Associations Between the Subjective Cognitive Decline-Questionnaire’s Scores,Gray Matter Volume,and Amyloid-β Levels

Abstract

Background:

Exploring the relationship between Alzheimer’s disease (AD) biomarkers and subjective cognitive decline (SCD) is needed for better defining its clinical meaning in preclinical AD (preAD).

Objective:

To assess the association between the Subjective Cognitive Decline Questionnaire (SCD-Q), gray matter (GM), and cerebrospinal fluid amyloid-β (Aβ).

Methods:

56 cognitively healthy older adults and their informants answered the SCD-Q. Correlations between GM and SCD-Q scores were explored using structural voxel-based morphometry models including Aβ levels. SCD-Q^*Aβ vectors were calculated with higher scores reflecting higher SCD and cerebral amyloid, simultaneously. Subjects were classified according to their perception of cognitive worsening in the last two years, exploring for GM differences between-groups.

Results:

Higher self-reported SCD-Q scores correlated with reduced GM in the right frontal lobe and increased volumes in the occipital lobe, calcarine sulcus, fusiform gyrus, and cerebellum, while higher informant’s scores correlated with increased GM in the right middle temporal gyrus. Correlations were more significant for SCD-Q language items, self-complaints, and more positive than negative correlations were found. The SCD-Q^*Aβ vectors were negatively associated with GM both in self and informant’s reports. Finally, lower Aβ levels related to lower GM in subjects who noticed cognitive worsening, but related to higher GM in subjects who have not noticed this decline.

Conclusions:

Our results suggest that SCD-Q scores relate with incipient brain changes that may be due to preAD. Independent studies are needed to confirm our observations.

Keywords

Amyloid-β gray matter preclinical Alzheimer’s disease subjective cognitive decline subjective cognitive decline-questionnaire

INTRODUCTION

Preclinical Alzheimer’s disease (preAD) is the earliest stage of AD and is defined by the presence of AD pathology in the absence of objective cognitive impairment [1]. There is great interest in characterizing preAD as it provides a critical opportunity for trialing disease modifying-therapies [2]. Although it has been proposed that the presence of amyloid-β (Aβ) burden may be sufficient to diagnose AD [1, 3], there is postmortem evidence of a significant proportion of individuals with amyloidosis who did not express clinical symptoms antemortem [4]. Thus, there is a contention of whether AD should be defined solely by the expression of AD biomarkers, highlighting the urge to identify additional markers capable of detecting disease progression in a relative short time frame before the onset of overt clinical symptoms [5].

Although subjects with preAD by definition do not meet the criteria for cognitive impairment, it is possible that they are already experiencing subtle cognitive decline [2]. Identifying these first cognitive changes may increase the prognostic value of the preAD definition, and is especially relevant to the development of preventive strategies. Currently, given the absence of definitive answers concerning the algorithm of progression to dementia in cognitively healthy subjects with a positive profile of AD biomarkers, the diagnosis of preAD is restricted to the research field. If we were capable of identifying a preAD profile with a higher likelihood of progression to definitive clinical AD, we could stop excluding a large number of patients who are not yet expressing a full-blown dementia from diagnosis and treatment [6].

In this context, there is an increasing interest in describing the presence of subjective cognitive decline (SCD) in preAD, as it could be the earliest symptomatic manifestation of AD. SCD is defined as the self-experienced worsening of cognitive abilities in subjects who are cognitively unimpaired [7, 8]. Recently, studies have shown an association between SCD and the presence of AD biomarkers, structural and functional brain changes in areas typically affected in AD, and higher risk of future dementia, in samples of cognitively healthy subjects [9 –13]. However, the nature of the link between SCD and AD pathology has not been comprehensively understood yet [14, 15]. While various studies have found structural and functional changes in AD-related brain areas in subjects with SCD [15 –17], other studies have failed to replicate these results [18 –21]. For example, Stewart et al. [17] found a significant association between SCD and reduction in global brain volume and hippocampus, whereas Rowe et al. [21] did not find volumetric differences between SCD and control subjects. Similarly, a number of studies have found a correlation between SCD and the presence of abnormal levels Aβ [13], while others have not found such link [18, 19].

This divergence is mainly due to the highly variable methodology used for defining and measuring SCD [20 , 22–25]. For instance, the concept of SCD has been used indistinctly and interchangeably with terms such as ‘subjective memory impairment’, ‘subjective memory decline’, ‘memory complaints’, and ‘subjective cognitive impairment’, among others [24]. At the same time, the methods used to explore and quantify the presence of SCD varies from the use of single questions to empirically and non-empirically derived questionnaires of diverse extension, that include a different array of cognitive domains, different points of reference to compare one’s current cognitive abilities, and different response options types [20, 23]. This way, an individual might be classified as SCD or non-SCD in different studies, depending on the operationalization used. For this reason, the Subjective Decline Initiative (SCD-I) workgroup has recently developed a common framework for research on SCD in preAD, which could enable comparability of research efforts across studies and settings [20]. According to this framework, SCD must be conceptualized as a self-experienced persistent decline, as opposed to impairment, in cognitive capacity, not restricted to memory solely, in comparison with a previously normal status in subjects who perform within the normal range in standardized cognitive tests.

On the other hand, the multi-factorial nature of SCD and its high prevalence in older adults (i.e., 25–50%) challenges its clinical value [5 , 27]. SCD may arise from numerous conditions normally affecting older adults, such as age-associated cognitive decline, physical health problems, chronic diseases, reduced activities of daily living, and sedentarism [26]. At the same time, studies have suggested a strong correlation between depressive and anxiety symptoms, and the presence of SCD [28, 29]. Therefore, SCD by itself may never be sufficient to diagnose preAD, and preAD is still defined uniquely by the presence of an AD biomarker profile [20].

Nevertheless, it could be possible that cognitive concerns in preAD differ from the concerns related with normal aging and psychological symptoms. By developing a collation of evidence from multiple observations, we could potentially produce a set of overlapping phenomenological features of SCD in preAD that uniquely signifies this condition [30]. For this reason, the SCD-I has encouraged researchers to identify specific SCD items by assessing their relationship to AD biomarkers, recognizing that the current knowledge is insufficient to comprehensively define the specific features of SCD in preAD [20].

Responding to this need, our research team developed and validated the Subjective Cognitive Decline-Questionnaire (SCD-Q) [31] based on the SCD-I guidelines [20]. The SCD-Q has emerged as a useful tool for measuring SCD incorporating the decliner and the informant’s perspective. It includes two parallel forms: MyCog (an abbreviation for “my cognition”), assessing the subjects’ perceptions of their cognitive decline, and TheirCog (an abbreviation of “their cognition”) representing informants’ perceptions. It distinguishes from the pre-existing SCD questionnaires in that it explores the perception of decline, as opposed to impairment, in a relatively short period of time (i.e., last 2 years), exploring the self-perceived performance in an array of daily life activities that involve multiple cognitive domains, instead of being restricted to memory. All of these properties increase the likelihood of detecting SCD due to preAD [20].

The questionnaire has been validated, showing high convergent validity, internal consistency, and discriminant power to distinguish between subjects with cognitive impairment and those without [31]. In previous works, we have explored the performance of the SCD-Q in preAD samples, finding a correlation between SCD-Q scores and the presence of cerebrospinal fluid (CSF) biomarkers of AD in otherwise cognitively healthy subjects, with a higher predictive value of informants’ reports compared to self-reports [32]. At the same time, our results have shown a higher specificity to preAD of SCD-Q items related with language and executive decline, compared to memory items [32, 33].

In the present study, we intend to extend our results by exploring the relationship between the SCD-Q scores and volumetric brain differences in a sample of cognitively healthy subjects. Given that volumetric brain differences are unspecific to preAD, the interaction between CSF AD-biomarkers, SCD-Q, and volumetric brain changes will be assessed. This study differentiates from previous literature in that: 1) it follows the SCD-I framework in the conceptualization and assessment of SCD; 2) it includes CSF AD biomarkers in the model; 3) it includes self and informant’s reports of SCD, and 4) it explores for a profile of specific SCD items with a higher predictive value for brain pathology. We expect to find a negative association between brain’s volume and SCD-Q scores, moderated by CSF Aβ levels. Based on our previous work [32, 33], we also expect to find stronger associations of language and executive SCD-Q components and gray matter (GM), compared with memory, both in self and informant’s reports of SCD.

