Limbic Network Derangement Mediates Unawareness of Apathy in Mild Cognitive Impairment due to Alzheimer’s Disease: Clues from [ 18 F]FDG PET Voxel-Wise Analysis

Abstract

Background:

Discrepancy between caregiver and patient assessments of apathy in mild cognitive impairment (MCI) is considered an index of apathy unawareness, independently predicting progression to AD dementia. However, its neural underpinning are uninvestigated.

Objective:

To explore the [¹⁸F]FDG PET-based metabolic correlates of apathy unawareness measured through the discrepancy between caregiver and patient self-report, in patients diagnosed with MCI.

Methods:

We retrospectively studied 28 patients with an intermediate or high likelihood of MCI-AD, progressed to dementia over an average of two years, whose degree of apathy was evaluated by means of the Apathy Evaluation Scale (AES) for both patients (PT-AES) and caregivers (CG-AES). Voxel-based analysis at baseline was used to obtain distinct volumes of interest (VOIs) correlated with PT-AES, CG-AES, or their absolute difference (DISCR-AES). The resulting DISCR-AES VOI count densities were used as covariates in an inter-regional correlation analysis (IRCA) in MCI-AD patients and a group of matched healthy controls (HC).

Results:

DISCR-AES negatively correlated with metabolism in bilateral parahippocampal gyrus, posterior cingulate cortex, and thalamus, PT-AES score with frontal and anterior cingulate areas, while there was no significant correlation between CG-AES and brain metabolism. IRCA revealed that MCI-AD patients exhibited reduced metabolic/functional correlations of the DISCR-AES VOI with the right cingulate gyrus and its anterior projections compared to HC.

Conclusions:

Apathy unawareness entails early disruption of the limbic circuitry rather than the classical frontal-subcortical pathways typically associated with apathy. This reaffirms apathy unawareness as an early and independent measure in MCI-AD, marked by distinct pathophysiological alterations.

Keywords

Alzheimer’s disease apathy awareness dementia mild cognitive impairment

INTRODUCTION

Apathy is a multi-dimensional syndrome mostly defined by reduction in goal-direct behavior. 1 It can affect up to 70% of patients with Alzheimer’s disease (AD). 2 In AD, apathy can precede cognitive disorder by up to 5 years, and it is associated with higher risk of conversion from mild cognitive impairment (MCI-AD) to dementia (ADD), faster cognitive decline and higher caregiver burden.3 –5

Several neuroimaging studies support the dysfunction of frontal-subcortical networks, as the neural underpinning of apathy in psychiatric and neurodegenerative disorders.6,7, 6,7 In particular, apathy in AD patients seem to involve more severely the anterior cingulate and orbitofrontal cortices. 7

The most consistent finding in apathy related to AD appears to be presence of dysfunction in the anterior cingulate cortex and related frontal-striatal circuits, although a reduced connectivity between the left insula and right superior parietal cortex has also been described. 8 Apathetic patients disclosed reduced gray matter volume in the medial/inferior frontal cortex, and the volume loss of the insular cortex has been suggested as a potential correlate of apathy. 4 In a recent meta-analysis, the grey matter volume (GMV) correlates of apathy in AD were further extended to the anterior cingulate, putamen, insula, inferior frontal gyrus, and middle temporal gyrus atrophy; moreover, there is a negative correlation between the Apathy Evaluation Scale (AES) scores 1 and volumes of the right putamen and middle temporalgyrus. 9

On the other hand, studies on the metabolic correlates by 2-deoxy-2-[fluorine-18]fluoro-D-glucose positron emission tomography ([¹⁸F]FDG PET) of apathy in AD are sparse. However, they consistently demonstrated an association of apathy metrics with hypometabolism in the orbitofrontal regions, anterior cingulate and striatal areas,10,11, 10,11 confirming previous findings using perfusion SPECT imaging.12 –14 Finally, either amyloid amyloid-β (Aβ) deposition as indicated by [¹¹C]-PIB retention 15 or neurofibrillary tangle burden in neuropathological samples 11 in the frontal, bilateral insular, and right anterior cingulate cortices, has been shown to correlate with severity of apathy in ADD.

By employing the caregiver-patient discrepancy approach, Jacus et al. 16 identified apathy as the foremost discriminative factor in evaluating disease awareness in AD. Such discrepancy in ratings may serve as a proxy for the patient’s unawareness of its condition, within the broader concept of anosognosia. Subsequently, in a prospective study involving 110 consecutive patients diagnosed with amnestic MCI (aMCI), our research group demonstrated that the discrepancy in apathy assessment, reported by patients and their respective caregivers based on the AES, significantly predicted conversion from MCI to ADD. 3 The clinical relevance was underscored by other researchers who, employing an alternative clinical metric— the Apathy Motivation Index (AMI), 17 have consistently demonstrated the correlation of discrepant apathy ratings with caregiver burden and cognitive deficits. 18

In this study, we sought to delve deeper into understanding of this distinctive aspect of the apathetic disturbance, whose neurobiological roots are still uninvestigated. More precisely, we aimed to explore the [¹⁸F]FDG PET brain metabolic correlates of the apathy unawareness, expressed by the discrepancy index (DISCR-AES, hereafter), among individuals diagnosed with MCI-AD. Additionally, we aimed to examine the disparities in metabolic connectivity between brain regions related with the DISCR-AES compared to those observed in healthy individuals, thereby providing insights into the underlying network of apathy unawareness in MCI-AD.

