Abstract

Vascular depression is a proposed subtype of late-life depression associated with cerebrovascular disease and defined clinically or by MRI. Two 1997 proposals named the construct, and the hypothesis behind it holds that cerebrovascular damage may raise vulnerability to some geriatric depressive syndromes, trigger them or keep them going. The account ties depression to white matter hyperintensities and executive dysfunction, yet prospective effects are modest or partial, some postmortem studies found no vascular link, and genetic designs suggest depression liability may precede vascular outcomes. Claims that lesions or executive deficits predict poorer treatment response remain contested, and late-life depression is associated with later dementia, with the largest pooled estimate for vascular dementia. Its validity as a distinct subtype remains debated, and it has no widely accepted diagnostic criteria and does not appear in DSM-5.

Keywords: vascular depression, late-life depression, white matter hyperintensities, executive dysfunction, cerebrovascular disease

Vascular depression is a proposed subtype of late-life depression in which cerebrovascular disease, or vascular risk, is thought to contribute to the illness. The founding hypothesis held that cerebrovascular disease could make some geriatric depressive syndromes more likely, set them off or sustain them, with ischemic lesions in prefrontal systems accumulating past a threshold (Alexopoulos et al., 1997a), and an MRI-based definition appeared the same year (Krishnan et al., 1997). MeSH lists subcortical ischemic depression as an entry term, the label Krishnan's group later applied to MRI-defined cases (Krishnan et al., 2004). A consensus report judged vascular depression a plausible subtype whose evidence is contested (Aizenstein et al., 2016). The topic matters to cognitive psychology because executive dysfunction is central to it: in older patients with major depression, a depression-executive dysfunction syndrome carried a distinct clinical profile and was tied to disability (Alexopoulos et al., 2002). This page covers the proposed late-life subtype linked to cerebrovascular disease; depressive mood in general is covered on the depression page, and the most common depressive diagnosis is covered on the major depressive disorder page.

The article first defines the construct and traces its history, then turns to its symptoms, diagnosis and measurement, to what the neighbouring first-person evidence reports about the experience, and to the types of vascular depression, from the clinical and MRI definitions to boundary presentations such as post-stroke depression. It then asks whether vascular depression is a distinct subtype, reports prevalence, course and burden, and sets out the founding hypothesis beside its rival and complementary accounts. Sections on executive dysfunction and cognition, on testing cause and effect, on biology, on brain networks and white matter, on brain and cognitive reserve and on the validity of lesion ratings and executive tests follow. The relation to dementia, co-occurring conditions, suicide and self-harm, and who has been studied come next. The treatment sections cover the trial evidence, medication safety in older adults, how treatments are thought to work, brain stimulation and other biological treatments, treatments under investigation, and prevention through vascular risk-factor management. The article closes with living with vascular depression, related conditions and how they differ, and current directions in research.

Key Takeaways

  • Two definitions appeared in 1997, one clinical and one MRI-based. Whether vascular depression is a distinct subtype is still unsettled: latent class and clinical-profile studies support a separate group, community cohorts and postmortem studies do not, and the condition has no widely accepted diagnostic criteria and is absent from DSM-5.
  • The vascular depression hypothesis proposes that cerebrovascular damage may raise vulnerability to some geriatric depressive syndromes, trigger an episode or keep one going; the lesion threshold was a prediction, not a measured result. White matter hyperintensities were more common in late-life depression than in controls in one meta-analysis, but another pooled analysis found the association weak.
  • Causality is not established, and genetic designs suggest depression liability may itself precede vascular and white matter outcomes. Late-life depression predicted later dementia in pooled cohorts, with the largest estimate for vascular dementia, but whether depression is a risk factor, a prodrome or a product of shared disease is unresolved.
  • Antidepressants beat placebo in late-life depression, though their advantage was smaller than in adults and was not significant at age 65 and older. Lesion burden and executive deficits predicted poorer response in several single studies, but meta-analyses support executive dysfunction as a predictor only for a few tests.
  • Problem-solving therapy outperformed supportive therapy in later weeks in one trial of older adults with depression and executive dysfunction, and four psychotherapies beat usual care in older adults with depression generally. A sham-controlled rTMS trial in vascular depression supported efficacy, most clearly at the higher dose; a single unreplicated tDCS trial is an early result; ECT evidence comes from non-randomized cohorts; and the nimodipine trials come from one group. No treatment is confirmed as specific to vascular depression, and certainty is low.
  • For depression after stroke, antidepressants beat placebo in a meta-analysis, and psychological treatments helped in mild but not moderate to severe cases, with low to very low certainty. For prevention, drugs did not differ from placebo after stroke, psychotherapy gave a small benefit, and in an observational cohort two blood-pressure drug classes did not differ in incident depression.
  • In an observational cohort of older depressed patients, SSRIs were associated with more falls and hyponatraemia than no antidepressant, and other antidepressants with several adverse outcomes, but confounding by indication means these hazards are not causal; SSRIs combined with oral anticoagulants were associated with more major bleeding. Stroke survivors had a higher risk of suicidal behaviour than people without stroke in observational studies, though none studied vascular depression itself.
  • The evidence comes mostly from older adults in clinical and community samples, no qualitative study cited here recruited people diagnosed with vascular depression, and how far the findings generalise across populations remains untested.

What Vascular Depression Is

MeSH defines vascular depression as late-life depression associated with clinical or MRI-defined cerebrovascular pathology (MeSH, D000088323). MeSH introduced the descriptor in 2022, indexing the topic under cerebrovascular disorders and depressive disorder before then, files it under both cerebrovascular disorders and depressive disorder, and lists subcortical ischemic depression as its entry term.

A Proposed Subtype of Late-Life Depression

The construct starts from a hypothesis. The hypothesis held that cerebrovascular disease could give rise to some depressive syndromes of later life, and it treated damage to prefrontal systems as the key pathway (Alexopoulos et al., 1997a). A later consensus report describes vascular depression as a plausible subtype of late-life depression, while stating that the evidence for it is contested (Aizenstein et al., 2016). Vascular depression is therefore best read as a proposed subtype, not a settled diagnostic entity.

Studies of the construct usually place it within geriatric psychiatry, where it contrasts with nonvascular depression in older patients. In an MRI-based comparison of 89 depressed patients, 32 were classed as vascular and 57 as nonvascular. The vascular group was older, more often had a late onset and more often had a nonpsychotic form of illness (Krishnan et al., 1997). The same authors concluded that the profile of these patients needed to be developed further, and that this was likely to carry therapeutic implications.

Two routes to a definition appear in this literature. One rests on clinical features together with vascular risk or established vascular disease. A comparison of 33 vascular and 32 nonvascular elderly depressed patients used the Cumulative Illness Rating Scale-Geriatrics in this spirit (Alexopoulos et al., 1997b). The other rests on MRI evidence of subcortical lesions, as in the classification used by Krishnan and colleagues (Krishnan et al., 1997). A review of the MRI-defined form judged its validity partly supported, with considerable supporting data but critical issues still open (Culang-Reinlieb et al., 2011). The two routes, and what each requires, are set out under Types of Vascular Depression.

The features most often described come from these comparisons. Relative to nonvascular depression, the vascular group showed greater cognitive impairment and disability, poorer fluency and naming, more retardation and more lack of insight, but less agitation and guilt (Alexopoulos et al., 1997b). In a case-control study of 139 depressed elderly subjects, ischemic depression was associated with greater age, a history of hypertension and lassitude, and with less family history of mental illness and less loss of libido. Age was the strongest predictor (Krishnan et al., 2004). Those authors suggested it may be a distinct late-life syndrome and noted that cognitive function and treatment implications needed further study.

A related profile centres on executive dysfunction. In 126 elderly patients with major depression, those with the depression-executive dysfunction syndrome showed more psychomotor retardation and paranoia, and fewer vegetative symptoms, than those without it. Psychomotor retardation, loss of interest and paranoia each predicted disability (Alexopoulos et al., 2002). Identifying the syndrome was suggested to help in close follow-up of patients.

Where It Sits in the Classifications

Vascular depression is not an established diagnostic category. The consensus report notes that no widely accepted diagnostic criteria exist and that the condition is absent from DSM-5 (Aizenstein et al., 2016).

Part of the reason lies in the criteria themselves. The review of MRI-defined vascular depression found diagnostic criteria inconsistent, the type and location of lesions unresolved and the course of the disorder unclear. It also stated that a causal relationship had not been established (Culang-Reinlieb et al., 2011). The consensus report likewise calls the causal relationship controversial (Aizenstein et al., 2016).

Reviewers disagree about how much the label should carry. One review treated the link between late-life depression and vascular brain disease as bidirectional and called the concept a useful framework that may be conceptually too restrictive, since inflammatory processes and non-traditional vascular risk factors could mediate the association (Baldwin, 2005). A later hypothesis paper argued that testing mechanisms was needed, with treatment implications to follow (Taylor et al., 2013). The evidence for and against treating the condition as a distinct subtype is weighed in Is Vascular Depression a Distinct Subtype?

History of the Concept

The idea that damaged brain vessels can produce depression in old age predates the term. Krishnan and McDonald (1995) set out the concept of arteriosclerotic depression, in which atherosclerosis of the brain leads to depressive illness. They argued that structural brain change counts for more than genetic or psychosocial factors in late-onset depression (Krishnan & McDonald, 1995). The imaging evidence arrived from a different direction. A case-control MRI comparison of elderly depressed patients receiving electroconvulsive therapy and normal controls found subcortical hyperintensities on T2-weighted scans both more frequent and more severe in the depressed group (Coffey et al., 1990). Five years later, a clinical MRI study of 39 inpatients with severe depression linked greater white matter hyperintensity severity to a poorer treatment response, which the authors described as preliminary evidence of prognostic significance (Hickie et al., 1995).

Two proposals in 1997 turned these observations into a named construct, each defining it differently. Alexopoulos and colleagues compared 33 elderly depressed patients classed as vascular with 32 classed as nonvascular. The vascular group showed more cognitive impairment and disability, and less agitation and guilt (Alexopoulos et al., 1997b). Krishnan and colleagues instead sorted 89 depressed patients by brain MRI, with 32 vascular and 57 nonvascular. The MRI-defined vascular group was older, more often late in onset and more often nonpsychotic, and the authors judged that the profile of these patients needed further development (Krishnan et al., 1997). The accompanying hypothesis paper proposed that cerebrovascular disease predisposes to some geriatric depression, with lesions accumulating past a threshold (Alexopoulos et al., 1997a). A clinical profile of its own followed. In 126 elderly patients with major depression, the depression-executive dysfunction syndrome marked out a distinct clinical picture with disability attached (Alexopoulos et al., 2002).

The idea was then examined rather than simply accepted. A reappraisal ten years after the original proposal took stock of the vascular depression hypothesis (Alexopoulos, 2006). A later narrative review framed vascular disease as a route to late-life depression and set out disconnection, inflammatory and hypoperfusion hypotheses, calling for tests of the mechanisms (Taylor et al., 2013). Its standing is still contested: a consensus report later judged vascular depression a plausible subtype of late-life depression, yet it described the evidence as disputed and the causal relationship at the heart of the hypothesis as controversial (Aizenstein et al., 2016).

The consensus report of Aizenstein and colleagues drew on about 4000 screened articles and reached a measured verdict. Vascular depression is a plausible subtype of late-life depression, but the evidence for it is contested, and its pathogenesis needs further study. It has no widely accepted diagnostic criteria and is absent from DSM-5 (Aizenstein et al., 2016). Vascular depression therefore stands as a contested research construct rather than an established diagnosis. The Medical Subject Headings thesaurus added a descriptor for it in 2022, a development taken up in What Vascular Depression Is.

Symptoms, Diagnosis and Measurement

Vascular depression is recognised through two kinds of evidence at once: a symptom picture assembled at the bedside and a brain-imaging picture read from MRI. The two do not always agree, and the instruments that capture each differ in kind, from self-report screens to clinician-rated scales, performance tasks and visual ratings of scans.

The Clinical Picture

Comparisons of elderly depressed patients with and without vascular disease show a profile weighted toward impairment. In a comparative clinical study of 33 vascular and 32 nonvascular patients, the vascular group showed more cognitive impairment and more disability, along with greater fluency and naming impairment, retardation and lack of insight. Agitation and guilt ran the other way, being lower in the vascular group (Alexopoulos et al., 1997b). The authors read this pattern as fitting lesions in striato-pallido-thalamo-cortical circuits.

A case-control study of 139 depressed elderly subjects, 75 of them (54%) in the ischemic group, found a similar drift. Subcortical ischemic depression was associated with older age, a history of hypertension and lassitude, and with less family history of mental illness and less loss of libido. Age was the strongest predictor. The authors suggested a distinct late-life syndrome, while noting that cognitive function and treatment implications needed further study (Krishnan et al., 2004).

Symptom burden also tracks the extent of lesions within a community-dwelling sample. Among 270 older adults in a multicenter longitudinal cohort of subcortical ischemic vascular disease (mean age 74.4 years), white matter lacunar volume was positively related to depressed mood, anhedonia, apathy and anergia, with cognitive status, age, gender and education controlled. Apathy rose from 1.9% in the cognitively normal to 16.4% in the impaired and 37.6% in the demented, and anergia rose from 9.7% to 25.7% to 45.2%. The analysis was exploratory, and the authors flagged concern about multiple hypothesis testing (Lavretsky et al., 2008).

A distinct profile has not always emerged. Two community-based cohorts of older Dutch adults, the Rotterdam Study and AMSTEL, failed to confirm a specific vascular-depression symptom profile in either sample. Among depressed people, vascular risk factors went with greater physical disability and a larger drop in energy, but the authors detected no significant association between vascular risk and either anhedonia or psychomotor retardation, and neither cohort showed a symptom profile specific to vascular depression (Naarding et al., 2007).

Instruments

The Geriatric Depression Scale (GDS) is a self-report screen built for older adults. Its authors selected 30 questions from a 100-item questionnaire, and in elderly subjects who were normal, mildly depressed or severely depressed it was reliable and valid. Its correlation with Research Diagnostic Criteria symptom counts exceeded that of the Zung Self-Rating Depression Scale and resembled that of the Hamilton Rating Scale for Depression, leading to its suggested use as a valid self-rating screen for the elderly (Yesavage et al., 1982).

Apathy, which features prominently in the lesion-related symptom picture, has its own instrument. The Apathy Evaluation Scale exists in clinician, informant and self-rated versions, and in 123 subjects aged 53 to 85 it proved reliable and construct-valid (Marin et al., 1991).

Performance measures of executive function supplement the ratings. Older adults with major depression performed worse than younger depressed adults, and worse than healthy comparison subjects, on executive tasks in a study of 40 depressed and 40 healthy adults; the age-by-depression interaction for selective and sustained attention was not significant (Lockwood et al., 2002). Initiation and perseveration scores, assessed with the Mattis Dementia Rating Scale alongside the Hamilton Depression Rating Scale and P300 latency, were more abnormal in 25 symptomatic elderly patients than in 24 remitted patients or in controls after six weeks of treatment, and they predicted response change. The authors cautioned that the result needed confirmation (Kalayam & Alexopoulos, 1999). A cross-sectional study of 126 elderly patients with major depression described a depression-executive dysfunction syndrome with a distinct clinical profile and disability. Psychomotor retardation and paranoia were more frequent in that group, and vegetative symptoms less so (Alexopoulos et al., 2002).

Imaging supplies the second axis. The Fazekas visual rating grew out of a 1.5 T MRI comparison of 12 patients with Alzheimer's disease, four with multiinfarct dementia and nine age-matched controls. A periventricular halo was more common in Alzheimer's disease than in controls (p = .024), whereas deep punctate lesions occurred at similar rates in both groups (60%) and were judged non-specific. The authors proposed that the halo might be of diagnostic importance in Alzheimer's disease, in a field where the type, frequency and extent of such signal abnormalities were a subject of controversy (Fazekas et al., 1987).

White Matter Hyperintensity Patterns

Choose a deep and a periventricular pattern to see how each looks on a schematic axial slice.

Deep white matter
Periventricular
Schematic axial slice (white areas = hyperintensity)

Deep: No deep white matter hyperintensity. Periventricular: No periventricular hyperintensity.

Schematic only, not a rating tool. The pattern names follow the deep and periventricular hyperintensity patterns described by Fazekas et al. (1987); the drawing is original and not to scale.

Terminology for the lesions themselves has been less settled than the rating. An international working group set out the STRIVE standards for imaging cerebral small vessel disease, covering white matter hyperintensities, lacunes, small subcortical infarcts, perivascular spaces, microbleeds and brain atrophy on MRI. The group noted that terms and protocols had varied widely, and that this inconsistency hampered progress. The standards were framed for research use but were described as usable clinically (Wardlaw et al., 2013).

The Experience of Vascular Depression

No qualitative study cited on this page recruited people diagnosed with vascular depression as such. The first-person material below therefore comes from neighbouring groups, older adults with depression and long-term physical illness and stroke survivors, and it describes what those groups report rather than what vascular depression is like.

Interviews with 41 primary care patients aged 55 to 75 supply the fullest account. Twelve had depression only, five also had coronary heart disease, ten type 2 diabetes and fourteen arthritis. The authors analysed the semistructured interviews thematically and concluded that core symptoms of depression were common to all participants, while the presence of a long-term condition added to the burden (Poole et al., 2022). Anhedonia and sadness did not differ between participants with and without such a condition, and somatic symptoms were similar across groups. Poor sleep and crying were worse in the arthritis and diabetes groups than in the depression-only group. Thoughts of death were more prominent among participants with heart disease than among those with diabetes or arthritis, yet suicidal thoughts did not differ by condition. Social withdrawal and low motivation interfered with managing the physical illness, and better self-management was in turn linked to improved mood and self-efficacy (Poole et al., 2022). The sample was small for heart disease, the arthritis group was mostly female, recruitment was confined to North London, and participants did not check the findings, so the account is suggestive rather than general.

Stroke survivors have been asked directly. A mixed-methods study of veterans discharged home after a new stroke set the Geriatric Depression Scale beside interviews and patient records to examine their emotional experiences, and it reported that poststroke depression was underrecognised and undertreated (Williams et al., 2005).

Two quantitative studies point to loss of initiative as a feature of the syndrome. Among 270 community-dwelling older adults, white matter lacunar volume was associated with depressed mood, anhedonia, apathy and anergia. Apathy rose from 1.9% in cognitively normal participants to 16.4% in impaired and 37.6% in demented participants, and anergia from 9.7% to 25.7% and 45.2%; the authors called the analysis exploratory and noted the risk of chance findings from multiple tests (Lavretsky et al., 2008). In 118 people with small vessel disease compared with 398 controls, apathy was related to reduced white matter integrity (r = -0.38, P ≤ 0.001), whereas the link for depression was weaker and not significant (r = -0.16, P = 0.09) (Hollocks et al., 2015). Neither study describes slowing in the person's own words, and neither offers a first-person account.

Types of Vascular Depression

Type in this section means a proposed operational definition or a neighbouring presentation. It does not mean a separate disease. MeSH lists no subtypes under this heading, so the divisions below come from the research literature, which offers competing definitions built on clinical signs, on MRI findings and on executive function.

Around those definitions sit boundary presentations that share a burden of vascular brain disease with them. Cognitive findings exist for some of the types and are thin for others, and each type below says which is the case.

Table 1

Proposed Types of Vascular Depression and Related Presentations

TypeStandingKey marker
Clinically defined vascular depressionProposed research criteriaRetardation; fluency and naming impairment
MRI-defined vascular depressionMRI-based research definitionMRI hyperintensities
Depression-executive dysfunction syndromeProposed syndromeExecutive dysfunction
Subthreshold vascular depressionProposed research criteriaModified Fazekas scale
Post-stroke depressionBoundary presentation–
Late-onset depressionBoundary presentationOnset at age 60 or later
Vascular apathyNeighbouring syndromeApathy
Depression in vascular cognitive impairment and vascular dementiaBoundary presentation–
Depression in CADASILBoundary presentation–

Two Ways to Define Vascular Depression

Describe a depressed older adult, then see which of the two 1997 research definitions would classify the case.

Clinically defined (Alexopoulos et al., 1997)

Does not meet the entry criterion

Grouping rested on vascular risk on a medical illness rating scale for older adults.

Compared with the nonvascular group: more cognitive impairment and disability, more retardation, less agitation and guilt.

MRI-defined (Krishnan et al., 1997)

Does not meet the entry criterion

Grouping rested on examination of brain magnetic resonance images.

Compared with the nonvascular group: older age, later onset and the nonpsychotic subtype more often.

Neither definition would classify this case as vascular depression.

A teaching sketch of how the two definitions sort cases, not a diagnostic tool. The group contrasts are those reported by Alexopoulos et al. (1997) and Krishnan et al. (1997).

Clinically Defined Vascular Depression

Alexopoulos and colleagues set out the clinical version of the construct by comparing 33 elderly depressed patients with vascular depression against 32 with nonvascular depression. The vascular group showed more cognitive impairment, more disability, more impairment of fluency and naming, and more retardation and lack of insight. They showed less agitation and guilt. The authors judged that this symptom picture fits lesions in striato-pallido-thalamo-cortical circuits (Alexopoulos et al., 1997b).

A community test did not bear the definition out. Two cohorts of older Dutch adults, the Rotterdam Study and AMSTEL, did not confirm a specific symptom profile, though some differences were found (Naarding et al., 2007).

Cognitive Implications

Cognitive and executive signs are central to this definition. Fluency and naming impairment, retardation and lack of insight set the vascular group apart from the nonvascular group, and the vascular group also showed more overall cognitive impairment (Alexopoulos et al., 1997b). The Executive Dysfunction and Cognition section takes up these findings in detail.

MRI-Defined Vascular Depression

A second definition rests on brain imaging. Krishnan and colleagues classified 89 depressed patients by MRI, 32 as vascular and 57 as nonvascular. The vascular group was older, had a later onset and more often had the nonpsychotic subtype, and it had somewhat less family history of illness. It also showed somewhat more anhedonia and disability. The authors concluded that the profile of patients with vascular depression needed to be developed further and was likely to carry therapeutic implications (Krishnan et al., 1997).

