Mesial temporal tau is related to worse cognitive performance and greater neocortical tau load in amyloid-β–negative cognitively normal individuals

Highlights

• Mesial temporal tau present in AB-negative cognitively normal individuals.

• Mesial temporal tau related to worse cognitive performance within normal range.

• Mesial temporal tau associated with subtle neocortical tau in AB-negative individuals.

• Tau pathology in the absence of amyloidosis (PART) not as benign as commonly thought.

Abstract

We examined whether mesial temporal (Me) tau relates to cognitive performance in 47 amyloid-β (Aβ)-negative, cognitively normal older adults (>60 years old). Me-tau was measured using [¹⁸F]flortaucipir–positron emission tomography standardized uptake value ratio. The effect of continuous and categorical (stratified at standardized uptake value ratio = 1.2 [21% Me-positive]) Me-tau on cognition (mini-mental state examination, pre-Alzheimer's cognitive composite, a memory composite, and a nonmemory composite score) was examined using general linear models, and associations between Me-tau and [¹⁸F]flortaucipir signal in the neocortex were assessed using voxelwise regressions (continuous) and voxelwise contrasts (categorical). In addition, we assessed the effect of age and Aβ burden on Me-tau. Both continuous and categorical Me-tau was associated with worse cognitive performance across all tests and with higher lateral temporal and parietal [¹⁸F]flortaucipir signal. Furthermore, we observed a marginal association between Me-tau and age, whereas there was no association with Aβ burden. Our findings indicate that Me-tau in Aβ-negative cognitively normal individuals, which is likely age-related (i.e., primary age-related tauopathy), might not be as benign as commonly thought.

Introduction

Recent reports suggest that age-related cognitive decline may be partly attributed to biological processes associated with neurodegenerative diseases (Ritchie et al., 2016; Salthouse, 2010; Schaie, 2005; Wilson et al., 2013), such as Alzheimer's disease (AD) (Baker et al., 2017). However, pathological changes, characteristic of AD, such as amyloid-β (Aβ) accumulation, are present in a significant number of older adults who perform within the normal limits of cognitive functioning (Jack, 2014; Jansen et al., 2015). Furthermore, in accordance with neuropathological findings (Nelson et al., 2012), recent tau-imaging (e.g., [¹⁸F]flortaucipir (FTP)-positron emission tomography (PET)) (Chien et al., 2013; Johnson et al., 2016; Schöll et al., 2016) studies have shown that many people performing within the range of normal cognitive functioning also show tau accumulation, mainly in mesial temporal (Me) regions (Schöll et al., 2016). When this Me-tau pathology occurs in the absence of detectable Aβ pathology, it is commonly referred to as primary age-related tauopathy (PART) (Crary et al., 2014). As many individuals have some degree of tau in the brain and perform within normal limits (Bennett et al., 2006; Davis et al., 1999; Schöll et al., 2016), PART is often regarded as a feature of the aging process (Crary et al., 2014; DeCarlo et al., 2014; Hofer and Sliwinski, 2001). Although one study indeed showed that there is no association between temporal or parietal tau and a cognitive composite score (Schultz et al., 2018), and another that tau is associated with memory decline only in those who are Aβ-positive (Sperling et al., 2019), most studies have shown a clear association between tau and worse cognitive performance (Aschenbrenner et al., 2018; Knopman et al., 2019; Maass et al., 2018; Ziontz et al., 2019). The discussion on whether PART should be viewed as a distinct tauopathy (Jellinger et al., 2015), or is indicative of, and will progress to, AD (Duyckaerts et al., 2015) is still ongoing. Regardless, the effects of tau pathology in Aβ-negative individuals warrant further investigation as this population is often overlooked because of the focus on Aβ-positivity.

As tau spread follows a stereotypical and hierarchical distribution in the brain with the Me lobe being one of the earliest affected regions (both in aging and AD (Braak and Braak, 1991; Delacourte et al., 1999)), we aimed at assessing the effects of Me-tau pathology on cognitive performance and neocortical tau load in Aβ-negative, cognitively healthy older adults.