The systemic inflammatory response to exercise has been extensively investigated, both in health and disease. The most pronounced marker of this response is IL-6 (Febbraio & Pedersen, 2002). Compared with healthy individuals, the IL-6 response to exercise is higher in patients with COPD, particularly after exercise of submaximal intensity (van Helvoort et al., 2006b). In keeping with previous data, the current study shows that exercise at 40 W engenders a significant rise of plasma IL-6 levels in COPD patients with low fat-free mass. Because investigations in healthy subjects made clear that contracting skeletal muscle is the main source of circulating IL-6 in response to exercise (Pedersen & Hoffman-Goetz, 2000), IL-6 is frequently referred to as a myokine. Interestingly, the amount of IL-6 released from muscles seems to depend on exercise intensity (Helge et al., 2003). However, in that respect, it seems paradoxical, that COPD patients show a higher IL-6 response at much lower workloads compared with healthy subjects (van Helvoort et al., 2006b). Findings from the current study provide a potential explanation for this apparent contradiction. The load on the peripheral muscles was identical between the two exercise side modes, i.e., 40 W, while the IL-6 response to exercise was almost completely abolished after exercise with NIV support. So the elevated IL-6 plasma levels upon exercise in COPD patients are not related to the load on the peripheral muscles. However, although we did not measure the load on the respiratory muscles in this pilot study, we expect to have unloaded the respiratory muscles by applying NIV. Accordingly, these data suggest that the main cause of an exaggerated IL-6-response in COPD patients during exercise is not the load on the peripheral muscles, but the load on the respiratory muscles. This postulation is supported by several previous observations. For example, Vassilakopoulos et al. demonstrated that increasing the load on the inspiratory muscles in healthy individuals leads to elevated IL-6 concentrations in plasma (Vassilakopoulos et al., 1999). Moreover, the load on the respiratory muscles during exercise is excessively increased in COPD patients compared with healthy subjects (Evison & Cherniack, 1968). Furthermore, Mercken and coworkers showed that localized leg muscle exercise of COPD patients did not result in elevated plasma IL-6 levels (Mercken et al., 2009b), suggesting that the respiratory muscles rather than the peripheral muscles produce IL-6 during whole body exercise. Indeed, in animals, it has been demonstrated that cytokine expression in the diaphragm is upregulated in response to loaded breathing (Sigala et al., 2011). An alternative explanation might be that the peripheral muscles start to produce IL-6 because of local vasoconstriction induced by increased load on the respiratory muscles, also known as the respiratory muscle metaboreflex (Harms et al., 1997; Dempsey et al., 2006). So, the present findings and previous studies provide evidence that increased loading of the respiratory muscles induces the supranormal IL-6 response to exercise observed in COPD patients. The exact origin of IL-6, either the peripheral or respiratory muscles, could be an interesting topic for future studies following up on the present pilot study.
In contrast to IL-6, NIV support could not prevent the rise of leukocytes and neutrophils in response to exercise. The extra amounts of circulating leukocytes that appear upon exercise probably arise from an increased blood flow through lymphoid tissue, the non-circulatory leukocyte pool. Increased heart rate and vasomotor tone as a response to stress hormones during exercise are thought to lead to the release of extra leukocytes into the circulation (Hoffman-Goetz & Pedersen, 1994). In the current study, we did not measure the effect of NIV on, for example, catecholamines, but we did show that NIV had no effect on the heart rate response to exercise. This could explain why the increase of leukocyte concentrations was not affected by NIV support. Although the main objective of the current study was to investigate the effect of NIV on markers of systemic inflammation, we additionally examined some markers of oxidative stress because oxidative stress and systemic inflammation have often been linked (Vassilakopoulos et al., 2003). For example, it is well known that leukocytes and in particular neutrophils are able to produce ROS under conditions of stress (Suzuki et al., 1996). As ROS modify the structure and function of proteins, fatty acids, and genetic material, their presence in the circulation has been proposed to be harmful. In our study, we found no effect of control exercise on the capacity of blood samples to produce ROS. In line with that, the plasma levels of carbonylated proteins, a footprint of oxidative protein damage, were not elevated upon control exercise. This is in contrast with previous studies from our lab, which did show an oxidative stress response to exercise in muscle-wasted COPD patients (van Helvoort et al., 2006a; van Helvoort et al. 2007). The discrepancy between the current study and those previous studies can probably be explained by less loss of fat-free mass in the patients that were currently studied, because indeed we previously demonstrated that the magnitude of the oxidative stress response is closely related to the severity of fat-free mass loss (van Helvoort et al., 2007). A causative role for oxidative stress in inducing the IL-6 response to exercise in COPD patients has previously been proposed (van Helvoort et al., 2006a; Jammes et al., 2008). The present study does not support that concept. First, IL-6 levels elevated upon control exercise, without increased presence of carbonylated proteins. Moreover, the abolished IL-6 response to exercise with NIV support was accompanied by lower arterial oxygen pressures and increased capacity of blood cells to produce ROS. So, the release of IL-6 upon exercise can be inhibited despite plasma states that favor oxidative stress. The present findings therefore do not support the concept that IL-6 response to exercise is caused by increased systemic oxidative stress but rather point to an important role for excessive loading of the respiratory muscles. Of note, we do not exclude the involvement of oxidative stress signaling within the respiratory muscles. Vassilakopoulos et al. have previously shown that increased loading of the respiratory muscles induces oxidative stress and administration of anti-oxidants can downregulate the expression and release of cytokines by the diaphragm (Vassilakopoulos et al., 2002).