Krishnan's group later used the label subcortical ischemic depression. In a case-control study of 139 depressed elderly people, age was the strongest predictor of the ischemic diagnosis. Lassitude and a past diagnosis of hypertension also raised the likelihood of the ischaemic diagnosis, while loss of libido and a family history of mental illness made it less likely. The authors suggested a distinct late-life syndrome (Krishnan et al., 2004). In one community cohort, MRI-defined vascular depression accounted for 56.2% of major depression, and a further 34.0% of non-major depression met the definition (Park et al., 2015). The Prevalence, Course and Burden section covers such figures.

A literature review found considerable support for the definition's validity but only partial support overall. It listed inconsistent diagnostic criteria, unresolved questions about lesion type and location, an unclear course, and no established causal relationship as open problems (Culang-Reinlieb et al., 2011).

Cognitive Implications

The review placed the type and location of lesions among the unresolved issues on which the validity of MRI-defined vascular depression depends (Culang-Reinlieb et al., 2011). The 2004 case-control study left cognitive function in subcortical ischemic depression for further study (Krishnan et al., 2004).

Depression-Executive Dysfunction Syndrome

The depression-executive dysfunction syndrome shifts the defining feature from the lesion to the cognitive profile. In 126 elderly patients with major depression, those with the syndrome showed more psychomotor retardation and more paranoia than those without it, and fewer vegetative symptoms. The syndrome was tied to disability, and the authors suggested that identifying it might help clinicians plan close follow-up (Alexopoulos et al., 2002). In an 8-week citalopram trial of 112 nondemented elderly patients, poor performance on initiation and perseveration items and on the Stroop test predicted a smaller change in symptoms (Alexopoulos et al., 2005). The Treatment Evidence section handles the debate over treatment response.

Cognitive Implications

The executive profile is the defining element, and the authors linked it to frontostriatal dysfunction. Within the 126 patients, psychomotor retardation, loss of interest and paranoia each predicted disability, and paranoia did so independently of executive dysfunction (Alexopoulos et al., 2002). See Executive Dysfunction and Cognition for the wider evidence.

Subthreshold Vascular Depression

Vascular criteria have also been applied to depression below the threshold for major depression. In a community study with a 3-year follow-up, 4.24% of participants had MRI-defined vascular non-major depression, which was 34.0% of all non-major depression (Park et al., 2015). This is the only community study on this page that reports the type, and its authors noted low response and follow-up rates.

Cognitive Implications

No study on this page reports cognitive findings specific to subthreshold vascular depression.

Post-Stroke Depression

Depression after stroke is a boundary presentation with a large literature of its own. A narrative review called it common and reported that antidepressants outperformed placebo in lowering its incidence (Robinson & Jorge, 2016), whereas a Cochrane review of prevention trials found no clear effect of drug therapy on preventing depression after stroke (Hackett et al., 2008). It also named white matter disease as a possible contributor, and it cautioned that some stroke studies were small and that meta-analyses had merged rigorously assessed patients with patients scoring above arbitrary rating-scale cutoffs (Robinson & Jorge, 2016).

Its frequency has been pooled across 61 studies of 25,488 people at 31%, a figure statistically indistinguishable from the 33% of an earlier review in 2005. Frequency fell to 25% at one to five years after stroke and to 23% at five years, and it varied considerably between studies (Hackett & Pickles, 2014).

An early hypothesis tied depression to the location of the lesion, especially the left anterior region. A systematic review of 48 eligible reports, drawn from 143, found no support for it. Left and right hemisphere lesions did not differ (RR 0.95), and a left anterior lesion did not differ significantly from other regions (RR 1.17) (Carson et al., 2000). The location hypothesis therefore lacks support, and the later narrative review pointed to white matter disease, inflammation and genetic variation as other possible contributors (Robinson & Jorge, 2016).

Cognitive Implications

The review reports that cognitive impairment was greater in major post-stroke depression than in minor depression (Robinson & Jorge, 2016).

Late-Onset Depression

Late-onset depression is a boundary construct. It overlaps with vascular depression because onset in later life has been used as a marker of vascular involvement, and the Related Conditions and How They Differ section places it among neighbouring conditions. In 73 patients aged 60 or older, early-onset and late-onset depression resembled each other in subtype, severity, phenomenology and previous episodes. Early onset carried more personality abnormalities, more family psychiatric history and more dysfunctional maternal relationships, so the authors concluded that the two forms are similar in presentation but differ in cause. The patients were tertiary referrals, the groups were not age-matched, and the data were self-reported (Brodaty et al., 2001).

Among 50 patients with late-onset major depression, measures of vascular disease did not differ between groups, whereas periventricular hyperintensity was greater in non-responders to treatment than in 35 controls. The authors raised the possibility of subtle cerebrovascular mechanisms (Baldwin et al., 2004).

Cognitive Implications

Neuropsychological performance did not differ between early-onset and late-onset patients (Brodaty et al., 2001). Within late-onset depression, cognition separated patients by treatment response. Non-responders (21) scored below controls on visuospatial, language and executive tests and below responders (29) on fluency and the Stroop test. Responders matched controls on those tests. Both patient groups fell below controls on verbal learning, and the authors suggested that treatment resistance may go with impaired executive function (Baldwin et al., 2004).

Vascular Apathy

Vascular apathy is a neighbouring syndrome. Its premise is that damage to white matter tracks apathy rather than depression. In 118 people with small vessel disease and 398 controls, lower fractional anisotropy was associated with apathy (r = -0.38) but not significantly with depression (r = -0.16, P = 0.09) (Hollocks et al., 2015). A cross-sectional study of 196 people with lacunar stroke found apathy more often than in 300 controls (27.6% against 10.7%), while depression did not differ significantly (12.8% against 9.7%) (Lohner et al., 2017).

A structured review of 14 population studies, 12 of which controlled for depression, found vascular markers associated with apathy. Most studies were cross-sectional and prospective data were scarce, so the direction of the link remains unclear (Wouts et al., 2020). The Related Conditions and How They Differ section sets apathy beside depression.

Cognitive Implications

In small vessel disease, apathy was associated with executive and processing-speed impairment (β = -.353, p < .001), and depression was not (β = .149, p = .118). The study was cross-sectional, and its authors treated causal claims with caution (Lohner et al., 2017). Apathy, and not depression, also predicted later dementia in two prospective cohorts (Tay et al., 2020). The dementia section returns to this finding.

Depression in Vascular Cognitive Impairment and Vascular Dementia

Depression can arise inside vascular cognitive impairment or vascular dementia, which sets it apart from depression that comes before the cognitive disorder. A meta-analysis of 35 studies with 5,805 participants pooled behavioural and psychological symptoms across subtypes. Depression was found in 43.48% overall, and it was more frequent in subcortical vascular cognitive impairment (52.11%) than in unspecified cases. Among people with vascular cognitive impairment without dementia, depression (44.97%) was commoner than irritability or anxiety. The review included cohort and cross-sectional studies, and it called for further research to clarify the underlying neurobiological mechanisms (Santos et al., 2025).

Cognitive Implications

That meta-analysis reports symptom frequencies by subtype of vascular cognitive impairment and no cognitive findings specific to the depressed group (Santos et al., 2025).

Depression in CADASIL

CADASIL is a small vessel disease in which mood disorders have been examined. A cross-sectional structured-interview study of 23 patients found mood disorders to be frequent, and its authors judged that the psychiatric aspects merit attention if confirmed in larger series (Valenti et al., 2011).

Cognitive Implications

No study on this page reports cognitive findings specific to depression in CADASIL.

Is Vascular Depression a Distinct Subtype?

A diagnostic subtype earns its place by two kinds of evidence. External validity asks whether the group differs from other cases in clinical profile, course and treatment response. Internal validity asks whether the data themselves fall into the proposed groups. A position paper on the diagnostic validity of vascular depression argued that external validity presupposes a defined construct, and that the subtype at that point lacked evidence of internal validity (Sneed et al., 2006). The studies below are sorted by whether they support the subtype or press against it.

The Case for a Subtype

Clinical comparisons gave the first support for a distinct profile. In a case-control study of 139 depressed elderly subjects, those with subcortical ischemic depression were older and more often had a history of hypertension and lassitude than non-ischemic cases. They less often reported a family history of mental illness or loss of libido (Krishnan et al., 2004). Age was the strongest predictor of ischemic depression in that analysis, and the authors concluded that the condition may be a distinct late-life syndrome. Cognitive function and treatment implications were left for further study.

A literature review on MRI-defined vascular depression reached a qualified verdict on external validity. Its verdict was that a considerable body of data backs the construct but that critical questions are still open, including the causal link, the diagnostic criteria, the type and location of lesions, and the course the disorder takes (Culang-Reinlieb et al., 2011). The causal relationship was not established, diagnostic criteria were inconsistent, lesion type and location were unresolved, and the course of the disorder was unclear.

Internal validity came from latent class analysis. The 2006 position paper had proposed taxometric and latent class cluster methods as the next step (Sneed et al., 2006). Two years later, investigators applied latent class analysis to two independent samples of late-life depressed patients, and in each one (NCODE: chi(2)(6) = 2.02, p = .90; Old-Old: chi(2)(6) = 7.024, p = .32) a solution with two classes, one vascular and one nonvascular, fitted well (Sneed et al., 2008). Deep white matter lesion burden, along with all the proposed features, predicted class membership. The authors presented this as the first support for internal validity.

The Case Against

Community cohorts did not reproduce the clinical profile. Two community-based studies of older Dutch residents, the Rotterdam Study and AMSTEL, found some differences between depressed people with and without vascular risk, but the specific vascular depression profile went unconfirmed in both (Naarding et al., 2007).

A review of the causal evidence described the link between late-life depression and vascular brain disease as bidirectional. It judged the vascular depression concept a useful framework that was perhaps conceptually too restrictive (Baldwin, 2005). Inflammatory processes, and non-traditional vascular risk factors, were named as possible mediators beyond the classic vascular risk factors.

The sharpest challenge came from postmortem work. In a population-based cohort of 153 non-demented brain donors, 36 had depression and 117 did not. Any vascular pathology was unrelated to depression (P = 0.98), as were infarction, lacunes and small vessel disease. Alzheimer pathology was also unrelated. Hippocampal neuronal loss and subcortical Lewy bodies were more frequent in the depressed donors, and the authors read the findings as implicating neuronal loss rather than vascular or Alzheimer pathology (Tsopelas et al., 2011). The P-values were unadjusted for multiple tests, severe lesions were rare, and few cases had Lewy bodies.

A clinical neuropathology study compared 38 patients with late-onset depression with 29 controls. Lacunes and microvascular lesions did not differ between the groups, and demyelination was not greater in the patients. Of the patients, 25 (66%) met vascular depression criteria, yet that subgroup showed no difference either (Santos et al., 2010). The study used a stricter significance threshold of P < 0.01 because of multiple comparisons, and it noted that autopsy material is hard to obtain. A narrative review with a neuropathology focus summarised the same pattern. Late-onset depression was not associated with lacunes or with periventricular or deep white matter demyelination, whereas subcortical lacunes were linked to post-stroke depression (Xekardaki et al., 2012). That review called the MRI causal link still controversial.

Where the Question Stands

The subtype remains unsettled. A consensus report that screened approximately 4000 articles judged vascular depression a plausible subtype of late-life depression whose evidence is contested (Aizenstein et al., 2016). It noted that no widely accepted diagnostic criteria exist, that the diagnosis is absent from DSM-5, and that the causal relationship remains controversial. Its conclusion was that the subtype is possible and that its pathogenesis needs further study. The question is open: latent class and clinical-profile evidence supports a distinct group, while community cohorts and postmortem studies do not.

The most recent appraisal, a narrative review from 2025, takes a firmer position on the definition. It treats vascular depression as a distinct depression subtype, separate from both late-life depression in general and major depressive disorder (Wu et al., 2025). The same review, however, notes that the condition has no standardized diagnostic criteria and no consensus, that some proposed mechanisms remain undetermined, and that patients often show resistance to conventional antidepressants. Little on the definitional side has therefore been settled since the 2016 report.

Prevalence, Course and Burden

Estimates of how common vascular depression is depend on who was sampled and how the syndrome was defined. A longitudinal community study with a three-year follow-up used a two-stage design: 783 people were surveyed and 122 received a diagnostic evaluation. Vascular depression was defined by MRI, using a modified Fazekas scale. Major depressive disorder of the vascular type was found in 2.39% of the sample and accounted for 56.2% of all major depression. Vascular depression without major depressive disorder was found in 4.24% and accounted for 34.0% of the non-major depressive disorders (Park et al., 2015). The authors noted a low response rate and a low follow-up rate, so the proportions are best read as estimates from a partial sample.

Post-stroke samples give a different kind of figure, because the depression is defined by the stroke rather than by the lesion rating. A meta-analysis of 61 observational studies covering 25 488 people pooled the frequency of depression after stroke at 31% (95% CI 28% to 35%). That did not differ significantly from the 33% of the same group's 2005 review, and frequency varied considerably across studies. The pooled estimate was higher than the 25% (95% CI 16 to 33%) seen at one to five years and the 23% (95% CI 14 to 31%) seen at five years (Hackett & Pickles, 2014). The authors concluded that the frequency had not fallen in practice and that screening and treatment were still needed.

Course and Disability in Late-Life Depression

Reviews of late-life depression in general, which are not specific to the vascular subtype, treat it as having a biopsychosocial aetiology that includes late-onset presentations and effects on quality of life (Blazer, 2003). A later narrative review described late-life depression as multifactorial, with vascular and psychosocial contributors. In that review chronic medical illness, cognitive impairment, arteriosclerosis and psychosocial adversity were each linked to depression, the role of heredity was uncertain, and antidepressants worked in older adults about as well as in younger ones (Alexopoulos, 2005); a later meta-analysis of placebo-controlled trials qualified that reading, finding a lower drug response in late life than in younger adults and no significant advantage over placebo at 65 and older (Tedeschini et al., 2011). These statements concern late-life depression as a whole and are not estimates for the vascular subtype.

Lesion Progression and Outcome

White matter lesion change tracks the course of illness. In a two-year volumetric MRI cohort of 133 adults aged 60 and over with major depression, hyperintensity volume grew by 11.5% in those who achieved sustained remission and by 31.6% in those who did not. Treatment followed a naturalistic somatic algorithm, and whether slowing lesion progression would improve outcome was not tested (Taylor et al., 2003). Faster progression was thus associated with poorer outcomes, though the design cannot show that it caused them. In a four-year prospective population cohort of 1658 adults aged 65 and over, lesion load and depression were strongly related, which the authors judged consistent with the vascular depression hypothesis (Godin et al., 2008).

Executive Dysfunction, Relapse and Response

Executive dysfunction has been linked to the later course. Among 58 remitted elderly patients in a randomized maintenance trial, abnormal initiation and abnormal perseveration each predicted relapse and recurrence. In that sample, outcome was not significantly predicted by memory impairment, disability, medical burden or social support, nor by a history of earlier episodes. The authors treated the finding as needing confirmation (Alexopoulos et al., 2000). In an eight-week citalopram trial of 112 nondemented elderly patients, abnormal initiation and perseveration scores and abnormal Stroop Color-Word performance each went with a poorer response, while other Dementia Rating Scale domains did not (Alexopoulos et al., 2005). Conversion to dementia is covered in the section Vascular Depression and Dementia.

The Vascular Depression Hypothesis

The vascular depression hypothesis is a proposal about causes, and it is the claim on which the whole concept rests. The three subsections below set out what the founding paper proposed, the mechanisms later reviews have added, and how the idea stands after more than two decades of testing.

Predispose, Precipitate, Perpetuate

The founding statement held that cerebrovascular disease may predispose to, precipitate or perpetuate some geriatric depressive syndromes (Alexopoulos et al., 1997a). The wording is deliberately plural. Damage could raise vulnerability before any episode, could trigger an episode, or could keep one going, and the paper did not commit to a single route. It was also restricted to some syndromes, not to late-life depression as a whole. The proposal named vascular risk factors and ischemic lesions as the agents of damage, with prefrontal systems as the structures affected (Alexopoulos et al., 1997a).

A second element of the proposal concerned accumulation. In its hypothesised form, lesions building up past a threshold would lead to depression (Alexopoulos et al., 1997a). This threshold idea was a prediction put forward by the paper, not a measured result. A later appraisal, titled as a view ten years on, revisited the hypothesis (Alexopoulos, 2006); the source is cited here only for its scope.

Proposed Pathways

Reviews since the founding paper have offered several accounts of how vascular disease might produce depression. Taylor et al. (2013) set out a disconnection hypothesis, an inflammatory hypothesis and a hypoperfusion hypothesis as routes by which vascular disease may lead to late-life depression, and they called for tests of the mechanisms and for attention to the treatment implications (Taylor et al., 2013). The same review linked cerebral hyperintensities and executive dysfunction to the clinical picture. Its predictions were that executive dysfunction goes with poorer antidepressant response, and that progression of hyperintensities, or of cognitive deficits, goes with a poorer course of depression (Taylor et al., 2013).

The consensus report by Aizenstein et al. (2016) places white matter hyperintensities and cerebral small vessel disease at the centre of the account, and it discusses frontal–subcortical–limbic networks as the circuits at issue (Aizenstein et al., 2016). Its conclusion was that a vascular subtype is possible but that its pathogenesis needs further study (Aizenstein et al., 2016). The evidence for each pathway is taken up in Biology of Vascular Depression and in Brain Networks and White Matter.

The causal arrow may also run in both directions. A narrative review that was explicitly not systematic found a clear bidirectional link, in which vascular disease predicts depression and depression predicts vascular disease, and it noted that the mechanisms still need clarifying (Thomas et al., 2004). A strictly one-way reading of the hypothesis therefore sits beside evidence for a two-way association.

Figure 1

The Vascular Depression Hypothesis: Three Proposed Roles of Cerebrovascular Damage

Schematic. A box on the left labelled cerebrovascular damage connects by arrows to three boxes in the middle, labelled predispose, precipitate and perpetuate, which connect by arrows to a box on the right labelled late-life depression. Predispose means the damage raises vulnerability before any episode, precipitate means it triggers an episode, and perpetuate means it keeps an episode going. Cerebrovascular damage Predispose raises vulnerability Precipitate triggers an episode Perpetuate keeps an episode going Late-life depression
Note. After Alexopoulos et al. (1997). Original schematic.

Standing

The hypothesis is contested, not settled. The most recent consensus report judged a vascular subtype of late-life depression plausible but its evidence contested, and it listed the absence of widely accepted diagnostic criteria and the absence of the syndrome from DSM-5 among the field's limits (Aizenstein et al., 2016). It also recorded that the causal relationship remains controversial, that white matter hyperintensities are common in the elderly, and that findings are inconsistent once comorbid conditions are considered (Aizenstein et al., 2016).

Two lines of evidence cut against the hypothesis. Recent autopsy findings challenge it (Aizenstein et al., 2016). In the European multicentre LADIS study, baseline white matter hyperintensity severity did not significantly predict depression at three years, although the report described hyperintensity severity as a possible biomarker for late-life depression (Aizenstein et al., 2016). The postmortem and cohort findings are weighed in the section Is Vascular Depression a Distinct Subtype?.

Rival and Complementary Accounts

The vascular account is one of several explanations offered for depression that begins late in life, and its competitors differ in standing. Some are proposals awaiting tests, one is contested, and one has support in population samples but little prospective evidence. Each is set out below with the status the cited work allows.

Depression as a Prodrome of Neurodegeneration

The prodromal account holds that late-life depression is an early sign of dementia, not a vascular syndrome in its own right, and its standing is contested. A narrative review found that depression earlier in life was consistently linked to later dementia, with a more than twofold increase in risk reported; studies of depression arising late in life were mostly positive but conflicting (Byers & Yaffe, 2011). The same review left open whether depression acts as a prodrome, a consequence or a risk factor, and whether treating it alters dementia risk (Byers & Yaffe, 2011). A review of the vascular concept likewise described the relationship to dementia as unclear (Baldwin & O'Brien, 2002).

A narrower proposal, the amyloid account, is a hypothesis and not an established finding. It proposes that amyloid accumulation compromises frontolimbic networks in a subgroup of older adults and may underlie treatment resistance, and its authors listed four supporting findings and presented it as a model that may guide the design of future studies (Mahgoub & Alexopoulos, 2016). A later systematic review and meta-analysis of 25 studies of older adults without dementia pooled amyloid-beta in cerebrospinal fluid, on PET imaging and in plasma. None of the three pooled associations with depression was significant, and the cerebrospinal fluid and PET results were trends only (Twait et al., 2024). In plasma studies, subgroup analyses found an association between amyloid-beta and depression among individuals with cognitive impairment, and the authors suggested that depressive symptoms may mark a prodromal stage of dementia; the evidence was cross-sectional (Twait et al., 2024). The amyloid account therefore remains a proposal, and it need not exclude a vascular contribution.

Inflammation

The inflammation hypothesis of geriatric depression proposes that neuroinflammation accompanying ageing may promote late-life depression. Its standing is proposed. A narrative review gathered evidence for it, including higher interleukin-1 beta in older people with major depression than in controls and a link between interleukin-6 and later depressive symptoms (Alexopoulos & Morimoto, 2011). The same review noted that the markers measured were mostly peripheral, that exploratory studies needed confirmation, that data on antidepressants were limited and that the hypothesis could not be tested in its entirety (Alexopoulos & Morimoto, 2011).

A genetic test of inflammation's causal role exists, though it comes from a general-population sample and not from older adults with depression. In a UK Biobank cohort of up to 144,890 participants, Mendelian randomisation indicated that genetically proxied interleukin-6 pathway activity raised depressive symptoms, whereas genetically proxied C-reactive protein showed the opposite direction of association (Ye et al., 2021). The authors treated the interleukin-6 pathway as a possible risk factor for depression, left the divergence between interleukin-6 and C-reactive protein unexplained and called for experimental studies to establish causality (Ye et al., 2021). Whether the finding extends to late-onset or vascular depression is not addressed by that sample.

Apathy, Not Depression

A third account treats white matter findings as markers of apathy and not of depression; the detail is taken up under Types of Vascular Depression. A structured review of 14 population studies found that vascular markers such as hypertension, cardiovascular disease, white matter hyperintensities and microbleeds were associated with apathy, and 12 of the 14 controlled for depression (Wouts et al., 2020). The review judged the evidence supportive but cross-sectional for the most part, with prospective data still needed (Wouts et al., 2020). In a case-control study of 118 people with small vessel disease and 398 controls, white matter microstructure was related to apathy but not significantly to depression (Hollocks et al., 2015). The vascular apathy hypothesis is thus supported, within those limits.

