Functional connectivity alterations in objectively defined subtle cognitive decline: A cross-sectional functional MRI study in cognitively healthy older adults

Abstract

Background

The aging of the population is leading to an increase in the incidence of neurocognitive disorders, particularly Alzheimer's disease, which has become a significant public health priority. Given the importance of early intervention in neurocognitive disorders, the identification of objectively defined subtle cognitive decline (Obj-SCD) may lead to early identification and better odds of slowing disease progression.

Objective

To characterize the connectivity patterns of subjects with Obj-SCD.

Methods

Fifty-one healthy adults (21 with Obj-SCD and 30 controls) over 55 years old underwent functional MRI and neuropsychological evaluations. MRI scans were conducted using a 3.0 Tesla scanner, and the data were preprocessed and denoised with CONN and SPM software, followed by independent component analysis (ICA) for identifying 20 brain networks and region-of-interest (ROI) analyses for assessing functional connectivity. The thresholds for the results were p < 0.05 for connections and FDR-corrected p < 0.05 for clusters.

Results

Compared with controls, individuals with Obj-SCD exhibited both hyperconnectivity and hypoconnectivity across key brain networks; increased activity was observed in the left angular and right lingual gyri, which showed greater connectivity with the language and visual networks but reduced connectivity with the somatosensory and dorsal attention networks. The default mode and central executive networks also showed functional connectivity alterations, whereas the salience network exhibited hypoconnectivity.

Conclusions

The connectivity alterations of individuals with Obj-SCD may reflect compensatory mechanisms, early network disruptions, or both. fMRI-based analyses could aid in detecting these early changes, providing opportunities for interventions that may slow or prevent further cognitive decline.

Keywords

Alzheimer's disease Alzheimer's disease predictors brain mapping cognitive aging cognitive dysfunction functional connectivity functional neuroimaging objectively defined subtle cognitive decline

Introduction

Population growth and aging are driving a significant demographic shift, with projections indicating that by 2030, the global population of adults over 60 years of age will reach 1400 million.¹ As life expectancy increases, so does the prevalence of age-related conditions, particularly neurocognitive disorders (NCDs), for which age is the primary risk factor.² Although aging is associated with a natural decline in cognitive function within normal parameters, pathological processes can exacerbate this decline, leading to the development of NCDs. Recognizing their profound emotional and economic impact, the World Health Organization (WHO) has identified these disorders as public health priorities.³ These conditions cause disability and dependence, affecting individuals, families, and healthcare systems. Given the lack of curative treatments, early detection and intervention are important for delaying onset or slowing disease progression.⁴

Understanding the progression of NCD requires distinguishing between distinct phases of cognitive decline. This continuum begins with the preclinical (asymptomatic) phase, advances to the prodromal Alzheimer's disease (AD) phase—which includes conditions such as mild cognitive impairment (MCI) due to AD, minor NCD, or very mild AD dementia—and ultimately ends in major NCD or dementia.^5–7

During the preclinical stage, individuals, particularly older adults, may experience subjective cognitive decline, which refers to a perceived worsening of cognitive abilities over time but with performance on standardized cognitive tests still within normal ranges as well as an inability to explain the individual's symptoms by the presence of psychiatric, neurological, or medical conditions. Studies have indicated that in dementia patients, SCD may appear up to a decade prior to an official diagnosis of the disease.

One quantitative approach for identifying SCD is the designation of objectively defined subtle cognitive decline (Obj-SCD), a neuropsychometric condition defined to improve predictive accuracy in detecting early cognitive impairment. This designation serves as a bridge between self-reported cognitive complaints and objectively measured deficits and can be used to identify individuals who exhibit subtle cognitive difficulties that may indicate the earliest stages of neurodegenerative disease. Furthermore, the presence of Obj-SCD has been shown to predict conversion from subjective cognitive decline to NCD.⁸

Individuals are classified as having Obj-SCD if they score more than one standard deviation below the normative average on one relevant test in two separate cognitive domains.^9–11 In contrast, prodromal AD is typically defined by impairments in two tests within the same domain or across multiple domains. By differentiating Obj-SCD from more advanced stages of impairment, this classification aims to identify individuals at higher risk of progressing to prodromal AD and, ultimately, dementia.

Research has demonstrated that individuals with Obj-SCD frequently exhibit neurobiological risk markers associated with preclinical AD, including the APOE ε4 allele, amyloid-β deposition, tau accumulation, entorhinal cortex thinning, and alterations in functional connectivity.^12–15

By analyzing correlations in the blood-oxygen-level-dependent (BOLD) signal across different regions on functional magnetic resonance imaging (fMRI), functional connectivity (FC) can be quantified, serving as a valuable metric for exploring the network organization of the brain. In the context of subtle cognitive decline, a potential early stage of AD and other dementias, FC on fMRI can be used to identify subtle changes in brain networks before significant cognitive deficits become apparent, especially in the three main large-scale neurocognitive networks described by Menon¹⁶ in his triple network model of psychopathology: the default mode network (DMN), the frontoparietal network/central executive network (FPN/CEN) and the salience network (SN).¹⁶ A key focus in subtle cognitive decline research is the DMN, a network that is active during rest and self-referential thinking and that often shows altered connectivity in individuals with subtle cognitive decline, suggesting early neurodegenerative processes.^12,14,17 In addition to those in the DMN, disruptions in other critical networks, such as the FPN/CEN and SN, which are responsible for task switching, attention, and executive control, have also been observed.¹⁶ These changes in FC patterns hold promise as potential biomarkers for predicting neurodegenerative disease progression to more severe cognitive impairment or dementia.^8,18 However, the differences in findings across studies, driven by differences in methodologies, definitions of subtle cognitive decline, and study populations, complicate the generalization of their results and highlight the need for further research to establish reliable and consistent FC-related biomarkers.^19,20 That said, existing evidence suggests that during the preclinical stage, FC within neurocognitive networks increases (possibly in a compensatory manner) and later decreases when cognitive decline becomes more severe.^21,22 Moreover, other evidence shows that in the preclinical stage, FC between neurocognitive networks increases, which translates to a desegregation of cognitive networks (i.e., a dedifferentiated brain), worse cognitive capacity,²³ and an increased risk of tau deposition and a progression toward cognitive decline.²⁴

Given the evidence that Obj-SCD, primarily with executive function impairment (Obj-SCD-EF), is as a well-established predictor of future NCD development but the lack of evidence regarding its relationship with brain FC, we investigated the FC patterns of healthy older adults with Obj-SCD-EF and compared them with those of cognitively unimpaired older adults to address relevant gaps in the literature. We hypothesized that compared with their control counterparts, participants with Obj-SCD-EF have higher within-network (i.e., DMN, CEN, and SN) connectivity and dedifferentiated brain FC (that is, greater between-network connectivity), reflecting adaptations for sustaining cognition from further decline.

