Moderate correlation between breath-holding and CO2 inhalation/hyperventilation methods for transcranial doppler evaluation of cerebral vasoreactivity

Background. Both CO2 inhalation followed by hyperventilation and breath-holding have been utilized to measure cerebral vasomotor reactivity (VMR) but their correlation has been poorly studied and understood. Methods. A retrospective study was conducted in 143 subjects (62.6 +/- 15.8 years old, 64% men) with transcranial Doppler ultrasonography measurement of mean flow velocity (MFV) at baseline, after 30 seconds of breath-holding, and after CO2 inhalation followed by hyperventilation, in the left and right middle cerebral artery. Breath-holding index (BHI) was calculated as the percentage of MFV increase from baseline per second of apnea. CO2 inhalation/hyperventilation index (CO2/HV) was calculated as the percentage of MFV difference between CO2 inhalation and hyperventilation. Results. There were 75 carotid arteries with >70% stenosis or occlusion, and 18 middle cerebral arteries with >50% stenosis or occlusion. The mean BHI was 0.93 +/- 0.7 and 0.89 +/- 0.6, whereas the mean CO2/HV was 61 +/- 26% and 60 +/- 26%, respectively, on the right and left sides. The correlation between BHI and CO2/HV was moderate on the right (r = 0.33; p < 0.01) and left sides (r = 0.38; p < 0.01). Multivariate linear regression analysis indicated that age (p = 0.01) and history of stroke (p = 0.007) were associated independently with an impaired VMR on the right as measured by CO2/HV. No predictors for impaired VMR by CO2/HV on the left and by BHI on either side were found. Conclusions. CO2/HV and BHI are only moderately correlated. Further studies are necessary to determine which method more accurately predicts clinical morbidity. (c) 2012 Wiley Periodicals, Inc. J Clin Ultrasound 2012; Published online in Wiley Online Library