Reverse Causation

The vascular hypothesis runs from vessel disease to depression, and the reverse direction has its own evidence. A narrative review that was not systematic described a bidirectional link between depression and vascular disease and said the mechanisms needed clarifying (Thomas et al., 2004). Another narrative review reported that depression raises the risk of cardiovascular disease, stroke and diabetes, and that these consequences could partly be due to dysregulations such as systemic inflammation and hypothalamic-pituitary-adrenal hyperactivity (Penninx et al., 2013). A meta-analysis of 28 prospective cohorts with 317,540 participants found that depression predicted total stroke (hazard ratio 1.45), and the hazard ratio was larger for fatal stroke (1.55) than for ischemic stroke (1.25) (Pan et al., 2011).

A bidirectional Mendelian randomization study pointed the same way. Depression may causally raise the risk of coronary artery disease and myocardial infarction, whereas cardiovascular disease showed a null effect on depression; adjustment for lipids and smoking attenuated the estimates, so these factors may mediate the link (Li et al., 2022). These designs concern depression arising across adulthood, not late-onset depression specifically. How such findings bear on the causal question is taken up under Testing Cause and Effect.

Psychological Models

Psychological models of depression, such as cognitive and behavioural accounts, are not applied to late-life or vascular depression in the sources reviewed for this article, so the evidence applying them here is thin.

Executive Dysfunction and Cognition

Cognitive findings in late-life depression centre on the frontal and subcortical systems. The studies below concern geriatric depression in general rather than vascular depression defined by imaging, so each result bears on the vascular account only as far as the proposed mechanism links the two.

Executive Dysfunction

Cognitive problems are common among older adults with depression, and difficulties with executive function stand out among them (Taylor et al., 2013). A case-control study of 40 adults with major depression and 40 healthy comparison adults, half of each group aged 20 to 60 and half 61 or older, found that depressed participants performed worse than healthy ones on executive tasks. Older depressed participants did worse than younger depressed participants, so the executive deficit grew with age. The tasks covered set shifting, problem solving, psychomotor speed and initiation of novel responses. The age-by-depression interaction was not significant for selective or sustained attention (Lockwood et al., 2002).

Initiation and perseveration abnormalities have also been tied to treatment outcome. In a prospective cohort, 49 depressed elderly participants (25 still symptomatic and 24 in remission after 6 weeks of treatment) were compared with 22 psychiatrically normal controls. The symptomatic group scored more abnormally on initiation/perseveration than both the remitted group and the controls, and the remitted group did not differ from controls. Initiation/perseveration, psychomotor retardation and P300 latency together predicted change in response, explaining 58% of the variance. Demographics and treatment intensity did not explain it. The authors framed the result as needing confirmation (Kalayam & Alexopoulos, 1999).

A cross-sectional clinical study of 126 elderly patients with major depression defined a depression-executive dysfunction (DED) syndrome. Patients with it showed more psychomotor retardation and more paranoia than other patients, but fewer vegetative symptoms. Psychomotor retardation and loss of interest each predicted disability, and so did paranoia, independently of executive dysfunction. The study tied the syndrome to frontostriatal dysfunction and suggested that identifying it could aid close follow-up (Alexopoulos et al., 2002). The standing of the DED syndrome is contested.

The proposed mechanism runs from vascular disease to the frontostriatal circuits that support executive control. A narrative review on the vascular depression hypothesis lists executive dysfunction among the predictors of poor antidepressant response, alongside progression of cerebral hyperintensities and of cognitive deficits as predictors of a poor course. The review called for mechanism testing (Taylor et al., 2013).

Processing Speed

Processing speed has been proposed as the deficit from which other impairments follow. In 155 patients with late-life major depression assessed across five cognitive domains, a regression analysis found that slowed processing mediated deficits in the other domains. The authors described it as a core deficit of late-life depression. Greater age, depression severity, less education, race and vascular risk factors were each associated with poorer cognition, whereas age of onset was not. The authors noted that mechanisms need clarifying and that the role of white matter was not examined (Sheline et al., 2006).

A second cross-sectional study compared 100 patients with late-life depression with 40 nondepressed controls. Patients scored below controls in every domain, and more than half fell under the control 10th percentile. Processing speed predicted the other deficits with standardized coefficients between .45 and .80. Vascular disease burden, medical burden, apolipoprotein E genotype, serum anticholinergicity and age at onset were not significant predictors of cognition. The study therefore supports slowing as a central feature but does not link it to measured vascular burden, and its authors judged a frontostriatal role plausible (Butters et al., 2004).

White matter microstructure offers a partial neural correlate for cognition in general, though not a clear one for speed. A cross-sectional case-control study used diffusion tensor imaging with tract-based spatial statistics in 51 depressed adults and 16 non-depressed controls, all aged 59 or older. Depressed participants showed higher radial diffusivity than controls and lower fractional anisotropy in the cingulate and the splenium, with axial diffusivity unchanged. Cingulate and uncinate measures showed positive relations with cognitive scores, while the corpus callosum showed only a weak relation. Across five neuropsychological factor scores, the depressed group scored below controls on episodic memory (p = 0.0228) and executive function (p = 0.0286). Processing speed did not differ significantly from controls (p = 0.0866), nor did language or working memory. The fractional anisotropy effect reached only p < 0.06, the control group was small, and remitters did not differ from non-remitters on the imaging. The authors judged myelin injury the probable cause, with ischemia possible (Mettenburg et al., 2012).

Memory, Persistence and the Contrast with Alzheimer Disease

Whether the deficits are a feature of the depressive state or a stable trait was tested over time. In a longitudinal cohort of 67 depressed adults aged 60 or older and 36 healthy comparison participants, assessed at baseline, 6 and 18 months (with some retested at 4 years), patients scored below comparison participants in all domains. Half of the patients showed generalized cognitive impairment that persisted at 18 months. Current mood, remission status and antidepressant treatment did not significantly explain the deficits, so the authors read them as trait effects rather than state effects. Late onset predicted memory and executive decline. Processing speed was a partial mediator but was insufficient to explain the group differences (Köhler et al., 2010). One earlier study listed remitted patients as needing further study (Butters et al., 2004).

Vascular risk and executive functioning may also act together on mood. In 77 geriatric rehabilitation patients followed for 18 months, those with two or more cerebrovascular risk factors and a low initiation/perseveration score had higher scores on the 15-item Geriatric Depression Scale than those with fewer risk factors and a high score, at baseline and at 18 months. The 6-month contrast was not significant. The authors suggested that executive scores may moderate the link between vascular risk and depression (Mast et al., 2004).

Testing Cause and Effect

The vascular depression hypothesis predicts an order of events: cerebrovascular damage comes first and depression follows. A cross-sectional association cannot test that order, so the case rests on designs that measure the brain before the mood outcome, on pooled estimates across many cohorts, and on genetic designs that approach the direction question from a different side. This section takes each in turn and reports where the results were null or partial.

Prospective Cohorts

Population cohorts that scanned older adults and then followed them give the most direct test. In a four-year prospective cohort of 1658 adults aged 65 and over, a higher baseline white matter lesion volume predicted incident depression, with an odds ratio of 1.3 for each quartile increase (95% CI 1.1-1.7) (Godin et al., 2008). The same cohort pointed the other way as well. Participants with lifetime major depression had a larger baseline lesion volume than the rest, and over follow-up their lesion load increased more (2.1 versus 1.5 cm³, p = .004), which the authors read as consistent with the hypothesis rather than as proof of it (Godin et al., 2008).

The European LADIS cohort, which enrolled 639 older adults aged 65 to 84 across 11 centres, supplies two further tests. At one year, greater baseline white matter change burden predicted higher Geriatric Depression Scale scores (P < 0.01; OR = 1.65, 95% CI 1.37-1.98), and a history of depression predicted later depression as well (Teodorczuk et al., 2007). The result for incident depressive episodes was weaker. Among 526 participants in the final analysis, 85 had an incident episode, and the link to baseline white matter change appeared only as a non-significant trend (OR = 1.63, P = 0.1) (Teodorczuk et al., 2007). The authors flagged several limits: only participants with white matter changes were enrolled, the study may have been underpowered for incident episodes, scale items overlap with dysexecutive symptoms, lesion progression was not assessed, and a two-way relationship could not be excluded (Teodorczuk et al., 2007).

The same LADIS sample was later followed for about three years, and 394 people had a repeat scan. Progression of white matter change was associated with incident depression at year 3 (P = 0.002), which the authors read as support for the vascular depression hypothesis while noting that a direct causal relationship is difficult to identify unambiguously (Firbank et al., 2012). Progression was not related to Geriatric Depression Scale score at three years, and it was not related to depression arising in years 1 to 3. Baseline depression history did not predict how fast lesions progressed (Firbank et al., 2012). Lesions were graded on a rating scale rather than measured by volume, participants dropped out over time, no standard diagnosis of depression was used, and the report itself noted that causality is hard to establish (Firbank et al., 2012).

The Age, Gene/Environment Susceptibility-Reykjavik Study, a cohort of 1,949 dementia-free older adults with mean age 74.6 years, added MRI markers of small vessel disease at baseline and follow-up. During follow-up, 10.1% developed depressive symptoms (a 15-item Geriatric Depression Scale score of 6 or more). Increases in white matter hyperintensity volume, new subcortical infarcts, new Virchow-Robin spaces, lower baseline brain volume and a decline in brain volume were each associated with onset (van Sloten et al., 2015). Only some baseline markers were significant, a qualification that bears on how much weight a single marker can carry (van Sloten et al., 2015).

Lesion progression has also been studied as a predictor of how depression runs its course. In 133 adults aged 60 and over with major depression, followed for two years with volumetric MRI, white matter hyperintensity volume increased by 11.5% in those who achieved and sustained remission and by 31.6% in those who did not (P = .01) (Taylor et al., 2003). Treatment followed a naturalistic algorithm, and whether slowing progression would improve outcomes was not tested (Taylor et al., 2003).

One cohort found that progression tracked apathy rather than depression. In the Lothian Birth Cohort 1936, assessed at mean ages 73, 76 and 79 years (672, 476 and 382 participants at the three visits), apathy was associated with white matter hyperintensity increase independently of age, which also predicted increase. Depression (β = -0.01, p = 0.78), anxiety and subjective memory complaints were not (Clancy et al., 2023). Among the apathy subscales, the executive and emotional ones were significant, while the initiation subscale was not (p = 0.08). Attrition across visits is a caveat, and the authors suggested apathy may mark small vessel disease (Clancy et al., 2023).

Meta-Analyses and Vascular Risk

Pooling the lesion evidence gives a smaller picture than the headline cohorts suggest. A meta-analysis of 19 reports found a weak but significant association between white matter hyperintensities and depression (Wang et al., 2014). Across cross-sectional data (N = 5876), the overall estimate was not significant (OR 1.12, 95% CI 0.96-1.30, p = 0.14). Deep lesions reached significance (OR 1.02, 95% CI 1.00-1.04, p = 0.02), and periventricular lesions fell just short (OR 1.08, p = 0.07). In longitudinal data the pooled analysis of 2015 participants was not significant (OR 1.12, p = 0.13), although the longitudinal white matter hyperintensity analysis (N = 1882) gave OR 1.22 (95% CI 1.06-1.4, p < 0.01) and the longitudinal deep-lesion estimate was not significant (Wang et al., 2014).

Onset age sharpens the signal. A random-effects meta-analysis of 30 eligible studies, drawn from 98 identified, found more white matter hyperintensities in late-life depression than in controls, and more again in late-onset than in early-onset depression, with odds above 4 or a difference of 0.7 to 0.8 standard deviations. Symptom duration and treatment duration showed no measurable effect on lesions, and the review supported a cerebrovascular form of late-onset depression (Herrmann et al., 2008).

A broader meta-analysis of 48 studies (43 600 participants, 9203 of them with depression; 8 studies longitudinal) examined microvascular dysfunction. Higher soluble ICAM-1 (OR 1.58), white matter hyperintensities (OR 1.29), cerebral microbleeds (OR 1.18) and cerebral (micro)infarctions (OR 1.30) were each associated with higher odds of late-life depression, and in the longitudinal studies hyperintensities predicted incident depression (OR 1.19, 95% CI 1.09-1.30). Albuminuria and retinal vessel diameter showed no significant association, and some markers rested on few studies (van Agtmaal et al., 2017).

Imaging markers of small vessel disease behave similarly when pooled. A systematic review spanning 36 studies related white matter hyperintensities, lacunes, microbleeds and cerebral atrophy to stroke, dementia and depression (nine studies, 9,538 individuals, for depression). Perivascular spaces were not pooled because only two studies existed, and the authors judged causality unproven (Rensma et al., 2018).

Risk factors outside the brain give a mixed result. In a meta-analysis of 26 studies, stroke (OR 2.11), cardiovascular disease (OR 1.76) and diabetes (OR 1.51) were clearly associated with late-life depression, and a composite of risk factors gave OR 1.49 (Valkanova & Ebmeier, 2013). Hypertension, dyslipidemia and the Framingham Stroke Risk Score showed no significant association, and smoking was weakly associated (OR 1.35, 95% CI 1.00-1.81). Heterogeneity was moderate to high, and publication bias was possible for the composite scores and diabetes (Valkanova & Ebmeier, 2013).

Genetic Designs

Mendelian randomization uses genetic variants as instruments, and the studies below use genome-wide association summary statistics. They matter because they can be run in both directions.

In the latest of these, a two-sample Mendelian randomization drawing on genome-wide data from 371,184 cases and 978,703 controls, greater genetic liability to depression went with a higher risk of small vessel stroke (odds ratio 1.36; 95% CI 1.14 to 1.62) and of deep intracerebral haemorrhage (odds ratio 1.82; 95% CI 1.14 to 2.89), and with larger white matter hyperintensity volume (beta 0.09; 95% CI 0.02 to 0.16). Cerebral microbleeds and perivascular spaces were not significant (P > 0.05). The authors described a possible causal link and said the mechanisms are unclear (Wang et al., 2025).

A phenome-wide analysis of polygenic risk for depression, with 10,674 discovery and 11,214 replication participants from the UK Biobank imaging sample, tested the same direction. Depression risk scores were associated with white matter microstructure and replicated, and the authors raised the possibility that these white matter differences are a consequence of depression liability (Shen et al., 2020).

A bidirectional analysis of cardiovascular disease and depression found that depression raised the risk of coronary artery disease and myocardial infarction (ORs 1.099 and 1.146), whereas the reverse direction, from cardiovascular disease to depression, was null. Adjusting for lipids and smoking attenuated the forward estimates, and the report noted that these factors may mediate it (Li et al., 2022).

These results need a fence. Their instrument is genetic liability to depression, so they speak to depression leading to vascular or white matter outcomes, the opposite order to the one the hypothesis proposes. They concern coronary disease, small vessel stroke and white matter markers, and nothing in them singles out late-onset depression or the clinical vascular subtype. A null reverse estimate for cardiovascular disease does not settle the direction for cerebral small vessel disease. The authors themselves used hedged wording (possible, possibly) and flagged mediation by lipids and smoking.

Where the Causal Question Stands

Causality is not established. The prospective cohorts give temporal order for lesion burden and later depressive symptoms, but their effects are modest, partial or null in places, and their authors name attrition, rating-scale limits, underpowered outcomes and the impossibility of excluding two-way influence. The meta-analyses differ in the strength of the associations they report, and a systematic review that found strong associations of small vessel disease markers with incident depression made its burden estimate conditional on those associations being causal (Rensma et al., 2018). The genetic results lean toward depression liability preceding small vessel and white matter outcomes, so the direction may run both ways. The consensus report on the Vascular Depression Hypothesis treats vascular depression as a candidate subtype of depression in later life whose causes still need investigation; it described the causal link as disputed, cited recent autopsy findings that challenge the hypothesis, and observed that white matter hyperintensities are common among older people (Aizenstein et al., 2016). The same report noted that in LADIS, hyperintensity severity did not predict depression at three years (Aizenstein et al., 2016). The standing of the hypothesis is therefore contested.

Biology of Vascular Depression

The biological case for vascular depression rests on several strands of evidence that differ in strength: imaging markers of small vessel disease, postmortem tissue, endothelial and microvascular measures, and inflammatory signals. One strand, from neuropathology in the general population, cuts against the others. Each is reported below with its standing, and the null results are given the same weight as the positive ones.

Imaging Markers and Standards

Research on the imaging markers depends on shared vocabulary. An international working group proposed the STRIVE standards, a set of common terms and protocols for describing cerebral small vessel disease on MRI, covering white matter hyperintensities, lacunes, small subcortical infarcts, perivascular spaces, microbleeds and brain atrophy (Wardlaw et al., 2013). The group began from the observation that terminology and protocols varied widely and that this inconsistency hampered progress. It framed the standards as meant for research yet usable in clinical work. This is a position paper that sets definitions, not a study that tests a mechanism.

Postmortem Evidence

The first neuropathological test of an ischemic origin of these lesions in depression comes from a postmortem study that paired MRI with neuropathology in 20 depressed elderly people and 20 elderly controls (Thomas et al., 2002). Ischemia accounted for every deep white matter lesion examined in the depressed group but for under one third of those in controls (P <.001). The difference lay in small lesions; for large lesions the depressed and control groups did not differ. Among the depressed, the dorsolateral prefrontal cortex carried more ischemic change than the anterior cingulate (P = .003) or the occipital cortex (P = .01). The authors read the findings as supporting the vascular depression hypothesis. The sample was small, so the result is best treated as a well-placed pointer, not a settled demonstration.

A population-based neuropathology study gave a contrasting picture. It examined 153 non-demented brains drawn from 456 donations, 36 of them from people with depression, and found no association between depression and any vascular pathology (P = 0.98), infarction (P = 0.51), lacunes (P = 0.33) or small vessel disease (P = 0.46) (Tsopelas et al., 2011). Alzheimer pathology was likewise unrelated. Depression was instead linked to hippocampal neuronal loss and to subcortical Lewy bodies. The authors cautioned that P-values were unadjusted for multiple tests, that Lewy body cases were few, and that severe lesions were rare enough for groups to be combined. The subtype section takes up what this null result means for the diagnosis.

The Vessel Wall

Beyond the white matter, work on the vessel wall itself points the same direction as the imaging literature. In 493 older adults from the Hoorn Study (mean age 68 years), a higher endothelial dysfunction score was independently associated with more depressive symptoms on the Center for Epidemiologic Studies Depression Scale, by 1.9 points (95% CI 0.7 to 3.1) per standard deviation (van Sloten et al., 2014). Low-grade inflammation and oxidative stress, tested in the same models, showed no significant association. A meta-analysis of 48 studies, with 43,600 participants of whom 9,203 had depression, extended the picture to microvascular dysfunction in both peripheral and cerebral forms, each linked to higher odds of late-life depression (van Agtmaal et al., 2017). The odds ratios were 1.58 for soluble ICAM-1 (95% CI 1.28 to 1.96), 1.29 for white matter hyperintensities, 1.18 for cerebral microbleeds and 1.30 for cerebral (micro)infarctions. White matter hyperintensities also predicted incident depression in the longitudinal studies (OR 1.19, 95% CI 1.09 to 1.30), of which there were 8. Albuminuria and retinal vessel diameter were not significantly related to depression, and the authors noted that some markers rested on few studies.

The relation between vessels and mood may run both ways. A narrative review, which was not systematic and did not include every relevant report, described a bidirectional link in which vascular disease predicts depression and depression predicts vascular disease, with the mechanisms still to be clarified (Thomas et al., 2004).

Cerebral Blood Flow and Hypoperfusion

Blood flow is the most direct test of a hypoperfusion pathway, and the one study of it reported here did not find reduced flow. Colloby and colleagues used arterial spin labelling MRI in a cross-sectional case-control design, comparing 38 people with late-life depression and 30 healthy controls (Colloby et al., 2012). White matter blood flow was higher in the depressed group, by 19% (F1,65 = 9.7, P = 0.003). Grey matter flow did not differ between the groups. Within the depressed group, higher white matter flow went with lower scores on both the Montgomery–Åsberg Depression Rating Scale and the Geriatric Depression Scale, so milder symptoms accompanied the higher flow.

Remission status gave a partial pattern. The 21 participants with scores of 10 or below on the Montgomery–Åsberg scale had higher white matter flow than controls (P = 0.001), whereas the 17 with scores of 11 or above did not differ from controls (P = 0.80). The direct contrast between the two depressed subgroups was not significant (P = 0.06). Age at onset, illness duration and antidepressant use were unrelated to flow, and early-onset and late-onset cases did not differ. The authors suggested the raised flow may reflect compensation or treatment response. They flagged that the grey matter comparison may have been underpowered, that the depressed cohort was heterogeneous, that some controls had depressive symptoms, and that striatal flow was not measured. For the hypoperfusion pathway, this means that in this sample white matter flow was raised, not lowered, and grey matter flow was unchanged, so the study offers no support for hypoperfusion; because the design was cross-sectional, the authors' predictor account still needs prospective testing.

Inflammation

Inflammation is the least settled of these strands. Alexopoulos and Morimoto (2011) proposed in a narrative review that neuroinflammation arising with ageing may promote geriatric depression (Alexopoulos & Morimoto, 2011). Several lines of evidence fit the idea: higher interleukin-6 predicted later depressive symptoms, and interleukin-1 receptor antagonist and interleukin-1β tracked symptoms or severity in older adults. Other results were less clear. Early studies of IL-1β, IL-6 and TNF-α in some settings were not significant, and the direction of antidepressant effects on cytokines remained unclear. The review noted that the markers were mostly peripheral, that exploratory studies needed confirmation, and that the hypothesis could not be tested in its entirety. It is therefore a proposal, not an established mechanism.

A recent multicenter study illustrates how these markers are being used, and how far they remain from mechanism. It built a machine-learning model, a light gradient boosting machine, on 602 participants from Wuhan (236 with vascular depression, 366 controls), and the model drew on eight features: white matter hyperintensity score, age, VEGF, interleukin-6, BDNF, TNF-α, lacune count and serotonin level (Zhang et al., 2025). It discriminated cases with an area under the curve of 0.937 internally, which fell to 0.896 in an external sample of 171. The model only marginally exceeded clinician-level performance. As an exploratory prediction tool, it shows that imaging and inflammatory markers carry joint information about the diagnosis. It does not show that inflammation causes the condition.