Methods

The results of this study are presented in accordance with the Best Practices in Data Analysis and Sharing in Neuroimaging using MRI, as outlined in the guidelines available at Equator Network. The protocol was approved by the Bioethics Committee of the Instituto de Neurobiología, UNAM ENES-Juriquilla (Reference No. 030-H-RM). The study complied with the principles of the Declaration of Helsinki, ensuring patient anonymity, and written informed consent was obtained from all participants.

Participants

Fifty-one participants (21 with Obj-SCD and 30 controls) successfully underwent functional MRI and neuropsychological evaluations and were included in the study. The participants were community-dwelling individuals who were recruited between April 2022 and January 2025. The inclusion criteria included relative healthiness (glucose, cholesterol, triglyceride, hemoglobin, and thyroid stimulant hormone levels within normal limits), age over 55 years, no history of neurological or psychiatric disorders, at least nine years of education, a Global Deterioration Scale score of 1 or 2, and performance above the age-adjusted cutoff on the Mexican translation (version 8.3) of the Montreal Cognitive Assessment (MoCA). Additional requirements included a score above 80 on the Wechsler Adult Intelligence Scale (WAIS-IV), a score below 5 on the Geriatric Depression Scale, and no functional impairment (i.e., a score of 8 on the Lawton and Brody Instrumental Activities of Daily Living Scale). While some participants scored one standard deviation below their normative group on some neuropsychological tests, none met the criterion for mild NCD as defined by the actuarial neuropsychological criteria put forward by Jak and Bondi: impairments in two tests within the same domain or across multiple domains.^25,26 This criterion has been shown to be more accurate in identifying older adults who progress to a major NCD than conventional diagnostic criteria are.²⁷

Participants were recruited through social media, radio advertisements, and referrals, with all providing written informed consent. A post hoc sensitivity analysis indicated that with the included sample sizes (Obj-SCD=21, control=30) and α=0.05, the study had 80% power in detecting large effect sizes (d > 0.80) in group comparisons.

Cognitive assessment

The standardized version of the WAIS-IV for the Mexican population, along with specific memory and executive function tests, was administered. Memory assessments included the Rey Complex Figure, story recall, and paired associations tests. Executive function tests included phonological and semantic fluency, mazes, the Tower of Hanoi, the Trail Making Test, and the Stroop test. These tests were drawn from the Neuropsi Attention and Memory battery and the Neuropsychological Battery of Executive Functions and Frontal Lobes, while adapted versions of the Trail Making Test and Stroop test were administered. All tests were conducted in the morning. Standardized scores were calculated for each measure using the appropriate norms to control for age-related effects.

Objectively defined subjective cognitive decline classification

Traditionally, memory decline is considered the first sign of cognitive impairment in AD, but recent evidence suggests that selective executive impairments occur in the preclinical stage.²⁸ In the pre-MCI stage, memory tests have revealed no significant differences between cognitively normal older adults, whereas differences in executive function have been detected.²⁸ Moreover, research has shown that compared with memory measures, executive function tests can better predict subsequent cognitive decline.²⁹ Importantly, pre-MCI individuals show executive function deficits that are intermediate between those of cognitively healthy individuals and MCI patients,³⁰ suggesting that the presence of multiple executive deficits indicates more severe impairment. Evidence from more than 800 participants consistently shows that multiple executive impairments reflect more advanced cognitive decline than single deficits do.^28,30 To characterize changes in FC in older adults at the early stages of pre-MCI, we classified participants as belonging to the Obj-SCD group according to a variation of the criteria proposed by Thomas et al.¹⁰ and Edmonds et al.,¹¹ wherein an individual meets the criteria for Obj-SCD-EF if they score more than 1.5 standard deviations below the norm in at most one executive function test. We selected a cutoff threshold of 1.5 standard deviations instead of 1 standard deviation to avoid including false positives (i.e., due to fatigue or a lack of effort). Moreover, a recent review by Thomson & Edmonds³¹ revealed that 6 studies defined Obj-SCD according to an abnormal score on only one test and subsequently established correlations between that score and AD biomarkers and progression to MCI.³¹

Imaging

Image acquisition

MRI scans were conducted using a 3.0 Tesla GE750 Discovery scanner (General Electric, Waukesha, WI, USA) equipped with a 32-channel head coil. High-resolution T1-weighted images were obtained using a 3D fast gradient echo sequence (3D FSPGR), providing whole-brain coverage with a voxel resolution of 1 × 1 × 1 mm³. The acquisition included 180 partitions and took 5.02 min, with the following parameters: repetition time (TR) = 8.1 ms, echo time (TE) = 3.1 ms, field of view (FoV) = 25.6 cm², matrix size = 256 × 256, flip angle = 12°, number of excitations (NEX) = 1, and bandwidth (BW) = 31.25 kHz.

For resting-state functional MRI (rs-fMRI), a T2*-weighted echoplanar imaging (EPI) sequence was used with the following settings: TR = 1000 ms, TE = 30 ms, FoV = 25.6 cm², 36 axial slices, flip angle = 50°, matrix size = 64 × 64, voxel size = 4 × 4 × 4 mm³, and 720 volumes. A multiband acceleration factor of 2 was applied, and the total acquisition time was 12 min. During the scan, participants were instructed to keep their eyes closed, avoid focusing on any specific thoughts, and remain still without falling asleep.

Overall fMRI image processing

Analyses of the imaging data were performed using CONN (RRID:SCR_009550) release 22.a and SPM (RRID:SCR_007037) release 12.7771, implemented in MATLAB 2024a. Preprocessing involved realignment, slice timing correction, outlier detection, segmentation, MNI space normalization, and smoothing of functional and anatomical data. Denoising included regression of confounding effects (e.g., white matter, cerebrospinal fluid, and motion parameters) and bandpass filtering (0.01–0.1 Hz). Group-level independent component analysis (ICA) revealed 20 temporally coherent networks from the fMRI data, which were reduced via singular value decomposition and analyzed using fast-ICA. Group-level analyses were performed with a general linear model (GLM) with voxelwise connectivity measures, and the results were evaluated using parametric statistics and cluster-level inferences (voxel p < 0.001 and cluster FDR-corrected p < 0.05). Following ICA component identification and between-group analysis, an ROI-to-ROI analysis that incorporated the identified independent components and a priori regions of interest was performed.

fMRI preprocessing

The data underwent a series of preparation steps to ensure accuracy. First, functional scans were aligned to a reference image to correct for head movement and distortions caused by the MRI scanner. Timing differences between slices of the brain were adjusted to ensure that all the data points matched correctly. Scans exhibiting excessive movement or unusual signal changes were flagged as outliers and excluded. The brain images were then standardized to a common space and divided into different tissue types (gray matter, white matter, and cerebrospinal fluid). Finally, the data were smoothed to reduce noise and improve signal quality.

