A Systematic Review of Neuroimaging Studies Comparing Individuals with Subjective Cognitive Decline to Healthy Controls

Abstract

Background:

Individuals with subjective cognitive decline (SCD) are hypothesized to be the earliest along the cognitive continuum between healthy aging and Alzheimer’s disease (AD), although more research is needed on this topic. Given that treatment approaches may be most effective pre-clinically, a primary objective of emerging research is to identify biological markers of SCD using neuroimaging methods.

Objective:

The current review aimed to comprehensively present the neuroimaging studies on SCD to date.

Methods:

PubMed and PsycINFO databases were searched for neuroimaging studies of individuals with SCD. Quality assessments were completed using the Appraisal tool for Cross-Sectional Studies.

Results:

In total, 62 neuroimaging studies investigating differences between participants with SCD and healthy controls were identified. Specifically, the number of studies were as follows: 36 MRI, 6 PET, 8 MRI/PET, 4 EEG, 7 MEG, and 1 SPECT. Across neuroimaging modalities, 48 of the 62 included studies revealed significant differences in brain structure and/or function between groups.

Conclusion:

Neuroimaging methods can identify differences between healthy controls and individuals with SCD. However, inconsistent results were found within and between neuroimaging modalities. Discrepancies across studies may be best accounted for by methodological differences, notably variable criteria for SCD, and differences in participant characteristics and risk factors for AD. Clinic based recruitment and cross-sectional study design were common and may bias the literature. Future neuroimaging investigations of SCD should consistently incorporate the standardized research criteria for SCD (as recommended by the SCD-Initiative), include more details of their SCD sample and their symptoms, and examine groups longitudinally.

Keywords

Healthy controls neuroimaging subjective cognitive decline systematic review

INTRODUCTION

Globally, dementia affects roughly 43.8 million older adults [1]. One of the most common forms of dementia includes Alzheimer’s disease (AD), an incurable neurodegenerative disorder that accounts for 60–70% of all dementia cases [2]. Currently, the disease-modifying treatments for AD aim to prevent symptom progression. However, much of this research has been focused on treating patients with amnestic mild cognitive impairment (aMCI) or AD itself who have already experienced significant neurodegeneration that has impacted their cognition and ability to carry out instrumental activities of daily living. In light of this, one objective of emerging research in AD has shifted towards identifying pre-clinical biomarkers in individuals at risk for developing AD. Broadly, the goal of the emerging research is to find a way of identifying individuals who will later develop AD before significant neurocognitive symptoms develop; a time point which is believed to be the period during which disease-modifying treatments are most effective [3 –5].

More recently, some studies in pre-clinical AD have begun to examine individuals with subjective cognitive decline (SCD). Individuals with SCD have been conceptualized as the group showing the earliest signs of decline on the cognitive continuum towards AD (i.e., between healthy aging and MCI) [6]. In 2014, an international working group on SCD was formed to propose a framework for research on SCD [6]. Specifically, Jessen et al. [6] describe individuals with SCD to self-report a decline in their own cognitive abilities, but these self-reports are not corroborated by impaired performance on neuropsychological assessment measures. Instead, these individuals perform within normal limits on these measures [6]. Within the literature, healthy controls (HC) are essentially the same as those with SCD, in terms of demographics and normal performance on neuropsychological assessment; the main difference between these two groups is the subjective report of cognitive decline.

In 2017, Molinuevo and colleagues of the SCD-Initiative [7] published a research article providing recommendations for operationalizing SCD and implementing the SCD criteria in research studies as outlined by Jessen et al. [6]. Notably, Jessen and colleagues [6] recognize that the experience of SCD is not specific to preclinical AD. In fact, SCD may be experienced for a variety of reasons such as normal aging, psychiatric conditions, other neurological conditions, the impact of substance use, or from medications. In recognition of these different causes of SCD, Jessen et al. [6, 8] have outlined additional criteria known as SCD-plus, which represents additional features of SCD that signal a higher risk of future development of AD. Such risk factors include but are not limited to concern of experiencing SCD, onset of SCD at age 60 years or older, persistence of SCD over time, and seeking medical attention for these symptoms [8]. Despite the creation of the research framework for SCD, advancement in SCD research has not progressed as rapidly as other areas of dementia research. This stalled progression may be attributed to a lack of common terminology and research standards across studies [9].

Being able to differentiate between those with SCD relative to HC is important but inherently difficult, given that individuals with SCD demonstrate normal levels of cognitive functioning as measured on neuropsychological assessment measures. As paper-and-pencil neuropsychological assessment measures are unable to differentiate between those with SCD and healthy individuals, there is a distinct need for brain-based biomarkers within the SCD group that could signal future risk of developing AD.

Given that biomarkers have potential to advance the diagnosis of AD, the NIA-AA released revised guidelines on the diagnosis of AD to include pathological criteria in addition to clinical observations [10, 11]. Currently, the diagnostic criteria for AD has integrated three different categories of biomarkers, including 1) accumulation of amyloid-β (Aβ) plaques or associated pathologic state (as measured by amyloid positron emission tomography (PET) imaging, low cerebrospinal fluid (CSF) Aβ₄₂, or CSF Aβ₄₀/Aβ₄₂ ratio), 2) accumulation of tau tangles or associated pathologic state (measured by tau PET or CSF phosphorylated tau (P-tau), and 3) neuronal injury or neurodegeneration (measured by structural magnetic resonance imaging (MRI), fluorodeoxyglucose (FDG) PET, or CSF total tau (T-tau)) [12]. Although not included in this research framework of AD, a multitude of different neuroimaging techniques have been used to study disease progression of AD. These methods include other MRI metrics (i.e., functional MRI (fMRI), diffusion-tensor imaging (DTI), etc.), electroencephalography (EEG), magnetoencephalography (MEG), and multi-modal methods incorporating a combination of neuroimaging techniques. Although these techniques are not included in the NIA-AA criteria, they may be especially relevant moving forward because MRI and EEG are widely available and thus ecologically relevant to clinicians and researchers compared to techniques such as PET which are often only available at major medical centers.

Importantly, a main thrust of biomarker research has focused on applying neuroimaging techniques to better understand brain-based changes in individuals with SCD relative to HC. Therefore, the major objectives of the current systematic review are: to comprehensively present the neuroimaging studies on SCD to date, to synthesize findings across studies and imaging modalities, and to identify steps for future research that will move this field forward. In the context of these aims, the current review sought to address four specific questions based on the current literature: 1) Can differences be detected between individuals with SCD and HC using a variety of neuroimaging methods?; 2) Are the detected differences consistent within each neuroimaging technique?; 3) Are the regions (structural and functional) affected in SCD consistent with those expected in early-stage AD?; and 4) Are there methodological characteristics/differences between the studies that could impact the findings?

METHODS

Eligibility criteria

English-language, peer-reviewed, human studies, that reported original research with full-text availability were considered eligible for inclusion. Additional inclusion criteria based on the research topic included the necessity of an SCD group (or equivalent condition group). It should be noted that given the variability in nomenclature prior to the recommendations of the International Working Group of the SCD-Initiative [6], the applied definition of SCD included individuals who were identified within an article to have cognitive complaints of any type (e.g., specific to memory, or more general cognitive complaints) in combination with neuropsychological testing sufficient to confirm that participants were functioning within normal limits. Studies were included when participants were aged 55 years and above, to account for the approximate 15-year timeframe to AD conversion that has been proposed [13]. Studies with a lower minimum age were included if they provided age-stratified data with a distinct 55 + age group. Finally, included studies must have included the use of a neuroimaging method, as well as an observational study design (no intervention studies were included). The review protocol for this study was registered with PROSPERO (CRD 42021235940).

Search strategy

PubMed and PsycINFO databases were searched using a combination of relevant keywords agreed upon by the authors (see the Supplementary Material for search terms in full). Following an initial screen to exclude ineligible articles and duplicates, the suitability of each article was independently adjudicated by two of the authors (two of AP, LO, VS, JRG). Where there was disagreement between the initial raters, the third author’s rating determined whether a given study was included.

Data analysis and synthesis

Studies were organized according to the neuroimaging methods employed: MRI (including anatomical MRI, DTI, fMRI, and multi-modal MRI), PET, SPECT, EEG, MEG, and multi-modal neuroimaging combining MRI and PET. For each study, the term/construct of interest (e.g., SCD, subjective memory impairment, subjective cognitive complaints, etc.), number of participants, mean sample age, mean sample education, number of males and females, prevalence of APOE ɛ4, recruitment source, and results relevant to SCD are tabulated.

Study quality ratings

Each study was also rated (by one of AP, LO, VS, CMS, JRG) for quality using the Appraisal Tool for Cross Sectional Studies (AXIS) tool [14], to evaluate risk of bias, study design quality, and quality of reporting. The AXIS tool is comprised of 20 questions and was not designed to yield a total quality score out of 20 as each item cannot be weighted equally [14]. We assessed and interpreted items individually to evaluate each study’s overall quality [14]. Although the AXIS tool [14] is designed to be used with cross-sectional studies, to maintain consistency we also used this scale when rating the quality of longitudinal studies.

RESULTS

Systematic review

The search was conducted on February 14, 2021. Initial search results yielded 619 studies, with 395 articles from PubMed and 224 articles from PsycINFO. Following manual removal of duplicates and evident irrelevant entries (e.g., meta-analyses, intervention studies), 505 abstracts were reviewed for inclusion and exclusion criteria. Following this initial review and manual retrieval of known articles that fit inclusion criteria, 296 full text articles were identified for more in-depth review. After the review of each article by two authors, discrepancies between authors were found for 82 articles. The nature of these 82 discrepancies pertained to correlational studies investigating severity of SCD symptoms in relation to neuroimaging, studies lacking a distinct HC group apart from those with SCD, and studies where the outcomes were focused on creation of novel methods. All discrepancies were successfully reconciled by a third reviewer who was not involved in the initial rating of that article. The exclusions primarily resulted from an undefined SCD group (e.g., lack of discrimination between those with SCD and MCI) or lack of reporting on neuroimaging biomarkers (e.g., reporting of cerebrospinal fluid biomarkers). Results of this process are summarized in Fig. 1 using the PRISMA Flow Diagram [15]. A list of the articles excluded from this review is provided in the Supplementary Material. A final sample of 62 articles published between 2010 and 2021 were included in the final review.

Fig. 1

PRISMA flow diagram for the current systematic review.

Table 1 shows the number of studies included in each neuroimaging modality. Table 2 provides brief descriptions of each neuroimaging technique as well as an overarching summary of the results. Grouped by neuroimaging modality, Tables 3–11 present more detailed information on each study including details on the participant sample and relevant SCD results for each included article. Table 12 shows a count of the significant and non-significant findings separated by neuroimaging modality considering the studies that incorporated multiple neuroimaging modalities.

