If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence to Sara A.J. van de Schraaf, MSc, Department of Internal Medicine, Geriatrics Section, Department of Care for Older People, Amsterdam UMC, Vrije Universiteit Amsterdam, De Boelelaan 1118, 1081 HV Amsterdam, the Netherlands.
Department of Internal Medicine, Geriatrics Section, Amsterdam University Medical Centers, Location VUmc, Amsterdam, the NetherlandsDepartment of Medicine for Older People, Amsterdam Public Health Research Institute, Amsterdam University Medical Centres, Location VUmc, Amsterdam, the Netherlands
Department of Internal Medicine, Geriatrics Section, Amsterdam University Medical Centers, Location VUmc, Amsterdam, the NetherlandsAlzheimer Center, Department of Neurology, Amsterdam University Medical Centers, Location VUmc, Amsterdam, the Netherlands
Department of Medicine for Older People, Amsterdam Public Health Research Institute, Amsterdam University Medical Centres, Location VUmc, Amsterdam, the Netherlands
Department of Internal Medicine, Geriatrics Section, Amsterdam University Medical Centers, Location VUmc, Amsterdam, the NetherlandsDepartment of Internal Medicine, Amstelland Hospital, Amstelveen, the Netherlands
Department of Medicine for Older People, Amsterdam Public Health Research Institute, Amsterdam University Medical Centres, Location VUmc, Amsterdam, the Netherlands
This study aimed to investigate the interrelation between slowing in walking, thinking and mood, and their relationship with cerebral small vessel disease (CSVD) in a geriatric population.
Design
Cross-sectional study.
Setting and Participants
566 geriatric outpatients from the Amsterdam Aging Cohort (49% female; age 79 ±6 years), who visited the Amsterdam UMC geriatric outpatient memory clinic.
Methods
Patients underwent a comprehensive geriatric assessment, brain imaging, and a neuropsychological assessment as part of medical care. Three slowing aspects were investigated: gait speed, processing speed, and apathy symptoms (higher scores indicating more advanced slowing). We visually rated CSVD [white matter hyperintensities (WMHs), strategic lacunes, and microbleeds] on brain imaging.
Results
Regression analyses showed that slowing in walking (gait speed) was associated with slowing in thinking [processing speed; β = 0.35, 95% confidence interval (CI) 0.22, 0.48] and slowing in mood (apathy symptoms; β = 0.21, 95% CI 0.13, 0.30), independent of important confounders. Large confluent areas of WMH (Fazekas 3) were associated with all slowing aspects: gait speed (β = 0.49, 95% CI 0.28, 0.71), processing speed (β = 0.36, 95% CI 0.19, 0.52) and apathy symptoms (β = 0.30, 95% CI 0.09, 0.51). In addition, in patients with more slowing aspects below predefined cutoffs, severe WMH was more common. Presence of ≥3 microbleeds was associated with apathy symptoms (β = 0.39, 95% CI 0.12, 0.66), whereas lacunes were not associated with slowing.
Conclusions and Implications
This study provides evidence that slowing in walking, thinking, and mood are closely related and associated with CSVD. This phenotype or geriatric syndrome could be helpful to identify and characterize patients with CSVD.
Slowing is a complex construct that acts on the continuum between normal aging and pathologic aging. The symptomatology of slowing is diverse and includes greater fatigability and lower physical and mental activity.
For instance, in healthy community-dwelling individuals, increasing age is associated with reductions in gait speed (walking), reductions in processing speed (thinking), and increased apathy (mood).
More insight into slowing is needed, as it is unclear in what manner slowing is related to normal aging or pathologic processes.
Limited evidence exists on the interrelation between different aspects of slowing in geriatric patients. A study of slowing aspects in community-dwelling older people showed that slowing in walking is associated with slowing in thinking.
Disruptions of the brain white matter network, most likely secondary to cerebral small vessel disease (CSVD), such as white matter hyperintensities (WMHs), lacunar infarcts, and microbleeds, may be partly responsible for different aspects of slowing. Several studies have shown associations between slowing in walking, thinking, and mood with vascular disease and risk factors
but have not studied all 3 slowing aspects at the same time.
