Introduction
Neuropsychiatric symptoms are highly prevalent in cognitive decline and dementia (Geda et al., Reference Geda, Roberts, Knopman, Petersen, Christianson, Pankratz, Smith, Boeve, Ivnik, Tangalos and Rocca2008; Steinberg et al., Reference Steinberg, Shao, Zandi, Lyketsos, Welsh-Bohmer, Norton, Breitner, Steffens and Tschanz2008). Since neuropsychiatric symptoms are associated with higher risk of institutionalisation and even death, they are an extremely important factor to consider in clinical settings (Okura et al., Reference Okura, Plassman, Steffens, Llewellyn, Potter and Langa2011).
Both cognitive decline and neuropsychiatric symptoms occur in cerebral small vessel disease (SVD). The consistent associations between SVD brain pathology (white matter hyperintensities [WMH], lacunes, cerebral microbleeds, and brain atrophy) and cognitive impairment have been established in previous studies (for a recent review see Hamilton et al., Reference Hamilton, Backhouse, Janssen, Jochems, Maher, Ritakari, Stevenson, Xia, Deary and Wardlaw2021). More severe WMH have been linked to apathy, fatigue and delirium as well as overall severity of neuropsychiatric symptoms (Clancy et al., Reference Clancy, Gilmartin, Jochems, Knox, Doubal and Wardlaw2021; Kim et al., Reference Kim, Kang, Kim, Kim, Jeon, Lee, Oh, Kim, Choe, Lee, Noh, Cho, Yoon, Chin, Cummings, Lee, Na and Seo2013). Apathy has been found to relate to SVD both in terms of disrupted white matter track connectivity and worse cognitive functioning (Lohner et al., Reference Lohner, Brookes, Hollocks, Morris and Markus2017; Tay et al., Reference Tay, Tuladhar, Hollocks, Brookes, Tozer, Barrick, Husain, de Leeuw and Markus2019). Overall SVD-related brain changes have been shown to relate to depression, and an association between white matter tract damage and depression has been found in patients with symptomatic SVD including stroke (Brookes et al., Reference Brookes, Herbert, Lawrence, Morris and Markus2014; Rensma et al., Reference Rensma, van Sloten, Launer and Stehouwer2018). WMH have also been shown to predict cognitive decline and an increase in depressive symptoms over time in Alzheimer’s disease (Puzo et al., Reference Puzo, Labriola, Sugarman, Tripodis, Martin, Palmisano, Steinberg, Stein, Kowall, McKee, Mez, Killiany, Stern and Alosco2019).
Previous research has shown that neuropsychiatric symptoms are associated with cognitive impairment. In community-dwelling subjects, cognitive impairment was related to higher incidence of neuropsychiatric symptoms and specific cognitive domains were associated with specific neuropsychiatric symptoms (executive functions and visuomotor speed with hyperactivity and affective symptoms; language and visual memory with psychosis symptoms, Xu et al., Reference Xu, Ang, Hilal, Chan, Wong, Venketasubramanian, Ikram and Chen2015). In overt SVD with dementia, higher prevalence of neuropsychiatric symptoms has been found to relate to worse cognitive outcomes (Chin et al., Reference Chin, Seo, Kim, Park, Ahn, Lee, Kang and Na2012).
In a memory clinic sample of elderly subjects, who were free from dementia at baseline, increased WMH over time was associated with more neuropsychiatric symptoms, and an increase in both WMH and neuropsychiatric symptoms was related to cognitive decline over a 2-year period (Kan et al., Reference Kan, Gyanwali, Hilal, Ng, Venketasubramanian, Chen and Xu2020). Pathological brain changes such as WMH and lacunes have been associated with higher incidence of neuropsychiatric symptoms in patients with subcortical vascular impairment (Kim et al., Reference Kim, Kang, Kim, Kim, Jeon, Lee, Oh, Kim, Choe, Lee, Noh, Cho, Yoon, Chin, Cummings, Lee, Na and Seo2013). In particular, WMH were found to relate to higher incidence of apathy and higher overall frequency of neuropsychiatric symptoms.
