The cross-sectional associations between enlarged perivascular spaces and mild behavioral impairment in dementia-free older adults

Abstract

Background

Enlarged perivascular spaces (EPVS) and mild behavioral impairment (MBI) are associated with greater dementia risk.

Objective

We investigated cross-sectional associations between EPVS burden and MBI presence and symptom severity.

Methods

Participants were dementia-free older adults in the Comprehensive Assessment of Neurodegeneration and Dementia (COMPASS-ND) study. EPVS were assessed using a validated visual rating scale applied to T2-weighted magnetic resonance imaging. MBI was measured using the informant-reported MBI Checklist (MBI-C) with a cut-point of >5 for MBI presence. Multivariable logistic and zero-inflated negative binomial regressions modelled EPVS associations with MBI presence and symptom severity, respectively, adjusting for age, education, Montreal Cognitive Assessment score, and presence of hypertension with an EPVS*sex interaction term.

Results

Among 363 participants (52.9% female), every 1-point rise in total EPVS score was associated with 1.28-fold greater odds of having MBI (95%CI:1.06–1.53, p = 0.009). When examining regional EPVS burden, those with higher centrum semiovale scores had greater odds of MBI (aOR = 1.47, 95%CI:1.08–2.00, p = 0.01), but this association was not significant for the basal ganglia (aOR = 1.37, 95%CI:0.95–1.97, p = 0.09) or midbrain (aOR = 1.49, 95%CI:0.80–2.79, p = 0.21). Total and regional EPVS scores were not associated with MBI symptom severity. None of the associations were moderated by sex.

Conclusions

Older adults with greater global and centrum semiovale EPVS burden are more likely to have MBI. These findings suggest that early microvascular pathology may contribute to later-life emergent and persistent behavioral symptoms, although longitudinal data are required, along with other indicators of vascular pathology.

Keywords

Alzheimer's disease cerebral small vessel disease dementia enlarged perivascular spaces mild behavioral impairment neuropsychiatric symptoms vascular disease

Introduction

Dementia is characterized by cognitive and behavioral impairment severe enough to affect daily life. While Alzheimer disease (AD) is the most common cause of dementia in older adults, there is a significant role of vascular factors in the trajectory of dementia.¹ Cerebrovascular risk factors and related diseases among older adults affect nearly one-third of patients with dementia.² Brain infarcts, arterial lipid deposits, or arteriolar wall thickening have been found in many patients diagnosed with AD in autopsy studies, indicating comorbid vascular pathology.³ Studies also implicate vascular disease in over 90% of cases of probable AD and mixed AD pathology.⁴

Cerebral small vessel disease (CSVD) is the most common vascular pathology in dementia. CSVD is a chronic and progressive vascular disease that affects arterioles, capillaries, and small veins of the brain.⁵ Several longitudinal studies have investigated the coexistence of CSVD with AD pathology in those with dementia, with some indicating an additive relationship and others illustrating a synergistic contribution to cognitive impairment.^6–9 Neuroimaging features of CSVD are associated with a greater risk of AD.¹⁰

Enlarged perivascular spaces (EPVS) are a neuroimaging feature of CSVD, alongside other markers such as white matter hyperintensities (WMHs) and lacunes.¹¹ EPVS has been reported to increase in number on magnetic resonance imaging (MRI) with age, along with other brain features of CSVD.^12,13 Similar to other CSVD features,^13,14 high EPVS burden predicts future risk of dementia.¹⁵ EPVS burden in the centrum semiovale is associated with a greater risk of incident cognitive decline and dementia, independent of other vascular risks, total brain volume, WMHs, and covert infarcts.¹⁶ Hence, EPVS may be a useful subclinical MRI marker to identify individuals who are at higher risk of developing dementia.

While behavioral symptoms are core features of dementia, these symptoms can also emerge in advance of dementia.^17,18 These behavioral changes have been incorporated into a validated clinical construct called mild behavioral impairment (MBI).¹⁷ MBI is characterized by the later-life onset of persistent neuropsychiatric symptoms (NPS) that represent change from longstanding behaviors, as an indicator of dementia risk. Numerous studies have supported MBI, whether concurrently presenting with mild cognitive impairment (MCI) or absent objective cognitive impairment in persons with subjective cognitive decline (SCD) or normal cognition, as a clinical marker of risk for incident dementia.^18–22

The link between MBI and specific neurodegenerative etiologies of dementia has been explored primarily with AD biomarkers, including amyloid-β (Aβ) and phosphorylated tau (p-tau), AD genetic risk factors like Apolipoprotein E, and AD patterns of brain atrophy or neurodegeneration.^23–33 However, other brain pathologies and atrophy patterns may contribute to later-life emergent and persistent behavioral symptoms.^34,35 There is some evidence supporting the association between cerebrovascular pathologies and behavioral impairment. Recent findings show that NPS are common in cerebral amyloid angiopathy (CAA), with severity linked to greater white matter hyperintensity volume.³⁶ Postmortem studies also reported associations between NPS and cerebrovascular disease.³⁷ Cardiovascular risk factors, like diabetes, hypertension, and stroke, additionally accelerate cognitive decline and exacerbate the presence of NPS, such as depression and apathy.^38,39 More recently, studies have found an association between greater burden of vascular risk factors (e.g., hypertension, diabetes, dyslipidemia, high body mass index) and MBI.^40–42

However, the association between CSVD markers and MBI remains poorly characterized, and EPVS burden has not been studied in this regard. Therefore, we aimed to investigate cross-sectional associations between EPVS burden and MBI presence (i.e., MBI + status) and symptom severity. Given that EPVS in different regions may reflect different underlying CSVD pathology, a secondary analysis examined associations involving regional EPVS burden in the basal ganglia, centrum semiovale, and midbrain EPVS.⁴³ Due to the high prevalence of cerebrovascular disease in older adults with dementia and the association between MBI and WMH volume,⁴⁴ we hypothesized that older adults with greater EPVS burden would be more likely to have MBI and exhibit greater MBI symptom severity.