Next, the data were cleaned to remove unwanted signals that could interfere with the analysis. This included removing signals from the white matter and cerebrospinal fluid and accounting for head motion, session effects, and linear trends. The cleaned data were filtered to focus on frequency ranges relevant to brain activity. This step ensured that the remaining signals were more likely to reflect true brain activity than noise.

Two further analyses were performed: 1) data-driven whole-brain analysis and 2) hypothesis-driven ROI-to-ROI analysis. In the first analysis, ICA was conducted to reduce data dimensionality. In the second, the triple network model of psychopathology described by Menon was explored.¹⁶

Independent component analysis (component determination)

Next, the analysis focused on identifying patterns of brain activity shared across all participants. As mentioned above, ICA identified 20 distinct brain region networks that showed synchronized activity. These networks were derived by combining data from all participants and reducing data complexity through mathematical techniques. The individual contribution of each participant to these networks was then calculated.

ICA between-group analysis

Finally, group-level analyses were conducted to compare brain connectivity patterns across participants. A statistical model was constructed to test for differences in connectivity between groups or conditions. The results were carefully evaluated to ensure that they were not due to random chance, with significant findings identified on the basis of strict thresholds. This approach allowed us to identify meaningful patterns of brain activity and connectivity across the groups.

Region of interest analysis

In the first-level analysis, ROI-to-ROI connectivity (RRC) matrices were computed to assess functional connectivity between the 32 regions from the Harvard–Oxford atlas and the 20 identified independent components. Functional connectivity strength was measured using Fisher-transformed bivariate correlation coefficients from a weighted GLM, which allowed estimations of the associations between BOLD signal time series for each ROI pair. To address transient magnetization effects at the start of each run, individual scans were weighted using a step function convolved with the canonical hemodynamic response function of the SPM. For group-level analyses, a GLM was applied to each connection, with first-level connectivity measures as dependent variables and group- or subject-level identifiers as independent variables. Connection-level hypotheses were tested using multivariate parametric statistics with random effects across subjects, and inferences were made at the cluster level using hierarchical clustering based on anatomical and functional proximity. The results were thresholded at p < 0.05 for individual connections and p-FDR < 0.05 for clusters to ensure statistical significance.

Results

The demographic and neuropsychological characteristics of the study sample are presented in Table 1. Briefly, the results of pairwise analyses indicated no significant differences between the groups in terms of age (risk: x̄ = 63.3 ± 4.8; control: x̄ = 62.6 ± 4.9; t = -0.48; p = 0.63), years of education (risk: x̄ = 21.4 ± 4.6; control: x̄ = 20.35 ± 3.2; t = -1.02; p = 0.31), or sex (risk female: 66.6%; control female: 80%; χ² = 1.15; p = 0.28). Moreover, except for those on the logical memory (encoding: U = 451.5, p = 0.008; retrieval: U = 444.0, p = 0.013), verbal paired association (encoding: t = 3.03, p = 0.004; retrieval: t = 2.54, p = 0.014), working memory (Digit Span: t = 2.66, p = 0.010; WMI: t = 2.65, 0.011) and planning tests (Hanoi 4: t = 9.54, p = 0.001), we did not find significant differences between the groups in the scores of any of the neuropsychological tests, including the memory, executive function and intelligence tests.

Table 1.

Differences in demographics and neuropsychological test results between the risk group and control group.

Demographics/Cognitive test	Risk group mean (SD)	Control group mean (SD)	Statistic value df (49)	p
Age	63.3 ± 4.8	62.6 ± 4.9	t = -0.48	0.633
Education	21.4 ± 4.6	20.35 ± 3.2	t = -1.025	0.310
Sex	66.6%	80%	χ² = 1.156	0.282
RCF
Copy	10.47 ± 1.69	10.86 ± 1.77	U = 364.500	0.328
Retrieval	10.85 ± 3.11	10.63 ± 2.34	t = -0.293	0.771
Logical memory
Encoding	12.42 ± 2.69	14.50 ± 2.06	U = 451.500	0.008*
Retrieval	11.95 ± 3.10	14.36 ± 2.45	U = 444.000	0.013*
PA
Encoding	10.95 ± 2.50	12.73 ± 1.70	t = 3.032	0.004*
Retrieval	11.57 ± 2.46	13.00 ± 1.55	t = 2.542	0.014*
Semantic fluency	12.42 ± 3.28	13.73 ± 3.43	t = 1.360	0.180
Phonological fluency	13.57 ± 3.45	14.13 ± 2.87	t = 0.632	0.530
Maze planning	10.42 ± 2.37	11.23 ± 1.52	t = 0.147	0.147
Hanoi 3	11.33 ± 2.19	11.83 ± 1.51	U = 350.500	0.480
Hanoi 4	4.00 ± 2.86	9.93 ± 1.55	t = 9.545	0.001*
TMT-A	79.66 ± 18.41	80.70 ± 18.91	U = 325.500	0.848
TMT-B	74.00 ± 11.84	74.06 ± 11.18	U = 308.500	0.908
Stroop interference	46.00 ± 19.31	47.53 ± 20.12	t = 0.272	0.787
Block design	10.76 ± 2.36	11.26 ± 2.27	t = 0.768	0.446
Similarities	12.95 ± 2.20	12.83 ± 2.11	U = 296.000	0.720
Digit span	10.57 ± 1.46	11.86 ± 1.85	t = 2.668	0.010*
Matrix reasoning	12.47 ± 1.91	12.50 ± 1.88	U = 336.000	0.690
Vocabulary	13.74 ± 1.52	14.03 ± 1.03	U = 350.000	0.495
Arithmetic	10.81 ± 2.15	12.16 ± 2.57	U = 418.000	0.047
Symbol search	11.38 ± 1.07	11.73 ± 1.66	U = 361.000	0.374
Visual puzzles	11.14 ± 2.22	11.80 ± 2.57	t = 0.947	0.348
Information	13.38 ± 1.43	13.33 ± 1.80	U = 333.500	0.724
Coding	11.52 ± 1.28	12.20 ± 1.40	t = 1.753	0.086
VCI	121.81 ± 10.11	122.23 ± 8.20	U = 312.500	0.969
PRI	108.38 ± 11.35	110.76 ± 10.89	t = 0.756	0.453
WMI	103.42 ± 7.99	110.06 ± 9.31	t = 2.652	0.011*
PSI	107.04 ± 5.49	110.13 ± 7.90	t = 1.544	0.129
FSIQ	112.23 ± 8.78	116.10 ± 7.67	t = 1.667	0.102

RCF: Rey–Osterrieth complex figure; PA: verbal paired associations; TMT: Trail Making Test, A and B; VCI: verbal comprehension index; PRI: perceptual reasoning index; WMI: working memory index; PSI: processing speed index; FSIQ: full-scale IQ; *p < 0.05. If the statistical value is > 0, the mean of the Obj-SCD group is greater than the mean of the CTL group, and vice versa.