METHODS

Subjects

Fifty-six cognitively unimpaired older adults were recruited through convenience sampling at Hospital Clinic’s memory clinic in Barcelona, Spain. Of the subjects recruited, 34 (61%) came to the clinic seeking medical assistance due to cognitive complaints. They were invited to participate in the study after a neurologist and a neuropsychologist at the clinic had excluded cognitive impairment, neurological disorders, and other medical conditions that may affect cognition. Therefore, the subjects had been diagnosed as cognitively unimpaired prior to being invited to join the study. The remaining 22 (39%) study participants came to the memory clinic as companions of patients. They did not seek medical assistance. All subjects were invited to join as cognitively unimpaired participants, and were asked to find a person who could act as their informant in this study.

Although a proportion of the participants came to the clinic seeking medical assistance while others were only companions of patients, the study subjects were not classified as SCD and non-SCD subjects for the purposes of this study. This was decided because not seeking medical assistance does not necessarily mean absence of SCD. Indeed, in our sample we found that 45% of the subjects who had not sought medical assistance for cognitive complaints answered ‘YES’ to the SCD-Q question 1 (i.e., Do you perceive memory or cognitive difficulties?) and to question 3 (i.e., In the last two years, has your cognition or memory declined?). Therefore, in this study we explored associations between the degree of SCD in cognitively healthy older adults, measured quantitatively by the SCD-Q and levels of AD biomarkers, without making a definition of SCD or dichotomizing the sample between SCD and non-SCD groups.

As a secondary analysis, we did classified subjects according to their answer to the SCD-Q question 3 (i.e., In the last two years, has your cognition or memory declined?). They were classified as perceiving cognitive worsening (CW, n = 27, 48.2%) if they answered ‘YES’, or as non-perceived cognitive worsening (no-CW, n = 29) if they answered ‘NO’ (n = 29, 51.8%). We chose this question of the SCD-Q, as we believe it is the closest to the SCD-I’s definition of SCD (i.e., self-experienced persistent decline in cognitive capacity in comparison with a previously normal status) [20].

The local ethics committee approved the study and subjects gave written informed consent before undergoing any assessment. The inclusion criteria for recruitment were: 1) Mini-Mental State Examination (MMSE)≥25 [34]; 2) age≥50, 3) absence of cognitive impairment defined as a score≤1.5 SD from the normative mean in every test of a neuropsychological battery (defined below); and 4) absence of functional impairment defined by the Functional Assessment Questionnaire (FAQ) [35]. Exclusion criteria included the presence of any neurological disease, psychiatric disease including major depression or anxiety disorders, or unstable medical condition that could affect cognition, and the presence of suspected non-amyloid pathologies (SNAPS) defined as the co-presence of abnormal CSF total tau (tau) or phosphorylated tau (ptau) levels and normal CSF Aβ levels [36]. SNAPS were excluded in this study because we are primarily interested in characterizing the cognitive changes of patients with AD pathology.

Neuropsychological battery

All subjects were assessed with a comprehensive neuropsychological battery administered by a trained neuropsychologist. It included the following tests: MMSE, Free and Cued Selective Reminding Test (FCSRT-IR) [37], Boston Naming Test (BNT) [38], BDAE’s semantic fluency test (animals) and comprehension of commands BDAE [39], VOSP Incomplete letters and Numbers’ location [40], Trail Making Test part A (TMT-A) [41], and phonetic fluencies (FAS) [42]. Scores were adjusted by age and educational level, and compared with normative values of the Spanish population [43]. Only subjects with a normal performance (i.e., within 1.5 SD from normative mean) in all tests were included in this study. Additionally, given the high correlation of SCD with anxiety and depressive symptoms [29], we quantified the levels of anxiety and depression in the study subjects using the Hospital’s Anxiety and Depression Scale (HADS) [44].

Subjective cognitive decline assessment

The degree of SCD was measured using the SCD-Q [31], a validated questionnaire that follows the SCD-I framework for research of SCD in preAD [20]. It has two parallel forms: one is for the subject to quantify their own self-perceived decline (MyCog, after ‘my cognition’), and the other is for the informant’s report of decline (TheirCog, after ‘their cognition’). The SCD-Q begins with three metacognitive questions: 1) Do you perceive to have a memory or cognitive problem?; 2) Would you ask a doctor about these difficulties?; and 3) In the last two years, has your cognition or memory declined? The first question intends to explore if the subject is concerned about their current cognitive performance, as previous studies have suggested that SCD associated with worry may be associated with a higher risk of conversion to any dementia [45]. The second question has the same intention, as it asks the participant if they consider the problem before-mentioned severe enough to seek medical assistance. The third question intends to explore for SCD following the SCD-I conceptual framework, by asking about decline instead of problems or impairment, and shortening the time frame of referral to exclude the perception of normal cognitive decline associated with aging.

These questions are followed by 24 items assessing perceived decline in instrumental activities of daily living that include memory, language, and executive tasks. Subjects must answer if they believe to be performing these activities worse than roughly two years ago, in a Yes/No format. The total score is computed from the 24 items (i.e., YES = 1, and NO = 0), with higher scores indicating greater perceived cognitive decline. For more details, see Rami et al. (2014) [31].

All participants in the study answered the SCD-Q. Informant reports were also available for 48 subjects. For exploring for SCD characteristics with a higher predictive value of preAD, we included the following SCD-Q components in the VBM analysis (MyCog and TheirCog separately):

SCD-Q total score [0–24]

SCD-Q memory items score (items 1–11) [0–11] [31]

SCD-Q language items score (items 12– 17) [0–6] [31]

SCD-Q executive items score (items 18–24) [0–8] [31]

SCD-Q Episodic Memory factor (EM; items 3,7,9,10,11) [0–5] [33]

SCD-Q Attentional factor (A; items 1,2,4,17,18,24) [0–6] [33]

SCD-Q Organizational skills factor (O; items 8 and 19) [0–2] [33]

SCD-Q Language factor (L; items 15 and 16) [0–2] [33]

SCD-Q Significant items (Sig-items; items 15,16,18,20,22,24) [0–6] [33]

The SCD-Q factors were derived from our previous work using an exploratory factor analysis (EFA) for exploring a preAD-SCD profile that differed significantly from normal aging [33]. In this study, cognitively healthy subjects were classified as preAD or controls according to the absence or presence of abnormal levels of CSF Aβ, respectively. This classification is consistent with the guidelines proposed by the National Institute on Aging and the Alzheimer’s Association (NIA-AA) for defining preAD for research purposes [46]. The Sig-items were also calculated in the same study through an item-analysis of the SCD-Q that compared the frequency of SCD (i.e., YES answers) in each item of the questionnaire between a preAD and a control group. Those items that showed a significantly higher frequency of SCD (p > 0.05) were labelled as Sig-items [33].

Determination of CSF biomarkers

All subjects in the study underwent a lumbar puncture between 9 a.m. and 12 p.m., and a 10-mL sample of CSF was collected. Samples were then centrifuged and stored in polypropylene tubes at –80°C within the first hour after extraction. CSF levels of Aβ, tau, and ptau, were measured by enzyme-linked immunosorbent assay kits (Innogenetics, Ghent, Belgium). In accordance with Hospital Clinic’s laboratory criteria, the following cutoff points for abnormality were considered for each biomarker: 1) Aβ≤550 pg/ml; 2) tau≥400 pg/ml for subjects between 50–70 years old, and≥450 pg/ml for subjects older than 70 years; and 3) ptau≥75 pg/ml. The study participants were blind to the CSF results.

Apolipoprotein E (APOE) analysis

For every participant included in the study, the Genomic DNA was extracted from peripheral blood using the QIAamp DNAblood minikit (Qiagen AG, Basel, Switzerland). APOE genotyping was performed by polymerase chain reaction amplification and HhaI restriction enzyme digestion. Possible APOE genotypes include the combination of alleles APOE ɛ2, APOE ɛ3, and APOE ɛ4. Given that APOE ɛ4 is widely recognized as one of the major genetic risk factors for AD [3], the frequency of carriers of this allele was explored within the total sample and between-groups. The study participants were blind to the APOE results.

Magnetic resonance imaging acquisition

All subjects included in the Study were examined on a 3T magnetic resonance imaging (MRI) scanner (Magnetom Trio Tim, Siemens Medical Solutions, Erlangen, Germany). A high-resolution three-dimensional structural dataset (T1-weighted magnetization-prepared rapid gradient-echo, repetition time = 2300 ms, echo time = 2.98 ms, 240 slices, field of view = 256 mm; matrix size = 256×256; slice thickness = 1 mm) was acquired for the 56 subjects. The mean time interval between the SCD-Q assessment and the MRI was 1 week (SD 11) with a maximum distance of 6 months between them.