Lesion Location After Stroke

The location of a lesion has been a further question, most closely studied after stroke. A meta-analysis of 48 eligible reports, drawn from 143 identified, found no support for the lesion-location hypothesis of post-stroke depression: left and right hemisphere lesions did not differ in risk (RR 0.95, 95% CI 0.83 to 1.10), and neither did left anterior lesions against other regions (95% CI 0.87 to 1.62) (Carson et al., 2000). The authors noted that prior evidence conflicted and that some reports drew on the same patients. The types section places this finding among the forms of the condition.

Brain Networks and White Matter

Much of the evidence for a vascular form of late-life depression concerns the brain's connecting tracts rather than any single region. The disconnection hypothesis, one of several mechanisms set out in a review of the vascular depression hypothesis, holds that focal vascular damage and the location of white matter lesions influence neural connectivity that contributes to clinical symptoms (Taylor et al., 2013). Diffusion tensor imaging (DTI) measures fractional anisotropy (FA), an index of how coherently water diffuses along fibres, and so gives a way to test that idea in living patients.

Microstructure in Late-Life Depression

An early case-control DTI study compared seventeen elderly depressed participants with 16 elderly controls and found reduced FA in right superior frontal white matter among the depressed group. The authors tied this region to late-life depression but left the mechanism linking it to clinical outcomes unclear (Taylor et al., 2004). The result bears on the frontal circuit account described later, because the tract in question forms part of the dorsolateral prefrontal white matter.

A later systematic review identified 15 eligible studies, of which nine entered a meta-analysis. Compared with controls, people with late-life depression showed lower FA in the dorsolateral prefrontal cortex (a large difference) and in the uncinate fasciculus (a medium one). No difference was reported for the corpus callosum or the cingulum. The authors read the pattern as frontal and frontal-limbic disruption that probably contributes to the illness. Heterogeneity and publication bias in the dorsolateral prefrontal estimate qualify that reading (Wen et al., 2014).

Circuit Function and Treatment Response

Functional imaging points the same way. In an event-related fMRI study with 13 patients and 13 elderly comparison participants, patients showed reduced dorsolateral prefrontal activity before treatment, and connectivity between that region and the dorsal anterior cingulate was lower than in controls. After 12 weeks of paroxetine, with Hamilton ratings falling from 19.7 to 7.5, right dorsolateral prefrontal activity rose. The connectivity reduction showed no significant change, which led the authors to describe it as a persistent component of the disorder and the activity deficit as an episodic one. The sample was small, group differences in dorsal anterior cingulate activity were not significant, and the authors raised vascular damage as a possible contributor (Aizenstein et al., 2009).

Tract integrity also bears on outcome. In 48 non-demented depressed older adults given open-label escitalopram for 12 weeks, those who did not remit had lower FA than those who did, in frontal and limbic areas and in the neostriatum and midbrain. The authors proposed that low-FA networks may predict a poor response (Alexopoulos et al., 2008). The wider hypothesis literature likewise links executive dysfunction to a poor antidepressant response, and hyperintensity progression to a poor course (Taylor et al., 2013).

Testing the Pathway and Apathy

The proposed route from vascular risk to depressive symptoms through white matter damage has been tested directly. A preregistered longitudinal mediation analysis of 6,964 middle-aged and older UK Biobank participants found that vascular risk factor burden predicted depressive symptoms, but only slightly (total effect = 0.042, 99% CI 0.020 to 0.068). Across 90 models using 15 white matter tracts, no white matter measure mediated the association, and the mean indirect effect was -0.0001. The authors concluded that white matter disconnection was unsupported. Caveats include a relatively healthy, population-based sample, tract-based measures of limited spatial specificity, and short, variable follow-up, so effects may have been underestimated (Blöchl et al., 2023).

Network damage may instead track a specific symptom. In 331 patients with cerebral small vessel disease, apathy, but not depression, was associated with white matter disconnection. Apathy was linked to lower network density and efficiency than in every other group, and path analysis showed network disruption mediating the link between small vessel disease markers and apathy. Apathy was also tied to reduced premotor and cingulate connectivity, so the authors suggested it may reflect damage to networks that initiate action (Tay et al., 2019).

Brain and Cognitive Reserve

Older adults with a similar amount of white matter damage do not all become depressed, and one proposed explanation is that some brains tolerate the damage better. Cognitive reserve is the usual name for that tolerance, and years of education serve as its stand-in measure. The question for the vascular depression hypothesis is whether reserve changes how strongly white matter hyperintensities (WMH) translate into depressive symptoms.

One cross-sectional study addressed this directly. It combined 3T MRI (T2-FLAIR) with neuropsychological testing in 54 patients with late-life depression (LLD) and 38 normal controls, with depressive symptoms scored on the 15-item Geriatric Depression Scale (Lin et al., 2020). The patients had fewer years of education than the controls (7.3 ± 2.5 vs 11.2 ± 4.3) and higher depressive scores. They also scored lower on processing speed, working memory and language, while their memory scores and WMH load did not differ significantly from the controls' (WMH load difference, p = 0.060) (Lin et al., 2020).

The moderation result was specific to the patients. Among those with LLD, education buffered the effect of WMH on depressive symptoms, and it did the same for language performance. In the controls the WMH-by-education interaction for depressive symptoms was not significant. A group-by-education-by-WMH test for depressive symptoms was also positive. The authors' conclusion was that cognitive reserve appears protective in LLD (Lin et al., 2020).

Cognitive Reserve and Lesion Load

Move the lesion load to compare the two illustrative lines.

Lesion load (arbitrary units)Depression severity (illustrative)Less educationMore education

Less education: 14.8. More education: 10. Gap: 4.8.

An illustrative model with invented values, not data. Its shape follows the direction of the education-by-lesion interaction that Lin et al. (2020) reported in 54 older adults with major depression; the published effect sizes are not shown.

This is the only study the article draws on for the claim, so it remains a single finding awaiting replication. The study's own caveats limit it further. The sample was small, and education is only a proxy for reserve. All participants were taking medication, the WMH loads were relatively low, and only WMH were examined, with periventricular and deep lesions not separated (Lin et al., 2020). A cross-sectional design also cannot show that reserve changes the course of the illness.

Measurement Validity: Lesion Ratings and Executive Tests

The vascular depression hypothesis rests on two kinds of measurement, a count of lesions on imaging and a profile of executive performance, and each has limits that bear on how far the hypothesis can be read.

Lesion ratings began as judgements about the appearance of signal on structural scans. A 1.5 T MRI comparison of 12 Alzheimer patients, four with multiinfarct dementia and nine age-matched controls found that periventricular hyperintensity, the halo, was more common in the Alzheimer group than in controls (p = .024), while deep punctate white-matter lesions did not separate the two groups (60%). The authors judged that the halo might have diagnostic value, and the study carries two cautions: its samples were small and the interpretation of such lesions was a subject of controversy (Fazekas et al., 1987). A pattern in a few dozen scans is a thin basis for treating any single lesion type as specific.

Standardised definitions address a different problem, the inconsistency of the vocabulary itself. An international working group produced STRIVE, a set of standard terms for imaging cerebral small vessel disease, covering white matter hyperintensities, lacunes, small subcortical infarcts, perivascular spaces, microbleeds and brain atrophy. The group noted that terminology and protocols varied widely, and that this inconsistency hampered progress. It described the standards as written for research but usable clinically (Wardlaw et al., 2013). Because the document is a position paper, it sets a convention and does not test validity.

A further question is whether a clinician's visual grade tracks a computed lesion volume. A retrospective cross-sectional study of 246 memory clinic patients, paired with a systematic review of 24 studies (13 entering the meta-analysis), compared the Fazekas visual rating scale with volumetric segmentation of white matter hyperintensities (Quantib). The two measures were strongly correlated (ρ = 0.83, p < 0.001). Visual ratings plus demographics explained 6 to 19% of the variance in cognition across six domains, and adding volumetrics explained at most 2% more. The gain was 2% for the MMSE and -1% for executive function, and it was inconsistent across domains. The authors concluded that visual rating performs about as well as volumetrics, with a small volumetric advantage. Cut-offs derived from the sample classified more accurately than published ones. The study was retrospective, drew on a single memory clinic cohort, and its review searched only until 30 September 2022 (van Dijk et al., 2026). For MRI-based definitions, this supports the view that a visual grade captures most of what volume adds for cognition in that clinic population. It does not show that either measure identifies vascular depression, because the sample was a memory clinic cohort and the outcome was cognition.

Executive tests raise the separate question of whether scores predict daily life. In a preliminary cross-sectional study of 105 nondemented elderly patients with a history of depression, deficits in initiation and perseveration, measured on the Mattis Dementia Rating Scale, were associated with the most impairment across nine instrumental activities of daily living. The link ran to shopping, meals, medicine and money, and the combination of initiation and perseveration deficits with depression was associated with shopping and meals. Attention, construction, conceptualisation and memory were linked to some activities. The authors suggested that interventions might target cognition-linked functional deficits (Kiosses & Alexopoulos, 2005).

Depression measurement also needs care after stroke: a mixed-methods study of veterans discharged home after a new stroke, using the Geriatric Depression Scale, interviews and patient records, concluded that poststroke depression was underrecognized and undertreated (Williams et al., 2005).

Vascular Depression and Dementia

Depression in later life and the onset of dementia travel together, and the open question is which way the arrow points. A systematic review and meta-analysis of 23 population-based prospective cohorts of older adults found that late-life depression predicted incident dementia of every type. The pooled estimate was 1.85 (95% CI 1.67-2.04) for all-cause dementia, 1.65 (95% CI 1.42-1.92) for Alzheimer's disease and 2.52 (95% CI 1.77-3.59) for vascular dementia (Diniz et al., 2013). The vascular figure exceeded the Alzheimer figure (P = 0.03), although that contrast drew on only five studies (Diniz et al., 2013). The ordering is what the vascular depression hypothesis would lead a reader to expect, but the small subgroup makes it suggestive rather than settled.

The link to Alzheimer disease has its own meta-analysis. Across 20 studies and 102 172 persons in 8 countries, a history of depression was associated with later Alzheimer disease, with an odds ratio of 2.02 (95% CI, 1.80-2.26) over all studies. Case-control designs (OR = 2.03) and cohort designs (OR = 1.90) agreed closely, and 19 of the 20 studies were positive (Ownby et al., 2006). The authors reported that the effect might be independent of other risk factors, and their caveats are substantial. Only crude odds ratios were pooled, heterogeneity was significant, possible publication bias was noted, and the absolute risk is small. Whether antidepressants protect remains speculative, and, separately, when the analysis was limited to the two prospective cohort studies that adjusted for a personal history of depression, the odds ratio was 1.58, and its confidence interval of 0.83 to 3.02 spans no effect (Ownby et al., 2006).

Three readings of the association compete. Depression may be a risk factor that contributes to dementia, an early prodrome of a dementia already under way, or a consequence of shared disease. A narrative review found a more than twofold rise in dementia risk after depression earlier in life, and a consistent link there, whereas studies of late-life depression were mostly positive but conflicting. It judged that whether depression is a prodrome, a consequence or a risk factor remains unclear, and that whether treating depression changes dementia risk is also unclear (Byers & Yaffe, 2011). The same review discussed vascular disease among several biological pathways that could link the two conditions (Byers & Yaffe, 2011). The prodromal account therefore remains contested.

Biomarker work bears on the prodromal reading without resolving it. A meta-analysis of 25 studies of older adults without dementia pooled cerebrospinal fluid, PET and plasma measures of amyloid-beta, and none of the pooled associations with depression was significant. The CSF and PET results showed only a trend, and subgroup analyses found an association between plasma amyloid-beta and depression among individuals with cognitive impairment. The authors suggested that depressive symptoms may be a dementia prodrome, while noting that the evidence was cross-sectional (Twait et al., 2024).

A single case report illustrates the clinical overlap. A 78-year-old woman with recurrent major depression, subcortical ischemic disease and cognitive deficits was discussed at a case conference, which concluded that antidepressant therapy can work despite such comorbidity (Steffens et al., 2003). A report on one patient cannot show progression from vascular depression to vascular dementia.

Small vessel disease shifts attention from depression to apathy. In two prospective cohorts, SCANS (n = 121) and RUN DMC (n = 352), baseline apathy predicted all-cause dementia (HR 1.49, 95% CI 1.05 to 2.11, in SCANS; HR 1.05, 95% CI 1.01 to 1.09, in RUN DMC), and increasing apathy did so in SCANS (HR 1.53, 95% CI 1.08 to 2.17). In people with small vessel disease, followed in the SCANS (n=121) and RUN DMC (n=352) cohorts, apathy predicted later all-cause dementia while depression did not, and Tay and colleagues took this to mean that apathy could be an early warning of dementia (Tay et al., 2020).

The current position is that late-life depression reliably accompanies a higher chance of later dementia, with the largest pooled estimate for vascular dementia. Why it does so is unresolved. Risk factor, prodrome and shared cause each retain support, and in small vessel disease apathy may carry the predictive signal that depression does not.

Co-occurring Conditions

Vascular depression rarely presents alone. The conditions that accompany it fall into three groups: overt cerebrovascular events, the covert small vessel changes that the hypothesis treats as its substrate, and the vascular and metabolic risk factors that precede both. The evidence runs in more than one direction, and the studies described below do not settle which way the dominant arrow points in any single patient.

Stroke is the clearest case. A meta-analysis of 61 observational studies, covering 25,488 people, pooled the frequency of depression after stroke at 31% (95% CI 28% to 35%), which was statistically indistinguishable from the 33% of the 2005 review (Hackett & Pickles, 2014). The authors read this as no real reduction in frequency over the interval, and as a case for screening and treatment. Frequency varied considerably across the included studies, and the pooled figure fell at longer follow-up: 25% (95% CI 16 to 33%) at one to five years and 23% (95% CI 14 to 31%) at five years, both below the overall estimate (Hackett & Pickles, 2014). A review of risk factors for late-life depression reached a compatible result from the other side. Prior stroke carried an odds ratio of 2.11 (95% CI 1.61 to 2.77), the largest of any vascular exposure it examined (Valkanova & Ebmeier, 2013).

The association also runs from mood to vessel. Across 28 prospective cohort studies with 317,540 participants and 8,478 stroke cases, depressed people had a higher risk of total stroke (HR = 1.45; 95% CI 1.29 to 1.63) than people without depression (Pan et al., 2011). The hazard ratio was higher for fatal stroke (1.55, from eight studies) than for ischemic stroke (1.25, from six studies), and follow-up ranged from 2 to 29 years (Pan et al., 2011). Earlier work on this question had been inconsistent, and the authors evaluated publication bias. Stroke and depression therefore appear to raise each other's likelihood, a bidirectionality that complicates any claim that vascular disease simply causes late-life depression.

Small vessel disease is the covert counterpart. Rensma and colleagues pooled studies in a systematic review and meta-analysis to test whether signs of cerebral small vessel disease visible on MRI forecast later stroke, dementia and depression (Rensma et al., 2018). For incident depression the pooled evidence came from nine studies with 9,538 individuals. White matter hyperintensities, lacunes, microbleeds and cerebral atrophy were each associated with worse outcomes, whereas perivascular spaces rested on only two studies and were not pooled. The authors judged causality unproven while noting a possibly large disease burden (Rensma et al., 2018). The imaging markers themselves are taken up under Biology of Vascular Depression and Brain Networks and White Matter.

Conventional vascular risk factors add a further layer. Across 26 cohort and case-control studies, the review of late-life depression found a composite risk factor score associated with depression (OR = 1.49; 95% CI 1.27 to 1.75), along with diabetes (OR = 1.51) and cardiovascular disease (OR = 1.76) (Valkanova & Ebmeier, 2013). Smoking showed a weak association (OR = 1.35; 95% CI 1.00 to 1.81). Hypertension, dyslipidemia and the Framingham Stroke Risk Score showed no significant association. Heterogeneity was moderate to high, publication bias was possible for the composite score and for diabetes, and the authors called for large longitudinal studies complemented by neuroimaging (Valkanova & Ebmeier, 2013). The null result for hypertension is notable, since hypertension is the exposure the vascular hypothesis is most often assumed to rest on.

A narrative review adds a mechanistic frame that is not specific to late life. It describes depression as predicting later cardiovascular disease, stroke, diabetes and obesity, possibly through biological dysregulation (Penninx et al., 2013). Across the work reviewed, HPA-axis hyperactivity and systemic inflammation tracked depression more consistently, while findings for autonomic dysregulation were less consistent. Atypical depression related more strongly to metabolic syndrome and inflammation, and melancholic depression to hypercortisolemia, so the heterogeneity of depression bears on which biology is found (Penninx et al., 2013).

Two further companions of vascular depression are treated elsewhere. Apathy, a prominent feature of the presentation, is described under Types of Vascular Depression. Cognitive impairment, and its relation to later dementia, is covered in Vascular Depression and Dementia.

Suicide and Self-Harm

Evidence on suicide in vascular depression is indirect. None of the studies discussed below examined suicidal behaviour in a sample defined as having vascular depression, so the two lines of work bear on the topic through neighbouring populations: stroke survivors, and older adults with depression in general.

A meta-analysis of 23 studies, drawn from 4093 articles and covering over 2 million stroke survivors, found that stroke was followed by a higher risk of suicidal behaviour than was seen in people without stroke (Vyas et al., 2021). Altogether 5563 people attempted suicide or died by suicide. The pooled risk ratio (RR) was 1.73 (95% CI 1.53–1.96). Attempts were the stronger signal, with an RR of 2.11 (1.73–2.56) against 1.61 (1.41–1.84) for death by suicide, and the difference between them was significant (P = 0.03). Risk fell with longer follow-up, by an RR of 0.97 (0.95–0.99) per year. The authors concluded that stroke merits recognition as a risk factor for suicide. Three caveats attach to the estimate. The studies were observational only, they were of fair quality, and heterogeneity was high (I2 = 93%). The review also treated depression and suicidal ideation as constructs, but it reports no finding specific to vascular depression.

The second source is a small cross-sectional study of 58 adults aged 65 years and older, which combined brain MRI with the Cognitive Abilities Screening Instrument (Lee et al., 2025). The group with a recent suicide attempt (RSA, n = 17) showed worse cognition and more periventricular white matter hyperintensities (WMH) than either a second group (NS, n = 20) or a healthy comparison group (HC, n = 21). Within the RSA group alone, heavier WMH load went with poorer cognition. The authors read the pattern as pointing to a possible link between cognition, WMH and attempt risk. The study was exploratory, the groups were small, and the design was cross-sectional, so the data cannot show that WMH or cognitive change leads to an attempt. Its subject was late-life depression broadly rather than a vascular subtype.

Anyone at risk can get immediate help from emergency services or a local crisis line.

Culture, Society and Who Is Studied

The evidence on vascular depression comes mostly from older adults, and the samples fall into three kinds. Clinical imaging studies enrolled small groups; the exploratory MRI and cognitive-testing study described under its own heading had 58 participants aged 65 and older (Lee et al., 2025). Community cohorts reached more people but with losses. A two-stage survey of 783 people, 122 of whom had a diagnostic evaluation, was followed for three years, and its authors reported a low response rate and a low follow-up rate, so its participants may not represent the whole community (Park et al., 2015).

Social context shows up in two ways. The first is a cohort study of 662 older adults in three groups, drawn from the Washington Heights-Inwood Columbia Aging Project. It examined socioeconomic status and white matter hyperintensities, and it found that subjective financial worry was associated with greater white matter hyperintensity volume. Education and income were not significantly associated with that volume. The worry association appeared in the Latinx group only. The Latinx group also had the highest worry and the lowest socioeconomic status, while the Black group had more white matter hyperintensities than the Latinx and White groups. The authors framed this as support for the minority poverty hypothesis, and the study's caveat is that traditional socioeconomic measures may misrepresent status. The study reported associations, so it does not show that worry causes the lesions (Morris et al., 2023).

The second is indirect. A qualitative study interviewed 17 family caregivers of stroke survivors and identified four themes, among them spirituality, religious coping and a balance between hope and hardship. Its subjects were caregivers, not people with vascular depression, so it speaks to the social world around vascular disease rather than to the depression (Pourhasan & Vasli, 2025).

The evidence base is narrower than the topic. The studies in this section name only a few settings: the Washington Heights-Inwood Columbia Aging Project, a community survey, a clinical imaging sample and a caregiver study. Only Morris et al. (2023) compares racial and ethnic groups, and none of the other studies reports such a comparison. Whether the findings hold across other countries, groups and care settings is therefore untested here, and the low response and follow-up rates add a further limit even within the community sample.

Treatment Evidence

The evidence on treating vascular depression is thin and, in places, contested. Most of it comes from studies of late-life depression as a whole, in which imaging findings and executive test scores are tested as predictors of who responds. Few trials were designed around vascular depression itself, and those that were come from one group.

Antidepressants and Vascular Burden

Against placebo, antidepressants worked in late-life major depression in a meta-analysis of 15 late-life and 59 adult trials (Tedeschini et al., 2011). Their advantage was significant at age 55 and older (P < .0001) but not at age 65 and older (P = .265). Drug response was lower in late life than in adults, while placebo response did not differ by age. Heterogeneity across studies was significant, and the reason for the lower efficacy remained unclear. A patient-level analysis of 10 placebo-controlled trials in adults aged 60 and older, with 2,283 patients in the seven trials that recorded every moderator, found that the drug-placebo difference depended on illness duration (Nelson et al., 2013). Drug beat placebo where illness had lasted more than 10 years and the Hamilton Depression Rating Scale score was 21 or higher, and the difference was not significant for short illness duration. Response rates were 46.3% on drug and 41.5% on placebo. Only seven of the trials carried every moderator, so the multivariate analyses rested on those seven.

Within this general picture, several studies link the burden of white matter hyperintensities to poorer outcome. In an early MRI study of 39 inpatients with severe depression, greater white matter hyperintensity severity went with poorer treatment response (r = -0.44, p < .01) (Hickie et al., 1995). The correlation held in the pharmacotherapy-only subgroup (n = 19; r = -.49, p < .05) but fell short of significance in the ECT subgroup (n = 20; r = -0.42, p = .06). The authors described the evidence as preliminary.