MATERIAL AND METHODS

Patients and controls

Out of 110 individuals diagnosed with aMCI enrolled in a previous paper, 3 we retrospectively selected 28 patients (21 females; age 76.2±4.9 years; education 8.72±4.0, Mini-Mental State Examination (MMSE) score 25.9±1.5) who had performed [¹⁸F]FDG PET scans within three months from the baseline visit, and converting to ADD during a mean follow-up of 18.8±10.5 months. They exhibited the typical hypometabolic pattern of AD, encompassing at least one brain region among precuneus/posterior cingulate (PC/PPC) and posterior temporoparietal cortex. 19 This fulfilled the criteria for MCI-AD with intermediate likelihood, according to the NIA-AA criteria. 20 Among the selected cohort, six patients had evidence of increased amyloid retention by [¹⁸F]Flutemetamol PET and could therefore be classified as MCI-AD with high likelihood. 20

Exclusively for the analysis of cerebral metabolism, we included a healthy control (HC) group composed of 40 volunteers, matched by age, sex, and education, who had undergone [¹⁸F]FDG-PET in previous research (20 females; age 75.6±5.0 years, range 54.9–85.1; education 10.7±3.6 years, range 5–17; MMSE score 28.9±1.0, range 27–29). Their health condition was verified through clinical history and examination, ensuring an MMSE score above 27, a CDR of 0, and a normal [¹⁸F]FDG-PET scan, as confirmed by the independent visual assessment of two expert nuclear medicine physicians.

All subjects gave their consent to use anonymized data according to the declaration of Helsinki and the study protocol met the approval of the local Ethics Committee.

Cognitive assessment and apathy evaluation

All subjects, underwent at baseline a standard cognitive screening (MMSE) and an extended neuropsychological test battery exploring attention, executive functions, verbal episodic memory, categorical and phonological verbal fluency, visuospatial abilities, and depression. To assess the degree of apathy, the Apathy Evaluation Scale (AES) was administered to both patients (PT-AES) and caregivers (CG-AES). 1 Complete description of protocol is detailed in previous paper. 3 We determined the AES discrepancy (DISCR-AES) by computing the absolute difference between the PT-AES and CG-AES without considering its sign (i.e., all values treated as positive). The approach here described differs from the previous paper, in which the DISCR-AES score was calculated subtracting CG-AES from the PT-AES. 3 This addresses a specific need in this study, namely, to focus on the apathy unawareness of the patient, regardless of whether it is an overestimation or underestimation compared to the caregiver’s perception.

[¹⁸F]FDG PET protocol and images preprocessing

[¹⁸F]FDG PET scans were acquired according to the European Association of Nuclear Medicine guidelines, 21 and images were reconstructed and preprocessed as detailed in Massa et al. (2022). 22

Statistical analysis

Voxel-based analysis (VBA) of [¹⁸F]FDG PET scans

Following preprocessing, the smoothed images underwent a whole-brain voxel-wise group analysis (Statistical Parametric Mapping, SPM12), including:

a two-sample t-test (with covariates including age and MMSE) to compare brain metabolism in MCI-AD patients against a control group of 40 HC, who performed [¹⁸F]FDG PET evaluations in previous research (20 females; age 75.6±5.0; education 10.7±3.6 years, MMSE score 28.9±1.0, CDR = 0). 22 The assessment of their health status involved clinical evaluations, and an expert nuclear medicine physician (SM) confirmed the normalcy of the [¹⁸F]FDG PET scans;

multiple linear regression analysis of the distinct apathy subscores (PT-AES, CG-AES, and DISCR-AES) with brain metabolism in MCI-AD patients with age and MMSE as covariates to explore the topography of brain metabolic correlates.

In all the analyses we maintained the default 0.8 gray matter threshold masking and utilized a grand mean scaling value of 50. We established a height threshold of p < 0.001, uncorrected for multiple comparisons at peak level. Only clusters with a minimum of at least 100 voxels were deemed statistically significant if they reached a p-value p < 0.05, family-wise corrected (FWE) for multiple comparisons at the cluster-level. In cases where these thresholds did not yield significant results, we explored fewer conservative thresholds up to p < 0.01. This flexibility aligns with the exploratory nature of our study, a methodology previously employed in several PET studies23 –25 to balance between type I and type II errors. This consideration becomes especially relevant given the relatively diminished sensitivity of [¹⁸F]FDG PET in repeated measures, particularly when dealing with a limited patient sample size.

We transformed the coordinates of the clusters into the Talairach 3D coordinate system, and the resulting cerebral areas were labeled according to the Brodmann classification. 25

Voxel-wise inter-regional correlation analysis (IRCA)

The regions whose brain metabolism correlated with the DIS-AES score were saved into a single volumetric region of interest, hereafter DISCR-AES VOI. The mean count densities of the DISCR-AES VOI were individually extracted using the MarsBaR (MARSeille Boîte À Région d’Intérêt) toolbox implemented in SPM12 for each patient. After scaling to the individual whole-brain count densities, they were utilized as a covariate in an inter-regional correlation analysis (IRCA) to identify brain regions showing significant voxel-wise correlations with the DISCR-AES VOI itself. This analysis was conducted separately for the MCI-AD and HC groups, following the validated procedure by Lee et al. (2008) 26 which has been previously applied in studies involving patients with MCI-AD. 25

The criteria for the analysis output and cluster coordinate extraction remained consistent with those previously outlined.

