The metabolic syndrome in a memory clinic population: Relation with clinical profile and prognosis

Journal of the Neurological Sciences 351 (2015) 18–23

Contents lists available at ScienceDirect

Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns

The metabolic syndrome in a memory clinic population: Relation with clinical profile and prognosis Lieza G. Exalto b,i,⁎, Wiesje M. van der Flier c,i, Caroline J.M. van Boheemen d, L. Jaap Kappelle b, Hugo Vrenken e,f,i, Charlotte Teunissen g,i, Ted Koene h,i, Phillip Scheltens a,i, Geert Jan Biessels b a

Department of Neurology, VU University Medical Centre, Amsterdam, The Netherlands Department of Neurology, Brain Centre Rudolf Magnus Institute, University Medical Centre Utrecht, Utrecht, The Netherlands Department of Epidemiology and Biostatistics, VU University, Amsterdam, The Netherlands d Department of Neurology, Medical Centre Haaglanden, The Hague, The Netherlands e Department of Radiology, VU University Medical Centre, Amsterdam, The Netherlands f Department of Physics and Medical Technology, VU University Medical Centre, Amsterdam, The Netherlands g Neurochemistry Lab and Biobank, Department of Clinical Chemistry, VU University Medical Centre, Amsterdam, The Netherlands h Department of Medical Psychology, VU University Medical Centre, Amsterdam, The Netherlands i Alzheimer Centre, VU University Medical Centre, Amsterdam, The Netherlands b c

a r t i c l e

i n f o

Article history: Received 14 August 2014 Received in revised form 21 January 2015 Accepted 2 February 2015 Available online 9 February 2015 Keywords: Metabolic syndrome Cognitive performance Dementia Vascular lesions Atrophy Cerebrospinal fluid (CSF) biomarkers Progression

a b s t r a c t Background: The metabolic syndrome (MetS) refers to a cluster of cardiovascular risk factors that is associated with an increased risk of cognitive impairment and dementia. It is unclear however, if the presence of the MetS is associated with a particular clinical profile or a different prognosis in patients with cognitive complaints or early dementia. Objectives: To compare 1) the clinical profile and 2) the prognosis of patients attending a memory clinic according to the presence or absence of MetS. Design: Longitudinal cohort. Setting: Memory clinic. Participants: We included and followed 86 consecutive patients (average age of 66.7 (SD 9.7)) from the Amsterdam Dementia Cohort with an MMSE N 22. Measurements: Clinical profile (neuropsychological examination, brain MRI, cerebrospinal fluid (CSF) biomarkers, clinical diagnosis) on an initial standardized diagnostic assessment was compared according to MetS status. Progression to dementia was assessed in initially nondemented patients (subjective complaints n = 40, mild cognitive impairment n = 24, follow-up available in 59). Results: 35 (41%) patients met the MetS criteria. Demographics were similar between patients with or without the MetS. At baseline, diagnosis, cognitive performance, severity of degenerative or vascular abnormalities on MRI, and CSF amyloid and tau levels did not differ between the groups (all p N 0.05). Among nondemented patients, however, MetS was associated with worse performance on executive function, attention & speed and visuoconstructive ability (z-scores, p b 0.05). During a mean follow-up of 3.4 years a similar proportion of patients with (4; 17%) and without (6; 17%) the MetS progressed to dementia (p = 0.45). Conclusion: Among nondemented patients presenting at a memory clinic MetS was associated with slightly worse cognitive performance (worse on tasks assessing executive functions, visuo-constructive ability, attention & speed), but conversion rate to dementia was not increased. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Cardiovascular risk factors are associated with an increased risk of dementia, which accounts for both Alzheimer's disease and vascular dementia (review: [1]). Cardiovascular risk factors frequently co-occur. ⁎ Corresponding author at: Department of Neurology, G03.228, University Medical Centre, PO Box 85500, 3508 GA Utrecht, The Netherlands. Fax: +31 30 2542100. E-mail address: [email protected] (L.G. Exalto).

http://dx.doi.org/10.1016/j.jns.2015.02.004 0022-510X/© 2015 Elsevier B.V. All rights reserved.

This clustering of risk factors has been captured by the concept of metabolic syndrome (MetS), which is defined as the presence of three or more of the following components: impaired glucose tolerance, obesity, hypertension, hypertriglyceridemia, and reduced high-density lipoprotein cholesterol [2]. The MetS and each of its components are associated with impaired cognitive performance and an increased risk of dementia (reviews: [3,1]). The strength of the association between MetS and dementia is dependent on the number of MetS components that is present in an

L.G. Exalto et al. / Journal of the Neurological Sciences 351 (2015) 18–23

individual [4], where the cumulative influence of multiple components is greater than the sum of its individual components [5]. The underlying mechanism of the association between MetS and dementia is, however, still uncertain. Epidemiological studies consistently link MetS with vascular dementia, while contrasting findings exist on the presence of an association between MetS and Alzheimer's disease (review: [6]). Hence, the association between MetS and dementia might be due to an acceleration of neurodegenerative changes, vascular lesions, or both. Furthermore, it is uncertain if the presence of MetS is associated with a faster progression of cognitive impairment [7,8]. Unravelling the nature of the relation between MetS and cognitive dysfunction will help in the development of potential prevention or treatment strategies. The first objective of this study was to determine whether patients with or without the MetS differ in their clinical profile (i.e. clinical diagnosis, cognitive performance, brain MRI abnormalities and cerebrospinal fluid (CSF) biomarkers) at first presentation at a memory clinic. The second aim was to evaluate whether the MetS is associated with the prognosis of these patients. 2. Patients and methods