Independent component analysis

Twelve independent components were identified through ICA (threshold z = 3.3, selected to maximize spatial specificity and reduce false positives during component identification, in line with standard practices in ICA-based resting-state fMRI analyses), corresponding to major brain networks, including the default mode, salience, frontoparietal, visual, language, and sensorimotor networks. Spatial overlap and network maps are shown in Supplemental Figures 1 and 2.

Figure 1.

Differences in brain activity between the Obj-SCD and control groups. Voxel-to-voxel brain display representation illustrating group differences between individuals with Obj-SCD and controls. Warmer colors represent greater activity in the at-risk group (i.e., Obj-SCD > controls). The left lateral view displays the angular gyrus cluster. The right medial view displays the lingual gyrus cluster.

Figure 2.

Connectome ring and glass brain visualization of default mode network (DMN) ROI-to-ROI connectivity in objectively defined subtle cognitive decline (Obj-SCD).

Differences in brain activity

Compared with the controls, the Obj-SCD participants exhibited significantly greater activity in the left angular gyrus (k = 160 voxels, p-FDR = 0.0337, peak activity coordinates: −36, −86, +34) and right lingual gyrus (k = 151 voxels, p-FDR = 0.0337, peak activity coordinates: +6, −52, +2), as depicted in Figure 1.

Independent analysis of functional connectivity differences

Seed-based analyses revealed that in the Obj-SCD group, both the left angular gyrus and right lingual gyrus exhibited altered connectivity with key functional networks, including increased connectivity with the language network and reduced connectivity with the default mode network. In the Obj-SCD group, the left angular gyrus showed increased connectivity with the language (F = 0.54985946) and cerebellar (F = 0.24836798) networks and decreased connectivity with the somatosensory (F = -0.8159726) and default mode (F = -1.3780019) networks. These effects are illustrated in Supplemental Figure 3a (connectivity bar graph) and Supplemental Figure 3b (voxel-wise seed correlation map).

Similarly, in individuals with Obj-SCD, the right lingual gyrus demonstrated increased connectivity with the language (F = 0.06636219) and visual (F = 2.05504778) networks and decreased connectivity with the dorsal attention (F = -0.7514507) and default mode (F = -1.7634643) networks. These differences in connectivity are illustrated in Supplemental Figure 4a and 4b.

Region of interest analysis

Analysis of 496 potential connections revealed 19 connections with uncorrected statistically significant differences (p < 0.05, uncorrected), primarily within the DMN, SN, and CEN. Compared with the controls, individuals with Obj-SCD showed greater connectivity in 10 connections and lower connectivity in 9 connections. Of these, only connections involving the subcallosal cortex survived FDR correction (MVPA Omnibus test). These differences were most pronounced in the default mode, salience, and central executive networks. The following results for the SN and CEN are presented as exploratory, uncorrected trends to aid in the formation of future hypotheses. Specific region-to-region connections, effect sizes, and network affiliations are summarized in Table 2.

Table 2.

Between-group connectivity effect size differences.

Connection		Effect directionality in reference to obj-SCD	Network	Effect size
1	Cingulate Gyrus posterior division and Frontal Medial Cortex	Hypoconnectivity	DMN	−0.123383673
2	Right Middle Temporal Gyrus posterior division and Cingulate Gyrus posterior division	Hypoconnectivity	DMN	−0.167719146
3	Left Paracingulate Gyrus and Subcallosal Cortex	Hyperconnectivity	DMN	0.132972816
4	Angular Gyrus Right and Right Superior Frontal Gyrus	Hypoconnectivity	DMN	−0.172225609
5	Left Insular Cortex and Right Superior Parietal Lobule	Hypoconnectivity	SN	−0.030835678
6	Right Paracingulate Gyrus and Right Supramarginal Gyrus posterior division	Hypoconnectivity	SN	−0.145169864
7	Right Insular Cortex and Right Superior Parietal Lobule	Hypoconnectivity	SN	−0.064825748
8	Right Hippocampus and Left Parahippocampal Gyrus anterior division	Hyperconnectivity	DMN	0.094388914
9	Right Paracingulate Gyrus and Right Insular Cortex	Hypoconnectivity	SN	−0.133012758
10	Left Frontal Orbital Cortex and Frontal Medial Cortex	Hyperconnectivity	CEN	0.120680895
11	Right Paracingulate Gyrus and Left Parahippocampal Gyrus anterior division	Hyperconnectivity	DMN	0.092765215
12	Cingulate Gyrus anterior division and Right Superior Parietal Lobule	Hypoconnectivity	SN	−0.094741605
13	Left Supramarginal Gyrus posterior division and Subcallosal Cortex	Hyperconnectivity	CEN	0.071496589
14	Left Middle Temporal Gyrus anterior division and Cingulate Gyrus anterior division	Hyperconnectivity	DMN	0.038060191
15	Left Angular Gyrus Left and Subcallosal Cortex	Hyperconnectivity	DMN	0.066384506
16	Left Middle Temporal Gyrus anterior division and Subcallosal Cortex	Hyperconnectivity	DMN	0.072438061
17	Left Middle Temporal Gyrus anterior division and Left Parahippocampal Gyrus posterior division	Hyperconnectivity	DMN	0.075103194
18	Right Parahippocampal Gyrus anterior division and Subcallosal Cortex	Hypoconnectivity	DMN	−0.055508926
19	Subcallosal Cortex and Left Supramarginal Gyrus anterior division	Hyperconnectivity	CEN	0.059347327

Obj-SCD: objectively defined subjective cognitive decline; DMN: default mode network; SN: salience network; CEN: central executive network.

Connectivity alterations in the Obj-SCD group

Individuals with Obj-SCD exhibited predominantly increased connectivity within the DMN, particularly involving the subcallosal cortex, anterior middle temporal gyrus, and anterior parahippocampal gyrus. Considering that subjects in the control group had normal connectivity for their age, hypoconnectivity and hyperconnectivity were defined as weaker and stronger connections, respectively, in the Obj-SCD group than in the control group. Hyperconnectivity is expected in healthy older adults with Obj-SCD because they are likely to develop an NCD. Therefore, in the preclinical phase, these individuals may exhibit increased connectivity to compensate for related deficiencies, according to the model proposed by Gorges et al.³² Compared with the controls, individuals with Obj-SCD exhibited predominantly higher connectivity within the DMN. Specifically, three key regions in the Obj-SCD group showed stronger connections: (1) the subcallosal cortex, (2) the left anterior middle temporal gyrus, and (3) the left anterior parahippocampal gyrus.

Within the DMN, Obj-SCD participants demonstrated increased connectivity between the subcallosal cortex and several regions, including the left angular gyrus (η² = 0.0664), left paracingulate gyrus (η² = 0.1330), and left middle temporal gyrus (η² = 0.0724). In addition, the left anterior parahippocampal gyrus showed greater connectivity with the right hippocampus (η² = 0.0944) and right paracingulate gyrus (η² = 0.0928). These relationships are illustrated in Figure 2.