The majority of the previous research has focused on symptomatic SVD. The combined effects of WMH and neuropsychiatric symptoms on cognitive and functional abilities are not known in the preclinical stages of SVD, where brain changes are detectible but no criteria are met for clinical diagnosis of stroke, dementia or significant disability (covert SVD).
We examined neuropsychiatric symptoms and domain-specific cognitive performance in subjects with varying degrees of WMH who were free from dementia and stroke and were independent in their basic activities of daily living. Specifically, we investigated whether (a) WMH are associated with neuropsychiatric symptoms, (b) neuropsychiatric symptoms are related to cognitive performance and instrumental activities of daily living (IADL), and (c) neuropsychiatric symptoms moderate the relationship between WMH and cognitive and functional outcomes.
Method
Subjects and design
The Helsinki Small Vessel Disease Study is a cohort study investigating covert cerebral SVD through brain imaging, clinical, and cognitive characteristics in older individuals. The study protocol has been described in detail previously (Arola et al., Reference Arola, Laakso, Pitkänen, Koikkalainen, Lötjönen, Korvenoja, Erkinjuntti, Melkas and Jokinen2021; Jokinen et al., Reference Jokinen, Laakso, Ahlström, Arola, Lempiäinen, Pitkänen, Paajanen, Sikkes, Koikkalainen, Lötjönen, Korvenoja, Erkinjuntti and Melkas2022). Briefly, the study included subjects who had recently undergone a brain scanning due to transient ischemic attack (26%), dizziness (18%), headache/migraine (10%), subjective cognitive complaints (4%), visual symptoms (11%), fall (7%), syncope (3%), or other reasons (20%). The subjects were recruited from the imaging registry of the Helsinki University Hospital in Finland between October 2016 and March 2020. A total of 152 subjects with varying degrees of WMH took part in the study including comprehensive neurological and neuropsychological assessments, self- and informant-filled questionnaires and a brain magnetic resonance imaging (MRI) with standard protocol carried out at three visits within approximately 1 month.
The inclusion criteria were: (a) age 65–75 years at the time of enrollment; (b) place of residence within the Helsinki and Uusimaa hospital district; (c) occurrence of not more than minor, temporary, and local neurological symptoms (having manifested 3 to 12 months before the enrollment), or no neurological symptoms at all; (d) functional independence in daily activities as defined by a modified Rankin Scale score (van Swieten, Koudstaal, Visser, Schouten, & van Gijn, Reference van Swieten, Koudstaal, Visser, Schouten and van Gijn1988) of 0–2; and (e) fluent Finnish language skills. The main exclusion criteria included significant neurological diseases, severe psychiatric disorders and factors that might hinder the administration of the neuropsychological tests (such as sight or hearing disabilities) or that might affect undergoing the MRI. Additional exclusion criteria based on MRI findings were defined as follows: (a) cortical infarct; (b) subcortical infarct larger than 15 mm (20 mm on diffusion-weighted images); (c) hemorrhage larger than microbleed (over 10 mm); (d) brain tumor; and (e) contusion, traumatic subarachnoid or intracranial hemorrhage, distinct diffuse axonal injury.
The current study included 134 subjects who had available evaluations of neuropsychiatric symptoms (see details below).
Ethics approval
The study was approved by the Ethics Committee of the Helsinki University Hospital and conducted according to the Declaration of Helsinki. Informed written consent was received from each subject.
Data sharing
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Magnetic resonance imaging
A 3T MRI scanner with 32-channel head coil was used for the imaging. The protocol consisted of fast three plane localizer, 3D FLAIR SPACE, 3D T2 SPACE, 3D T1 MPRAGE, 3D gradient echo susceptibility weighted imaging sequence, 3D gradient echo sequence with magnetization transfer pulse on and off.