Voxel-based morphometry (VBM) analysis

T1 images were segmented into GM, white matter, and CSF using the new segment function implemented in Statistical Parametrical Mapping software (SPM 12, Wellcome Department of Imaging Neuroscience, London, UK). The DARTEL toolbox was used to generate a reference template object of the sample, which was warped into a standard MNI space. The generated flow fields and normalization parameters were then implemented to normalize the GM maps to the MNI space. Jacobian determinants were applied to preserve the local native amount of GM (modulated images). Finally, images were spatially smoothed with an 8 mm full-width at half maximum (FWHM) Gaussian kernel. Total intracranial volume (TIV) was computed by summing the values of GM, white matter, and CSF maps for each individual.

Statistical analysis

Statistical analyses were performed using SPSS package for MAC (V.20). All statistical analyses considered p < 0.05 for significance. Demographical variables and CSF AD biomarkers of the whole sample were described with descriptive statistics, and differences between subjects who have perceived cognitive worsening in the last two years (CW) and those who have not (nCW) were explored using t-Student test for independent samples and Chi-squared analyses. Cross-sectional scores in the SCD-Q were compared between the former using Analysis of Covariance (ANCOVA) after adjusting for age, years of education, and anxiety and depression levels. Bonferroni confidence interval adjustments were included in all ANCOVA to reduce multiple comparisons biases.

Voxel-wise analyses were performed using the general linear model as implemented in SPM 12. The modulated and smoothed GM images were entered in a full-factorial design and correlations between GM volumes and SCD-Q scores, as well as differences between CW and no-CW groups were explored. An uncorrected p-value (<0.001) and a minimum cluster size (k = 100) were used as significance thresholds. Subsequently, a series of multiple regression models were constructed to explore the relationship between GM volume and SCD-Q’s scores (cfr supra). Age, sex, and TIV were used as nuisance variables in all models. CSF Aβ values were also introduced in additional models to explore whether the observed associations varied as a function of amyloid measures. Finally, SCD-Q^*Aβ interaction vectors (for MyCog and TheirCog) were calculated by multiplying the normalized SCD-Q total score (normalized score = score/range) by the normalized CSF Aβ (normalized Aβ=(Aβ-min)/650-min). A value of 650 was used as a constant to define the cut-off for no relevant amyloid pathology. To facilitate the interpretation of such vectors the result was subtracted from 1. Thus, the higher the vector the higher the complaint and Aβ burden (i.e., lower CSF Aβ levels), simultaneously.

RESULTS

Characteristics of the sample

The final sample included 56 cognitively healthy older adults with a mean age of 64.5 (SD 6.9) years and a mean educational level of 11.1 (SD 4.7) years (Table 1). Females represented 69.6% of the total sample. A total of 15 subjects (26,8%) were classified as preAD based on an abnormal level of CSF Aβ (i.e., ≤550 pg/ml), and 7 (12.5%) subjects were carriers of APOE ɛ4. Sample means in the MMSE, HADS anxiety and depressive scores, and CSF AD biomarkers levels are shown in Table 1.

Table 1

Sample characteristics. Data are presented as means±standard deviation. YOE, years of education; MMSE, Mini-Mental State Examination Score (max. 30); HADS-A, Hospital Anxiety and Depression, anxiety subscale; HADS-D, Hospital Anxiety and Depression, depression subscale; CSF Aβ, cerebrospinal fluid levels of amyloid-β isoform 42; CSF tau, cerebrospinal levels of total protein tau; CSF ptau, cerebrospinal fluid levels of phosphorylated protein tau; APOE ɛ4, Apolipoprotein allele 4 carriers; preAD, preclinical Alzheimer’s disease defined as abnormal levels of CSF Aβ (≤550 pg/ml); Total MyCog, total score in self-report of SCD-Q (max.24); Total TheirCog, total score in informant-report of SCD-Q (max. 24)

Total Study Sample (n56)
	Mean (SD)	Minimum	Maximum
Age	64.5 (6.9)	51	80
YOE	11.1 (4.7)	1	26
Females (%)^a	39 (69.6)	–	–
MMSE	28.7 (1.4)	25	30
HADS Anx.	5.3 (2.6)	0	11
HADS Depr.	3.1 (2.4)	0	9
CSF A β	735.9 (267.5)	228.5	1350.7
CSF tau	234.6 (77.6)	92.3	429
CSF ptau	51.6 (14.8)	13.9	86.5
APOEɛ4 (%)^a	7 (13.2)	–	–
preAD (%)^a	15 (26.8)	–	–
Total MyCog	8.3 (6.8)	0	24
Total TheirCog	4.8 (5.7)	0	18

^aData represents total number of subjects.

Gray matter, Aβ, and SCD-Q scores

GM volume and SCD-Q performance

Higher MyCog total scores (i.e., self-reported SCD) correlated with reduced volume in posterior regions of the right middle frontal and precentral gyrus, and increased volumes in the left inferior occipital lobe, fusiform gyrus, cerebellum, and right cuneus, occipital lobe, and calcarine sulcus (Fig. 1a). On the other hand, a positive relationship between TheirCog total scores (i.e., informant’s report of SCD) and GM volume in the right posterior middle temporal gyrus was found, with higher scores correlating with increased GM volume in this area (Fig. 1b). This cluster was initially under the cluster-size significant threshold [k = 94] but after adjusting for CSF Aβ levels it increases its size to k = 118. Overall, correlations between SCD-Q and GM were more abundant and significant for MyCog (i.e., self-report of SCD) than TheirCog (i.e., informant’s reports of SCD).

Fig.1

MyCog total scores and GM in the whole sample. A) Higher MyCog total scores (i.e., higher self-perceived SCD) correlate with increased volumes in the left inferior occipital lobe, fusiform gyrus, cerebellum, and right cuneus, occipital lobe, and calcarine sulcus. B) Higher Mycog total score is associated with reduced volume in posterior regions of the right middle frontal and the precentral gyrus.

Correlations between GM volume and the SCD-Q components are shown in Table 2. In MyCog, the larger clusters and most significant peaks were found for the positive associations between language items and GM volume in posterior cortices (i.e., left occipital/fusiform gyri and medial portion of parietal lobe). Positive relationships in these areas, mainly in the occipital regions, were also found for memory, executive, and Sig-items (i.e., the items that were more frequently endorsed by subjects in preAD). Smaller clusters in frontal areas were also found to be positively associated to language, EM and A factors, and some clusters in the left temporal lobe show association to EM factor. On the other hand, negative relationships (the higher scores, the less the GM) were observed for executive items and regions of the frontal and parietal lobes, and for Sig-items and frontal regions.

Table 2

Voxel-based morphometry results on the linear relationship between SCD-Q components and gray matter volume

