ABSTRACT Background and Purpose Arterial spin labeling (ASL)‐MRI measures cerebral blood flow (CBF) by magnetically labeling red blood cells (RBC). ASL‐MRI shows reduced CBF in diseases including Alzheimer's Disease (AD). ASL‐measured CBF is affected by hematocrit, the proportion of blood volume composed of RBC. Hematocrit affects the ASL signal by affecting the T1 relaxation time of blood, and also physiologically, via rheological effects and autoregulation. This may confound comparisons between groups with differing hematocrit levels, such as men and women. We evaluated the influence of hematocrit on sex differences in ASL‐MRI‐measured CBF using two independent datasets. Methods 403 participants (252 women) from a local dataset and 119 participants (67 women) from ADNI underwent ASL‐MRI to measure cortical CBF and blood sampling for hematocrit. Using multiple regression, we assessed sex differences in CBF with and without accounting for hematocrit. Results Hematocrit was strongly and inversely correlated with CBF calculated using two different ASL‐MRI methods. Without adjusting for hematocrit, women appeared to have higher CBF than men, but this difference was no longer statistically significant when hematocrit was included in the model. Conclusions Hematocrit significantly affects ASL‐MRI‐measured CBF and must be considered when comparing groups with differing hematocrit. We demonstrate this for sex differences, and discuss its relevance to menopause, which is associated with increased hematocrit due to cessation of menstruation. Since female sex and post‐menopause are important AD risk factors, accounting for hematocrit is essential when using ASL‐MRI to investigate how sex and menopause influence brain perfusion, function, and vulnerability to AD.
Butler et al. (Thu,) studied this question.