Method

Study design

Data were from the Comprehensive Assessment of Neurodegeneration and Dementia (COMPASS-ND), a Canada-wide prospective cohort study aiming to investigate several causes of neurodegeneration and their interactions (ClinicalTrials.gov NCT03402919).^45,46 COMPASS-ND participants included older adults with dementia, MCI, SCD, and cognitively normal individuals (CN). For this study, we defined CN as those who performed in the normal range on neuropsychological testing, which includes those who were CN and had SCD. Data collected included sociodemographic information, clinical and cognitive data, structural brain imaging, biomarkers, and measures of sensory function. Extensive descriptions of COMPASS-ND recruitment and data collection procedures have been published elsewhere.⁴⁵ The COMPASS-ND study was approved by the Jewish General Research Ethics Board.⁴⁵ Data were downloaded on October 2, 2023. Data were collected with ethical approval by institutional review boards at each of the 13 study sites affiliated with the study, with written informed consent provided by participants.⁴⁷

Participants

The complete sample from COMPASS-ND consisted of 1284 participants. However, the Mild Behavioral Impairment Checklist (MBI-C) was only introduced into the study about two years after its initial launch.^48,49 Therefore, only participants enrolled thereafter were administered the MBI-C (n = 531). From this subset of participants, further exclusions were made if participants (1) had a diagnosis of dementia, indicating that they no longer met criteria for MBI (n = 115)¹⁷; (2) were missing T2 MRI scans (n = 36); (3) had poor quality T2 scans, where EPVS severity could not be accurately assessed (n = 15); and (4) were missing data for covariates such as sex, education, or Montreal Cognitive Assessment (MoCA) score (n = 2). The final sample size was 363 (Figure 1).

Figure 1.

Participant flow-diagram. Demographics included age, sex, education, and MoCA score. COMPASS-ND: Comprehensive Assessment of Neurodegeneration and Dementia; MBI-C: Mild Behavioral Impairment Checklist; MRI: magnetic resonance image; EPVS: enlarged perivascular spaces; MoCA: Montreal Cognitive Assessment.

MRI acquisition

Brain MRIs were acquired using the previously validated Canadian Dementia Imaging Protocol (http://www.cdip-pcid.ca).^50,51 Across data collection sites, participants underwent MRI scanning on a 3 T magnet machine.⁴⁵ The sequences used included three-dimensional T1, proton density/T2, fluid-attenuated inversion recovery, gradient echo, resting-state functional MRI, and diffusion tensor imaging.

For this study, data on EPVS were collected from T2-weighted MRI scans which allow for more accurate detection of EPVS due to the high contrast between EPVS and the white matter.⁵² Three-dimensional inversion-prepared T2-weighted images were acquired using the following parameters: repetition time (TR)/echo time (TE) = 650/20 ms, flip angle = 20°, voxel size = 0.94 × 0.94 × 3.0 mm3, field of view (FOV) = 240 × 240 mm, resolution = 0.94 × 0.94 × 3.0 mm.^50,51

Enlarged perivascular spaces rating

EPVS were assessed by three raters (DG, AG, GM) using a validated visual rating scale, which evaluates the number of EPVS in the basal ganglia and centrum semiovale, as well as the presence/absence of EPVS in the midbrain.⁵³ This scale was used in this study as it incorporates all relevant EPVS sites and allows the rating of all grades of EPVS severity on standard structural brain MRI.⁵³ EPVS severity is rated from 0–4 in the basal ganglia and centrum semiovale. A rating of 0 indicates no EPVS, 1 indicates 1–10 EPVS, 2 indicates 11–20 EPVS, 3 indicates 21–40 EPVS, and 4 indicates more than 40 EPVS. For the midbrain, only presence or absence is recorded, with a maximum score of 1. For each rating, both hemispheres and all relevant slices are assessed; however, only the slice and hemisphere with the highest number of EPVS are used for the rating. The total EPVS score is obtained from the sum of the three regional ratings.

Inter-rater reliability was established through a training program wherein AG (neurologist with 10 years of experience in neuroimaging research) trained DG and GM on EPVS ratings using 20 training cases graded in person. All three raters then independently rated 20 unseen testing cases. Gwet's agreement coefficient (AC, particularly suitable for samples with greater frequencies of certain values, e.g., 0–1) and percentage agreement with AG were calculated, with Gwet's AC > 0.6 and agreement >70% set a priori as thresholds for DG and GM to attain in order to proceed with independent ratings.⁵⁴

Mild behavioral impairment

MBI was measured using the informant-reported MBI-C.⁴⁸ The MBI-C comprises five domains assessed by 34 items: decreased motivation (6 questions), emotional dysregulation (6 questions), impulse dyscontrol (12 questions), social inappropriateness (5 questions), and abnormal perceptions (5 questions). Each question assesses the severity of symptoms and can be answered on a scale from 0–3, with 0 indicating no symptoms and 3 indicating severe symptoms. The full MBI-C is available for free at https://www.mbitest.org/. MBI presence was ascertained based on an MBI-C total score >5, as this cut-off can effectively distinguish between participants with and without MBI in CN and mixed CN and MCI samples.^55,56 The MBI total score, ranging from 0–102, was calculated by summing the scores for each item, with higher scores indicating more severe symptoms.

