Promising Role of Neuromodulation in Predicting the Progression of Mild Cognitive Impairment to Dementia

Abstract

The differential diagnosis of mild cognitive impairment (MCI) and Alzheimer’s disease (AD) is not always straightforward, and the rate of progression of MCI to dementia is not negligible. Thus, there is a need for para-clinical approaches that can improve the differential diagnosis and identify patients that are at risk of progression. There is a growing interest, at present, in the role of the deterioration of brain oscillations as a predictor of MCI-to-AD conversion. For this reason, we experimentally modulated γ-band oscillations (GBO) in a sample of MCI and AD patients and an age-matched healthy elderly group, using a transcranial alternating current stimulation (tACS) protocol that was applied to different cortical sites. We correlated the after-effects of tACS on the GBO and the neuropsychological data, in an attempt to differentiate MCI from AD patients and identify, among the MCI patients, those that could be at potential risk of MCI-to-dementia conversion. MCI patients showed a partial GBO increase and improvement in some neuropsychological tests whereas AD individuals did not show significant tACS after-effects. Notably, some MCI subjects lacked significant neuropsychological and electrophysiological after-effects, similar to AD individuals. In a two-year follow-up, such MCI individuals had converted into AD. Therefore, our data suggest that tACS may support the clinical differential diagnosis of MCI and AD and identify MCI patients who could be at risk of developing dementia. This prediction index may help the clinician to adopt a better prevention/follow-up strategy in such a disabling neurodegenerative disease.

Keywords

Alzheimer’s disease gamma-band oscillations mild cognitive impairment transcranial alternating current stimulation

INTRODUCTION

Alzheimer’s disease (AD) is one of the leading causes of gradual loss of cognitive function [1]. For the reason of its devastating effects on the quality of life of patients and family members, AD requires an accurate early diagnosis so that cognitive neurorehabilitation and the most appropriate patient management can be activated [2]. In particular, it is of foremost importance to detect any early clinical sign of cognitive decline (including memory, language, thinking, and judgment impairment), configuring a clinical picture of mild cognitive impairment (MCI). The latter describes a patient who has an abnormal cognitive decline compared to a healthy person of the same age. However, the symptoms are not severe enough to interfere significantly with daily life activities in MCI patients, thus not configuring a clinical picture of dementia [3]. It is noteworthy to mention that the annual MCI-to-AD conversion rate is about 10–15% [4].

Neuropsychological tests represent the most straightforward way to diagnose MCI and AD [5]. Nevertheless, both the differential diagnosis and identification of MCI patients who may convert to dementia are often complex, especially when patients reach borderline scores in neuropsychological tests [6–7]. Several para-clinical approaches, structural and functional neuroimaging, neurophysiological, and laboratory tests [8–9] are employed to differentiate MCI from early-AD individuals and to identify those among MCI patients who could developdementia.

To this end, there is great interest concerning the role of connectivity and brain oscillatory activity deterioration in cognitive impairment development [10 –13]. Indeed, a correlation between the amount of cognitive impairment and alterations of brain oscillations has been proposed [14 –17], both depending on amyloid-β (Aβ) aggregate formation [18–19]. Nonetheless, some individuals may show normal connectivity and oscillatory patterns during resting conditions, but a deteriorated pattern under specific conditions (e.g., when performing a test) may manifest.

Taking into account that the performance of cognitive tasks may be biased by many factors, including patient’s cooperation or capacity in understanding the task, it could be helpful to simulate a connectivity and oscillatory pattern modulation using passive approaches. To this end, transcranial alternating current stimulation (tACS) has proved to be an efficient and safe tool in specifically triggering the brain oscillations associated with cognitive functions [20 –25]. A suitable way to evaluate the network alterations in AD and MCI patients may be represented by the assessment the γ-band oscillations (GBO), which have been associated with many cognitive processes [26] due to their role in temporal binding processes, cortico-cortical transmission, and information integration across neural networks [27 –31].

It has been demonstrated that GBO is abnormal in AD patients in comparison to age-matched healthy elderly and MCI individuals. Nonetheless, it has also been reported that GBO during the resting condition unpredictably correlates to cognitive decline [32 –37]. Therefore, the possibility of entraining cortical rhythms using tACS may allow the clinician to identify subtle changes in GBO, which could be undetectable during the resting condition, or possibly misinterpreted during the administration of cognitive tasks.

AD and MCI patients suffer from a different degree of connectivity impairment within large-scale cortical-thalamocortical networks, which is paralleled by a different degree of cognitive impairment [38 –40]. Thus, we may hypothesize that AD patients would not show a modulable GBO, whereas MCI individuals would. In addition, we may argue that the MCI subjects who do not show GBO modulation could be at high risk of developing dementia. Thus, the aim of our study was to find a possible correlation between tACS-induced GBO modulations and the neuropsychological profile in a sample of clinically defined MCI and AD, in an attempt to identify those among MCI patients who could develop dementia within a period of two years, thus the need for a careful follow-up and management.

MATERIALS AND METHODS

Subjects

We enrolled 35 AD and 25 MCI right-handed individuals between January and February 2014. The clinical diagnosis was achieved according to current international diagnostic guidelines [41 –46]. The exclusion criteria were: i) evidence of other neurological or psychiatric diseases leading to cognitive impairment; ii) uncontrolled or complicated systemic diseases or traumatic brain injuries; iii) epileptic history or electroencephalographic (EEG) epileptiform activity; iv) presence of electro/mechanic devices within head and neck (including pacemaker, aneurysms clips, neurostimulator, brain/subdural electrodes); and v) use of any psychoactive medication other than antidepressants and donepezil. Detailed demographic and clinical characteristics are summarized in Table 1. For the control group, we enrolled 27 age-matched, right-handed, healthy elderly individuals (NOLD) without a history of neurologic or psychiatric disease. The Local Ethics Committee approved the present study, and both the NOLD and the patients gave their written informed consent.