Table 1

Summary of neuroimaging modalities used in the articles included (n = 62)

Neuroimaging modality	Table number	Number of studies
MRI
Anatomical MRI	3	11
DTI	4	6
fMRI	5	9
Multi-Modal MRI	6	10
PET	7	6
MRI & PET	8	8
EEG	9	4
MEG	10	7
SPECT	11	1

DTI, diffusion tensor imaging; EEG, electroencephalography; fMRI, functional magnetic resonance imaging; MEG, magnetoencephalography; MRI, magnetic resonance imaging; PET, positron emission tomography; SPECT, single-photon emission computed tomography

Table 2

Description of brain imaging modalities and summary of results

Modality	Description	Brain regions with detectable difference in SCD-equivalent group versus HC
Anatomical MRI	Anatomical MRI detects atrophy or volume loss as well as changes in tissue characteristics through measuring the nuclear magnetic resonance of protons (i.e., hydrogen atoms in water) to create a computerized 3D image of tissues. Common techniques include voxel-based morphometry and cortical thickness analysis.	Hippocampus [16–19 , 66]; amygdala [17]; dorsal frontal cortex [26]; entorhinal cortex [21–23 , 28]; cholinergic basal forebrain [67]; medial frontal cortex [16]; orbitofrontal gyrus [25]; inferior frontal gyrus [25]; right calcarine gyrus [25]; inferior temporal gyrus [25]; lingual gyrus [25]; precentral gyrus [16]; precuneus [16, 25]; cuneus [16]; anterior cingulate cortex [16]; mid-cingulate areas [25]; temporal-parietal regions [62]; frontal regions [27]
DTI	DTI uses diffusion weighted imaging to detect directionality and magnitude of diffusion which reflects underlying tissue boundaries as well as the microstructural integrity of white matter inferred from the diffusion properties (i.e., Brownian motion) of water molecules. Common metrics include apparent diffusion coefficient, fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AxD), radial diffusivity (RD).	Corpus callosum [25 , 31]; retrosplenial cortices [68]; forceps major and minor [30, 31]; corticospinal tract [29 –31]; superior and inferior longitudinal fasciculi [29, 31]; uncinate fasciculi [28, 69]; internal capsule [29]; corona radiata [29, 31]; thalamic radiation [29–31 , 70]; inferior fronto-occipital fascilulus [30]; left cingulum [30, 32]; cerebellum [31]; entorhinal white matter [23]; hippocampus [23 , 31]; cingulum [25, 69]; posterior cingulate [68]; middle temporal [68]
fMRI	fMRI measures differences in the magnetic properties of oxygenated versus deoxygenated hemoglobin as an indirect measure of brain activity (i.e., blood oxygen level dependent (BOLD) signal). Common methods include resting state functional connectivity (FC) analyses as well as task-based analyses (i.e., correlation between the time course of the BOLD signal between brain regions or with a given task).	Increases in SCD relative to HC: frontal medial cortex with retrosplenial cortex [71]; left thalamus [72]; left inferior temporal lobe [72]; bilateral superior frontal gyrus [60]; right superior temporal gyrus [73]; occipital gyrus [73]; cerebellum posterior lobe [73]; inferior parietal lobe [73]; default mode network [16, 33]; medial visual network [16]
		Decreases in SCD relative to HC: left orbital frontal gyrus [72]; precuneus [60]; retrosplenial cortex and the dorsal medial prefrontal cortex and anterior cingulate cortex [69]; retrosplenial cortex and precuneus [61]; default mode network [61]; medial visual network [16]
		Task-based fMRI differences: Increased activation in the cingulate cortex, precuneus, and middle frontal gyrus during the N-back task [34]; increased activation in the right dorsolateral prefrontal cortex and decreased activation in the right hippocampus during an episodic memory task [74]; increased activation in the bilateral caudate, posterior cingulate, thalamus, and left medial temporal lobe during a divided attention task [35]
PET	PET can detect changes in cellular metabolism or the presence of particular proteins through measuring the spatial localization of a given radiotracer (neuroanatomical location is determined through combination with CT or MRI). Common radiotracers include 18-flourodeoxyglucose (FDG), Pittsburgh compound B (PiB) and 18-flutemetamol.	Higher mean global SUVR values [39]; higher SUVR values in the frontal-, parietal-, and lateral-temporal cortex [40]; higher SUVR values in the precuneus and posterior cingulate cortex [41]; Aβ in various regions [37, 38]; glucose hypometabolism in periventricular regions [42]; glucose hypometabolism in the right precuneus [24]; glucose hypermetabolism in the right medial temporal lobe [24]; higher phosphorylated-tau values [27 , 41]; higher total-tau values [41]
EEG	EEG measures the electrical activity of neurons (primarily the pyramidal neurons of the cortex). Common metrics include waveforms (i.e., alpha, beta, theta, delta, gamma) as well as task-based components associated with specific information processing (e.g., N170, P300).	Abnormal EEG rhythms in alpha, theta, delta cortical sources [75]; topographical differences in regions associated with attention networks on microstate maps [76]; Task-based EEG: Differences in N170 amplitudes when presented with fear stimuli [77]; Differences in P300 amplitudes during an attentional control task [63]
MEG	MEG measures the magnetic field signature of cortical neuronal activity and can detect the temporal sequence of neuronal activation within a neural system as well as the location of activation, when paired with structural imaging techniques (e.g., anatomical MRI).	Lower synchronization values [78]; decreased alpha relative power bilaterally in prefrontal areas and in middle and superior temporal lobes [79]; increased functional connectivity in the anterior network [80]; decreased functional connectivity in the posterior network [80]; decreased network clustering in beta and theta bands [81]; decreased modularity in the alpha band [81]; increased nodal degree in left postcentral node [81]; Task-based MEG: Greater activity in parietal, temporal, occipital-temporal, and prefrontal regions during a memory task [82]
SPECT	SPECT can measure regional cerebral blood flow using radioactive tracers and can detect different variables based on the specific tracer used (e.g., can be used to infer specific neurotransmitters)	Caudate, thalamus, and bilateral temporal regions [44]

Table 3

Sample characteristics and results of anatomical MRI studies (n = 11)

Study	Study characteristics							Neuroimaging results in context of SCD versus HC
	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Archer et al.	SNCI	28¹;	61.9 (8.7)¹;	14.3 (2.8)¹;	13/15¹;	11¹; 2²	Clinic	Compared with HC, those with SNCI (both progressors and non-progressors) did not show significantly greater rates of whole brain atrophy, ventricular change, or hippocampal atrophy. When analyzed separately, SNCI progressors had greater hippocampal atrophy compared to HC, while SNCI non-progressors showed significantly less whole brain atrophy compared to HC.
		4²	70.8 (4.8)²	15.8 (2.5)²	2/2²
Kim et al. [17]	SMI	90	65.8 (8.5)	10.4 (5.1)	28/62	–	Clinic	Those in the SCD group had smaller hippocampal and amygdala volumes compared to HC.
Kuhn et al. [26]	SCD	23³;	71.7 (6.6)³;	12.65 (4.1)³;	9/14³;	6³; 4⁴	Clinic & Community (IMAP+)	Compared to HC, the SCD-clinic group had greater atrophy progression over time in the dorsal frontal cortex. No significant differences were found between the SCD-community group and HC in atrophy progression.
		27⁴	68.3 (8.0)⁴	12.85 (3.6)⁴	16/11⁴
Lauriola et al. [83]	SCD	32	64.8 (6.3)	11.6 (4.2)	11/21	–	Community	No significant differences between SCD and HC groups.
Lindberg et al. [20]	SCD	183	70.5 (5.7)	–	83/100	73	Clinic (BioFINDER)	No significant differences in brain volume were found between SCD and HC groups.
Marcotte et al. [21]	SCD	66	71 (6.4)	–	29/37	–	Clinic & Community	Only the left entorhinal cortex showed greater cortical thickness and volume in HC relative to the SCD group, but effect sizes were limited. No other significant differences were found between SCD and HC.
Meiberth et al. [22]	SMI	41	68.9 (7.2)	14.8 (3.5)	29/12	–	Clinic	Those with SMI showed significant decreases in the left entorhinal cortex compared with HC.
Perrotin et al. [66]	SCD	17	69.1 (8.5)	12.8 (3.9)	12/5	–	Clinic	Compared with HC, those with SCD had significant TIV-normalized volume decreases in global hippocampal volume. Those with SCD also showed significantly greater hippocampal atrophy in the CA1 subfield as well as slight involvement in the subiculum.
Platero et al. [84]	SCD	87	71.7 (5.1)	12.9 (5.5)	19/68	–	Clinic	No significant differences between SCD and HC groups.
Scheef et al. [67]	SCD	24	67 (6.1)	15.0 (3.6)	18/6	7	Clinic	Relative to HC, those with SCD showed a significant volume reduction in the cholinergic basal forebrain, with greatest effects seen in the medial septal nucleus (Ch1), the vertical nucleus of the diagonal band of Broca (Ch2), and the cholinergic component of the nucleus basalis Meynert (Ch4).
Striepens et al. [18]	SMI	21	66.3 (6.2)	14.6 (3.6)	15/6	5 (ɛ4+);	Clinic	Compared with HC, those with SCD showed significant volume reductions in the right amygdala, bilateral hippocampus, and bilateral entorhinal cortex. No significant differences were found in total brain volume between HC and SCD groups.
16 (ɛ4–)

Bold text denotes statistical significance found between SCD and HC groups; ¹SCD non-progressors; ²SCD progressors; ³Recruited from the community; ⁴Recruited from a clinic; –, not reported; APOE ɛ4, Apolipoprotein E4; CA1, Cornu Ammonis 1; Edu, education; F, female; HC, healthy controls; M, male; MRI, magnetic resonance imaging; SCD, subjective cognitive decline; SD, standard deviation; SNCI, symptoms of memory loss but no objective cognitive impairment; SMI, subjective memory impairment; TIV, total intracranial volume

Table 4

Characteristics of DTI Studies (n = 6)

Study	Study characteristics							Neuroimaging results in context of SCD versus HC
	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Li et al. [29]	SCD	27	65.3 (8.0)	11.4 (3.9)	9/18	–	Clinic	Those with SCD showed decreased FA and increased MD in widespread regions, including but not limited to the corpus callosum, corticospinal tracts, superior and inferior longitudinal fasciculi. Those with SCD also showed increased AxD in areas including but not limited to the internal capsule, corona radiata, and posterior thalamic radiation.
Luo et al. [30]	SCD	38	66.42 (6.69)	12.24 (3.412)	17/21	–	Clinic	Compared to HC, those with SCD had significantly decreased FA values in the forceps major and forceps minor, and increased MD in the bilateral anterior thalamic radiation, forceps minor, left corticospinal tract, and inferior fronto-occipital fasciculus, and left cingulum (cingulate gyrus and hippocampus).
Ohlhauser et al. [31]	SCD	30	72.9 (4.8)	16.4 (2.6)	10/20	10	Clinic (ADNI)	Compared to HC, those with SCD showed significantly decreased FA values in the bilateral anterior thalamic radiations, cerebellum, corpus callosum, corticospinal tracts, forceps major and minor, superior and inferior longitudinal fasciculi, and hippocampus. Further, those with SCD showed significantly increased MD values in the bilateral corpus callosum, corticospinal tracts, superior corona radiata, and superior and inferior longitudinal fasciculi relative to HC.
Shao et al. [32]	SMI	20	62.4 (5.3)	10.5 (2.9)	10 : 10	–	Community (BABRI)	Individuals with SCD had lower white matter integrity (e.g., lower FA and higher MD) compared to HC in the left cingulum of the hippocampus.
Shu et al. [70]	SCD	36	63.5 (8.7)	11.5 (4.6)	15/21	11¹	Clinic	Those with SCD demonstrated lower global and local efficiency compared to HC, especially linked to bilateral prefrontal regions and the left thalamus.
Wang et al. [85]	SCD	21	62.9 (9.2)	10.7 (4.1)	6/15	–	Clinic	No significant differences were found between SCD and HC groups. Though not significant, those with SCD fell between HC and those with aMCI on most metrics.

