerequisite for identifying pathological immune signatures associated with inflammatory or autoimmune diseases.
Methods
Here, 47 plasma cytokines, chemokines, and growth factors were quantified with a bead-based multiplex-assay (Merck HCYTA-60 K) using a FLEXMAP 3D™ instrument in 1175 individuals of the Study of Health in Pomerania (SHIP; TREND cohort, 532 men and 643 women, age: 20 to 81, BMI: 17.7 to 53.6). Associations of cytokine concentrations with age, sex, BMI, season, and blood cell parameters (BCP) were examined by multivariate regression models.
Results
The physiological cytokine concentrations differed strongly between analytes, with median concentrations ranging from 0.6 to 7820 pg/mL. Many cytokine levels showed a large dynamic range within the study population. Higher levels of the pro-inflammatory cytokines and chemokines IL-6, IL-8, CXCL9, CXCL10, IL-12p40, CCL2, CCL4, CCL11, IL-27, FLT3LG, and TNFα were significantly associated with increasing age. The strongest age-associated effects were seen for CXCL9 (βst = 0.4, p < 0.001) and CXLC10 (βst = 0.3, p < 0.001). Significant sex differences were detected for CCL2, CCL3, CCL4, CCL11, CCL22, IL-12p40, IL-1RA, IL-18, IL-27, and TNFα levels among which CCL11 showed the strongest effect (βst = −0.24, p < 0.001) with a lower level in women compared to men. Moreover, seven cytokines and chemokines, i.e. CCL4, CCL22, CXCL10, IL-1RA, IL-18, IL-6, and TNFα, displayed higher levels with increasing BMI. Among those, the strongest effect was seen for IL-1RA (βst = 0.19, p < 0.001), CCL4 (βst = 0.16, p < 0.001) and CXCL10 (βst = 0.14, p < 0.001). Only CCL11 (βst = −0.17, p < 0.001) decreased with increasing BMI. Subjects categorized as obese exhibited significantly elevated levels of CCL4, CCL22, CXCL10, and IL-1RA, while only CCL11 showed significantly reduced levels compared to normal weight. Certain cytokines such as IL-6, IL-18, or TNFα showed decreased significance levels after adjustment for blood cell components indicating blood cell components (BCPs) as potential confounders. We observed no significant non-linear seasonal effects for the investigated cytokines.
Conclusion
The generated cytokine atlas provides detailed information on cytokine variations in the general population and will provide a reference base for disease-related studies in the future. Furthermore, BCPs should be considered as potential confounders in association studies based on plasma cytokine levels.