If and how slowing in walking, thinking, and mood are interrelated in geriatric patients, and how slowing is associated with the presence of CSVD in this population, remains unclear. In this study, we investigated slowing in a geriatric outpatient population, in which we studied (1) the interrelation between slowing in walking, thinking, and mood and (2) their relationship with CSVD.
is an ongoing longitudinal cohort of geriatric outpatients visiting the Amsterdam UMC, location VUmc. For this study, we included patients who visited our memory clinic between February 1, 2016, and December 1, 2020. As part of routine medical care, all patients received a comprehensive geriatric assessment and underwent neuropsychological assessment. In 93% of the patients, brain imaging was available (n = 525). The Medical Ethics Committee of the Amsterdam UMC, location VUmc, approved the cohort study. All patients provided written informed consent to use clinical data for research purposes.
Materials
Slowing Aspects
Slowing in walking
We operationalized slowing in walking as gait speed. Gait speed was measured in meters per second, derived from the fastest of 2 attempts to complete a 4-m walk, a subtask of the Short Physical Performance Battery.
A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission.
Patients were instructed to walk at usual pace and allowed to use a walking aid. We scored inability to walk as 0 m/s. We inverted this score by computing −1 × score, so that a higher score represented more slowing in walking.
Slowing in thinking
We operationalized slowing in thinking as processing speed. Processing speed was measured by computing a composite z score of the raw scores of 3 neuropsychological test outcomes: the Trail Making Test, Part A
To substantiate our choice for the composite score, we performed a principal components analysis. This revealed that a first component, which we interpreted as representing processing speed, accounted for 60% of the variance in the 3 test scores; all variables had an equally high loading on this component (data not shown). A higher score represented more slowing in thinking.
Slowing in mood
We operationalized slowing in mood as apathy symptoms. Apathy symptoms were represented by the Geriatric Depression Scale 3A (GDS-3A) score (range 0-3)—a 3-item subset of the 15-item Geriatric Depression Scale (GDS-15)—which has been used previously in research settings.
It was measured by taking the sum of questions 2 (“Have you dropped many of your activities and interests?”), 9 (“Do you prefer to stay at home, rather than going out and doing new things?”), and 13 (“Do you feel full of energy?”—reverse coded).
Patients underwent magnetic resonance imaging (MRI) on a 1.5-tesla (T) or 3-T scanner or computed tomography (CT) (when MRI was not available or possible). Two trained raters visually rated the images independently (H.R. or S.v.d.S., and the neuroradiology department).
Cerebral small vessel disease
White matter hyperintensity (WMH) severity was rated on MRI T2-weighted fluid-attenuated inversion recovery or CT using the Fazekas scale
for deep white matter (range 0-3: 0 = no WMH; 1 = punctate foci; 2 = beginning confluent areas; 3 = large confluent areas). Lacunes are small infarcts of 3 to 15 mm, which appear as (hyperintense) foci on T2-weighted images or CT. For this study, we recoded lacunes located in the basal ganglia and the thalami, as these are associated with cognitive and motor dysfunction,
as present (0) or not present (1). We rated microbleeds on MRI gradient-echo sequencing images, where they appear as 2 to 10 mm hypointense punctate foci; microbleeds are not visible on CT. For this study, we dichotomized microbleeds into 0-2 or ≥3 microbleeds.
(range 1-7: 1 = less than 6 years of primary education to 7 = university degree). To display the characteristics of the population, we obtained smoking status, medication use, and medical history of cardiovascular risk factors and cardiovascular diseases from patient records. We coded the use of sedative medication and otherwise performance-reducing medication: sedative antihistamines, benzodiazepine agonists, opioids, antidepressants, and antipsychotics.
Cardiovascular risk factors included diabetes, hypertension, and hypercholesterolemia. Cardiovascular diseases included acute coronary syndrome, heart failure, cerebrovascular accident, and transient ischemic attack. Global cognitive functioning was assessed using the Mini Mental State Examination (MMSE).
Cognitive diagnosis was agreed on in a consensus meeting, including geriatricians, a neurologist, a neuropsychologist, and a psychiatrist. Mild cognitive impairment was diagnosed using the Petersen criteria,
The diagnosis of dementia due to Alzheimer's disease: Recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease.
and vascular dementia using the International Workshop of the National Institute of Neurological Disorders and Stroke (NINDS) and the Association Internationale pour la Recherche et l'Enseignement en Neurosciences (AIREN) criteria criteria.