Contrast	Association direction	Cluster #	Label	MNI coordinates			Cluster	Peak
				x	y	z	Size	Z-score
MyCog Subscores/Factors
Memory Items	+	Cluster #1	Right Insula	45	6	0	114	3.70
		Cluster #2	Left Middle Occipital	–45	–78	–2	180	3.46
Language Items	+	Cluster #1	Left Inferior Occipital	–54	–75	–6	1133	4.76
			Left Middle Occipital
			Left Fusiform
		Cluster #2	Right Frontal Middle	38	38	23	502	3.95
		Cluster #3	Right Cuneus	8	–87	17	248	3.94
			Left Cuneus
			Right Calcarine
		Cluster #4	Left Rectus	–8	44	–21	100	3.70
			Left Medial Orbitofrontal
		Cluster #5	Right Cuneus	12	–75	23	447	3.58
			Right Precuneus
			Left Calcarine
		Cluster #6	Right Angular	62	–53	36	136	3.49
			Right Supamarginal
			Right Superior Temporal
		Cluster #7	Left Insula	–33	–14	14	114	3.47
Executive Items	+	Cluster #1	Left Middle Occipital	–36	–89	–2	251	3.81
	–	Cluster #1	Right Precentral	47	11	32	524	4.10
			Right Frontal Operculum
			Right Middle Frontal
		Cluster #2	Left Postcentral	–15	–41	75	179	3.80
			Left Superior Parietal
			Left Precuneus
EM factor	+	Cluster #1	Left Middle Temporal	–65	–39	–2	160	4.50
		Cluster #2	Right Insula	47	6	–2	265	4.12
			Right Inferior Frontal Operculum
			Right Rolandic Operculum
		Cluster #3	Left Middle Occipital	–45	–78	–2	526	4.08
			Left Inferior Occipital
			Left Middle Temporal
		Cluster #4	Right Middle Temporal	42	–52	11	237	3.64
A Factor	+	Cluster #1	Right Precentral	50	11	33	151	3.80
			Right Inferior Frontal Operculum
L Factor	+	Cluster #1	Left Middle Temporal	–41	–81	–2	252	4.33
			Left Inferior Occipital
Sig-items	+	Cluster #1	Left Middle Occipital	–42	–83	–2	387	4.26
		Cluster #2	Left Inferior Temporal	–50	–47	–9	118	4.05
			Left Middle Temporal
	–	Cluster #1	Right Inferior Frontal Operculum	44	9	35	234	3.82
			Right Precentral
			Right Middle Frontal
TheirCog Subscores/Factors
Memory Items	+	Cluster #1	Right Middle Temporal	56	–63	17	170	4.17
Language Items	–	Cluster #1	Left Cerebellum Crus 2	–3	–78	–47	285	3.81
A Factor	+	Cluster #1	Right Inferior Occipital	39	–92	–8	141	3.68
O Factor	–	Cluster #1	Right Middle Cingulate	3	15	35	177	3.75
			Left Middle Cingulate
		Cluster #2	Left Middle Frontal	–30	12	57	110	3.56

Automated Anatomical Labeling (AAL) labels are used; MNI, Montreal Neurological Institute;+, positive association; -, negative association; MyCog, self-report of SCD-Q; TheirCog, informant’s report of SCD-Q; memory, SCD-Q items 1–11; language, items 12–17; executive, items 18–24; EM-factor, episodic memory factor; L-factor, language factor; A-factor, attentional factor; O-factor, organizational factor; Sig-items, language and executive items with significantly higher prevalence of complain in preAD compared to controls.

Regarding the informant-reported scores (i.e., TheirCog), lesser significant relationships were found and the significant clusters were smaller compared to MyCog. Positive associations were found between memory items and right middle temporal lobe, and between A factor and right occipital lobe. O factor showed a negative relationship with GM volume in cingulate and lateral frontal regions.

Aβ effect on GM and SCD-Q interaction

The former correlations remained significant after adjusting for CSF Aβ levels, with the exception of MyCog executive items and left occipital lobe, TheirCog A-factor and right occipital lobe, and TheirCog L-factor with left cerebelus. For further exploring the SCD-Q x CSF Aβ effect on GM volume, we calculated a vector by multiplying the normalized SCD-Q total score by the normalized CSF Aβ level. The higher the vector, the higher the complaint and lower the CSF Aβ simultaneously (for more details, refer to Statistical analysis in the Methods section). The resulting vector for MyCog scores and Aβ levels was significantly associated with GM volume, with higher scores showing lower GM volumes in the right angular gyrus, right parietal inferior lobe, and right frontal lobe (Fig. 2a). Also, higher vector scores for the interaction between TheirCog and Aβ levels correlated with lower GM volume in bilateral caudate gyrus, pallidum, putamen, and the right temporal lobe (Fig. 2b).

Fig.2

GM volume and SCD-Q x CSF Aβ interaction. A) The resulting vector for MyCog scores and Aβ levels is associated with GM volume, with higher scores (i.e., higher SCD and lower CSF Aβ simultaneously) associated with lower GM volumes in the right angular gyrus, right parietal inferior lobe, and right frontal lobe. B) Higher vector scores for the interaction between TheirCog and Aβ levels correlate with lower GM volume in bilateral caudate gyrus, pallidum, putamen, and the right temporal lobe.

Fig.3

Cognitive worsening, Aβ, and GM relation. A) In the cognitive worsening (CW) group (i.e., who perceive that their memory or cognition has decline in the last two years) lower levels of CSF Aβ levels relate to lower volume in the anterior portion of the left superior temporal region. B) In the no-CW group, lower CSF Aβ levels related to higher volume in the medial portion of the cingulate cortex.

Perceived cognitive worsening versus non-perceived cognitive worsening

Subjects were classified as perceiving cognitive worsening (CW, n = 27, 48.2%) if they answered ‘YES’, or as non-perceived cognitive worsening (no-CW) if they answered ‘NO’ (n = 29, 51.8%) to the SCD-Q question: “In the last two years, has your cognition or memory declined?” Between-group analyses revealed a significantly lower MMSE score in the CW group compared to the no-CW (t = 2.6; df = 54; p = 0.01). No other significant differences were observed between-groups (Table 3). 29.6% of CW subjects were classified as preAD compared to 24.1% in nCW (p = 0.64), and 8% were carriers of APOE ɛ4 compared to 17.9% in no-CW (p = 0.29).

Table 3

CW and no-CW groups

	CW (n = 27)	no-CW (n = 29)
Age	64.7 (6.9)	64.2 (7.1)
YOE	11.6 (4.8)	10.6 (4.6)
Females^a	21 (77.8%)	18 (62.1%)
MMSE	28.2 (1.4)^*	29.1 (1.3)^*
HADS Anx.	5.9 (2.6)	4.8 (2.5)
HADS Depr.	3.6 (2.5)	2.7 (2.2)
CSF Aβ	742.3 (281.5)	729.8 (258.6)
CSF tau	249.3 (73.7)	221.1 (79.9)
CSF ptau	52.8 (12.7)	50.5 (16.6)
APOEɛ4 carriers^a	2 (8%)	5 (17.9%)
preAD^a	8 (29.6%)	7 (24.1%)
Total MyCog	12.5 (6.3)^**	4.4 (4.6)^**
MyCog memory	6.9 (3.3)^**	1.9 (2.7)^**
MyCog language	3.6 (1.7)^**	1.6 (1.6)^**
MyCog executive	2.9 (2.2)^**	0.9 (1.1)^**
MyCog EM-factor	3.2 (1.8)^**	0.9 (1.4)^**
MyCog L-factor	1.3 (0.6)^**	0.5 (0.6)^**
MyCog A-factor	3.5 (1.9)^**	1.2 (1.6)^**
MyCog O-factor	0.6 (0.8)^*	0.2 (0.4)^*
MyCog Sig-items	2.9 (1.7)	1.1 (1.1)
Total TheirCog	5.9 (5.3)	3.8 (5.9)
TheirCog memory	2.8 (2.7)	1.9 (3.4)
TheirCog language	1.7 (1.7)	0.9 (1.6)
TheirCog executive	1.5 (1.7)	0.9 (1.5)
TheirCog EM-factor	1.0 (1.2)	1.0 (1.6)
TheirCog L-factor	0.7 (0.8)^*	0.2 (0.5)^*
TheirCog A-factor	1.8 (1.6)	0.9 (1.7)
TheirCog O-factor	0.3 (0.5)	0.2 (0.4)
TheirCog Sig-items	1.6 (1.4)^*	0.7 (1.1)^*

Data are presented as means±standard deviation. SCD-Q scores were compared between-groups using ANCOVA with Bonferroni adjustment and controlling for age, YOE, and HADS. YOE, years of education; MMSE, Mini-Mental State Examination score (max. 30); HADS-A, Hospital Anxiety and Depression, anxiety subscale; HADS-D, Hospital Anxiety and Depression, depression subscale; CSF Aβ, cerebrospinal fluid levels of amyloid-β isoform 42; CSF tau, cerebrospinal levels of total protein tau; CSF ptau, cerebrospinal fluid levels of phosphorylated protein tau; APOE ɛ4, Apolipoprotein allele 4 carriers; preAD, preclinical Alzheimer’s disease defined as abnormal levels of CSF Aβ (≤550 pg/ml); MyCog, self-report of SCD-Q; TheirCog, informant’s report of SCD-Q; memory, SCD-Q items 1-11; language, items 12–17; executive, items 18–24; EM-factor, episodic memory factor; L-factor, language factor; A-factor, attentional factor; O-factor, organizational factor; Sig-items, language and executive items with significantly higher prevalence of complain in preAD compared to controls; ^*p < 0.05; ^**p < 0.01. ^aData represents total number of subjects

The SCD-Q scores of the CW and nCW groups can be found on Table 2. As expected, significantly higher scores in MyCog were found for every component analyzed in the CW group, with a mean total score of 12.5 (SCD 6.3) compared to 4.5 (SD 4.6) in the no-CW group (t =–5.2; df = 54; p < 0.001). The differences remained significant after adjusting for age, anxiety, depression levels, and CSF AD biomarkers. Regarding the performance in TheirCog, significantly higher scores were found in the CW group on the L-factor (t =–2.4; df = 47; p = 0.02) and Sig-items (t =–2.4; df = 46; p = 0.02) (Table 3). The results did not change after adjusting for age, anxiety, and depression levels. TheirCog L-factor and Sig-items differences were lost after adjusting for CSF Aβ, tau, and ptau levels.

