The effect of amyloid-β pathology on cortical atrophy and
cognition in cognitively normal elderly
A. Lamé
Student number: 11657421
Bachelor Psychobiology, Faculty of Science, University of Amsterdam
Supervisors: dr. A. den Braber, dr. B. Tijms
Alzheimer Center, Department of Neurology, VU University Medical Center Date: 19-06-2020
Abstract
There is a growing population of elderly with Alzheimer’s disease (AD) but there is no sufficient treatment yet. AD related pathology can precede cognitive decline by decades. Therefore, preventative therapies that intervene with the progression of the disease when neurodegeneration is still limited are promising. A probable cause of AD is the development of amyloid-β pathology, followed by cortical atrophy and resulting in deteriorating cognition. To investigate the earliest stage of AD development, this paper takes an exploratory look at the effect of amyloid-β pathology on the cognitive performance in cognitively normal elderly and whether the effect of this pathology on cognition is mediated by its effect on (sub)cortical atrophy. Therefore, 202 cognitively normal monozygotic twins included in the European Medical Information Framework for AD-PreclinAD are investigated on memory performance and cognition, (sub)cortical atrophy and amyloid-β pathology.
It is found that amyloid-β pathology is associated with a poorer overall cognitive performance and visuospatial memory, a thinner lateral orbitofrontal cortex (LOFC), precuneus, supramarginal gyrus and a lower putamen volume. Significant associations are also seen between a thinner LOFC, precuneus and supramarginal gyrus and poorer visuospatial memory. No mediation effect of (sub)cortical atrophy is found on the effect of amyloid-β pathology on visuospatial memory. However, a suggestive mediation effect of the supramarginal gyrus on visuospatial memory is found. In short, the cognitively normal participants with AD related pathology show poorer cognitive functioning than participants with no pathology.
Keywords: Amyloid-β pathology, Cortical atrophy, Cognitively normal, Memory performance, Monozygotic Twins.
1. Introduction
Alzheimer’s disease (AD) is a degenerative brain disease that is associated with the deterioration of cognitive functioning as the disease progresses (Alzheimer’s Association, 2019). Gradually, symptoms as memory loss, disorientation, impaired communication, loss of motor functions and personality changes arise and get worse over time, which leads to death eventually. Although the amount of deaths due to AD is increasing worldwide every year, there is yet no sufficient treatment (Alzheimer’s Association, 2019). This growing population of elderly with AD is an increasing social and economic burden on society. In order to be able to stop this growth of people with AD, the development of preventative strategies is required. Early interventions may reduce the progression of the disease when neurodegeneration is still limited. Therefore, precise and early diagnosis of elderly with AD is necessary.
AD is believed to start developing more than two decades before the first cognitive symptoms become apparent (Villemagne et al., 2013). For example, cellular pathologies related with sporadic AD can be found in the brain starting from 40 years and older (Braak, Thal, Ghebremedhin, & Tredici, 2011), while symptoms may manifest much later. Patients in preclinical and prodromal phases of AD are possibly most susceptible for interventions, as the pathology is developing but cognitive symptoms are not visible yet (Insel, Hansson, Mackin, Weiner, & Mattsson, 2018). There are many factors that are associated with the progression of AD. One of the probable causes of the neurodegeneration in AD is the aggregation of the misfolded amyloid-β (aβ) protein, resulting in the formation of amyloid-β plaques (Kumar, Singh, & Ekavali, 2015), which is toxic for the surrounding neuronal cells (Serrano-Pozo, Betensky, Frosch, & Hyman, 2016). The main component of amyloid-β plaques is the 42 amino acid form of amyloid-amyloid-β (aamyloid-β42) (Blennow, Zetterberg, & Fagan, 2012). This process of amyloid-β deposition leading to neuronal cell death as an explanation of the development of AD is called ‘the amyloid cascade hypothesis’ (Karran, Mercken, & Strooper, 2011).
According to this amyloid cascade hypothesis, a volumetric loss of certain brain areas arises from the neurodegeneration due to amyloid-β pathology (Becker et al., 2011). This cortical and subcortical atrophy is found in the brains of AD patients (Dickerson et al., 2009). There are certain regions of the brain that are more vulnerable for neurodegeneration and these areas show atrophy in earlier stages of the disease. Cortical areas that show thinning relatively early are the anterior and posterior cingulate cortex (ACC and PCC), the entorhinal cortex, the lateral and medial orbitofrontal cortex (LOFC and MOFC), precuneus, parahippocampal gyrus and supramarginal gyrus (Gispert et al., 2016; Krumm et al., 2016; Thomas, Sheelakumari, Kannath, Sarma, & Menon, 2019; Traschütz et al., 2020). The classical subcortical areas that show volumetric loss in AD are the hippocampus (Tentolouris-Piperas, Ryan, Thomas, & Kinnunen, 2017) and nearby amygdala, which are regions that are also found to be vulnerable for atrophy in cognitively normal individuals (Bernard et al., 2014). Additionally, some studies show that the caudate, thalamus and putamen are susceptible for amyloid deposition and shrinkage (Chételat et al., 2005; Cho et al., 2014; Madsen et al., 2010).
