Abstract DP271: Sex Differences Exist in Cerebrovascular Reactivity in Children

Background: Multiple studies have found a male prevalence in childhood ischemic stroke regardless of age. Causes for this sex difference are not fully understood. We hypothesized there may be potential differences in hemodynamic reserve, measured by cerebrovascular reactivity (CVR). Methods: Children ages 8-24 years with and without sickle cell anemia (SCA) underwent cerebrovascular reactivity (CVR) measurement using computer-controlled carbon dioxide manipulation during blood-oxygen level dependent (BOLD) MRI, along with MRI- measurements of cerebral blood flow (CBF; multi-post-label delay arterial spin labeling) and oxygen extraction (asymmetric spin echo). Cerebrovascular reactivity (CVR) was assessed voxel-wise within the gray matter (GM) in two ways: Magnitude CVR (mCVR) was defined as the slope of Δ %BOLD signal per Δ mmHg end-tidal CO 2 (ETCO 2 ). We then modeled the dynamic response as the convolution of ETCO 2 with an exponential hemodynamic response function (HRF), yielding two metrics: the time constant τ, representing the latency and speed of the vascular response, and the steady-state CVR (ssCVR), defined as the slope of BOLD signal versus the convolved ETCO 2 regressor, reflecting time-independent reactivity. Group comparisons used the Mann–Whitney U test for continuous variables and Fisher’s exact test for categorical variables, with Benjamini–Hochberg correction for multiple comparisons. A bidirectional stepwise linear regression modeled each CVR, allowing for entry/removal of all potential predictors, with candidate variables including age, sex, SCA status, hemoglobin, GM CBF, GM arterial transit time (ATT), GM OEF, GM cerebral metabolic rate of oxygen utilization (GM CMRO 2 ). Results: CVR data was available in 69 children median age 15 years [IQR 12-18], 46 females. The 17 SCA and 52 control participants were age (p=0.34) and sex (p=0.18) matched. Males had lower mCVR (p=0.009) and ssCVR (p=0.01) than females. On stepwise variable selection for multivariable modeling of mCVR, hemoglobin (p=0.006), sex (p=0.002) and GM arterial transit time (p=0.02) remained significant. In ssCVR, only hemoglobin (p