Demographical clinical data

As an additional analysis, we divided the patients into two subgroups based on ‘high’ or ‘low’ discrepancy values, determined by the median score of the DISCR-AES (= 7). Subsequently, we compared demographical data as well as neuropsychological test scores at baseline between these two groups using a two-tailed t-test. This allowed for exploration of any differences in cognitive domain alterations that might be driving the discrepancy in ratings between patients and caregivers, i.e. the patient unawareness of apathy, in MCI-AD.

RESULTS

Cognitive assessment and apathy evaluation

As depicted in Table 1, there were no significant statistical differences in the demographic features and neuropsychological test scores at baseline between patients with ‘high’ or ‘low’ DISCR-AES. The ‘high’ DISCR-AES group had a significantly lower PT-AES score (p = 0.01) and a slight trend toward significance for a shorter conversion time to dementia from symptom onset (p = 0.08), compared to the ‘low’ DISCR-AES group.

Table 1

Demographical data, neuropsychological test and apathy assessment test scores

Demographical and cognitive variables	MCI-AD	High DISCR-AES	Low DISCR-AES	p
	(n = 28)	(n = 14)	(n = 14)
Age (y)	76.18±4.9	77.8±5.0	74.6±4.6	0.088
Sex (M:F)	7 : 14	3 : 14	4 : 14
Education	8.4±3.68	8.6±4.0	8.3±3.5	0.842
MMSE	26.0±1.44	25.6±1.4	26.4±1.4	0.195
Progression to dementia (months)	18.8±10.55	15.4±10.2	22.3±10.1	0.082
PT-AES	35.3±7.6	31.7±5.2	38.9±8.0	0.010
CG-AES	41.8±7.9	43.9±6.6	39.6±8.7	0.148
GDS-15	2.6±2.24	2.4±1.7	2.0±2.7	0.624
Rey Auditory Verbal Learning test – Immediate	17.3±6.9	15.7±5.0	19.0±8.2	0.209
Rey Auditory Verbal Learning test – Delayed	1.66±2.13	1.2±1.6	2.1±2.6	0.314
Babcock Story Recall Test	4.9±2.8	5.0±3.3	4.8±2.3	0.818
Trail making test A	91.7±57.5	102.9±68.8	80.5±43.0	0.312
Stroop color	31.6±6.8	30.5±7.3	32.7±6.3	0.395
Stroop color-word	7.8±4.1	7.3±3.2	8.2±5.0	0.564
Phonemic Fluency	24.7±8.3	24.0±7.3	25.4±9.5	0.658
Semantic Fluency	24.1±7.8	23.1±7.2	25.0±8.5	0.538
Freehand copying of drawings	8.2±2.2	8.3±2.0	8.1±2.4	0.802

MCI-AD, Mild cognitive impairment due to Alzheimer’s disease; DISCR-AES, Discrepancy on Apathy Evaluation Scale of the caregivers and patients; GDS, Geriatric Depression Scale; MMSE, Mini-Mental State Examination, PT-AES, patients Apathy Evaluation Scale, CG-AES, caregiver Apathy Evaluation Scale.

Voxel-based analysis (VBA)

Comparison between patients and HC

The clusters of significant relative hypometabolism in MCI-AD compared to HC mostly encompassed the temporo-parietal junction in the left hemisphere.

Brain metabolic correlates of apathy scores in MCI-AD patients

We found significant negative correlations of metabolic values with i) DISCR-AES in bilateral parahippocampal gyri and thalami, right posterior cingulate cortex and putamen, and ii) PT-AES score with frontal (superior and medial) and anterior cingulate areas, prominent in the left hemisphere. To note, for the latter result, the statistical threshold was less stringent both at the voxel- (p < 0.01) and the cluster-level (p = 0.01 uncorrected), which does not make the results significant from a statistical point of view. Conversely, CG-AES score disclosed no statistically significant metabolic correlates.

The statistical thresholds, as well as the extent (number of voxels), coordinates, and Brodmann areas of the clusters resulting from these analyses, are summarized in Table 2 and displayed in Fig. 1.

Table 2

Significant brain regions identified in comparisons of brain metabolism between MCI-AD and HC groups, and relationship with apathy scores