Network-specific connectivity changes

Hypoconnectivity in the SN and hyperconnectivity in the CEN further suggest an imbalance between early network disruptions and compensatory activation in individuals with Obj-SCD. In the salience network, Obj-SCD participants had reduced connectivity between the right superior parietal lobule and anterior cingulate cortex (η² = -0.0947) and between the right paracingulate gyrus and both the right insular cortex (η² = 0.1330) and right supramarginal gyrus (η² = 0.1452). The right paracingulate cortex also showed reduced connectivity with the left and right insular cortices (η² = -0.0308 and η² = 0.0648, respectively), indicating possible early-stage deficits in salience processing. These findings are depicted in Figure 3.

Figure 3.

Connectome ring and glass brain visualization of salience network (SN) ROI-to-ROI connectivity in objectively defined subtle cognitive decline (Obj-SCD). (A) Connectome ring representation showing significant reductions in ROI-to-ROI functional connectivity within the SN in the Obj-SCD group relative to cognitively unimpaired controls. The blue lines indicate decreased functional connectivity, with line thickness representing the strength of group differences. (B) Glass brain rendering in MNI space displaying the spatial distribution and directionality of disrupted SN connections across three anatomical perspectives: right medial, anterior, and superior views. (C) Circular schematic summarizing group differences in SN connectivity with emphasis on hemispheric laterality and frontal-to-occipital directional patterns. Affected regions included the anterior cingulate cortex (AC), insula (IC), superior parietal lobule (SPL), and posterior supramarginal gyrus (pSTG), with consistent reductions in connectivity involving the right paracingulate gyrus (PaCG). (A & C) Node color represents the directionality of group effects (blue = decreased), while edge thickness scales with effect size. The results are shown at p < 0.05, FDR corrected. Visualizations were generated using the CONN toolbox; (C) Image generated using BioRender (Agreement number: SB28DP1Y2A).

Within the CEN, Obj-SCD participants showed greater connectivity between the subcallosal cortex and the left paracingulate gyrus (η² = 0.1330) and the left middle temporal gyrus (η² = 0.0724). Increased connectivity was also observed between the left frontal orbital cortex and frontal medial cortex (η² = 0.1207). These relationships are visualized in Figure 4.

Figure 4.

Connectome ring and glass brain visualization of central executive network (CEN) ROI-to-ROI connectivity in objectively defined subtle cognitive decline (Obj-SCD). (A) Connectome ring representation showing significant increases in ROI-to-ROI functional connectivity within the CEN in the Obj-SCD group compared with the cognitively unimpaired controls. Red lines denote increased connectivity, with the line thickness reflecting the strength of the group effect. (B) Glass brain renderings in MNI space presenting the spatial configuration of CEN connections across three anatomical perspectives: left medial, anterior, and superior views. (C) Circular schematic summarizing group differences in CEN connectivity with emphasis on hemispheric laterality and frontal-to-occipital directional patterns. Notable increases in connectivity are observed among regions such as the left anterior supramarginal gyrus (aSMG), posterior supramarginal gyrus (pSMG), medial frontal cortex (MedFC), subcallosal cortex (SubCaC), and orbitofrontal cortex (FOrb). (A & C) Node color represents the directionality of group effects (red = increased), while edge thickness scales with effect size. The results are shown at p < 0.05, FDR corrected. Visualizations were created using the CONN toolbox; (C) Image generated using BioRender (Agreement number: LV28DP5QT4).

Corrected connectivity findings

Following FDR correction (MVPA Omnibus test), only connections involving the subcallosal cortex remained significantly altered, highlighting its potential as a central hub in early functional reorganization in individuals with Obj-SCD. Specifically, these individuals showed greater connectivity between the subcallosal cortex and the anterior and posterior left supramarginal gyri, left angular gyrus, left paracingulate gyrus, and right anterior parahippocampal cortex. Reduced connectivity was observed only between the subcallosal cortex and right anterior parahippocampal cortex. These corrected findings emphasize the central role of the subcallosal cortex in early connectivity changes in individuals with Obj-SCD and are summarized in Figure 5.

Figure 5.

Connectome ring and glass brain visualization of the subcallosal cortex (SubCaC) ROI-to-ROI connectivity in objectively defined subtle cognitive decline (Obj-SCD). (A) Connectome ring representation showing significant alterations in ROI-to-ROI functional connectivity between the subcallosal cortex (SubCaC) and multiple cortical regions in the Obj-SCD group compared with the cognitively unimpaired controls. Red lines denote increased connectivity, and blue lines denote decreased connectivity, with the line thickness representing the strength of the group effect. (B) Glass brain renderings in MNI space illustrating the spatial distribution of SubCaC connectivity changes across three anatomical perspectives: left medial, right medial, and superior views. (C) Circular schematic summarizing group differences in SubCaC connectivity with emphasis on hemispheric laterality and frontal-to-occipital directional patterns. Altered connectivity is observed between the SubCaC and regions such as the posterior supramarginal gyrus (pSMG), angular gyrus (AG), anterior middle temporal gyrus (aMTG), anterior parahippocampal cortex (aPaHC) and paracingulate gyrus (PaCG). (A & C) Node color represents the directionality of group effects (red = increased, blue = decreased, purple = bilateral or mixed effects), while edge thickness scales with effect size. The results are shown at p < 0.05, FDR corrected. Visualizations were created using the CONN toolbox; (C) Image generated using BioRender (Agreement number: BQ28DP974O).

Discussion

In this study, two groups of older adults were compared: one with Obj-SCD and the other with typical neuropsychological assessment results. The Obj-SCD group showed deficiencies in three cognitive processes: memory, planning and working memory. As cognitive ability naturally decreases with aging, compensatory changes occur in the brain to maintain cognitive function. The connectivity pattern of the control group is considered to reflect this normal compensation, whereas the altered brain connectivity pattern observed in the Obj-SCD group could result from brain disruptions, insufficient compensatory processes, or a combination of both.

This study examined functional connectivity alterations in individuals with Obj-SCD, focusing on the left angular gyrus and right lingual gyrus, which were selected because their connectivity with other brain areas was significantly greater in the Obj-SCD group. Greater activation in a brain region has been interpreted as either greater recruitment of neural resources or less efficient use of available resources. The right lingual gyrus is involved in visual scene recognition and visual memory³³; therefore, disruptions in this region could lead to topographical disorientation,³⁴ a common complaint among older adults, and visual memory difficulties,³⁵ which were not evident in the Obj-SCD group. In contrast, the left angular gyrus is a node of the DMN involved in various cognitive functions, including number and semantic processing, memory retrieval, attention, spatial cognition, reasoning, social cognition, working memory and planning.^36–38 Notably, in our study, cognitive differences were observed between the Obj-SCD and control groups in terms of executive function, specifically planning, working memory, and episodic memory (encoding and retrieval), particularly in tasks involving logical memory and verbal paired associates.

