Effects of Gray Matter Normalization on Cerebrovascular Reactivity in Middle‐Aged Adults at Elevated Risk of Alzheimer's Disease

Abstract

Cerebrovascular reactivity is a measure of cerebrovascular health and attenuated cerebrovascular reactivity is associated with Alzheimer's disease (AD). Gray matter (GM) atrophy occurs with advancing age and is greater in individuals with AD compared to age-matched controls. Normalizing cerebrovascular reactivity to GM to account for individual differences in GM volume may provide a more accurate measure of cerebrovascular reactivity. Therefore, the purpose of this study was to investigate the effects of GM normalization on cerebrovascular reactivity in middle-aged adults with and without a family history of AD. Seventy-five adults with a family history of AD (FH, n = 50, men = 17, women = 33, age = 62 ± 4 yrs) and without a family history of AD (NoFH, n = 25, men = 7, women = 18, age = 63 ± 4 yrs) participated in this study. Family history of AD was determined using a validated Dementia Questionnaire or autopsy report when available. To determine brain volumes, participants underwent a T1-weighted scan on a 3T MRI scanner. GM was segmented in SPM12 and normalized to intracranial volume. Cerebrovascular reactivity to hypercapnia was measured by continuously recording middle cerebral artery velocity (MCAv) with a transcranial Doppler ultrasound, mean arterial pressure (MAP) with a finometer, and end-tidal carbon dioxide (ETCO2) with a nasal cannula during stepwise elevations of CO2. Cerebrovascular conductance index (CVCi) was calculated using the equation (MCAv/MAP)*100. Cerebrovascular reactivity was calculated as the linear relationship between the change in ETCO2 and the change in MCAv or CVCi. GM normalization was performed by dividing reactivity slopes by GM volume. There were no differences in MCAv reactivity (FH: 2.1 ± 0.1 cm/s/mmHg vs. NoFH: 1.9 ± 0.2 cm/s/mmHg; p > 0.05) between groups; however, there was a trend for greater CVCi reactivity in adults with a family history of AD (FH: 1.5 ± 0.1 cm/s/mmHg2 vs. NoFH: 1.2 ± 0.1 cm/s/mmHg2; p = 0.09). There were no differences in GM volume (FH: 0.47 ± 0.03 l vs. NoFH: 0.46 ± 0.03 l; p > 0.05) between groups. After normalizing for GM, there were no differences in normalized MCAv reactivity (FH: 4.4 ± 0.2 cm/s/mmHg/l vs. NoFH: 4.0 ± 0.3 cm/s/mmHg/l; p > 0.05) between groups. The trend for greater CVCi reactivity in adults with a family history of AD persisted after normalizing for GM (FH: 3.2 ± 0.2 cm/s/mmHg2/l vs. NoFH: 2.5 ± 0.3 cm/s/mmHg2/l; p = 0.09). Normalized MCAv and CVCi reactivity were greater than MCAv and CVCi reactivity (p < 0.05). These findings suggest that GM normalization yields greater cerebrovascular reactivity slopes compared to raw slopes; however, GM normalization did not impact the effect of a family history of AD on cerebrovascular reactivity. Normalization of cerebrovascular reactivity values during accelerating GM atrophy with advancing age and AD progression may be important when evaluating cerebrovascular health.