Furthermore, the amyloid cascade hypothesis also states that as a result of the amyloid-β pathology caused neurodegeneration, patients with AD related pathology will show declining cognition. Cognition can be assessed with neuropsychological testing. For several of these tests covering memory and cognitive functioning, there is a strong association between the existence of amyloid-β pathology and lower performance in cognitively normal elderly. First of all, a lower score on the Rey Complex Figure Recall (RCFR) task for visuospatial memory is associated with amyloid-β pathology in cognitively normal elderly, especially when looking at the 3 minutes delayed recall score (Konijnenberg et al., 2019; Snitz et al., 2013). Further, a poorer performance on the Face-Name associative memory examination (FNAME) shows a correlation with amyloid-β pathology as well (Papp et al., 2014). Also, the Cambridge Neuropsychological Test Automated Battery Paired Associate Learning task with total errors adjusted (CANTAB PAL-ER) for visual associative memory is correlated with amyloid-β pathology in cognitively normal individuals (Reijs et al., 2017). A test for cognitive functioning rather than memory is the Mini-Mental State Examination (MMSE), which is often used in the diagnosis of AD patients and consists of questions that are designed to test a range of daily mental skills (Pinto et al., 2019). A
lower MMSE total score indicates dementia. However, it is found that amyloid-β pathology is associated with a lower MMSE score in individuals with mild cognitive impairment (MCI), but not in cognitively normal elderly (Jansen et al., 2018).
Although the amyloid cascade hypothesis can explain the pathway of the development in the disease roughly, there are some apparent inconsistencies. First of all, the existence of the amyloid-β pathology in the brain is not necessarily directly related to a deterioration in cognitive performance. This is supported by the fact that plaques are also found postmortem in the brains of cognitively normal functioning elderly (Jansen et al., 2015). Additionally, the reduction of amyloid burden in patients with mild AD in a clinical trial shows no cognitive benefits (Karran et al., 2011). Also, there is a large variation in the degree of cognitive decline for patients with comparable pathology (Vos et al., 2013). This implies that these pathologies and declining cognition might be separate processes rather than that they are following consequently. Furthermore, this variation in the progression of the disease can be explained by factors like age, gender and educational level (Xu et al., 2015). It is known that higher age, the female gender and a lower educational level increase the risk at a poorer cognitive outcome in AD (Daviglus et al., 2010; Laws, Irvine, & Gale, 2018). On top of that, cortical atrophy, which is a great predictor of cognitive decline, does not only appear in patients with amyloid-β pathology in AD. Cognitively normal functioning elderly show increasing atrophy as they age (Mattsson, Insel, Nosheny, et al., 2014). Thus, other factors may play a more important role in the development of the disease than that is taken into account with the amyloid cascade hypothesis.
In order to get better insight into the earliest stages of disease development, it is needed to take an exploratory look at these other factors that play an important role in the development of the disease, such as the cortical atrophy in certain areas. Hence, possibilities for interference with the progression of the disease can be sought and might ultimately lead to intervention strategies in patients with a risk of developing AD while neurodegeneration is still limited. Therefore, the effect of amyloid-β pathology on the cognitive performance in cognitively normal elderly and whether the effect of this pathology on cognition is mediated by its effect on (sub)cortical atrophy is investigated in this paper. Based on the amyloid cascade hypothesis, it is expected that amyloid-β pathology and cortical atrophy are associated with a poorer cognitive performance. But this deteriorating effect of amyloid-β pathology on cognition is expected to be greater when there is more (sub)cortical atrophy present. This way, this paper can contribute to a better understanding of the early stages of Alzheimer’s disease.
2. Methods
2.1. Participants
In this research, data from 202 cognitively normal monozygotic twins included in the European Medical Information Framework for AD-PreclinAD study was used (Konijnenberg et al., 2018). Inclusion criteria for this study were age of 60 years and older, normal cognitive functioning, with no signs of depression, dementia or other disorders that could affect cognitive functioning. All physical and neuropsychological examinations were performed at the VU University Medical Center in Amsterdam. Not all participants were able to run through all measurements, due to measuring failures, time restraints, the participant’s refusal or contra-indications. This resulted in different numbers of participants for each test.
2.2. Amyloid-β measurements
Amyloid-β pathology is detectable in cerebrospinal fluid (CSF) (Mattsson, Insel, Landau, et al., 2014). When pathology occurs, a reduction in the ratio of aβ1-42/aβ1-40 in CSF will be found (Toombs et al., 2018). Amyloid-β pathology is also perceptible with positron emission tomography (PET) (Mattsson, Insel, Landau, et al., 2014). Flutemetamol labeled with radioactive fluorine 18 ([18F]flutemetamol) is a derivative of the 11 C-Pittsburgh compound B (PiB) and highly sensitive and specific in the in-vivo detection of amyloid-β pathology (Curtis et al., 2015). These radiotracers are ligands that bind with aβ42 and can be measured with an amyloid
PET scanner. Both CSF and PET measurements of amyloid-β plaques are correlated with clinical AD signs. Individuals are categorized as amyloid ‘positive’ when pathology is present either visible in the CSF or with PET.
2.2.1. Cerebrospinal fluid collection
A lumbar puncture for CSF collection was executed in 126 of the participants. About 20 mL of CSF was collected between 10 AM and 2 PM, after at least 2 hours of fasting. A 25-gauge needle was placed in one of the intervertebral spaces between L3 and S1. The CSF samples were mixed and centrifuged at 1300 – 2000 g for 10 minutes at 4 C. Within 2 hours of the lumbar puncture, supernatants of the samples were frozen until further analysis. In these samples, the levels of different amyloid protein were analyzed with kits from ADx Neurosciences/Euroimmun according to manufacturer’s protocol (Konijnenberg et al., 2018).