Patients were labeled with subjective cognitive decline when cognitive complaints were not confirmed by cognitive testing and criteria for cognitive, neurologic, or psychiatric diagnoses were unmet.
Statistical Methods
Characteristics of the population
We displayed baseline characteristics for the total population. Continuous variables were displayed as mean ± SD or median (interquartile range) depending on the distribution; categorical variables were displayed as n (%). We performed intergroup comparisons for the baseline characteristics stratified for the cumulative score of slowing aspects below predefined cutoffs (slowing sum score) with analysis of variance, Kruskal Wallis test for ordinal data and non-normally distributed data, and Fisher exact test for contingency tables. For gait speed, we used the cutoff <0.8 m/s, based on the EWGSOP criteria for sarcopenia.
Because a processing speed cutoff was not readily available for the composite score or the individual measures, we choose the lowest tertile (lowest 33.3%) on the processing speed composite score as below cutoff. We performed appropriate post hoc tests (Tukey test, Dunn test, and pairwise Fisher test respectively) in case of significant results, adjusted with the Bonferroni-Holm method.
Multiple imputation
Next, we used multiple imputation using chained equations
to impute missing values. Missing data ranged from 0% for medication and disease history to 17% for microbleeds. We did not consider the data missing completely at random, but sufficiently missing at random when the other observed variables were taken into account. All displayed values from linear regressions represent the pooled results of analyses of 17 imputed data sets, based on a rule of thumb: the highest percentage of missing values as the number of imputations.
For validation, we also performed all analyses with complete cases only, which yielded similar results (data not shown).
Interrelation between slowing aspects
First, we assessed the percentage of overlap between the different slowing aspects below the predefined cutoffs. Next, we used multiple linear regression to investigate the association between the (continuous) slowing aspects, using the following predictor–outcomes structures: processing speed–gait speed, apathy–gait speed, and apathy–processing speed.
Relationship between slowing and cerebral small vessel disease
We used multiple linear regression to associate CSVD (WMH, microbleeds, and lacunes) as determinants with gait speed, processing speed, and apathy symptoms as outcomes. The CSVD determinants were dummy coded; with Fazekas score 0/1, no lacunes in the basal ganglia or thalamus, and 0-2 microbleeds as the reference category. Finally, we displayed the distribution of WMH severity (Fazekas scale) at different levels of a cumulative score slowing below predefined cutoffs (slowing sum score) in a graph.
A P value <.05 was considered statistically significant. Analyses were performed in Rstudio 1.3.959, using R, version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria). We displayed β, 95% confidence intervals (CIs), and significance levels for all linear regressions. All regression models were adjusted for age, sex (and education-level in analyses with processing speed) in model 1, additionally adjusted for depressive symptoms (on the GDS) and global cognitive functioning (MMSE score) in model 2, and sedative or otherwise performance-reducing medication in model 3.
Sensitivity analyses
Finally, we performed exploratory sensitivity analyses to assess the possibility of differential effects at different levels of cognitive functioning. Therefore, we stratified our linear regression results of adjusted model 3 (without correcting for MMSE scores) for dementia diagnosis.
Results
Characteristics of the Population
In this geriatric outpatient population (N = 566), mean age was 79.6 ±5.9 years and 49% were women (Table 1). Almost half (45%) had a higher education degree. Mean MMSE score was 25, with an interquartile range of 21-27. More than one-third had a diagnosis of Alzheimer’s disease (36%), which was the most common cognitive diagnosis, followed by mild cognitive impairment (27%) and subjective cognitive decline (15%). Nine percent of patients had vascular dementia. More than half of the population had (beginning) confluent areas of WMH (Fazekas 2 and 3; 53%), 13% had lacunes in the basal ganglia or thalamus, and 11% had ≥3 microbleeds. Between-group differences in the population stratified for slowing sum score showed that patients with higher slowing sum scores were older, used more (sedative) drugs, and more often had a history of hypertension or diabetes. In addition, higher slowing sum scores were related to lower MMSE and GDS scores and a higher prevalence of vascular dementia (Supplementary Table 1).