Relative Hypometabolism in MCI-AD with respect to HC group
Cluster extent	Peak level threshold	Cluster peaks coordinates			Brain region	BA
	(Cluster Significance)	x	y	z
1,609	p < 0.001	–54.6	–35.5	–14.42	L Inferior Temporal	20
	(p < 0.05 FWE corr)	–45.06	–6.03	–27.69	L inferior Temporal	20
1,167	p < 0.001	–40.52	–59.26	35.81	L Angular	39
	(p < 0.05 FWE corr)	–36.79	–57.24	34.27	L Angular	39
		–53.39	–69.5	25.62	L Middle Temporal	39
		–53.32	–70.84	20.09	L Middle Temporal	39
		–49.58	–54.09	21.74	L Superior Temporal	39
		–55.1	–68.44	14.88	L Middle Temporal	39
		–49.74	–73.59	28.89	L Middle Temporal	39
		–55.01	–69.61	7.56	L Middle Occipital	19
Metabolic Correlate with DISCR-AES score in MCI-AD patients (DISCR-AES VOI)
1,338	p < 0.005	–12.27	–41.18	3.77	L Parahippocampal	30
	(p < 0.05 FWE corr)	24.79	–30.02	3.65	R Thalamus – Pulvinar
		8.05	–45.19	5.54	R Posterior Cingulate	29
		13.67	–37.42	2.77	R Parahippocampal	27
		–21.5	–29.96	4.68	L Thalamus – Pulvinar	–
		19.39	–36.23	–9.64	R Parahippocampal	35
		–12.17	–20.34	2.14	L Thalamus – Mammillary body	–
		30.47	–18.18	–2.33	R Lentiform Nucleus – Putamen	–
		–21.31	–21.29	–7.11	L Parahippocampal	28
Metabolic Correlate with PT-AES score in MCI-AD patients
581	p < 0.01	–8.9	29.49	50.16	L Superior Frontal	6
	(p = 0.01 uncorr)	–6.91	39.67	42.15	L Superior Frontal	8
		–7.07	–5.4	41.48	L Cingulate	24
		–14.74	–1.86	63.31	L Superior Frontal	6
		–7.21	11.84	57.53	L Superior Frontal	6
		4.26	47.41	39.47	R Medial Frontal	8
		–18.5	–15.06	63.8	L Precentral	6

Peak coordinates and cortical regions in each cluster are ordered downwards from the highest peak Z-score. BA, Brodmann area; HC, healthy controls; L, left; R, right; MCI-AD, Mild cognitive impairment due to Alzheimer’s disease; DISCR-AES VOI, disease-related-volumes of interest VOI; PT-AES, patients Apathy Evaluation Scale.

Fig. 1

The two-dimensional representation (axial cuts) and three-dimensional rendering showing i) comparison of cerebral metabolism of MCI-AD patients and HC (displayed in red), ii) metabolic correlate of DISCR-AES score in green, iii) metabolic correlate of PT-AES score (yellow) in MCI-AD patients, superimposed on the MNI reference atlas (using MRIcroGL software, https://www.nitrc.org/projects/mricrogl). In the 3D rendering, mid-line sagittal and coronal cuts were used to enhance the visualization of the VOIs. HC, healthy controls; MCI-AD, Mild cognitive impairment due to Alzheimer’s disease; DISCR-AES VOI, disease-related-volumes of interest VOI; PT-AES, patients Apathy Evaluation Scale.

Voxel-wise inter-regional correlation analysis (IRCA)

In the MCI-AD group, the IRCA revealed that the DISCR-AES VOI whole brain scaled count densities significantly correlated with metabolism in areas of the limbic network, specifically the left parahippocampal and middle temporal gyrus, as well as thalamus bilaterally (Table 3). In the HC group, the DISCR-AES VOI whole brain scaled count densities correlated bilaterally with metabolism values of the thalamus and cingulate gyrus, especially in the posterior region.

Table 3

Significant brain areas related with the DISCR-AES score resulting from inter-regional correlation analysis (IRCA) in MCI-AD and HC

IRCA MCI-AD patients
Cluster extent	Peak level threshold	Cluster peaks coordinates			Region
	(Cluster Significance)
		x	y	z
2604	p < 0.001	15.57	–27.94	1.89	R Thalamus
	(p < 0.05 FWE corr)	–14.11	–35.59	4.27	L Parahippocampal Gyrus
		–14.07	–29.82	3.02	L Thalamus
	–40.12	–45.29	8.32	L Middle Temporal Gyrus
	37.63	–40.10	10.12	R Superior Temporal Gyrus
IRCA HC
4535	p < 0.001	15.54	–29.98	3.50	R Thalamus
	(p < 0.05 FWE corr)	–15.94	–28.12	4.95	L Thalamus
		7.83	–15.78	31.74	R Cingulate Gyrus
		–8.69	–49.35	10.26	L Posterior Cingulate
		5.99	–19.32	29.57	R Cingulate Gyrus

The statistical thresholds, as well as the extent (number of voxels), coordinates, and Brodmann areas of the clusters resulting from the IRCA analyses, are summarized in Table 3 and displayed in Fig. 2.

Fig. 2

The two-dimensional representation (axial cuts) and three-dimensional rendering show the voxel-wise inter-regional correlation analysis (IRCA) of the DISCR-AES VOIs: In particular, the HC group (in green) showed greater connectivity in the cingulate gyrus projecting to the anterior frontal areas compared with MCI-AD patients compared with MCI-AD patients (in blue), where a prominent autocorrelation within the limbic network, mostly in thalamic and parahippocampal regions, is seen. DISCR-AES VOIs, DIS-AES score-volumes of interest VOI. Others as in Fig. 1.