Experimental procedure

Each participant underwent a neuropsychological test battery and an EEG recording at baseline (T₀).Then, a tACS conditioning protocol was applied to the primary motor (M1), premotor area (PMA), supp-lementary motor area (SMA), dorsolateral (DLPFC), and dorsomedial prefrontal cortex (DMPFC) within the left hemisphere. A sham-tACS was also performed over M1. Each participant was randomly administered all the conditioning protocols (thus receiving at least one protocol every seven days). An EEG recording and the same neuropsychological battery were carried out immediately after the end of the conditioning protocol (T₁). Moreover, the content of some tests within the neuropsychological battery varied, in order to prevent learning effects between different types of stimulations (e.g., the names that had to be promptly repeated during the Mini-Mental State Examination (MMSE), the content of Attentive Matrices and Phonemic Letter and Verbal Fluency, and so on). After 60 minutes (T_1 +60), we re-recorded the EEG to evaluate tACS after-effect duration and rule out a casual effect of tACS on brain oscillations (i.e., if we found the tACS after-effects only at T₁). The neuropsychological assessment was performed only at T₁ to avoid excessive stress on the subjects. Both the patients and the experimenters who analyzed the data were blinded on the experimental procedure carried out by each patient. Nonetheless, patients were not asked during each session whether they thought they were receiving real or sham stimulation. At a two-year follow-up, all the subjects were re-evaluated from a clinical (neuropsychologicalbattery) and electrophysiological (EEG) point of view (T₂).

Neuropsychological and functional status assessment

All the participants underwent a semi-structured interview with the patient and his or her relative (usually, the patient’s spouse or an offspring of the patient), and physical and neurological examinations. Thereafter, each individual was administered a neuropsychological battery assessing the global cognitive status, frontal functions, verbal and non-verbal memory, language, attention, and executive functions. The battery included the mmse, Reversal Motor Learning (rml), verbal Digit Span forward (ds), Clock Drawing Test (cdt), Attentive Matrices (am), Phonemic Letter and Verbal Fluency (CVF and LVF) [47]. In addition, we assessed the Activities Of Daily Living (adl) and Instrumental Activities Of Daily Living (iadl). Even though a more extensive battery would have been helpful in detecting early and more subtle deficits, it should be considered that a larger battery would have been more time-consuming, and could have biased the global level of attention in subjects with cognitive impairment.

tACS

The entire stimulation procedure complied with the well-established safety guidelines regarding transcranial direct and alternating current stimulation [48]. tACS was delivered by a battery-driven BrainStim stimulator (E.M.S., Bologna, Italy), through a pair of saline-soaked rectangular sponge electrodes. Electrodes were held-in-place by rubber bands and positioned according to formerly used electrode montages [49]. The active electrode (25 cm²) was placed on the M1 (C3), DLPFC (AF3-AF7), DMPFC (AF3-F1), PMA (FC3), or SMA (FCz) of the left hemisphere, according to the International 10% -System for EEG, whereas the reference one (35 cm²) was fixed on the right mastoid. The reference electrode size was larger than the active one to reduce current density and limit stimulation effects under active surface [50]. The stimulator monitored the electrode impedance at ≤ 5 kΩ. Stimulation was applied for 10 minutes at γ-band frequency (i.e., continuously and randomly ranging from 40 to 120 Hz, with the same number of discrete steps delivered at 20 Hz), with zero-degree phase-lag [51] and at a peak-to-peak current intensity of 1 mA (40 μA/cm²) [22, 34]. This approach was chosen because it covers the whole spectrum of GBO that can be recruited by tACS [52]. Indeed, alternating currents can entrain spiking activity by using frequencies close to the endogenous one (or double that frequency) and at very low intensities, whereas higher intensities induce pacing at half of the stimulation frequency [53]. The stimulator was switched off after 30 s of stimulation when performing the sham-tACS, in which the tACS-electrodes were put on C3 (active) and C4 (reference). However, the induced current flows may substantially vary between individuals. In fact, the induced electric fields are influenced by the geometry and depth of sulci and gyri, the volume of the cerebrospinal fluid, the thickness of the skull, the distance of cortical layers from the electrodes generating the electric field, and the possible drift of the electrodes during the stimulation [54 –56]. Therefore, we employed a GUI Computational Model of Transcranial Electrical Stimulation (BONSAI and SPHERES) to estimate the electric fields induced by the tACS electrode montages employed, according to individual MRI findings [57]. The latter was also used to check the tACS electrode positioning, in analogy to a previous study [58].

EEG recording systems and studied parameters

EEG was recorded at T₀, T₁, and T_1 +60 while the subjects were sitting in a comfortable recliner chair in a dim and quiet room, keeping their eyes closed. Since our study was designed to be easy application in clinical setting, we used a standard 19 Ag-AgCl ring electrode arrangement according to 10-20-International System (Fp1, Fp2, F7, F3, Fz, F4, F8, T7, C3, Cz, C4, T8, P7, P3, Pz, P4, P8, O1, O2), plus a ground (forehead) and a reference (left mastoid) electrode. Eye movements and blinks were detected through additional electrodes placed on the right peri-orbital region. Data were acquired using a Brain-Quick System (Micromed, Mogliano Veneto, Italy), sampled at 512 Hz, and filtered at a 0.3/120 Hz bandwidth (+50 Hz notch). Skin-electrode impedance was kept ≤ 5 kΩ . Data were stored on a personal computer for offline analysis accessibility through a free license of EEGLAB toolbox [59]. Raw data were processed to eliminate artifacts from neurophysiological components (visual inspection and independent component analysis). The focal and brief high power-high frequency components were isolated and removed from an independent component analysis to avoid possible contamination of the gamma-band with muscle high-frequency activity. Thus, a 3-minute artifact-free EEG was segmented into 1-s epochs (180 epochs) at each time T (Fig. 1). For each epoch, we quantified the gamma-band POW by applying a Fast Fourier Transform (Hamming window, frequency resolution 1 Hz). The mean absolute band-POW was calculated by averaging single POW values. The temporal POW variations at T₁ and T_1 +60 were calculated as compared to T₀, according to the Pfurtscheller’s formula [60], in three groups of electrodes: frontal (Fp1, Fp2, F7, F3, Fz, F4, F8), central (T7, C3, Cz, C4, T8), and parietooccipital (P7, P3, Pz, P4, P8, O1, O2).

