BOLD hemodynamic response function changes significantly with healthy aging

ABSTRACT Functional magnetic resonance imaging (fMRI) has been used to infer age‐differences in neural activity from the hemodynamic response function (HRF) that characterizes the blood‐oxygen‐level‐dependent (BOLD) signal over time. BOLD literature in healthy aging lacks consensus in age‐related HRF changes, the nature of those changes, and their implications for measurement of age differences in brain function. Between‐study discrepancies could be due to small sample sizes, analysis techniques, and/or physiologic mechanisms. We hypothesize that, with large sample sizes and minimal analysis assumptions, age‐related changes in HRF parameters could reflect alterations in one or more components of the neural‐vascular coupling system. To assess HRF changes in healthy aging, we analyzed the large population‐derived dataset from the Cambridge Center for Aging and Neuroscience (CamCAN) study (Shafto et al., 2014). During scanning, 74 younger (18–30 years of age) and 173 older participants (54–74 years of age) viewed two checkerboards to the left and right of a central fixation point, simultaneously heard a binaural tone, and responded via right index finger button‐press. To assess differences in the shape of the HRF between younger and older groups, HRFs were estimated using FMRIB's Linear Optimal Basis Sets (FLOBS) to minimize a priori shape assumptions. Group mean HRFs were different between younger and older groups in auditory, visual, and motor cortices. Specifically, we observed increased time‐to‐peak and decreased peak amplitude in older compared to younger adults in auditory, visual, and motor cortices. Changes in the shape and timing of the HRF in healthy aging, in the absence of performance differences, support our hypothesis of age‐related changes in the neural‐vascular coupling system beyond neural activity alone. More precise interpretations of HRF age‐differences can be formulated once these physiologic factors are disentangled and measured separately. HIGHLIGHTSThere are age‐related changes in the shape and timing of the HRF in healthy aging.Small sample sizes and diverse methods weaken meaningful inter‐study conclusions.BOLD signal results from a complex underlying physiology.Results indicate that age‐differential BOLD is not due to neural activity alone.Results support the hypothesis of age‐related alterations in neural‐vascular coupling.