WMH of presumed vascular origin were defined on FLAIR sequences as hyperintense areas in the white matter without cavitation (Wardlaw et al., Reference Wardlaw, Smith, Biessels, Cordonnier, Fazekas, Frayne, Lindley, O’Brien, Barkhof, Benavente, Black, Brayne, Breteler, Chabriat, Decarli, de Leeuw, Doubal, Duering, Fox and Dichgans2013). A board certified neuroradiologist first evaluated WMH visually using the modified Fazekas scale (Pantoni et al., Reference Pantoni, Basile, Pracucci, Asplund, Bogousslavsky, Chabriat, Erkinjuntti, Ferro, O’Brien, Scheltens, Visser, Wahlund, Wallin, Fazekas, Hennerici, Waldemar and Inzitari2004); 0 = none, 1 = mild, 2 = moderate, and 3 = severe, accounting for deep and subcortical WMH. Total WMH and gray matter (GM) volumes as well as periventricular and deep WMH volumes were segmented on FLAIR images with an automated multistage segmentation method based on the Expectation-Maximization algorithm (Wang et al., Reference Wang, Catindig, Hilal, Soon, Ting, Wong, Venketasubramanian, Chen and Qiu2012). Three step method described earlier was used for the segmentation (Koikkalainen et al., Reference Koikkalainen, Rhodius-Meester, Tolonen, Barkhof, Tijms, Lemstra, Tong, Guerrero, Schuh, Ledig, Rueckert, Soininen, Remes, Waldemar, Hasselbalch, Mecocci, van der Flier and Lötjönen2016). Intracranial volume was used to normalize the total WMH and GM volume (ml) and the non-normality of the distribution was accounted for using a logarithmic transformation.
Evaluation of cognitive functioning
Cognitive functioning was evaluated using a comprehensive neuropsychological assessment comprised of both paper-and-pencil and computerised tests described in detail previously (Jokinen et al., Reference Jokinen, Laakso, Ahlström, Arola, Lempiäinen, Pitkänen, Paajanen, Sikkes, Koikkalainen, Lötjönen, Korvenoja, Erkinjuntti and Melkas2022). Briefly, the following domains were evaluated; processing speed, executive functions, working memory, memory and learning, visuospatial perception, and verbal reasoning (see Supplemental Table 1 for specific cognitive tests). Domain scores were constructed as follows: (1) the distributions of the raw scores were checked for outliers and non-normality, and square root or logarithmic transformations were applied where appropriate, (2) variables were standardized into z scores based on the data of the present sample, (3) the scales were inverted, when needed, so that higher values represented better performance in all variables, (4) composite measures were calculated by taking the mean of the standardized test scores within each domain. A global cognition score consisted of a standardized mean score comprised of the six domain scores. In this study, we focused on global cognitive score as the primary measure of cognitive performance. In addition, we examined three specific cognitive domains, namely, processing speed, executive functions and memory, as the core domains most susceptible in the covert stages of SVD (Du et al., Reference Du, Zhu, Zhou, Lu, Qiu, Yu, Cao, Zhi, Yang, Xu, Sun and Zhou2020; Knopman et al., Reference Knopman, Griswold, Lirette, Gottesman, Kantarci, Sharrett, Jack, Graff-Radford, Schneider, Windham, Coker, Albert and Mosley2015; Prins et al., Reference Prins, van Dijk, den Heijer, Vermeer, Jolles, Koudstaal, Hofman and Breteler2005). Additionally, we used Montreal Cognitive Assessment (MoCA) for descriptive purposes to indicate the level of general cognitive status of the subjects (Nasreddine et al., Reference Nasreddine, Phillips, Bédirian, Charbonneau, Whitehead, Collin, Cummings and Chertkow2005).
Evaluation of neuropsychiatric symptoms
Neuropsychiatric symptoms were reported by a close informant using the Neuropsychiatric Inventory Questionnaire (NPI-Q, Cummings et al., Reference Cummings, Mega, Gray, Rosenberg-Thompson, Carusi and Gornbein1994). NPI-Q includes items on 12 neuropsychiatric symptoms (delusions, hallucinations, aggression, depression, anxiety, euphoria, apathy, disinhibition, irritability, motor disturbance, night time disruptive behaviors, and appetite). For each symptom, the first question is whether the subject has had the symptom (yes or no) in the last month. We did not include the sentence “Please answer the following questions based on changes that have occurred since the patient first began to experience memory problems” from the original instruction, since not all of our subjects had experienced memory problems. NPI-Q total score is the sum of the occurrence of the 12 symptoms. The severity of the NPI-Q ranges between 1 and 3 (1 = mild, 2 = moderate, and 3 = severe) for each neuropsychiatric symptom. A total of 134 informants filled in the NPI-Q.