19

available for n = 45. Patients who did not undergo a lumbar puncture did not differ in sociodemographic characteristics, MMSE score and clinical diagnosis from those who did. 2.3. Definition of the metabolic syndrome MetS was defined according to the National Cholesterol Education Program/Adult Treatment Panel III (NCEP/ATPIII) criteria revised AHA/ NHLBI [2]. Patients were considered to have MetS if they met three or more of the following five criteria: 1) plasma glucose ≥ 6.1 mmol/l or antidiabetic drug use, 2) triglyceride concentration ≥ 1.69 mmol/l or antihypertriglyceridemia drug use; 3) high-density lipoprotein cholesterol b 1.29 mmol/l in women and b 1.03 mmol/l in men; 4) blood pressure ≥ 130/85 mm Hg or antihypertensive drug use; and 5) obesity, because waist circumference was not measured at the time of data acquisition, it was substituted by BMI N 27 in women and BMI N 29 in men [13]. Even with missing data on one or two MetS components, the presence or absence of MetS could be established for patients in whom at least three risk factors from the metabolic syndrome could be classified as absent (without MetS) or three as present (MetS).

2.1. Study population 2.4. Neuropsychological examination Patients were included from the memory clinic based Amsterdam Dementia Cohort of the VU University Medical Centre (VUmc). Patients received a standardized one day diagnostic assessment including an interview on medical history, medication use and education level, a physical examination, neurological examination, extensive neuropsychological testing, an MRI-scan of the brain and blood tests. Clinical diagnoses were made at a multidisciplinary consensus meeting based on internationally established diagnostic criteria, without knowledge of the CSF biomarker results. The medical records of all patients who were evaluated at the memory clinic for the first time between 01/08/2004 and 30/06/2006 were examined and followed until 01/01/2012. Patients from the following four diagnostic categories were eligible for this study: 1) possible or probable Alzheimer's disease [9], 2) vascular dementia [10], 3) mild cognitive impairment [11], and 4) subjective complaints without cognitive impairment on neuropsychological assessment (subjective cognitive complaints, i.e. criteria for mild cognitive impairment not fulfilled and no psychiatric diagnosis). We selected all patients who were at least 45 years of age, had a Mini Mental State Examination (MMSE) score above 22 (n = 108), and from whom sufficient data was available to establish the presence or absence of MetS (i.e. either the presence or absence of at least three out of the five risk factors). This resulted in a study population consisting of 86 patients. 2.2. Biologic samples Plasma fasting glucose, high-density lipoprotein cholesterol and triglyceride levels were determined at the Department of Clinical Chemistry of the VUmc. For Apolipoprotein E (ApoE) genotyping, DNA was isolated from 10 ml EDTA blood by the QIAamp DNA blood isolation kit from Qiagen. ApoE genotype was determined with the light cycler ApoE mutation detection method (Roche Diagnostics GmbH, Mannheim, Germany). Subjects were classified as ApoE ε4 carriers if they had one or two ε4 alleles, and as non-carriers if they had no ε4 alleles. All patients who are eligible for a lumbar puncture were offered CSF analyses as part of their evaluation. CSF was obtained by lumbar puncture between the L3/L4 or L4/L5 vertebra, using a 25-gauge needle, and collected in 10 ml polypropylene tubes. CSF levels of amyloid β 1–42 (Aβ42), tau and tau phosphorylated at threonine 181 (p-tau) were measured by commercially available sandwich ELISA (Innotest β-amyloid, Innotest hTAU-Ag, and Innotest Phosphotau; Innogenetics, Ghent, Belgium), as described previously [12]. The team involved in the CSF analysis was not aware of the clinical diagnoses. CSF was