DISCUSSION

Apathy is a complex behavioral symptom in AD patients, whose relationship with frontal and anterior cingulate functioning has been confirmed by several studies,4,7,8,27,28 , 4,7,8,27,28 in fact, apathetic AD patients showed greater atrophy in inferior frontal with a decreased function of fronto-basal regions in modulating the behavioral initiation and reward mechanisms. 7 A key characteristic of patients with AD, that can complicate the assessment of apathy, is the diminished awareness regarding their clinical condition.29,30, 29,30 This intercepts with the concept of anosognosia and is prominently assessed through discrepancy scores comparing the patient’s self-evaluation of performance with either caregiver’s or objective measures.16,18,31–33 , 16,18,31–33

By exploring the metabolic correlates of apathy unawareness, particularly focusing on the correlation between DISCR-AES and [¹⁸F]FDG PET metrics, we discovered involvement of brain regions beyond the frontal-striatal circuitry. While the correlation between the patient apathy scale (PT-AES) and reduced metabolism in frontal and anterior cingulate regions was somewhat anticipated, the findings regarding the metabolic correlate of DISCR-AES prompt further exploration into alternative networks, associated with apathy awareness in patients diagnosed with MCI-AD. Indeed, DISCR-AES, which measures unawareness of apathy by comparing patients’ ratings to those provided by caregivers, exhibited correlations with regions within the limbic network, namely the thalamus, bilateral parahippocampal gyrus, posterior cingulate, and right putamen. The involvement of the limbic network in apathy unawareness underscores the distinct neural circuits implicated in apathy and its lack of awareness in prodromal AD. This suggests that these two aspects of apathy may arise from different underlying mechanisms within the brain. Specifically, apathy itself affects the frontal network, whereas the unawareness of this condition involves thelimbic network.

The correlation between limbic network areas and unawareness has been previously documented, shedding light on the neural basis of anosognosia in AD, with the cingulate cortex playing a significant role. 34 Additionally, thalamic dysfunction has been widely linked to anosognosia, even in non-neurodegenerative diseases.35,36, 35,36 It is also established that dysfunction of the parahippocampal gyrus occurs early in patients with AD, correlating with psychiatric symptoms like depression, and poorer attribution and monitoring of the spatial and temporal context.37 –39 Furthermore, decreased functional connectivity in the parahippocampal gyrus is associated with misperception of cognitive functions and abilities. 40 Noteworthy, in our study the metabolic correlates of apathy unawareness were independent of the relative hypometabolic regions observed in MCI-AD compared to healthy controls (HC), primarily involving the left temporo-parietal junction.

To deepen our understanding of the neural underpinnings of apathy unawareness in MCI-AD, we employed the Inter-Regional Correlation Analysis (IRCA) of resting [¹⁸F]FDG uptake in both patient and HC groups. This approach is recognized for depicting brain metabolic connectivity during rest, assuming the regions correlated with the DISCR-AES (i.e., the DISCR-AES VOI) serve as the central hub of the apathy unawareness network. Compared to the functional activation observed during specific tasks in techniques like functional MRI and perfusion PET, our approach allowed us to investigate the dysfunction of a particular neuronal network in the early stages of AD, akin another previous study from our group. 25 Our analysis uncovered that the metabolic values within the DISCR-AES VOI in MCI-AD patients showed a reduced connection with the right cingulate gyrus and its anterior projections with respect to HC. Prior research has highlighted the relationship between unawareness in AD and impaired functioning of the cingulate cortex, particularly in the right hemisphere, leading to a decrease in cingulo-frontal connectivity.41,42, 41,42 These areas are crucial components of the Default Mode Network (DMN), which is involved in self-monitoring, social functions, and self-awareness. 2 Previous studies have demonstrated an association between apathy scores and regions within the DMN, such as hypothalamus, precuneus, and posterior cingulate cortex (PCC), indicating altered connectivity in AD-associated apathy. 9 Early derangement of this within-network connectivity may contribute to a lack of initiation and self-activating behaviors, along with a failure in self-perceived apathy by patients. This occurs whether there is significant hypometabolism of the circuit nodes, aligning with the concept of AD as an early network disconnection pathology.43,44, 43,44

We observed that apathy unawareness does not correlate with the worsening of cognitive impairment, and specifically, it does not appear to be a proxy for patient memory disturbance. In fact, our additional analysis found no differences in neuropsychological test scores between patients with ‘high’ or ‘low’ DISCR-AES. Albeit the small sample size of the subgroups needs to be taken into account, this result further suggests that apathy unawareness is an early measure in MCI-AD, independent from cognitive decline. In line with recent studies,45,46, 45,46 it is plausible that a distinct neural circuitry is involved compared to other cognitive domains (e.g., memory) and apathy itself.

All patients included in this study received a diagnosis based on biomarkers, indicating at least an intermediate likelihood of AD according to current diagnostic NIA-AA criteria, 20 and had a final confirmation of ADD at follow-up. This meticulous patient selection helps alleviate concerns about the relatively small sample size. However, it is worth noting that the limited number of participants can also be attributed to the non-routine utilization of the AES and specifically of the DISCR-AES in clinical practice. Nevertheless, the availability of the DISCR-AES score and brain metabolism data enhances the value of our cohort for exploring the neural mechanisms underlying apathy unawareness. This phenomenon, which previous research has indicated as a highly valuable predictive indicator in patients with prodromal AD, 3 should increasingly find application in clinical settings.