Statistical analysis

The differences between the groups in clinical, neuropsychological, and GBO power parameters were assessed by one-way ANOVA, employing group (three levels: NOLD, AD, and MCI) as between-subject factor. For POW values, the sphericity assumption was assessed by Mauchley’s test, and the Huynh-Feldt correction factor ɛ was applied to compensate for possible effects of non-sphericity in the compared measurements; this factor reduces the degrees of freedom of the F-test. In all the conditions, the normal data distribution was evaluated with the Kolmogorov-Smirnov test. Concerning tACS after-effects, we applied a four-way ANOVA for repeated measures (rmANOVA) with time (three levels: T₀, T₁, and T_1 +60), protocol (six levels: M1, DLPFC, DMPFC, PMA, SMA, and sham), group of electrode (three levels: frontal, central, and parietooccipital) as within-subject factors, and group (three levels: NOLD, AD, and MCI) as between-subject factor. Age, education, gender, MMSE, disease duration, and drug intake were employed as covariates in the rmANOVAs. A p-value <0.05 was considered significant. Post-hoc paired-sample t-tests were carried out to assess the significance of interactions, applying the Bonferroni correction for multiple comparisons. Possible correlations among neuropsychological and tACS after-effects were assessed through a Pearson’s correlation (for each site of tACS application).

Receiver operating characteristics (ROC) analysis was used to analyze the relative predictive power of GBO power modulation at T₁ to predict MCI-to-dementia conversion at the two-year follow-up. We calculated the area under the curves (AUC) and the sensitivities and specificities for dementia [61].

RESULTS

The entire experimental procedure was well tolerated by each individual, without reporting any significant side effect, including visual disturbances.

Baseline

AD patients showed a poorer performance in nearly all the neuropsychological tests than the MCI patients did (Tables 1 and 2, Supplementary Table 1). Both groups of patients showed a decrease in α and beta activity at the parietooccipital lobes and an increase of θ, δ, parietooccipital gamma activities, and of θ/gamma ratio (more in AD than MCI patients) (Fig. 1, Table 2, Supplementary Table 1) as compared to the EEG features in NOLD and the literature data [8 , 62–63].

tACS after-effects

DLPFC, DMPFC, and M1 tACS induced a significant increase in GBO power at T₁ and T_1 +60 in the NOLD group (Fig. 2, Table 3, Supplementary Table 1). In detail, GBO increase was distributed over all the groups of electrodes after DLPFC stimulation, limited to frontal-central electrodes after DMPFC stimulation, and limited to central electrodes after M1-tACS. The other sites of stimulation had no effect. Such electrophysiological findings were paralleled by a significant improvement in RML, DS, CVF, LVF, and AM after DLPFC- and DMPFC-tACS (Fig. 3, Table 4, Supplementary Table 1). To this end, we observed a significant correlation between GBO power magnitude and AM increase and GBO power magnitude and CDT increase (r = 0.510, p = 0.02). None of the covariates influenced rmANOVAfindings.

Twenty-one out of 25 MCI patients, namely responder MCI (MCI-R), showed a GBO power increase in all the electrodes at T₁ (as well as at T_1 +60) after DMPFC-tACS (Fig. 2, Table 3, Supplementary Table 1). These electrophysiological findings were paralleled by a mild but significant improvement in RML, DS, CVF, LVF, and AM (Fig. 3, Table 4, Supplementary Table 1). To this end, we observed a significant correlation between frontal and central GBO power magnitude and AM increase (r = 0.706, p = 0.04) and between frontal and central GBO power magnitude and CDT increase (r = 0.845, p = 0.03) after DMPFC-tACS. On the other hand, four MCI patients, namely non-responder MCI (MCI-NR), had a clinical and neuropsychological profile similar to that of the MCI-R individuals at T₀, but they did not show any significant DMPFC-tACS after-effect at T₁ and T_1 +60 (Supplementary Table 1, Fig. 4).

Finally, there were no significant clinical and electrophysiological after-effects in the AD sample (Tables 3 and 4, Supplementary Table 1, Figs. 2 and 3).

Two-year follow-up

All the NOLD, all the MCI-NR, 17 out of 21 MCI-R, and 32 out of 35 AD subjects were included in the two-year follow-up; one MCI-R and two AD patients died, one MCI-R and one AD refused the follow-up, and two MCI-R were lost. All the NOLD and the 17 MCI-R individuals remained substantially stable at the neuropsychological tests (Table 1, Supplementary Table 1). Instead, the MCI-NR individuals were diagnosed with AD (Table 1, Supplementary Table 1). In addition, they had a significant increase in GBO power magnitude when compared to T₀, whereas MCI-R individuals did not show such an increase (time×group interaction F_(1,20) = 21, p < 0.002; MCI-NR t_(1,3) = 5.7, p = 0.01; MCI-R p = 0.3) (time: T₀ and T₂; group: MCI-R and MCI-NR) (Fig. 4). Hence, the MCI patients who were not responsive to tACS protocol (MCI-NR) converted to AD within two years. Moreover, such patients showed a GBO increase while recording EEG in resting conditions at T₂ (Fig. 4). ROC analysis showed that a clear prediction of MCI-to-AD conversion (AUC = 0.88; SE = 0.083) was obtained with a GBO power magnitude increase of at least + 15% after tACS application (i.e., at T₁) (Table 5, Fig. 5).

