Cerebral amyloid-β, glucose metabolism and hippocampal volume in major depression with and without suspected non-Alzheimer pathophysiology (SNAP).

Abstract

Background: Major depressive disorder (MDD) can encompass suspected non-Alzheimer's disease pathophysiology (SNAP), characterized by negative cerebral Amyloid-β (Aβ) and positive neurodegeneration (ND) biomarkers. Relationships among cerebral Aβ deposition, glucose metabolism, and hippocampal volume in SNAP and non-SNAP MDD groups remain unclear.

Methods: We conducted cross-sectional analyses of 136 MDD patients (enrolment age ≥ 50 years), classified into SNAP (n = 71) and non-SNAP (n = 65) groups. Dual-phase ¹⁸F-florbetapir (AV45) PET/MR scans provided early-phase (as a proxy for glucose metabolism) and late-phase (reflecting Aβ deposition) images. Bilateral adjusted hippocampal volume (HVa) was measured using MRI. Partial correlations and multiple regression analyses examined the associations among Aβ deposition, metabolism, HVa, and depression-related variables.

Results: In the non-SNAP MDD, greater Aβ deposition was associated with reduced glucose metabolism and smaller HVa, while HVa positively correlated with metabolism. Glucose metabolism partially mediated the relationship between Aβ deposition and both hippocampal atrophy and cognition function. In the SNAP MDD group, lower Aβ deposition was associated with lower glucose metabolism. While more lifetime depressive episodes were related to smaller HVa, depression severity showed a positive correlation with HVa. Additional subgroup analyses stratified by age and Aβ/ND status were also conducted.

Conclusions: These findings suggest distinct biological patterns in SNAP and non-SNAP MDD. Age- and biomarker (Aβ/ND)-stratified analyses yielded consistent Aβ-metabolism-atrophy profiles; however, the biomarker-stratified results should be interpreted with caution due to limited sample sizes. Notably, the non-SNAP group comprises biologically heterogeneous subgroups, which may obscure important distinctions and limit the interpretability of pooled findings. These results underscore the need for subgroup-specific approaches to better elucidate the pathophysiology of MDD.