Table 1Characteristics of the Total Population (N = 566)
GDS-3A score, measured by taking the sum of questions 2, 9, and 13 (reverse-coded) of the 15-item Geriatric Depression Scale.
1 [0, 2]
Cerebral small vessel disease
White matter hyperintensities: Fazekas score ≥2
281 (52.8)
Lacunes, basal ganglia and thalamus
69 (13.3)
Microbleeds ≥ 3
50 (10.6)
Continuous variables are displayed as mean ± SD or median [interquartile range] depending on the distribution; categorical variables are displayed as n (%). The table is based on nonimputed data: number of missing values ranged from 0 in cardiovascular disease and risk factors and 97 in microbleeds.
∗ Based on the Dutch Verhage Scale of Education Level (range 1-7); levels 6-7 represent a higher education degree.
† Stroke: cerebrovascular accident or transient ischemic attack.
‡ Other = other dementias, neurologic and psychiatric disease.
§ Fastest of 2 attempts at a 4-m walk.
‖ GDS-3A score, measured by taking the sum of questions 2, 9, and 13 (reverse-coded) of the 15-item Geriatric Depression Scale.
There was an even overlap of slowing aspects below predefined cutoffs: all slowing aspects co-occurred at approximately the same rate. Of the total population, the prevalence of no, 1, 2, or 3 slowing aspects was 34%, 40%, 19%, and 7%, respectively.
Slowing in walking (gait speed) was associated with slowing in thinking (processing speed; β = 0.35, 95% CI 0.22, 0.48) and slowing in mood (apathy symptoms; β = 0.21, 95% CI 0.13, 0.30), independent of age, sex, educational level, depressive symptoms, MMSE score, and sedative medication (Figure 1). No association was found between apathy and processing speed (β = 0.06, 95% CI –0.01, 0.13). In Supplementary Table 2, standardized regression coefficients of corrected models 1, 2, and 3 are displayed.
Fig. 1Interrelation between slowing aspects: associations between gait speed, processing speed, and apathy symptoms. Values displayed are standardized regression coefficients (β) with 95% confidence intervals of linear regression models, from model 3 [adjusted for age, sex, education level (in the case of processing speed), depressive symptoms on the 15-item Geriatric Depression Scale, Mini Mental State Examination score, and sedative medications]. The relationships did not change when adding confounder variables; see Supplementary Table 2. ∗P < .001.
Relationship Between Slowing and Cerebral Small Vessel Disease
The presence of large confluent areas of WMH (Fazekas 3) was associated with slowing in all aspects: gait speed (β = 0.49, 95% CI 0.28, 0.71), processing speed (β = 0.36, 95% CI 0.18, 0.55), and apathy symptoms (β = 0.31, 95% CI 0.10, 0.52) independent of aforementioned confounders (model 3; Table 2). Presence of beginning confluent areas of WMH (Fazekas 2) was associated with processing speed (β = 0.22, 95% CI 0.08, 0.36). We found no associations between gait speed, processing speed and apathy symptoms with lacunes in the basal ganglia. Having ≥3 microbleeds was associated with more apathy symptoms (β = 0.39, 95% CI 0.13, 0.66).
Table 2Relationship Between Slowing Aspects and Cerebral Small Vessel Disease
Composite z score of the raw scores of 3 neuropsychological tests: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
Values displayed are standardized regression (β) coefficients with 95% confidence intervals of linear regression models. Predictor values (white matter hyperintensities, lacunes, and microbleeds were dummy-coded with Fazekas score 0/1, with no lacunes in the basal ganglia or thalamus and 0-2 microbleeds as the reference categories). Model 1: crude model adjusted for age, sex, and (in case of processing speed in regression model) education level; model 2: model 1 adjusted for depressive symptoms on the 15-item Geriatric Depression Scale and Mini Mental State Examination score; model 3: model 2 adjusted for sedative medication.
∗ Fastest of 2 attempts at a 4-m walk, inverted score by computing −1 × score, resulting in a higher score representing more slowing.
† Composite z score of the raw scores of 3 neuropsychological tests: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
‡ GDS-3A score, measured by taking the sum of questions 2, 9, and 13 (reverse-coded) of the 15-item Geriatric Depression Scale.