A 12-week sertraline study of 217 adults aged 60 and older, a two-site, prospective, nonrandomized trial, found that Fazekas ratings predicted depression scores, although not their decrease once baseline scores were controlled (Sheline et al., 2010). Remission (MADRS of 7 or below) occurred in 33%. Remitters differed from non-remitters in cognition and in vascular risk factor scores, and neuropsychological scores correlated negatively with Fazekas ratings of hyperintensity severity. The study excluded participants with cognitive impairment, and its authors read the results as supporting the vascular depression hypothesis.

Deep white matter hyperintensities gave the largest odds ratio for non-remission in a secondary analysis of a 12-week randomized trial in 38 late-life depression patients (Sneed et al., 2011). A high deep white matter burden was associated with non-remission (7.14 times, p = 0.02), whereas periventricular volume was not significantly related to it. Total hyperintensity volume reached p = 0.05. Adjusting for age left these effects unchanged, and the authors concluded that lesion location may not matter. The sample was small and predominantly White.

Treatment resistance has also been linked to executive function in a case-control study of 50 late-onset depression patients and 35 controls, 29 of whom responded to treatment and 21 of whom did not (Baldwin et al., 2004). Non-responders scored below controls on visuospatial, language and executive tests, while responders performed like controls. Non-responders were also below responders on verbal fluency and the Stroop test, and had more periventricular hyperintensity than controls. Vascular disease measures and atrophy did not differ between groups. The authors suggested that treatment-resistant depression in late life may be associated with impaired executive function, and that subtle cerebrovascular mechanisms may be involved.

Executive Dysfunction and Response

A claim that executive dysfunction predicts poor antidepressant response grew from three studies in older patients. In 49 depressed elderly people followed for 6 weeks, with 22 controls, initiation/perseveration scores, psychomotor retardation and P300 latency together explained 58% of the variance in response change (F3 = 20.1, P < .001) (Kalayam & Alexopoulos, 1999). Patients who stayed symptomatic scored worse than those in remission on initiation/perseveration, while remitted patients resembled controls. The authors said the finding needed confirmation. In an 8-week citalopram trial of 112 nondemented elderly patients (target daily dose 40 mg), abnormal initiation/perseveration (scores of 35 or below) and abnormal Stroop Color-Word performance (22 or below) went with less symptom change (Alexopoulos et al., 2005). Other Dementia Rating Scale domains were not significant. A literature review later tied deficits in verbal fluency and response inhibition to worse response (Pimontel et al., 2012). It proposed processing speed as a mediator, and noted that the semantic strategy finding rested on a single study and that the results needed replication.

The pooled evidence is weaker. A meta-analysis of 17 studies and 1,269 individuals pooled the 30 of its 90 cognitive tests and subtests that more than one study had used, and five of them discriminated well between responders and non-responders: one executive test and four non-executive ones (McLennan & Mathias, 2010). The authors concluded that the depression-executive dysfunction model lacked strong support, with the caveat that few studies were available and that age and treatment duration could not be tested statistically. A Bayesian meta-analysis of eight studies and 25 executive measures across six domains, in outpatients aged 50 and older, found a link to non-response only for planning and organization (g = 0.91, 95% CI 0.32 to 1.58) (Pimontel et al., 2016). None of the remaining measures reached significance as predictors of response: not cognitive flexibility, response inhibition, selective attention or verbal fluency, and not the Initiation/Perseveration composite of the Dementia Rating Scale. The planning estimate came from one study and had a wide interval. The studies differed in duration, treatment and measures, and there were too few for moderator analyses.

The current position is that executive dysfunction as a predictor of response remains contested. Early single studies supported it, and the pooled analyses support it only for a few tests, with the strongest estimate resting on one study.

Problem-Solving Therapy

Psychotherapy has been tested in older adults who have depression and executive dysfunction. One trial compared problem-solving therapy with supportive therapy in geriatric major depression with executive dysfunction (Alexopoulos et al., 2003). Only its scope is available here, because no abstract could be retrieved for it.

Two later reports describe a randomized trial of 221 adults aged 60 and older with depression and executive dysfunction, enrolled from 653 referred, with 12 weekly sessions. Problem-solving therapy produced greater symptom reduction than supportive therapy in weeks 9 to 12 and no difference in weeks 1 to 6 (Areán et al., 2010). Response at week 9 was 47.1% against 29.3%, and at week 12 it was 56.7% against 34.0%. Remission at week 9 was 37.9% against 21.7%, and at week 12 it was 45.6% against 27.8%. The authors judged that problem-solving therapy may be an alternative to pharmacotherapy.

The 36-week follow-up report put the groups at 110 and 111, with 201 participants (91%) completing 12 weeks and 10 dropouts in each arm (Alexopoulos et al., 2011). Disability favoured problem-solving therapy in weeks 9 to 12, with equal disability in the first 6 weeks, and the advantage was retained after treatment. The advantage was larger with more cognitive impairment (lower Mini Mental State Examination scores) and more previous episodes. Disability nevertheless increased after treatment ended. Therapists knew the randomization, and ratings relied on clinical judgment.

Psychological treatment of late-life depression has also been compared across many research groups. A network meta-analysis of 86 randomized trials in 10,165 older adults covered five types of treatment: cognitive behavior therapy, behavioral activation, problem-solving therapy, life review therapy and supportive therapy (Cuijpers et al., 2026). The first four each outperformed care as usual (SMD 0.52 to 0.99) and a waiting list (SMD 0.62 to 1.10), whereas supportive therapy did not differ significantly from either. Cognitive behavior therapy, problem-solving therapy and life review therapy beat supportive therapy (SMD 0.43 to 0.80), and behavioral activation did not differ significantly from it. Life review therapy appeared superior to cognitive behavior therapy and behavioral activation, but it lost that advantage once an outlier was removed. The four active therapies are therefore probably comparable. Direct and indirect evidence differed significantly, only treatments with five or more trials were included, and the participants were older adults with depression generally, not vascular depression specifically.

Vasoactive Drugs

Two double-blind trials tested augmentation with the calcium channel blocker nimodipine in vascular depression, and both came from one research group. In the first, 84 patients received nimodipine (n = 40) or vitamin C (n = 44) for up to 300 days (Taragano et al., 2001). Hamilton scores improved more with nimodipine (F(1,81) = 8.64, p = 0.004). Remission was 45% against 25% (p = 0.054) and recurrence 7.4% against 32% (p = 0.058), and both differences were marginal.

The second trial randomized 101 patients to nimodipine (n = 50) or placebo (n = 51) added to fluoxetine for 8 months (Taragano et al., 2005). Nimodipine gave greater Hamilton improvement, full remission of 54% against 27% (p = 0.006; NNT = 4) and recurrence of 3.7% against 35.7% (NNT = 3). Side effects did not differ significantly (p = 0.133). Neither trial reports independent replication, and the first trial's key differences were marginal, so the evidence on these drugs is limited.

Treating Post-Stroke Depression

Depression that is already present after stroke has its own treatment trials, separate from the prophylaxis question taken up under Prevention and Vascular Risk-Factor Management. A meta-analysis of 16 randomized trials in 1,320 patients with post-stroke depression, covering studies from 1984 to 2006, found antidepressants superior to placebo (Chen et al., 2006). Response rates were 65.18% (234 of 359) on active drug and 44.37% (138 of 311) on placebo, a rate difference of 0.23 (95% CI 0.03 to 0.43). Symptom improvement also favoured active treatment and was related to treatment duration (Spearman rho = -0.93, p = 0.001), which led the authors to suggest that longer use may work better. Effects on neurologic recovery and on activities of daily living were inconsistent, with no clear difference from placebo. The authors noted that earlier efficacy data had conflicted.

Psychological treatment was assessed in a network meta-analysis of 38 randomized trials with 3,106 stroke survivors, comparing 11 intervention types across seven databases (Cao et al., 2026). Third-wave therapies, which in this analysis meant mindfulness and acceptance and commitment therapy, ranked best against treatment as usual, both after the intervention and at follow-up, and also exceeded cognitive training, and at follow-up psychoeducation and cognitive behavioural therapy. Baseline severity mattered. In mild post-stroke depression, relaxation therapy, family therapy and cognitive behavioural therapy each beat treatment as usual, and relaxation beat psychoeducation. In moderate to severe depression, no intervention differed significantly from treatment as usual, though only nine such studies were available. The caveats are substantial. Heterogeneity was high (I2 = 83.00%), certainty was low to very low, and the biofeedback ranking came from one trial. Trials came predominantly from China, the UK and Australia, and scores on different scales were converted to the PHQ-9.

Bottom Line

No treatment in the evidence above is confirmed as specific to vascular depression. Antidepressants show a smaller advantage in late life than in adults, imaging and executive measures predict response only inconsistently, and the problem-solving and nimodipine trials each rest on one group's work. A narrative review concluded that the efficacy of several treatments within the syndromes of late-life depression still needs to be studied (Alexopoulos, 2019). Certainty in this area is therefore low.

Medication Safety in Older Adults

Vascular depression is a late-life condition, so the safety of antidepressant classes in older people bears directly on its treatment. The largest relevant evidence is a UK primary care cohort of 60,746 depressed patients aged 65 to 100, drawn from 570 general practices in the QResearch database. Of these, 54,038 (89.0%) received at least one of 1,398,359 antidepressant prescriptions during follow-up (Coupland et al., 2011). The cohort was not specific to vascular depression, and the study reported classes of drug rather than lesion-defined subtypes.

Compared with no antidepressant treatment, selective serotonin reuptake inhibitors (SSRIs) were associated with more falls (hazard ratio 1.66, 1.58 to 1.73) and more hyponatraemia (hazard ratio 1.52, 1.33 to 1.75). A residual group of other antidepressants was associated with five adverse outcomes, among them all-cause mortality (hazard ratio 1.66, 1.56 to 1.77), attempted suicide or self-harm (hazard ratio 5.16, 3.90 to 6.83) and epilepsy or seizures (hazard ratio 2.24, 1.60 to 3.15). One-year mortality was 7.04% in untreated patients, 8.12% with tricyclic antidepressants, 10.61% with SSRIs and 11.43% with the other drugs. Tricyclics did not carry the highest hazard ratio for any outcome, and the classes differed significantly on seven outcomes (Coupland et al., 2011).

The design is observational. Since the cohort was observational, its hazard ratios cannot be read as causal: patients may have been steered toward particular drugs by their indication or by channelling, and residual confounding remains, so Coupland and colleagues asked for more research and for each older patient's risks to be balanced against the expected benefit (Coupland et al., 2011).

A separate hazard concerns SSRIs taken together with oral anticoagulants. A systematic review and random-effects meta-analysis of 14 studies (7 cohort and 7 nested case-control) pooled eight of them, covering 98,070 patients. The pooled hazard ratio for major bleeding was 1.35 (95% CI 1.14 to 1.58) for SSRIs combined with an oral anticoagulant, set against an oral anticoagulant alone. For SSRIs combined with a direct oral anticoagulant specifically, the pooled hazard ratio was 1.47 (95% CI 1.03 to 2.10). The review suggested that physicians may need to tailor SSRI treatment to individual bleeding risk in people who use oral anticoagulants (Rahman et al., 2023). The evidence supplied here does not describe cerebrovascular disease populations, so how far this applies to people with vascular depression is not established by it.

How Treatments Are Thought to Work

Treatment ideas in this field start from the depression-executive dysfunction syndrome and its proposed frontostriatal basis. In a cross-sectional study of 126 elderly patients with major depression, the syndrome showed a distinct clinical profile and was tied to disability; psychomotor retardation, loss of interest and paranoia each predicted disability, and paranoia did so independently of executive dysfunction (Alexopoulos et al., 2002). Because impaired executive function is the target, a therapy that works at the level of problem solving has an obvious rationale, and one trial set problem-solving therapy against supportive therapy in geriatric major depression with executive dysfunction (Alexopoulos et al., 2003).

A later randomized trial gave 12 weekly sessions to 221 adults aged 60 and over who had depression and executive dysfunction. The two arms did not differ over weeks 1 to 6. In weeks 9 to 12, problem-solving therapy produced greater symptom reduction, with response at week 12 of 56.7% against 34.0% and remission of 45.6% against 27.8% (Areán et al., 2010). The authors' conclusion was that it may be an alternative to pharmacotherapy.

The same frontostriatal and cognitive-control account is invoked to explain why antidepressants may help less in this group. A narrative review lists frontostriatal abnormalities and the cognitive control network among the mechanisms and treatment targets of late-life depression (Alexopoulos, 2019). Cognitive remediation follows from the same logic as a training approach; a small comparison of 11 treatment-resistant older adults with 33 matched historical controls found computerized remediation about as effective as escitalopram for depressive symptoms, and better for executive function (Morimoto et al., 2014).

The evidence for mediators is thin. These trials report outcomes; they do not test whether improved executive function carries the benefit of therapy. The 2019 review itself says that the effectiveness of streamlined psychotherapies needs further investigation (Alexopoulos, 2019), and the remediation result rests on historical controls. Treatment accounts remain hypotheses about problem solving and frontostriatal function, not established pathways.

Brain Stimulation and Other Biological Treatments

Two brain-stimulation approaches have been tested in older adults whose depression comes with cerebrovascular change. Evidence for repetitive transcranial magnetic stimulation comes from randomized sham-controlled work, whereas the evidence for electroconvulsive therapy rests on naturalistic cohorts in which white matter lesions were rated on MRI.

Jorge and colleagues reported two randomized sham-controlled experiments of repetitive transcranial magnetic stimulation in 92 patients with clinically defined vascular depression, about half of whom met MRI-based criteria (Jorge et al., 2008). The experiments differed in total cumulative dose, 12,000 pulses in one (TCD-12K) and 18,000 in the other (TCD-18K). Outcomes were scored on the 17-item Hamilton Depression Rating Scale, and a sham coil served as the control. In the lower-dose experiment, active stimulation produced a larger fall in scores than sham (33.1% vs 13.6%, P = .04). The response rate (33.3% vs 6.7%, P = .08) and the remission rate (13.3% vs 6.7%, P = .50) did not differ significantly (Jorge et al., 2008).

The higher-dose experiment gave clearer separation. Scores fell by 42.4% with active stimulation against 17.5% with sham (P < .001). Response reached 39.4% against 6.9% (P = .003), and remission 27.3% against 3.5% (P = .01) (Jorge et al., 2008). Taken together, the trial supports the efficacy of the technique in geriatric vascular depression. Within the same report, older age predicted a poorer response, and greater frontal gray matter volume predicted a better one (Jorge et al., 2008).

For electroconvulsive therapy, the question is whether lesion burden forecasts the outcome. Oudega and colleagues followed 81 elderly depressed patients in a naturalistic, non-randomized cohort who underwent MRI before treatment. Patients with moderate or severe medial temporal lobe atrophy improved less on the Montgomery-Asberg Depression Rating Scale (a 37.9% decrease against 66.2% without such atrophy, P = .008), and patients without atrophy reached remission at about three times the rate (HR = 3.22, 95% CI 1.30 to 7.69, P = .01). White matter hyperintensities and global cortical atrophy showed no differences in response (Oudega et al., 2011). The authors' reading is that medial temporal atrophy, rather than white matter lesions, may be the better predictor of response (Oudega et al., 2011).

Cognitive side effects were examined in a prospective naturalistic cohort of 80 patients with late-life depression. Those with moderate-to-severe white matter hyperintensities scored lower on the Mini Mental State Examination than patients without severe lesions, at every phase of the course (F(1,75.54) = 5.42, p = 0.02). The cognitive trajectory itself did not differ by lesion status, as the time-by-lesion interaction was not significant (F(4,65.85) = 1.9, p = 0.25). Scores fell below baseline during treatment and rose above baseline afterwards, with follow-up lasting 6 months post-ECT. Medial temporal atrophy and global cortical atrophy were not associated with cognitive impairment (Wagenmakers et al., 2021). The study concluded that electroconvulsive therapy remains viable despite severe brain changes, and it described the impairment during treatment as transient (Wagenmakers et al., 2021).

Neither electroconvulsive cohort was randomized, so these results describe who responded and how cognition moved, not what the treatment caused.

Treatments Under Investigation

Three lines of treatment research remain at an early or mixed stage in vascular depression: brain stimulation added to an antidepressant, computerised training of executive skills, and agents aimed at the vessels themselves.

Transcranial direct current stimulation has the more direct evidence. A double-blind, randomised, sham-controlled trial enrolled 93 inpatients over 60 with vascular depression and gave stimulation for 10 consecutive working days alongside sertraline (Zanardi et al., 2020). Depressive symptoms, scored on the 21-item Hamilton depression rating scale, fell with a significant time-by-treatment interaction (F = 14, p < 0.001), and a second interaction appeared for cognitive functioning, which the trial scored with the MODA (F = 3.31, p = 0.04). The two active groups did not respond in the same way early on. At the first assessment one active group scored better than sham (p < 0.001) while the other did not differ from sham. By the second assessment both active groups were ahead of sham, and the difference between one active group and sham was p = 0.005 (Zanardi et al., 2020). On cognitive scores, both stimulation groups beat sham at the second assessment (MODA p < 0.001 in one group and p = 0.007 in the other), and the two did not differ from each other. A schedule with two daily sessions was linked to better Hamilton scores. The authors judged augmentation with stimulation a potentially useful strategy (Zanardi et al., 2020). Standing: early. The result comes from a single trial of inpatients that has not been replicated.

Cognitive remediation delivered by computer has weaker design support. A neuroplasticity-based programme for geriatric depression, nCCR-GD, was given to 11 treatment-resistant older adults and compared with 33 matched historical controls who had received escitalopram (20 mg, over 12 weeks). Of the 11, 91% completed the programme. Depressive symptoms improved about as much as with the drug, and the programme reached that point in 4 weeks instead of 12. Executive function favoured the programme over escitalopram. The authors concluded that remediation may be equally effective as the drug (Morimoto et al., 2014). Standing: early and tentative. The comparison group was historical, not randomised, the sample was small, and the study does not describe its participants as having vascular depression. The finding therefore bears on executive dysfunction in late-life depression, a feature that vascular depression is thought to share (see Executive Dysfunction and Cognition), more than on the subtype as such.

Agents that act on the vessels, such as treatment of vascular risk factors, are covered under Treatment Evidence and under Prevention and Vascular Risk-Factor Management, and are not re-weighed here. Their evidence is set out there, and none of it is borrowed for the two lines above.

Taken together, both lines are better described as candidates than as options with a settled place in care. The stimulation trial is the stronger of the two on design, and the remediation study the weaker, so neither yet shows that its benefit is specific to vascular depression.

Prevention and Vascular Risk-Factor Management

Prevention research bears on the vascular depression hypothesis in two ways. One line of work asks whether drugs that treat vascular risk factors, chiefly high blood pressure, also lower the chance of becoming depressed. The other asks whether depression can be forestalled after a stroke, the clearest vascular insult to the brain. Neither line has produced a firm preventive strategy, and the trials that exist include null results. A third body of work, on statins and exercise, tested the treatment of depression that was already present, so it does not speak to prevention.

Blood-Pressure Drugs and Incident Depression

A large propensity-matched, new-user cohort drawn from the Clinical Practice Research Datalink and Hospital Episodes Statistics compared two commonly prescribed drug classes in hypertensive adults. It included 12,938 matched pairs (mean age 67.6 years, 54.7% women), formed from 36,781 eligible users of ACE inhibitors or angiotensin receptor blockers and 13,052 eligible users of thiazide diuretics. Depression arose at 6.2 per 1000 person years in one arm and 6.6 in the other, a hazard ratio of 0.96 (95% CI 0.79; 1.15) across 467 primary outcome cases (van Sloten et al., 2022). The authors therefore found no protective effect in people with hypertension. The result held when the analysis was limited to treated depression, and the two ACE inhibitor and receptor blocker subclasses did not differ from each other (van Sloten et al., 2022).

Drug lipophilicity, duration of use and daily dose were each unrelated to risk as well (van Sloten et al., 2022). Two limits apply. The self-harm outcome had too few events to analyse, with fewer than five in the control arm, and the estimate rests on observational data (van Sloten et al., 2022).

Preventing Depression After Stroke

A Cochrane review of randomized trials pooled fourteen trials with 1515 participants who had had a stroke. Ten trials tested pharmaceutical agents (12 comparisons) and four tested psychotherapy. Drugs did not differ from placebo in preventing depression, a non-significant result. Psychotherapy outperformed standard care, though the benefit was small (Hackett et al., 2008). The review also examined mood and physical function as outcomes, and it concluded that more evidence was required before any approach could be recommended for routine use (Hackett et al., 2008).

Statins and Exercise: Treatment, Not Prevention

Two meta-analyses addressed other interventions, and both enrolled people who already had depression. Their outcome was symptom severity, not the onset of depression, so they tested treatment rather than prevention of vascular depression.

A meta-analysis of three double-blind randomized trials added a statin to an antidepressant in 165 adults with moderate to severe depression (82 on a statin, 83 on placebo). On the Hamilton Depression Rating Scale, the statin groups improved more than the placebo groups, with a standardized mean difference of -0.73 (95% CI -1.04 to -0.42, p < 0.001), over trials lasting 6 to 12 weeks (Salagre et al., 2016). The authors concluded that adjunctive statins could be useful. Only three trials were included, more randomized trials were needed, and no serious adverse effects were reported (Salagre et al., 2016).

In older adults, a systematic review and meta-analysis of randomized trials covered 9 trials with 667 participants, of which 7 were meta-analysed. Structured, tailored exercise was associated with a small reduction in depression severity compared with control, a standardized mean difference of -0.34 (95% CI -0.52 to -0.17) (Bridle et al., 2012). The estimate was similar when eligibility rested on a clinician diagnosis (-0.38) or on a symptom checklist (-0.34), and it was unchanged when only trials at low risk of bias were analysed. It was larger against an active control (-0.44) and unchanged at extended follow-up (-0.32) (Bridle et al., 2012). The caveats are considerable. Most trials carried a moderate-to-high risk of bias, only published studies were searched, participants may not have been representative, and the evidence on Tai Chi and Qi Gong style exercise was insufficient (Bridle et al., 2012).

Living with Vascular Depression

The consequences of vascular depression reach beyond mood. Evidence on how the condition shapes everyday functioning comes mostly from studies of late-life depression with executive dysfunction and from studies of small vessel disease, and the evidence on families comes from work on stroke caregivers and on dementia. Each body of work is fenced below by what it sampled.