2.2.2. Positron emission tomography scanning
The dynamic amyloid-PET scans were executed on a Philips Ingenuity Time-of-Flight PET-MRI scanner. 191 participants were scanned while resting with a fixed head to reduce movements for 30 minutes, immediately after an injection of 185 MBq [18F]flutemetamol. Hereafter, the participants rested for 60 minutes in a nearby room. Then the participants were scanned for an additional 20 minutes, which is the recommended time window for the assessment of amyloid-β pathology (Curtis et al., 2015). The dynamic scans were reconstructed into 18 frames with increasing lengths and into 4 frames of 300 seconds each. These frames were then summed to get a static image and parametric non-displaceable binding potential (BPND) images were made to visualize amyloid-β pathology. These images were visually categorized as normal or abnormal by the consensus of 3 experienced readers, who were blinded tot the clinical and demographic data.
2.3. Assessment of (sub)cortical structures
195 participants were scanned in a 3 T Philips Achieva MRI scanner equipped with an eight-channel head coil. The MRI protocol included structural 3D-T1, 3D fluid-attenuated inversion recovery (FLAIR), pseudocontinuous arterial spin labeling (ASL), susceptibility weighted imaging (SWI), diffusion tensor imaging (DTI) and 6 minutes of resting state functional MRI (fs-fMRI). MRI settings are presented in the paper of Konijnenberg et al. (2018). Cortical thickness and subcortical volumes of brain structures were obtained from 3D-T1 images using the Freesurfer image analysis suite, which is freely available (http://surfer.nmr.mgh.harvard.edu/).
2.4. Measuring cognitive performance
All participants were tested on numerous neuropsychological tests and answered several questionnaires about their life and activities in about 4 hours. Different domains of cognitive functioning and overall cognitive functioning were tested. Scores on four cognitive tests that are associated with amyloid-β pathology were selected for further analysis in this paper. These tests evaluate visuospatial (RCFR task) and visual associative memory (FNAME, CANTAB PAL-ER task) and overall cognitive functioning (MMSE).
2.5. Statistical analysis
Statistical analyses were performed in RStudio version 3.6.1. The amyloid-β pathology was described as a dichotomous and continuous variable, where an amyloid-β 1-42/1-40 ratio ≤ 0.066 and/or positive visual read of the amyloid-PET BPND indicated amyloid-β pathology (Konijnenberg et al., 2019). Cortical thickness and volumetric measures from left and right hemispheres were averaged for each structure. Across-participant associations were evaluated using Generalized Estimating Equations with the amyloid-β pathology assessment (dichotomous values, positive indicates amyloid-β pathology) or the amyloid-β 1-42/1-40 ratio (continuous values) as predictors and cognitive performance, cortical thickness or subcortical volumes as outcome variable. All equations were corrected for clustering of twins within pairs as exchangeable. When significant associations were found between the amyloid-β pathology and outcome variables of cognition and (sub)cortical atrophy,
these variables were used for further testing. The (sub)cortical areas of interest were then used as a predictor of the cognitive test outcomes.
Lastly, to test the hypothesis that a lower amyloid-β 1-42/1-40 ratio is associated with a poorer cognitive performance and that this effect can be explained by the effect of the ratio on (sub)cortical atrophy, a mediation analysis was executed. The continuous values of the amyloid-β 1-42/1-40 ratio were used as a predictor for the cognitive test outcome of interest and with different (sub)cortical areas of interest, while correcting for clustering of twins within pairs. All described analyses were run without (Model 1) and with (Model 2) the covariates age and gender. Educational Verhage years were also added in the equations when regarding cognition, and intracranial volume (ICV) was added when regarding subcortical volumes.
3. Results
3.1. Demographics
202 participants were included in the study, of which were 99 complete twin pairs and 4 participants without their twin available. 126 participants had CSF measurements available, 191 had PET BPND data, 195 participants had (sub)cortical assessments and the amount of participants with cognitive outcomes differed for each cognitive test. The participants are on average 70.5 (SD = 7.7) years old and 117 (58%) are female. Demographics are shown in Table 1.
Table 1: Demographics of the participants
Number of subjects (%) Mean (± SD)
Gender: female 117 (57.92)
-Age (years) 202 (100.00) 70.50 (7.68)
Educational Verhage Years
202 (100.00) 11.41 (2.61)
MMSE total score 202 (100.00) 28.93 (1.19)
RCF delayed recall score (3 mins)
202 (100.00) 18.30 (5.55)
F-NAME total score 182 (90.10) 52.40 (17.58)
PAL score total errors adjusted
201 (99.50) 28.92 (16.27)
Amyloid-β 1-42/1-40 ratio
126 (62.38) 0.096 (0.03)
PET BPND read: positive 25 (12.38)
-Amyloid-β pathology: positive 31 (15.35) -Average cortical thickness (mm) 195 (96.53) 2.34 (0.10) ICV (mm3) 196 (97.03) 1.33
∙
1011 (4.13∙
1010) 3.2. Across-participant associations3.2.1. Relation between amyloid-β pathology and cognitive performance
A lower amyloid-β 1-42/1-40 ratio was associated with lower MMSE total scores (b = 10.21, p = 0.0199), which evaluates overall cognitive functioning. Also, a lower ratio was associated with lower RCF organization scores (b = 11.73, p = 0.0420), RCF copy scores (b = 23.60, p = 0.0480) and lower RCF delayed recall scores of 3 minutes (b = 54.50, p = 0.0007) and 20 minutes (b = 45.10, p = 0.0014), which evaluates visuospatial memory. Amyloid positivity was only associated with lower RCF delayed recall scores of 3 minutes (b = -3.20, p = 0.0001). Results are shown in Table 2A and 2B (Model 1) and Figure 1.