Bold text denotes statistical significance found between SCD and HC groups; ¹: Only 22 participants (of 36) had information on APOE status; –, not reported; aMCI, amnestic mild cognitive impairment; APOE ɛ4, Apolipoprotein E4, AxD, axial diffusivity; BABRI, Beijing Aging Brain Rejuvenation Initiative; Edu, education; F, female; FA, fractional anisotropy; HC, healthy controls; M, male; MD, mean diffusivity; SCD, subjective cognitive decline; SD, standard deviation; SMI, subjective memory impairment.

Table 5

Characteristics of fMRI studies (n = 9)

Study	Study characteristics							Results
	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Contreras et al. [86]	SCD	27¹;	70.11 (9.92)¹;	17.11 (2.36)¹;	9/18¹;	–	Clinic	No significant FC pattern differences were found in SCD compared to HC.
	16²	73.38 (7.95)²	17.37 (1.93)²	8/8²
Dillen et al. [71]	SCI	27	65.7 (7.9)	13.6 (4.2)	12/15	8	Clinic	Compared with HC, those with SCD showed greater levels of FC in the frontal medial cortex using the retrosplenial cortex as the seed region. No differences in FC were seen between SCD and HC groups when using the posterior cingulate cortex as the seed region.
Dillen et al. [87]	SCD	28	65.8 (7.8)	13.6 (4.2)	13/15	8	Clinic	No significant differences were found between HC and SCD groups in frontoparietal coupling.
Dong et al. [72]	SCD	63	65.8 (5.0)	11.9 (2.8)	43/20	18	Clinic	Compared with HC, those with SCD showed significant differences in temporal flexibility and spatiotemporal diversity. Specifically, those with SCD showed decreases in the left orbital frontal gyrus and increases in multiple areas including but not limited to the left thalamus and left inferior temporal lobe.
Dumas et al. [34]	CC	12	57.1 (2.3)	16.0 (2.4)	0/12	–	Community	During a visual-verbal version of the N-back task, those with SCD showed increased activation in the cingulate gyrus, precuneus, and middle frontal gyrus relative to HC. Compared to HC, those with SCD would also recruit additional areas of the working memory network on the N-back task when it increased in difficulty.
Erk et al. [74]	SMI	19	68.4 (5.7)	14.5 (3.5)	14/5	–	Clinic	Compared with HC, those with SCD showed reduced activation in the right hippocampus and increased activation in the right dorsolateral prefrontal cortex during a task of episodic memory recall. No differences in activation were found between SCD and HC groups on a task of working memory.
Rodda et al. [35]	SCI	11	64.0 (59.0–66.0)³	11.0 (10.0–13.0)³	6/5	–	Clinic	Compared with HC, those with SCD showed increased functional activation in the bilateral caudate, posterior cingulate, thalamus, and left medial temporal lobe during a divided attention task.
Van Rooden et al. [88]	SCI	18	66.5 (11.2)	3.5⁴ (3)⁵	13/5	–	Clinic	No significant differences were found between HC and SCD groups in cortical phase shift.
Xue et al. [60]	SCD	10	63.10 (8.77)	13.85 (1.83)	4/10	–	Clinic & Community	Compared with HC, those with SCD showed increased FC in the bilateral superior frontal gyrus and decreased FC in the precuneus.

Bold text denotes statistical significance found between SCD and HC groups; ¹Recruited from the Indiana Alzheimer’s Disease Center; ²Recruited from the Indiana Memory and Aging Study; ³Range; ⁴Median; ⁵Verhage Scale; 𠀓, not reported; APOE ɛ4, Apolipoprotein E4; CC, cognitive complaints; Edu, education; F, female; FC, functional connectivity; HC, healthy controls; M, male; SCD, subjective cognitive decline; SCI, subjective cognitive impairment; SD, standard deviation; SMI, subjective memory impairment.

Table 6

Characteristics of multi-modal MRI studies (n = 10)

Study	Study characteristics								Results
	Imaging modality	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Fan et al. [28]	MRI DTI	SCD	43	66.1 (7.0)	13.6 (2.9)	21/22	–	Clinic	Compared to HC, those with SCD had decreased cortical thickness in medial temporal cortical areas including the bilateral perirhinal cortex, bilateral parahippocampal cortex, and left entorhinal cortex. Further, those with SCD had significantly decreased FA in the right and left uncinate fasciculi compared to HC.
Hafkemeijer et al. [16]	MRI fMRI	SMC	25	71.4 (9.2)	15.3 (2.5)	14/11	–	Clinic	Compared with HC, those SCD showed atrophy in various areas including the hippocampus, medial prefrontal cortex, precentral gyrus, precuneus, cuneus, and the anterior cingulate cortex. Further, those with SCD showed increased FC in areas of the default mode network and medial visual network.
Hong et al. [25]	MRI DTI	SMI	28	70.9 (6.23)	9.2 (5.7)	7/19¹	–	Clinic	Compared with HC, those with SCD showed decreased FA and increased MD in the hippocampal body, anterior cingulum, posterior cingulum, anterior corpus callosum, and posterior corpus callosum. Further, those with SCD showed grey matter atrophy in the left orbito-frontal gyrus, inferior frontal gyrus, right calcarine gyrus, precuneus, lingual gyrus, inferior temporal gyrus, and both mid-cingulate areas, compared with HC.
Kiuchi et al. [89]	MRI DTI	SCI	28	70.5 (7.3)	12.36 (2.23)	9/19	–	Clinic	No significant differences between HC and SCD groups.
Parker et al. [33]	MRI fMRI	SCD	23	72.9 (5.4)	16.7 (3.0)	11/12	7	Clinic (ADNI)	No significant structural differences in grey matter density between SCD and HC as assessed by VBM. Individuals with SCD showed increased and decreased FC in various regions relative to HC in the default mode network.
Ryu et al. [23]	MRI DTI	SMI	18	69.89 (6.26)	10.08 (5.59)	5/13	5	Clinic	Compared with HC, those with SCD showed significantly decreased volume in the entorhinal cortex (but not in the hippocampus). Further, those with SMI showed lower FA and higher MD in entorhinal white matter and in the hippocampal body relative to HC.
Scarapicchia et al. [90]	MRI fMRI	SCD	19	72.2 (5.2)	16.3 (3.16)	9/10	–	Clinic (ADNI)	No significant differences between individuals with SCD and HC in WMH burden or in whole brain resting state BOLD variability in white or grey matter.
Selnes et al. [68]	MRI DTI	SCI	16	59.2 (45–71)²	–	5/11	–	Clinic	Compared with HC, those with SCD had significantly higher MD and radial diffusivity values in the middle temporal, posterior cingulate, and retrosplenial cortices. No significant differences were found between HC and SCD groups for cortical thickness.
Sun et al. [73]	MRI fMRI	SCD	25	65.52 (6.12)	10.64 (4.10)	14/11	–	Clinic	Compared with HC, those with SCD showed increased ALFF values in the right superior temporal gyrus, inferior and middle occipital gyrus, cerebellum posterior lobe, and bilateral inferior parietal lobe. Further, no significant differences in grey matter volume were found between SCD and HC groups.
Viviano et al. [61]	MRI fMRI DTI	SCD	35	68.51 (7.66)	–	13/22	–	Clinic & Community	Relative to HC, those with SCD showed lower FC in posterior memory systems, whose structures largely overlap with the default mode network. The SCD group also demonstrated lower FC between the retrosplenial cortex and the precuneus, relative to HC. Further, SCD and HC groups did not differ on DTI metrics.

Bold text denotes statistical significance found between SCD and HC groups; ¹This study reported 28 participants in the SMI group, but only 19/26 of the participants being female; ²Range; –, not reported; ALFF, amplitude of low-frequency fluctuations; APOE ɛ4, Apolipoprotein E4; BOLD, blood-oxygen-level-dependent; Edu, education; F, female; FA, fractional anisotropy; FC, functional connectivity; HC, healthy controls; M, male; MD, mean diffusivity; SCD, subjective cognitive decline; SCI, subjective cognitive impairment; SD, standard deviation; SMC, subjective memory complaints; SMI, subjective memory impairment; WMH, White Matter Hyperintensity.

Table 7

Characteristics of PET studies (n = 6)

Study	Study characteristics							Results
	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Brugnolo et al. [43]	SMC	32	70.47 (7.68)	10.06 (4.27)	15/17	–	Clinic	No significant differences between HC and SCD groups in glucose metabolism.
Jimenez-Bonilla et al. [39]	SMC	8	65 (11)¹	–	–	–	Clinic	Compared to HC, those with SCD showed significantly higher mean global SUVR.
Rodda et al. [91]	SCI	5	64.2 (3.53)	–	–	–	Clinic	No significant differences were found between HC and SCD groups in PiB uptake ratio.
Snitz et al. [40]	SCD	14	68.1 (4.0)	17.6 (2.07)	5/9	4	Clinic	Compared to HC, those with SCD showed significantly higher SUVR values in the frontal, parietal, and lateral temporal cortex.
Song et al. [42]	SMI	68	69.94 (6.44)	11.72 (3.65)	25/43	–	Clinic	Compared to HC, those with SCD showed glucose hypometabolism in periventricular regions.
Voevodskaya et al. [41]	SCD	49	70.7 (5.7)	12.1 (3.8)	22/27	35	Clinic	Relative to HC (both with and without abnormal levels of CSF Aβ₄₂ levels), those with SCD possessed significantly higher tau values. Relative to HC (without abnormal CSF Aβ₄₂ levels), those with SCD possessed significantly higher SUVR of the precuneus and posterior cingulate cortex using [¹⁸F]-flutemetamol PET.