Evaluation of functional abilities
Instrumental activities of daily living were used as a measure of functional abilities. They were assessed with the short version of the Amsterdam Instrumental Activities of Daily Living (A-IADL) questionnaire also filled in by a subject’s close informant (Jutten et al., Reference Jutten, Peeters, Leijdesdorff, Visser, Maier, Terwee, Scheltens and Sikkes2017). The short version includes 30 questions assessing complex activities on areas such as taking care of household chores, work or finances, using appliances or a computer, as well as leisure activities. Scoring ranges from 0 (no difficulty) to 4 (unable to do activity) based on how performance is currently compared to the past performance. Item response theory is used to get a total score and higher scores show better functioning. A total of 132 informants filled in the A-IADL.
Statistical analyses
First, we used Spearman’s correlations to examine the bivariate associations between NPI-Q total score and sex, age, and years of educations as well as the MRI volumetrics (total, periventricular, deep WMH and GM volumes). Second, we ran a logistic regressions with the MRI volumetrics as the independent variables and the neuropsychiatric symptoms (present/absent) as the dependent variable. Third, we ran linear regression analyses with the neuropsychiatric symptoms (present/absent) as the independent variable and the cognitive or functional outcomes as the dependent variable (five analyses: global cognition, processing speed, executive functioning, memory and A-IADL). Fourth, we repeated the linear regression analyses by adding in the interaction terms (neuropsychiatric symptoms [present/absent] x WMH), as the independent variables and cognitive composite scores as well as A-IADL as the dependent variables. We then repeated the third and fourth steps for the most common neuropsychiatric symptoms (depression, irritability, night time disruptive behaviours, apathy, and changes in appetite) as the independent variables. All analyses were controlled for sex, age, and education (years). The false discovery rate (FDR) correction was used for multiple analysis accounting for the five dependent variables in each set of analyses.
Results
Subject characteristics
The present study sample consisted of 134 subjects with informant-filled NPI-Q, who did not differ from those with missing data (n = 18) in sex, age, education, WMH volume, A-IADL or MoCA scores (all p-values > 0.05). Exact number of subjects also varied depending on the NPI-Q item, most had 18 missing values, but night time disruptive behaviors had 21 missing values (usually due to the informant being unsure how to answer that particular question). Demographic and clinical information can be found in Table 1. A-IADL score ranged between 43.6 and 70.0. Descriptives of the cognitive, A-IADL and NPI-Q scores in groups based on WMH severity (Fazekas score none/mild vs. moderate/severe) are presented in Supplemental Table 2. Of the subjects, 62% had no neuropsychiatric symptoms, 16% had one symptom and 22% had two or more symptoms. The most common neuropsychiatric symptoms in our sample were depression (25%), irritability (21%), night time disruptive behaviors (15%), apathy (11%) and changes in appetite (11%). The occurrence of each neuropsychiatric symptom of the NPI-Q are detailed in Figure 1. The distribution of the NPI-Q severity score was as follows: 0, n = 85 (63%); 1, n = 21 (16%); 2, n = 12 (9%); 3, n = 6 (5%); 4, n = 6 (5%), 8, n = 2 (2%), 9, n = 1 (1%), and 13, n = 1 (1%).
Table 1. Characteristics for 134 subjects with NPI-Q data
A-IADL = Amsterdam Instrumental Activities of Daily Living questionnaire; MoCA = Montreal Cognitive Assessment; WMH = white matter hyperintensities.
Figure 1. Occurrence of different neuropsychiatric symptoms (in percentages) based on the Neuropsychiatric Inventory Questionnaire (NPI-Q) from most common to least common clockwise.