Subjects reporting CW displayed decreased GM volumes in the left orbitofrontal cortex, extending to the temporal pole, and in the right frontal inferior operculum, On the other hand, a volume increase was observed in the left postcentral gyrus in the CW group. Although these associations did not change after adjusting by CSF Aβ levels, CW and no-CW groups showed different patterns of GM volume and CSF Aβ levels associations in the intragroup analyses: while in the CW group lower levels of CSF Aβ levels related to lower volume in the anterior portion of the left superior temporal region, in the no-CW group lower CSF Aβ levels related to higher volume in the medial portion of the cingulate cortex (Fig. 1). Given the proximity of the cingulate cortex to the corpus callosum, an additional VBM analysis was done on the white matter to confirm that the cluster was not related to a partial volume effect. In the analysis, no significant association with white matter was found.

DISCUSSION

In this article, we have explored the relationship between brain volume and the degree of SCD measured with the SCD-Q, while studying the moderating effect of CSF Aβ in the former. As a main result, we found that higher self-reported SCD (i.e., MyCog scores) correlated with reduced GM in the right frontal lobe, left parietal lobe, and left precuneus, and increased GM in the occipital lobes, temporal lobes, insula, cerebellum, and right precuneus. At the same time, higher informant-reported SCD (i.e., TheirCog scores) correlated with increased GM in the right middle temporal gyrus and right occipital lobe, and decreased volume in the cerebellum, cingulate cortex, and left frontal lobe. Correlations remained significant after adjusting for CSFAβ values. However, SCD-Q^*Aβ vectors were negatively associated with GM both in self and informant’s reports. Finally, lower Aβ levels related to lower GM in subjects who noticed cognitive worsening, but related to higher GM in subjects who have not noticed this decline.

The associations observed in our study include brain areas that are part of the cortical signature of AD, such as the cingulate, precuneus, and temporoparietal neocortex [47]. However, a heterogeneous pattern of volumetric change was found in this study, with some areas showing an increase in volume and others showing volumetric reduction. In general, volume increases were observed in the posterior cortices, especially of the right cortex, while volumetric reductions were observed primarily in the left frontal cortex.

The heterogeneous direction of the correlations between SCD-Q scores and GM could be reflecting a complex interplay between AD pathology and brain changes in preAD. For example, Pegueroles et al. [48] compared the changes in cortical thickness in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) cohort during a 2-year follow-up study of preAD subjects, finding a biphasic trajectory of changes in brain structure where stage 1 subjects (i.e., abnormal Aβ and normal tau levels) showed reduced atrophy rates and stage 2/3 subjects (i.e., abnormal Aβ and abnormal tau levels) presented accelerated atrophy rates. Similarly, in a previous study, Fortea et al. [49] found that cognitively preserved subjects with transitional CSF Aβ levels had thicker cortex in AD vulnerable areas, observing a non-linear relationship between Aβ and GM in preAD. These findings suggest the existence of an inverted U shape in cortical brain changes in this stage of the disease. On the other hand, it has been suggested that at a given timepoint, different brain areas can be at different pathological stages, which could further explain the differences in volumetric changes observed in our study [50].

Regarding the observation of volumetric reductions, our findings are in line with previous studies observing atrophy of neocortical and limbic regions in subjects with SCD [12 , 51]. For example, Saykin et al. [16] found reduced GM density in subjects with cognitive complains relative to controls in distributed brain regions including the frontal lobes, resembling the pattern of atrophy observed in MCI subjects. In another study, Peter et al. [12] found that the GM volumetric pattern in SCD subjects resembled that found in AD-dementia patients, and further observed that this pattern was predictive of an episodic memory decline in the former.

The volumetric reductions found in our study involve brain areas that are affected early in the disease and that are related with episodic memory functioning and executive functions, such as the precuneus, cingulate cortex, and frontal lobes. However, we did not observe significant reductions in the medial temporal lobes (MTL). This was unexpected given that neuroimaging studies have consistently reported MTL affection in early AD [47 , 53]. One possible explanation for this might be the occurrence of an exponential volumetric reduction in the MTL with normal aging, which could challenge the discrimination of brain changes due to preAD in the MTL especially in a small sample size like the present [54, 55].

Surprisingly, we found numerous positive correlations between SCD-Q scores and GM especially in posterior cortices, suggesting that an increased degree of SCD might be associated with an increase in volume in certain brain areas. This result was unexpected and may be due to cofounder variables that could not be controlled given the small sample size and cross-sectional nature of the data. However, after thoroughly reviewing the literature, we did find numerous studies observing brain volume increase and hyperactivation in SCD and preAD samples [48 , 56–60]. For example, Chetelat et al. [56] found that the temporal volume of preAD subjects was larger than that of controls, reflecting larger volume compared to standard volume rather than a lack of atrophic process compared to MCI and AD dementia.

There are various possible explanations for an increase in volume in early AD, such as the presence of inflammatory responses, or the effect of cognitive reserve and compensatory brain mechanisms [48]. Interestingly, hippocampal hypertrophy of the neuronal nuclei and cell bodies have been evidenced at autopsy in the brains of cognitively healthy older adults with amyloidosis, which could be interpreted as an early compensatory cellular response to the underlying neuropathological process, allowing the brain to function normally despite the presence of amyloid plaques [57, 58].

It could be possible, then, that the presence of brain’s compensatory mechanisms may be active in subjects with SCD, allowing them to remain cognitively healthy despite the initiation of AD pathology, and that these mechanisms become less efficient with increasing levels of pathology making the cognitive decline more perceivable. The former hypothesis is consistent with our observation that the vector resulting from the interaction between SCD-Q scores and Aβ levels was significantly associated with lower GM volumes, with higher scores (i.e., reflecting higher SCD and lower CSF Aβ levels) showing lower GM volumes in AD-related areas (i.e., right angular gyrus, right frontal lobe, left parietal lobe, bilateral caudate gyrus, pallidum, putamen, and middle temporal gyrus). Additionally, we found a different pattern of interaction between CSF Aβ levels and GM between the CW (i.e., subjects who have perceived cognitive worsening in the last 2 years) and no-CW groups (i.e., subjects who have not perceived cognitive worsening in the last 2 years): while in the CW group lower levels of CSF Aβ levels related to lower volume in the anterior portion of the left superior temporal region, in the no-CW group lower CSF Aβ levels related to higher volume in the medial portion of the cingulate cortex.

Unexpectedly, the correlations found in this study seemed to be unrelated to CSF Aβ levels. This result was surprising, as we expected the SCD-brain interaction to be mediated by early AD pathological events occurring in the cortex. This result could be explained by different means. First, it is possible that Aβ accumulation and brain atrophy occur as independent events, with subtle brain changes even preceding significant amyloidosis [61, 62]. For example, Braak et al. [61] studied more than two thousand brains from 1- to 100-year-old individuals and concluded that tauopathy, a biomarker of downstream neuronal degeneration, preceded the accumulation of Aβ. Similarly, Amariglio et al. [63] found that both Aβ and reduced hippocampal/entorhinal volumes were independently associated with SCD and that the interaction between Aβ and volume reduction was not significant, suggesting that each biomarker provides an independent and additive association with SCD.

Another objective of the present study was to explore for a specific profile of SCD items that related to GM changes. Given the multifactorial nature of SCD, scientific efforts have been directed toward defining the specific features of SCD that may predict the presence of preAD [20, 64]. In this study, we found more correlations between self-reports (i.e., MyCog) and GM, compared to informant’s reports (i.e., TheirCog). This was unexpected, given our previous study where we found that only TheirCog scores correlated with CSF AD biomarkers in a cognitively healthy sample [32]. Previous studies have mostly found that informant’s reports of SCD have a higher accuracy for identifying objective cognitive difficulties in non-demented adults, and the SCD-I criteria has included the informant’s confirmation of SCD as a SCD-plus criteria that increases the likelihood of preAD in SCD [20 , 66]. There are some studies, though, that have suggested a higher sensitivity of self-reports of SCD for subtle decline, compared to informant’s reports, in the very earliest stages of cognitive decline [67]. Therefore, the relative value in terms of sensitivity and specificity for preAD of self versus informant’s reports of SCD is still a question to be resolved.