Bold text denotes statistical significance found between SCD and HC groups; ¹Value for the entire study (not just the SCD group); –, not reported; APOE ɛ4, Apolipoprotein E4; CSF, cerebrospinal fluid; Edu, education; HC, healthy controls; F, female; M, male; PET, positron emission tomography; PiB, Pittsburgh Compound B; SCD, subjective cognitive decline; SCI, subjective cognitive impairment; SD, standard deviation; SMC, subjective memory complaints; SUVR, standardized uptake value ratio

Table 8

Characteristics of multi-modal MRI & PET studies (n = 8)

Study	Study characteristics								Results
	Imaging modality	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Chetelat et al. [62]	MRI PET	SCI	30¹;	72.1 (7.1)¹;	13.7 (3.5)¹;	17/13¹;	1¹;	Community (AIBL)	In those with SCD, there was increased atrophy in temporoparietal regions and in the anterior and posterior cingulate cortex in the high ¹¹C-PIB group compared to the low ¹¹C-PIB group. They also found significant differences in grey matter atrophy in the temporal lobes of SCD individuals with high Aβ deposition compared to HC with high Aβ.
			19²	76.7 (6.5)²	12.7 (3.4)²	7/12²	12²
Chetelat et al. [92]	MRI PET	SCI	49	73.9 (7.2)	13.3 (3.5)	24/25	–	Community (AIBL)	No significant differences were found in global neocortical PiB-SUVR or grey matter volume in SCI compared with HC.
Perrotin et al. [36]	MRI PET	SCD	35³;	70.8 (7.5)³;	12.6 (4.2)³;	14/21³;	8³;	Clinic & Community	Both SCD groups had higher Aβ deposition compared to HC but did not show significant differences between each other. No differences in brain structure were found between HC and SCD from the community. No reported results on brain structure between HC and SCD from clinic sample.
			28⁴	67.6 (7.7)⁴	13.3 (3.4)⁴	17/11⁴	4⁴
Risacher et al. [37]	MRI PET	SMCon	71⁵;	72.5 (5.7)⁵;	16.6 (2.7)⁵;	31/40⁵;	0⁵;	Clinic (ADNI)	No significant differences in glucose metabolism or medial temporal lobe atrophy were found in SCD (ɛ4 + or ɛ4–) compared to HC. However, those with SCD ɛ4 + had greater amyloid deposition across multiple brain regions and higher P-tau levels compared to HC (both with ɛ4 + and ɛ4–).
			33⁶	70.3 (5.2)⁶	17.2 (2.0)⁶	12/21⁶	33⁶
Rowe et al. [38]	MRI PET	SCD	96	71.2 (7.4)	–	47/49	36	Clinic	At the group level, no significant differences were found between SCD and HC in hippocampal volumes and Aβ burden. However, those with SCD ɛ4 + showed significantly higher Aβ burden compared to those with SCD ɛ4– and HC (both with ɛ4 + and ɛ4–). No differences in hippocampal volumes were found when groups were stratified by APOEɛ4 carrier status.
Scheef et al. [24]	MRI PET	SMI	31⁷;	67.6 (6.2)⁷;	14.4 (3.4)⁷;	22/9⁷;	9⁷;	Clinic	SCD group showed hypometabolism in the right precuneus and hypermetabolism in the right medial temporal lobe as well as increased grey matter atrophy in the right hippocampus compared to HC.
			27⁸	67.4 (6.5)⁸	14.7 (2.9)⁸	19/8⁸	9⁸
Toledo et al. [27]	MRI PET	SMC	71⁹;	71.6 (5.2)⁹;	-⁹;	25/46⁹;	24⁹;	Clinic (ADNI)	Results found those with SCD had greater levels of brain atrophy with a frontal pattern of atrophy compared to HC. Those with SCD had higher levels of phosphorylated tau₁₈₁ compared to HC.
			65¹⁰	71.3 (5.2)¹⁰	-¹⁰	22/43¹⁰	21¹⁰
Yasuno et al. [69]	DTI fMRI PET	SCI	23	69.6 (8.0)	13.7 (2.1)	13/10	–	Clinic	Compared with HC, those with SCD had decreased FC between the retrosplenial cortex and the dorsal medial prefrontal cortex and anterior cingulate cortex. SCD subjects also had lower FA in the superior longitudinal fasciculus and higher FA in the left cingulum relative to HC. No significant differences in global cortical mean PiB binding potential were found between HC versus SCD.

Bold text denotes statistical significance found between SCD and HC groups; ¹Low-PiB subjective cognitive impairment group; ²High-PiB subjective cognitive impairment group; ³SCD recruited from community; ⁴SCD recruited from clinic; ⁵APOE ɛ4- group; ⁶APOE ɛ4 + group; ⁷at baseline; ⁸at follow-up; ⁹5^th percentile; ¹⁰10^th percentile; –, not reported; APOE ɛ4, Apolipoprotein E4; Edu, education; F, female; FA, fractional anisotropy; HC, healthy controls; M, male; MD, mean diffusivity; PiB, Pittsburgh Compound B; SCD, subjective cognitive decline; SCI, subjective cognitive impairment; SD, standard deviation; SMC, subjective memory complaints; SMCon, subjective memory concerns; SMI, subjective memory impairment; SUVR, standardized uptake value ratio

Table 9

Characteristics of EEG studies (n = 4)

Study	Study characteristics							Results
	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Babiloni et al. [75]	SMC	53	69 (0.99¹)	12 (0.49¹)	30/23	–	Clinic (DESCRIPA)	Compared to HC, those with SCD showed abnormal EEG rhythms in alpha, delta, and theta cortical sources.
Lazarou et al. [77]	SMI	14	69.9 (6.83)	11.7 (2.58)	7/7	–	Clinic	Compared to HC, those with SCD showed significantly larger N170 amplitudes when viewing facial stimuli of “fear,” no difference in N170 amplitude was seen between these groups when viewing facial stimuli of “anger.” Further, no significant differences between HC and SCD were seen in N170 latencies when viewing facial stimuli of either “anger” or “fear.”
Smailovic et al. [76]	SCD	210	60.0 (6.1)	13.3 (3.6)	79/131	–	Clinic	Topographical between group comparison of microstate maps revealed significant differences between SCD and HC (in regions associated with attention networks).
Smart et al. [63]	SCD	17	69.47 (2.38)	16.53 (2.55)	5/12	–	Community	Compared to HC, those with SCD had weaker P300 amplitudes for rare NoGo and frequent Go stimuli during an attention control task.

Bold text denotes statistical significance found between SCD and HC groups; ¹Standard Error; ²Median; –, not reported; APOE ɛ4, Apolipoprotein E4; Edu, education; EEG, electroencephalography; F, female; HC, healthy controls; M, male; SCD, subjective cognitive decline; SD, standard deviation; SMC, subjective memory complaints; SMI, subjective memory impairment

Table 10

Characteristics of MEG studies (n = 7)

Study	Study characteristics							Results
	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Bajo et al. [78]	SCD	12	72.5 (6)	8.3 (–)	3/9	–	Clinic	Compared with HC, those with SCD had a similar pattern of connectivity, but a lower synchronization value.
Lopez-Sanz et al. [79]	SCD	41	71.6 (4.5)	14.2 (5.8)	9/32	–	Clinic	Compared with HC, those with SCD showed decreased alpha relative power bilaterally in prefrontal areas and in the middle and superior temporal lobes. No significant differences in hippocampal volumes or alpha peak frequency were found between those with SCD and HC.
Lopez-Sanz et al. [80]	SCD	41	71.6 (4.5)	–	–	–	Clinic	Compared to HC, those with SCD showed increased FC in the anterior network and decreased FC in the posterior network.
Lopez-Sanz et al. [81]	SCD	55	71 (5.0)	3.95 (1)	13/42	–	Clinic	Compared to HC, those with SCD showed a decrease in global network clustering in beta and theta bands as well as a decrease in modularity in the alpha band. Further, in the left postcentral node, those with SCD showed a nodal degree increase relative to HC. No differences were found between those with SCD and HC in small-world topology or in hippocampal volumes.
Maestu et al. [82]	SCD	12	72.5 (6)	8.3 (–)	3/9	–	Clinic	Compared to HC, those with SCD showed greater activity during a memory task in the parietal, temporal, occipital-temporal, and prefrontal regions.
Serrano et al. [93]	SCD	49	72.0 (4.4)	14.1 (5.5)	11/38	–	Clinic	No significant differences were found between HC and SCD groups.
Shumbayawonda et al. [94]	SCD	33	71.88 (4.29)	–	12/21	–	Clinic	Compared to HC, those with SCD showed greater mean global complexity in MEG signal source space, although this finding was not statistically significant.

Bold text denotes statistical significance found between SCD and HC groups; –, not reported; APOE ɛ4, Apolipoprotein E4; Edu, education; F, female; FC, functional connectivity; HC, healthy controls; M, male; MEG, magnetoencephalography; SCD, subjective cognitive decline; SD, standard deviation

Table 11

Characteristics of SPECT studies (n = 1)

Study	Study characteristics							Results
	SCD label	N	Mean age (SD)	Mean Edu. (SD)	Sex: M/F	APOEɛ4	Recruitment source
Niwa et al. [44]	SCC	32	68.4 (59–75¹)	12.7 (9–16¹)	17/15	–	Clinic	Compared with HC, those with SCD showed hypoperfusion in the caudate, thalamus, and bilateral temporal regions.

Bold text denotes statistical significance found between SCD and HC groups; ¹Range; –, not reported; APOE ɛ4, Apolipoprotein E4; Edu, education; F, female; HC, healthy controls; M, male; SCC, subjective cognitive concerns; SCD, subjective cognitive decline; SD, standard deviation; SPECT, single-photon emission computed tomography

Table 12

Breakdown of significant and non-significant findings across neuroimaging modalities

Neuroimaging type	Total number of incorporated analyses	Number of sig. findings	Number of non-sig. findings
Anatomical MRI	28	16	12
DTI	13	10	3
fMRI	15	11	4
PET	14	10	4
EEG	4	4	0
MEG	7	5	2
SPECT	1	1	0

The total number of analyses exceeds the number of studies included in this review because this table accounts for studies that incorporated multiple neuroimaging analyses. DTI, diffusion tensor imaging; EEG, electroencephalography; fMRI, functional magnetic resonance imaging; MEG, magnetoencephalography; MRI, magnetic resonance imaging; PET, positron emission tomography; SPECT, single-photon emission computed tomography.

Anatomical MRI results

16 of 28 studies (eight from Table 3, five from Table 6, and three from Table 8) found significant differences between those with SCD compared to HC using anatomical MRI neuroimaging analyses. Across studies that examined brain atrophy in SCD, reductions in grey matter volume were commonly found in the amygdala [16 –18], hippocampus [16 –24], entorhinal cortex [18 , 21–23], precuneus [16, 25], and various frontal areas ([16 , 25–27]. Cortical thickness analyses revealed significantly reduced parenchymal tissue in the left entorhinal cortex [21 , 28].

DTI results

In investigations of the microstructural characteristics of white matter, using DTI, 10 of 13 studies (five from Table 4, four from Table 6, and one from Table 8) found widespread differences between SCD and HC groups in various metrics (such as fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD), or axial diffusivity (AxD)). Although loss of white matter integrity was widespread, common regions of decreased FA in those with SCD included the corpus callosum [25 , 29–31], superior longitudinal fasciculi [29, 31], hippocampus [23 , 31], and uncinate fasciculus [28, 29]. Increased MD in those with SCD were also widespread across studies, with common regions including the corpus callosum [25 , 29–31], superior longitudinal fasciculus [29, 31]), hippocampus [23 , 32], and cingulum [25, 29, 30].

fMRI results

When investigating differences between HC and SCD using fMRI techniques, 11 of 15 studies (six from Table 5, four from Table 6, and one from Table 8) found significant results. Considering resting-state fMRI studies, individuals with SCD exhibited differences in functional connectivity in areas associated with the default mode network (DMN) compared to HC [16, 33]. In studies using task-based fMRI, increased functional activity was found in the precuneus during a divided attention task [34], as well as in the posterior cingulate and thalamus during a working memory task [35]. Interestingly, increased functional connectivity was also seen in the precuneus and thalamus in the studies conducted by Hafkemeijer et al. [16] and Parker et al. [33] using resting-state fMRI.

PET results

Ten of 14 studies (four from Table 7 and six from Table 8) looking at PET yielded significant findings. Six of nine studies that examined Aβ found that individuals with SCD had significantly higher Aβ deposition relative to HC [36 –41]. All three studies that investigated tau levels in SCD versus HC found significantly higher tau values in those with SCD [27 , 41]. Finally, three of four studies found significant glucose hypometabolism in those with SCD compared to their healthy counterparts [24 , 43].