Table 2A: Predictor: Amyloid-β 1-42/1-40 ratio
Table 2B: Predictor: Amyloid positivity
Outcome variables: Model 1Estimate p-value Model 2Estimate p-value Model 1Estimate p-value Model 2Estimate p-value MMSE total 10.21 0.0199 * 7.01 0.0820 -0.26 0.1900 -0.16 0.4780 RCF organization score 11.73 0.0420 * 10.42 0.1200 -0.28 0.4200 -0.13 0.7190 RCF copy score 23.60 0.0480 * 16.19 0.1310 -0.48 0.3900 -0.03 0.9600 RCF delayed recall (3 mins) 54.50 0.0007 *** 31.66 0.0364 * -3.20 0.0001 *** -2.25 0.0053 ** RCF delayed recall (20 mins) 45.10 0.0014 ** 26.21 0.0590 -1.60 0.1000 -1.00 0.2800
F-NAME score names 42.52 0.1500 25.40 0.4221 -0.33 0.7600 -0.59 0.6494
F-NAME score occupation
51.47 0.0960 18.61 0.5381 -1.62 0.2900 -1.41 0.3390
F-NAME total score 80.09 0.1500 27.32 0.6411 -1.32 0.5600 -1.59 0.4944
PAL score total errors adjusted
-64.46 0.2300 -24.75 0.6410 3.13 0.3600 1.13 0.7305
Figure 1. Scatterplot of RCF delayed recall of 3 minutes score related to the amyloid-β 1-42/1-40 ratio, stratified by amyloid-βpathology with amyloid positivity indicated in red.
When the covariates age, gender and education were added (Table 2A and 2B, Model 2), only significant relations were found between lower RCF delayed recall scores of 3 minutes and a lower amyloid-β 1-42/1-40 ratio (b = 31.66, p = 0.0364) and amyloid positivity (b = -2.25, p = 0.0053).
3.2.2. Relation between amyloid-β pathology and (sub)cortical structures
No associations were found between a lower amyloid-β 1-42/1-40 ratio and cortical thickness or subcortical volumes. However, there were significant relations between amyloid positivity and a thinner lateral orbitofrontal cortex (b = -0.060, p = 0.0004), precuneus (b = -0.046, p = 0.0011), supramarginal gyrus (b = -0.033, p = 0.0240) and a lower putamen volume (L/R, b = 201, p = 0.041). Results are shown in Table 3A and 3B (Model 1) and Figure 2 and 3.
Table 3A: Predictor: Amyloid-β 1-42/1-40 ratio
Table 3B: Predictor: Amyloid positivity
Outcome variables Model 1Estimate p-value Model 2Estimate p-value Model 1Estimate p-value Model 2Estimate p-value
ACC 0.08 0.9100 0.30 0.6900 0.02 0.7100 0.02 0.6190 PCC 0.41 0.2800 0.08 0.8437 -0.02 0.4600 -4.54
∙
10-4 0.9846 Entorhinal cortex 0.31 0.7600 -0.57 0.5472 -0.05 0.3600 -0.02 0.6580 LOFC 0.56 0.1400 0.18 0.6000 -0.06 0.0004 *** -0.04 0.0062 ** MOFC -0.14 0.7300 -0.51 0.1400 -0.04 0.0680 -0.02 0.3200 Precuneus 0.38 0.1700 0.10 0.7056 -0.05 0.0011 ** -0.03 0.0230 * Parahippocampal gyrus -0.18 0.8000 -1.15 0.0810 0.01 0.8600 0.05 0.1800 Supramarginal gyrus 0.63 0.1300 0.40 0.2550 -0.03 0.0240 * -0.02 0.1800 Thalamus (L/R) -345 0.8900 -1410 0.5600 -30.6 0.8200 11.8 0.9280 Caudatus (L/R) 616 0.5400 557 0.5582 -14.7 0.7800 -29.5 0.5870 Hippocampus (L/R) 1476 0.2500 -742 0.5321 -20.0 0.7000 49.7 0.3090 Amygdala (L/R) -421 0.4900 -971 0.0930 -57.1 0.0960 -36.1 0.3090 Putamen (L/R) 2258 0.2100 1370 0.4210 -201 0.0410 * -190 0.0563Figure 2. Scatterplot of the LOFC thickness related to the amyloid-β 1-42/1-40 ratio, stratified by amyloid-β pathology with amyloid positivity indicated in red.
Figure 3. Scatterplot of the putamen volume related to the amyloid-β 1-42/1-40 ratio, stratified by amyloid-β pathology with amyloid positivity indicated in red.
When the covariates age, gender and intracranial volume were added (Table 3A and 3B, Model 2), only significant relations were found between amyloid positivity and a thinner LOFC (b = -0.04, p = 0.0062) and precuneus (b = -0.03, p = 0.0230).
3.2.3. Relation between (sub)cortical structures and cognitive performance
There were significant relations between a lower RCF delayed recall 3 minutes score and a thinner LOFC (b = 5.58, p = 0.0410), precuneus (b = 8.34, p = 0.0250) and the supramarginal gyrus (b = 11.70, p = 0.0012). Results are shown in Table 3 (Model 1) and Figure 4. However, none of the observed associations remained significant after adjusting for age and gender (and ICV for the putamen) (Model 2).