Statistical analyses

Multivariable logistic regressions modelled the cross-sectional associations between total, basal ganglia, centrum semiovale, and midbrain EPVS scores (predictors) and MBI presence (outcome). Multivariable zero-inflated negative binomial regression modelled the cross-sectional associations between total, basal ganglia, centrum semiovale, and midbrain EPVS scores (predictors) and MBI symptom severity (outcome). This model was used because the distribution of MBI-C scores resembled an overdispersed (variance > mean and variance increases quadratically) Poisson distribution with an excess of zeroes. Raw coefficients, which are in the natural logarithmic scale, were exponentiated to facilitate interpretation along the original MBI-C scale of measurement. The corresponding coefficients reported in this paper represent the adjusted multiplicative change in the MBI-C total score for each one-unit rise in EPVS score.

The analyses were controlled for age, sex, education, MoCA score, and presence of hypertension; sex was explored as a moderator for all associations as per the COMPASS-ND publications and data access policy. R version 4.4.0 was used to conduct all statistical analyses.⁵⁷

Results

Participant characteristics

Participant characteristics for all analyses are summarized in Table 1. Of the total sample (n = 363), the mean age was 74.6 (SD = 7.2, range = 54.8–95.1), 192 (52.9%) were females, and mean years of education was 15.7 (SD = 3.4, range = 5.0–30.0). Regarding cognition, 180 (49.6%) participants had MCI, and 183 (50.4%) were CN, with a mean MoCA score of 25.3 (SD = 3.6, range = 13.0–30.0). There were no significant differences in vascular risk factor presence between the MBI presence and absence groups with the exception of hypertension, which was more common in the MBI group. The median EPVS score was 5.0 (IQR = 4.0–6.0), with a median basal ganglia EPVS score of 2.0 (IQR = 1.0–2.0), centrum semiovale EPVS score of 2.0 (IQR = 2.0–3.0), and midbrain EPVS score of 1.0 (IQR = 1.0–1.0). Lastly, the median MBI-C score was 2.0 (0–8.0); 125 (34.4%) participants were MBI + based on the MBI-C total score cut-off >5.

Table 1.

Participant demographics.

	Total	MBI Present	MBI Absent	p
n	363	125	238
Age (y)	74.6 (7.2), 54.8–95.1	75.0 (8.1), 54.8–94.2	74.4 (6.6), 58.4–95.1	0.30^w
Sex				<0.001^c
Female	192 (52.9%)	51 (40.8%)	141 (59.2%)
Male	171 (47.1%)	74 (59.2%)	97 (40.8%)
Education (Years)	15.7 (3.4), 5.0–30.0	15.3 (3.5), 7.0–25.0	15.9 (3.4), 5.0–30.0	0.057^w
MoCA Score	25.3 (3.6), 13.0–30.0	24.1 (3.9), 13.0–30.0	25.9 (3.3), 14.0–30.0	<0.001^w
Cognitive Diagnosis				<0.001^c
MCI	180 (49.6%)	85 (68.0%)	95 (39.9%)
CN	183 (50.4%)	40 (32.0%)	143 (60.1%)
Vascular Risk Factors
Hypertension Present	122 (33.6%)	52 (41.6%)	70 (29.4%)	0.02^c
Hyperlipidemia Present	152 (41.9%)	54 (43.2%)	98 (41.2%)	0.73^c
Type I Diabetes Present	4 (1.1%)	2 (1.6%)	2 (0.8%)	0.51^c
Type 2 Diabetes Present	40 (11.0%)	19 (15.2%)	21 (8.8%)	0.06^c
MBI-C Score	2.0 (0–8.0)	12.0 (8.0–20.0)	0 (0–1.0)	<0.001^w
Total EPVS Score	5.0 (4.0–6.0)	5.0 (4.0–6.0)	4.5 (4.0–6.0)	0.001^w
Basal Ganglia Score	2.0 (1.0–2.0)	2.0 (1.0–2.0)	1.0 (1.0–2.0)	0.011^w
Centrum Semiovale Score	2.0 (2.0–3.0)	3.0 (2.0–3.0)	2.0 (2.0–3.0)	0.004^w
Midbrain Score	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	0.18^w

Values for continuous variables are shown in mean (SD), range, except for MBI and EPVS scores, which are indicated in median (IQR). Values were rounded to one decimal place where appropriate. Categorical variables are shown as n (%). Five participants were missing data for the vascular risk factors: hyperlipidemia (n = 1), type I diabetes (n = 2), type II diabetes (n = 2). MBI presence was defined as an MBI total score of >5. MBI: mild behavioral impairment; MoCA: Montreal Cognitive Assessment; MCI: mild cognitive impairment; CN: cognitively normal; EPVS: enlarged perivascular space. ^wWilcoxon Rank Sum (Mann-Whitney) Test; ^cChi-Squared Test

Inter-rater reliability of EPVS ratings

Compared to the senior expert rater, each rater attained the pre-set thresholds of Gwet's AC > 0.60 and agreement > 70% based on 20 independently rated test cases from the dataset. Specifically, Gwet's AC and percentage agreement were, respectively, 0.76–0.80 and 85.0–85.7% for midbrain ratings, 0.74–0.94 and 80.0–95.2% for the basal ganglia, and 0.72–0.79 and 81.0–85.0% for the centrum semiovale.