EEG results

There were four studies that used various EEG techniques (e.g., EEG rhythms, amplitudes, and microstate maps) to examine differences between HC and SCD groups (see Table 9). All four studies found significant differences; however, little commonality was found between studies due to the disparate nature of their research questions of interest. See Table 8 for a summary of research findings from EEG studies.

MEG results

Five of the seven studies that used MEG to investigate differences between SCD and HC found significant results. Interestingly, all seven of these studies were comprised of participants from the same recruitment sources, however their sample size increased over time. As each of these studies focused on evaluating unique research questions, the findings from each study yielded unique results which are described in Table 10.

SPECT results

Only one study used SPECT to study differences between HC and those with SCD (see Table 11). This study by Niwa et al. [44] found that individuals with SCD showed hypoperfusion in the caudate, thalamus, and bilateral temporal regions compared to HC.

Study quality ratings

The AXIS tool was used to assess study quality in cross-sectional and longitudinal studies. The AXIS tool does not yield a total numerical score, as a total score would not meaningfully communicate the level of quality across studies as study quality is assessed over 20 components [14]. Table 13 shows a breakdown of the quality ratings for the 62 included studies for each of the 20 items on the AXIS tool. Information on the quality assessment for each individual study can be found in the Supplementary Material.

Table 13

Breakdown of ratings for each item on the AXIS tool across the 62 included studies

AXIS Tool Items:	Y	N	U	N/A
1. Were the aims/objectives of the study clear?	62	0	0	0
2. Was the study design appropriate for the stated aim(s)?	62	0	0	0
3. Was the sample size justified?	26	36	0	0
4. Was the target/reference population clearly defined? (Is it clear who the research was about?)	60	1	0	1
5. Was the sample frame taken from an appropriate population base so that it closely represented the target/reference population under investigation?	49	8	4	1
6. Was the selection process likely to select subjects/participants that were representative of the target/reference population under investigation?	52	6	4	0
7. Were measures undertaken to address and categorize non-responders?	3	13	10	36
8. Were the risk factor and outcome variables measured appropriate to the aims of the study?	60	0	1	1
9. Were the risk factor and outcome variables measured correctly using instruments/measurements that have been trialed, piloted, or published previously?	59	2	1	0
10. Is it clear what was used to determine statistical significance and/or precision estimates? (e.g., p values, CIs)	61	1	0	0
11. Were the methods (including statistical methods) sufficiently described to enable them to be repeated?	57	4	1	0
12. Were the basic data adequately described?	58	4	0	0
13. Does the response rate raise concerns about non-response bias?^*	2	21	13	26
14. If appropriate, was information about non-responders described?	2	19	4	37
15. Were the results internally consistent?	54	4	2	2
16. Were the results for the analyses described in the methods presented?	61	0	1	0
17. Were the authors discussions and conclusions justified by the results?	60	1	1	0
18. Were the limitations of the study discussed?	50	12	0	0
19. Were there any funding sources or conflicts of interest that may affect the authors’ interpretation of the results?^*	8	52	2	0
20. Was ethnical approval or consent of participants attained?	61	0	1	0

AXIS tool, Appraisal Tool for Cross-Sectional Studies; Y, yes; N, no; U, unclear; N/A, not applicable; ^*Item is reverse scored (i.e., N rating is positive).

DISCUSSION

With the goal of synthesizing the existing neuroimaging studies on SCD and identifying steps for future research that will move this field forward, the current study sought to address four specific questions that will be discussed herein.

Can differences be detected between individuals with SCD compared to HC using neuroimaging methods?

A major purpose of this review was to synthesize the neuroimaging literature comparing individuals with SCD to HC. Of the 62 included articles, there were significant findings across each imaging modality. In total, 48 articles identified significant differences between SCD and HC groups. Notably, significant differences were not detected between HC and individuals with SCD in every study. For instance, some studies found significant differences between SCD and other patient groups, such as MCI or AD, while there were no significant differences between the HC and SCD groups. Further, some studies only found differences between HC and SCD when controlling for other variables such as high PiB, or APOE ɛ4, etc. Notably, methodological considerations were very important to the outcomes of each study. Nonetheless, it is clear that significant differences between HC and individuals with SCD have been found using each of the neuroimaging techniques included in the current review. This evidence suggests that there are detectable structural and functional differences in the brains of cognitively normal individuals who report symptoms of SCD compared to those who do not.

Are the detected differences consistent within each neuroimaging technique?

Taken together, 26 of 41 studies that incorporated structural neuroimaging analyses and 31 of 41 studies that incorporated functional neuroimaging analyses found significant differences between HC and those with SCD (Note: The number of incorporated analyses is greater than the total number of studies included in this review as many studies incorporated more than one neuroimaging technique). Overall, there were some common findings within neuroimaging techniques, particularly for the structural imaging approaches. Functional studies that took similar approaches seemed to have consistent results, although the small number of studies and wide variety of analysis techniques employed led to unique findings across studies. Replication of these findings and applying multimodal approaches for the same participants represent important steps for future research.

Are the regions affected in SCD consistent with those expected in early-stage AD?

Across structural MRI studies, we saw areas of atrophy in regions that are consistent with early AD, such as the hippocampus [45, 46] and entorhinal cortex [47 –49]. Although atrophy was detected in brain regions associated with AD, studies also reported structural atrophy beyond the hippocampus and entorhinal cortex, including frontal regions and widespread white matter tracts. It is possible that the additional regions of atrophy identified in these SCD studies relates to the aforementioned variability in etiology of SCD; not all individuals with SCD will convert to AD.

Functional MRI studies of preclinical AD represent a relatively new and emerging area of research. Currently, fMRI biomarkers are not included in the AT(N) biomarker grouping outlined by Jack et al. [12]. Although there are no guidelines for AD specific biomarkers such as altered functional connectivity through analysis of fMRI, previous research has noted the usefulness of resting-state fMRI in its ability to distinguish between groups of HC, MCI, and AD [50 –52]. A recent review examining the potential of functional imaging biomarkers for neurodegenerative disorders examined 95 papers focused on functional connectivity in AD and reported that evidence for reduced connectivity in the DMN (including the precuneus, posterior cingulate, and prefrontal cortex) in AD has been well established [53]. In the current systematic review, the findings for individuals with SCD showed decreases in functional connectivity in the DMN consistent with AD. Remarkably, increases in functional connectivity were also detected, mostly in frontal regions, setting the findings for individuals with SCD apart from those with AD. It is possible that at earlier stages of the continuum of cognitive decline, individuals with SCD are able to compensate for cognitive challenges, as demonstrated by increased functional connectivity (prior to observable structural atrophy). These findings are congruent with the Posterior-Anterior-Shift with Aging (PASA) model, which characterizes a shift in functional activity from posterior regions of the brain to frontal regions of the brain, reflective of executive based compensation [54]. However, at later stages such as AD, these compensatory mechanisms may no longer subsist. To investigate these possibilities, more research is needed to capture individuals with SCD, MCI, and AD using the same methods longitudinally.

Similar to fMRI, EEG, and MEG are not included in the current AT(N) biomarker grouping [12]. Recently, Horvath et al. [55] published a critical review of EEG/ERP biomarkers for AD that included more than 300 citations focused on a wide variety of methodologies and types of dementia. They concluded that a majority of studies detected alterations in AD, with variations in latency, amplitude, and source localization. They noted that “outcome variability is a serious issue” that may primarily relate to sample size. The current review included four EEG studies focused on individuals with SCD that had a wide variety of methods and sample sizes; therefore, it is not yet possible to compare findings across EEG/ERP studies on SCD to findings in AD. Similarly, with all the MEG studies on SCD coming from one sample and only one published report using SPECT, the outcomes of these studies cannot yet be thoroughly compared to findings from studies focused on AD. Overall, techniques such as EEG and MEG offer complementary information to MRI methods. In future dementia research, it will be important for techniques such as these to be anchored to other methods that we have a greater knowledge base in (such as MRI or PET).

Within the PET literature, it has been well established that markers of amyloid and tau are measurable hallmarks of AD. The current review revealed mixed findings within the SCD PET literature, with some studies revealing greater tau and amyloid values in SCD relative to HC groups, but other studies finding no significant differences. These mixed findings are almost certainly attributable to the methodological differences between studies, discussed next.

Are there methodological characteristics/differences between the studies that could impact the findings?

In synthesizing neuroimaging studies on SCD versus HC, several methodological differences that could lead to discrepant findings became apparent. Below we discuss several methodological differences including differences in definitions and characterization of SCD (i.e., lack of standardized terms and criteria for SCD), APOE ɛ4 status, participant recruitment location, study design, and inclusion/exclusion criteria which may have led to inconsistent findings and may be exerting a broader impact on the SCD neuroimaging literature.

Definitions and characterization of SCD

Across the studies in this systematic review, there was a lack of consistency in the terminology and criteria characterizing individuals with subjective reports of cognitive decline. Across the 62 studies, eight different terms were used to describe those with subjective cognitive decline or SCD-equivalent classifications. Overall, the term SCD was used in just over half of the included studies (see Table 14 for the frequency counts of each of these eight terms). Almost all of the studies that used the term SCD were published after 2014, which may suggest these studies are following the guidance of Jessen and colleagues of the SCD-Initiative, an international working group focused on the advancement of knowledge in the realm of SCD. However, some of the studies that used other terms to refer to SCD such as SMI or SCI were also published after 2014, and do not follow the recommendations set out by the SCD-Initiative [6]. As a result, some studies may have only included individuals with subjective complaints in the domain of memory, while others may have included individuals with subjective decline in other domains, such as executive function (which SCD is intended to capture).

Table 14

Frequency counts of each SCD-equivalent classifications

SCD-Equivalent Classification	Abbreviation	Frequency
Subjective Cognitive Decline	SCD	34
Subjective Memory Impairment	SMI	10
Subjective Cognitive Impairment	SCI	9
Subjective Memory Complaints	SMC	5
Cognitive Complaints	CC	1
Subjective Cognitive Concerns	SCC	1
Subjective Memory Concerns	SMCon	1
Symptoms of memory loss but No Cognitive Impairment	SNCI	1
	Total	62

Studies often lacked detailed descriptions of SCD reports such as severity of cognitive concerns, which cognitive domains are affected, onset and length of SCD, the persistence of SCD symptoms, and corroboration of decline from an informant. For example, in documenting and determining inclusion criteria for SCD, some studies only asked one or two questions about whether the person experiences SCD. Meanwhile other studies included completion of questionnaires with specific cut-off scores (e.g., Cognitive Change Index and SCD-Questionnaire), however the rationale for the specified cut-off scores is not typically provided. Furthermore, some studies used comprehensive assessments with several cognitive measures to ensure objective cognitive performance of the participants fell within the normal range, while others only administered cognitive screeners or select subtests from different cognitive measures.

APOE ɛ4 status

In the literature, it is widely accepted that the presence of the APOE ɛ4 allele is associated with an increased risk of developing late-onset AD [56 –59]. However, in this review, there was variability across studies in reporting APOE ɛ4 status in their groups. Out of the 62 included articles, only 20 articles included information on APOE ɛ4 status. These were seen across anatomical MRI, DTI, fMRI, multi-modal MRI, PET, and MRI & PET studies. No articles reported APOE ɛ4 status that utilized EEG, MEG, or SPECT.

