The Trouble Interpreting fMRI Studies in Populations with Cerebrovascular Risk: The Use of a Subject-Specific Hemodynamic Response Function in a Study of Age, Vascular Risk, and Memory

Abstract

Functional magnetic resonance imaging (fMRI) is commonly used to investigate the neural bases of behavior ranging from basic cognitive mechanisms to aging to psychological disorders.However, the BOLD signal captured by fMRI is an indirect measure of neural function and is affected by many factors that are non-neural in origin. These non-neural factors, however, do affect brain vasculature such as the shape and timing of the hemodynamic response function (HRF) during task-evoked fMRI that, in turn, can cause inappropriate and/or misleading interpretations of fMRI differences between groups. In this study, we tested the proposition that vascular health risks, which often go unmeasured in neuroimaging studies, and aging interact to modify the shape and/or timing of the HRF (height, time-to-peak, width), which then affect the differences in patterns of brain activity in a task-evoked memory encoding paradigm. Adult participants (aged 20-74) answered questions about their health history and underwent two fMRI tasks: viewing of a flashing checkerboard using a slow event-related design and a paired associates memory encoding task during a fast event-related design. We found that aging and vascular risk had the largest impacts on the maximum peak value of the HRF. Using a subjectspecific HRF resulted in an overall dampening of the estimated brain activity in both taskpositive and task-negative regions due to a reduction in the inter-individual variance of that activity. Across three vascular risk factors, using a subject-specific HRF resulted in more consistent brain regions that reached significance and larger effect sizes compared with the canonical HRF. A slight advantage in the reliability of brain-behavior correlations also was found. The findings from this study have far reaching consequences for the interpretation of taskevoked fMRI activity, especially in populations known to experience alterations to brain vasculature including adults of all ages that have higher vascular risk, the majority of older AGE AND VASCULAR RISK EFFECTS ON HRF 3 adults, and people with neurocognitive disorders in which vasculature differences may play a role including dementia. Highlights• Older age was associated with smaller maximum peak of the hemodynamic response.• Younger and middle-aged adults with more vascular risk had higher HRF peaks.• Using a subject-specific HRF resulted in a "dampening" of brain activity.• A subject-specific HRF resulted in more consistent aging and vascular risk effects.

FMRI hemodynamic response function (HRF) as a novel marker of brain function: applications for understanding obsessive-compulsive disorder pathology and treatment response

“…Although it is intuitive to associate lower RH with larger TTP or FWHM, it may not always be the case. For example, a recent study found lower RH in aging and in those with vascular risk but no altered TTP or FWHM in those cases [12]. Likewise, in our study the majority of regions showing RH changes were not associated with TTP or FWHM changes.…”

The confound of hemodynamic response function variability in human resting-state functional MRI studies

“…Obtaining simultaneous rs-EEG/fMRI data and considering EEG as the neural input to deconvolve fMRI [using AFNI's 3dDeconvolve (Feige et al, 2017)] is problematic because generative fMRI models do not consider scalp EEG as properly representing BOLD-inducing neural activity (Logothetis et al, 2001). A viable alternative is estimating HRF latency from a hypercapnic challenge because breathhold causes vasodilation and modulates cerebral blood flow (CBF) (Thomason et al, 2007;Hall et al, 2016;McDonough et al, 2019) independent of neural activity, allowing us to measure vascular latency. Chang et al (Chang et al, 2008) utilized this to correct for vascular latency prior to connectivity analysis.…”

The Confound of Hemodynamic Response Function Variability in Resting-State Functional MRI Studies

“…Obtaining simultaneous rs-EEG/fMRI data and considering EEG as the neural input to deconvolve fMRI (using AFNI's 3dDeconvolve [28]) is problematic because generative fMRI models do not consider scalp EEG as properly representing BOLD-inducing neural activity [4]. A viable alternative is estimating HRF latency from a hypercapnic challenge because breath-hold causes vasodilation and modulates cerebral blood flow (CBF) [29][30] [31] in the absence of neural activity, allowing us to measure vascular latency. Chang et al [32] utilized this to correct for vascular latency prior to connectivity analysis.…”