The association between EPVS score and MBI status

As shown in Table 2 and Figure 2, for every 1-point rise in EPVS total score, the odds of having MBI increased by a factor of 1.28 (95%CI:1.06–1.53, p = 0.009). When examining subregion EPVS scores, those with higher centrum semiovale scores had greater odds of MBI (aOR = 1.47, 95%CI:1.08–2.00, p = 0.01). However, this association was not observed for the basal ganglia (aOR = 1.37, 95%CI:0.95–1.97, p = 0.09) or midbrain (aOR = 1.49, 95%CI:0.80–2.79, p = 0.21). The associations between total and subregion EPVS and MBI status were not dependent on sex, based on our moderation analyses (EPVS total*sex aOR = 0.83, 95%CI:0.58–1.17, p = 0.29; EPVS basal ganglia*sex aOR = 0.73, 95%CI:0.37–1.46, p = 0.38; EPVS centrum semiovale*sex aOR = 0.80, 95%:0.44–1.48, p = 0.48; EPVS midbrain*sex aOR = 0.88, 95%CI:0.25–3.08, p = 0.84).

Figure 2.

Violin plots for the distribution of the total EPVS score, basal ganglia EPVS score, centrum semiovale EPVS score, and midbrain EPVS score for participants with MBI absence and presence. MBI presence was determined using an MBI-C total score cut-off of >5. The white line indicates the interquartile range. Data points were jittered randomly by up to 0.4 units along the x-axis and 0.3 units along the y-axis to reduce overlap. Wilcoxon rank-sum test p-values are provided above each plot representing the difference between the MBI absence and presence distributions. EPVS: enlarged perivascular spaces; MBI: mild behavioral impairment; MBI-C: mild behavioral impairment checklist (n = 363).

Table 2.

Multivariable regression model outputs for the association between MBI presence, MBI total score, and EPVS ratings.

Exposure Variable	Outcome Variable	aOR	95%CI	p
Total EPVS	MBI Presence	1.28	1.06–1.53	0.009*
Basal Ganglia EPVS	MBI Presence	1.37	0.95–1.97	0.09
Centrum Semiovale EPVS	MBI Presence	1.47	1.08–2.00	0.01*
Midbrain EPVS	MBI Presence	1.49	0.80–2.79	0.21

Exposure Variable	Outcome Variable	b	95%CI	p
Total EPVS	MBI Total Score	1.08	0.94–1.24	0.28
Basal Ganglia EPVS	MBI Total Score	1.06	0.79–1.41	0.71
Centrum Semiovale EPVS	MBI Total Score	1.13	0.89–1.44	0.31
Midbrain EPVS	MBI Total Score	1.30	0.81–2.08	0.28

For the associations between EPVS (total, basal ganglia, centrum semiovale, and midbrain) and MBI presence multivariable logistic regressions were used. For the associations between EPVS (total, basal ganglia, centrum semiovale, and midbrain) and MBI total score multivariable zero-inflated negative binomial models were used. Coefficients for MBI total score (b) are exponentiated to reflect the original scale (n = 363). aOR: adjusted odds ratio; CI: confidence interval; EPVS: enlarged perivascular spaces; MBI: mild behavioral impairment.

The association between EPVS score and MBI symptom severity

As shown in Table 2 and Figure 3, the total EPVS score was not associated with MBI-C total score (b = 1.08, 95%CI:0.94–1.24, p = 0.28). Additionally, none of the basal ganglia (b = 1.06, 95%CI:0.79–1.41, p = 0.71), centrum semiovale (b = 1.13, 95%CI:0.89–1.44, p = 0.31), or midbrain (b = 1.30, 95%CI:0.81–2.08, p = 0.28) EPVS scores were associated with MBI-C total score. The associations between total and subregion EPVS and MBI symptom severity were not dependent on sex, based on our moderation analyses (EPVS total*sex b = 0.89, 95%CI:0.68–1.17, p = 0.41; EPVS basal ganglia*sex b = 0.70, 95%CI:0.41–1.21, p = 0.21; EPVS centrum semiovale*sex b = 0.88, 95%:0.55–1.42, p = 0.60; EPVS midbrain*sex b = 1.05, 95%CI:0.40–2.74, p = 0.92).

Figure 3.

Boxplot for the distribution of the EPVS score quartiles as a function of the total MBI-C score. Kruskal-Wallis test p-value represents differences between quartiles. EPVS: enlarged perivascular spaces; MBI-C: Mild Behavioral Impairment Checklist (n = 363).

Discussion

This study examined the cross-sectional associations between EPVS burden and MBI presence and severity to understand the relationship between these dementia risk markers. We found that participants with greater total and centrum semiovale EPVS burden had greater odds of having MBI.

Our exploration of the relationship between EPVS and MBI adds to the previous literature on the association of cerebrovascular pathology with neuropsychiatric symptoms. First, there is a link between NPS preceding dementia and vascular disease.⁵⁸ For instance, the worsening of NPS is a clinical indicator of CSVD progression and progression in NPS and CSVD are associated with cognitive decline.⁵⁹ Postmortem studies have found an overlap between the presence of NPS and cerebrovascular disease in non-dementia populations.³⁷ Studies have also found that the number of vascular comorbidities, like arterial hypertension, diabetes, and dyslipidemia, were positively associated with MBI-C scores greater than 6.5.⁴⁰ In memory clinic patients with MCI, patients with MBI had a 9.3% greater WMH volume.⁴⁴ NPS are also common in CAA, where their severity is linked to white matter hyperintensity, suggesting white matter damage as a contributing factor.³⁶ However, no previous studies, to our knowledge, investigate the relationship between EPVS and MBI.

