Two methods for determining plasma IL-6 in humans at rest and following exercise

We compared two methods to determine IL-6 at rest and following exercise. On two occasions, plasma IL-6 was determined in 10 men and 13 women before and immediately following 90 min of cycling at 65% $$ \dot{V}{\text{O}}_{2\max } $$ using a cytometric bead array (CBA) and an ELISA. The CBA detected IL-6 in 63% of pre-exercise samples and 84% of post-exercise samples; the ELISA detected IL-6 in all samples. In samples with detectable IL-6, resting values were comparable between the CBA and ELISA (2.3 ± 1.0 and 2.4 ± 1.2 pg/mL, respectively; p = 0.99); the IL-6 response to exercise was underestimated (p = 0.0003) by the CBA by ~45%, compared with the ELISA. CBA- and ELISA-determined IL-6 values were moderately correlated (r = 0.66, p < 0.0001), but exhibited poor agreement (mean bias ± 1SD = −1.4 ± 2.1 pg/mL). The CBA provided concordant IL-6 concentrations at rest, when detected, but underestimated exercise-induced increases in IL-6 versus the ELISA method.