Participant recruitment location

Most of the articles included in the current review recruited their samples from memory clinics. Of the 62 articles included in this systematic review, only 11 articles investigated differences between HC and individuals with SCD in community samples, rather than memory clinics. In many studies, it appeared that recruitment from a memory clinic was part of the inclusion criteria for their SCD group. The results of the 11 studies that utilized community recruitment (or both clinic and community recruitment) for their SCD groups were mixed. Eight of these 11 studies found significant differences between HC and individuals with SCD [21 , 60–63]. Community recruitment was not utilized in stand-alone PET, MEG, or SPECT studies. Studies that recruited individuals with SCD from memory clinics may yield participant samples with greater levels of concern about their SCD compared to studies that recruited individuals with SCD from the community who were not actively seeking medical care [7]. Furthermore, as most SCD participants were recruited from memory clinics, most were located within larger cities and these studies tended to include participants with higher levels of education (e.g., a mean of 16 years of education was not uncommon). It is unclear if this represents a sampling bias or whether individuals with higher levels of education are more likely to report and seek care for experiences of SCD. It is also possible that individuals with SCD and lower levels of education may yield different findings.

Study design

Nearly all the studies included in this review were cross-sectional. It is possible that different results would be revealed if within-subject longitudinal study designs were utilized. Investigating SCD longitudinally will be important in understanding who is more likely to progress from SCD to AD-related or non-AD related dementia in the future.

There was also a wide range in sample size observed in the literature. In the current systematic review, sample sizes for the SCD groups ranged from 5 to 210, which may also have impacted the statistical power to detect differences from HC groups.

Inclusion/exclusion criteria

Many of the studies included in this review listed exclusion criteria regarding their sample such as history of a mental disorder, depression, prescription of psychotropic or psychoactive medications, history of stroke, history of addiction, history of traumatic brain injury that could cause cognitive impairment, or structural brain abnormalities such as white matter hyperintensities or vascular problems. Careful consideration of inclusion/exclusion criteria is imperative, given that some of the above mentioned criteria (e.g., depression, vascular problems) may contribute the presentation of SCD and can occur independent of, or be considered integral to, the presentation of AD. Molinuevo and colleagues [7] recommended three areas of information that should be included in research on SCD including information on the measurement approach, defined cut-off scores for subjective report measures of SCD, and defined cut-off scores for measures evaluating objective cognition. In the current review, a large majority of studies adequately described their measurement approach such that they included information on the research environment and recruitment source, the names of the measures used, who completed the measures, how the measures were administered, the comparison group (i.e., HC), and the domains of cognition that were assessed. Although a large majority of the studies reviewed provided detailed information on their measurement approach, several of these studies did not provide defined cut-off scores the subjective and objective report measures. While some studies provided defined cut-off scores for the subjective report measures and/or objective cognitive measures., many simply stated that performance by the SCD and HC groups were within normal limits.

Limitations

The search strategy in the current review was limited to published studies written in English available on the PubMed and PsycINFO databases. Systematic searches were not conducted outside of these databases nor were unpublished studies solicited. It is possible that studies included in this review are skewed towards those with significant findings due to the “file drawer problem” [64]. Although the current systematic review provides a comprehensive synthesis of neuroimaging findings in individuals with SCD, there were several types of studies that were not included. Specifically, this review sought to detail group differences between individuals with SCD compared to HC. This specific criterion led to the exclusion of studies that did not include a separate SCD group from HC (i.e., control groups comprised of healthy individuals with or without reports of SCD). This criterion also excluded studies without a distinct control group (i.e., healthy individuals without SCD) but investigated differences between individuals with SCD relative to a cognitively impaired group, such as MCI or AD. Similarly, studies that only investigated relationships or associations between the severity of SCD symptoms and neuroimaging metrics were also excluded if they did not report group differences in neuroimaging metrics between individuals with SCD and HC.

Future directions

There are several specific recommendations for future research that can be made based on the existing literature. In particular, studies investigating SCD should engage in standardized research practices outlined by the SCD-Initiative, using both the proposed standardized terminology and group classification of individuals with SCD. Further, SCD-plus criteria should be applied to specifically examine individuals with SCD that have an increased risk of future cognitive decline. Notably, the criteria for SCD-plus has been recently updated by Jessen et al. [8] and recommendations for operationalizing SCD have been well described by Molinuevo et al [7]. Relatedly, the inclusion of more details on the SCD groups being studied will be useful in contextualizing the findings of each study. Specifically, including data on the nature of the self-experienced cognitive decline such as persistence of symptoms, length of time experiencing SCD, and the cognitive domain(s) affected, will help contextualize results based on participant characteristics. It will also be important for future studies to include information on the diversity of the samples as well as consider whether there are cultural contributions to the presentation and/or identification of SCD. Collecting information on APOE ɛ4 from participants should be continued in order to investigate the relationship between APOE, SCD, and AD. Additionally, longitudinal study designs should be used to provide valuable information regarding conversion rates and distinguishing between individuals with SCD who progress from subjective to objective cognitive decline based on neuropsychological measures and those who remain cognitively stable. Longitudinal designs would also help determine whether SCD is indicative of early AD or if it is related to other forms of dementia. Furthermore, multiple measurements could help to model the progression of change over time (e.g., Ashford & Schmitt [65]). Finally, along with consistency, multimodal methods and replications with large sample sizes will be crucial for translating research findings on individuals with SCD into practice.

Conclusions

Biomarker research has focused on the identification of pre-clinical indicators of AD. We have completed the most comprehensive overview of published neuroimaging studies comparing individuals with SCD to HC to date. The current review found that differences can be detected between individuals with SCD and HC across all included neuroimaging modalities, anatomical MRI, DTI, fMRI, PET, EEG, MEG, SPECT, and multi-modal neuroimaging. These findings suggest that neuroimaging is capable of detecting differences between individuals experiencing both normal and non-normal aging, even during SCD where no objective impairments in cognition have yet taken place. Although significant differences were found between individuals with SCD and HC across the neuroimaging modalities included in this review, results were inconsistent both within and across neuroimaging modalities likely due to methodological variations. Due to the inconsistencies in methodology, the level of generalizability of the findings across studies is unknown. Since SCD may arise from other etiologies apart from preclinical AD, it is important that future research in this area is designed and conducted purposefully, recognizing the impact of methodology on the study outcome. Implementation of the recommendations for future neuroimaging studies on SCD are imperative to understanding the earliest changes in the brain that relate to cognitive decline and potential future development of AD.

DISCLOSURE STATEMENT

Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/21-5249r2).

Footnotes

The supplementary material is available in the electronic version of this article: .

References

Nichols

, Szoeke

CEI

, Vollset

, Abbasi

, Aichour

MTE

, Akinyemi

, Asgedom

, Awasthi

, Barker-Collo

, Baune

, Béjot

, Belachew

, Bennett

, Biadgo

, Bin Sayeed

, Brayne

, Carpenter

, Cerin

, Djalalinia

, Dubey

, Duken

, Faro

, Fernandes

, Filip

, Gnedovskaya

, Gupta

, Hachinski

, Hamidi

, Hankey

, Haro

, Hay

, Irvani

SSN

, Jonas

, Kalani

, Karch

, Kasaeian

, Khader

, Khalil

, Khan

, Khanna

, Khoja

TAM

, Khubchandani

, Kissimova-Skarbek

, Kivimäki

, Koyanagi

, Majdan

, Malekzadeh

, Mengistu

, Meretoja

, Mohammadi-Khanaposhtani

, Mokdad

, Mondello

, Moradi

, Nagel

, Naghavi

, Naik

, Nguyen

, Nirayo

, Ogbo

, Olagunju

, Owolabi

, Passos

VMdA

, Pereira

, Pinilla-Monsalve

, Poustchi

, Qorbani

, Radfar

, Reiner

, Robinson

, Roshandel

, Rostami

, Safari

, Safiri

, Sawhney

, Sepanlou

, Shaikh

, Sahraian

, Shiri

, Shiue

, Silva

, Sobhani

, Stein

, Tabarés-Seisdedos

, Tovani-Palone

, Tran

, Tsegay

, Ullah

, Vlassov

, Wang

Y-P

, Westerman

, Wyper

GMA

, Yonemoto

, Yousefifard

, Zaidi

, Zare

, Vos

, Feigin

, Murray

CJL

, Collaborators

GBDD

(2019) Global, regional, and national burden of Alzheimer’s disease and other dementias, 1990–2016: A systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 18, 88–106.

World Health Organization, Dementia - Key Facts, https://www.who.int/news-room/fact-sheets/detail/dementia

Berti

, Polito

, Lombardi

, Ferrari

, Sorbi

, Pupi

(2016) Rethinking on the concept of biomarkers in preclinical Alzheimer’s disease. Neurol Sci 37, 663–672.

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

Jr. , Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wagster

, Phelps

(2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 280–292.

Imtiaz

, Tolppanen

A-M

, Kivipelto

, Soininen

(2014) Future directions in Alzheimer’s disease from risk factors to prevention. Biochem Pharmacol 88, 661–670.

Jessen

, Amariglio

, van Boxtel

, Breteler

, Ceccaldi

, Chételat

, Dubois

, Dufouil

, Ellis

, van der Flier

, Glodzik

, van Harten

, de Leon

, McHugh

, Mielke

, Molinuevo

, Mosconi

, Osorio

, Perrotin

, Petersen

, Rabin

, Rami

, Reisberg

, Rentz

, Sachdev

, de la Sayette

, Saykin

, Scheltens

, Shulman

, Slavin

, Sperling

, Stewart

, Uspenskaya

, Vellas

, Visser

, Wagner

(2014) A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement 10, 844–852.

Molinuevo

, Rabin

, Amariglio

, Buckley

, Dubois

, Ellis

, Ewers

, Hampel

, Klöppel

, Rami

, Reisberg

, Saykin

, Sikkes

, Smart

, Snitz

, Sperling

, van der Flier

, Wagner

, Jessen

(2017) Implementation of subjective cognitive decline criteria in research studies. Alzheimers Dement 13, 296–311.

Jessen

, Amariglio

, Buckley

, van der Flier

, Han

, Molinuevo

, Rabin

, Rentz

, Rodriguez-Gomez

, Saykin

, Sikkes

SAM

, Smart

, Wolfsgruber

, Wagner

(2020) The characterisation of subjective cognitive decline. Lancet Neurol 19, 271–278.

Rabin

, Smart

, Amariglio

(2017) Subjective cognitive decline in preclinical Alzheimer’s disease. Annu Rev Clin Psychol 13, 369–396.

10.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr. , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269.

11.

Sosa-Ortiz

, Acosta-Castillo

, Prince

(2012) Epidemiology of dementias and Alzheimer’s disease. Arch Med Res 43, 600–608.

12.

Jack

Jr. , Bennett

, Blennow

, Carrillo

, Dunn

, Haeberlein

, Holtzman

, Jagust

, Jessen

, Karlawish

, Liu

, Molinuevo

, Montine

, Phelps

, Rankin

, Rowe

, Scheltens

, Siemers

, Snyder

, Sperling

, Elliott

, Masliah

, Ryan

, Silverberg

(2018) NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 14, 535–562.

13.

