Modeling oxygenation in venous blood and skeletal muscle in response to exercise using near-infrared spectroscopy

Noninvasive, continuous measurements in vivo are commonly used to make inferences about mechanisms controlling internal and external respiration during exercise. In particular, the dynamic response of muscle oxygenation (Sm(O(2))) measured by near-infrared spectroscopy (NIRS) is assumed to be correlated to that of venous oxygen saturation (Sv(O(2))) measured invasively. However, there are situations where the dynamics of Sm(O(2)) and Sv(O(2)) do not follow the same pattern. A quantitative analysis of venous and muscle oxygenation dynamics during exercise is necessary to explain the links between different patterns observed experimentally. For this purpose, a mathematical model of oxygen transport and utilization that accounts for the relative contribution of hemoglobin (Hb) and myoglobin (Mb) to the NIRS signal was developed. This model includes changes in microvascular composition within skeletal muscle during exercise and integrates experimental data in a consistent and mechanistic manner. Three subjects (age 25.6 +/- 0.6 yr) performed square-wave moderate exercise on a cycle ergometer under normoxic and hypoxic conditions while muscle oxygenation (C(oxy)) and deoxygenation (C(deoxy)) were measured by NIRS. Under normoxia, the oxygenated Hb/Mb concentration (C(oxy)) drops rapidly at the onset of exercise and then increases monotonically. Under hypoxia, C(oxy) decreases exponentially to a steady state within approximately 2 min. In contrast, model simulations of venous oxygen concentration show an exponential decrease under both conditions due to the imbalance between oxygen delivery and consumption at the onset of exercise. Also, model simulations that distinguish the dynamic responses of oxy-and deoxygenated Hb (HbO(2), HHb) and Mb (MbO(2), HMb) concentrations (C(oxy) = HbO(2) + MbO(2); C(deoxy) = HHb + HMb) show that Hb and Mb contributions to the NIRS signal are comparable. Analysis of NIRS signal components during exercise with a mechanistic model of oxygen transport and metabolism indicates that changes in oxygenated Hb and Mb are responsible for different patterns of Sm(O(2)) and Sv(O(2)) dynamics observed under normoxia and hypoxia.