Cardiovascular risk factors are associated with lower posterior-medial network functional connectivity in older adults

Abstract

Background: Cortico-hippocampal functional networks, specifically the anterior-temporal (AT) and posterior-medial (PM) systems, are crucial for memory and highly vulnerable to aging and Alzheimer's disease (AD). While modifiable cardiovascular risk factors may offer prevention opportunities to preserve brain aging, their effects on AT/PM functional connectivity remain unknown. This study aims to investigate these associations in older adults, considering major risk categories and exploring potential interactions with protective lifestyle habits and AD risk factors.

Methods: One hundred thirty-one community-dwelling cognitively unimpaired adults aged 65 + were selected from the Age-Well trial, a French monocentric population-based study conducted from 2016 to 2020. Resting-state fMRI and cardiovascular risk assessments were performed at baseline and 18-month follow-up. Functional connectivity within the AT and PM networks was derived from seed-based analyses using the perirhinal and parahippocampal cortices as individual seeds, respectively. Generalized additive and linear mixed models assessed the effects of cardiovascular risk factors on AT/PM functional connectivity, including interactions with protective lifestyle habits and AD risk factors.

Results: Baseline mean age was 69 (65-84) years, with 63.5% women. Higher abdominal fat (95% CI: -0.00118, -0.00005; F = 5.39; P =.02), higher LDL cholesterol (95% CI: -0.01642, -0.00345; F = 10.40; P =.001), longer smoking duration (95% CI: NA; F = 3.89; P =.03) and greater alcohol consumption (95% CI: -0.01134, -0.00045; F = 4.66; P =.02) were consistently associated with lower PM connectivity, collectively explaining 11.4% of the variance. However, only LDL cholesterol survived multiple comparisons, possibly reflecting a more direct involvement in cardiovascular mechanisms affecting functional connectivity. No association was found with AT connectivity. Exploratory analyses showed that these relationships were independent of cerebral Aβ-positivity or APOE-ε4 carrier status and were unaffected by physical activity and Mediterranean diet when considered separately.

Discussion: This study highlights converging associations between higher cardiovascular risk factors and lower functional connectivity in cognitively unimpaired older adults, specifically affecting the PM-but not AT-network, and independent of AD risk. Targeting these specific modifiable factors may prevent age-related network alterations to promote cognitive health in aging.

Trial registration information: The Age-Well trial was registered with ClinicalTrials.gov on November 25, 2016 (identifier: NCT02977819).

Keywords: Aging; Aslzheimer’s disease; Cardiovascular risk factors; Functional connectivity; Functional magnetic resonance imaging; Medial temporal lobe.

Fig. 1

Flow Diagram of the inclusion process. Following data quality assessment, our study involved 131 of the 157 participants from the Age-Well clinical trial. The Age-Well monocentric randomized trial was originally designed to assess the impact of 18-month non-pharmacological interventions, with participants assigned to one of three parallel arms: (i) the meditation arm receiving meditation-based training, (ii) the non-native language arm undergoing structurally matched language training and (iii) the no-intervention control arm. The Age-Well design is presented for contextual background, but our study was purely observational, using baseline and follow-up data solely to increase sensitivity of our analysis. Abbreviations: (f)MRI, (functional) magnetic resonance imaging

Fig. 2

The anterior-temporal (AT) and posterior-medial (PM) networks. Detailed procedure is described in Supplementary methods. Abbreviations: AT, anterior-temporal network; FWE, family-wise error; PHC, parahippocampal cortex; PM, posterior-medial network; PRC, perirhinal cortex; TFCE, Threshold-Free Cluster Enhancement

Fig. 3

Cardiovascular risk factors associated with PM, but not AT, functional connectivity. (A) Illustration of the masks for the anterior-temporal (AT – in red) and posterior-medial (PM – in blue) networks. (B) Relationships between cardiovascular risk factors and the mean perirhinal/parahippocampal functional connectivity within the AT/PM network, respectively. Solid lines represent significant associations and dashed lines nonsignificant ones (P >.05). Regression lines indicate partial effects based on model-derived estimates. Shaded areas represent 95% confidence intervals. Raw data points are presented in Figures S2-5. (C) Bar graph of the explained variance of the fixed effects for each model. The y-axis values represent the marginal pseudo-R-squared multiplied by 100. Abbreviations: AT, anterior-temporal network; AUDIT-QF, Alcohol Use Disorder Identification Test-Quantity Frequency; LDL, Low-density lipoprotein cholesterol; PM, posterior-medial network

Fig. 4

APOE-ε4 and Aβ-positivity modulate the association between PM connectivity and smoking or drinking habits. Solid lines represent significant associations and dashed lines nonsignificant ones, based on post-hoc t-tests using estimated marginal mean values (P >.05). Regression lines indicate model-derived estimates. Shaded areas represent 95% confidence intervals. Raw data are presented in Figures S7-9. Abbreviations: Aβ = β-amyloid; APOE4, Apolipoprotein E-ε4; AUDIT-QF, Alcohol Use Disorder Identification Test-Quantity Frequency; PM, posterior-medial network