Arterial partial pressure of CO
2 (PaCO
2) has a strong influence on CBF (Willie et al.
2014). In the present study however, we measured but did not clamp ETCO
2. In many studies, ETCO
2 is used as a proxy for PaCO
2 as the latter requires arterial catheterization and cannot be measured continuously. ETCO
2 is typically reduced with LBNP-induced central hypovolemia, which is often interpreted to reflect a reduction in PaCO
2. Changes in ETCO
2 are, however, associated with changes in cardiac output (Joseph et al.
2003; Lakhal et al.
2017) (Appendix, Fig. C). When we entered ETCO
2 as an explanatory variable in the multivariable analysis with MAP, cardiac output and MCAV, the effect of cardiac output was reduced (Appendix, Regression 7). This, along with the close association between cardiac output and ETCO
2 (Appendix, Fig C) indicates statistical multicollinearity, and that ETCO
2 is redundant in the model when cardiac output is already entered. This further indicates that, in our model, changes in ETCO
2 exert little independent effect when changes in cardiac output have been accounted for.
At least three mechanisms could explain a reduction in ETCO
2 during hypovolemia: (1) a reduction in PaCO
2 due to hyperventilation and/or reduction in peripheral CO
2-production, (2) reduced PaCO
2 due to sequestering of CO
2 in peripheral tissues (decreased wash-out) (Garnett et al.
1989), or (3) reduced pulmonary blood flow, increased dead-space ventilation and increased alveolo-arterial CO
2-difference. In the third case, a reduction in ETCO
2 does not need to be associated with a reduction in PaCO
2. Several studies indicate that this mechanism partly explains the reduction in ETCO
2 with experimental (Dubin et al.
2000) and clinical hypovolemia (Campion et al.
2019; Tyburski et al.
2003), questioning the assumption that ETCO
2 can substitute PaCO
2. This finding is supported by studies applying orthostatic stress, where both LBNP (Zhang and Levine
2007) and head-up-tilt (Immink et al.
2006; Serrador et al.
2006) induced greater reductions in ETCO
2 than PaCO
2. Also, while controlling for the effect of PaCO
2, they found an effect of cardiac output on CBF. A recent study reported stable PaCO
2 in the LBNP model (van Helmond et al.
2018). As in other studies, we found a reduction in ETCO
2 with increasing hypovolemia. To explore possible mechanisms behind this reduction, we plotted ETCO
2 and cardiac output over time at release of LBNP (Appendix, Fig. N). The figure shows that both cardiac output and ETCO
2 increase within 5 s, which seems hard to reconcile with mechanism (1), as PaCO
2-production would increase much more slowly, and also mechanism (2), as even the wash-out and transport of CO
2 from peripheral tissues to the lungs would be expected to take more time than what was observed. Hence, the reduction in ETCO
2 in our study seems to be, at least in part, caused by increased alveolar dead-space ventilation. The relationship between ETCO
2 and PaCO
2 is, however, complicated, and may differ between acute changes and steady-state conditions (Isserles and Breen
1991). It is therefore difficult to state if or to what extent PaCO
2 is changed in the present study and it would also question the value of clamping ETCO
2. Further studies on the relationship between ETCO
2 and PaCO
2 in the LBNP model are warranted.