There may be several mechanisms underlying this association between global EPVS burden and the presence of MBI. First, PVS are involved in several pathways that aid in clearing brain waste, like the perivascular,⁶⁰ glymphatic,⁶¹ and intermural periarterial drainage pathways.⁶² Enlargement of PVS is a marker of dysfunction in these drainage pathways.⁶³ Thus, impaired clearance, as indicated by EPVS, could lead to the accumulation of toxic metabolites such as Aβ, increasing dementia risk. Given the association between MBI and Aβ,²⁶ MBI might be a clinical marker of poor Aβ clearance. Additionally, EPVS are associated with neuroinflammation, possibly due to reduced clearance of pro-inflammatory cytokines and increased blood-brain barrier (BBB) permeability.^64,65 Dysfunctions of glymphatic clearance may exacerbate neuroinflammation by impairing the clearance of pro-inflammatory cytokines.⁶⁶ Because the BBB is the internal boundary of the PVS,⁶⁷ inflammation in the PVS can disrupt the BBB, leading to the infiltration of other immune cells and furthering neuroinflammation. NPS have also been shown to have associations with neuroinflammation and oxidative stress,^68–71 suggesting that EPVS and MBI may both be markers of neuroinflammation.

This study also found that the EPVS scores in the centrum semiovale were associated with MBI status. However, EPVS scores in basal ganglia and midbrain were neither associated with MBI presence nor its severity. This finding may relate to the fact that EPVS in different regions have different underlying etiologies. For instance, EPVS in the basal ganglia are associated with hypertensive angiopathy.^72–74 In contrast, EPVS in the centrum semiovale are associated with advanced age and cerebral amyloid angiopathy,⁷⁵ which is frequently associated with AD,^76–79 and has shown associations with MBI.³⁶ Thus, it is possible that MBI is more strongly associated with amyloid deposition, including vascular amyloid, than other vascular disease, as only centrum semiovale EPVS were significantly associated with MBI in this study.

On the other hand, EPVS score was not associated with the severity of MBI, as reflected by the MBI-C total score. Considering the fact that EPVS are one of the earliest markers of CSVD, preceding vascular pathologies like lacunes, cerebral microbleeds,⁶³ and WMHs,^11,80 these early signs of EPVS alone may not be associated with severe MBI symptoms. However, this finding was unexpected, and future studies may seek to study the relationship of MBI symptom severity with more granular assessments of EPVS with emerging technology, such as automated quantification of EPVS numbers or of combining EPVS with other markers of CSVD. Additionally, it may be worth studying the relationship between EPVS and MBI symptom severity in populations with more severe CSVD burden.

This study is not without limitations. First, these are cross-sectional data, so no causal associations can be inferred. Future studies should investigate the longitudinal relationships between these markers to monitor changes in MBI over time as a function of EPVS severity. Second, the rating of EPVS severity is subject to human error and may not reflect the precise extent of vascular pathology in the participants. However, we used a rigorous system of training and testing of inter-rater reliability to ensure a high quality of our EPVS ratings. Lastly, cerebrospinal fluid (CSF) data were only available in a limited number of participants, and thus information about AD biomarkers such as amyloid-β 42, phosphorylated tau, and total tau were not included in the models. Exploring AD biomarkers in the context of EPVS is another avenue for future studies.

Conclusion

This study explored the relationship between EPVS severity and MBI presence and severity. We hypothesized that a greater EPVS burden would be associated with greater odds of MBI presence and increased MBI symptom severity. We found that greater global and centrum semiovale EPVS burdens were associated with MBI presence; however, we found no associations between EPVS burden and MBI symptom severity or between basal ganglia and midbrain EPVS scores and MBI presence or severity. These findings suggest that EPVS may contribute to later-life emergent and persistent behavioral symptoms, consistent with known associations between EPVS and cognition, and merit further validation in longitudinal studies. Future studies should investigate MBI associations with other CSVD markers.

Footnotes

Acknowledgements

Eric E. Smith and Aravind Ganesh are members of Team 7 and Zahinoor Ismail is a member of Team 11, which are a part of the Canadian Consortium on Neurodegeneration in Aging (CCNA).

Ethical considerations

The COMPASS-ND study was approved by the Jewish General Research Ethics Board.⁴⁵ Data were collected with ethical approval by institutional review boards at each of the 13 study sites affiliated with the study, with written informed consent provided by participants.⁴⁷

Consent to participate

All participants and their study partners provided written informed consent at the start of the COMPASS-ND study.

Consent for publication

Not applicable

Author contributions

Dinithi Mudalige: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Writing – original draft.

Dylan Guan: Conceptualization; Methodology; Supervision; Writing – review & editing.

Graham McLeod: Data curation; Methodology; Writing – review & editing.

Eric Smith: Conceptualization; Writing – review & editing.

Aravind Ganesh: Conceptualization; Methodology; Supervision; Writing – review & editing.

Zahinoor Ismail: Conceptualization; Resources; Supervision; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The Canadian Consortium on Neurodegeneration in Aging is supported by a grant from the Canadian Institutes of Health Research with funding from several partners. ZI is supported by CIHR (Grant BCA 527734) and the NIHR UK Exeter Biomedical Research Centre.