Relationships between NPI-Q and MRI volumetrics
NPI-Q total score correlated significantly with total WMH volume (r s = 0.20, p = 0.019), and more specifically with periventricular WMH volume (r s = 0.20, p = 0.024) and deep WMH volume (r s = 0.19, p = 0.028), but it was not related to total GM volume (r s = 0.04), age, sex or years of education (all p > 0.05). Total WMH volume was significantly associated with the presence of neuropsychiatric symptoms (binary variable) also after adjusting for age, sex, and education (OR 2.89, CI 95% 1.13–7.43, p = 0.027).
Relationships between NPI-Q, cognitive scores and A-IADL
NPI-Q total score did not have significant bivariate correlations with cognitive composite scores (r s values between −0.08 and −0.13, all p-values > 0.05). After adjusting for sex, age, and education (years), there were no statistically significant associations between the presence of neuropsychiatric symptoms and cognitive functioning (standardized ß between −0.06 and −0.13, all p-values > 0.05). However, there were significant synergistic interactions between the presence of neuropsychiatric symptoms and WMH volume on all cognitive composite scores (Figure 2). Subjects with presence of any neuropsychiatric symptom together with higher levels of WMH volume had disproportionately low cognitive abilities. The synergistic interaction between the presence of neuropsychiatric symptoms and WMH volume on A-IADL was not significant (standardized ß −0.18, p > 0.05).
Figure 2. Interactions between white matter hyperintensities volume, presence or absence of informant-evaluated neuropsychiatric symptoms based on the Neuropsychiatric Inventory Questionnaire (NPI-Q) and cognitive functioning. Gray areas represent 95% confidence intervals. All interactions were significant also after False Discovery Rate corrections.
NPI-Q total score correlated inversely with A-IADL score (r s = −0.41, p < 0.001). After adjusting for sex, age, and education, there was a significant association between the presence of neuropsychiatric symptoms and A-IADL (standardized ß −0.37, p < 0.001). The interaction between the presence of neuropsychiatric symptoms and WMH volume on A-IADL was not significant (p-value > 0.05).
Associations between most common NPI-Q symptoms, cognitive scores and A-IADL
The presence of depression was significantly associated with global cognition (standardized ß −0.19, p < 0.014, f2 = 0.05) and A-IADL (standardized ß −0.43, p < 0.001, f2 = 0.21), but not with specific cognitive domains (standardized ß between −0.11 and −0.14, all p-values > 0.05). There were significant interactions between the presence of depression and WMH volume on all the cognitive composite scores and A-IADL (Figures 3 and 4). Presence of depression together with higher levels of WMH volume associated with disproportionately poor cognitive composite and A-IADL scores. Irrespective of WMH, the presence of apathy was significantly associated with global cognition (standardized ß −0.29, p < 0.001, f2 = 0.06), processing speed (standardized ß −0.23, p = 0.005, f2 = 0.05), executive functions (standardized ß −0.19, p = 0.017, f2 = 0.03), and memory (standardized ß −0.22, p = 0.005, f2 = 0.04) as well as A-IADL (standardized ß −0.46, p < 0.001, f2 = 0.25). The interactions between the presence of apathy, WMH and all the outcome variables were not significant (standardized ß between −0.07 and −0.19, all p-values > 0.05). The associations between the other most common neuropsychiatric symptoms (irritability, night time disruptive behaviors and changes in appetite) and cognitive and functional outcomes were statistically nonsignificant (all p-values > 0.05 after FDR correction).
Figure 3. Interactions between white matter hyperintensity (WMH) volume and depressive symptoms group (based on the informant-evaluated Neuropsychiatric Inventory Questionnaire, NPI-Q) on cognitive functioning. Gray areas represent 95 % confidence intervals. All interactions were significant also after False Discovery Rate corrections.
Figure 4. Interaction between white matter hyperintensity (WMH) volume and depressive symptoms group (based on the informant-evaluated Neuropsychiatric Inventory Questionnaire, NPI-Q) on functional abilities assessed with the Amsterdam Instrumental Activities of Daily Living questionnaire (A-IADL). Gray areas represent 95 % confidence intervals. Interaction was significant also after False Discovery Rate correction.