Table 3: Outcome variable: RCF (3 mins)
Predictors Model 1Estimate p-value Model 2Estimate p-value
Precuneus 8.34 0.0250 * 3.29 0.3950 Supramarginal gyrus 11.70 0.0012 ** 6.47 0.0790 Putamen (L/R) 4.19
∙
10 -4 0.5900 3.59∙
10 -5 0.9600Figure 4. Scatterplot of RCF delayed recall of 3 minutes score related to the supramarginal gyral thickness, stratified by amyloid-β pathology with amyloid positivity indicated in red.
3.3. Mediation analysis
The mediation analysis of the effect of the amyloid-β 1-42/1-40 ratio on the RCF delayed recall (3 minutes) score was done with three possible mediating cortical structures: LOFC, precuneus and supramarginal gyrus. Mainly, a direct effect (ADE) of the ratio on the cognitive outcome was found, shown in Table 4 (Model 1). However, no significant effects of the mediator (ACME) were observed, but the estimate and p-value approaching significance for the supramarginal gyrus (9.223, p = 0.066) might suggest a trend of a possible mediation of the supramarginal gyrus on the effect of the amyloid-β 1-42/1-40 ratio on RCF delayed recall (3 minutes) score.
Table 4: Mediation analysis with different (sub)cortical structures as mediators. ACME, Average Causal Mediation effect; ADE, Average Direct Effect.
Mediators
Model 1
Estimate ACME (p-value)
Estimate ADE (p-value) Estimate Total Effect (p-value)
Estimate Prop. Med. (p-value) LOFC 2.935 (0.336) 51.158 (0.006 **) 54.093 (0.004 **) 0.042 (0.340) Precuneus 2.390 (0.422) 51.360 (0.010 *) 53.750 (0.008 **) 0.032 (0.426) Supramarginal gyrus 9.223 (0.066) 45.201 (0.018 *) 54.428 (0.002 **) 0.164 (0.068) Model 2 LOFC 0.228 (0.893) 32.930 (0.078) 33.158 (0.076) 0.002 (0.893) Precuneus 0.105 (0.970) 32.222 (0.080) 32.327 (0.080) 0.001 (0.950) Supramarginal gyrus 3.890 (0.362) 29.273 (0.110) 33.163 (0.072) 0.0933 (0.394)
When covariates age, gender and education were added, no direct effects were found of the amyloid-β 1-42/1-40 ratio on RCF delayed recall (3 minutes) score (Table 4, Model 2).
Looking closer into the mediation analysis, the relation between the amyloid-β 1-42/1-40 ratio, RCF delayed recall (3 minutes) score and supramarginal gyral thickness is visualized in Figure 5. The supramarginal gyral thickness is divided in three groups. The mediation effect of the supramarginal gyrus is mainly visible in
the group with the thickest cortex, as they seem to perform as low as the group with the thinnest cortex when amyloid-β pathology is present. However, they tend to perform better than the group with the thinnest cortex when there is no amyloid-β pathology present.
Figure 5. Scatterplot of RCF delayed recall of 3 minutes score related to the amyloid-β 1-42/1-40 ratio, stratified by supramarginal gyral thickness divided in three groups of thickness (small, medium and large).
4. Discussion
This study showed that, in cognitively normal elderly, a lower amyloid-β 1-42/1-40 ratio was associated with a poorer overall cognitive performance (MMSE) and a poorer visuospatial memory (RCFR task). Participants that were categorized as amyloid positive, based on PET and CSF results, showed a lower score on the visuospatial memory task, but only for the delayed recall of 3 minutes. This effect of amyloid-β pathology on the delayed recall score (3 mins) remained when the covariates age and gender were added. Furthermore, amyloid positivity was associated with a thinner LOFC, precuneus, supramarginal gyrus and a lower putamen volume. There were significant associations between a thinner LOFC, precuneus and supramarginal gyrus and a poorer visuospatial memory score (RCF delayed recall of 3 minutes), but these effects disappeared when adding covariates. No mediation effect of (sub)cortical atrophy was found on the effect of amyloid-β 1-42/1-40 ratio on visuospatial memory. However, the direct effect of amyloid-β pathology on visuospatial memory was evident. Thus, the effect of (sub)cortical atrophy on cognition was not that strong that it mediated the effect of the amyloid-β pathology on cognition. This can be explained by the fact that there were no relations between the amyloid-β 1-42/1-40 ratio and any (sub)cortical atrophy areas. There were only between significant relations found between amyloid p and some areas, but the amyloid-β 1-42/1-40 ratio was used in the mediation analysis. In short, the participants with AD related pathology showed poorer cognitive performance than the participants with no pathology, which is in accordance with the amyloid cascade hypothesis and previous research.
