Autonomic Cardiac Function in Preclinical Alzheimer’s Disease

Abstract

To explore early autonomic cardiac changes in pre-clinical Alzheimer’s disease (AD), we have evaluated electrocardiologic measures of vagal tone for 63 adults (ages 55–75) at rest, during cognitive testing, and then again at rest. All subjects had multiple risk factors for AD, and all completed amyloid PET scans (¹⁸F-Florbetapir) to determine amyloid positivity (Aβ+). No change in electrocardiographic measures were observed for Aβ+ participants under each testing condition, whereas Aβ–subjects showed an expected increase in vagal tone during the cognitive stress condition. These findings suggest an early relationship between cortical Aβ accumulation, a precursor to AD development, and autonomic cardiac function.

Keywords

Aging Alzheimer’s disease cardiac heart rate variability resting sinus arrhythmia vagal tone

INTRODUCTION

A recent review of the complex relationship between Alzheimer’s disease (AD), cardiovascular disease, and cerebrovascular disease [1] highlights the many shared risk factors and pathophysiologic mechanisms across these diseases and, as a result, the preclinical stages of these diseases may often concurrently affect the same at-risk population [1]. Clinical studies demonstrate that even mild cerebrovascular pathology results in reduced cognitive performance in very early AD [2, 3]. Additionally, there is evidence that vascular/cerebrovascular pathology can accelerate the progression of AD [4, 5]. Moreover, the blood-brain barrier seems to be disrupted early in AD, causing dysfunction in brain perfusion [6] and reduced cerebral blood flow [7]. It remains unclear just how early in the disease course the shared pathological relationships between AD, cardiovascular disease, and cerebrovascular disease can be observed. The exploration of the mechanistic relationships between these disease clusters in a preclinical population is important to better understand the complicated co-occurrence of these three diseases.

We recently examined the relationship between central nervous system amyloid-β (Aβ) burden, and a simple measure of myocardial oxygen consumption: rate pressure product, a well-known cardiac workload marker that reflects myocardial stress based on the number of times that the heart needs to beat per minute in relation to the arterial blood pressure that it is pumping against, in cognitively normal participants at high-risk for the development of AD [8]. This work supports a relationship between elevated neocortical Aβ aggregation and inefficient myocardial oxygen consumption in the absence of significant metabolic demands (after accounting for expected effects of age).

In the current study, we sought to explore if the changes in cardiac workload that we see in preclinical AD during cognitive testing are due to centrally-mediated changes in vagal tone measured by vagal ratio and the resting sinus arrhythmia (RSA). Both of these indices become less reactive and they decrease in amplitude with normal aging, due to declines in parasympathetic tone [9, 10]. The vagal ratio is the ratio between low and high frequency components of the Heart Rate Variability (HRV) and the RSA is a measure of the HRV in synchrony with respiration. In this study, rapid changes in these two indices of vagal tone were measured before, during, and immediately after the administration of a cognitive stressor, to explore the adaptive flexibility of parasympathetic autonomic function in response to a mild cognitive stressor (performance on a hidden maze learning task).

METHODS

Participants

Sixty-three adults aged between 55 and 75 years (mean age = 62.79 years), all at some risk for development of AD based on the presence of two risk factors for the disease (a first-degree family history and subjective memory complaints), were recruited for a longitudinal study of individuals with possible preclinical AD [11]. From this larger sample, we identified 15 participants who were categorized as presenting with preclinical stage disease based on evidence of both: 1) elevated neocortical amyloid-β burden as determined by PET amyloid imaging [12, 13]; and 2) relative cognitive impairments in response to a challenge with a very low-dose muscarinic anticholinergics [11, 14]. Potential participants were excluded from this study if they presented with diagnoses of either mild cognitive impairment or AD, following NIA-AA diagnostic criteria [13, 15], diabetes, or histories of any neurological or psychiatric disorders. With respect to cardiovascular risk, four of 15 subjects in the preclinical AD group (26.7%) and 12 of 48 subjects in the control group (25%) presented with hypercholesterolemia that was under medication control. Two individuals in the preclinical AD group (13.3%) and 8 persons in the control group (16.6%) were under treatment for hypertension. For all subjects, a score on the Mini-Mental State Examination (MMSE) >27 was required for study entry.

With respect to genetic risk for AD, eight of 15 individuals (53%) in the preclinical AD group (Florbetapir PET standardized uptake value ratio (SUVr) scores≥1.10) had at least one copy of the APOE ɛ4 allele, whereas 20 of 45 individuals (45%) in the control group (Florbetapir PET SUVr scores <1.10) had at least one copy of the APOE ɛ4 allele. Hence, roughly half of the subjects in each group presented with this additional risk factor for AD, but due to small sample sizes we were not able to further evaluate the specific effect of APOE genetic risk on the relationship between vagal tone and cognitive performance. Subject demographic information for both groups are provided in Table 1.