Discussion
We assessed the occurrence of neuropsychiatric symptoms and their relationship with MRI volumetrics and cognitive and functional abilities in covert SVD. The most common neuropsychiatric symptoms in our sample were depressive symptoms, irritability, night time disruptive behaviors, apathy, and changes in appetite. Neuropsychiatric symptoms had significant correlations with total WMH volume, and with deep and periventricular WMH volume, but not with total GM volume. Neuropsychiatric symptoms were not directly related to cognitive functioning. Instead, the presence of neuropsychiatric symptoms together with higher total WMH volume was associated with exacerbated cognitive impairment in all the cognitive composite scores (global cognition, processing speed, executive functions, and memory). Furthermore, neuropsychiatric symptoms were related to worse functional abilities.
Depressive symptoms were the most common symptoms in our sample. Depressive symptoms together with higher levels of WMH volumes were related to worse cognitive and functional abilities. Apathy symptoms were related to worse cognitive and functional performance irrespective of WMH severity. The other most common neuropsychiatric symptoms (irritability, night time disruptive behaviors, and changes in appetite) were not related to cognitive or functional outcomes.
Previous meta-analyses have shown consistent associations between SVD brain pathology (WMH, lacunes, cerebral microbleeds, and stroke) and depression, apathy, fatigue, and delirium (Clancy et al., Reference Clancy, Gilmartin, Jochems, Knox, Doubal and Wardlaw2021; Rensma et al., Reference Rensma, van Sloten, Launer and Stehouwer2018). In healthy individuals, there is tentative evidence with a small sample to suggest that WMH are related to significant preclinical neuropsychiatric symptoms (Spalletta et al., Reference Spalletta, Iorio, Vecchio, Piras, Ciullo, Banaj, Sensi, Gianni, Assogna, Caltagirone and Piras2020). We also found evidence to suggest that WMH were related to the presence of neuropsychiatric symptoms, depressive symptoms in particular, already in covert SVD without presence of significant clinical impairment. In our study, the overall occurrence of neuropsychiatric symptoms was 38%, which is similar to what is found in mild cognitive impairment (MCI) research (Mariani, Monastero, & Mecocci, Reference Mariani, Monastero and Mecocci2007).
Increasing neuropsychiatric symptoms have been previously shown to associate with SVD disease progression and cognitive decline over a 2-year period in a dementia-free elderly sample (Kan et al., Reference Kan, Gyanwali, Hilal, Ng, Venketasubramanian, Chen and Xu2020). Kan et al. found that at baseline, higher neuropsychiatric symptom incidence was associated with higher SVD-related brain pathology burden. The authors argued that neuropsychiatric symptoms could be used as a clinical marker of SVD progression. Our results were in line with these findings with significant associations between WMH volume and presence of neuropsychiatric symptoms. Furthermore, our results suggest that preclinical depressive symptoms accelerate worse cognitive impairment in covert SVD and are related to worse functional ability overall.
Previous research has identified differing results between apathy and depression in SVD. More precisely, apathy but not depression, has been shown to relate to white matter track connectivity and cognitive functioning (Hollocks et al., Reference Hollocks, Lawrence, Brookes, Barrick, Morris, Husain and Markus2015; Lohner et al., Reference Lohner, Brookes, Hollocks, Morris and Markus2017). Contrary to these previous results, we found depressive symptoms to be related to worse cognitive performance in the presence of higher levels of WMH, whereas apathy symptoms were related to worse cognitive outcomes irrespective of WMH. These previous studies used self-report questionnaires to determine the presence of apathy or depression. In our study, a close informant reported binary responses for each neuropsychiatric symptom. Neither of the evaluations represent a clinical diagnosis. The distinction between depressive and apathy symptoms in the NPI-Q is based on whether the subject seems sad or says they are depressed (depressive symptoms) or whether they seem less interested in their usual activities or in the activities and plans of others (apathy symptoms) according to the informant. Although we used solely the presence or absence of neuropsychiatric symptoms in our analyses, we still found significant associations, highlighting the importance of informant interviews already at an early stage of SVD assessments in order to get a more thorough view on a patient’s ability to function.