MMSE was used as measure of global cognitive function. Memory was assessed with the Rey Auditory 15-word Verbal Learning Test and the Visual Association Test [14,15]. Language was assessed using animal fluency (60 s) and Visual Association Test object naming [14]. Furthermore, speed & attention was assessed with the Trail Making Test A, the forward condition of Digit Span and Stroop cards I and II [16–18]. Executive functioning was tested by the third card of the Stroop test, the Trail Making Test B and the backwards condition of Digit Span [16–18]. Visuoconstructive abilities were tested by the Rey–Osterrieth Complex Figure Test [19]. All neuropsychological data were standardized into z-scores. The scores on the Trail Making Test and Stroop were inverted by computing − 1 ∗ z-score, because higher scores imply a worse performance. In order to create the five cognitive domains, the mean z-scores of the available tests in every domain were calculated. Depressive symptoms were assessed by the Geriatric Depression Scale, a questionnaire specifically developed as a screening instrument for the presence of depressive symptoms in older populations [20]. 2.5. Brain imaging MR imaging was performed on a 1.0-T Siemens Magnetom Impact Expert scanner (Siemens AG, Erlangen, Germany) and included coronal T1 and T2-weighted 3D MPRAGE volumes (magnetization prepared rapid acquisition gradient echo; single slab 168 slices; matrix 256 × 256; FOV 250 mm; voxel size 1 mm × 1 mm × 1.5 mm; repetition time = 15 ms; echo time = 7 ms; inversion time = 300 ms; and flip angle 15) and a fast fluid attenuated inversion recovery sequence (axial, 5-mm contiguous slice, and 1-mm in-plane resolution). Three patients did not undergo an MRI. 2.5.1. Visual rating Medial temporal lobe atrophy (MTA) was evaluated by using a visual rating scale [21] on coronal T1-weighted images based on the choroid fissure width, the temporal horn width, and the hippocampal height (possible range of scores for each side, 0 to 4). Scores of the right and left sites were averaged. Lacunes were defined as T1-hypointense and T2-hyperintense CSF-like lesions that were surrounded by white matter or subcortical gray matter and not located in areas with a high prevalence of widened perivascular spaces (e.g., anterior commissure region) [22]. White matter lesions (WML) were rated according to a slight modification of the Scheltens rating scale [23] and divided in periventricular

20

L.G. Exalto et al. / Journal of the Neurological Sciences 351 (2015) 18–23

WML (PWML) (range 0–12) and deep WML (DWML) (range 0–36). Image assessment was performed blinded to clinical information, with the use of digital image files. There was 1 missing MTA score and 1 missing WML score.

2.5.2. Brain volume Whole-brain volume was measured with an adaptation of Structural Image Evaluation, using Normalisation, of Atrophy (SIENA) called SIENAX, a fully automated technique part of FSL (for a detailed explanation see: http://www.fmrib.ox.ac.uk/analysis/research/siena/) [24]. This adapted method is described in more detail elsewhere and provides a normalized brain volume (NBV) in cm3 [25]. All individual scans, registration results, and SIENAX output were reviewed by an investigator who was blinded to the diagnosis. Twelve scans were excluded in the SIENAX process because of an inappropriate result or inadequate quality of the scan. The distribution of the diagnostic groups of these 12 patients was similar compared to the cohort overall.

2.6. Follow up Most patients had a yearly visit at the memory clinic. The patients who had not visited the memory clinic again at 1/10/2008 after their first visit (between 01/08/2004 and 30/06/2006) received a phone call to inquire about their (cognitive) well-being between 1/10/2008 and 1/7/2009. During the phone call, their cognitive performance was assessed by the Dutch version of the Telephone Interview Cognitive Status [26] of which the results were evaluated during a consensus meeting. When the telephone interview indicated possible clinically significant modification of the cognitive complaints, the patient was re-invited at the memory clinic to establish a diagnosis. In January 2012 we reviewed the medical files of all included patients again to gather the MMSE score and diagnosis from follow-up visits. A diagnosis at follow-up is known from 79 patients (92%); three patients died before a follow-up diagnoses was established and 4 were lost to follow up. The range of the duration of follow-up was 0.6–4.6 years.

2.7. Statistical analysis Statistical analyses were performed using SPSS 17.0 for windows. Because of a non-normal distribution, NBV, tau, p-tau 181, Aβ42, test results of the Stroop and Trail Making Test were log-transformed for statistical analyses. Baseline differences between subjects with and without MetS were compared with Student's t test or chi-square test. The adjusted mean difference on the cognitive performance, NBV and the CSF biomarkers by MetS status were calculated with linear regression analyses. The associations between MetS and either infarcts, or MTA, or WMLs were analysed with logistic regression. These imaging variables were dichotomized for logistic regression analyses: lacunes absent (0) or present (number ≥ 1), left/right average MTA scores b 1.5 or at least 1.5, PWML b 7 or at least 7 and DWML b 9 or at least 9. The cut-points of PWML and DWML represented the highest tertile in this dataset. The cognitive performance analyses were adjusted for age, gender, clinical diagnosis and education. The MRI and CSF biomarker analyses were adjusted for age, gender and diagnosis. The analyses were repeated for the subgroup of nondemented patients. Finally, for the nondemented patients at baseline we fitted a Cox proportional hazard model to compare the risk of conversion to dementia during follow-up by MetS status with age and gender as covariates. As post-hoc analysis, the NPO, MRI and CSF biomarker analyses were repeated with an additional adjustment for depression, measured by the Geriatric Depression Scale. The level of statistical significance was set at p b 0.05.

