Association of AD biomarker levels with functional connectivity within and between large‐scale brain networks among cognitively normal individuals

Abstract

AbstractBackgroundIt is well documented that amyloid pathology accumulates many years prior to the emergence of clinical symptoms of Alzheimer’s disease (AD), when individuals are cognitively normal. Few prior studies have investigated the impact of AD pathology on brain connectivity among cognitively normal individuals. We investigated the impact of AD biomarkers on functional connectivity within and between large‐scale brain networks important for cognitive function.MethodAnalyses included 123 cognitively normal participants (mean age=69 years) from the BIOCARD Study who completed clinical and cognitive assessments, resting state functional magnetic resonance imaging (rs‐fMRI) scans, and cerebrospinal fluid (CSF) measures of AD biomarkers [abeta1‐42 (AB42), abeta1‐40 (AB40), total tau (ttau), and phosphorylated tau 181 (ptau)] during their annual visit. Among these, 106 participants also had 11C‐PiB PET imaging to assess amyloid burden. Functional connectivity within 7 large‐scale, non‐overlapping networks was measured and averaged to compute global within‐network connectivity. Between‐network connectivity was measured using graph theory‐based metrics of modularity and segregation. Linear regressions (covarying age, sex, education, whole brain volume, and vascular risks) tested the associations between the connectivity measures and the CSF and PET AD biomarkers.ResultOlder age was associated with reduced global connectivity within networks and reduced network modularity and segregation (all p’s <0.05), indicating reduced functional specialization with older age, consistent with prior studies. Amyloid positivity on PET, as well as lower AB42/AB40 and higher ttau/AB42 and ptau/AB42 ratios were associated with higher network modularity and segregation (all p’s <=0.01). When AB42/40, ttau, and ptau were simultaneously entered in the regressions, only AB42/AB40 remained significantly associated with modularity and segregation (p’s<0.01). There were no associations between the CSF or PET biomarkers and global within‐network connectivity.ConclusionAmong older individuals with normal cognition, more abnormal AD biomarker levels were related to greater network modularity and segregation. Given that higher modularity and segregation are important for maintaining cognitive function with age, these results suggest that the associations with rs‐fMRI connectivity may reflect a compensatory response of the brain to accumulating pathology. These findings are consistent with autopsy studies showing compensatory increases in neuronal size (neuronal hypertrophy) among asymptomatic individuals with AD pathology.