Furthermore, the increased connectivity of the left angular gyrus with regions from other networks, including the semantic network, DMN, reading network, and episodic memory network, and the increased connectivity of the right lingual gyrus with key nodes of the visual, facial identification, reading, and orthographic networks may reflect compensatory recruitment of additional neural resources to maintain cognitive ability.^39,40 This interpretation aligns with the scaffolding theory of aging and cognition.⁴¹ On the other hand, the reduced connectivity of both the left angular gyrus and right lingual gyrus may suggest early-stage network inefficiencies.

Emerging research highlights the importance of identifying and monitoring subjective cognitive decline as a potential early marker of neurodegenerative diseases, offering opportunities for timely intervention and prevention. Within this context, Obj-SCD has gained attention as a potential early indicator of pathological changes in the brain.

Despite its promise, the literature on fMRI-based investigations of Obj-SCD remains limited, underscoring the need for additional research to establish reliable neural markers of early cognitive decline. Zhang et al.¹³ demonstrated that individuals with Obj-SCD exhibit heightened functional connectivity in key brain regions, including the left precuneus and superior temporal gyrus. These changes were positively correlated with cognitive performance, possibly reflecting early adaptations in brain network integrity. Similarly, Thomas et al.⁴² reported that subtle cognitive difficulties predict future amyloid accumulation and neurodegeneration, reinforcing the link between Obj-SCD and early AD pathology.

Overall, these findings suggest that fMRI-based connectivity metrics may serve as valuable, noninvasive biomarkers for the early detection of subtle cognitive impairments. Such advancements could help identify individuals at risk for neurodegenerative diseases such as AD and support the development of timely interventions to slow disease progression.

To further investigate connectivity changes in individuals with Obj-SCD within the core neurocognitive networks of the triple network model, we conducted a seed-based analysis. Our findings revealed distinct functional connectivity alterations in individuals with Obj-SCD, particularly within the DMN, SN, and CEN.

The DMN, which is critical for self-referential processing, memory consolidation, and cognitive integration, exhibited mostly increased connectivity in the Obj-SCD group.¹⁶ Stronger connectivity was observed between key DMN hubs, primarily in the left hemisphere, including the subcallosal cortex, anterior middle temporal gyrus, and anterior parahippocampal gyrus, which showed increased connections with regions such as the angular gyrus, hippocampus, and paracingulate gyrus. These results are consistent with those of Zhang et al.¹³ Several decreased connections were also observed in the Obj-SCD group within the DMN, particularly in the posterior cingulate cortex, frontal medial cortex, and right superior parietal lobule. These reductions may indicate early inefficiencies in information transfer across the DMN, a pattern previously associated with AD progression.¹¹ This coexistence of hyperconnectivity and hypoconnectivity suggests that while compensatory mechanisms may be active in individuals with Obj-SCD, they may not be sufficient to offset emerging network disruptions.

The SN, which mediates switching between the DMN and CEN by detecting and integrating relevant internal and external stimuli,¹⁶ exhibited significantly decreased connectivity in individuals with Obj-SCD. Specifically, reduced connectivity was detected between the right superior parietal lobule and anterior cingulate cortex and both insular cortices, as well as between the right paracingulate gyrus and the insular cortex and supramarginal gyrus. The SN is involved in attentional control and emotional regulation,⁴³ both of which are critical for cognitive functions that were impaired in the Obj-SCD group. The disrupted connectivity of the SN could impair switching efficiency between networks. Given the association of SN dysfunction with neurodegenerative diseases such as AD and frontotemporal dementia,⁴³ these connectivity reductions may serve as early indicators of cognitive decline in individuals with Obj-SCD.

In contrast, the CEN demonstrated increased connectivity in individuals with Obj-SCD, with stronger connections between the subcallosal cortex and the left paracingulate gyrus, middle temporal gyrus, and frontal medial cortex. The CEN supports higher-order functions such as working memory and cognitive control,¹⁶ and its increased connectivity may reflect compensatory engagement in response to objectively assessed cognitive inefficiencies. This pattern is consistent with prior findings indicating that individuals with subtle cognitive impairments recruit additional neural resources to preserve their cognitive performance.¹²

In summary, our findings reveal that older adults with Obj-SCD exhibit predominantly increased connectivity in the DMN and CEN and decreased connectivity in the SN. This imbalance may reflect early network disruptions in the coordination between these networks and could underlie subtle cognitive impairments. We also found a lateralized pattern of increased left-hemispheric connectivity and decreased right-hemispheric connectivity in individuals with Obj-SCD. This may represent a transitional adaptive mechanism in the progression toward NCD. Asymmetric amyloid deposition, with a left-lateralized pattern in early NCD,⁴⁴ could drive this pattern. Amyloid plaques are known to induce hyperexcitability and increased connectivity within affected brain networks⁴⁵; thus, increased left-hemispheric functional connectivity may indicate early synaptic dysfunction, whereas right-sided reductions may reflect deteriorating network integrity. Furthermore, the hemispheric asymmetry reduction in older adults (HAROLD) model posits that healthy aging involves bilateral recruitment for cognitive function,⁴⁶ whereas persistent lateralization may signal inefficient resource use and worse outcomes. Accordingly, the asymmetric pattern observed in individuals with Obj-SCD may indicate early signs of subtle decline.

The combination of increased and decreased connectivity across major networks in individuals with Obj-SCD supports its characterization as an intermediate state between healthy aging and mild cognitive impairment. Although compensatory mechanisms may temporarily uphold cognitive function, underlying inefficiencies, especially within the DMN, SN, and CEN, could drive progression to more severe stages of decline. These results underscore the potential of fMRI-based connectivity metrics as early neural markers of Obj-SCD, offering valuable tools for refining AD risk assessments.

Although it could be considered a weakness of the study to consider a single test below a cutoff as a criterion for Obj-SCD, it should be noted that in a review, Thomas & Edmonds³¹ reported that a score in a single test below a specified cutoff could establish differences between good performance and the beginning of subtle decline. On the other hand, if a single low score already reveals functional connectivity differences consistent with cognitive decline, this criterion may be advantageous because it enables earlier detection. Therefore, in addition to the above rationale, we selected a cutoff threshold of 1.5 standard deviations instead of 1 standard deviation to avoid including false positives (e.g., due to fatigue or lack of effort), thereby potentially yielding more robust FC differences between groups. Moreover, a 1.5 SD cutoff is stricter than a 1 SD cutoff is and is very close to the threshold corresponding to p = 0.05 in a one-tailed (lower) Normal distribution. Thus, although only one low score was required in this study, a more conservative cutoff (1.5 SD) was applied to reduce false positives.