3. Results There were 35 (41%) patients who met the MetS criteria. Patient characteristics by MetS status are summarized in Table 1. There was no difference in age, sex, educational background and ApoE ε4 status between patients with and without the MetS. As expected the components of the MetS and a history of vascular disease and diabetes mellitus, were more common among the patients with MetS. There was no difference in smoking status (p = .34) or alcohol consumption (p = .53) between the groups (data not shown). The scores of the Geriatric Depression Scale also did not differ between patients with and without MetS. The distribution of the four clinical diagnoses — subjective memory complaints, mild cognitive impairment, Alzheimer's disease and vascular dementia — was similar between patients with (46%, 26%, 26% and 3%, respectively, n = 35,) and without MetS (47%, 29%, 22% and 2%; n = 51). The MMSE score did not differ between patients with (27.3 ± 1.7) and without (27.6 ± 1.9) MetS. Overall, patients with and without MetS performed similarly on the five cognitive domains and MMSE (Table 2). The post-hoc analyses that were additionally adjusted for depression, showed similar results (data not shown). In the subgroup of nondemented patients, however, patients with MetS performed worse on executive function, attention & speed and visuoconstructive ability. Also the compound z-score of overall cognitive function was significantly lower in the patients with MetS compared to patients without MetS. The distribution of MTA, DWML and PWML grades by MetS status is presented in Fig. 1. The proportion of patients with severe MTA, DWML or PWML was marginally, but not statistically significantly higher in patients with MetS compared to those without MetS. There was neither a difference in the presence of infarcts nor in NBV between patients with and without MetS (Table 3). Also the levels of the CSF biomarkers tau, p-tau and Aβ42 did not differ according to MetS status (Table 3). The post-hoc analyses that were additionally adjusted for depression, showed similar results (data not shown). Repeated analysis on the MRI and CSF biomarker measures in the subgroup of nondemented patients showed similar results (Table 3). For our second aim we obtained follow-up in 59 of the 64 nondemented patients at baseline, with a mean follow up time of Table 1 Baseline characteristics of patients with or without MetS. Without MetS (n = 51)

MetS (n = 35)

p value

Demographic characteristics Female Mean age, years Level of educationa (range 1–7) ApoE ε4 carriers (n = 83) Depressive symptomsb (range 0–15)

24 (47%) 66.4 ± 9.7 5 (4–6) 22 (44%) 1 (0.8–3)

10 (29%) 67.3 ± 10 5 (3–6) 12 (36%) 2 (1–4)

.09 .66 .20 .49 .17

Mets criteria High BMI (n = 75) Elevated triglycerides Low HDL-cholesterol Elevated fasting glucose (n = 84) Hypertension (n = 84)

6 (13%) 2 (4%) 0 2 (4%) 41 (82%)

6 (21%) 30 (86%) 25 (71%) 19 (56%) 33 (97%)

.38 b.00 b.00 b.00 b.00

Medical history Diabetes mellitus Atrial fibrillation Stroke Macro vascular event

1 (2%) 4 (8%) 0 (0%) 0 (0%)

9 (26%) 3 (9%) 1 (3%) 11 (31%)

b.00 .90 .23 b.00

Data are presented as number of patients with the variable present (column %), or mean ± standard deviation, or median (interquartile range). Comparison of data by t test or χ2 when appropriate and Mann Whitney test for level of education and depression scale. There was no missing data for variables were the number (n) is not specifically mentioned. HDL-cholesterol = high-density lipoprotein cholesterol. a Level of education using Verhage's classification [40]. b Depressive symptoms measured by Geriatric Depression Scale.

L.G. Exalto et al. / Journal of the Neurological Sciences 351 (2015) 18–23

21

Table 2 Cognitive performance in patients with or without MetS. Domain

z-Score

MMSE Attention & speed Memory Language Executive functions Visuoconstructive Overall cog. function

Adjusted mean difference (95% CI)

Without MetS

MetS

All

Without dementia

(N = 51)

(N = 35)

(N = 86)

(N = 64)

0.05 ± 1.04 0.09 ± 0.82 0.04 ± 0.91 0.01 ± 0.86 0.14 ± 0.84 0.08 ± 1.01 0.14 ± 0.63

−0.07 ± 0.95 −0.12 ± 0.57 0.00 ± 0.83 −0.01 ± 0.83 −0.21 ± 0.62 −0.13 ± 0.99 −0.02 ± 0.56

.03 (−.34 .40) −.12 (−.42 .17) .05 (−.20 .29) .07 (−.26 .40) −.25 (−.51 .003) −.20 (−.64 .24) −.13 (−.30 .04)

.04 (−.37 .44) −.33 (−.66 −.01)⁎ .06 (−.15 .27) .09 (−.26 .45) −.43 (−.71 −.15)⁎ −.38 (−.69 −.07)⁎ −.20 (−.36 −.04)⁎

Data are presented as mean z-scores ± standard deviation and adjusted mean difference with 95% confidence interval (95% CI). Analyses were adjusted for age, gender, clinical diagnosis and education. The mean z-scores for the subgroup without dementia is not shown. Please note that for all domains a higher z-score signifies better performance. MMSE = mini mental state examination. ⁎ p value b 0.05.