Finally, all the AD patients remained substantially stable concerning neuropsychological assessment (Table 1, Supplementary Table 1), except a deterioration in MMSE (p < 0.001), RML (p < 0.001), and AM (p < 0.001) performances. In addition, all the AD patients showed a GBO power increase while recording EEG in resting conditions (Fig. 4).

DISCUSSION

To the best of our knowledge, this is the first study employing tACS in a sample of AD and MCI patients. Our data suggest two significant issues: 1) tACS could be a safe and efficient tool in supporting the clinical differential diagnosis of AD and MCI, and 2) tACS could identify those among MCI patients who could have a high risk of developing dementia.

tACS-based AD/MCI differential diagnosis

DMPFC-tACS was able to differentiate MCI and AD patients at individual and group level. In fact, the AD sample did not show any significant tACS after-effect. Nonetheless, four MCI patients showed very mild tACS after-effects (MCI-NR).

The tACS-induced neuropsychological improvement in NOLD and MCI-R individuals was correlated with the GBO power increase, similarly to the growth of GBO during the execution of cognitive tasks that has been reported in other studies [31, 64]. In fact, GBO power increase is associated with many cognitive processes [26] due to its role in temporal binding processes, cortico-cortical transmission, and information integration among neural networks [27 –31]. On the other hand, AD and MCI-NR patients did not show any tACS-induced GBO increase.

Notably, MCI-R patients showed a maladaptive DMPFC-tACS induced GBO increase as compared to NOLD individuals. In fact, GBO increase was mild and associated with a significant DLPFC and DMPFC connectivity dysfunction. In addition, an increase in GBO power may reflect the increased effort in maintaining rapid information processing and mentation in resting conditions in cognitively declining patients [65, 66]. Recent experimental reports have further suggested the importance of the correlation between GBO impairment and cortico-cortical connectivity breakdown regarding cognitive decline [67, 68]. Indeed, tACS induces an increase in coherent local activity, thus enhancing information transfer and processing within the subset of oscillating networks subserving a function [51, 69]. Since we observed that the more deteriorated the cognitive performance, the more limited and poorer organized the tACS-induced GBO modulations, we may argue that the partial GBO activation associated with the DLPFC/DMPFC connectivity dysfunction may represent a marker of the degree of cognitive and connectivity impairment. This issue is further corroborated by the fact that the DLPCF/DMPFC dysfunction foresaw the cognitive decline at the two-year follow-up. Therefore, the growing impairment of frontal and frontoparietal connectivity, expressed by the mild tACS-induced GBO increase, may account for the cognitive and behavioral impairment [70 –73].

Hence, MCI-R patients could have engaged a unique network involving the DMPFC, thus ensuring a residual cognitive performance probably through a compensatory reallocation or a de novo recruitment of cognitive resources within frontoparietal networks, namely cognitive reserve [74 –78]. On the other hand, AD subjects who lacked such compensatory mechanisms because of a more impaired cortico-cortical functional connectivity were devoid of the tACS-induced potentiation of cognitive performances. These observations agree with previous data suggesting that the so-called default mode is more persistently active in AD than MCI patients, probably owing to an increased GBO modulation that might be associated to an exaggeration of cortical facilitator processes [31 , 79–81]. Therefore, MCI-R patients may have a residual capacity of modulating GBO, which allows a better neuropsychological performance in comparison to AD individuals (who totally lack such capacity). On the other hand, the MCI-NR patients may have taken advantage of their cognitive reserve. Therefore, they were able to perform the neuropsychological tests in resting condition but were unable to improve when GBO and frontoparietal connectivity were perturbed using tACS, by reason of the subthreshold frontoparietal network connectivity impairment.

The specificity of GBO modulation concerning cognitive impairment is suggested by the fact that the EEG features remained stable at the two-year follow-up, except the GBO power magnitude, which increased significantly in AD and MCI-NR individuals. The shorter follow-up duration and the relatively long disease duration (5±1 for MCI, 9±2 for AD) could justify the preservation of the EEG features at T₂ as compared to previous studies that demonstrated a deterioration of other brain rhythms [63]. In addition, tACS did not significantly entrain other brain rhythms, as instead it has been reported in another study employing γ-range tACS [82]. This discrepancy was probably due to the specifications of our tACS procedure (including site, duration, and intensity of stimulation, alternating current properties).

MCI to AD progression

The MCI-NR patients (i.e., those among MCI patients who did not show significant clinical and electrophysiological DMPFC after-effects) converted to AD over a period of two years. Notably, these MCI-NR patients had baseline clinical scores and GBO power magnitudes that were similar to the other MCI subjects. Therefore, we may argue that the MCI patients who do not show significant DMPFC-tACS after-effects on GBO power may be at high risk of developing dementia, as suggested by the strong classifier discriminant power of DMPFC-tACS GBO modulation, and the significant effect of GBO power demodulation at baseline (T₀) on the GBO power at the two-year follow-up (T₂). To this end, we may hypothesize that the cognitive reserve was already used at baseline (T₀) in these MCI patients that other external amounts of plasticity (i.e., tACS-delivered at T₁) could not have further perturbed GBO [83]. In keeping with these findings, neuroimaging studies showed an increase in prefrontal activity, a reduced correlation between blood oxygen level-dependent activity and the activity of several circuits within the frontoparietal networks at rest, and an abnormal functional coupling among resting state cortical EEG rhythms in some early AD and MCI patients, as compared to NOLD controls [74–76 , 84]. Indeed, a persistent and subthreshold hyper-activation of neural networks could favor the progression of MCI to dementia since Aβ deposition is increased in those areas showing high metabolic activity [85–86]. The prominent involvement of DLPFC and DMPFC in MCI-to-AD progression may depend on specific cytoarchitectonic and connectivity properties of these regions [87–88], with particular regard to frontoparietal and temporal-limbic networks, which show a high degree of anterograde or retrograde degenerative processes [86 , 88–91]. Nonetheless, these issues should be verified by studies with larger samples and longer follow-up.