Reisberg

, Gauthier

(2008) Current evidence for subjective cognitive impairment (SCI) as the pre-mild cognitive impairment (MCI) stage of subsequently manifest Alzheimer’s disease. Int Psychogeriatr 20, 1–16.

14.

Downes

, Brennan

, Williams

, Dean

(2016) Development of a critical appraisal tool to assess the quality of cross-sectional studies (AXIS). BMJ Open 6, e011458.

15.

Moher

, Liberati

, Tetzlaff

, Altman

(2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 6, e1000097.

16.

Hafkemeijer

, Altmann-Schneider

, Oleksik

, van de Wiel

, Middelkoop

HAM

, van Buchem

, van der Grond

, Rombouts

SARB

(2013) Increased functional connectivity and brain atrophy in elderly with subjective memory complaints. Brain Connect 3, 353–362.

17.

Kim

M-J

, Seo

, Kim

, Lee

J-M

, Qiu

, Na

(2013) Less depressive symptoms are associated with smaller hippocampus in subjective memory impairment. Arch Gerontol Geriatr 57, 110–115.

18.

Striepens

, Scheef

, Wind

, Popp

, Spottke

, Cooper-Mahkorn

, Suliman

, Wagner

, Schild

, Jessen

(2010) Volume loss of the medial temporal lobe structures in subjective memory impairment. Dement Geriatr Cogn Disord 29, 75–81.

19.

Archer

, Kennedy

, Barnes

, Pepple

, Boyes

, Randlesome

, Clegg

, Leung

, Ourselin

, Frost

, Rossor

, Fox

(2010) Memory complaints and increased rates of brain atrophy: Risk factors for mild cognitive impairment and Alzheimer’s disease. Int J Geriatr Psychiatry 25, 1119–1126.

20.

Lindberg

, Mårtensson

, Stomrud

, Palmqvist

, Wahlund

L-O

, Westman

, Hansson

(2017) Atrophy of the posterior subiculum is associated with memory impairment, Tau- and Aβ pathology in non-demented individuals. Front Aging Neurosci 9, 306.

21.

Marcotte

, Potvin

, Collins

, Rheault

, Duchesne

(2019) Brain atrophy and patch-based grading in individuals from the CIMA-Q study: A progressive continuum from subjective cognitive decline to AD. Sci Rep 9, 13532.

22.

Meiberth

, Scheef

, Wolfsgruber

, Boecker

, Block

, Träber

, Erk

, Heneka

, Jacobi

, Spottke

, Walter

, Wagner

, Hu

, Jessen

(2015) Cortical thinning in individuals with subjective memory impairment. J Alzheimers Dis 45, 139–146.

23.

Ryu

, Lim

, Na

, Shim

, Cho

, Yoon

, Hong

, Yang

(2017) Hippocampal and entorhinal structures in subjective memory impairment: A combined MRI volumetric and DTI study. Int Psychogeriatr 29, 785–792.

24.

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.

25.

Hong

, Yoon

, Shim

, Ahn

, Yang

, Lee

J-H

(2015) Gray and white matter degenerations in subjective memory impairment: Comparisons with normal controls and mild cognitive impairment. J Korean Med Sci 30, 1652–1658.

26.

Kuhn

, Moulinet

, Perrotin

, La Joie

, Landeau

, Tomadesso

, Bejanin

, Sherif

, De La Sayette

, Desgranges

, Vivien

, Poisnel

, Chételat

(2019) Cross-sectional and longitudinal characterization of SCD patients recruited from the community versus from a memory clinic: Subjective cognitive decline, psychoaffective factors, cognitive performances, and atrophy progression over time. Alzheimers Res Ther 11, 61.

27.

Toledo

, Bjerke

, Chen

, Rozycki

, Jack

Jr. , Weiner

, Arnold

, Reiman

, Davatzikos

, Shaw

, Trojanowski

(2015) Memory, executive, and multidomain subtle cognitive impairment: Clinical and biomarker findings. Neurology 85, 144–153.

28.

Fan

L-Y

, Lai

Y-M

, Chen

T-F

, Hsu

Y-C

, Chen

P-Y

, Huang

K-Z

, Cheng

T-W

, Tseng

W-YI

, Hua

M-S

, Chen

Y-F

, Chiu

M-J

(2018) Diminution of context association memory structure in subjects with subjective cognitive decline. Hum Brain Mapp 39, 2549–2562.

29.

X-Y

, Tang

Z-C

, Sun

, Tian

, Liu

Z-Y

, Han

(2016) White matter degeneration in subjective cognitive decline: A diffusion tensor imaging study. Oncotarget 7, 54405–54414.

30.

Luo

, Li

, Qin

, Chen

, Yang

, Huang

, Liu

, Xu

, Bai

, Zhao

(2020) White matter microstructural damage as an early sign of subjective cognitive decline. Front Aging Neurosci 11, 378.

31.

Ohlhauser

, Parker

, Smart

, Gawryluk

(2019) White matter and its relationship with cognition in subjective cognitive decline. Alzheimers Dement (Amst) 11, 28–35.

32.

Shao

, Li

, Zhang

, Yang

, Tao

, Zhang

, Peng

(2019) White matter integrity disruption in the pre-dementia stages of Alzheimer’s disease: From subjective memory impairment to amnestic mild cognitive impairment. Eur J Neurol 26, 800–807.

33.

Parker

, Smart

, Scarapicchia

, Gawryluk

(2020) Identification of earlier biomarkers for Alzheimer’s disease: A multimodal neuroimaging study of individuals with subjective cognitive decline. J Alzheimers Dis 77, 1067–1076.

34.

Dumas

, Kutz

, McDonald

, Naylor

, Pfaff

, Saykin

, Newhouse

(2013) Increased working memory-related brain activity in middle-aged women with cognitive complaints. Neurobiol Aging 34, 1145–1147.

35.

Rodda

, Dannhauser

, Cutinha

, Shergill

, Walker

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

36.

Perrotin

, La Joie

, de La Sayette

, Barré

, Mézenge

, Mutlu

, Guilloteau

, Egret

, Eustache

, Chételat

(2017) Subjective cognitive decline in cognitively normal elders from the community or from a memory clinic: Differential affective and imaging correlates. Alzheimers Dement 13, 550–560.

37.

Risacher

, Kim

, Nho

, Foroud

, Shen

, Petersen

, Jack

, Beckett

, Aisen

, Koeppe

, Jagust

, Shaw

, Trojanowski

, Weiner

, Saykin

(2015) APOE effect on Alzheimer’s disease biomarkers in older adults with significant memory concern. Alzheimers Dement 11, 1417–1429.

38.

Rowe

, Ellis

, Rimajova

, Bourgeat

, Pike

, Jones

, Fripp

, Tochon-Danguy

, Morandeau

, O’Keefe

, Price

, Raniga

, Robins

, Acosta

, Lenzo

, Szoeke

, Salvado

, 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.

39.

Jiménez-Bonilla

, Banzo

, De Arcocha-Torres

, Quirce

, Martínez-Rodríguez

, Sánchez-Juan

, Carril

(2016) Amyloid imaging with 11C-PIB in patients with cognitive impairment in a clinical setting: A visual and semiquantitative analysis. Clin Nucl Med 41, e18–e23.

40.

Snitz

, Lopez

, McDade

, Becker

, Cohen

, Price

, Mathis

, Klunk

(2015) Amyloid-β imaging in older adults presenting to a memory clinic with subjective cognitive decline: A pilot study. J Alzheimers Dis 48, S151–S159.

41.

Voevodskaya

, Sundgren

, Strandberg

, Zetterberg

, Minthon

, Blennow

, Wahlund

L-O

, Westman

, Hansson

(2016) Myo-inositol changes precede amyloid pathology and relate to APOE genotype in Alzheimer disease. Neurology 86, 1754–1761.

42.

Song

I-U

, Choi

, Oh

, Chung

Y-A

, Chung

S-W

(2016) Alteration patterns of brain glucose metabolism: Comparisons of healthy controls, subjective memory impairment and mild cognitive impairment. Acta Radiol 57, 90–97.

43.

Brugnolo

, Morbelli

, Arnaldi

, De Carli

, Accardo

, Bossert

, Dessi

, Famà

, Ferrara

, Girtler

, Picco

, Rodriguez

, Sambuceti

, Nobili

(2014) Metabolic correlates of Rey auditory verbal learning test in elderly subjects with memory complaints. J Alzheimers Dis 39, 103–113.

44.

Niwa

, Kondo

, Sakurada

, Nakagawa

, Imanishi

, Mizuno

(2016) Regional cerebral blood flow in [123]I-IMP single-photon emission computed tomography and the Wechsler Memory Scale-revised in nondemented elderly subjects with subjective cognitive impairment. Intern Med 55, 3571–3578.

45.

Cherbuin

, Sargent-Cox

, Easteal

, Sachdev

, Anstey

(2015) Hippocampal atrophy is associated with subjective memory decline: The path through life study. Am J Geriatr Psychiatry 23, 446–455.

46.

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.

47.

Dickerson

, Goncharova

, Sullivan

, Forchetti

, Wilson

, Bennett

, Beckett

, deToledo-Morrell

(2001) MRI-derived entorhinal and hippocampal atrophy in incipient and very mild Alzheimer’s disease. Neurobiol Aging 22, 747–754.

48.

Jessen

, Feyen

, Freymann

, Tepest

, Maier

, Heun

, Schild

H-H

, Scheef

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

49.

Killany

, Gomez-Isla

, Moss

, Kikinis

, Sandor

, Jolesz

, Tanzi

, Jones

, Hyman

, Albert

(2000) Use of structural magnetic resonance imaging to predict who will get Alzheimer’s disease. Ann Neurol 47, 430–439.

50.

Rombouts

SARB

, Barkhof

, Goekoop

, Stam

, Scheltens

(2005) Altered resting state networks in mild cognitive impairment and mild Alzheimer’s disease: An fMRI study. Hum Brain Mapp 26, 231–239.

51.

Damoiseaux

(2012) Resting-state fMRI as a biomarker for Alzheimer’s disease? Alzheimers Res Ther 4, 8.

52.

Vemuri

, Jones

, Jack

Jr. (2012) Resting state functional MRI in Alzheimer’s disease. Alzheimers Res Ther 4, 2.

53.

Hohenfeld

, Werner

, Reetz

(2018) Resting-state connectivity in neurodegenerative disorders: Is there potential for an imaging biomarker? Neuroimage Clin 18, 849–870.

54.

Davis

, Dennis

, Daselaar

, Fleck

, Cabeza

(2007) Qué PASA? The Posterior–Anterior Shift in Aging. Cereb Cortex 18, 1201–1209.

55.

Horvath

, Szucs

, Csukly

, Sakovics

, Stefanics

, Kamondi

(2018) EEG and ERP biomarkers of Alzheimer’s disease: A critical review. Front Biosci 23, 183–220.

56.

Corder

, Saunders

, Strittmatter

, Schmechel

, Gaskell

, Small

, Roses

, Haines

, Pericak-Vance

(1993) Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer’s disease in late onset families. Science 261, 921–923.

57.

Hohman

, Dumitrescu

, Barnes

, Thambisetty

, Beecham

, Kunkle

, Gifford

, Bush

, Chibnik

, Mukherjee

, De Jager

, Kukull

, Crane

, Resnick

, Keene

, Montine

, Schellenberg

, Haines

, Zetterberg

, Blennow

, Larson

, Johnson

, Albert

, Bennett

, Schneider

, Jefferson

, Alzheimer’s Disease Genetics Consortium and the Alzheimer’s Disease Neuroimaging Initiative (2018) Sex-specific association of apolipoprotein E with cerebrospinal fluid levels of tau. JAMA Neurol 75, 989–998.