3.4 years (SD 1.1). The follow-up could be obtained from 23 (92%) with and 36 (92%) without MetS. Four (17%) patients with MetS and six (17%) without MetS converted to dementia. The fitted Cox proportional hazard model showed no significant difference in conversion to dementia between patients with and without MetS (hazard ratio (HR) 0.5; 95% CI 0.1–2.4). Follow up MMSE assessment was available from 20 (57%) patients with and 33 (65%) patients without MetS. Although the mean MMSE score decrease per year for the patients with MetS (− 1.2 SD 2.4) was almost twice as large as patients without MetS (− 0.7 SD 1.8), this difference was not statistically significant (p = .45). 4. Discussion At the first presentation at a memory clinic the clinical profile of patients did not differ according to the presence or absence of MetS. There was no difference between patients with and without MetS in clinical diagnosis, brain MRI abnormalities or CSF biomarkers. Among nondemented patients, however, patients with MetS did perform worse on tasks assessing executive functions, visuoconstructive ability, attention & speed and overall cognitive function compared to patients without MetS. In contrast, the relative frequency of progression to a diagnosis of dementia between patients with and without MetS three years after the first visit did not differ. This is the first study that describes the association of MetS and the overall clinical profile (i.e. clinical diagnosis, cognitive performance, brain MRI abnormalities and CSF biomarkers) in a memory clinic population. Previous studies reported on the association of MetS and either clinical diagnosis, cognitive performance or brain MRI abnormalities.

A.MTA

B. DWML

Our results are in line with most of these previous studies. A study of an outpatient geriatric clinic population, also observed no difference in the distribution of clinical diagnosis of cognitive complaints between patients with and without MetS [27]. In another study in subjects without dementia attending an outpatient geriatric clinic, MetS was associated with lower MMSE scores [4]. The neuropsychological profile in relation to MetS status has only been studied in population-based cohorts. These results are in line with ours. In subjects without dementia, MetS is associated with specific cognitive domains including memory, information processing speed, attention, visuospatial ability and executive functioning (review: [28]). Although the whole construct of the MetS itself has not been studied in relation to imaging markers in a memory clinic setting before, individual vascular risk factors have. Hypertension and smoking were observed to be associated with WML in patients attending a memory clinic, but diabetes and hypercholesterolemia were not [29]. Another study reported that hypertension and plasma glucose are not associated with MTA [30]. The results from previous studies on the association between MetS and progression of cognitive decline are inconsistent. The Italian Longitudinal Study on Aging showed that the presence of MetS in patients with mild cognitive impairment was associated with a higher risk of progression to dementia [7]. Another study compared the relation between MetS and cognitive decline between patients with early Alzheimer's disease with healthy controls. They showed that in healthy controls, MetS did not significantly predict cognitive performance, though higher insulin predicted poorer cognitive performance outcomes. In the Alzheimer's disease group, the higher presence of MetS components predicted better cognitive outcomes [8]. In the general

C. PWML

Fig. 1. Distributions of MTA and WML grades by MetS status. A. Distribution of MTA (0–4) grades by MetS status. The MTA score is an average score of left and right, rounded to the closest integer. White is no MTA (0) and the darker the shade the higher MTA score. B. Distribution of DWML (0–36). For clarity of presentation DWML scores are compiled in categories: 0, 1–4, 5– 8, and N8. White is no DWML and the darker the shade the higher DWML score. C. Distribution of PWML (0–12). For clarity of presentation PWML scores are compiled in categories: 0, 1–5, and N6. The darker the shade the higher DWML score. MTA = medial temporal lobe atrophy. DWML = deep white matter lesions. PWML = periventricular white matter lesion.

22

L.G. Exalto et al. / Journal of the Neurological Sciences 351 (2015) 18–23

Table 3 Brain imaging and CSF biomarkers in patients with or without MetS. All Without MetS (n = 50)

Without dementia MetS (n = 32)

(n = 64) Odds ratio (95% CI)

MRI ≥1 infarcts MTA PWML DWML

7 (14%) 0.5 (0–1) 6 (5–7) 6 (3–9)

4 (13%) 0.3 (0–1.4) 6 (6,7) 7 (4–9.8)

NBV (ml) CSF biomarkers tau pg/mla p-tau pg/mlb Aβ42 pg/mlc

1502 ± 133 (n = 30) 497 ± 320 66 ± 29 668 ± 269

1491 ± 110 (n = 15) 476 ± 293 61 ± 24 675 ± 281

0.6 (0.1 3.5) 2.1 (0.5 8.1) 1.5 (0.5 4.4) 1.8 (0.6 5.1) Adjusted mean difference (95% CI) 3.7 (−47.8 −55.3) −21.6 (−175.0 131.7) −4.3 (−19.0 10.4) 20.4 (−86.9 127.6)

0.8 (0.1 5.0) 1.7 (0.2 12.4) 1.8 (0.5 6.9) 3.3 (0.9 11.5) −7.6 (−58.4 43.3) (n = 32) −42.3 (−185.6 101.1) −3.9 (−19.6 11.9) −15.4 (−173.0 142.3)