Regarding the identification of the cognitive domains of the SCD-Q with a stronger relationship with brain volume, we found that language and executive items were more related to GM than memory items. This is consistent with our previous work, where we found that executive and language items of the SCD-Q, but not memory items, discriminated preAD from a control sample [33]. Although studies have shown that episodic memory is one of the first cognitive functions to decline in preAD [68], memory complaints are highly present in normal aging with 60–70% of older adults perceiving memory difficulties, undermining their specificity to AD pathology [27, 28].

Previous studies have found executive decline in preAD samples, suggesting that clinical effects from early amyloid pathophysiology may precede those from hippocampal atrophy [69 –71]. At the same time, there is previous evidence suggesting that spoken language could be a sensitive measurement for predicting future AD development in cognitively healthy subjects [72 –76]. Our findings support the former findings and the hypothesis that the earliest symptoms of AD are not limited to memory decline [77, 78]. At the same time, they are consistent with the SCD-I workgroup’s suggestion of investigating cognitive decline as opposed to only memory decline in preAD samples [20].

The present study has some limitations that need to be considered. First, given our relatively small sample size, validation in independent samples is needed to confirm our observations. On the other hand, SCD might be over-represented due to the recruitment setting (i.e., memory clinic) and the profile of complaints observed might not represent that found in the community. Another important limitation is the absence of longitudinal follow-up of the subjects, because while cross-sectional studies such as this one provide useful information, ultimately more powerful evidence for the relationship between SCD and brain structural changes throughout the course of preAD will be obtained from serial longitudinal assessments. Our group is following this cohort and is pursuing longitudinal analyses in the future. At last, the CW and no-CW groups must not be interpreted directly as SCD and non-SCD groups. For example, in the CW group, 26% of the subjects did not consider to have a cognitive or memory problem despite the perception of decline, while in the no-CW group, 41% considered to have a cognitive or memory problem although it seemed to remain stable in the last years. Moreover, 45% of no-CW subjects have sought medical assistance for their perceived cognitive problems. Despite the latter, we have chosen to compare both groups following the following the SCD-I framework for research of SCD in preAD, which suggests that because subtle and progressive cognitive decline occurs as part of normal aging processes, reporting decline within a short time frame may have a higher predictive value for preAD, and perceiving decline rather than difficulties may have a higher specificity and sensitivity to AD [20]. Finally, it is recognized that the results obtained may depend on the instrument chosen to measure SCD (i.e., the SCD-Q). It would be interesting to replicate this study using a different instrument designed to study SCD that also follows the SCD-I, in order to confirm our observations.

As a conclusion, our results are promising as we found evidence of an association between SCD-Q scores and incipient brain changes in adults who score within the normal range in an exhaustive battery of standard neuropsychological tests. At the same time, the presence of positive correlations between SCD-Q scores and GM volumes, as well as different GM and CSF Aβ relationships between CW and no-CW groups, could suggest the presence of compensatory brain mechanisms that might explain the absence of cognitive impairment on preAD. Future studies in independent samples are needed to confirm our results.

Footnotes

ACKNOWLEDGMENTS

This study was funded by research grants from the Instituto Carlos III (FIS080036) through the PI14/00563 project of the Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016, co-founded by the Fondo Europeo de Desarrollo Regional (FEDER) “Una manera de hacer Europa”. Dr. Lorena Rami is the recipient of a Miguel Servet Grant as a senior investigator from the Ministry of Science, Spain (CP08/00147). Finally, we would like to thank all the volunteers for their participation in this study.

Authors’ disclosures available online ().

References

Jack

, Albert

, Knopman

, McKhann

, Sperling

, Carrillo

, William

, Phelps

(2011) Introduction to the recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 257–262.

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

Jr , Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wag- ster

, Phelps

(2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging and the Alzheimer’s Association workgroup. Alzheimers Dement 7, 1–13.

Liu

, Kanekiyo

, Xu

, Bu

(2013) Apolipoprotein E and Alzheimer disease: Risk, mechanisms and therapy. Nat Rev Neurol 9, 106–118.

Bennett

, Schneider

, Arvanitakis

, Kelly

, Aggar-wal

, Shah

, Wilson

(2006) Neuropathology of older persons without cognitive impairment from two community-based studies. Neurology 66, 1837–1844.

Dubois

, Hampel

, Feldman

, Scheltens

, Aisen

, Andrieu

, Bakardijan

, Benali

, Bertram

, Blennow

, Broich

, Cavedo

, Crutch

, Dartigues

, Duyckaerts

, Epelbaum

, Frisoni

, Gauthier

, Genthon

, Gouw

, Habert

, Holtzman

, Kivipelto

, Lista

, Molinuevo

, O’Bryant

, Rabinovici

, Rowe

, Salloway

, Schenider

, Sperling

, Teichmann

, Carrillo

, Cummins

, Jack

Jr (2016) Preclinical Alzheimer’s disease: Definition, natural history, and diagnostic criteria. Alzheimers Dement 12, 292–323.

Dubois

(2018) The emergence of a new conceptual framework for Alzheimer’s disease. J Alzheimers Dis 62, 1059–1066.

Tandetnik

, Farrell

, Cary

, Cines

, Emrani

, Karlawish

, Cosentino

(2015) Ascertaining subjective cognitive decline: A comparison of approaches and evidence for using an age-anchored reference group.. J Alzheimers Dis 48, S43–S55.

Colijn

, Grossberg

(2015) Amyloid and tau biomarkers in subjective cognitive impairment. J Alzheimers Dis 47, 1–8.

Jessen

(2014) Subjective and objective cognitive decline at the pre-dementia stage of Alzheimer’s disease. Eur Arch Psychiatry Clin Neurosci 264, 3–7.

10.

Studart Neto

, Nitrini

(2016) Subjective cognitive decline: The first clinical manifestation of Alzheimer’s disease? Dement Neuropsychol 10, 170–177.

11.

Reisberg

, Prichep

, Mosconi

, John

, Glodzik-Sobanska

, Boksay

, Monteiro

, Torossian

, Vedvyas

, Ashraf

, Jamil

, de Leon

(2008) The pre-mild cognitive impairment, subjective cognitive impairment stage of Alzheimer’s disease. Alzheimers Dement 4, S98–S108.

12.

Peter

, Scheef

, Abdulkadir

, Boecker

, Heneka

, Wagner

, Koppara

, Klöppel

, Jessen

, Alzheimer’s Disease Neuroimaging Initiative (2014) Gray matter atrophy pattern in elderly with subjective memory impairment. Alzheimers Dement 10, 99–108.

13.

Perrotin

, Mormino

, Madison

, Hayenga

, Jagust

(2012) Subjective cognition and amyloid deposition imaging: A Pittsburgh compound B positron emission tomography study in normal elderly individuals. Arch Neurol 69, 223–229.

14.

Lista

, Molinuevo

, Cavedo

, Rami

, Amouyel

, Teipel

, Garaci

, Toschi

, Habert

, Blennow

, Zetterberg

, O’Bryant

, Johnson

, Galluzzi

, Bokde

, Broich

, Herholz

, Bakardjian

, Dubois

, Jessen

, Carrillo

, Aisen

, Hampel

(2015) Evolving evidence for the value of neuroimaging methods and biological markers in subjects categorized with subjective cognitive decline. J Alzheimers Dis 48(S1), S171–S191.

15.

Scheef

, Spottke

, Daerr

, Joe

, Striepens

, Kölsch

, Popp

, Daamen

, Gorris

, Heneka

, Boecker

, Biersack

, Maier

, Schild

, Wagner

, Jessen

(2012) Glucose metabolism, gray matter structure, and memory decline in subjective memory impairment. Neurology 79, 1332–1339.

16.

Saykin

, Wishart

, Rabin

, Santulli

, Flashman

, West

, McHugh

, Mamourian

(2006) Older adults with cognitive complaints show brain atrophy similar to that of amnestic MCI. Neurology 67, 834–842.