Figure 2 shows that in patients with a higher slowing sum score, severe WMH (Fazekas 3) was more common. Consequently, no or punctate WMH (Fazekas 0/1) was less common at higher slowing sum scores.
Fig. 2Distribution of WMH severity (Fazekas scale) according to slowing sum score. Slowing: slowing sum score represents impaired slowing aspects using the predefined cutoffs. Gait speed <0.8 m/s, GDS-3A ≥2, and the lowest tertile (33.3%) of the processing speed composite z score.
In the sensitivity analyses, effect estimates were similar to the main analyses in the groups stratified for dementia and no dementia, although confidence intervals were widened because of limited sample size (Supplementary Tables 3 and 4).
Discussion
In this study, we investigated slowing and its relation with CSVD. We provide evidence that slowing in walking, thinking, and mood are closely related. Moreover, they occur and coincide more frequently in patients with CSVD, particularly in those patients with large confluent areas of WMH.
Our finding that slowing aspects in different domains are interrelated in a geriatric outpatient population is in line with studies in community-dwelling older individuals. These studies showed that slowing in walking (gait speed) is associated to slowing in thinking (processing speed)
This could be due to the differences in cerebrovascular pathology and the differences in measurement of apathy symptoms (see limitations) between our population and the stroke populations.
This study supports the hypothesis that slowing and CSVD are closely related. All slowing aspects were associated with CSVD in the form of WMH, independent of important confounders like global cognition and depressive symptoms. This is in line with previous research.
Cerebral white matter consists of a complex network of fiber connections, which are largely composed of myelinated axons. The myelin sheath insulates axons to increase the speed of electrical signal conduction. The extent to which the brain can efficiently transfer information between regions depends on the integrity and the organization of these white matter connections.
It is therefore not surprising that white matter injury, which can be seen on MRI as confluent areas of WMH, reduces speed of signal conduction, resulting in slowing of walking, thinking, and/or mood depending on the extent and location of white matter injury. Several previous studies relating brain network properties to slowing found that specific structural subnetworks exist relating impaired connectivity in multiple white matter tracts (including those in basal ganglia and corpus callosum) to slowing in walking and thinking
Future research should focus more on the brain subnetworks in relation to the different slowing aspects.
Slowing in mood was the only slowing aspect related to microbleeds. This could be explained by the fact that apathy symptoms are not only associated with vascular disease, but are also a neuropsychiatric symptom of Alzheimer’s disease,
the most common cognitive diagnosis in our sample (36%). Microbleeds, especially lobar microbleeds, often have a different pathophysiology than WMH and lacunes, and are associated with cerebral amyloid angiopathy and Alzheimer’s disease.
This suggests that Alzheimer’s pathology rather than vascular pathology could explain the relationship between slowing in mood and microbleeds in our sample.
One might argue that slowing is a normative feature of human aging, as slowing aspects are indeed amplified in older age.
Moreover, some studies have suggested that successful aging is perceived by individuals as maintaining a positive attitude and accepting that functional changes, such as increased slowing and fatigability, are inevitable.
However, our results show that slowing is at least partly related to pathologic brain changes in the form of CSVD, even when adjusting for age as a confounding factor. As CSVD is associated with lifestyle behaviors, some degree of slowing should be preventable within normal aging.
Slowing is possibly closely related to fatigue in older adults. Fatigability increases with normal ageing and impacts physical, mental, and cognitive function owing to a lack of available energy.
In addition, higher use of sedatives in the patients with more slowing could cause increased fatigability (Supplementary Table 1), although adjusting for sedative medication use did not change our results. Alternatively, higher use of sedatives could represent overall higher incidence of (psychiatric) comorbidities. Because fatigability is increased in individuals at higher risk of cardiovascular disease
it could be hypothesized that fatigability is a mediating factor in the relationship between CSVD and slowing. Unfortunately, investigating the complex relationship between fatigability and slowing was beyond the scope of our study. Future studies should address the relationship of increased fatigability with the concept of slowing.