Limitations of this study include the relatively small sample size. With a larger sample the occurrence of neuropsychiatric symptoms would be higher and there would be more statistical power to find possible effects of other, less common neuropsychiatric symptoms. It is also important to emphasize that this study is about neuropsychiatric symptoms and not about neuropsychiatric diagnoses. Symptom severity was predominantly evaluated low by the informants, suggesting that these symptoms might not reach the level required for a clinical diagnosis. Another limitation is the cross-sectional design, which only permits inferences of how the neuropsychiatric symptoms are related to WMH at one point in time. More research is needed to examine these relationships over time in covert SVD.
A meta-analysis in MCI has suggested that neuropsychiatric symptoms in the elderly might be an early sign of brain pathology and as such should be routinely screened (Martin & Velayudhan, Reference Martin and Velayudhan2020). Apathy and irritability in particular have been linked to Alzheimer’s disease progression and cognitive decline. In Alzheimer’s disease, WMH have been shown to associate with significant cognitive impairment and increase in depressive symptoms over time, whereas no relation to an increase in informant-evaluated neuropsychiatric symptoms has been found (Puzo et al., Reference Puzo, Labriola, Sugarman, Tripodis, Martin, Palmisano, Steinberg, Stein, Kowall, McKee, Mez, Killiany, Stern and Alosco2019). Our results were different from the aforementioned previous research possibly due to differences in our sample populations (Alzheimer’s/SVD). We found neuropsychiatric symptoms to be present already in covert SVD and to have a significant role on subjects’ cognitive and functional abilities. Due to the potential to use neuropsychiatric symptoms as an early sign for future decline, more longitudinal research is needed.
Considerable amount of research exists on the topic of the “vascular depression hypothesis”, suggesting that cerebrovascular disease may “predispose, precipitate or perpetuate” some geriatric depressive syndromes (Taylor, Aizenstein, & Alexopoulos, Reference Taylor, Aizenstein and Alexopoulos2013). Cognitive impairment in these cases is common and response to antidepressant medication is often poor. Taylor, Aizenstein and Alexopoulos (Reference Taylor, Aizenstein and Alexopoulos2013) have suggested a hypothesis, wherein a disconnection due to focal vascular damage and WMH location is crucial in the potential development of vascular depression symptomatology. More recent meta-analyses support the vascular cognitive hypothesis (Rensma et al., Reference Rensma, van Sloten, Launer and Stehouwer2018; van Agtmaal et al., Reference van Agtmaal, Houben, Pouwer, Stehouwer and Schram2017). Our study gives tentative evidence to suggest these processes may commence already at the preclinical stages of SVD.
To conclude, we found that neuropsychiatric symptoms in general, and depressive symptoms in particular, associated with WMH, and importantly, seemed to accelerate cognitive impairment related to preclinical SVD. Especially depressive symptoms interact with WMH resulting in worse cognitive and functional abilities, whereas apathy relates to worse cognitive and functional outcomes irrespective of WMH. Our results imply a need to screen neuropsychiatric symptoms at an early stage of SVD assessment, as they are associated with worse cognitive and functional outcomes. Finally, our results highlight the importance of having an informant present at an evaluation, since even at the covert stage of SVD, an informant can aid in recognizing those patients that are at a higher risk for worse outcomes.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1355617722000480
Acknowledgments
We thank Dr. Sietske S. A. Sikkes, Department of Clinical, Neuro and Developmental Psychology, VU University; Alzheimer Center Amsterdam, Department of Neurology, Amsterdam Neuroscience, Amsterdam UMC, Amsterdam, The Netherlands, for providing the A-IADL scale, and Dr. Teemu Paajanen and Dr. Jussi Virkkala from the Research and Service Centre, Finnish Institute of Occupational Health, Helsinki, Finland, for providing the Flexible Attention Test, to the use of Helsinki Small Vessel Disease study.
Funding statement
The study was funded by the Helsinki and Uusimaa Hospital District, University of Helsinki and Päivikki and Sakari Sohlberg Foundation, Finland.
Conflicts of interest
JK and JL are shareholders at Combinostics Oy. JL has received lecture fees from Merck and Sanofi (paid to the employer). The other authors report no competing of interests.
Open access