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: EES reported consulting (unpaid) for Alnylam Pharmaceuticals and Eli Lilly, and an advisory board (unpaid) for Eisai. ZI has served as an advisor/consultant to CADTH, EISAI, Lilly, Lundbeck/Otsuka, Novo Nordisk, and Roche. ZI is an Editorial Board Member of this journal but was not involved in the peer-review process of this article nor had access to any information regarding its peer-review. The remaining authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data used in this study were obtained from CCNA (). Data are available free of charge to bona fide researchers who submit a research proposal.

ORCID iDs

Dinithi Mudalige

Zahinoor Ismail

References

Deckers

van Boxtel

Schiepers

, et al. Target risk factors for dementia prevention: a systematic review and Delphi consensus study on the evidence from observational studies. Int J Geriatr Psychiatry 2015; 30: 234–246.

Power

Mormino

Soldan

, et al. Combined neuropathological pathways account for age-related risk of dementia. Ann Neurol 2018; 84: 10–22.

Boyle

Leurgans

, et al. Attributable risk of Alzheimer's dementia attributed to age-related neuropathologies. Ann Neurol 2019; 85: 114–124.

Kapasi

DeCarli

Schneider

. Impact of multiple pathologies on the threshold for clinically overt dementia. Acta Neuropathol 2017; 134: 171–186.

Chojdak-Łukasiewicz

Dziadkowiak

Zimny

, et al. Cerebral small vessel disease: a review. Adv Clin Exp Med 2021; 30: 349–356.

Akoudad

Wolters

Viswanathan

, et al. Association of cerebral microbleeds with cognitive decline and dementia. JAMA Neurol 2016; 73: 934–943.

Jellinger

Attems

. Neuropathology and general autopsy findings in nondemented aged subjects. Clin Neuropathol 2012; 31: 87–98.

Schneider

Bennett

. Where vascular meets neurodegenerative disease. Stroke 2010; 41: S144–S146.

White

Edland

Hemmy

, et al. Neuropathologic comorbidity and cognitive impairment in the Nun and Honolulu-Asia aging studies. Neurology 2016; 86: 1000–1008.

10.

Liu

, et al. Cerebrolysin alleviates cognitive deficits induced by chronic cerebral hypoperfusion by increasing the levels of plasticity-related proteins and decreasing the levels of apoptosis-related proteins in the rat hippocampus. Neurosci Lett 2017; 651: 72–78.

11.

Duering

Biessels

Brodtmann

, et al. Neuroimaging standards for research into small vessel disease—advances since 2013. Lancet Neurol 2023; 22: 602–618.

12.

Wardlaw

Smith

Biessels

, et al. Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 2013; 12: 822–838.

13.

Francis

Ballerini

Wardlaw

. Perivascular spaces and their associations with risk factors, clinical disorders and neuroimaging features: a systematic review and meta-analysis. Int J Stroke 2019; 14: 359–371.

14.

Debette

Schilling

Duperron

M-G

, et al. Clinical significance of magnetic resonance imaging markers of vascular brain injury: a systematic review and meta-analysis. JAMA Neurol 2019; 76: 81–94.

15.

Smeijer

Ikram

Hilal

. Enlarged perivascular spaces and dementia: a systematic review. J Alzheimers Dis 2019; 72: 247–256.

16.

Jose

Adlin

Hugo

, et al. MRI-visible perivascular spaces and risk of incident dementia. Neurology 2022; 99: e2561.

17.

Ismail

Smith

Geda

, et al. Neuropsychiatric symptoms as early manifestations of emergent dementia: provisional diagnostic criteria for mild behavioral impairment. Alzheimers Dement 2016; 12: 195–202.

18.

Wise

Rosenberg

Lyketsos

, et al. Time course of neuropsychiatric symptoms and cognitive diagnosis in national Alzheimer's coordinating centers volunteers. Alzheimers Dement (Amst) 2019; 11: 333–339.

19.

Matsuoka

Ismail

Narumoto

. Prevalence of mild behavioral impairment and risk of dementia in a psychiatric outpatient clinic. J Alzheimers Dis 2019; 70: 505–513.

20.

Ismail

McGirr

Gill

, et al. Mild behavioral impairment and subjective cognitive decline predict cognitive and functional decline. J Alzheimers Dis 2021; 80: 459–469.

21.

McGirr

Nathan

Ghahremani

, et al. Progression to dementia or reversion to normal cognition in mild cognitive impairment as a function of late-onset neuropsychiatric symptoms. Neurology 2022; 98: e2132–e2139.

22.

Ghahremani

Smith

Ismail

. Improving dementia prognostication in cognitively normal older adults: conventional versus novel approaches to modelling risk associated with neuropsychiatric symptoms. Br J Psychiatry 2025; 226: 129–136.

23.

Johansson

Stomrud

Insel

, et al. Mild behavioral impairment and its relation to tau pathology in preclinical Alzheimer’s disease. Transl Psychiatry 2021; 11: 76.

24.

Lussier

Pascoal

Chamoun

, et al. Mild behavioral impairment is associated with β-amyloid but not tau or neurodegeneration in cognitively intact elderly individuals. Alzheimers Dement 2020; 16: 192–199.

25.

Creese

Arathimos

Brooker

, et al. Genetic risk for Alzheimer's disease, cognition, and mild behavioral impairment in healthy older adults. Alzheimers Dement (Amst) 2021; 13: e12164.

26.

Ismail

Leon

Creese

, et al. Optimizing detection of Alzheimer’s disease in mild cognitive impairment: a 4-year biomarker study of mild behavioral impairment in ADNI and MEMENTO. Mol Neurodegener 2023; 18: 50.

27.