A strength of our study is that our large cohort of geriatric outpatients has been extensively phenotyped in a standardized manner, including a geriatric and neuropsychological assessment and brain imaging. In addition, all patients were included after being referred to our outpatient memory clinic, making our cohort as “real-life” as possible. There are several limitations to this study. First, we measured slowing in mood with 3 questions of the GDS, the GDS-3A. Although the GDS-3A has been used repeatedly in studies on apathy,
it might have limited sensitivity to detect all aspects of apathy, which is increasingly considered a neuropsychiatric symptom distinct from depression.
Future studies with more elaborate testing of apathy symptoms with other methods should verify the robustness of our results. In addition, we used a 4-m walking test as a measure of psychomotor slowing (gait speed). A recent study has proposed the Moberg Picking-up Test as an alternative to measuring psychomotor slowing.
This test needs less space to administer and seems more suitable for patients in hospitalized settings that are more physically frail (unable to walk). Future research should consider this test as it might deal with the problem of noncompleters and therefore measurement accuracy. Another limitation might be the presence of several neurodegenerative diseases in our sample, as the true causal pathways are possibly confounded by comorbid disease and cognitive impairment. However, correcting for MMSE score and stratifying for dementia diagnosis did not materially change our results. An additional limitation is our lack of data on ethnicity, as one might expect racial or cultural differences in the expression of mood, and therefore slowing. However, it is estimated that more than 90% of our cohort is white and of European descent. We believe this homogeneity makes our results generalizable to geriatric patients of this background. Finally, we did not use quantitative, voxel-based, measures of brain volumes, which potentially decreased the power of our results. Nevertheless, visual scores have been reported to be accurate and associated with cognitive impairment.
From a clinical perspective, our findings suggest that the presence of slowing in one aspect could prompt a health care specialist to be aware of the presence of other slowing aspects, and actively search for them. On top of that, slowing in any domain might be a phenotype of older people with CSVD, especially WMH severity. Our findings can be helpful in settings where state-of-the-art neuroimaging or other diagnostic tools are unavailable, as in primary care or nursing homes, as a marker of CSVD. On the other hand, performing neuroimaging in patients with slowing would provide a possible explanation of the symptoms, and thus treatment options. Presence of slowing, with accompanying CSVD on neuroimaging, could entice the clinician to provide targeted treatment and care. This treatment and care should focus on cardiovascular risk management to prevent progression of CSVD and guidance for patients and caregivers on how to handle the problems arising from slowing. Furthermore, health care specialists should be aware of the possibility of slowing when neuroimaging shows presence of moderate or severe WMH. Patients and their informal caregivers could benefit from more information about the “hidden” consequences of CSVD, like slowing. Especially, slowing in mood, or apathy, is an often misinterpreted symptom that is associated with high caregiver burden.
This study provides initial evidence for the presence of a phenotype of “slowing,” and potentially a geriatric syndrome seen in CSVD. The term geriatric syndrome is used for highly prevalent, co-occurring symptoms with shared risk factors that are present in older people, especially in cognitively or functionally impaired individuals, but that do not fit in traditional disease categories.
Another core feature of geriatric syndromes is that they predict future adverse outcomes. Although individual slowing aspects have been related to adverse outcomes,
no studies investigated the relationship between adverse outcomes and slowing in multiple domains. Future research should aim to investigate adverse outcomes related to the slowing phenotype. In the meantime, health care professionals in different settings should be aware of slowing as a marker of CSVD.
Acknowledgments
The authors thank Peter Alders, Greetje Asma, Gerda Bolink, Anouk Burger, Elske Gieteling, Bart Homan, Rosa de Jager, Emma Kleipool, Gooke Lagaay, Petra Moens, Astrid Verburg-Bakker, Barbara Verhaar, and Kathelijn Versteeg for their efforts and contribution to the Amsterdam Aging Cohort.
Supplementary Data
Supplementary Table 1Baseline Characteristics Stratified for Slowing Sum Score
Intergroup comparisons were done with analysis of variance, Kruskal-Wallis or Fisher exact test where appropriate with post hoc tests adjusted with the Bonferroni-Holm method for multiple comparisons.
Based on the Dutch Verhage scale of education level (range 1-7), levels 6-7 represent a higher education degree. Measured on ordinal scale: intergroup differences calculated with Kruskal-Wallis test.