CONCLUSIONS

Even though we limited our study to GBO analysis, our data suggest that tACS may be of some help in the clinical differential diagnosis of MCI and AD, in particular among those patients showing borderline scores at the neuropsychological assessment. In fact, the evaluation of the sensitivity of GBO to external perturbations may offer a reliable index of cognitive reserve, which may be used to monitor its relative impact on cognitive domains and to predict the evolution of prodromal stages of AD [84]. Indeed, the individuals showing an abnormal GBO modulation within frontoparietal networks may be considered at potential risk of developing dementia. This issue could help clinicians to adopt a better prevention/management of such disabling neurodegenerative disease.

Footnotes

ACKNOWLEDGMENTS

We would like to thank Prof. Dorothy Mckinley and Ms. Antonina Donato for having edited the paper in English language.

Authors’ disclosures available online ().

Appendix

References

Cornutiu

(2015) The epidemiological scale of Alzheimer’s disease, J Clin Med Res 7, 657–666.

Andrieu

, Coley

, Lovestone

, Aisen

, Vellas

(2015) Prevention of sporadic Alzheimer’s disease: Lessons learned from clinical trials and future directions, Lancet Neurol 14, 926–944.

Janoutová

, Šerý

, Hosák

, Janout

(2015) Is mild cognitive impairment a precursor of Alzheimer’s disease? Short review, Cent Eur J Public Health 23, 365–367.

Cooper

, Sommerlad

, Lyketsos

, Livingston

(2015) Modifiable predictors of dementia in mild cognitive impairment: A systematic review and meta-analysis, Am J Psychiatry 172, 323–334.

Snyder

, Kahle-Wrobleski

, Brannan

, Miller

, Schindler

, DeSanti

, Ryan

, Morrison

, Grundman

, Chandler

, Caselli

, Isaac

, Bain

, Carrillo

(2014) Assessing cognition and function in Alzheimer’s disease clinical trials: Do we have the right tools?, Alzheimers Dement 10, 853–860.

Vega

, Newhouse

(2014) Mild cognitive impairment: Diagnosis, longitudinal course, and emerging treatments, Curr Psychiatry Rep 16, 490.

Hugo

, Ganguli

(2014) Dementia and cognitive impairment: Epidemiology, diagnosis, and treatment, Clin Geriatr Med 30, 421–442.

Drago

, Babiloni

, Bartrés-Faz

, Caroli

, Bosch

, Hensch

, Didic

, Klafki

, Pievani

, Jovicich

, Venturi

, Spitzer

, Vecchio

, Schoenknecht

, Wiltfang

, Redolfi

, Forloni

, Blin

, Irving

, Davis

, Hårdemark

, Frisoni

(2011) Disease tracking markers for Alzheimer’s disease at the prodromal (MCI) stage, J Alzheimers Dis 263, 159–199.

Bertè

, Lamponi

, Calabrò

, Bramanti

(2014) Elman neural network for the early identification of cognitive impairment in Alzheimer’s disease, Funct Neurol 29, 57–65.

10.

Hamm

, Héraud

, Cassel

, Mathis

, Goutagny

(2015) Precocious alterations of brain oscillatory activity in Alzheimer’s disease: A window of opportunity for early diagnosis and treatment, Front Cell Neurosci 9, 491.

11.

Salgado-Puga

, Pena-Ortega

(2015) Cellular and network mechanisms underlying memory impairment induced by amyloid β protein, Protein Pept Lett 22, 303–321.

12.

Kim

, Yoon

, Basak

, Kim

(2014) Apolipoprotein E in synaptic plasticity and Alzheimer’s disease: Potential cellular and molecular mechanisms, Mol Cells 37, 767–776.

13.

Birch

(2014) The contribution of astrocytes to Alzheimer’s disease, Biochem Soc Trans 42, 1316–1320.

14.

Yener

, Başar

(2013) Brain oscillations as biomarkers in neuropsychiatric disorders: Following an interactive panel discussion and synopsis, Suppl Clin Neurophysiol 62, 343–363.

15.

Başar

, Güntekin

, Tülay

, Yener

(2010) Evoked and event related coherence of Alzheimer patients manifestdifferentiation of sensory-cognitive networks, Brain Res 1357, 79–90.

16.

Yener

, Güntekin

, Örken

, Tülay

, Forta

, Başar

(2012) Auditory delta event-related oscillatory responses are decreased in Alzheimer’s disease, Behav Neurol 25, 3–11.

17.

Yener

, Başar

(2010) Sensory evoked and event related oscillations in Alzheimer’s disease: A short review, Cogn Neurodyn 4, 263–274.

18.

Sheng

, Sabatini

, Südhof

(2012) Synapses and Alzheimer’s disease, Cold Spring Harb Perspect Biol 4, 5.

19.

Uhlhaas

, Singer

(2006) Neural synchrony in brain disorders: Relevance for cognitive dysfunctions and pathophysiology, Neuron 52, 155–168.

20.

Moliadze

, Antal

, Paulus

(2010) Boosting brain excitability by transcranial high frequency stimulation in the ripple range, J Physiol 588, 4891–4904.

21.

Kuo

, Nitsche

(2012) Effects of transcranial electrical stimulation on cognition, Clin EEG Neurosci 43, 192–199.

