Multivariate modeling of cognitive-motor stimulation on neurovascular coupling: transcranial Doppler used to characterize myogenic and metabolic influences

Abstract

Neural activation induces changes in cerebral blood flow velocity (CBFV) with separate contributions from resistance-area product (V(RAP)) and critical closing pressure (V(CrCP)). We modeled the dependence of V(RAP) and V(CrCP) on arterial blood pressure (ABP), end-tidal CO(2) (EtCO(2)), and cognitive stimulation to test the hypothesis that V(RAP) reflects myogenic activity while V(CrCP) reflects metabolic pathways. In 14 healthy subjects, CBFV was measured with transcranial Doppler ultrasound, ABP with the Finapres device and EtCO(2) with infrared capnography. Two different paradigms (word or puzzle) were repeated 10 times (30 s on-off), and the corresponding square-wave signal was used, together with ABP and EtCO(2), as inputs to autoregressive-moving average (ARMA) models, which allowed identification of the separate contributions of the three inputs to either V(RAP) or V(CrCP). For both paradigms, the contribution of ABP was mainly manifested through V(RAP) (P < 0.005 for word; P < 0.004 for puzzle), while stimulation mainly contributed to V(CrCP) (P < 0.002 for word; P < 0.033, for puzzle). The contribution of EtCO(2) was relatively small (<10%) with greater contribution to V(CrCP) (P < 0.01 for puzzle; not significant for word). Separate step responses were also obtained for each of the three inputs. ARMA modeling of V(RAP) and V(CrCP) allows the separation of the effects of cerebral autoregulation and CO(2) reactivity from the main effects of cognitive-motor stimulation and have the potential to improve the diagnostic value of neurovascular coupling testing in physiological and clinical studies.