Disability and Daily Function

Executive deficits track the loss of practical independence. In a preliminary cross-sectional clinical study of 105 nondemented elderly patients with a history of depression, deficits on the initiation/perseveration measure of the Mattis Dementia Rating Scale affected more instrumental activities of daily living than any other cognitive domain, interfering with shopping, preparing meals, taking medicine and managing money (Kiosses & Alexopoulos, 2005). Depression severity, measured on the 24-item Hamilton Depression Rating Scale, interacted with initiation/perseveration for shopping and meals. Attention, construction, conceptualization and memory were linked to some activities as well. The authors concluded that interventions might target the functional deficits that accompany these cognitive problems (Kiosses & Alexopoulos, 2005).

Disability is also a treatment outcome, and the randomized trial that measured it in adults with depression and executive dysfunction is reported in Treatment Evidence (Alexopoulos et al., 2011). Apathy likewise bears on functional outcome separately from low mood, and the small vessel disease findings on apathy, white matter and quality of life are reported under Vascular Apathy in Types of Vascular Depression (Hollocks et al., 2015).

Carers and Families

Family members carry part of the load. A qualitative study from Iran interviewed 17 family caregivers of stroke survivors in depth, in semistructured interviews of 30 to 60 minutes between December 2023 and June 2024, and analysed them by conventional content analysis (Pourhasan & Vasli, 2025). Four themes emerged. They covered spirituality and religious coping, psychological coping strategies, social and relational dynamics, and the challenges and struggles of care, with caregivers balancing hope against hardship. The authors called for support programmes that reduce the burden of care (Pourhasan & Vasli, 2025).

A second qualitative study turned to apathy, which is a symptom pattern that carers meet. It interviewed six dyads, each a person with dementia and a carer, using semi-structured interviews and interpretative phenomenological analysis, and it reported three superordinate themes (Chang et al., 2021). Apathy was a burden for the carers, who described guilt, frustration and acceptance. The authors suggested that greater societal awareness may help carer forums. Only the carers' perspectives were reported (Chang et al., 2021).

Both studies concerned other samples and reported themes, not rates. The first addressed stroke survivors in one country and the second people with dementia, and neither addressed vascular depression specifically. They show what caring in these settings can involve and do not measure the burden of caring for someone with vascular depression. The wider concept of caregiver burden frames this strain in general terms.

Vascular depression shares its presentation, its age group and some of its brain findings with several neighbouring conditions. Late-life depression as a whole is multifactorial, with chronic medical illness, cognitive impairment, arteriosclerosis and psychosocial adversity each predicting depression (Alexopoulos, 2005), so no single feature separates one cause from another. Each boundary below therefore gets one rule that a reader can apply.

Apathy Syndrome

The rule is to ask what is being measured. Apathy is a construct in its own right, defined through goal-directed behaviour and scored on the Apathy Evaluation Scale, which exists in clinician, informant and self-rated versions. In a validation study of 123 people aged 53 to 85, the scale proved reliable and construct-valid, and the work set apathy beside depression and anxiety as separate constructs (Marin et al., 1991).

The distinction matters because the two track brain damage differently. In 118 people with small vessel disease and 398 controls, the white matter signal tracked apathy rather than depression: lower fractional anisotropy correlated with apathy at r = -0.38 (P ≤ 0.001) and with depression only at r = -0.16 (P = 0.09), which fell short of significance, and with apathy controlled the link between white matter and depression was not significant (Hollocks et al., 2015). An exploratory analysis of 270 older adults found white matter lacunar volume related to depressed mood, anhedonia, apathy and anergia alike, so the symptoms overlap in practice. Apathy prevalence rose from 1.9% in cognitively normal participants to 16.4% in impaired and 37.6% in demented participants (Lavretsky et al., 2008). A low-drive presentation in a person with white matter disease is thus not automatically depression, and an apathy score is the first discriminator.

Early Alzheimer Disease and Mild Cognitive Impairment

The rule is to ask whether cognitive deficits follow mood or stand apart from it. In a cohort of 67 depressed adults aged 60 or older and 36 controls, patients scored below controls in every cognitive domain, and the deficits were not significantly related to current mood, remission status or antidepressant treatment. The authors read them as trait effects rather than state effects (Köhler et al., 2010). Persistence after mood improves is therefore not, on that evidence, a sign of early dementia in itself.

The link to dementia is real but its meaning is unsettled. Depression earlier in life has been consistently associated with a more than twofold rise in dementia risk, whereas late-life studies mostly agree but conflict. Whether late-life depression is a prodrome, a consequence or a risk factor remains unclear (Byers & Yaffe, 2011). Separating vascular depression from early Alzheimer disease on clinical grounds alone is therefore not supported by the studies above, and none of them compares mild cognitive impairment directly.

Post-Stroke Depression

The rule is to ask what the depression follows. Post-stroke depression is depression arising after a stroke, whereas vascular depression is the broader proposal that cerebrovascular disease, often without a clinical stroke, underlies late-life depression; the Types of Vascular Depression section sets out the forms. The two share biology. Post-stroke depression is common, antidepressants outperformed placebo in reducing its incidence, and white matter disease is a possible contributor (Robinson & Jorge, 2016). The evidence base carries caveats: stroke samples were small, meta-analyses merged diagnostic categories, and rating-scale cutoffs were arbitrary.

Early-Onset Recurrent Depression in Late Life

The rule is to ask when the first episode occurred, because the picture will not say. In 73 patients aged 60 or older drawn from 407 consecutive clinic attenders, those whose depression began before 60 resembled those with later onset in subtype, severity, phenomenology, previous episodes and neuropsychological performance. They differed in aetiology, with more personality abnormalities, more family psychiatric history and more dysfunctional maternal relationships (Brodaty et al., 2001). The sample was of moderate size, not age-matched and drawn from tertiary referrals, and the data were self-report. Late onset, by contrast, predicted memory and executive decline over follow-up (Köhler et al., 2010), which is the pattern the vascular account would anticipate.

Worked Example

Reading a Pooled Frequency

A meta-analysis of 61 observational studies, covering 25 488 people, put the pooled frequency of depression after stroke at 31%, with a 95% confidence interval of 28% to 35% (Hackett & Pickles, 2014). Consider a hypothetical stroke service that discharges 200 patients in a year. Multiplying the pooled figure through (0.31 x 200) gives about 62 patients who might be expected to have depression after their stroke. Applying the two ends of the interval in the same way gives a range of roughly 56 to 70 patients. The service is invented for illustration, and the arithmetic carries a further limit: the pooled value is an average over studies whose frequencies varied considerably, so a real service could sit well outside that range (Hackett & Pickles, 2014). The figure is a starting estimate for planning screening, not a forecast for any one clinic.

Why Shares Differ by Definition and Setting

Two studies that both used brain magnetic resonance imaging to define vascular depression arrive at very different shares. In a clinical comparison of 89 depressed patients, 32 were classified as vascular and 57 as nonvascular, so the vascular group made up 36.0% of the sample (Krishnan et al., 1997). In a longitudinal community study, which surveyed 783 older people and gave a diagnostic evaluation to 122 of them, MRI-defined vascular depression accounted for 56.2% of the cases of major depressive disorder (Park et al., 2015). The two percentages answer different questions. One describes a group of patients already identified as depressed; the other describes cases found in a community sample through a two-stage design, with a low response rate and a low follow-up rate (Park et al., 2015). Each also rests on its own imaging definition of vascular. Neither figure is therefore a prevalence of vascular depression in general, and the gap between them mostly reflects who was sampled and how the subtype was defined.

Discussion

Some findings about vascular depression are reasonably secure. White matter hyperintensities are more frequent in late-life depression than in controls, and more frequent again when depression begins late (Herrmann et al., 2008), although pooled across studies the association between these lesions and depression is significant but weak (Wang et al., 2014). Executive dysfunction is common in older depressed patients, as the section on Executive Dysfunction and Cognition sets out. What remains disputed is what these two observations mean.

The first dispute is whether vascular depression forms a subtype at all. Latent class and clinical-profile evidence point toward a separate group, while community cohorts and postmortem work point away from one (see Is Vascular Depression a Distinct Subtype?). The postmortem challenge is the sharper of the two: in a population-based sample of non-demented brain donors, depression was unrelated to any vascular pathology and was instead linked to neuronal loss (Tsopelas et al., 2011). One review called the concept a useful framework that may be too restrictive conceptually, because mediators beyond the classic vascular risk factors, inflammation among them, may carry the link (Baldwin, 2005). The current position is the consensus report's: a vascular subtype is possible, the evidence is contested and the pathogenesis needs further study (Aizenstein et al., 2016).

The second dispute is the direction of cause. The hypothesis runs from vessel disease to mood, and prospective cohorts do place lesion burden before later depressive symptoms, but the effects are modest or partial and the genetic designs lean the other way, as Testing Cause and Effect reports. The review that set out disconnection, inflammatory and hypoperfusion pathways said that testing them would inform the link between vascular disease and depression and guide the possible repurposing of drugs (Taylor et al., 2013). The question stays open until those mechanisms are tested in designs that can separate the two directions.

The third dispute concerns treatment response. Early single studies tied executive deficits and lesion burden to poorer outcomes, but the most recent pooled analysis linked non-response only to planning and organization, an estimate that rested on one study and carried a wide interval, and found the other executive domains non-significant (Pimontel et al., 2016). The standing of the depression-executive dysfunction syndrome as a predictor of poor response is therefore contested, and the later pooled evidence carries more weight than the single studies that preceded it.

The fourth dispute concerns dementia. Late-life depression predicts later dementia, and the pooled estimate is highest for vascular dementia, although that contrast drew on only five studies (Diniz et al., 2013). Whether depression is a cause, an early sign or a fellow product of the same disease is taken up in Vascular Depression and Dementia, and it remains unresolved.

Several kinds of study bear on these disputes. Prospective imaging cohorts with standard diagnoses can test temporal order, genetic designs run in both directions can test the arrow of cause, and trials that sample enough patients with and without heavy vascular burden can test whether lesions change the response to treatment. The evidence reviewed here contains few trials of the last kind.

Current Directions

Several lines of work are testing the vascular depression hypothesis with designs that earlier decades lacked. Genetic designs are one. A two-sample Mendelian randomization drawing on 371,184 cases and 978,703 controls found that genetic liability to depression predicted a higher risk of small vessel stroke (OR 1.36, 95% CI 1.14-1.62), deep intracerebral hemorrhage (OR 1.82, 95% CI 1.14-2.89) and greater white matter hyperintensity volume, whereas cerebral microbleeds and perivascular spaces showed no significant association (Wang et al., 2025). The authors read this as a possible causal link whose mechanisms remain unclear, and it points the arrow opposite to the one the hypothesis draws. A polygenic risk study with 10,674 discovery and 11,214 replication participants likewise tied depression risk scores to white matter microstructure, a pattern that may be a consequence of depression liability (Shen et al., 2020). The direction of causation is taken up in Testing Cause and Effect.

Large longitudinal cohorts are testing the proposed pathway through white matter. In 6,964 UK Biobank participants, a preregistered mediation analysis found that a greater burden of vascular risk factors predicted more depressive symptoms, though the total effect was small (0.042, 99% CI 0.020-0.068). None of the white matter measures mediated that effect, so the white matter disconnection account went unsupported (Blöchl et al., 2023). The sample was population-based and relatively healthy, and follow-up intervals were short, so the effects may have been underestimated. Brain Networks and White Matter places this result beside the imaging evidence. In the Lothian Birth Cohort 1936, apathy, but not depression, anxiety or subjective memory complaints, was associated with white matter hyperintensity progression over the preceding six years, which suggests that apathy may mark small vessel disease (Clancy et al., 2023). Attrition across visits is a caveat there.

A rival marker is amyloid. A meta-analysis of 25 studies of older adults without dementia found no significant pooled association between depression and amyloid-beta measured in cerebrospinal fluid, by PET or in plasma. The authors proposed that depressive symptoms may be a prodrome of dementia, and the evidence was cross-sectional only (Twait et al., 2024). The question is developed further in Vascular Depression and Dementia.

Two applied lines are also under way. In a double-blind, sham-controlled trial of 93 inpatients over 60 with vascular depression, transcranial direct current stimulation added to sertraline over 10 consecutive working days improved depressive symptoms and cognitive functioning, so the authors judged it a potentially useful augmentation strategy (Zanardi et al., 2020); Brain Stimulation and Other Biological Treatments reviews it with the wider evidence. Prediction models are the other. Among 602 participants (236 with vascular depression and 366 controls), a light gradient boosting machine using eight features, among them white matter hyperintensity score, age, interleukin-6, VEGF, BDNF, TNF-alpha, lacune counts and serotonin level, reached an AUROC of 0.937 and 0.896 in an external sample of 171, only marginally exceeding clinician-level performance (Zhang et al., 2025). Biology of Vascular Depression discusses the markers themselves.

Common Misconceptions

Depression in older adults is a normal part of ageing.
Reviews treat late-life depression as an illness with its own causes, among them vascular and psychosocial contributors, and they report that antidepressants work in older adults about as well as in younger ones (Alexopoulos, 2005). A later meta-analysis of placebo-controlled trials narrowed that claim: drug response was lower in late life than in younger adults, and at 65 and older the advantage over placebo was not significant (Tedeschini et al., 2011). A broad review likewise describes a biopsychosocial aetiology and effects on quality of life (Blazer, 2003). The belief draws on a real observation: chronic medical illness and cognitive impairment, both common in later life, each predict depression, so the mood disorder is easily mistaken for the illness around it.
White matter lesions on a scan mean a person has vascular depression.
White matter hyperintensities are common in elderly people, and the consensus report notes that the causal relationship between such lesions and late-life depression remains controversial (Aizenstein et al., 2016). In a population-based postmortem study, depression was unrelated to vascular pathology of any kind (Tsopelas et al., 2011), and two community cohorts found no specific vascular depression symptom profile, only minor differences (Naarding et al., 2007). The belief rests on the real but weak association between lesions and depression in pooled data, which describes groups, not individual diagnoses.
Vascular depression is an established diagnosis in the diagnostic manuals.
No widely accepted diagnostic criteria exist, and the condition is absent from DSM-5 (Aizenstein et al., 2016). One review judged the concept a useful framework but perhaps too restrictive conceptually (Baldwin, 2005). The label is a research construct; it appears in MeSH as an indexing term, which is not a diagnostic endorsement.
Apathy and depression are the same thing.
Apathy is a construct of its own, defined through goal-directed behaviour and measured reliably and validly with the Apathy Evaluation Scale (Marin et al., 1991). In small vessel disease, reduced white matter integrity was related to apathy but not significantly to depression (Hollocks et al., 2015). The two overlap in practice, since loss of drive and low mood can occur together, which is why they are so often merged.

Glossary

Apathy.
A loss of goal-directed behaviour, rated with instruments such as the Apathy Evaluation Scale in clinician, informant and self-rated versions; in small vessel disease it tracks white matter damage more closely than depression does.
CADASIL.
A cerebral small vessel disease in which a structured-interview study of 23 patients found mood disorders to be frequent, a result its authors said needed confirmation in larger series.
Cerebral small vessel disease.
Disease of the brain's small vessels, recognised on MRI through markers that include white matter hyperintensities, lacunes and microbleeds, for which the STRIVE standards set common terms.
Clinically defined vascular depression.
The definition that identifies vascular depression from clinical features together with vascular risk or vascular disease, introduced in a 1997 comparison of vascular and nonvascular elderly depressed patients.
Cognitive reserve.
The proposed capacity of some brains to tolerate damage, usually indexed by years of education; one cross-sectional study found that education buffered the effect of white matter hyperintensities on depressive symptoms in late-life depression.
Depression-executive dysfunction syndrome.
A proposed late-life syndrome of major depression with executive dysfunction, marked by more psychomotor retardation and paranoia, fewer vegetative symptoms and greater disability; its value as a predictor of poor treatment response is contested.
Diffusion tensor imaging.
An MRI method that measures how water diffuses along nerve fibres, used to estimate fractional anisotropy and so the integrity of white matter tracts.
Executive dysfunction.
Impairment of control processes such as set shifting, problem solving and the initiation of novel responses, a common cognitive problem in late-life depression.
Fazekas scale.
A visual MRI rating of white matter hyperintensities that grew out of a small 1987 comparison of Alzheimer patients and controls; modified versions have been used to define vascular depression by MRI.
Fractional anisotropy.
A diffusion tensor imaging index of how coherently water diffuses along fibres; lower values in frontal and frontal-limbic tracts have been reported in late-life depression.
Frontostriatal circuit.
The pathways linking frontal cortex with the striatum that support executive control, proposed in the vascular account as the circuits that vascular damage disrupts.
Lacune.
A small subcortical lesion and one of the MRI markers of cerebral small vessel disease; postmortem work found lacunes unrelated to late-onset depression but linked to post-stroke depression.
Late-onset depression.
Depression whose first episode begins in later life; it carries more white matter hyperintensities than early-onset depression but a similar clinical picture.
Mendelian randomization.
A design that uses genetic variants as instruments, here drawn from genome-wide association summary statistics, to test the direction of a causal link; applied to depression, it suggested that depression liability may precede vascular outcomes.
MRI-defined vascular depression.
The definition that classes late-life depression as vascular from MRI evidence of subcortical lesions; a review found its validity partly supported, with critical issues unresolved.
Post-stroke depression.
Depression arising after a stroke, a boundary presentation of vascular depression whose pooled frequency across 61 studies was 31%.
Problem-solving therapy.
A structured psychotherapy tested in older adults with depression and executive dysfunction, in whom one randomized trial found it ahead of supportive therapy in weeks 9 to 12 but not earlier.
Processing speed.
The rate at which information is handled; in cross-sectional studies of late-life depression, slowing predicted deficits in other cognitive domains and was described as a core deficit.
Subcortical ischemic depression.
The MeSH entry term for vascular depression and the label used for MRI-defined cases in a case-control study in which older age, a history of hypertension and lassitude predicted the diagnosis.
Vascular apathy.
The hypothesis that vascular damage to the brain tracks apathy rather than depression, supported in population studies that were mostly cross-sectional.
Vascular depression hypothesis.
The 1997 proposal that cerebrovascular disease may raise vulnerability to some geriatric depressive syndromes, trigger them or sustain them; its standing is contested.
White matter hyperintensity.
An area of increased signal in white matter on T2-weighted or FLAIR MRI, common in elderly people and weakly associated with depression in pooled analyses.

Key Researchers

Howard J. Aizenstein

Geriatric psychiatrist at the University of Pittsburgh who studies neuroimaging of late-life depression and led the vascular depression consensus report. ORCID - Faculty Page

George S. Alexopoulos

Psychiatrist at Weill Cornell Medicine who proposed the vascular depression hypothesis and the depression-executive dysfunction syndrome and tested problem-solving therapy in older adults. ORCID - Faculty Page

K. Ranga Rama Krishnan

Psychiatrist at Rush University and professor emeritus at Duke University who proposed MRI-defined vascular depression and described subcortical ischemic depression. ORCID - Faculty Page

John T. O'Brien

Professor of old age psychiatry at the University of Cambridge who studied white matter changes and late-life depression in longitudinal cohorts and in postmortem tissue. ORCID

Yvette I. Sheline

Psychiatrist at the University of Pennsylvania Perelman School of Medicine who tested the vascular depression hypothesis in a two-site antidepressant trial. Faculty Page

Joel R. Sneed

Clinical psychologist at Queens College, City University of New York, who tested the internal validity of the vascular depression subtype. Faculty Page

David C. Steffens

Geriatric psychiatrist and chair of psychiatry at UConn Health who studied the progression of subcortical ischemic disease from vascular depression to vascular dementia. ORCID - Faculty Page

Warren D. Taylor

Psychiatrist and chair of psychiatry at the University of North Carolina at Chapel Hill who studies white matter, vascular disease and late-life depression. Faculty Page

Frequently Asked Questions

What is vascular depression?

Vascular depression is a proposed subtype of late-life depression in which cerebrovascular disease is thought to contribute to the illness. A consensus report judged it a plausible subtype, though the evidence for it is contested and its pathogenesis needs further study (Aizenstein et al., 2016).

How is vascular depression identified?

Researchers have used two routes, one based on clinical features together with vascular risk or disease and one based on MRI evidence of subcortical lesions. In an MRI-based classification of 89 depressed patients, the 32 classed as vascular were older and more often had a late onset than the 57 classed as nonvascular (Krishnan et al., 1997).

Is vascular depression an official diagnosis?

No. The consensus report notes that the condition has no widely accepted diagnostic criteria and is absent from DSM-5, so it remains a research construct (Aizenstein et al., 2016).

Do white matter lesions cause depression?

That has not been shown. A systematic review found MRI markers of cerebral small vessel disease strongly associated with incident depression, and its authors made the burden attributable to small vessel disease conditional on the associations being causal (Rensma et al., 2018). Genetic studies suggest the link may also run from depression to vascular change.

How does vascular depression affect thinking?

Executive function is the domain most often affected in older depressed patients. In one case-control study, depressed adults did worse than healthy adults on executive tasks, and older depressed adults did worse than younger ones (Lockwood et al., 2002). Most such studies concern geriatric depression in general rather than vascular depression defined by imaging.

Does vascular depression respond to treatment?

The claim that vascular burden or executive deficits predict a poor response is contested. A meta-analysis linked non-response only to planning and organization, an estimate drawn from one study, and found the other executive domains non-significant (Pimontel et al., 2016). No treatment is confirmed as specific to vascular depression.

Which treatments for vascular depression have trial evidence?

Few trials were designed around vascular depression itself. In two randomized sham-controlled experiments of repetitive transcranial magnetic stimulation in 92 patients with clinically defined vascular depression, the higher-dose experiment gave clearer benefit, with response of 39.4% against 6.9% on sham (Jorge et al., 2008). A single unreplicated tDCS trial and two nimodipine trials from one group add limited support, so certainty about treatment remains low.

Does vascular depression lead to dementia?

Late-life depression predicted later all-cause dementia, Alzheimer disease and vascular dementia in pooled population cohorts, with the highest estimate for vascular dementia, though that contrast rested on only five studies (Diniz et al., 2013). Whether depression is a risk factor, an early sign or a product of shared disease is unresolved.

