Association of critically short telomeres with brain and blood markers of ageing and Alzheimer's disease in older adults

Abstract

Background: Accumulation of critically short telomeres (CST) is implicated in decreased tissular regenerative capacity and increased susceptibility to degenerative diseases such as Alzheimer's disease (AD). Telomere shortening has also been associated with age-related brain changes. However, it remains unclear whether CST accumulation is directly associated with AD markers or instead amplifies age-related effects, potentially increasing susceptibility of developing AD in cognitively healthy older adults.

Methods: This cross-sectional study used baseline data of 129 community-dwelling cognitively healthy older adults from the Age-Well trial (NCT02977819), aged 65 years and older enrolled between 2016 and 2018, in France. Using linear regressions, we analyzed the relationship between an innovative marker of telomere shortening, the percentage of CST (%CST), structural, functional and molecular neuroimaging outcomes, and multiple blood-based biomarkers related to AD pathophysiology. The effect of apolipoprotein E ε4 genotype (APOE4) was assessed on these relationships using interaction analysis.

Results: A higher %CST was associated with lower global kurtosis fractional anisotropy (β = -.230; P = .010), particularly in frontal and temporal regions. A higher %CST was also related to higher plasma levels of Neurofilament light chain (β = .195; P = .020) and a lower subiculum volume (β = -.206; P = .020), although these associations did not meet the threshold for multiple comparisons. %CST was not associated with AD-related neuroimaging markers, including the AD-sensitive gray matter pattern (β = -.060; P = .441), glucose metabolism pattern (β = -.099; P = .372), brain perfusion pattern (β = -.106; P = .694) or hippocampus volume (β = -.106; P = .194). In APOE4 carriers, higher %CST was associated with lower subiculum (β = -.423; P = 0.003), DG (β = -.410; P = 0.018) and CA1 volumes (β = -.373; P = 0.024), even though associations with DG and CA1 volumes did not survive multiple comparison.

Conclusions: Although an increase in %CST does not appear to be directly linked to the pathophysiology of AD in cognitively healthy older adults, it could heighten the susceptibility of APOE4 carriers to develop AD plausibly due to greater vulnerability to age-related effects. However, longitudinal studies would be necessary to determine whether %CST influences the development and progression of AD later in life.

Keywords: Ageing; Alzheimer’s disease; Biomarkers; Neuroimaging; Telomeres.

Fig. 1

Flow diagram of the study. Abbreviations: DKI, Diffusion Kurtosis Imaging; GFAP, Glial Fibrillary Acidic Protein; p-Tau181, phosphorylated-Tau181, Aβ; β-Amyloid

Fig. 2

Voxelwise associations between the %CST and kFA. Negative voxel-wise multiple regression between %CST and kFA are presented, controlling for age, sex, education and BMI, in 127 healthy older adults. Results are presented at a p_uncorrected < 0.005 threshold combined with a cluster-level multiple comparisons correction. (1) Superior frontal gyrus white matter, Cingulum, Corpus Callosum; (2) Left inferior longitudinal fasciculus, uncinate fasciculus, fornix. Abbreviations: %CST, Percentage of Critically Short Telomeres; kFA, kurtosis Fractional Anisotropy; BMI, Body mass index

Fig. 3

Associations of the %CST with hippocampal subfield volumes, in cognitively healthy older adults and according to APOE4 status. A Scatterplots of linear regression between %CST and hippocampal subfield volumes, in 125 healthy older adults. B Scatterplots of linear regression between %CST and hippocampal subfields volumes, according to APOE4 status. Interaction and Post-hoc analyses are indicated. Analysis are corrected by age, sex, education and BMI. Detailed statistics of the analyses are summarized in Table 6 and Table 8 in supplementary 3. 3D representations of hippocampus subfields were obtained using 3D slicer software based on an ASHS segmentation. All hippocampal subfield volumes are TIV normalized. *P < 0.05. **P < 0.01. ***P < 0.001. Abbreviations: %CST, Percentage of critically short telomeres; APOE4, Apolipoprotein E ε4; CA, Cornu Ammonis

Fig. 4

Associations of the %CST with neurodegeneration and AD physiopathology-related blood-based markers in cognitively healthy older adults and according to APOE4 status. A Scatterplots of linear regression between %CST and blood-based markers, in 122 healthy older adults. B Scatterplots of linear regression between %CST and blood-based markers, according to APOE4 status. Interaction and post-hoc analyses are indicated. Analysis are corrected by age, sex, education and BMI. Detailed statistics of the analyses are summarized in Table 7 and Table 9 in supplementary 3. *P < 0.05. **P < 0.01. ***P < 0.001. Abbreviations: %CST, Percentage of Critically Short Telomeres; APOE4, Apolipoprotein E ε4; NfL, Neurofilament Light chain, GFAP; Glial Fibrillary Acidic Protein, p-Tau181; phosphorylated-Tau181, Aβ; β-Amyloid