One limitation of this study is the sample size; while it was sufficient to detect large-effect alterations, future studies with larger cohorts are needed to identify more subtle effects and improve generalizability. Our findings may have implications beyond detection for the development of targeted interventions. The identified hubs of dysconnectivity, particularly the subcallosal cortex—which emerged as a central node of alteration—could serve as candidate targets for neuromodulation therapies such as transcranial magnetic stimulation (TMS). By modulating activity in this key region or normalizing the imbalance among the SN, CEN, and DMN, bolstering network resilience and slowing cognitive decline may be possible. Although fMRI itself may not be the most accessible screening tool, it can play a crucial role in identifying target networks for more scalable interventions such as TMS, which could then be deployed on the basis of simpler, initial plasma biomarker screening. Evidence suggests that reversing cognitive dysfunction in individuals with MCI to normal cognition with TMS is in fact possible,⁴⁷ and FC has even been suggested as a biomarker for predicting cognitive improvement with TMS.⁴⁷ However, these interpretations remain speculative. Without longitudinal data linking these connectivity patterns to future clinical progression, hyperconnectivity could equally reflect early, maladaptive dysregulation, and the observed network imbalance may indicate initial breakdown rather than successful compensation. Future research should explore the longitudinal trajectory of these network alterations to assess their ability to predict disease progression. Integrating connectivity data with other biomarkers, such as amyloid and tau, may yield a more comprehensive understanding of the neural mechanisms underlying Obj-SCD.

Conclusion

These findings support the association of Obj-SCD with functional connectivity alterations that are consistent with a potential transitional phase between normal cognition and MCI. Increased connectivity in certain networks may function as an initial compensatory response to neuronal stress, whereas reductions in connectivity may signal the early breakdown of network integrity. These results support the view that Obj-SCD involves both adaptive and maladaptive neural changes, thereby capturing a transitional phase in the progression of cognitive decline. However, longitudinal validation is essential to confirm whether these neural changes indeed predict clinical progression.

Functional MRI offers a promising approach for detecting these early alterations in brain connectivity. By identifying changes that precede overt cognitive symptoms, neuroimaging may aid in the formation of strategies for early intervention aimed at slowing or preventing further decline. As such, fMRI-based connectivity metrics may prove valuable in identifying individuals at heightened risk for neurodegenerative diseases, facilitating timely and potentially more effective clinical responses.

Supplemental Material

sj-docx-1-alz-10.1177_13872877261435619 - Supplemental material for Functional connectivity alterations in objectively defined subtle cognitive decline: A cross-sectional functional MRI study in cognitively healthy older adults

Supplemental material, sj-docx-1-alz-10.1177_13872877261435619 for Functional connectivity alterations in objectively defined subtle cognitive decline: A cross-sectional functional MRI study in cognitively healthy older adults by Jorge Sigg-Alonso, Jaime D. Mondragón, Mauricio González-López, Erick H. Pasaye Alcaraz and Thalía Fernández in Journal of Alzheimer's Disease

Footnotes

Acknowledgements

The authors express their gratitude to César Arturo Domínguez Frausto, Héctor Belmont, Paloma Arlet Roa Rojas, Luisa Mariana Pérez Figueroa, Teresa Álvarez, Gina Lorena Quirarte, Nuri Aranda López, Daniela Roldán García and Bertha Esquivel for their technical support. This article constitutes a part of Jorge Sigg-Alonso's PhD thesis project.

ORCID iDs

Jorge Sigg-Alonso

Jaime D. Mondragón

Mauricio González-López

Thalía Fernández

Ethical considerations

The protocol was approved by the Bioethics Committee of the Instituto de Neurobiología, Universidad Nacional Autónoma de México (Reference No. 030-H-RM). The study was conducted in accordance with the Declaration of Helsinki.

Consent to participate

Written informed consent was obtained from all participants for both participation and publication, with full assurance of anonymity.

Consent for publication

Not applicable

Author contribution(s)

Jorge Sigg-Alonso: Conceptualization; Methodology; Formal analysis; Investigation; Data curation; Writing, Original Draft, Review & Editing.

Jaime D. Mondragón: Methodology; Software; Formal analysis; Data Curation; Writing, Original Draft, Review & Editing; Visualization.

Mauricio González-López: Investigation; Writing, Review & Editing).

Erick H. Pasaye Alcaraz: Investigation; Writing, Review & Editing.

Thalía Fernández: Conceptualization; Methodology; Investigation; Resources; Writing, Review & Editing; Supervision; Project administration; Funding acquisition.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by grants IT201123 and IT201125 from the Universidad Nacional Autónoma de México (DGAPA-PAPIIT) and CBF-2025-I-560 from la Secretaría de Ciencia, Humanidades, Tecnología e Innovación. During this study, Jorge Sigg-Alonso (CVU: 765024) was a recipient of a fellowship from the Consejo Nacional de Humanidades, Ciencia y Tecnología (CONAHCYT).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data supporting the findings of this study are available upon request from the corresponding authors. The data are not publicly available because of privacy or ethical restrictions.

Supplemental material

Supplemental material for this article is available online.

References

United Nations, Department of Economic and Social Affairs, Population Division. World population ageing 2017: highlights. Report No.: ST/ESA/SER.A/397. New York: United Nations, 2017.

Alzheimer’s Disease International. Policy brief: Risk factors for dementia [Internet]. London: Alzheimer’s Disease International. https://www.alzint.org/u/policy-brief-risk-factors-for-dementia.pdf (2012 , accessed February 22, 2026).

World Health Organization. Dementia: a public health priority [Internet]. Geneva: World Health Organization. https://www.who.int/publications/i/item/dementia-a-public-health-priority (2012 , accessed February 22, 2026).

Rasmussen

Langerman

. Alzheimer's disease: why we need early diagnosis. Degener Neurol Neuromuscul Dis 2019; 9: 123–130.

Reisberg

Ferris

de Leon

, et al. The Global Deterioration Scale for assessment of primary degenerative dementia. Am J Psychiatry 1982; 139: 1136–1139.

Scharre

. Preclinical, prodromal, and dementia stages of Alzheimer’s disease. Pract Neurol 2019; 15: 36–47.

Vermunt

Sikkes

SAM

van den Hout

, et al. Duration of preclinical, prodromal, and dementia stages of Alzheimer's disease in relation to age, sex, and APOE genotype. Alzheimers Dement 2019; 15: 888–898.

Shen

Kuo

Yang

, et al. Subtle cognitive impairment as a marker of Alzheimer's pathologies and clinical progression in cognitively normal individuals. Alzheimers Dement (Amst) 2021; 13: e12198.

Rabin

Smart

Amariglio

. Subjective cognitive decline in preclinical Alzheimer's disease. Annu Rev Clin Psychol 2017; 13: 369–396.

10.

Thomas

Edmonds

Eppig

, et al. Using neuropsychological process scores to identify subtle cognitive decline and predict progression to mild cognitive impairment. J Alzheimers Dis 2018; 64: 195–204.