Miao

Chen

H-Y

Gill

, et al. Plasma β-amyloid in mild behavioural impairment – neuropsychiatric symptoms on the Alzheimer’s continuum. J Geriatr Psychiatry Neurol 2021; 35: 434–441.

28.

Ghahremani

Wang

Chen

, et al. Plasma phosphorylated tau at threonine 181 and neuropsychiatric symptoms in preclinical and prodromal Alzheimer disease. Neurology 2023; 100: e683–e693.

29.

Naude

Wang

Leon

, et al. Tau-PET in early cortical Alzheimer brain regions in relation to mild behavioral impairment in older adults with either normal cognition or mild cognitive impairment. Neurobiol Aging 2024; 138: 19–27.

30.

Matsuoka

Imai

Narumoto

. Neuroimaging of mild behavioral impairment: a systematic review. PCN Rep 2023; 2: e81.

31.

Guan

Rehman

Nathan

, et al. Neuropsychiatric symptoms: risk factor or disease marker? A study of structural imaging biomarkers of Alzheimer's disease and incident cognitive decline. Hum Brain Mapp 2024; 45: e70016.

32.

Sharif

Guan

Bodnar

, et al. Neuropsychiatric symptoms and progression to pathologically confirmed Alzheimer's disease. Brain 2025: 1–15. Epub ahead of print 25 April 2025. DOI: 10.1093/brain/awaf156.

33.

Ghahremani

Leon

Smith

, et al. Exploring the association between mild behavioral impairment and plasma p-tau217: implications for early detection of Alzheimer's disease. Alzheimers Dement (Amst) 2025; 17: e70119.

34.

Matsuoka

Narumoto

Morii-Kitani

, et al. Contribution of amyloid and putative Lewy body pathologies in neuropsychiatric symptoms. Int J Geriatr Psychiatry 2023; 38: e5993.

35.

Orso

Mattei

Arnaldi

, et al. Clinical and MRI predictors of conversion from mild behavioural impairment to dementia. Am J Geriatr Psychiatry 2020; 28: 755–763.

36.

Smith

Crites

Wang

, et al. Cerebral amyloid angiopathy is associated with emotional dysregulation, impulse dyscontrol, and apathy. J Am Heart Assoc 2021; 10: e022089.

37.

Cao

Sun

, et al. Association of cerebral small vessel disease burden with neuropsychiatric symptoms in non-demented elderly: a longitudinal study. J Alzheimers Dis 2022; 89: 583–592.

38.

Dhar

Barton

. Depression and the link with cardiovascular disease. Front Psychiatry 2016; 7: 33.

39.

Eurelings

Jaccard

Moll van Charante

, et al. The mediating role of cardiovascular risk factors in the relationship between symptoms of apathy and incident cardiovascular disease in community-dwelling older individuals. Int Psychogeriatr 2016; 28: 669–679.

40.

Stella

Pais

Loureiro

, et al. Neuropsychiatric symptoms and cerebrovascular risk in non-demented elders: cross-sectional study using the mild behavioural impairment checklist (MBI-C). Psychogeriatrics 2022; 22: 55–66.

41.

Guan

Aundhakar

Tomaszewski Farias

, et al. Vascular risk factor associations with subjective cognitive decline and mild behavioural impairment. Brain Commun 2025; 7: fcaf163.

42.

Soo

Wong

, et al. The association between diabetes mellitus and mild behavioral impairment among mild cognitive impairment: findings from Singapore. J Alzheimers Dis 2021; 82: 411–420.

43.

Trolli

Cipollini

Moci

, et al. Perivascular unit: this must be the place. The anatomical crossroad between the immune, vascular and nervous system. Front Neuroanat 2020; 14: 17.

44.

Miao

Chen

H-Y

Robert

, et al. White matter hyperintensities and mild behavioral impairment: findings from the MEMENTO cohort study. Cereb Circ Cogn Behav 2021; 2: 100028.

45.

Chertkow

Borrie

Whitehead

, et al. The comprehensive assessment of neurodegeneration and dementia: Canadian cohort study. Can J Neurol Sci 2019; 46: 499–511.

46.

Mohaddes

Das

Abou-Haidar

, et al. National neuroinformatics framework for Canadian consortium on neurodegeneration in aging (CCNA). Front Neuroinform 2018; 12: 85.

47.

Smith

Duchesne

Gao

, et al. Vascular contributions to neurodegeneration: protocol of the COMPASS-ND study. Can J Neurol Sci 2021; 48: 799–806.

48.

Ismail

Agüera-Ortiz

Brodaty

, et al. The mild behavioral impairment checklist (MBI-C): a rating scale for neuropsychiatric symptoms in pre-dementia populations. J Alzheimers Dis 2017; 56: 929–938.

49.

Patten

Charlton

, et al. Validating the mild behavioral impairment checklist in a cognitive clinic: comparisons with the neuropsychiatric inventory questionnaire. J Geriatr Psychiatry Neurol 2023; 36: 107–120.

50.

Duchesne

Chouinard

Potvin

, et al. The Canadian dementia imaging protocol: harmonizing national cohorts. J Magn Reson Imaging 2019; 49: 456–465.

51.

Smith

Duchesne

Gao

, et al. Vascular contributions to neurodegeneration: protocol of the COMPASS-ND study. Can J Neurol Sci 2021; 48: 799–806.

52.

Barisano

Law

Custer

, et al. Perivascular space imaging at ultrahigh field MR imaging. Magn Reson Imaging Clin N Am 2021; 29: 67–75.

53.