Data of the cohort overall are presented as n (column %), median (IQR) or mean ± SD by MetS status. Logistic regression and linear regression were used for comparison, adjusted for age, gender and diagnosis. Dichotomization: left/right average MTA scores b 1.5 or at least 1.5, PWML b 7 or at least 7 and DWML b 9 or at least 9. NBV, tau, p-tau and Aβ42 were log transformed for statistical analysis, however the reported adjusted mean differences are based on the raw data. MTA = medial temporal atrophy. PWML = periventricular white matter lesions. DWML = deep white matter lesions. NBV = normalized brain volume. Aβ42 = amyloid beta. a Normal range b 375 pg/ml. b Normal range b 52 pg/ml. c Normal range N 550 pg/ml.

population most studies found an association between MetS and increased progression of cognitive impairment (Review [31]). The latter is in line with the observed association between MetS and elevated dementia risk in the general population. There are several possible explanations for the apparent discrepancies between the findings in memory clinic patients and the findings of previous studies in the general population. A factor that predisposes to dementia at the population level does not necessarily have to be associated with more severe symptoms, faster progression or more brain abnormalities at the time of presentation in a memory clinic setting. Nevertheless, the strength of studies in memory clinic populations is that it can help to identify specific cognitive or clinical phenotypes of presumed etiological factors for cognitive decline and dementia, in a population in which early features of dementia are common. The absence of both an MRI and a CSF biomarker profile in the current study, suggests that once cognitive complaints have commenced MetS is not associated with a specific type of pathology (i.e. neurodegenerative changes, or vascular lesions). Moreover, once cognitive complaints of any severity are reason to attend a memory clinic, patients with MetS don't have a faster progression of their cognitive complaints compared to patients without MetS. The absence of a significant relation might be explained by limited power of the current study. However it is more likely, that it is explained by the timing (i.e. the presence of cognitive complaints) of the current study. Overall, the period where an unfavourable vascular health status contributes to dementia might be early in the disease process. It has been reported that the association between individual vascular risk factors and dementia is age-dependent and some associations (hypertension, hypercholesterolemia, and obesity) are known to invert later in life [3]. In the oldest-old (N 80 years old), MetS is no longer associated with cognitive performance [32,33] suggesting that also the association between MetS and dementia inverts later in life. Importantly, this cannot explain the absence of a clinical profile in the current study, as the average age of the cohort was 66.7 (SD9.7) and the nondemented patients with MetS did perform worse compared to the patients without MetS. The construct of the MetS is related to several underlying metabolic abnormalities, including insulin resistance, inflammation and a prothrombotic state, as well as predisposing lifestyle factors such as sedentary behaviour. These metabolic abnormalities, as well as the factors that predispose to the MetS, as well as the vascular risk factors that

constitute the components of the MetS are all potentially modifiable factors that may provide leads for the prevention of dementia. A review [34] regarding potentially modifiable risk factors showed that a 10% reduction in midlife obesity prevalence could potentially prevent about 67,000 Alzheimer's disease cases worldwide. The Honolulu Asia Aging Study showed that about 27% of the dementia cases could be attributed to untreated hypertension, while it was about −12% for treated hypertension [35,36]. Preliminary results from the FINRISK study indicate that lipid-lowering drugs may have a beneficial effect in dementia prevention in the general population [36]. Furthermore, a systematic review showed that a Mediterranean diet and physical exercise decrease the risk of dementia in the general population [41]. There is also interest in the potential of treatment of the MetS or some of its associated factors, to slow cognitive decline in patients with early dementia or MCI. A multicentre randomized controlled clinical trial showed that there was no difference in decline between elderly patients with mild dementia who received strict cardiovascular treatment versus patients who had received standard care [37]. A recent systematic review showed that physical exercise in older subjects with MCI has some positive effects on cognition, while for older subjects with dementia physical exercise showed no effect on cognition [42]. Furthermore, nutritional strategies (like ω−3 fatty acid or Souvenaid) have not shown to delay the rate of cognitive decline in patients with mild to moderate AD, however positive effects were observed in a very small subgroup with very mild cognitive dysfunction [43]. This together with our findings suggest that, in order to be successful, intervention strategies should be timed early on in the disease process, preferably in a pre-symptomatic stage. A strength of this study is the extensive standardized diagnostic evaluation of the patients. Another strength is the high (92%) response rate of the follow-up. The used NCEP–ATP-III criteria are the most commonly used criteria to define MetS in studies on cognition [6]. A possible limitation is the substitution of waist circumference by BMI. Waist circumference reflects central obesity more accurately and abdominal obesity has been shown to be a better predictor of dementia than general obesity measured by BMI [38]. Unfortunately, data on lifestyle variables, such as physical exercise and diet, that are potentially related to both the MetS and cognition and might affect their interrelation were not collected in a standardized fashion for the present cohort. Although all the applied MRI measures are well-established, the WML scoring is a rather crude method, which could potentially mask small effect sizes.

