Effects of cerebral ischemia on human neurovascular coupling, CO₂ reactivity, and dynamic cerebral autoregulation

<jats:p> Cerebral blood flow (CBF) regulation can be impaired in acute ischemic stroke but the combined effects of dynamic cerebral autoregulation (CA), CO<jats:sub>2</jats:sub> cerebrovascular reactivity (CVR), and neurovascular coupling (NVC), obtained from simultaneous measurements, have not been described. CBF velocity in the middle cerebral artery (MCA) (CBFv, transcranial Doppler), blood pressure (BP, Finometer), and end-tidal Pco<jats:sub>2</jats:sub> (Pet<jats:sub>CO<jats:sub>2</jats:sub></jats:sub>, infrared capnography) were recorded during a 1-min passive movement of the arm in 27 healthy controls [mean age (SD) 61.4 (6.0) yr] and 27 acute stroke patients [age 63 (11.7) yr]. A multivariate autoregressive-moving average model was used to separate the contributions of BP, arterial Pco<jats:sub>2</jats:sub> (Pa<jats:sub>CO<jats:sub>2</jats:sub></jats:sub>), and the neural activation to the CBFv responses. CBFv step responses for the BP, CO<jats:sub>2</jats:sub>, and stimulus inputs were also obtained. The contribution of the stimulus to the CBFv response was highly significant for the difference between the affected side [area under the curve (AUC) 104.5 (4.5)%] and controls [AUC 106.9 (4.3)%; P = 0.008]. CBFv step responses to CO<jats:sub>2</jats:sub> [affected hemisphere 0.39 (0.7), unaffected 0.55 (0.8), controls 1.39 (0.9)%/mmHg; P = 0.01, affected vs. controls; P = 0.025, unaffected vs. controls] and motor stimulus inputs [affected hemisphere 0.20 (0.1), unaffected 0.22 (0.2), controls 0.37 (0.2) arbitrary units; P = 0.009, affected vs. controls; P = 0.02, unaffected vs. controls] were reduced in the stroke group compared with controls. The CBFv step responses to the BP input at baseline and during the paradigm were not different between groups ( P = 0.07), but Pet<jats:sub>CO<jats:sub>2</jats:sub></jats:sub> was lower in the stroke group ( P < 0.05). These results provide new insights into the interaction of CA, CVR, and NVC in both health and disease states. </jats:p>