Table 1

Demographic characteristics

	Main Outcome	Full sample	Preclinical AD	Healthy Controls
		(n = 63)	(n = 15)	(n = 48)
		N (%)	N (%)	N (%)	p
Sex	No. of female	39 (61.9%)	11 (73.3%)	28 (58.3%)	0.296
APOE	Number of ɛ4 carriers	29 (46%)	8 (53.3%)	21(43.8%)	0.516
Hypertension	Number of participants	10 (15.9%)	2 (13.3%)	8 (16.6%)	0.758
Hypercholesterolemia	Number of participants	16 (25.4%)	4 (26.7%)	12 (25%)	0.897
		Mean (SD)	Mean (SD)	Mean (SD)	p	Cohen’s d
Age	No. of years	62.79 (5.35)	63.93 (6.31)	62.44 (5.04)	0.349	0.28
Education	No. of years	17.21 (2.77)	17.47 (3.46)	17.14 (2.55)	0.689	0.12
Florbetapir PET SUVr	Standardized Uptake Value ratio	1.04 (0.19)	1.30 (0.20)	0.95 (0.08)	0.000	2.95
(ACC region of interest)
GDS	Total Score	1.86 (2.16)	1.60 (1.45)	1.94 (2.35)	0.602	–0.16
DASS Depression Subscale	Total Depression Subscale Score	3.56 (6.70)	2.60 (2.59)	3.87 (7.56)	0.526	–0.19
DASS Anxiety Subscale	Total Anxiety Subscale Score	2.73 (4.53)	2.40 (3.58)	2.83 (4.83)	0.752	–0.09
DASS Stress Subscale	Total Stress Subscale Score	6.73 (6.77)	6.67 (4.55)	6.74 (7.38)	0.969	–0.01
Body Mass Index	Body Mass Index	26.69 (5.50)	28.66 (7.95)	26.07 (4.41)	0.113	0.48
MMSE	Total Score	29.05 (1.02)	28.93 (1.16)	29.08 (0.99)	0.624	–0.15
ISLT Total Recall	Total words recalled	25.59 (4.22)	24.73 (4.46)	25.85 (4.16)	0.374	–0.26

SUVr, standardized uptake value ratio; ACC, anterior cingulate cortex; GDS, Geriatric Depression Scale; DASS, Depression, Anxiety, and Stress Scale; MMSE, Mini-Mental State Examination; ISLT, International Shopping List Test; bolded value is significant at the p < 0.001 level.

Neuroimaging

PET scans were performed at baseline, with 370 MBq (10 mCi±10%) bolus injection of ¹⁸F-florbetapir administered intravenously. PET standardized uptake value (SUV) data were summed and normalized to the whole cerebellum SUV, resulting in a region-to-cerebellum ratio termed SUVr. We defined amyloid positivity (Aβ+) as individuals with elevated ligand binding (SUVr≥1.1) in the anterior cingulate (ACC). Consistent with Lim et al. [14], this specific region of interest was selected because: 1) the young age of participants enrolled in the study [16, 17], and as such, we would not expect widespread neocortical amyloidosis in this very early preclinical disease stage; 2) the relationship between the ACC and early changes in cholinergic tone [18, 19]; and 3) emerging research that suggests that increased Aβ burden, in the ACC specifically, is related to memory changes in early AD [20, 21]. The SUVr calculation was performed using MIMneuro software [22]. For all cases, Aβ positivity was confirmed by consensus over-read by two board-certified radiologists who were also board-certified in NuclearMedicine [14].

Cardiac measures

Electrocardiographic (ECG) measures of RSA and vagal ratio (described above) were recorded by BioPac Systems [23] for each participant: 1) at rest, sitting quietly in a semi-supine position (120 s); 2) during the performance of the Groton Maze Learning Test (GMLT; approx. 190 s); and 3) post-test at rest in a semi-supine position (120 s).

Both measures were calculated using the HRV analysis of fixed-width intervals around events, which follows the European Heart Journal guidelines [24] for the spectral method with the Multi-epoch HRV processing [25].

RSA is computed using the peak-valley method, which uses both a recorded ECG signal and a respiration signal. By using respiration information, this analysis method can provide breath-to-breath analysis that does not require parameter tweaking for individual subjects [23, 26].

