Abstract 125: Enlarged Perivascular Spaces In The Centrum Semiovale Are Associated With Incident Late-life Depression: Framingham Heart Study

• Published in: Stroke (1970) Vol. 54; no. Suppl_1; p. A125
• Main Authors: Medved, Sara, Pinheiro, Adlin, parva, pedram, Himali, Jayandra J, Seshadri, Sudha, Demissie, Serkalem, Romero, Jose R
• Format: Journal Article
• Language: English
• Published: Lippincott Williams & Wilkins 01-02-2023
• ISSN: 0039-2499; 1524-4628
• DOI: 10.1161/str.54.suppl_1.125

IntroductionPerivascular spaces visible on brain MRI (PVS) are considered cerebral small vessel disease (CSVD) markers, possibly reflecting impaired brain fluid and metabolite clearance, and could contribute to depression in old age. This study aims to determine the association of PVS with prevalent and incident depression in the community-based Framingham Heart Study (FHS). MethodsFHS participants free of stroke and dementia with available Center for Epidemiological Studies Depression (CES-D) scale and PVS ratings in the basal ganglia (BG) and centrum semiovale (CSO) were included. Depression was defined as CES-D score≥16 or use of antidepressants. PVS were rated on a scale from I-IV in the BG and CSO using a validated method, with grades III-IV representing high PVS burden. A mixed score was created to describe high burden in the BG and/or CSO (0=no high ePVS burden, 1=high burden in the BG, 2=high burden in the CSO, 3=high burden in both regions). Logistic regression analysis was used to assess the association between PVS in the CSO, BG, and mixed regions and prevalent and incident depression. Models adjusted for age, sex, FSH cohort, time interval between CES-D score and MRI measurements, body mass index, smoking status, prevalent hypertension and diabetes. ResultsAmong 3815 participants included (mean age 57±14, male 47%), incident depression was observed in 19% over the follow up period ranging from 8 to 15 years subsequent to PVS rating. In multivariable adjusted analyses, high burden PVS in the CSO was associated with incident depression (OR 1.75 [1.18, 2.60], p=0.005). In the mixed score, concurrent high burden PVS in both regions (OR 2.10 [1.13, 3.91], p=0.02) and high burden in the CSO only (OR 1.65 [1.05, 2.58], p=0.03) were associated with incident depression. PVS were not associated with prevalent depression. ConclusionIn conclusion, high PVS burden was associated with incident depression in a community-based sample free of stroke and dementia. The association was predominantly driven by high burden of CSO PVS, suggesting that lobar forms of CSVD may play role in late life depression, possibly reflecting disrupted neuroanatomic circuits involved in mood control, and differing from BG PVS.