17.

Stewart

, Godin

, Crivello

, Maillard

, Mazoyer

, Tzourio

, Dufouil

(2011) Longitudinal neuroimaging correlates of subjective memory impairment: 4-year prospective community study. Br J Psychiatry 198, 199–205.

18.

Hollands

, Lim

, Buckley

, Pietrzak

, Snyder

, Ames

, Ellis

, Harrington

, Lautenschlager

, Martins

, Masters

, Villemagne

, Rowe CC Maruff

(2015) Amyloid-ß related memory decline is not associated with subjective or informant rated cognitive impairment in healthy adults. J Alzheimers Dis 43, 677–686.

19.

Pike

, Ellis

, Villemagne

, Good

, Chetelat

, Ames

, Szoeke

, Laws

, Verdile

, Martins

, Masters

, Rowe

(2011) Cognition and beta-amyloid in preclinical Alzheimer’s disease: Data from the AIBL study. Neuropsychologia 49, 2384–2390.

20.

Jessen

, Amariglio

, Van Boxtel

, Breteler

, Ceccaldi

, Chételat

, Dubois

, Defouil

, Ellis

, van der Flier

, Glodzik

, van Harten

, de Leon

, McHugh

, Mielke

, Molinuevo

, Mosconi

, Oosorio

, Perrotin

, Petersen

, Rabin

, Rami

, Reisberg

, Rentz

, Sachdev

, de la Sayette

, Saykin

, Scheltens

, Schulman

, SLavin

, Sperling

, Stewart

, Uspenskaya

, Vellas

, Visser

, Wagner

, Subjective Cognitive Decline (SCD-I) Working Group (2014) A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement 10, 844–852.

21.

Rowe

, Ellis

, Rimajova

, Bourgeat

, Pike

, Jones

, Fripp

, Tochon-Danguy

, Morandeau

, O’Keefe

, Price

, Raniga

, Robins

, Acosta

, Lenzo

, Szoeke

, Slavado

, Head

, Martins

, Masters

, Ames

, Villemagne

(2010) Amyloid imaging results from the Australian Imaging, Biomarkers and Lifestyle (AIBL) study of aging. Neurobiol Aging 31, 1275–1283.

22.

Tales

, Jessen

, Butler

, Wilcock

, Phillips

, Bayer

(2015) Subjective cognitive decline. J Alzheimers Dis 48(S1), S1–S3.

23.

Rabin

, Smart

, Crane

, Amariglio

, Berman

, Boada

, Buckley

, Chételat

, Dubois

, Ellis

, Gifford

, Jefferson

, Jessen

, Katz

, Lipton

, Luck

, Maruff

, Mielke

, Molinuevo

, Naeem

, Perrotin

, Petersen

, Rami

, Resiberg

, Rentz

, Riedel-Heller

, Risacher

, Rodriguez

, Sachdev

, Saykin

, Slavin

, Snitz

, Sperling

, Tandetnik

, van der Flier

, Wagner

, Wolfsgruber

, Sikkes

(2015) subjective cognitive decline in older adults: An overview of self-report measures used across 19 international research studies. J Alzheimers Dis 48(S1), S63–86.

24.

Abdulrab

, Heun

(2008) Subjective memory impairment. A review of its definitions indicates the need for a comprehensive set of standardised and validated criteria. Eur Psychiatry 23, 321–330.

25.

Cheng

, Chen

, Chiu

(2017) From mild cognitive impairment to subjective cognitive decline: Conceptual and methodological evolution. Neuropsychiatr Dis Treat 13, 491–498.

26.

Mark

, Sitskoorn

(2013) Are subjective cognitive complaints relevant in preclinical Alzheimer’s disease? A review and guidelines for healthcare professionals. Rev Clin Gerontol 23, 61–74.

27.

Jonker

, Geerlings

, Schmand

(2000) Are memory complaints predictive for dementia? A review of clinical and population-based studies. Int J Geriatr Psychiatry 15, 983–991.

28.

Slavin

, Brodaty

, Kochan

, Crawford

, Trollor

, Draper

, Sachdev

(2010) Prevalence and predictors of “subjective cognitive complaints” in the Sydney Memory and Ageing Study. Am J Geriatr Psychiatry 18, 701–710.

29.

Jorm

, Butterworth

, Anstey

, Christensen

, Easteal

, Maller

, Mather

, Turakulov

, Wen

, Sachdev

(2004) Memory complaints in a community sample aged 60-64 years: Associations with cognitive functioning, psychiatric symptoms, medical conditions, APOE genotype, hippocampus and amygdala volumes, and white-matter hyperintensities. Psychol Med 34, 1495–1506.

30.

Buckley

, Saling

, Frommann

, Wolfsgruber

, Wagner

(2015) Subjective cognitive decline from a phenomenological perspective: A review of the qualitative literature. J Alzheimers Dis 48, S125–S140.

31.

Rami

, Mollica

, García-Sanchez

, Saldafia

, Sanchez

, Sala

, Valls-Pedret

, Castellví

, Olives

, Molinuevo

(2014) The subjective cognitive decline questionnaire (SCD-Q): A validation study. J Alzheimers Dis 41, 453–466.

32.

Valech

, Mollica

, Olives

, Tort

, Fortea

, Lleo

, Belén

, Molinuevo

, Rami

(2015) Informant’s perception of subjective cognitive decline helps to discriminate preclinical Alzheimer’s disease from normal aging. J Alzheimers Dis 48(S1), S87–S98.

33.

Valech

, Tort-Merino

, Coll-Padrós

, Olives

, León

, Rami

, Molinuevo

(2018) Executive and language subjective cognitive decline complaints discriminate preclinical Alzheimer’s disease from normal aging. J Alzheimers Dis 61, 689–703.

34.

Folstein

, Folstein

, Mchugh

(1975) “Mini-mental state”: A practical method for grading the cognitive state of patients for the clinician. J Psychiatry Res 12, 189–198.

35.

Pfeffer

, Kurosaki

, Harrah

, Chance

, Filos

(1982) Measurement of functional activities in older adults in the community. J Gerontol 37, 323–329.

36.

Jack

, Knopman

, Chetelat

, Dickson

, Fagan

, Frisoni

, Jagust

, Mormino

, Petersen

, Sperling

, van der Flier

, Villemagne

, Vlasser

, Vos

SJB

(2016) Suspected non-Alzheimer disease pathophysiology-concept and controversy. Nat Rev Neurol 12, 117–124.

37.

Grober

, Ocepek-Welikson

, Teresi

(2009) The Free and Cued Selective Reminding Test: Evidence of psychometric adequacy. Psychol Sci Q 51, 266–282.

38.

Kaplan

, Goodglass

, Weintraub

(1983) Boston Naming Test, Lea & Febiger, Philadelphia.

39.

Goodglass

, Kaplan

, Barresi

(2000) Boston Diagnostic Aphasia Examination-III Edition. Pearson Canada Assessment Inc., Canada.

40.

Warrington

, James

(1991) Visual Object and Space Perception Battery (VOSP). Thames Valley Test Co, Bury St. Edmunds.

41.

Reitan

(1994) Trail MakingTest (TMT). Reitan Neuropsychology Laboratory, USA.

42.

Benton

, Hamsher

(1976) Multilingual Aphasia Examination Manual, University of Iowa, Iowa City.

43.

Peña-Casanova , Blesa

, Aguilar

, Gramunt-Fombuena

, Gómez-Ansón

, Oliva

, Molinuevo

, Robles

, Barquero

, Antúnez

, Martínez-Parra

, Frank-García

, Fernández

, Alfonso

, Sol

(2009) Spanish Multicenter Normative Studies (NEURONORMA project): Methods and sample characteristics. Arch Clin Neuropsych 24, 307–319.

44.

Zigmond

, Snaith

(1983) The hospital anxiety and depression scale. Acta Psychiatr Scand 67, 361–370.

45.

Jessen

, Wiese

, Bachman

, Eifflaender-Goerfer

, Haller

, Kölsh

, Luck

, Mösch

, van den Bussche

, Wagner

, Wollny

, Zimmermann

, Pentzek

, Riedel-Heller

, Romber

, Weyerer

, Kaduszkiewicz

, Maier

, Bickel

(2010) Prediction of Dementia by Subjective Memory Impairment: Effects of severity and temporal association with cognitive impairment. Arch Gen Psychiatry 67, 414–422.

46.