When looking more closely into the mediation analysis, there was no mediation effect of the supramarginal gyral thickness on the effect of the amyloid-β 1-42/1-40 ratio on cognition. However, the p-values of the ACME and estimate proportion mediated were low (p > 0.07) for the supramarginal gyrus in comparison with the other brain areas (Table 4, Model 1). This suggestive mediation effect of the supramarginal gyral thickness is shown in Figure 5. As expected, participants with the thinnest cortex have the lowest cognitive score, independent from the presence of amyloid-β pathology. But participants with the thickest supramarginal gyrus appear to be more influenced by the existence of amyloid-β pathology, as they perform lower when amyloid-β pathology is present. This might be explained by the thickness of other brain areas or effects of covariates on the mediation, because the suggestive mediation effect of the supramarginal gyrus disappears
when these were added (Table 4, Model 2). The supramarginal gyrus is previously related to a decrease in grey matter volume in early stages of AD (Gispert et al., 2016). It is also known that the supramarginal gyrus is a key player in the preservation of serial order in short-term memory (Guidali, Pisoni, Bolognini, & Papagno, 2019). A reduction in thickness of the supramarginal gyrus might therefore explain a lower score on a short-term memory task, like the RCFR task measuring visuospatial memory. But the executed mediation analysis showed that the amyloid-βpathology probably affects visuospatial memory more than that this process is mediated through the thickness of the supramarginal gyrus at this early stage of pathology, which is in accordance with the amyloid cascade hypothesis. It would be interesting to research whether the supramarginal gyrus mediates the effect of amyloid-β pathology on cognition at a later stage of the disease, deviating from the hypothesis. This could be achieved by monitoring cognitively normal elderly with supramarginal gyral atrophy and amyloid-β pathology in a longitudinal study.
There is a large variation in the degree of cognitive decline for patients with comparable pathology (Vos et al., 2013) which can not be explained by the amyloid cascade hypothesis only. Previous research also suggests that these pathologies and declining cognition might be separate processes rather than following consequently (Jansen et al., 2015; Mattsson, Insel, Nosheny, et al., 2014). As seen in this study, some of the found associations between amyloid-β pathology and cognition or cortical atrophy disappeared when covariates were added. Regarding the covariate age, the group of participants with amyloid-β pathology is on average 74.3 (SD = 6.8) years old, which is almost 4 years older than the mean. Given that higher age is a great predictor of a lower cognitive performance and cortical atrophy (Clark et al., 2018; Fjell, McEvoy, Holland, Dale, & Walhovd, 2014), the associations that were found might have been mostly influenced by age rather than only the amyloid-β pathology.
In this study, no corrections for multiple testing were executed as most effects would then disappear, which can be rectified by the fact that the aim of this paper was to explore the influence of other factors in early stages of the development of AD. It also must be taken into account that the sample size of the participants with amyloid-β pathology is small (n = 31). Because the research was conducted with monozygotic twins, all analysis had to be corrected for twin dependency which might have reduced statistical power. Furthermore, all participants were cognitively normal, so differences amongst participants in cognitive performance and AD related pathology in the brain can be subtle. Only including cognitively normal individuals might have therefore limited the ability to detect associations. However, this research showed that individuals with AD related pathology already show cognitive deviations in the earliest stages of the disease. Also, a large sample of cognitively normal elderly with amyloid-β biomarker data available with PET was used and relatively large part of the participants had CSF data available.
Finally, this study helped understanding the early pathophysiology of AD. However, more longitudinal research is needed to see if people at risk due to amyloid-β pathology and/or cortical atrophy develop AD and how these pathologies relate to or mediate each other at later stages of the disease. This way, elderly individuals at risk for AD related pathology and cognitive decline can be identified and the progression of the disease can possibly be delayed when neurodegeneration is still limited.
5. References
2019 Alzheimer’s disease facts and figures. (2019). Alzheimer’s & Dementia, 15(3), 321–387. https://doi.org/10.1016/j.jalz.2019.01.010 Becker, J. A., Hedden, T., Carmasin, J., Maye, J., Rentz, D. M., Putcha, D., … Johnson, K. A. (2011). Amyloid-β associated cortical thinning in
clinically normal elderly. Annals of Neurology, 69(6), 1032–1042. https://doi.org/10.1002/ana.22333
Bernard, C., Helmer, C., Dilharreguy, B., Amieva, H., Auriacombe, S., Dartigues, J. F., … Catheline, G. (2014). Time course of brain volume changes in the preclinical phase of Alzheimer’s disease. Alzheimer’s and Dementia, 10(2), 143-151.e1.
https://doi.org/10.1016/j.jalz.2013.08.279
Blennow, K., Zetterberg, H., & Fagan, A. M. (2012). Fluid biomarkers in Alzheimer disease. Cold Spring Harbor Perspectives in Medicine, 2(9), a006221. https://doi.org/10.1101/cshperspect.a006221
From 1 to 100 Years. Retrieved from https://academic.oup.com/jnen/article-abstract/70/11/960/2917408
Chételat, G., Landeau, B., Eustache, F., Mézenge, F., Viader, F., De La Sayette, V., … Baron, J. C. (2005). Using voxel-based morphometry to map the structural changes associated with rapid conversion in MCI: A longitudinal MRI study. NeuroImage, 27(4), 934–946. https://doi.org/10.1016/j.neuroimage.2005.05.015
Cho, H., Kim, J. H., Kim, C., Ye, B. S., Kim, H. J., Yoon, C. W., … Seo, S. W. (2014). Shape changes of the basal ganglia and thalamus in Alzheimer’s disease: A three-year longitudinal study. Journal of Alzheimer’s Disease, 40(2), 285–295. https://doi.org/10.3233/JAD-132072
Clark, L. R., Berman, S. E., Norton, D., Koscik, R. L., Jonaitis, E., Blennow, K., … Carlsson, C. M. (2018). Age-accelerated cognitive decline in asymptomatic adults with CSF β-amyloid. https://doi.org/10.1212/WNL.0000000000005291
Curtis, C., Gamez, J. E., Singh, U., Sadowsky, C. H., Villena, T., Sabbagh, M. N., … Salloway, S. (2015). Phase 3 trial of flutemetamol labeled with radioactive fluorine 18 imaging and neuritic plaque density. JAMA Neurology, 72(3), 287–294.