22.

Santarnecchi

, Muller

, Rossi

, Sarkar

, Polizzotto

, Rossi

, Cohen Kadosh

(2016) Individual differences and specificity of prefrontal gamma frequency-tACS on fluid intelligence capabilities, Cortex 75, 33–43.

23.

Hoy

, Bailey

, Arnold

, Windsor

, John

, Daskalakis

, Fitzgerald

(2015) The effect of γ-tACS on working memory performance in healthy controls, Brain Cogn 101, 51–56.

24.

Cabral-Calderin

, Anne Weinrich

, Schmidt-Samoa

, Poland

, Dechent

, Bähr

, Wilke

(2016) Transcranial alternating current stimulation affects the BOLD signal in a frequency and task-dependent manner, Hum Brain Mapp 37, 94–121.

25.

Vosskuhl

, Huster

, Herrmann

(2015) Increase in short-term memory capacity induced by down-regulating individual theta frequency via transcranial alternating current stimulation, Front Hum Neurosci 9, 257.

26.

Abuhassan

, Coyle

, Maguire

(2011) Employing neuronal networks to investigate the pathophysiological basis of abnormal cortical oscillations in Alzheimer’s disease, Conf Proc IEEE Eng Med Biol Soc 2011, 2065–2068.

27.

Booth

, Ridler

, Murray

, Ward

, de Groot

, Goodfellow

, Phillips

, Randall

, Brown

(2016) Electrical and network neuronal properties are preferentially disrupted in dorsal, but not ventral, medial entorhinal cortex in a mouse model of tauopathy, J Neurosci 36, 312–324.

28.

Missonnier

, Herrmann

, Michon

, Fazio-Costa

, Gold

, Giannakopoulos

(2010) Early disturbances of gamma band dynamics in mild cognitive impairment, J Neural Transm (Vienna) 117, 489–498.

29.

Moretti

, Frisoni

, Fracassi

, Pievani

, Geroldi

, Binetti

, Rossini

, Zanetti

(2011) MCI patients’ EEGs show group differences between those who progress and those who do not progress to AD, Neurobiol Aging 32, 563–571.

30.

Park

, Lee

, An

, Lee

, Kim

, Namkoong

. Gamma oscillatory activity in relation to memory ability in older adults, Int J Psychophysiol 86, 58–65.

31.

vanDeursen

, Vuurman

, Verhey

, van Kranen-Mastenbroek

, Riedel

(2008) Increased EEG gamma band activity in Alzheimer’s disease and mild cognitive impairment, J Neural Transm (Vienna) 115, 1301–1311.

32.

Babiloni

, Carducci

, Lizio

, Vecchio

, Baglieri

, Bernardini

, Cavedo

, Bozzao

, Buttinelli

, Esposito

, Giubilei

, Guizzaro

, Marino

, Montella

, Quattrocchi

, Redolfi

, Soricelli

, Tedeschi

, Ferri

, Rossi-Fedele

, Ursini

, Scrascia

, Vernieri

, Pedersen

, Hardemark

, Rossini

, Frisoni

(2013) Resting state cortical electroencephalographic rhythms are related to gray matter volume in subjects with mild cognitive impairment and Alzheimer’s disease, Hum Brain Mapp 34, 1427–1446.

33.

Babiloni

, Lizio

, Carducci

, Vecchio

, Redolfi

, Marino

, Tedeschi

, Montella

, Guizzaro

, Esposito

, Bozzao

, Giubilei

, Orzi

, Quattrocchi

, Soricelli

, Salvatore

, Baglieri

, Bramanti

, Cavedo

, Ferri

, Cosentino

, Ferrara

, Mundi

, Grilli

, Pugliese

, Gerardi

, Parisi

, Vernieri

, Triggiani

, Pedersen

, Hårdemark

, Rossini

, Frisoni

(2011) Resting state cortical electroencephalographic rhythms and white matter vascular lesions in subjects with Alzheimer’s disease: An Italian multicenter study, J Alzheimers Dis 26, 331–346.

34.

Hsiao

, Chen

, Wang

, Yan

, Lin

(2014) Altered source-based EEG coherence of resting-state sensorimotor network in early-stage Alzheimer’s disease compared to mild cognitive impairment, Neurosci Lett 558, 47–52.

35.

McBride

, Zhao

, Munro

, Smith

, Jicha

, Jiang

(2013) Resting EEG discrimination of early-stage Alzheimer’s disease from normal aging using inter-channel coherence network graphs, Ann Biomed Eng 41, 1233–1342.

36.

Moretti

, Frisoni

, Binetti

, Zanetti

(2011) Anatomical substrate and scalp EEG markers are correlated in subjects with cognitive impairment and Alzheimer’s disease, Front Psychiatry 1, 152.

37.

Rossini

, Del Percio

, Pasqualetti

, Cassetta

, Binetti

, Dal Forno

, Ferreri

, Frisoni

, Chiovenda

, Miniussi

, Parisi

, Tombini

, Vecchio

, Babiloni

(2006) Conversion from mild cognitive impairment to Alzheimer’s disease is predicted by sources and coherence of brain electroencephalography rhythms, Neuroscience 143, 793–803.

38.

Vecchio

, Miraglia

, Marra

, Quaranta

, Vita

, Bramanti

, Rossini

. Human brain networks in cognitive decline: A graph theoretical analysis of cortical connectivity from EEG data, J Alzheimers Dis 41, 113–127.

39.

Tam

, Dansereau

, Badhwar

, Orban

, Belleville

, Chertkow

, Dagher

, Hanganu

, Monchi

, Rosa-Neto

, Shmuel

, Wang

, Breitner

, Bellec

, the Alzheimer’s Disease Neuroimaging Initiative (2015) Common effects of amnestic mild cognitive impairment on resting-state connectivity across four independent studies, Front Aging Neurosci 7, 242.