How This Article Was Prepared

This is a narrative reference article, not a systematic review. Its sources were identified through a survey of the 25 most-cited works on the subject, the most-cited research of five of its key researchers and two logged searches of OpenAlex, which matched 68 records, of which the 17 most cited were screened by title, on 22 November 2026.

- Considered and excluded: 13 candidate works, each with a recorded reason (for example, superseded by a later review, a single case report, or outside the article's scope). - Included: 121 sources, each confirmed by its DOI against the publisher's or indexer's own record. 33 read in full text, 86 from their abstracts and two cited without a retrievable abstract. Abstracts came from PubMed, Crossref and Scite. - Identifiers: 119 references carry a PubMed ID and 45 a PubMed Central ID, each confirmed by the National Library of Medicine against the cited DOI. - Retractions: 121 sources were checked for retractions and editorial notices; none was retracted. - Publication dates: two from the 1980s, eight from the 1990s, 39 from the 2000s, 46 from the 2010s and 26 from 2020 or later. - Checked against sources: each of the 323 cited sentences was matched to a passage in its source (311 supported by that passage, four resting on a source without a retrievable abstract), and an independent reviewer checked the article against those sources over three passes before publication. - Limits: a narrative review reflects choices about which studies to include, so it can over-represent some lines of work more than a systematic review would. To limit this, each of the field's 25 most-cited works was considered; every logged search and screening decision is recorded; and the reviewer searched each cited source for contrary evidence before accepting a sentence.

See our Editorial Policy for the full method.

References

Aizenstein, H. J., Baskys, A., Boldrini, M., Butters, M. A., Diniz, B. S., Jaiswal, M. K., Jellinger, K. A., Kruglov, L. S., Meshandin, I. A., Mijajlovic, M. D., Niklewski, G., Pospos, S., Raju, K., Richter, K., Steffens, D. C., Taylor, W. D., & Tene, O. (2016). Vascular depression consensus report – a critical update. BMC Medicine, 14, 161. https://doi.org/10.1186/s12916-016-0720-5 PMID: 27806704 · PMCID: PMC5093970

Aizenstein, H. J., Butters, M. A., Wu, M., Mazurkewicz, L. M., Stenger, V. A., Gianaros, P. J., Becker, J. T., Reynolds, C. F., & Carter, C. S. (2009). Altered functioning of the executive control circuit in late-life depression: Episodic and persistent phenomena. American Journal of Geriatric Psychiatry, 17(1), 30-42. https://doi.org/10.1097/jgp.0b013e31817b60af PMID: 19001356 · PMCID: PMC2626170

Alexopoulos, G. S., & Morimoto, S. S. (2011). The inflammation hypothesis in geriatric depression. International Journal of Geriatric Psychiatry, 26(11), 1109-1118. https://doi.org/10.1002/gps.2672 PMID: 21370276 · PMCID: PMC3160498

Alexopoulos, G. S. (2005). Depression in the elderly. The Lancet, 365(9475), 1961-1970. https://doi.org/10.1016/S0140-6736(05)66665-2 PMID: 15936426

Alexopoulos, G. S. (2006). The vascular depression hypothesis: 10 years later. Biological Psychiatry, 60(12), 1304-1305. https://doi.org/10.1016/j.biopsych.2006.09.006 PMID: 17157096

Alexopoulos, G. S. (2019). Mechanisms and treatment of late-life depression. Translational Psychiatry, 9, 188. https://doi.org/10.1038/s41398-019-0514-6 PMID: 31383842 · PMCID: PMC6683149

Alexopoulos, G. S., Kiosses, D. N., Heo, M., Murphy, C. F., Shanmugham, B., & Gunning-Dixon, F. (2005). Executive dysfunction and the course of geriatric depression. Biological Psychiatry, 58(3), 204-210. https://doi.org/10.1016/j.biopsych.2005.04.024 PMID: 16018984

Alexopoulos, G. S., Kiosses, D. N., Klimstra, S., Kalayam, B., & Bruce, M. L. (2002). Clinical presentation of the "depression-executive dysfunction syndrome" of late life. The American Journal of Geriatric Psychiatry, 10(1), 98-106. https://doi.org/10.1176/appi.ajgp.10.1.98

Alexopoulos, G. S., Meyers, B. S., Young, R. C., Campbell, S., Silbersweig, D., & Charlson, M. (1997). 'Vascular depression' hypothesis. Archives of General Psychiatry, 54(10), 915-922. https://doi.org/10.1001/archpsyc.1997.01830220033006 PMID: 9337771

Alexopoulos, G. S., Meyers, B. S., Young, R. C., Kakuma, T., Silbersweig, D., & Charlson, M. (1997). Clinically defined vascular depression. American Journal of Psychiatry, 154(4), 562-565. https://doi.org/10.1176/ajp.154.4.562 PMID: 9090349

Alexopoulos, G. S., Meyers, B. S., Young, R. C., Kalayam, B., Kakuma, T., Gabrielle, M., Sirey, J. A., & Hull, J. (2000). Executive dysfunction and long-term outcomes of geriatric depression. Archives of General Psychiatry, 57(3), 285-290. https://doi.org/10.1001/archpsyc.57.3.285 PMID: 10711915

Alexopoulos, G. S., Murphy, C. F., Gunning-Dixon, F. M., Latoussakis, V., Kanellopoulos, D., Klimstra, S., Lim, K. O., & Hoptman, M. J. (2008). Microstructural white matter abnormalities and remission of geriatric depression. American Journal of Psychiatry, 165(2), 238-244. https://doi.org/10.1176/appi.ajp.2007.07050744 PMID: 18172016

Alexopoulos, G. S., Raue, P., & Areán, P. (2003). Problem-solving therapy versus supportive therapy in geriatric major depression with executive dysfunction. The American Journal of Geriatric Psychiatry, 11(1), 46-52. https://doi.org/10.1097/00019442-200301000-00007

Alexopoulos, G. S., Raue, P. J., Kiosses, D. N., Mackin, R. S., Kanellopoulos, D., McCulloch, C., & Areán, P. A. (2011). Problem-solving therapy and supportive therapy in older adults with major depression and executive dysfunction: Effect on disability. Archives of General Psychiatry, 68(1), 33-41. https://doi.org/10.1001/archgenpsychiatry.2010.177 PMID: 21199963 · PMCID: PMC3018861

Areán, P. A., Raue, P., Mackin, R. S., Kanellopoulos, D., McCulloch, C., & Alexopoulos, G. S. (2010). Problem-solving therapy and supportive therapy in older adults with major depression and executive dysfunction. American Journal of Psychiatry, 167(11), 1391-1398. https://doi.org/10.1176/appi.ajp.2010.09091327 PMID: 20516155 · PMCID: PMC2998516

Baldwin, R. C., & O'Brien, J. (2002). Vascular basis of late-onset depressive disorder. British Journal of Psychiatry, 180(2), 157-160. https://doi.org/10.1192/bjp.180.2.157 PMID: 11823328

Baldwin, R. C. (2005). Is vascular depression a distinct sub-type of depressive disorder? A review of causal evidence. International Journal of Geriatric Psychiatry, 20(1), 1-11. https://doi.org/10.1002/gps.1255 PMID: 15578670

Baldwin, R., Jeffries, S., Jackson, A., Sutcliffe, C., Thacker, N., Scott, M., & Burns, A. (2004). Treatment response in late-onset depression: relationship to neuropsychological, neuroradiological and vascular risk factors. Psychological Medicine, 34(1), 125-136. https://doi.org/10.1017/s0033291703008870 PMID: 14971633

Blazer, D. G. (2003). Depression in late life: Review and commentary. The Journals of Gerontology: Series A, 58(3), M249-M265. https://doi.org/10.1093/gerona/58.3.M249 PMID: 12634292

Blöchl, M., Schaare, H. L., Kumral, D., Gaebler, M., Nestler, S., & Villringer, A. (2023). Vascular risk factors, white matter microstructure, and depressive symptoms: A longitudinal analysis in the UK Biobank. Psychological Medicine, 54(1), 125-135. https://doi.org/10.1017/S0033291723000697 PMID: 37016768

Bridle, C., Spanjers, K., Patel, S., Atherton, N. M., & Lamb, S. E. (2012). Effect of exercise on depression severity in older people: Systematic review and meta-analysis of randomised controlled trials. The British Journal of Psychiatry, 201(3), 180-185. https://doi.org/10.1192/bjp.bp.111.095174 PMID: 22945926

Brodaty, H., Luscombe, G., Parker, G., Wilhelm, K., Hickie, I., Austin, M. P., & Mitchell, P. (2001). Early and late onset depression in old age: different aetiologies, same phenomenology. Journal of Affective Disorders, 66(2-3), 225-236. https://doi.org/10.1016/s0165-0327(00)00317-7 PMID: 11578676

Butters, M. A., Whyte, E. M., Nebes, R. D., Begley, A. E., Dew, M. A., Mulsant, B. H., Zmuda, M. D., Bhalla, R., Meltzer, C. C., Pollock, B. G., Reynolds, C. F., & Becker, J. T. (2004). The nature and determinants of neuropsychological functioning in late-life depression. Archives of General Psychiatry, 61(6), 587-595. https://doi.org/10.1001/archpsyc.61.6.587 PMID: 15184238

Byers, A. L., & Yaffe, K. (2011). Depression and risk of developing dementia. Nature Reviews Neurology, 7(6), 323-331. https://doi.org/10.1038/nrneurol.2011.60 PMID: 21537355 · PMCID: PMC3327554

Cao, X., Chen, R., Qu, Y., Ban, Y., & Zhang, Q. (2026). Ranking psychological interventions for post-stroke depression: A systematic review and network meta-analysis. Psychological Medicine, 56, e230. https://doi.org/10.1017/s0033291726105273 PMID: 42478229 · PMCID: PMC13434865

Carson, A. J., MacHale, S., Allen, K., Lawrie, S. M., Dennis, M., House, A., & Sharpe, M. (2000). Depression after stroke and lesion location: a systematic review. The Lancet, 356(9224), 122-126. https://doi.org/10.1016/S0140-6736(00)02448-X PMID: 10963248

Chang, C. Y. M., Baber, W., Dening, T., & Yates, J. (2021). "He just doesn't want to get out of the chair and do it": The impact of apathy in people with dementia on their carers. International Journal of Environmental Research and Public Health, 18(12), 6317. https://doi.org/10.3390/ijerph18126317 PMID: 34207955 · PMCID: PMC8296153

Chen, Y., Guo, J. J., Zhan, S., & Patel, N. C. (2006). Treatment effects of antidepressants in patients with post-stroke depression: A meta-analysis. Annals of Pharmacotherapy, 40(12), 2115-2122. https://doi.org/10.1345/aph.1h389 PMID: 17119102

Clancy, U., Radakovic, R., Doubal, F., Valdés Hernández, M. del C., Muñoz Maniega, S., Taylor, A. M., Corley, J., Chappell, F. M., Russ, T. C., Cox, S. R., Bastin, M. E., Deary, I. J., & Wardlaw, J. M. (2023). Are neuropsychiatric symptoms a marker of small vessel disease progression in older adults? Evidence from the Lothian Birth Cohort 1936. International Journal of Geriatric Psychiatry, 38(1), e5855. https://doi.org/10.1002/gps.5855 PMID: 36490272 · PMCID: PMC10108049

Coffey, C. E., Figiel, G. S., Djang, W. T., & Weiner, R. D. (1990). Subcortical hyperintensity on magnetic resonance imaging: A comparison of normal and depressed elderly subjects. American Journal of Psychiatry, 147(2), 187-189. https://doi.org/10.1176/ajp.147.2.187 PMID: 2301657

Colloby, S. J., Firbank, M. J., He, J., Thomas, A. J., Vasudev, A., Parry, S. W., & O'Brien, J. T. (2012). Regional cerebral blood flow in late-life depression: Arterial spin labelling magnetic resonance study. The British Journal of Psychiatry, 200(2), 150-155. https://doi.org/10.1192/bjp.bp.111.092387 PMID: 22194184

Coupland, C., Dhiman, P., Morriss, R., Arthur, A., Barton, G., & Hippisley-Cox, J. (2011). Antidepressant use and risk of adverse outcomes in older people: population based cohort study. BMJ, 343, d4551. https://doi.org/10.1136/bmj.d4551 PMID: 21810886 · PMCID: PMC3149102

Cuijpers, P., Miguel, C., Karyotaki, E., & Harrer, M. (2026). Psychological treatment of depression in older adults: A network meta-analysis. International Psychogeriatrics, 38(5), 100213. https://doi.org/10.1016/j.inpsyc.2026.100213 PMID: 41991348

Culang-Reinlieb, M. E., Johnert, L. C., Brickman, A. M., Steffens, D. C., Garcon, E., & Sneed, J. R. (2011). MRI-defined vascular depression: A review of the construct. International Journal of Geriatric Psychiatry, 26(11), 1101-1108. https://doi.org/10.1002/gps.2668 PMID: 21192018

Diniz, B. S., Butters, M. A., Albert, S. M., Dew, M. A., & Reynolds, C. F., III. (2013). Late-life depression and risk of vascular dementia and Alzheimer's disease: Systematic review and meta-analysis of community-based cohort studies. The British Journal of Psychiatry, 202(5), 329-335. https://doi.org/10.1192/bjp.bp.112.118307 PMID: 23637108 · PMCID: PMC3640214

Fazekas, F., Chawluk, J. B., Alavi, A., Hurtig, H. I., & Zimmerman, R. A. (1987). MR signal abnormalities at 1.5 T in Alzheimer's dementia and normal aging. American Journal of Roentgenology, 149(2), 351-356. https://doi.org/10.2214/ajr.149.2.351 PMID: 3496763

Firbank, M. J., Teodorczuk, A., van der Flier, W. M., Gouw, A. A., Wallin, A., Erkinjuntti, T., Inzitari, D., Wahlund, L.-O., Pantoni, L., Poggesi, A., Pracucci, G., Langhorne, P., & O'Brien, J. T. (2012). Relationship between progression of brain white matter changes and late-life depression: 3-year results from the LADIS study. The British Journal of Psychiatry, 201(1), 40-45. https://doi.org/10.1192/bjp.bp.111.098897 PMID: 22626634

Godin, O., Dufouil, C., Maillard, P., Delcroix, N., Mazoyer, B., Crivello, F., Alpérovitch, A., & Tzourio, C. (2008). White matter lesions as a predictor of depression in the elderly: The 3C-Dijon study. Biological Psychiatry, 63(7), 663-669. https://doi.org/10.1016/j.biopsych.2007.09.006 PMID: 17977521

Hackett, M. L., & Pickles, K. (2014). Part I: Frequency of depression after stroke: An updated systematic review and meta-analysis of observational studies. International Journal of Stroke, 9(8), 1017-1025. https://doi.org/10.1111/ijs.12357 PMID: 25117911

Hackett, M. L., Anderson, C. S., House, A., & Halteh, C. (2008). Interventions for preventing depression after stroke. Cochrane Database of Systematic Reviews, (3), CD003689. https://doi.org/10.1002/14651858.CD003689.pub3 PMID: 18646094

Herrmann, L. L., Le Masurier, M., & Ebmeier, K. P. (2008). White matter hyperintensities in late life depression: a systematic review. Journal of Neurology, Neurosurgery & Psychiatry, 79(6), 619-624. https://doi.org/10.1136/jnnp.2007.124651 PMID: 17717021

Hickie, I., Scott, E., Mitchell, P., Wilhelm, K., Austin, M. P., & Bennett, B. (1995). Subcortical hyperintensities on magnetic resonance imaging: Clinical correlates and prognostic significance in patients with severe depression. Biological Psychiatry, 37(3), 151-160. https://doi.org/10.1016/0006-3223(94)00174-2 PMID: 7727623

Hollocks, M. J., Lawrence, A. J., Brookes, R. L., Barrick, T. R., Morris, R. G., Husain, M., & Markus, H. S. (2015). Differential relationships between apathy and depression with white matter microstructural changes and functional outcomes. Brain, 138(12), 3803-3815. https://doi.org/10.1093/brain/awv304 PMID: 26490330 · PMCID: PMC4655344

Jorge, R. E., Moser, D. J., Acion, L., & Robinson, R. G. (2008). Treatment of vascular depression using repetitive transcranial magnetic stimulation. Archives of General Psychiatry, 65(3), 268-276. https://doi.org/10.1001/archgenpsychiatry.2007.45 PMID: 18316673

Kalayam, B., & Alexopoulos, G. S. (1999). Prefrontal dysfunction and treatment response in geriatric depression. Archives of General Psychiatry, 56(8), 713-718. https://doi.org/10.1001/archpsyc.56.8.713 PMID: 10435605

Kiosses, D. N., & Alexopoulos, G. S. (2005). IADL functions, cognitive deficits, and severity of depression: A preliminary study. The American Journal of Geriatric Psychiatry, 13(3), 244-249. https://doi.org/10.1176/appi.ajgp.13.3.244 PMID: 15728756

Köhler, S., Thomas, A. J., Barnett, N. A., & O'Brien, J. T. (2010). The pattern and course of cognitive impairment in late-life depression. Psychological Medicine, 40(4), 591-602. https://doi.org/10.1017/S0033291709990833 PMID: 19656429

Krishnan, K. R. R., & McDonald, W. M. (1995). Arteriosclerotic depression. Medical Hypotheses, 44(2), 111-115. https://doi.org/10.1016/0306-9877(95)90081-0 PMID: 7596303

Krishnan, K. R. R., Hays, J. C., & Blazer, D. G. (1997). MRI-defined vascular depression. American Journal of Psychiatry, 154(4), 497-501. https://doi.org/10.1176/ajp.154.4.497 PMID: 9090336

Krishnan, K. R. R., Taylor, W. D., McQuoid, D. R., MacFall, J. R., Payne, M. E., Provenzale, J. M., & Steffens, D. C. (2004). Clinical characteristics of magnetic resonance imaging-defined subcortical ischemic depression. Biological Psychiatry, 55(4), 390-397. https://doi.org/10.1016/j.biopsych.2003.08.014 PMID: 14960292

Lavretsky, H., Zheng, L., Weiner, M. W., Mungas, D., Reed, B., Kramer, J. H., Jagust, W., Chui, H., & Mack, W. J. (2008). The MRI brain correlates of depressed mood, anhedonia, apathy, and anergia in older adults with and without cognitive impairment or dementia. International Journal of Geriatric Psychiatry, 23(10), 1040-1050. https://doi.org/10.1002/gps.2030 PMID: 18412291 · PMCID: PMC2575050

Lee, Y.-T., Huang, L.-K., Sajatovic, M., Hu, C.-J., Lee, H.-C., Lin, C.-Y., Wang, C.-A., Tsai, S.-Y., & Lin, Y.-K. (2025). Cognition, white matter hyperintensities and suicide risk in late-life depression patients: an exploratory study. BMC Geriatrics, 25, 686. https://doi.org/10.1186/s12877-025-06358-x PMID: 40903731 · PMCID: PMC12409947

Li, G. H.-Y., Cheung, C.-L., Chung, A. K.-K., Cheung, B. M.-Y., Wong, I. C.-K., Fok, M. L.-Y., Au, P. C.-M., & Sham, P.-C. (2022). Evaluation of bi-directional causal association between depression and cardiovascular diseases: a Mendelian randomization study. Psychological Medicine, 52(9), 1765-1776. https://doi.org/10.1017/s0033291720003566 PMID: 33032663

Lin, C., Huang, C.-M., Fan, Y.-T., Liu, H.-L., Chen, Y.-L., Aizenstein, H. J., Lee, T. M.-C., & Lee, S.-H. (2020). Cognitive reserve moderates effects of white matter hyperintensity on depressive symptoms and cognitive function in late-life depression. Frontiers in Psychiatry, 11, Article 249. https://doi.org/10.3389/fpsyt.2020.00249 PMID: 32322221 · PMCID: PMC7158948

Lockwood, K. A., Alexopoulos, G. S., & van Gorp, W. G. (2002). Executive dysfunction in geriatric depression. American Journal of Psychiatry, 159(7), 1119-1126. https://doi.org/10.1176/appi.ajp.159.7.1119 PMID: 12091189

Lohner, V., Brookes, R. L., Hollocks, M. J., Morris, R. G., & Markus, H. S. (2017). Apathy, but not depression, is associated with executive dysfunction in cerebral small vessel disease. PLOS ONE, 12(5), e0176943. https://doi.org/10.1371/journal.pone.0176943 PMID: 28493898 · PMCID: PMC5426624

Mahgoub, N., & Alexopoulos, G. S. (2016). Amyloid hypothesis: Is there a role for antiamyloid treatment in late-life depression? The American Journal of Geriatric Psychiatry, 24(3), 239-247. https://doi.org/10.1016/j.jagp.2015.12.003 PMID: 26946981 · PMCID: PMC4801691

Marin, R. S., Biedrzycki, R. C., & Firinciogullari, S. (1991). Reliability and validity of the Apathy Evaluation Scale. Psychiatry Research, 38(2), 143-162. https://doi.org/10.1016/0165-1781(91)90040-v PMID: 1754629

Mast, B. T., Yochim, B., MacNeill, S. E., & Lichtenberg, P. A. (2004). Risk factors for geriatric depression: The importance of executive functioning within the vascular depression hypothesis. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 59(12), 1290-1294. https://doi.org/10.1093/gerona/59.12.1290 PMID: 15699528

McLennan, S. N., & Mathias, J. L. (2010). The depression-executive dysfunction (DED) syndrome and response to antidepressants: A meta-analytic review. International Journal of Geriatric Psychiatry, 25(10), 933-944. https://doi.org/10.1002/gps.2431 PMID: 20872927

Mettenburg, J. M., Benzinger, T. L., Shimony, J. S., Snyder, A. Z., & Sheline, Y. I. (2012). Diminished performance on neuropsychological testing in late life depression is correlated with microstructural white matter abnormalities. NeuroImage, 60(4), 2182-2190. https://doi.org/10.1016/j.neuroimage.2012.02.044 PMID: 22487548 · PMCID: PMC3337766

Morimoto, S. S., Wexler, B. E., Liu, J., Hu, W., Seirup, J., & Alexopoulos, G. S. (2014). Neuroplasticity-based computerized cognitive remediation for treatment-resistant geriatric depression. Nature Communications, 5, Article 4579. https://doi.org/10.1038/ncomms5579 PMID: 25093396 · PMCID: PMC4139707