L.G. Exalto et al. / Journal of the Neurological Sciences 351 (2015) 18–23

Then again, a recent study that used a more sensitive technique to calculate the WML volume, segmentation, reported no association between WML and MetS in a population based cohort [39]. Nonetheless, it should be noted that the 95% confidence intervals of the logistic regression analysis were relatively wide, suggesting a potential lack of power to detect an association between MetS and the dichotomized visual rating scales. Importantly, our negative findings on the association between MetS and either NBV and CSF biomarkers are not likely to be due to limited statistical power. In these linear regression analyses, the point estimates were close to zero and the 95% confidence intervals were relatively narrow. Of note, the evaluation of the contributions of the individual risk factors of the MetS was not evaluated in our study, because the sample size of the current cohort (n = 86) did not provide sufficient statistical power to analyse each of the 5 individual MetS components, while adjusting for the other 4. Long-term maintenance of vascular health may delay or prevent dementia [34]. The present findings demonstrate that MetS as an expression of an unfavourable vascular heath status is associated with worse performance on cognitive tests in patients without dementia, but overall MetS is not associated with a specific clinical profile or faster progression once a patient noticed changes in their cognitive functioning. This is suggestive of a possible role of MetS early on in the disease process leading to cognitive impairment. Further research is needed to unravel the life-time perspective on vascular risk factors for dementia. Conflict of interest statement G. J. Biessels consults for and receives research support from Boehringer Ingelheim and consults for Takeda pharmaceuticals. Dr. Biessels receives no personal compensation for these activities. Dr. Scheltens serves/has served on the advisory boards of: Genentech, Novartis, Pfizer, Roche, Danone, Nutricia, Jansen AI, Baxter and Lundbeck. He has been a speaker at symposia organised by Lundbeck, Lilly, Merz, Pfizer, Jansen AI, Danone, Novartis, Roche and Genentech. He serves on the editorial board of Alzheimer's Research & Therapy and Alzheimer's Disease and Associated Disorders, and is a member of the scientific advisory board of the EU Joint Programming Initiative and the French National Plan Alzheimer. The Alzheimer Centre receives unrestricted funding from various sources through the VUmc Fonds. Dr. Scheltens receives no personal compensation for the activities mentioned above. The other authors have no conflict of interest to declare. Acknowledgments The Alzheimer Centre VUmc is supported by Alzheimer Nederland and Stichting VUmc fonds. The clinical database structure was developed with funding from Stichting Dioraphte. This study was supported by VIDI grant 91711384 from ZonMw, The Netherlands Organisation for Health Research and Development and a High Potential grant of Utrecht University to Geert Jan Biessels. References [1] Panza F, Frisardi V, Seripa D, et al. Metabolic syndrome, mild cognitive impairment, and dementia. Curr Alzheimer Res 2011;8:492–509. [2] Grundy SM, Brewer Jr HB, Cleeman JI, et al. Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Arterioscler Thromb Vasc Biol 2004;24:e13–8. [3] Kloppenborg RP, van den Berg E, Kappelle LJ, et al. Diabetes and other vascular risk factors for dementia: which factor matters most? A systematic review. Eur J Pharmacol 2008;585: 97–108. [4] Viscogliosi G, Andreozzi P, Chiriac IM, et al. Screening cognition in the elderly with metabolic syndrome. Metab Syndr Relat Disord 2012;10(5):358–62. [5] Yaffe K. Metabolic syndrome and cognitive decline. Curr Alzheimer Res 2007;4:123–6. [6] Panza F, Solfrizzi V, Logroscino G, et al. Current epidemiological approaches to the metabolic–cognitive syndrome. J Alzheimers Dis 2012;30 Suppl 2:S31–75.