As a main limitation of the study, we acknowledge the permissive statistical threshold in the correlations of apathy scores, especially PT-AES which can be considered formally not significant, with brain metabolism. Nevertheless, we decided to present this result for emphasizing the involvement of the limbic network in apathy unawareness compared to the frontal network typically associated with apathy itself. For this reason, we must regard our results not as conclusively significant, but rather as explorative in nature. They should be interpreted cautiously and replicated in larger samples using more conservative statistical thresholds to confirm their validity.

Exploring neural network differences between patients who underestimate or overestimate their apathy compared to the caregiver’s perception could be intriguing. However, the limited number of the latter group in our study precluded a structured analysis. Additionally, our study lacks a characterization of caregivers in terms of age, sex, education, caregiver burden, and the quality of the relationship between patient and caregiver. Such factors may influence the accurate estimation of the discrepancy score between patients and caregiver reports. Therefore, future research should consider incorporating these caregiver characteristics to provide a more comprehensive understanding of the dynamics involved. Moreover, in the future, it could be interesting to incorporate sex as a main variable in assessing differences in metabolic correlates.

Finally, expanding the evaluation of metabolic correlates to other determinants of unawareness 16 and other etiologies could enrich our understanding of MCI patients at risk of dementia progression, alongside our study focused on apathy awareness in MCI with at least intermediate likelihood of AD.

Conclusions

The current exploratory results suggest a functional role of the limbic circuitry, particularly the thalamic and parahippocampal regions, in apathy unawareness in the early stages of AD. This underscores a significant difference from the circuitry typically associated with apathy, which involves the frontal-striatal network, described in pathological and neuroimaging studies in both healthy individuals and neurodegenerative diseases. They also support the hypothesis that patients with AD exhibit disruptions in functional connectivity already in the MCI stage, with reduced interconnectivity of the thalamic-hippocampal network with regions involved in the DMN, particularly the cingulate cortex and anterior projections. In this context, apathy unawareness may reveal a more severe connectivity failure that transcends cognitive impairment itself, representing an independent metric in the symptomatic profile of prodromal AD patients at risk of clinical progression.

AUTHOR CONTRIBUTIONS

Wendy Kreshpa (Conceptualization; Data curation; Formal analysis; Writing – original draft; Writing – review & editing); Stefano Raffa (Data curation; Supervision); Nicola Girtler (Data curation; Investigation; Supervision); Andrea Brugnolo (Data curation; Investigation; Supervision); Pietro Mattioli (Methodology; Supervision); Beatrice Orso (Investigation; Supervision); Francesco Calizzano (Methodology; Supervision); Dario Arnaldi (Methodology; Supervision; Validation); Enrico Peira (Data curation; Formal analysis); Andrea Chincarini (Data curation; Formal analysis; Supervision); Luca Tagliafico (Supervision); Fiammetta Monacelli (Methodology; Supervision); Pietro Calcagno (Methodology; Supervision); Gianluca Serafini (Methodology; Supervision); Fabio Gotta (Methodology; Supervision); Paola Mandich (Data curation; Supervision); Stefano Pretta (Supervision); Massimo Del Sette (Conceptualization; Methodology; Supervision; Validation); Luca Sofia (Data curation); Gianmario Sambuceti (Supervision); Silvia Daniela Morbelli (Conceptualization; Data curation; Formal analysis; Supervision; Validation); Angelo Schenone (Conceptualization; Supervision; Validation); Federico Massa (Conceptualization; Data curation; Formal analysis; Methodology; Supervision; Validation; Writing – original draft; Writing – review & editing); Matteo Pardini (Conceptualization; Data curation; Funding acquisition; Methodology; Supervision; Validation).

Footnotes

ACKNOWLEDGMENTS

This work was developed within the framework of the DINOGMI Department of Excellence of MIUR 2018–2022 (legge 232 del 2016).

FUNDING

This work was partially supported by a grant from the Italian Ministry of Health to IRCCS Ospedale Policlinico San Martino [Fondi per la Ricerca Corrente, and Italian Neuroscience network (RIN)], supported by #NEXTGENERATIONEU (NGEU) and funded by the Ministry of University and Research (MUR), National Recovery and Resilience Plan (NRRP), project MNESYS (PE0000006) – A Multiscale integrated approach to the study of the nervous system in health and disease (DN. 1553 11.10.2022).

CONFLICT OF INTEREST

Federico Massa has received speaker and consultant honoraria from Roche Diagnostic S.p.A and Eli Lilly S.p.A, Dario Arnaldi received fees from Fidia, Bruno, Italfarmaco, Idorsia for lectures and board participation; Silvia Morbelli has received speaker Honoraria from G.E. Healthcare, Life Molecular imaging and Eli Lilly; Matteo Pardini receives research support from Novartis and Nutricia, received fees from Novartis, Merck. Massimo Del Sette received speaker and consultant honoraria form PIAM, Lundbeck, Revalma.

Federico Massa and Fiammetta Monacelli are Editorial Board members of this Journal as Associate editors but are not involved in the peer-review process of this article nor have access to any information regarding its peer-review.

The other authors have nothing to disclose.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author.

References

Marin

. Apathy: a neuropsychiatric syndrome. J Neuropsychiatry Clin Neurosci 1991; 3: 243–254.