Jack

, Knopman

, Weigand

, Wiste

, Vemuri

, Lowe

, Kantarci

, Gunter

, Senjem , Ivnik

, Roberts

, Rocca

, Boeve

, Petersen

(2012) An operational approach to National Institute on Aging-Alzheimer’s Association criteria for preclinical Alzheimer disease. Ann Neurol 71, 765–775.

47.

Dickerson

, Bakkour

, Salat

, Feczko

, Pacheco

, Greve

, Grodstein

, Wright

, Blacker

, Rosas

, Sperling

, Atri

, Growdon

, Hyman

, Morris

, Fischl

, Buckner

(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. Cereb Cortex 19, 497–510.

48.

Pegueroles

, Vilaplana

, Montal

, Sampedro

, Alcolea

, Carmona-Iragui

, Clarimon

, Blesa

, Lleó

, Fortea

(2017) Longitudinal brain structural changes in preclinical Alzheimer’s disease. Alzheimers Dement 13, 499–509.

49.

Fortea

, Sala-Llonch

, Bartrés-Faz

, Lladó

, Solé-Padullés

, Bosch

, Antonell

, Olives

, Sanchez-Valle

, Molinuevo

, Rami

(2011) Cognitively preserved subjects with transitional cerebrospinal fluid β-amyloid 1-42 values have thicker cortex in Alzheimer’s disease vulnerable areas. Biol Psychiatry 70, 183–190.

50.

Jack

Jr , Knopman

, Jagust

, Shaw

, Aisen

, Weiner

, Petersen

, Trojanowski

(2010) Hypothetical model of dynamic biomarkers of the Alzheimer’s pathological cascade. Lancet Neurol 9, 119–128.

51.

Jessen

, Feyen

, Freymann

, Tepest

, Maier

, Heun

, Schild

, Scheef

(2006) Volume reduction of the entorhinal cortex in subjective memory impairment. Neurobiol Aging 27, 1751–1756.

52.

Kim

, Kim

, Lee

(2013) Hippocampus-precuneus functional connectivity as an early sign of Alzheimer’s disease: A preliminary study using structural and functional magnetic resonance imaging data. Brain Res 1495, 18–29.

53.

Squire

, Stark

CEL

, Clark

(2004) The medial temporal lobe. Annu Rev Neurosci 27, 279–306.

54.

Sullivan

(2017) Patterns of brain atrophy on magnetic resonance imaging and the boundary between ageing and Alzheimer’s disease. Rev Clin Gerontol 19, 295–307.

55.

Jernigan

, Archibald

, Fennema-Notestine

, Gamst

, Stout

, Bonner

, Hesselink

(2001) Effects of age on tissues and regions of the cerebrum and cerebellum. Neurobiol Aging 22, 581–594.

56.

Chételat

, Villemagne

, Pike

, Baron

, Bourgeat

, Jones

, Faux

, Ellis

, Salvado

, Szoeke

, Martins

, Ames

, Masteres

, Rowe

, Australian Imaging Biomarkers and Lifestyle Study of Ageing (AIBL) Research Group (2010) Larger temporal volume in elderly with high versus low beta-amyloid deposition. Brain 133, 3349–3358.

57.

Riudavets

, Iacono

, Resnick

, O’Brien

, Zonderman

, Martin

, Rudow

, Pletnikova

, Troncoso

(2007) Resistance to Alzheimer’s pathology is associated with nuclear hypertrophy in neurons. Neurobiol Aging 28, 1484–1492.

58.

Iacono

, Markesbery

, Gross

, Pletnikova

, Rudow

, Zandi

, Troncoso

(2009) The Nun study: Clinically silent AD, neuronal hypertrophy, and linguistic skills in early life. Neurology 73, 665–673.

59.

Rodda

, Dannhauser

, Cutinha

, Shergill

, Walker

(2009) Subjective cognitive impairment: Increased prefrontal cortex activation compared to controls during an encoding task. Int J Geriatr Psychiatry 24, 865–874.

60.

Rodda

, Dannhauser

, Cutinha

, Shergill

, Walker

(2011) Subjective cognitive impairment: Functional MRI during a divided attention task. Eur Psychiatry 26, 457–462.

61.

Braak

, Thal

, Ghebremedhin

, Del Tredici

(2011) Stages of the pathologic process in Alzheimer disease: Age categories from 1 to 100 years. J Neuropathol Exp Neurol 70, 960–969.

62.

Jack

, Holtzman

(2013) Biomarker modeling of Alzheimer’s disease. Neuron 80, 1347–1358.

63.

Amariglio

, Mormino

, Pietras

, Marshall

, Vannini

, Johnson

, Sperling

, Rentz

(2015) Subjective cognitive concerns, amyloid-β, and neurodegeneration in clinically normal elderly. Neurology 85, 56–62.

64.

Sun

, Yang

, Lin

, Han

(2015) Biochemical and neuroimaging studies in subjective cognitive decline: Progress and perspectives. CNS Neurosci Ther 21, 768–775.

65.

Rabin

, Wang

, Katz

, Derby

, Buschke

, Lipton

(2012) Predicting Alzheimer’s disease: Neuropsychological tests, self reports, and informant reports of cognitive difficulties. J Am Geriatr Soc 60, 1128–1134.

66.

Carr

, Gray

, Baty

, Morris

(2000) The value of informant versus individual’s complaints of memory impairment in early dementia. Neurology 55, 1724–1726.

67.

Slot

RER

, Verfaillie

SCJ

, Overbeek

, Timmers

, Wesselman

LMP

, Teunissen

, Dols

, Bouwman

, Prins

, Brakhof

, Lammertsma

, van Berckel

BNM

, Scheltens

, Sikkes

SAM

, van der Flier

(2018) Subjective Cognitive Impairment Cohort (SCIENCe): Study design and first results. Alzheimers Res Ther 10, 1–13.

68.

Hedden

, Oh

, Younger

, Patel

(2013) Ovid: Meta-analysis of amyloid-cognition relations in cognitively normal older adults. Neurology 89, 1341–1348.

69.

Harrington

, Chiang

, Pogoda

, Gomez

, Thomas

, Marion

, Miller

, Siddarth

, Yi

, Zhou

, Lee

, Arakaki

, Cowan

, Tran

, Charleswell

, Ross

, Fonteh

(2013) Executive function changes before memory in preclinical Alzheimer’s pathology: A prospective, cross-sectional, case control study. PLoS One 8, e79378.

70.

Van Harten

, Smits

, Teunissen

, Visser

, Koene

, Blankenstein

, Scheltens

, van der Flier

(2013) Preclinical AD predicts decline in memory and executive functions in subjective complaints. Neurology 81, 1409–1416.

71.

Grober

, Hall

, Lipton

, Zonderman

, Resnick

, Kawas

(2008) Memory impairment, executive dysfunction, and intellectual decline in preclinical Alzheimer’s disease. J Int Neuropsychol Soc 14, 266–278.

72.

Fleming

(2014) Early detection of cognitive-linguistic change associated with mild cognitive impairment. Commun Disord Q 35, 146–157.

73.

Hoffmann

, Nemeth

, Dye

, Pákáski

, Irinyi

, Káiman

(2010) Temporal parameters of spontaneous speech in Alzheimer’s disease. Int J Speech Lang Pathol 12, 29–34.

74.

Schröder

, Wendelstein

, Felder

(2011) Language in the preclinical stage of Alzheimer’s Disease: Content and complexity in biographical narratives. Alzheimers Dement 7, S602.

75.

Cuetos

, Arango-Lasprilla

, Uribe

, Valencia

, Lopera

(2007) Linguistic changes in verbal expression: A preclinical marker of Alzheimer’s disease. J Int Neuropsychol Soc 13, 433–439.

76.

Papp

, Mormino

, Amariglio

, Munro

, Dagley

, Schultz

, Johnson

, Sperling

, Rentz

(2016) Biomarker validation of a decline in semantic processing in preclinical Alzheimer’s disease. Neuropsychology 30, 624–630.

77.

Bäckman

(2008) Memory and cognition in preclinical dementia: What we know and what we do not know. Can J Psychiatry 53, 354–361.

78.

Petersen

, Wiste

, Weigand

, Rocca

, Roberts

, Mielke

, Lowe

, Knopman

, Pankratz

, Machulda

, Gede

, Jack

Jr (2016) Association of elevated amyloid levels with cognition and biomarkers in cognitively normal people from the community. JAMA Neurol 73, 85–92.