40.

Babiloni

, Vecchio

, Lizio

, Ferri

, Rodriguez

, Marzano

, Frisoni

, Rossini

(2011) Resting state cortical rhythms in mild cognitive impairment and Alzheimer’s disease: Electroencephalographic evidence, J Alzheimers Dis 26S3, 201–214.

41.

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

Jr , Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wagster

, Phelps

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

42.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

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

43.

Morris

(1993) The Clinical Dementia Rating (CDR): Current vision and scoring rules, Neurology 43, 2412–2414.

44.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) 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 7, 270–279.

45.

Jack

Jr , Albert

, Knopman

, McKhann

, Sperling

, Carrillo

, Thies

, Phelps

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

46.

Petersen

, Doody

, Kurz

, Mohs

, Morris

, Rabins

, Ritchie

, Rossor

, Thal

, Winblad

(2001) Current concepts in mild cognitive impairment, Arch Neurol 58, 1985–1992.

47.

Lezak

, Howieson

, Loring

(2004), Neuropsychological Assessment (4th edition). Oxford University Press.

48.

Iyer

, Mattu

, Grafman

, Lomarev

, Sato

, Wassermann

(2005) Safety and cognitive effect of frontal DC brain polarization in healthy individuals, Neurology 64, 872–875.

49.

Wach

, Krause

, Moliadze

, Paulus

, Schnitzler

, Pollok

(2013) Effects of 10Hz and 20Hz transcranial alternating current stimulation tACS on motor functions and motor cortical excitability, Behav Brain Res 241, 1–6.

50.

Nitsche

, Doemkes

, Karaköse

, Antal

, Liebetanz

, Lang

, Tergau

, Paulus

(2007) Shaping the effects of transcranial direct current stimulation of the human motor cortex, J Neurophysiol 97, 3109–3117.

51.

Polanía

, Nitsche

, Korman

, Batsikadze

, Paulus

(2012) The importance of timing in segregated theta phase-coupling for cognitive performance, Curr Biol 22, 1314–1318.

52.

Veniero

, Vossen

, Gross

, Thut

(2015) Lasting EEG/MEG aftereffects of rhythmic transcranial brain stimulation: Level of control over oscillatory network activity, Front Cell Neurosci 9, 477.

53.

Reato

, Rahman

, Bikson

, Parra

(2013) Effects of weak transcranial alternating current stimulation on brain activity-a review of known mechanisms from animal studies, Front Hum Neurosci 7, 687.

54.

Miranda

, Mekonnen

, Salvador

, Ruffini

(2013) The electric field in the cortex during transcranial current stimulation, Neuroimage 70, 48–58.

55.

Windhoff

, Opitz

, Thielscher

(2013) Electric field calculations in brain stimulation based on finite elements: An optimized processing pipeline for the generation and usage of accurate individual head models, Hum Brain Mapp 34, 923–935.

56.

Opitz

, Zafar

, Bockermann

, Rohde

, Paulus

(2014) Validating computationally predicted TMS stimulation areas using direct electrical stimulation in patients with brain tumors near precentral regions, Neuroimage Clin 4, 500–507.

57.

Truong

, Hüber

, Xie

, Datta

, Rahman

, Parra

, Dmochowski

, Bikson

(2014) Clinician accessible tools for GUI computational models of transcranial electrical stimulation: BONSAI and SPHERES, Brain Stimul 7, 521–524.

58.

Lee

, Choi

, Oh

, Sohn

, Lee

(2016) Treatment of Alzheimer’s disease with repetitive transcranial magnetic stimulation combined with cognitive training: A prospective, randomized, double-blind, placebo-controlled study, J Clin Neurol 12, 57–64.

59.

Delorme

, Makeig

(2004) EEGLAB: An open source toolbox for analysis of single trial EEG dynamics including independent component analysis, J Neurosci Meth 134, 921.

60.

Pfurtscheller

, Lopes da Silva

(1999) Event-related EEG/MEG synchronization and desynchronization: Basic principles, Clin Neurophysiol 110, 1842–1857.

61.

Maroco

, Silva

, Rodrigues

, Guerreiro

, Santana

, de Mendonça

(2011) Data mining methods in the prediction of Dementia: A real-data comparison of the accuracy, sensitivity and specificity of linear discriminant analysis, logistic regression, neural networks, support vector machines, classification trees and random forests, BMC Res Notes 4, 299.

62.

Moretti

(2015) Conversion of mild cognitive impairment patients in Alzheimer’s disease: Prognostic value of Alpha3/Alpha2 electroencephalographic rhythms power ratio, Alzheimers Res Ther 7, 80.

63.

Moretti

(2015) Theta and alpha EEG frequency interplay in subjects with mild cognitive impairment: Evidence from EEG, MRI, and SPECT brain modifications, Front Aging Neurosci 7, 31.

64.

van Deursen

, Vuurman

, van Kranen-Mastenbroek

, Verhey

, Riedel

(2011) 40Hz steady state response in Alzheimer’s disease and mild cognitive impairment, Neurobiol Aging 32, 24–30.

65.

Wang

, Xia

, Dai

, Liang

, Song

, He

, Li

(2015) Differentially disrupted functional connectivity of the subregions of the inferior parietal lobule in Alzheimer’s disease, Brain Struct Funct 220, 745–762.

66.

Becker

, Mintun

, Aleva

, Wiseman

, Nichols

, DeKosky

(1996) Compensatory reallocation of brain resources supporting verbal episodic memory in Alzheimer’s disease, Neurology 46, 692–700.

67.

Ahnaou

, Huysmans

, Jacobs

, Drinkenburg

(2014) Cortical EEG oscillations and network connectivity as efficacy indices for assessing drugs with cognition enhancing potential, Neuropharmacology 86, 362–377.

