Total respiratory input impedance with the upper airway wall shunt minimized.

Respiratory impedances (Zrs) measured with the cheeks manually supported (Zsc) demonstrate an antiresonance (AR) at approximately 170 Hz, whereas those measured with the head generator (Zhg) have an AR at significantly lower frequencies (150 Hz). The differences in the AR between methods were previously attributed to upper airway wall (UAW) shunting in Zsc, and Zhg was believed to represent shunt-free impedance. To test this hypothesis, we measured four independent estimates of Zrs (4-256 Hz) in five healthy adults. Applying the oscillations at the mouth, we measured Zrs with the cheeks unsupported (Zuc) and when the subject's head and neck were enclosed in a rigid chamber completely filled with water (Zwa). We also measured Zhg with the oscillations applied around the head. Because water is incompressible, Zwa should minimize UAW shunt if not completely eliminate it. There were no significant differences in any of the AR features (e.g., location and bandwidth) of Zuc, Zsc, and Zwa. Conversely, all AR features of Zwa and Zhg were significantly different. To assess wall motion during both forms of oscillations (i.e., applied at the mouth and around the head), we measured two UAW local impedances by use of accelerometers placed on the cheeks (Zch) and submental (Zsm) region. Above 40 Hz, Zch and Zsm estimates were significantly dependent on the forcing method. Furthermore, Zch and Zsm made with the standard oscillations were consistent with the lumped UAW impedances measured with a head plethysmograph, whereas those made with the oscillations around the head were not. Therefore we conclude that, at > 32 Hz, Zsc more closely represents shunt-free impedance than does Zhg and hence should be preferred when Zrs are measured.