Long‐Term Blood Pressure Variability, Hippocampal Subfield Volume and Cognitive Decline: A Population‐Based Prospective
MRI
Study
Fei Li, Jing Chen, Huijing Shi, Ling Yang, Xianyu Zhu, Rui Li, Xinyu Zhao, Xiaoliang Liang, Shun Zhang, Yaping Zhang, Han Lv, Jing Li, Shuohua Chen, Yuntao Wu, Ying Hui, Shouling Wu, Zhenchang Wang, Xiaoshuai Li ABSTRACT
Background
Long‐term blood pressure variability (BPV) is associated with cognitive decline, yet the underlying mechanisms have not been fully elucidated.
Purpose
To investigate the associations between long‐term BPV and hippocampal subfield volume and cognitive function and to assess the mediation effect of hippocampal subfield volume.
Study Type
Prospective.
Population
One thousand three hundred sixty‐six adults attended at least 3 surveys between 2006 and 2020 as part of the Kailuan Study.
Field Strength/Sequence
3T; 3D T1‐weighted sequence (T1WI).
Assessment
Cognitive function was evaluated by Montreal Cognitive Assessment (MoCA) at the last (7th) survey visit and brain MRI acquired. The hippocampus was segmented into 12 subfields via FreeSurfer software: molecular layer, cornu ammonis (CA)1, CA2‐3, CA4, subiculum, presubiculum, parasubiculum, hippocampal amygdalar transition area (HATA), dentate gyrus (DG), fimbria, hippocampal tail, and hippocampal fissure. BPV was defined as the standard deviation of manual blood pressure (BP) readings acquired at each survey visit. Associations between BPV, hippocampal subfield volume, and cognitive function were assessed.
Statistical Tests
Associations were evaluated using generalized linear models. To explore whether hippocampal subfield volume mediated the association between BPV and cognitive function, mediation analysis was employed.
Results
Higher systolic BPV (SBPV) was significantly associated with lower volumes in CA1, CA2‐3, molecular layer, and HATA ( β = −8.99, −3.15, −7.18, and −1.27, respectively). Elevated diastolic BPV (DBPV) was significantly associated with lower molecular layer volume ( β = −5.71). SBPV and DBPV were significantly negatively associated with MoCA scores ( β = −0.24 and −0.20, respectively). Decreased volumes in the molecular layer, CA4, CA2‐3, CA1, subiculum, DG, fimbria, and hippocampal tail were significantly associated with lower MoCA scores ( β = −0.32, −0.29, −0.25, −0.28, −0.26, −0.31, −0.53, and −0.26, respectively). The association between SBPV and MoCA scores was mediated by molecular layer volume (proportion: 9.45%).
Data Conclusions
Elevated BPV was associated with hippocampal subfield atrophy and cognitive decline. The association between SBPV and cognitive decline appears to be mediated by molecular layer volume.
Evidence Level
3.
Technical Efficacy
3.