58.

Altmann

, Tian

, Henderson

, Greicius

, Alzheimer’s Disease Neuroimaging Initiative (2014) Sex modifies the APOE-related risk of developing Alzheimer disease. Ann Neurol 75, 563–573.

59.

Liu

C-C

, Kanekiyo

, Xu

, Bu

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

60.

Xue

, Yuan

, Yue

, Xu

, Wang

, Wu

, Ji

, Zhou

, Zhao

, Rao

, Yang

, Xiao

, Chen

(2019) Distinct disruptive patterns of default mode subnetwork connectivity across the spectrum of preclinical Alzheimer’s disease. Front Aging Neurosci 11, 307.

61.

Viviano

, Hayes

, Pruitt

, Fernandez

, van Rooden

, van der Grond

, Rombouts

SARB

, Damoiseaux

(2019) Aberrant memory system connectivity and working memory performance in subjective cognitive decline. Neuroimage 185, 556–564.

62.

Chételat

, Villemagne

, Pike

, Baron

J-C

, Bourgeat

, Jones

, Faux

, Ellis

, Salvado

, Szoeke

, Martins

, Ames

, Masters

, Rowe

(2010) Larger temporal volume in elderly with high versus low beta-amyloid deposition. Brain 133, 3349–3358.

63.

Smart

, Segalowitz

, Mulligan

, MacDonald

SWS

(2014) Attention capacity and self-report of subjective cognitive decline: A P3 ERP study. Biol Psychol 103, 144–151.

64.

Rosenthal

(1979) The file drawer problem and tolerance for null results. Psychol Bull 86, 638–641.

65.

Ashford

, Schmitt

(2001) Modeling the time-course of Alzheimer dementia. Curr Psychiatry Rep 3, 20–28.

66.

Perrotin

, de Flores

, Lamberton

, Poisnel

, La Joie

, de la Sayette

, Mézenge

, Tomadesso

, Landeau

, Desgranges

, Chételat

(2015) Hippocampal subfield volumetry and 3D surface mapping in subjective cognitive decline. J Alzheimers Dis 48, S141–S150.

67.

Scheef

, Grothe

, Koppara

, Daamen

, Boecker

, Biersack

, Schild

, Wagner

, Teipel

, Jessen

(2019) Subregional volume reduction of the cholinergic forebrain in subjective cognitive decline (SCD). Neuroimage Clin 21, 101612.

68.

Selnes

, Fjell

, Gjerstad

, Bjørnerud

, Wallin

, Due-Tønnessen

, Grambaite

, Stenset

, Fladby

(2012) White matter imaging changes in subjective and mild cognitive impairment. Alzheimers Dement 8, S112–S121.

69.

Yasuno

, Kazui

, Yamamoto

, Morita

, Kajimoto

, Ihara

, Taguchi

, Matsuoka

, Kosaka

, Tanaka

, Kudo

, Takeda

, Nagatsuka

, Iida

, Kishimoto

(2015) Resting-state synchrony between the retrosplenial cortex and anterior medial cortical structures relates to memory complaints in subjective cognitive impairment. Neurobiol Aging 36, 2145–2152.

70.

Shu

, Wang

, Bi

, Zhao

, Han

(2018) Disrupted topologic efficiency of white matter structural connectome in individuals with subjective cognitive decline. Radiology 286, 229–238.

71.

Dillen

KNH

, Jacobs

HIL

, Kukolja

, von Reutern

, Richter

, Onur

ÖA

, Dronse

, Langen

K-J

, Fink

(2016) Aberrant functional connectivity differentiates retrosplenial cortex from posterior cingulate cortex in prodromal Alzheimer’s disease. Neurobiol Aging 44, 114–126.

72.

Dong

, Yang

, Li

C-sR

, Wang

, Zhang

, Du

, Han

, Tang

(2020) Dynamic network connectivity predicts subjective cognitive decline: The sino-longitudinal cognitive impairment and dementia study. Brain Imaging Behav 14, 2692–2707.

73.

Sun

, Dai

, Li

, Sheng

, Li

, Wang

, Chen

, He

, Han

(2016) Subjective cognitive decline: Mapping functional and structural brain changes-a combined resting-state functional and structural MR imaging study. Radiology 281, 185–192.

74.

Erk

, Spottke

, Meisen

, Wagner

, Walter

, Jessen

(2011) Evidence of neuronal compensation during episodic memory in subjective memory impairment. Arch Gen Psychiatry 68, 845–852.

75.

Babiloni

, Visser

, Frisoni

, De Deyn

, Bresciani

, Jelic

, Nagels

, Rodriguez

, Rossini

, Vecchio

, Colombo

, Verhey

, Wahlund

L-O

, Nobili

(2010) Cortical sources of resting EEG rhythms in mild cognitive impairment and subjective memory complaint. Neurobiol Aging 31, 1787–1798.

76.

Smailovic

, Koenig

, Laukka

, Kalpouzos

, Andersson

, Winblad

, Jelic

(2019) EEG time signature in Alzheimer's disease: Functional brain networks falling apart. Neuroimage Clin 24, 102046.

77.

Lazarou

, Adam

, Georgiadis

, Tsolaki

, Nikolopoulos

, Yiannis Kompatsiaris

, Tsolaki

(2018) Can a novel high-density EEG approach disentangle the differences of visual event related potential (N170), elicited by negative facial stimuli, in people with subjective cognitive impairment? J Alzheimers Dis 65, 543–575.

78.

Bajo

, Castellanos

, López

, Ruiz

, Montejo

, Montenegro

, Llanero

, Gil

, Yubero

, Baykova

, Paul

, Aurtenetxe

, Del Pozo

, Maestu

(2012) Early dysfunction of functional connectivity in healthy elderly with subjective memory complaints. Age (Dordr) 34, 497–506.

79.

López-Sanz

, Bruña

, Garcés

, Camara

, Serrano

, Rodríguez-Rojo

, Delgado

, Montenegro

, López-Higes

, Yus

, Maestú

(2016) Alpha band disruption in the AD-continuum starts in the subjective cognitive decline stage: A MEG study. Sci Rep 6, 37685.

80.

López-Sanz

, Bruña

, Garcés

, Martín-Buro

, Walter

, Delgado

, Montenegro

, Higes

, Marcos

, Maestú

(2017) Functional connectivity disruption in subjective cognitive decline and mild cognitive impairment: A common pattern of alterations. Front Aging Neurosci 9, 109.

81.

López-Sanz

, Garcés

, Álvarez

, Delgado-Losada

, López-Higes

, Maestú

(2017) Network disruption in the preclinical stages of Alzheimer’s disease: From subjective cognitive decline to mild cognitive impairment. Int J Neural Syst 27, 1750041.

82.

Maestu

, Baykova

, Ruiz

, Montejo

, Montenegro

, Llanero

, Solesio

, Gil

, Yubero

, Paul

, Pozo

, Nevado

(2011) Increased biomagnetic activity in healthy elderly with subjective memory complaints. Clin Neurophysiol 122, 499–505.

83.

Lauriola

, Esposito

, Delli Pizzi

, de Zambotti

, Londrillo

, Kramer

, Rabinovici

, Tartaro

(2017) Sleep changes without medial temporal lobe or brain cortical changes in community-dwelling individuals with subjective cognitive decline. Alzheimers Dement 13, 783–791.

84.

Platero

, López

, Tobar

MdC

, Yus

, Maestu

(2019) Discriminating Alzheimer’s disease progression using a new hippocampal marker from T1-weighted MRI: The local surface roughness. Hum Brain Mapp 40, 1666–1676.

85.

Wang

X-N

, Zeng

, Chen

G-Q

, Zhang

Y-H

, Li

X-Y

, Hao

X-Y

, Yu

, Zhang

, Sheng

, Li

Y-X

, Sun

, Li

H-Y

, Song

, Li

K-C

, Yan

T-Y

, Tang

X-Y

, Han

(2016) Abnormal organization of white matter networks in patients with subjective cognitive decline and mild cognitive impairment. Oncotarget 7, 48953–48962.

86.

Contreras

, Avena-Koenigsberger

, Risacher

, West

, Tallman

, McDonald

, Farlow

, Apostolova

, Goñi

, Dzemidzic

, Wu

Y-C

, Kessler

, Jeub

, Fortunato

, Saykin

, Sporns

(2019) Resting state network modularity along the prodromal late onset Alzheimer’s disease continuum. Neuroimage Clin 22, 101687.

87.

Dillen

KNH

, Jacobs

HIL

, Kukolja

, Richter

, von Reutern

, Onur

ÖA

, Langen

K-J

, Fink

(2017) Functional disintegration of the default mode network in prodromal Alzheimer’s disease. J Alzheimers Dis 59, 169–187.

88.

van Rooden

, Buijs

, van Vliet

, Versluis

, Webb

, Oleksik

, van de Wiel

, Middelkoop

HAM

, Blauw

, Weverling-Rynsburger

AWE

, Goos

JDC

, van der Flier

, Koene

, Scheltens

, Barkhof

, van de Rest

, Slagboom

, van Buchem

, van der Grond

(2016) Cortical phase changes measured using 7-T MRI in subjects with subjective cognitive impairment, and their association with cognitive function. NMR Biomed 29, 1289–1294.

89.

Kiuchi

, Kitamura

, Taoka

, Yasuno

, Tanimura

, Matsuoka

, Ikawa

, Toritsuka

, Hashimoto

, Makinodan

, Kosaka

, Morikawa

, Kichikawa

, Kishimoto

(2014) Gray and white matter changes in subjective cognitive impairment, amnestic mild cognitive impairment and Alzheimer’s disease: A voxel-based analysis study. PLoS One 9, e104007.

90.

Scarapicchia

, Garcia-Barrera

, MacDonald

, Gawryluk

(2019) Resting state BOLD variability is linked to white matter vascular burden in healthy aging but not in older adults with subjective cognitive decline. Front Hum Neurosci 13, 429.

91.

Rodda

, Okello

, Edison

, Dannhauser

, Brooks

, Walker

(2010) ¹¹C-PIB PET in subjective cognitive impairment. Eur Psychiatry 25, 123–125.

92.

Chételat

, Villemagne

, Bourgeat

, Pike

, Jones

, Ames

, Ellis

, Szoeke

, Martins

, O’Keefe

, Salvado

, Masters

, Rowe

(2010) Relationship between atrophy and β-amyloid deposition in Alzheimer disease. Ann Neurol 67, 317–324.

93.

Serrano

, López-Sanz

, Bruña

, Garcés

, Rodríguez-Rojo

, Marcos

, Crespo

, Maestú

(2020) Spatiotemporal oscillatory patterns during working memory maintenance in mild cognitive impairment and subjective cognitive decline. Int J Neural Syst 30, 1950019.

94.

Shumbayawonda

, López-Sanz

, Bruña

, Serrano

, Fernández

, Maestú

, Abasolo

(2020) Complexity changes in preclinical Alzheimer’s disease: An MEG study of subjective cognitive decline and mild cognitive impairment. Clin Neurophysiol 131, 437–445.