https://doi.org/10.1001/jamaneurol.2014.4144
Daviglus, M. L., Bell, C. C., Berrettini, W., Bowen, P. E., Connolly, E. S., Cox, N. J., … Trevisan, M. (2010). National Institutes of Health state-of-the-science conference statement: Preventing Alzheimer disease and cognitive decline. In Annals of Internal Medicine (Vol. 153, pp. 176–181). American College of Physicians. https://doi.org/10.7326/0003-4819-153-3-201008030-00260
Dickerson, B. C., Bakkour, A., Salat, D. H., Feczko, E., Pacheco, J., Greve, D. N., … Buckner, R. L. (2009). The Cortical Signature of Alzheimer’s Disease: Regionally Specific Cortical Thinning Relates to Symptom Severity in Very Mild to Mild AD Dementia and is Detectable in Asymptomatic Amyloid-Positive Individuals. Cerebral Cortex March, 19, 497–510. https://doi.org/10.1093/cercor/bhn113
Fjell, A. M., McEvoy, L., Holland, D., Dale, A. M., & Walhovd, K. B. (2014, June 1). What is normal in normal aging? Effects of aging, amyloid and Alzheimer’s disease on the cerebral cortex and the hippocampus. Progress in Neurobiology. Elsevier Ltd. https://doi.org/10.1016/j.pneurobio.2014.02.004
Gispert, J. D., Monté, G. C., Falcon, C., Tucholka, A., Rojas, S., Sánchez-Valle, R., … Molinuevo, J. L. (2016). CSF YKL-40 and pTau181 are related to different cerebral morphometric patterns in early AD. Neurobiology of Aging, 38, 47–55.
https://doi.org/10.1016/j.neurobiolaging.2015.10.022
Guidali, G., Pisoni, A., Bolognini, N., & Papagno, C. (2019). Keeping order in the brain: The supramarginal gyrus and serial order in short-term memory. Cortex, 119, 89–99. https://doi.org/10.1016/j.cortex.2019.04.009
Insel, P. S., Hansson, O., Mackin, R. S., Weiner, M., & Mattsson, N. (2018). Amyloid pathology in the progression to mild cognitive impairment. Neurobiology of Aging, 64, 76–84. https://doi.org/10.1016/j.neurobiolaging.2017.12.018
Jansen, W. J., Ossenkoppele, R., Knol, D. L., Tijms, B. M., Scheltens, P., Verhey, F. R. J., … Zetterberg, H. (2015). Prevalence of cerebral amyloid pathology in persons without dementia: A meta-analysis. JAMA - Journal of the American Medical Association, 313(19), 1924–1938. https://doi.org/10.1001/jama.2015.4668
Jansen, W. J., Ossenkoppele, R., Tijms, B. M., Fagan, A. M., Hansson, O., Klunk, W. E., … Zetterberg, H. (2018). Association of cerebral amyloid-β Aggregation with cognitive functioning in persons without dementia. JAMA Psychiatry, 75(1), 84–95.
https://doi.org/10.1001/jamapsychiatry.2017.3391
Karran, E., Mercken, M., & Strooper, B. De. (2011, September 19). The amyloid cascade hypothesis for Alzheimer’s disease: An appraisal for the development of therapeutics. Nature Reviews Drug Discovery. Nature Publishing Group. https://doi.org/10.1038/nrd3505 Konijnenberg, E., Carter, S. F., Ten Kate, M., Den Braber, A., Tomassen, J., Amadi, C., … Visser, P. J. (2018, August 4). The EMIF-AD PreclinAD
study: Study design and baseline cohort overview. Alzheimer’s Research and Therapy. BioMed Central Ltd. https://doi.org/10.1186/s13195-018-0406-7
Konijnenberg, E., den Braber, A., ten Kate, M., Tomassen, J., Mulder, S. D., Yaqub, M., … Visser, P. J. (2019). Association of amyloid pathology with memory performance and cognitive complaints in cognitively normal older adults: a monozygotic twin study. Neurobiology of
Aging, 77, 58–65. https://doi.org/10.1016/j.neurobiolaging.2019.01.006
Krumm, S., Kivisaari, S. L., Probst, A., Monsch, A. U., Reinhardt, J., Ulmer, S., … Taylor, K. I. (2016). Cortical thinning of parahippocampal subregions in very early Alzheimer’s disease. Neurobiology of Aging, 38, 188–196.
https://doi.org/10.1016/j.neurobiolaging.2015.11.001
Kumar, A., Singh, A., & Ekavali. (2015, April 1). A review on Alzheimer’s disease pathophysiology and its management: An update.
Pharmacological Reports. Elsevier. https://doi.org/10.1016/j.pharep.2014.09.004
Laws, K. R., Irvine, K., & Gale, T. M. (2018, March 1). Sex differences in Alzheimer’s disease. Current Opinion in Psychiatry. Lippincott Williams and Wilkins. https://doi.org/10.1097/YCO.0000000000000401
Madsen, S. K., Ho, A. J., Hua, X., Saharan, P. S., Toga, A. W., Jack, C. R., … Thompson, P. M. (2010). 3D maps localize caudate nucleus atrophy in 400 Alzheimer’s disease, mild cognitive impairment, and healthy elderly subjects. Neurobiology of Aging, 31(8), 1312–1325. https://doi.org/10.1016/j.neurobiolaging.2010.05.002
Mattsson, N., Insel, P. S., Landau, S., Jagust, W., Donohue, M., Shaw, L. M., … Weiner, M. (2014). Diagnostic accuracy of CSF Ab42 and florbetapir PET for Alzheimer’s disease. Annals of Clinical and Translational Neurology, 1(8), 534–543.
