Age differences in brain signal variability are robust to multiple vascular controls

Douglas D Garrett,Claudine J Gauthier,Richard D Hoge,Ulman Lindenberger

doi:10.1038/s41598-017-09752-7

Abstract

A host of studies support that younger, better performing adults express greater moment-to-moment blood oxygen level-dependent (BOLD) signal variability (SDBOLD) in various cortical regions, supporting an emerging view that the aging brain may undergo a generalized reduction in dynamic range. However, the exact physiological nature of age differences in SDBOLD remains understudied. In a sample of 29 younger and 45 older adults, we examined the contribution of vascular factors to age group differences in fixation-based SDBOLD using (1) a dual-echo BOLD/pseudo-continuous arterial spin labeling (pCASL) sequence, and (2) hypercapnia via a computer-controlled gas delivery system. We tested the hypothesis that, although SDBOLD may relate to individual differences in absolute cerebral blood flow (CBF), BOLD cerebrovascular reactivity (CVR), or maximum BOLD signal change (M), robust age differences in SDBOLD would remain after multiple statistical controls for these vascular factors. As expected, our results demonstrated that brain regions in which younger adults expressed higher SDBOLD persisted after comprehensive control of vascular effects. Our findings thus further establish BOLD signal variability as an important marker of the aging brain.

Highlights

The study of lifespan development, cognition, and brain signal variability continues to gain momentum in cognitive neuroscience[1,2,3,4,5,6,7] via EEG, MEG, and fMRI
cerebral blood flow (CBF), blood oxygen level-dependent (BOLD) cerebrovascular reactivity (CVR), and M represent a comprehensive index of potential vascular contributions to BOLD, allowing us to address how accounting for such vascular factors may impact age group differences in BOLD signal variability
One of several BOLD signal variance measures utilized in fMRI research, we focused on a modified voxel-wise time series standard deviation from fixation block data (SDBOLD_fix)

Summary

Introduction

The study of lifespan development, cognition, and brain signal variability continues to gain momentum in cognitive neuroscience[1,2,3,4,5,6,7] via EEG, MEG, and fMRI. Age differences in vascular properties could provide one potential reason why BOLD signal variability appears generally reduced in older adults[1, 13, 14]. More direct measures of vascular factors may be needed to support principled interpretations of age differences in BOLD signal variability. It remains to be seen whether controlling for vascular rigidity and reactivity[20, 21] would eliminate observed age group differences in BOLD signal variability. CBF, BOLD CVR, and M represent a comprehensive index of potential vascular contributions to BOLD, allowing us to address how accounting for such vascular factors may impact age group differences in BOLD signal variability. A host of studies have examined how various types of vascular scaling impact standard analyses of age differences in mean BOLD signals[17, 33, 34], no study to date has examined whether typically found age differences in BOLD signal variability[1, 13] remain robust after comprehensive control of vascular factors

Methods

Results

Conclusion