23

[7] Solfrizzi V, Scafato E, Capurso C, et al. Metabolic syndrome, mild cognitive impairment, and progression to dementia. The Italian Longitudinal Study on Aging. Neurobiol Aging 2011; 32:1932–41. [8] Watts AS, Loskutova N, Burns JM, et al. Metabolic syndrome and cognitive decline in early Alzheimer's disease and healthy older adults. J Alzheimers Dis 2013;35:253–65. [9] McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology 1984;34:939–44. [10] Roman GC, Tatemichi TK, Erkinjuntti T, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology 1993;43: 250–60. [11] Petersen RC, Doody R, Kurz A, et al. Current concepts in mild cognitive impairment. Arch Neurol 2001;58:1985–92. [12] Mulder C, Verwey NA, van der Flier WM, et al. Amyloid-beta(1–42), total tau, and phosphorylated tau as cerebrospinal fluid biomarkers for the diagnosis of Alzheimer disease. Clin Chem 2010;56:248–53. [13] Zhu S, Wang Z, Heshka S, et al. Waist circumference and obesity-associated risk factors among whites in the third National Health and Nutrition Examination Survey: clinical action thresholds. Am J Clin Nutr 2002;76:743–9. [14] Lindeboom J, Schmand B, Tulner L, et al. Visual association test to detect early dementia of the Alzheimer type. J Neurol Neurosurg Psychiatry 2002;73:126–33. [15] Van der Elst W, van Boxtel MP, van Breukelen GJ, et al. Rey's verbal learning test: normative data for 1855 healthy participants aged 24–81 years and the influence of age, sex, education, and mode of presentation. J Int Neuropsychol Soc 2005;11:290–302. [16] Lindeboom J, Matto D. [Digit series and Knox cubes as concentration tests for elderly subjects]. Tijdschr Gerontol Geriatr 1994;25:63–8. [17] Reitan R. Validity of the Trail Making Test as an indicator of organic brain damage. Percept Mot Skills 1958;8:271–6. [18] Stroop JR. Studies of interference in serial verbal reactions. J Exp Psychol 2010;18:643–62. [19] Rey A. L'examen psychologique dans les cas d'encephalopathie traumatique: Les problemes. Arch Psychol 1941:286–340. [20] Yesavage JA. Geriatric Depression Scale. Psychopharmacol Bull 1988;24:709–11. [21] Scheltens P, Leys D, Barkhof F, et al. Atrophy of medial temporal lobes on MRI in “probable” Alzheimer's disease and normal ageing: diagnostic value and neuropsychological correlates. J Neurol Neurosurg Psychiatry 1992;55:967–72. [22] van Straaten EC, Scheltens P, Knol DL, et al. Operational definitions for the NINDS-AIREN criteria for vascular dementia: an interobserver study. Stroke 2003;34:1907–12. [23] Manschot SM, Brands AM, van der Grond J, et al. Brain magnetic resonance imaging correlates of impaired cognition in patients with type 2 diabetes. Diabetes 2006;55:1106–13. [24] Smith SM, Zhang Y, Jenkinson M, et al. Accurate, robust, and automated longitudinal and cross-sectional brain change analysis. Neuroimage 2002;17:479–89. [25] Sluimer JD, Vrenken H, Blankenstein MA, et al. Whole-brain atrophy rate in Alzheimer disease: identifying fast progressors. Neurology 2008;70:1836–41. [26] Kempen GI, Meier AJ, Bouwens SF, et al. The psychometric properties of the Dutch version of the Telephone Interview Cognitive Status (TICS). Tijdschr Gerontol Geriatr 2007;38: 38–45. [27] Isik AT, Cankurtaran M, Bozoglu E, et al. Is there any relation between insulin resistance and cognitive function in the elderly? Int Psychogeriatr 2007;19:745–56. [28] Crichton GE, Elias MF, Buckley J, et al. Metabolic syndrome, cognitive performance, and dementia. J Alzheimers Dis 2011;30 Suppl 2:S77–87. [29] Benedictus MR, Goos JD, Binnewijzend MA, et al. Specific risk factors for microbleeds and white matter hyperintensities in Alzheimer's disease. Neurobiol Aging 2013;34:2488–94. [30] Korf ES, van Straaten EC, de Leeuw FE, et al. Diabetes mellitus, hypertension and medial temporal lobe atrophy: the LADIS study. Diabet Med 2007;24:166–71. [31] Panza F, Frisardi V, Capurso C, et al. Metabolic syndrome and cognitive impairment: current epidemiology and possible underlying mechanisms. J Alzheimers Dis 2010;21: 691–724. [32] Laudisio A, Marzetti E, Pagano F, et al. Association of metabolic syndrome with cognitive function: the role of sex and age. Clin Nutr 2008;27:747–54. [33] van den Berg E, Biessels GJ, de Craen AJ, et al. The metabolic syndrome is associated with decelerated cognitive decline in the oldest old. Neurology 2007;69:979–85. [34] Barnes DE, Yaffe K. The projected effect of risk factor reduction on Alzheimer's disease prevalence. Lancet Neurol 2011;10:819–28. [35] Launer LJ, Hughes T, Yu B, et al. Lowering midlife levels of systolic blood pressure as a public health strategy to reduce late-life dementia: perspective from the Honolulu Heart Program/Honolulu Asia Aging Study. Hypertension 2010;55:1352–9. [36] Solomon A, Sippola R, Soininen H, et al. Lipid-lowering treatment is related to decreased risk of dementia: a population-based study (FINRISK). Neurodegener Dis 2010;7:180–2. [37] Richard E, Kuiper R, Dijkgraaf MG, et al. Vascular care in patients with Alzheimer's disease with cerebrovascular lesions—a randomized clinical trial. J Am Geriatr Soc 2009;57: 797–805. [38] Whitmer RA, Gustafson DR, Barrett-Connor E, et al. Central obesity and increased risk of dementia more than three decades later. Neurology 2008;71:1057–64. [39] Sala M, de Roos A, van den Berg A, et al. Microstructural brain tissue damage in metabolic syndrome. Diabetes Care 2013;37(2):493–500. [40] Verhage F. Intelligentie en Leeftijd: Onderzoek bij Nederlanders van Twaalf tot Zevenenzeventig Jaar (Intelligence and age: study with Dutch people aged 12 to 77). ; 2010. [41] Plassman BL, Williams Jr JW, Burke JR, et al. Systematic review: factors associated with risk for and possible prevention of cognitive decline in later life. Ann Intern Med 2010;153: 182–93. [42] Ohman H, Savikko N, Strandberg TE, et al. Effect of physical exercise on cognitive performance in older adults with mild cognitive impairment or dementia: a systematic review. Dement Geriatr Cogn Disord 2014;38:347–65. [43] Mi W, van WN Cansev M, et al. Nutritional approaches in the risk reduction and management of Alzheimer's disease. Nutrition 2013;29:1080–9.