Dolphin

, Dyer

, McHale

, et al. An update on apathy in Alzheimer’s disease. Geriatrics Basel 2023; 8: 75.

Grisanti

, Massa

, Chincarini

, et al. Discrepancy between patient and caregiver estimate of apathy predicts dementia in patients with amnestic mild cognitive impairment. J Alzheimers Dis 2023; 93: 75–86.

Lanctôt

, Agüera-Ortiz

, Brodaty

, et al. Apathy associated with neurocognitive disorders: Recent progress and future directions. Alzheimers Dement 2017; 13: 84–100.

Lyketsos

, Carrillo

, Ryan

, et al. Neuropsychiatric symptoms in Alzheimer’s disease. Alzheimers Dement 2011; 7: 532.

Kos

, van Tol

, Marsman

JBC

, et al. Neural correlates of apathy in patients with neurodegenerative disorders, acquired brain injury, and psychiatric disorders. Neurosci Biobehav Rev 2016; 69: 381–401.

Cotta Ramusino

, Imbimbo

, Capelli

, et al. Role of fronto-limbic circuit in neuropsychiatric symptoms of dementia: clinical evidence from an exploratory study. Front Psychiatry 2024; 15: 1231361.

Jones

, De Marco

, Manca

, et al. Altered frontal and insular functional connectivity as pivotal mechanisms for apathy in Alzheimer’s disease. Cortex 2019; 119: 100–110.

Chaudhary

, Zhornitsky

, Chao

, et al. Hypothalamic functional connectivity and apathy in people with Alzheimer’s disease and cognitively normal healthy controls. J Alzheimers Dis 2022; 90: 1615.

10.

Holthoff

, Beuthien-Baumann

, Kalbe

, et al. Regional cerebral metabolism in early Alzheimer’s disease with clinically significant apathy or depression. Biol Psychiatry 2005; 57: 412–421.

11.

Marshall

, Fairbanks

, Tekin

, et al. Neuropathologic correlates of apathy in Alzheimer’s disease. Dement Geriatr Cogn Disord 2006; 21: 144–147.

12.

Benoit

, Koulibaly

, Migneco

, et al. Brain perfusion in Alzheimer’s disease with and without apathy: a SPECT study with statistical parametric mapping analysis. Psychiatry Res Neuroimaging 2002; 114: 103–111.

13.

Lanctôt

, Moosa

, Herrmann

, et al. A SPECT study of apathy in Alzheimer’s disease. Dement Geriatr Cogn Disord 2007; 24: 65–72.

14.

Robert

, Darcourt

, Koulibaly

, et al. Lack of initiative and interest in Alzheimer’s disease: a single photon emission computed tomography study. Eur J Neurol 2006; 13: 729–735.

15.

Mori

, Shimada

, Shinotoh

, et al. Apathy correlates with prefrontal amyloid β deposition in Alzheimer’s disease. J Neurol Neurosurg Psychiatry 2014; 85: 449–455.

16.

Jacus

J-P

, Voltzenlogel

, Antoine

, et al. apathy is the best dimension to consider for awareness assessment in Alzheimer’s disease. J Alzheimers Dis 2022; 87: 149–154.

17.

Ang

Y-S

, Lockwood

, Apps

MAJ

, et al. Distinct subtypes of apathy revealed by the Apathy Motivation Index. PLoS One 2017; 12: e0169938.

18.

Klar

, Ang

, Lockwood

, et al. Assessment of apathy in neurological patients using the Apathy Motivation Index caregiver version. J Neuropsychol 2022; 16: 236.

19.

Morbelli

, Brugnolo

, Bossert

, et al. Visual versus semi-quantitative analysis of 18F-FDG-PET in amnestic MCI: an European Alzheimer’s Disease Consortium (EADC) Project. J Alzheimers Dis 2015; 44: 815–826.

20.

Albert

, DeKosky

, Dickson

, et al. The diagnosis of mild cognitive impairment 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 2011; 7: 270.

21.

Guedj

, Varrone

, Boellaard

, et al. EANM procedure guidelines for brain PET imaging using [18F]FDG, version 3. Eur J Nucl Med Mol Imaging 2022; 49: 632.

22.

Massa

, Halbgebauer

, Barba

, et al. Exploring the brain metabolic correlates of process-specific CSF biomarkers in patients with MCI due to Alzheimer’s disease: preliminary data. Neurobiol Aging 2022; 117: 212–221.

23.

Desgranges

, Baron

, Lalevée

, et al. The neural substrates of episodic memory impairment in Alzheimer’s disease as revealed by FDG– PET: relationship to degree of deterioration. Brain 2002; 125: 1116–1124.

24.

Okazawa

, Ikawa

, Jung

, et al. Multimodal analysis using [11C]PiB-PET/MRI for functional evaluation of patients with Alzheimer’s disease. EJNMMI Res 2020; 10: 30.

25.

Massa

, Grisanti

, Brugnolo

, et al. The role of anterior prefrontal cortex in prospective memory: an exploratory FDG-PET study in early Alzheimer’s disease. Neurobiol Aging 2020; 96: 117–127.

26.

Lee

, Kang

, Kim

, et al. Metabolic connectivity by interregional correlation analysis using statistical parametric mapping (SPM) and FDG brain PET; methodological development and patterns of metabolic connectivity in adults. Eur J Nucl Med Mol Imaging 2008; 35: 1681–1691.