Based on the Dutch Verhage scale of education level (range 1-7), levels 6-7 represent a higher education degree. Measured on ordinal scale: intergroup differences calculated with Kruskal-Wallis test.
Other: other dementias, neurological and psychiatric diseases.
15 (7.9)
27 (12.0)
26 (23.6)
3 (7.3)
ACS, acute coronary syndrome; AD, Alzheimer’s dementia; MCI, mild cognitive impairment; MMSE, Mini Mental State Examination; ns, not significant after post hoc testing; SCD, subjective cognitive decline; VAD, vascular dementia.
Continuous variables are displayed as mean ± SD or median [interquartile range] depending on the distribution; categorical variables are displayed as n (%). Boldface indicates significance. Slowing sum score was calculated by adding the number of impaired slowing aspects (0-3) below predefined cutoffs: gait speed < 0.8 m/s, GDS-3A ≥ 2, and the lowest tertile (33.3%) of a processing speed composite score. The table is based on nonimputed data: number of missing values ranged from 0 in cardiovascular disease and risk factors and 97 in microbleeds.
∗ Intergroup comparisons were done with analysis of variance, Kruskal-Wallis or Fisher exact test where appropriate with post hoc tests adjusted with the Bonferroni-Holm method for multiple comparisons.
† Based on the Dutch Verhage scale of education level (range 1-7), levels 6-7 represent a higher education degree. Measured on ordinal scale: intergroup differences calculated with Kruskal-Wallis test.
‡ Stroke: cerebrovascular accident or transient ischemic attack.
§ Other: other dementias, neurological and psychiatric diseases.
Composite z score of the raw scores of 3 neuropsychological test outcomes: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
Values displayed are standardized regression coefficients (β) with 95% confidence intervals of linear regression models. Model 1: crude model adjusted for age, sex, and (in case of processing speed in regression model) education level; model 2: model 1 adjusted for depressive symptoms on the 15-item Geriatric Depression Scale and Mini Mental State Examination score; model 3: model 2 adjusted for sedative medication.
∗ Fastest of 2 attempts at a 4-m walk, inverted score by computing −1 × score resulting in a higher score representing more slowing.
† Composite z score of the raw scores of 3 neuropsychological test outcomes: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
‡ P < .001.
§ GDS-3A score, measured by taking the sum of questions 2, 9, and 13 (reverse-coded) of the 15-item Geriatric Depression Scale.
Composite z score of the raw scores of 3 neuropsychological test outcomes: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
Values displayed are standardized regression coefficients (β) with 95% confidence intervals of linear regression models. All linear regression models were adjusted for age, sex, education level (in case of processing speed in regression model), depressive symptoms on the 15-item Geriatric Depression Scale, and sedative medication.
∗ Fastest of 2 attempts at a 4-m walk, inverted score by computing −1 × score.
† Composite z score of the raw scores of 3 neuropsychological test outcomes: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
‡ P < .001.
§ GDS-3A score, measured by taking the sum of questions 2, 9, and 13 (reverse coded) of the 15-item Geriatric Depression Scale.
Composite z score of the raw scores of 3 neuropsychological tests: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
Values displayed are standardized regression coefficients (β) with 95% confidence intervals of linear regression models. Predictor values (white matter hyperintensities, lacunes, and microbleeds were dummy coded with Fazekas score 0/1, with no lacunes in the basal ganglia or thalamus and 0-2 microbleeds as the reference categories). All linear regressions models were adjusted for age, sex, education level (in case of processing speed in regression model), depressive symptoms on the 15-item Geriatric Depression Scale and sedative medication.
∗ Fastest of 2 attempts at a 4-m walk, inverted score by computing −1 × score, resulting in a higher score representing more slowing.
† Composite z score of the raw scores of 3 neuropsychological tests: the Trail Making Test, Part A; the Stroop I, Word Card; and the Stroop II, Color Card.
‡ GDS-3A score, measured by taking the sum of questions 2, 9, and 13 (reverse coded) of the 15-item Geriatric Depression Scale.
A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission.
The diagnosis of dementia due to Alzheimer's disease: Recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease.
00678-2&id=gr1.jpg){kind=link}
00678-2&id=gr2.jpg){kind=link}