Mattsson, N., Insel, P. S., Nosheny, R., Tosun, D., Trojanowski, J. Q., Shaw, L. M., … Weiner, M. W. (2014). Emerging β-amyloid pathology and accelerated cortical atrophy. JAMA Neurology, 71(6), 725–734. https://doi.org/10.1001/jamaneurol.2014.446
Papp, K. V., Amariglio, R. E., Dekhtyar, M., Roy, K., Wigman, S., Bamfo, R., … Rentz, D. M. (2014). Development of a Psychometrically Equivalent Short Form of the Face–Name Associative Memory Exam for use Along the Early Alzheimer’s Disease Trajectory. The
Clinical Neuropsychologist, 28(5), 771–785. https://doi.org/10.1080/13854046.2014.911351
Pinto, T. C. C., Machado, L., Bulgacov, T. M., Rodrigues-Júnior, A. L., Costa, M. L. G., Ximenes, R. C. C., & Sougey, E. B. (2019, April 1). Is the Montreal Cognitive Assessment (MoCA) screening superior to the Mini-Mental State Examination (MMSE) in the detection of mild cognitive impairment (MCI) and Alzheimer’s Disease (AD) in the elderly? International Psychogeriatrics. Cambridge University Press. https://doi.org/10.1017/S1041610218001370
Reijs, B. L. R., Ramakers, I. H. G. B., Köhler, S., Teunissen, C. E., Koel-Simmelink, M., Nathan, P. J., … Visser, P. J. (2017). Memory Correlates of Alzheimer’s Disease Cerebrospinal Fluid Markers: A Longitudinal Cohort Study. Journal of Alzheimer’s Disease, 60(3), 1119–1128. https://doi.org/10.3233/JAD-160766
Serrano-Pozo, A., Betensky, R. A., Frosch, M. P., & Hyman, B. T. (2016). Plaque-associated local toxicity increases over the clinical course of Alzheimer disease. American Journal of Pathology, 186(2), 375–384. https://doi.org/10.1016/j.ajpath.2015.10.010
Snitz, B. E., Weissfeld, L. A., Lopez, O. L., Kuller, L. H., Saxton, J., Singhabahu, D. M., … DeKosky, S. T. (2013). Cognitive trajectories associated with β-amyloid deposition in the oldest-old without dementia. Neurology, 80(15), 1378–1384.
https://doi.org/10.1212/WNL.0b013e31828c2fc8
Tentolouris-Piperas, V., Ryan, N. S., Thomas, D. L., & Kinnunen, K. M. (2017, January 15). Brain imaging evidence of early involvement of subcortical regions in familial and sporadic Alzheimer’s disease. Brain Research. Elsevier B.V.
https://doi.org/10.1016/j.brainres.2016.11.011
Thomas, B., Sheelakumari, R., Kannath, S., Sarma, S., & Menon, R. N. (2019). Regional cerebral blood flow in the posterior cingulate and precuneus and the entorhinal cortical atrophy score differentiate mild cognitive impairment and dementia due to Alzheimer disease.
American Journal of Neuroradiology, 40(10), 1658–1664. https://doi.org/10.3174/ajnr.A6219
Toombs, J., Foiani, M. S., Wellington, H., Paterson, R. W., Arber, C., Heslegrave, A., … Zetterberg, H. (2018). Amyloid β peptides are differentially vulnerable to preanalytical surface exposure, an effect incompletely mitigated by the use of ratios. Alzheimer’s and
Dementia: Diagnosis, Assessment and Disease Monitoring, 10, 311–321. https://doi.org/10.1016/j.dadm.2018.02.005
Traschütz, A., Enkirch, S. J., Polomac, N., Widmann, C. N., Schild, H. H., Heneka, M. T., & Hattingen, E. (2020). The Entorhinal Cortex Atrophy Score Is Diagnostic and Prognostic in Mild Cognitive Impairment. Journal of Alzheimer’s Disease, Preprint(Preprint), 1–10.
https://doi.org/10.3233/jad-181150
Villemagne, V. L., Burnham, S., Bourgeat, P., Brown, B., Ellis, K. A., Salvado, O., … Masters, C. L. (2013). Amyloid β deposition,
neurodegeneration, and cognitive decline in sporadic Alzheimer’s disease: A prospective cohort study. The Lancet Neurology, 12(4), 357–367. https://doi.org/10.1016/S1474-4422(13)70044-9
Vos, S. J. B., Van Rossum, I. A., Verhey, F., Knol, D. L., Soininen, H., Wahlund, L. O., … Visser, P. J. (2013). Prediction of Alzheimer disease in subjects with amnestic and nonamnestic MCI. Neurology, 80(12), 1124–1132. https://doi.org/10.1212/WNL.0b013e318288690c Xu, W., Tan, L., Wang, H. F., Jiang, T., Tan, M. S., Tan, L., … Yu, J. T. (2015). Meta-analysis of modifiable risk factors for Alzheimer’s disease.