Potter

Chappell

Morris

, et al. Cerebral perivascular spaces visible on magnetic resonance imaging: development of a qualitative rating scale and its observer reliability. Cerebrovasc Dis 2015; 39: 224–231.

54.

Gwet

. Computing inter-rater reliability and its variance in the presence of high agreement. Br J Math Stat Psychol 2008; 61: 29–48.

55.

Kassam

Chen

Nosheny

, et al. Cognitive profile of people with mild behavioral impairment in brain health registry participants. Int Psychogeriatr 2023; 35: 643–652.

56.

Ghahremani

Nathan

Smith

, et al. Functional connectivity and mild behavioral impairment in dementia-free elderly. Alzheimers Dement (N Y) 2023; 9: e12371.

57.

R Core Team . R: A Language and Environment for Statistical Computing. Austria: R Foundation for Statistical Computing, 2022.

58.

Clancy

Gilmartin

Jochems

, et al. Neuropsychiatric symptoms associated with cerebral small vessel disease: a systematic review and meta-analysis. Lancet Psychiatry 2021; 8: 225–236.

59.

Kan

Gyanwali

Hilal

, et al. Neuropsychiatric correlates of small vessel disease progression in incident cognitive decline: independent and interactive effects. J Alzheimers Dis 2020; 73: 1053–1062.

60.

Rennels

Gregory

Blaumanis

, et al. Evidence for a ‘paravascular’ fluid circulation in the mammalian central nervous system, provided by the rapid distribution of tracer protein throughout the brain from the subarachnoid space. Brain Res 1985; 326: 47–63.

61.

Iliff

Wang

Liao

, et al. A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid β. Sci Transl Med 2012; 4: 147ra111.

62.

Diem

MacGregor Sharp

Gatherer

, et al. Arterial pulsations cannot drive intramural periarterial drainage: significance for Aβ drainage. Front Neurosci 2017; 11: 274820.

63.

Wardlaw

Benveniste

Nedergaard

, et al. Perivascular spaces in the brain: anatomy, physiology and pathology. Nat Rev Neurol 2020; 16: 137–153.

64.

Ineichen

Okar

Proulx

, et al. Perivascular spaces and their role in neuroinflammation. Neuron 2022; 110: 3566–3581.

65.

Yao

Gao

Bai

, et al. Enlarged perivascular spaces, neuroinflammation and neurological dysfunction in NMOSD patients. Front Immunol 2022; 13: 966781.

66.

Filiano

Gadani

Kipnis

. How and why do T cells and their derived cytokines affect the injured and healthy brain? Nat Rev Neurosci 2017; 18: 375–384.

67.

Gouveia-Freitas

Bastos-Leite

. Perivascular spaces and brain waste clearance systems: relevance for neurodegenerative and cerebrovascular pathology. Neuroradiology 2021; 63: 1581–1597.

68.

Esch

Stefano

. Proinflammation: a common denominator or initiator of different pathophysiological disease processes. Signature 2002; 8: 9.

69.

Esch

Stefano

Fricchione

, et al. The role of stress in neurodegenerative diseases and mental disorders. Neuro Endocrinol Lett 2002; 23: 199–208.

70.

Büttiker

Weissenberger

Esch

, et al. Dysfunctional mitochondrial processes contribute to energy perturbations in the brain and neuropsychiatric symptoms. Front Pharmacol 2023; 13: 1095923.

71.

Clark

Richiardi

Maréchal

, et al. Systemic and central nervous system neuroinflammatory signatures of neuropsychiatric symptoms and related cognitive decline in older people. J Neuroinflammation 2022; 19: 127.

72.

Hansen

Cain

Thomas

, et al. Dilated perivascular spaces in the Basal Ganglia are a biomarker of small-vessel disease in a very elderly population with dementia. AJNR Am J Neuroradiol 2015; 36: 893–898.

73.

Zhuo

Peng

, et al. The correlation between enlarged perivascular spaces and cognitive impairment in Parkinson’s disease and vascular parkinsonism. BMC Neurol 2022; 22: 282.

74.

Yamasaki

Ikawa

Ichihara

, et al. Factors associated with the location of perivascular space enlargement in middle-aged individuals undergoing brain screening in Japan. Clin Neurol Neurosurg 2022; 223: 107497.

75.

Charidimou

Meegahage

Fox

, et al. Enlarged perivascular spaces as a marker of underlying arteriopathy in intracerebral haemorrhage: a multicentre MRI cohort study. J Neurol Neurosurg Psychiatry 2013; 84: 624–629.

76.

Charidimou

Hong

Jäger

, et al. White matter perivascular spaces on magnetic resonance imaging. Stroke 2015; 46: 1707–1709.

77.

Charidimou

Boulouis

Pasi

, et al. MRI-visible perivascular spaces in cerebral amyloid angiopathy and hypertensive arteriopathy. Neurology 2017; 88: 1157–1164.

78.

Banerjee

Kim

Fox

, et al. MRI-visible perivascular space location is associated with Alzheimer's disease independently of amyloid burden. Brain 2017; 140: 1107–1116.

79.

Jeong

Cha

Park

, et al. Association of enlarged perivascular spaces with amyloid burden and cognitive decline in Alzheimer disease continuum. Neurology 2022; 99: e1791–e1802.

80.

Barnes

Ballerini

Valdés Hernández

MdC

, et al. Topological relationships between perivascular spaces and progression of white matter hyperintensities: a pilot study in a sample of the lothian birth cohort 1936. Front Neurol 2022; 13: 889884.