Oxygen diffusing capacity estimates derived from measured VA/Q distributions in man

Data from eighteen subjects, studied in hypoxia (minimum PIO2 = 80 Torr) both at rest and during exercise, were analyzed using computer models which estimate O2 diffusing capacity from measured VA/Q distributions (obtained using the multiple inert gas elimination technique 'MIGET') and measured O2 exchange. Two of these models assigned the distribution of the diffusing capacity (D) in proportion to either the perfusion (DLO2-Qwt) or ventilation (DLO2-Vwt) distributions from MIGET, and thus modeled the effects of VA/Q and D/Q beta (where Q beta is the perfusive conductance) inequalities respectively. The third model (DLO2-3C) assigned all the diffusing capacity to a single homogeneous compartment. At rest DLO2 was 41.1 +/- 4.8, 41.1 +/- 5.4 and 30.2 +/- 2.1 ml X min-1 X Torr-1 for the Qwt, Vwt and 3C models respectively. These rose to 93.7 +/- 2.6, 109.3 +/- 4.5 and 81.1 +/- 1.9 ml X min-1 X Torr-1 respectively at maximal exercise, all significantly different from rest (P less than 0.001 for each). The effects of measured VA/Q and theoretical D/Q beta inhomogeneities on diffusing capacity estimates were significant even in normal lungs. Both types of inequality caused an appreciable underestimation of DLO2. These multi-compartment model estimates, using real data, are consistent with published theoretical predictions of the effects of V, Q and D inequalities. The results during exercise come close to morphometric predictions of maximal oxygen diffusing capacity in man.