11.

Edmonds

Delano-Wood

Galasko

, et al. Subtle cognitive decline and biomarker staging in preclinical Alzheimer's disease. J Alzheimers Dis 2015; 47: 231–242.

12.

Hafkemeijer

Altmann-Schneider

Oleksik

, et al. Increased functional connectivity and brain atrophy in elderly with subjective memory complaints. Brain Connect 2013; 3: 353–362.

13.

Zhang

Zeng

Wang

, et al. Alteration of functional connectivity network in population of objectively-defined subtle cognitive decline. Brain Commun 2024; 6: fcae033.

14.

Hafkemeijer

van der Grond

Rombouts

. Imaging the default mode network in aging and dementia. Biochim Biophys Acta 2012; 1822: 431–441.

15.

O'Donoghue

Murphy

Zamboni

, et al. APOE Genotype and cognition in healthy individuals at risk of Alzheimer's disease: a review. Cortex 2018; 104: 103–123.

16.

Menon

. Large-scale brain networks and psychopathology: a unifying triple network model. Trends Cogn Sci 2011; 15: 483–506.

17.

Wang

Laviolette

O'Keefe

, et al. Intrinsic connectivity between the hippocampus and posteromedial cortex predicts memory performance in cognitively intact older individuals. Neuroimage 2010; 51: 910–917.

18.

Thomas

Brier

Bateman

, et al. Functional connectivity in autosomal dominant and late-onset Alzheimer disease. JAMA Neurol 2014; 71: 1111–1122.

19.

Molinuevo

Rabin

Amariglio

, et al. Implementation of subjective cognitive decline criteria in research studies. Alzheimers Dement 2017; 13: 296–311.

20.

van Harten

Mielke

Swenson-Dravis

, et al. Subjective cognitive decline and risk of MCI: the mayo clinic study of aging. Neurology 2018; 91: e300–e312.

21.

Schultz

Chhatwal

Hedden

, et al. Phases of hyperconnectivity and hypoconnectivity in the Default Mode and Salience Networks track with amyloid and tau in clinically normal individuals. J Neurosci 2017; 37: 4323–4331.

22.

Viviano

Damoiseaux

. Functional neuroimaging in subjective cognitive decline: current status and a research path forward. Alzheimers Res Ther 2020; 12: 23.

23.

Fountain-Zaragoza

Liu

Benitez

. Functional network alterations associated with cognition in pre-clinical Alzheimer's disease. Brain Connect 2023; 13: 275–286.

24.

Hojjati

Butler

Luchsinger

, et al. Increased between-network connectivity: a risk factor for tau elevation and disease progression. Neurosci Lett 2024; 840: 137943.

25.

Seo

Kim

Lee

, et al. Altered executive function in pre-mild cognitive impairment. J Alzheimers Dis 2016; 54: 933–940.

26.

Bondi

Jak

Delano-Wood

, et al. Neuropsychological contributions to the early identification of Alzheimer's disease. Neuropsychol Rev 2008; 18: 73–90.

27.

Jak

Bondi

Delano-Wood

, et al. Quantification of five neuropsychological approaches to defining mild cognitive impairment. Am J Geriatr Psychiatry 2009; 17: 368–375.

28.

Bondi

Edmonds

Jak

, et al. Neuropsychological criteria for mild cognitive impairment improves diagnostic precision, biomarker associations, and progression rates. J Alzheimers Dis 2014; 42: 275–289.

29.

Clark

Schiehser

Weissberger

, et al. Specific measures of executive function predict cognitive decline in older adults. J Int Neuropsychol Soc 2012; 18: 118–127.

30.

Duara

Loewenstein

Greig

, et al. Pre-MCI and MCI: neuropsychological, clinical, and imaging features and progression rates. Am J Geriatr Psychiatry 2011; 19: 951–960.

31.

Thomas

Edmonds

. Objective criteria for subtle cognitive decline in aging and preclinical Alzheimer disease: a systematic review. Neurology 2025; 104: e213536.

32.

Gorges

Müller

Lulé

, et al. To rise and to fall: functional connectivity in cognitively normal and cognitively impaired patients with Parkinson's disease. Neurobiol Aging 2015; 36: 1727–1735.

33.

Bogousslavsky

Miklossy

Deruaz

, et al. Lingual and fusiform gyri in visual processing: a clinico-pathologic study of superior altitudinal hemianopia. J Neurol Neurosurg Psychiatry 1987; 50: 607–614.

34.

Mendez

Cherrier

. Agnosia for scenes in topographagnosia. Neuropsychologia 2003; 41: 1387–1395.

35.

Mitchell

Cusack

; Cam-CAN group. Visual short-term memory through the lifespan: preserved benefits of context and metacognition. Psychol Aging 2018; 33: 841–854.

36.

Mitani

Rathnayake

, et al. Brain activity associated with the planning process during the long-time learning of the Tower of Hanoi (ToH) task: a pilot study. Sensors (Basel) 2022; 22: 8283.

37.

Seghier

. The angular gyrus: multiple functions and multiple subdivisions. Neuroscientist 2013; 19: 43–61.

38.

Vatansever

Manktelow

Sahakian

, et al. Angular default mode network connectivity across working memory load. Hum Brain Mapp 2017; 38: 41–52.

39.

Garcia

Cohen

Porges

, et al. Functional connectivity of brain networks during semantic processing in older adults. Front Aging Neurosci 2022; 14: 814882.

40.

Langer

O'Neal

Garcia

, et al. Functional connectivity in distributed cortical networks associated with semantic processing in healthy older adults. Front Aging Neurosci 2025; 17: 1479861.

41.

Park

Reuter-Lorenz

. The adaptive brain: aging and neurocognitive scaffolding. Annu Rev Psychol 2009; 60: 173–196.

42.

Thomas

Bangen

Weigand

, et al. Objective subtle cognitive difficulties predict future amyloid accumulation and neurodegeneration. Neurology 2020; 94: e397–e406.

43.

Seeley

Menon

Schatzberg

, et al. Dissociable intrinsic connectivity networks for salience processing and executive control. J Neurosci 2007; 27: 2349–2356.

44.

Kjeldsen

Parbo

Hansen

, et al. Asymmetric amyloid deposition in preclinical Alzheimer's disease: a PET study. Aging Brain 2022; 2: 100048.

45.

Wales

Leung

. The effects of amyloid and tau on functional network connectivity in older populations. Brain Connect 2021; 11: 599–612.

46.

Cabeza

. Hemispheric asymmetry reduction in older adults: the HAROLD model. Psychol Aging 2002; 17: 85–100.

47.

Guo

Jiang

, et al. Repetitive transcranial magnetic stimulation may promote the reversion of mild cognitive impairment to normal cognition. Front Psychiatry 2025; 16: 1544728.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

5.48 MB

0.00 MB