27.

Eggins

, Wong

, Wei

, et al. A shared cognitive and neural basis underpinning cognitive apathy and planning in behavioural-variant frontotemporal dementia and Alzheimer’s disease. Cortex 2022; 154: 241–253.

28.

Starkstein

and Brockman

. The neuroimaging basis of apathy: Empirical findings and conceptual challenges. Neuropsychologia 2018; 118: 48–53.

29.

Prigatano

. Anosognosia: Clinical and ethical considerations. Curr Opin Neurol 2009; 22: 606–611.

30.

Salmon

, Meyer

, Genon

, et al. Neural correlates of impaired cognitive processes underlying self-unawareness in Alzheimer’s disease. Cortex 2024; 171: 1–12.

31.

Kalbe

, Salmon

, Perani

, et al. Anosognosia in very mild Alzheimer’s disease but not in mild cognitive impairment. Dement Geriatr Cogn Disord 2005; 19: 349–356.

32.

Salmon

, Perani

, Herholz

, et al. Neural correlates of anosognosia for cognitive impairment in Alzheimer’s disease. Hum Brain Mapp 2006; 27: 588.

33.

Vogel

, Hasselbalch

, Gade

, et al. Cognitive and functional neuroimaging correlate for anosognosia in mild cognitive impairment and Alzheimer’s disease. Int J Geriatr Psychiatry 2005; 20: 238–246.

34.

Andrade

, Guieysse

, Medani

, et al. The dual-path hypothesis for the emergence of anosognosia in Alzheimer’s disease. Front Neurol 2023; 14: 1239057.

35.

Bertagnoli

, Pacella

, Rossato

, et al. Disconnections in personal neglect. Brain Struct Funct 2022; 227: 3161.

36.

Healton

, Navarro

, Bressman

, et al. Subcortical neglect. Neurology 1982; 32: 776–778.

37.

Uchida

, Matsuda

, Nakanishi

, et al. Parahippocampal atrophy is associated with depressive symptoms in Alzheimer’s disease. Osaka City Med J 2016; 62: 59–67.

38.

McLachlan

, Bousfield

, Howard

, et al. Reduced parahippocampal volume and psychosis symptoms in Alzheimer’s disease. Int J Geriatr Psychiatry 2018; 33: 389–395.

39.

Trimble

, Mendez

and Cummings

. Neuropsychiatric symptoms from the temporolimbic lobes. J Neuropsychiatry Clin Neurosci 2006; 9: 429–438.

40.

Sharma

, Murari

, Vandermorris

, et al. Functional connectivity between the posterior default mode network and parahippocampal gyrus is disrupted in older adults with subjective cognitive decline and correlates with subjective memory ability. J Alzheimers Dis 2021; 82: 435–445.

41.

Amanzio

, Torta

, Sacco

, et al. Unawareness of deficits in Alzheimer’s disease: role of the cingulate cortex. Brain 2023; 134: 1061–1076.

42.

Hanyu

, Sato

, Akai

, et al. Neuroanatomical correlates of unawareness of memory deficits in early Alzheimer’s disease. Dement Geriatr Cogn Disord 2008; 25: 347–353.

43.

Brier

, Thomas

and Ances

. Network dysfunction in Alzheimer’s disease: refining the disconnection hypothesis. Brain Connect 2014; 4: 299.

44.

Delbeuck

, Van der Linden

, Collette

. Alzheimer’s disease as a disconnection syndrome? Neuropsychol Rev 2003; 13: 79–92.

45.

Mehak

, Shivakumar

, Saraf

, et al. Apathy in Alzheimer’s disease: A neurocircuitry based perspective. Ageing Res Rev 2023; 87: 101891.

46.

Grisanti

, Massa

, Chincarini

, et al. Discrepancy between patient and caregiver estimate of apathy predicts dementia in patients with amnestic mild cognitive impairment. J Alzheimers Dis 2023; 93: 75–86.

Limbic Network Derangement Mediates Unawareness of Apathy in Mild Cognitive Impairment due to Alzheimer’s Disease: Clues from [ 18 F]FDG PET Voxel-Wise Analysis

Abstract

Background:

Objective:

Methods:

Results:

Conclusions:

Keywords

INTRODUCTION

MATERIAL AND METHODS

Patients and controls

Cognitive assessment and apathy evaluation

[18F]FDG PET protocol and images preprocessing

Statistical analysis

Voxel-based analysis (VBA) of [18F]FDG PET scans

Voxel-wise inter-regional correlation analysis (IRCA)

Demographical clinical data

RESULTS

Cognitive assessment and apathy evaluation

Voxel-based analysis (VBA)

Comparison between patients and HC

Brain metabolic correlates of apathy scores in MCI-AD patients

Voxel-wise inter-regional correlation analysis (IRCA)

DISCUSSION

Conclusions

AUTHOR CONTRIBUTIONS

Footnotes

ACKNOWLEDGMENTS

FUNDING

CONFLICT OF INTEREST

DATA AVAILABILITY

References

[¹⁸F]FDG PET protocol and images preprocessing

Voxel-based analysis (VBA) of [¹⁸F]FDG PET scans