Is later-life depression a risk factor for Alzheimer's disease or a prodromal symptom: a study using post-mortem human brain tissue?

Abstract

Background: Depression and dementia are both common diseases. Although new cases of depression are more common in younger adults, there is a second peak at the age of 50 years suggesting a different pathological process. Late-life depression (LLD) is associated with dementia. However, it remains unclear whether depression represents a dementia prodrome or is a true risk factor for its development. LLD is thought to have a vascular component and this may be a possible link between depression and dementia. We hypothesised that later-life depression is a prodromal manifestation of dementia and would therefore be associated with more AD, and/or ischaemic brain abnormalities that are present in earlier-life depression or in age- and sex-matched controls.

Methods: We assessed post-mortem orbitofrontal cortex and dorsolateral pre-frontal cortex from 145 individuals in 4 groups: 28 18-50-year-olds with depression, 30 older individuals (ages 51-90) with depression, 28 with early AD (Braak tangle stages III-IV) and 57 matched controls (17 early-life, 42 later-life). Levels of Aβ, phospho-tau and α-synuclein were assessed by immunohistochemistry and ELISA. To quantify chronic ischaemia, VEGF, MAG and PLP1 were measured by ELISA. To assess pericyte damage, PDGFRB was measured by ELISA. For blood-brain barrier leakiness, JAM-A, claudin 5 and fibrinogen were measured by ELISA. To quantity endothelial activation, the ratio of ICAM1:collagen IV was assessed by immunohistochemistry.

Results: There was no evidence of chronic cerebral hypoperfusion or increased Aβ/tau in either depression group. There was also no indication of pericyte damage, increased blood-brain barrier leakiness or endothelial activation in the OFC or DLPFC in the depression groups.

Conclusions: Contrary to some previous findings, we have not found evidence of impaired vascular function or increased Aβ in LLD. Our study had a relatively small sample size and limitations in the availability of clinical data. These results suggest that depression is a risk factor for dementia rather than an early manifestation of AD or a consequence of cerebral vascular insufficiency.

Keywords: Alzheimer’s disease; Amyloid; Blood–brain barrier; Dementia; Depressive disorder; Tau; Vascular depression.

Fig. 1

Markers of neurodegeneration. Insoluble Aβ 40 was increased in the early AD group in both the OFC (A) and DLPFC (C). Insoluble Aβ 42 was also increased in the early AD group in the OFC (B) and DLPFC (D). Conversely, insoluble α-synuclein was decreased in the early AD groups in the OFC (E) and DLPFC (F). In the DLPFC, insoluble Aβ 42 (D) was increased in both the early AD and early AD control groups compared to all other groups. There was no difference between the early AD and early AD control groups in insoluble Aβ 42

Fig. 2

Aβ and tau loads measured by immunohistochemistry and field fraction analysis in the DLPFC (A, B), medial temporal lobe (C, D) and orbitofrontal cortex (E, F). In all areas, the expected increase of both proteins was seen in the early AD group compared to controls

Fig. 3

Markers of perfusion and angiogenesis. The expected reduction in MAG:PLP ratio was not seen in either the OFC (A) or DLPFC (B). Neither was there evidence of increased angiogenesis (C, D) or microvessel density (E, F) or pericyte damage (G, H)

Fig. 4

Blood–brain barrier integrity and endothelial activation. There was no evidence of a between-group difference in JAM-A (A), occludin (B, D), claudin-5 (C, E) or ICAM-1:collagen ratio (F)