Morris, E. P., Turney, I. C., Palms, J. D., Zaheed, A. B., Sol, K., Amarante, E., Beato, J., Chesebro, A. G., Morales, C. D., Manly, J. J., Brickman, A. M., & Zahodne, L. B. (2023). Racial and ethnic differences in the relationship between financial worry and white matter hyperintensities in Latinx, non-Latinx Black, and non-Latinx White older adults. Neurobiology of Aging, 129, 149-156. https://doi.org/10.1016/j.neurobiolaging.2023.05.008 PMID: 37331245 · PMCID: PMC10878173

Naarding, P., Tiemeier, H., Breteler, M. M. B., Schoevers, R. A., Jonker, C., Koudstaal, P. J., & Beekman, A. T. F. (2007). Clinically defined vascular depression in the general population. Psychological Medicine, 37(3), 383-392. https://doi.org/10.1017/s0033291706009196 PMID: 17076919

Nelson, J. C., Delucchi, K. L., & Schneider, L. S. (2013). Moderators of outcome in late-life depression: A patient-level meta-analysis. American Journal of Psychiatry, 170(6), 651-659. https://doi.org/10.1176/appi.ajp.2012.12070927 PMID: 23598969

Oudega, M. L., van Exel, E., Wattjes, M. P., Comijs, H. C., Scheltens, P., Barkhof, F., Eikelenboom, P., de Craen, A. J. M., Beekman, A. T. F., & Stek, M. L. (2011). White matter hyperintensities, medial temporal lobe atrophy, cortical atrophy, and response to electroconvulsive therapy in severely depressed elderly patients. The Journal of Clinical Psychiatry, 72(1), 104-112. https://doi.org/10.4088/jcp.08m04989blu PMID: 20816035

Ownby, R. L., Crocco, E., Acevedo, A., John, V., & Loewenstein, D. (2006). Depression and risk for Alzheimer disease: Systematic review, meta-analysis, and metaregression analysis. Archives of General Psychiatry, 63(5), 530-538. https://doi.org/10.1001/archpsyc.63.5.530 PMID: 16651510 · PMCID: PMC3530614

Pan, A., Sun, Q., Okereke, O. I., Rexrode, K. M., & Hu, F. B. (2011). Depression and risk of stroke morbidity and mortality: A meta-analysis and systematic review. JAMA, 306(11), 1241-1249. https://doi.org/10.1001/jama.2011.1282 PMID: 21934057 · PMCID: PMC3242806

Park, J. H., Lee, S. B., Lee, J. J., Yoon, J. C., Han, J. W., Kim, T. H., Jeong, H.-G., Newhouse, P. A., Taylor, W. D., Kim, J. H., Woo, J. I., & Kim, K. W. (2015). Epidemiology of MRI-defined vascular depression: A longitudinal, community-based study in Korean elders. Journal of Affective Disorders, 180, 200-206. https://doi.org/10.1016/j.jad.2015.04.008 PMID: 25913805

Penninx, B. W. J. H., Milaneschi, Y., Lamers, F., & Vogelzangs, N. (2013). Understanding the somatic consequences of depression: biological mechanisms and the role of depression symptom profile. BMC Medicine, 11, 129. https://doi.org/10.1186/1741-7015-11-129 PMID: 23672628 · PMCID: PMC3661358

Pimontel, M. A., Culang-Reinlieb, M. E., Morimoto, S. S., & Sneed, J. R. (2012). Executive dysfunction and treatment response in late-life depression. International Journal of Geriatric Psychiatry, 27(9), 893-899. https://doi.org/10.1002/gps.2808 PMID: 22009869

Pimontel, M. A., Rindskopf, D., Rutherford, B. R., Brown, P. J., Roose, S. P., & Sneed, J. R. (2016). A meta-analysis of executive dysfunction and antidepressant treatment response in late-life depression. The American Journal of Geriatric Psychiatry, 24(1), 31-41. https://doi.org/10.1016/j.jagp.2015.05.010 PMID: 26282222 · PMCID: PMC4928373

Poole, L., Frost, R., Rowlands, H., & Black, G. (2022). Experience of depression in older adults with and without a physical long-term condition: findings from a qualitative interview study. BMJ Open, 12(2), e056566. https://doi.org/10.1136/bmjopen-2021-056566 PMID: 35210344 · PMCID: PMC8883274

Pourhasan, K., & Vasli, P. (2025). Explaining the multifaceted experiences of family caregivers of stroke survivors: a qualitative study in Iran. BMJ Open, 15(5), e097463. https://doi.org/10.1136/bmjopen-2024-097463 PMID: 40425250 · PMCID: PMC12107558

Rahman, A. A., He, N., Rej, S., Platt, R. W., & Renoux, C. (2023). Concomitant use of selective serotonin reuptake inhibitors and oral anticoagulants and risk of major bleeding: A systematic review and meta-analysis. Thrombosis and Haemostasis, 123(1), 54-63. https://doi.org/10.1055/a-1932-8976 PMID: 36037829

Rensma, S. P., van Sloten, T. T., Launer, L. J., & Stehouwer, C. D. A. (2018). Cerebral small vessel disease and risk of incident stroke, dementia and depression, and all-cause mortality: A systematic review and meta-analysis. Neuroscience & Biobehavioral Reviews, 90, 164-173. https://doi.org/10.1016/j.neubiorev.2018.04.003 PMID: 29656031 · PMCID: PMC6123527

Robinson, R. G., & Jorge, R. E. (2016). Post-stroke depression: A review. American Journal of Psychiatry, 173(3), 221-231. https://doi.org/10.1176/appi.ajp.2015.15030363 PMID: 26684921

Salagre, E., Fernandes, B. S., Dodd, S., Brownstein, D. J., & Berk, M. (2016). Statins for the treatment of depression: A meta-analysis of randomized, double-blind, placebo-controlled trials. Journal of Affective Disorders, 200, 235-242. https://doi.org/10.1016/j.jad.2016.04.047 PMID: 27148902

Santos, A. P., Lôbo, A. O. M., Santos, C. B. A. L., Carvalho, M. G. N. C., Ribeiro, M. M. C., Rodrigues, A. E. O., Santos, S. M., Costa Filho, B. H. C. C., Welkovic Junior, S., de Melo, E. S., & Barbosa, B. J. A. P. (2025). Behavioral and psychological symptoms of dementia associated with vascular impairment: a systematic review and meta-analysis. Brazilian Journal of Psychiatry, 47, e20254225. https://doi.org/10.47626/1516-4446-2025-4225 PMID: 41043167 · PMCID: PMC12815398

Santos, M., Gold, G., Kövari, E., Herrmann, F. R., Hof, P. R., Bouras, C., & Giannakopoulos, P. (2010). Neuropathological analysis of lacunes and microvascular lesions in late-onset depression. Neuropathology and Applied Neurobiology, 36(7), 661-672. https://doi.org/10.1111/j.1365-2990.2010.01101.x PMID: 20609111 · PMCID: PMC2962688

Sheline, Y. I., Barch, D. M., Garcia, K., Gersing, K., Pieper, C., Welsh-Bohmer, K., Steffens, D. C., & Doraiswamy, P. M. (2006). Cognitive function in late life depression: Relationships to depression severity, cerebrovascular risk factors and processing speed. Biological Psychiatry, 60(1), 58-65. https://doi.org/10.1016/j.biopsych.2005.09.019 PMID: 16414031

Sheline, Y. I., Pieper, C. F., Barch, D. M., Welsh-Bohmer, K., McKinstry, R. C., MacFall, J. R., ... Doraiswamy, P. M. (2010). Support for the vascular depression hypothesis in late-life depression: results of a 2-site, prospective, antidepressant treatment trial. Archives of General Psychiatry, 67(3), 277-285. https://doi.org/10.1001/archgenpsychiatry.2009.204 PMID: 20194828 · PMCID: PMC2838210

Shen, X., Howard, D. M., Adams, M. J., et al. (2020). A phenome-wide association and Mendelian Randomisation study of polygenic risk for depression in UK Biobank. Nature Communications, 11, 2301. https://doi.org/10.1038/s41467-020-16022-0 PMID: 32385265 · PMCID: PMC7210889

Sneed, J. R., Culang-Reinlieb, M. E., Brickman, A. M., Gunning-Dixon, F. M., Johnert, L., Garcon, E., & Roose, S. P. (2011). MRI signal hyperintensities and failure to remit following antidepressant treatment. Journal of Affective Disorders, 135(1-3), 315-320. https://doi.org/10.1016/j.jad.2011.06.052 PMID: 21802739

Sneed, J. R., Rindskopf, D., Steffens, D. C., Krishnan, K. R. R., & Roose, S. P. (2008). The vascular depression subtype: evidence of internal validity. Biological Psychiatry, 64(6), 491-497. https://doi.org/10.1016/j.biopsych.2008.03.032 PMID: 18490003 · PMCID: PMC2597784

Sneed, J. R., Roose, S. P., & Sackeim, H. A. (2006). Vascular depression: A distinct diagnostic subtype? Biological Psychiatry, 60(12), 1295-1298. https://doi.org/10.1016/j.biopsych.2006.06.018 PMID: 16996483

Steffens, D. C., Taylor, W. D., & Krishnan, K. R. R. (2003). Progression of subcortical ischemic disease from vascular depression to vascular dementia. American Journal of Psychiatry, 160(10), 1751-1756. https://doi.org/10.1176/appi.ajp.160.10.1751 PMID: 14514484

Taragano, F. E., Allegri, R., Vicario, A., Bagnatti, P., & Lyketsos, C. G. (2001). A double blind, randomized clinical trial assessing the efficacy and safety of augmenting standard antidepressant therapy with nimodipine in the treatment of 'vascular depression'. International Journal of Geriatric Psychiatry, 16(3), 254-260. https://doi.org/10.1002/gps.340 PMID: 11288158

Taragano, F. E., Bagnatti, P., & Allegri, R. F. (2005). A double-blind, randomized clinical trial to assess the augmentation with nimodipine of antidepressant therapy in the treatment of "vascular depression". International Psychogeriatrics, 17(3), 487-498. https://doi.org/10.1017/s1041610205001493 PMID: 16252380

Tay, J., Morris, R. G., Tuladhar, A. M., Husain, M., de Leeuw, F.-E., & Markus, H. S. (2020). Apathy, but not depression, predicts all-cause dementia in cerebral small vessel disease. Journal of Neurology, Neurosurgery & Psychiatry, 91(9), 953-959. https://doi.org/10.1136/jnnp-2020-323092 PMID: 32651249 · PMCID: PMC7476304

Tay, J., Tuladhar, A. M., Hollocks, M. J., Brookes, R. L., Tozer, D. J., Barrick, T. R., Husain, M., de Leeuw, F. E., & Markus, H. S. (2019). Apathy is associated with large-scale white matter network disruption in small vessel disease. Neurology, 92(11), e1157-e1167. https://doi.org/10.1212/wnl.0000000000007095 PMID: 30737341 · PMCID: PMC6511108

Taylor, W. D., Aizenstein, H. J., & Alexopoulos, G. S. (2013). The vascular depression hypothesis: Mechanisms linking vascular disease with depression. Molecular Psychiatry, 18(9), 963-974. https://doi.org/10.1038/mp.2013.20 PMID: 23439482 · PMCID: PMC3674224

Taylor, W. D., MacFall, J. R., Payne, M. E., McQuoid, D. R., Provenzale, J. M., Steffens, D. C., & Krishnan, K. R. R. (2004). Late-life depression and microstructural abnormalities in dorsolateral prefrontal cortex white matter. American Journal of Psychiatry, 161(7), 1293-1296. https://doi.org/10.1176/appi.ajp.161.7.1293 PMID: 15229065

Taylor, W. D., Steffens, D. C., MacFall, J. R., McQuoid, D. R., Payne, M. E., Provenzale, J. M., & Krishnan, K. R. R. (2003). White matter hyperintensity progression and late-life depression outcomes. Archives of General Psychiatry, 60(11), 1090-1096. https://doi.org/10.1001/archpsyc.60.11.1090 PMID: 14609884

Tedeschini, E., Levkovitz, Y., Iovieno, N., Ameral, V. E., Nelson, J. C., & Papakostas, G. I. (2011). Efficacy of antidepressants for late-life depression: A meta-analysis and meta-regression of placebo-controlled randomized trials. Journal of Clinical Psychiatry, 72(12), 1660-1668. https://doi.org/10.4088/jcp.10r06531 PMID: 22244025

Teodorczuk, A., O'Brien, J. T., Firbank, M. J., Pantoni, L., Poggesi, A., Erkinjuntti, T., Wallin, A., Wahlund, L.-O., Gouw, A., Waldemar, G., Schmidt, R., Ferro, J. M., Chabriat, H., Bazner, H., & Inzitari, D. (2007). White matter changes and late-life depressive symptoms: longitudinal study. The British Journal of Psychiatry, 191, 212-217. https://doi.org/10.1192/bjp.bp.107.036756 PMID: 17766760

Thomas, A. J., Kalaria, R. N., & O'Brien, J. T. (2004). Depression and vascular disease: what is the relationship? Journal of Affective Disorders, 79(1-3), 81-95. https://doi.org/10.1016/s0165-0327(02)00349-x PMID: 15023483

Thomas, A. J., O'Brien, J. T., Davis, S., Ballard, C., Barber, R., Kalaria, R. N., & Perry, R. H. (2002). Ischemic basis for deep white matter hyperintensities in major depression: A neuropathological study. Archives of General Psychiatry, 59(9), 785-792. https://doi.org/10.1001/archpsyc.59.9.785 PMID: 12215077

Tsopelas, C., Stewart, R., Savva, G. M., Brayne, C., Ince, P., Thomas, A., & Matthews, F. E. (2011). Neuropathological correlates of late-life depression in older people. The British Journal of Psychiatry, 198(2), 109-114. https://doi.org/10.1192/bjp.bp.110.078816 PMID: 21282780

Twait, E. L., Wu, J. H., Kamarioti, M., Basten, M., van der Flier, W. M., Gerritsen, L., & Geerlings, M. I. (2024). Association of amyloid-beta with depression or depressive symptoms in older adults without dementia: a systematic review and meta-analysis. Translational Psychiatry, 14(1), 25. https://doi.org/10.1038/s41398-024-02739-9 PMID: 38225253 · PMCID: PMC10789765

Valenti, R., Pescini, F., Antonini, S., Castellini, G., Poggesi, A., Bianchi, S., Inzitari, D., Pallanti, S., & Pantoni, L. (2011). Major depression and bipolar disorders in CADASIL: A study using the DSM-IV semi-structured interview. Acta Neurologica Scandinavica, 124(6), 390-395. https://doi.org/10.1111/j.1600-0404.2011.01512.x PMID: 21428968

Valkanova, V., & Ebmeier, K. P. (2013). Vascular risk factors and depression in later life: A systematic review and meta-analysis. Biological Psychiatry, 73(5), 406-413. https://doi.org/10.1016/j.biopsych.2012.10.028 PMID: 23237315

van Agtmaal, M. J. M., Houben, A. J. H. M., Pouwer, F., Stehouwer, C. D. A., & Schram, M. T. (2017). Association of microvascular dysfunction with late-life depression: A systematic review and meta-analysis. JAMA Psychiatry, 74(7), 729-739. https://doi.org/10.1001/jamapsychiatry.2017.0984 PMID: 28564681 · PMCID: PMC5710252

van Dijk, G. G. C. E. A., Box, C. V. J., Faghir, F., Martens, J., Papma, J. M., Mattace Raso, F. U. S., van den Berg, E., Vernooij, M. W., & Wolters, F. J. (2026). Visual rating versus volumetric segmentation of white matter hyperintensities for cognitive assessment: A memory clinic-based comparison and systematic review. Cerebral Circulation - Cognition and Behavior, 11, 100564. https://doi.org/10.1016/j.cccb.2026.100564 PMID: 42898208 · PMCID: PMC13677418

van Sloten, T. T., Schram, M. T., Adriaanse, M. C., Dekker, J. M., Nijpels, G., Teerlink, T., Scheffer, P. G., Pouwer, F., Schalkwijk, C. G., Stehouwer, C. D. A., & Henry, R. M. (2014). Endothelial dysfunction is associated with a greater depressive symptom score in a general elderly population: the Hoorn Study. Psychological Medicine, 44(7), 1403-1416. https://doi.org/10.1017/S0033291713002043 PMID: 23942242

van Sloten, T. T., Sigurdsson, S., van Buchem, M. A., Phillips, C. L., Jonsson, P. V., Ding, J., ... Launer, L. J. (2015). Cerebral small vessel disease and association with higher incidence of depressive symptoms in a general elderly population: the AGES-Reykjavik Study. American Journal of Psychiatry, 172(6), 570-578. https://doi.org/10.1176/appi.ajp.2014.14050578 PMID: 25734354 · PMCID: PMC4451386

van Sloten, T. T., Souverein, P. C., Stehouwer, C. D. A., & Driessen, J. H. M. (2022). Angiotensin-converting enzyme inhibitors and angiotensin receptor blockers and risk of depression among older people with hypertension. Journal of Psychopharmacology, 36(5), 594-603. https://doi.org/10.1177/02698811221082470 PMID: 35388727 · PMCID: PMC9112619

Vyas, M. V., Wang, J. Z., Gao, M. M., & Hackam, D. G. (2021). Association between stroke and subsequent risk of suicide: A systematic review and meta-analysis. Stroke, 52(4), 1460-1464. https://doi.org/10.1161/strokeaha.120.032692 PMID: 33691505

Wagenmakers, M. J., Vansteelandt, K., van Exel, E., Postma, R., Schouws, S. N. T. M., Obbels, J., Rhebergen, D., Bouckaert, F., Stek, M. L., Barkhof, F., Beekman, A. T. F., Veltman, D. J., Sienaert, P., Dols, A., & Oudega, M. L. (2021). Transient cognitive impairment and white matter hyperintensities in severely depressed older patients treated with electroconvulsive therapy. American Journal of Geriatric Psychiatry, 29(11), 1117-1128. https://doi.org/10.1016/j.jagp.2020.12.028 PMID: 33454176

Wang, L., Leonards, C. O., Sterzer, P., & Ebinger, M. (2014). White matter lesions and depression: a systematic review and meta-analysis. Journal of Psychiatric Research, 56, 56-64. https://doi.org/10.1016/j.jpsychires.2014.05.005 PMID: 24948437

Wang, M., Daghlas, I., & Wei, N. (2025). Genetic liability to depression and cerebral small vessel disease: A Mendelian randomization study. Journal of Stroke and Cerebrovascular Diseases, 34(8), 108385. https://doi.org/10.1016/j.jstrokecerebrovasdis.2025.108385 PMID: 40570994

Wardlaw, J. M., Smith, E. E., Biessels, G. J., Cordonnier, C., Fazekas, F., Frayne, R., Lindley, R. I., O'Brien, J. T., Barkhof, F., Benavente, O. R., et al. (2013). Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. The Lancet Neurology, 12(8), 822-838. https://doi.org/10.1016/S1474-4422(13)70124-8 PMID: 23867200 · PMCID: PMC3714437

Wen, M.-C., Steffens, D. C., Chen, M.-K., & Zainal, N. H. (2014). Diffusion tensor imaging studies in late-life depression: Systematic review and meta-analysis. International Journal of Geriatric Psychiatry, 29(12), 1173-1184. https://doi.org/10.1002/gps.4129 PMID: 24798480

Williams, L. S., Rittman, M. R., Boylstein, C., Faircloth, C., & Haijing, Q. (2005). Qualitative and quantitative measurement of depression in veterans recovering from stroke. Journal of Rehabilitation Research and Development, 42(3), 277-290. https://doi.org/10.1682/jrrd.2004.02.0017 PMID: 16187241

Wouts, L., van Kessel, M., Beekman, A. T. F., Marijnissen, R. M., & Oude Voshaar, R. C. (2020). Empirical support for the vascular apathy hypothesis: A structured review. International Journal of Geriatric Psychiatry, 35(1), 3-11. https://doi.org/10.1002/gps.5217 PMID: 31617249 · PMCID: PMC6916153

Wu, S., Zhang, Y., Lu, Y., Yin, Y., Yang, C., Tang, W., Song, T., Tao, X., & Wang, Q. (2025). Vascular depression: A comprehensive exploration of the definition, mechanisms, and clinical challenges. Neurobiology of Disease, 211, 106946. https://doi.org/10.1016/j.nbd.2025.106946 PMID: 40349857

Xekardaki, A., Santos, M., Hof, P., Kövari, E., Bouras, C., & Giannakopoulos, P. (2012). Neuropathological substrates and structural changes in late-life depression: The impact of vascular burden. Acta Neuropathologica, 124(4), 453-464. https://doi.org/10.1007/s00401-012-1021-5 PMID: 22836715

Ye, Z., Kappelmann, N., Moser, S., Davey Smith, G., Burgess, S., Jones, P. B., & Khandaker, G. M. (2021). Role of inflammation in depression and anxiety: Tests for disorder specificity, linearity and potential causality of association in the UK Biobank. EClinicalMedicine, 38, 100992. https://doi.org/10.1016/j.eclinm.2021.100992 PMID: 34505025 · PMCID: PMC8413248

Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, O., Huang, V., Adey, M., & Leirer, V. O. (1982). Development and validation of a geriatric depression screening scale: A preliminary report. Journal of Psychiatric Research, 17(1), 37-49. https://doi.org/10.1016/0022-3956(82)90033-4 PMID: 7183759

Zanardi, R., Poletti, S., Prestifilippo, D., Attanasio, F., Barbini, B., & Colombo, C. (2020). Transcranial direct current stimulation: A novel approach in the treatment of vascular depression. Brain Stimulation, 13(6), 1559-1565. https://doi.org/10.1016/j.brs.2020.08.013 PMID: 32896644

Zhang, R., Li, T., Fan, F., He, H., Lan, L., Sun, D., Xu, Z., Peng, S., Cao, J., Xu, J., Peng, X., Lei, M., Song, H., & Zhang, J. (2025). Identification and validation of an explainable machine learning model for vascular depression diagnosis in the older adults: a multicenter cohort study. BMC Medicine, 23(1), 448. https://doi.org/10.1186/s12916-025-04283-9 PMID: 40745611 · PMCID: PMC12315376