68.

Liu

, Bai

, Wang

, Tian

(2016) An aberrant link between gamma oscillation and functional connectivity in Aβ1-42-mediated memory deficits in rats, Behav Brain Res 297, 51–58.

69.

Fries

(2005) A mechanism for cognitive dynamics: Neuronal communication through neuronal coherence, Trends Cogn Sci 9, 474–480.

70.

Schroeter

, Stein

, Maslowski

, Neumann

(2009) Neural correlates of Alzheimer’s disease and mild cognitive impairment: A systematic and quantitative meta-analysis involving 1351 patients, Neuroimage 47, 1196–1206.

71.

Babiloni

, Lizio

, Marzano

, Capotosto

, Soricelli

, Triggiani

, Cordone

, Gesualdo

, Del Percio

(2016) Brain neural synchronization and functional coupling in Alzheimer’s disease as revealed by resting state EEG rhythms, Int J Psychophysiol 103, 88–102.

72.

Coyle

, Bhattacharya

, Zou

, Wong-Lin

, Abuhassan

, Maguire

(2014) Neural circuit models and neuropathological oscillations. In Handbook of Bio/Neuroinformatics, Kasabov

, ed. Springer, Berlin Heidelberg, pp. 673–702.

73.

Sen Bhattacharya

, Coyle

, Maguire

(2011) A thalamocortical neural mass model to study brain rhythms in Alzheimer’s disease, Neurocomp Mod Brain Disord 24, 631–645.

74.

Barulli

, Stern

(2013) Efficiency, capacity, compensation, maintenance, plasticity: Emerging concepts in cognitive reserve, Trends Cogn Sci 17, 502–509.

75.

Ferrari

, Nacmias

, Bagnoli

, Piaceri

, Lombardi

, Pradella

, Tedde

, Sorbi

(2014) Imaging and cognitive reserve studies predict dementia in presymptomatic Alzheimer’s disease subjects, Neurodegener Dis 13, 157–159.

76.

Grady

, McIntosh

, Beig

, Keightley

, Burian

, Black

(2003) Evidence from functional neuroimaging of a compensatory prefrontal network in Alzheimer’s disease, J Neurosci 23, 986–893.

77.

Whalley

, Staff

, Fox

, Murray

(2016) Cerebral correlates of cognitive reserve, Psychiatry Res 247, 65–70.

78.

, Yu

, Tan

(2015) Cognitive reserve and Alzheimer’s disease, Mol Neurobiol 51, 187–208.

79.

Di Lazzaro

, Oliviero

, Pilato

, Saturno

, Dileone

, Marra

, Daniele

, Ghirlanda

, Gainotti

, Tonali

(2004) Motor cortex hyperexcitability to transcranial magnetic stimulation in Alzheimer’s disease, J Neurol Neurosurg Psychiatry 75, 555–559.

80.

Nardone

, Bratti

, Tezzon

(2006) Motor cortex inhibitory circuits in dementia with Lewy bodies and in Alzheimer’s disease, J Neural Transm 113, 1679–1684.

81.

Nardone

, Marth

, Ausserer

, Bratti

, Tezzon

(2006) Reduced short latency afferent inhibition in patients with Down syndrome and Alzheimer-type dementia, Clin Neurophysiol 117, 2204–2210.

82.

Santarnecchi

, Polizzotto

, Godone

, Giovannelli

, Feurra

, Matzen

, Rossi

(2013) Frequency-dependent enhancement of fluid intelligence induced by transcranial oscillatory potentials, Curr Biol 23, 1449–1453.

83.

Algarabel

, Sales

, Pitarque

, Meléndez

, Escudero

, Mayordomo

(2016) Associative and implicit memoryperformance as a function of cognitive reserve in elderly adults with and without mild cognitive impairment.E, Span J Psychol 19, 4.

84.

Teipel

, Grothe

, Zhou

, Sepulcre

, Dyrba

, Sorg

, Babiloni

(2016) Measuring cortical connectivity in Alzheimer’s disease as a brain neural network pathology: Toward clinical applications, J Int Neuropsychol Soc 22, 138–163.

85.

Habeck

, Risacher

, Lee

, Glymour

, Mormino

, Mukherjee

, Kim

, Nho

, DeCarli

, Saykin

, Crane

, Alzheimer’s

Disease Neuroimaging Initiative

(2012) Relationship between baseline brain metabolism measured using 18FFDG-PET and memory and executive function in prodromal and early Alzheimer’s disease, Brain Imaging Behav 6, 568–583.

86.

Cerami

, Della Rosa

, Magnani

, Santangelo

, Marcone

, Cappa

, Perani

(2014) Brain metabolic maps in mild cognitive impairment predict heterogeneity of progression to dementia, Neuroimage Clin 7, 187–194.

87.

Salat

, Kaye

, Janowsky

(2001) Selective preservation and degeneration within the prefrontal cortex in aging and Alzheimer disease, Arch Neurol 58, 1403–1408.

88.

van Veluw

, Sawyer

, Clover

, Cousijn

, De Jager

, Esiri

, Chance

(2012) Prefrontal cortex cytoarchitecture in normal aging and Alzheimer’s disease: A relationship with IQ, Brain Struct Funct 217, 797–808.

89.

Agosta

, Pievani

, Geroldi

, Copetti

, Frisoni

, Filippi

(2012) Resting state fMRI in Alzheimer’s disease: Beyond the default mode network, Neurobiol Aging 33, 1564–1578.

90.

Dhanjal

, Wise

(2014) Fronto-parietal cognitive control of verbal memory recall in Alzheimer’s disease, Ann Neurol 76, 241–251.

91.

Stam

(2014) Modern network science of neurological disorders, Nat Rev Neurosci 15, 683–695.