The vagal ratio measure accounts for the fact that parasympathetic control via the vagus nerve is believed to primarily modulate the R-R intervals at a rate of between 0.15 and 0.4 cycles per second. Frequency based analysis of HRV in AcqKnowledge computes vagal ratio as the amount of power in the parasympathetic band normalized to an approximation of the total power in the signal [25].

Cognitive assessment

All participants completed the GMLT (http://www.cogstate.com) as a mild cognitive stressor, during the ECG recording. The GMLT is a computerized hidden spatial maze learning task, developed by one of the authors (P.J.S.) as a measure of working memory, learning efficiency, and reasoning/problem solving [27]. This GMLT composite score was used to explore group differences in performance. All subjects completed a practice trial of the GMLT, to allow for initial task familiarity; and after a 10-min break, all subjects repeated the GMLT as a baseline assessment (at which time ECG recordings were obtained, as described above).

Additionally, all subjects completed a test of verbal episodic memory: International Shopping List Test [28], the MMSE as a measure of general cognitive function, the 15-item Geriatric Depression Scale, and the Depression, Anxiety and Stress Scale. Subjective memory impairment was determined using the Memory Complaint Questionnaire. All measures were performed by trained staff supervised by a licensed clinical neuropsychologist.

Statistical treatment of data

Generalized estimating equations (GEEs) were used to compare within-subject changes in vagal ratio and RSA across the three time points (rest à cognitive task à rest) between preclinical AD subjects (n = 15) and the control subjects (n = 48). GEEs are generalized linear models wherein the within-subjects (or other) nesting is accounted for in the residual error of the model (R-side), as contrasted with constructing random effects separate from the residual (G-side as in random, mixed, or hierarchical models) [29]. In our case, a compound symmetry variance-covariance matrix was used across testing phases, which assumes a common variance for each phase as well as a common covariance between all phases. Despite best intentions and approach, there can be non-obvious imprecision in these specification and so classical sandwich estimation [30] was used to adjust for any misspecification in this structure by subsequently adjusting the variances and covariances in the structure based on the actual distribution of the data (inverse of the empirical variance “sandwiched” between model variance(s)). The Shaffer step-down adjustment was used to maintain alpha at 0.05 across all comparisons within each model. Figures 2 and 3 were designed to simultaneously illustrate between- and within-participant variation.

Fig.1

GMTL composite z-scores for participants with low (left) versus high (right) for Aβ aggregation in the anterior cingulate region of interest (PET amyloid imaging) (p > 0.05).

Fig.2

Plots of vagal ratio in patients testing negative (blue, left) and positive (red, right) for Aβ (ACC) as a function of pre-, during, and post-GMLT testing. Least squares means (large circles) and 95% confidence intervals were adjusted for age. Small circles indicate the raw individual patient values for each phase of testing in order to convey the variability in values between patients. Light gray lines illustrate the individual changes relative to pre-GMLT testing (normalized to pre-GMLT by subtraction) in order to convey the consistency of within-subject changes. Aβ–group showed an increase in vagal ratio during the cognitive task condition, relative to either of the two at-rest conditions (Pre-GMLT: p = 0.0061; adj. p = 0.0304; Post-GMLT: p < 0.0001; adj. p = 0.0007), with post-GMLT vagal ratio not differing significantly from pre-GMLT levels (p = 0.2283; adj. p = 0.9185). Aβ+ group did not consistently show any change in vagal ratio at any point during testing (pre- versus GMLT p = 0.1837, GMLT versus post- p = 0.2221, post- versus pre-GMLT p = 0.7883; adj. p = 0.9985 for all). Further, the increased vagal ratio during the cognitive stress condition (p = 0.0099; adj. p = 0.0304) and subsequent decreases (p = 0.0013; adj. p = 0.0007) for the Aβ–group differed significantly from the Aβ+ group.

Fig.3

Plots of RSA in participants testing negative (blue, left) and positive (red, right) for Aβ (ACC) as a function of pre-, during, and post-GMLT testing. Least squares means (large circles) and 95% confidence intervals were adjusted for age. Small circles indicate the raw individual patient values for each phase of testing in order to convey the variability in values between patients. Light gray lines illustrate the individual changes relative to pre-GMLT testing (normalized to pre-GMLT by subtraction) in order to convey the consistency of within-subject changes. Aβ–subjects showed a significant increase in RSA during the cognitive task condition, relative to either of the two at-rest conditions (Pre-GMLT: p < 0.0001; adj. p < 0.0001; Post-GMLT: p≤0.0001; adj. p = 0.0003), with post-GMLT RSA not differing significantly from pre-GMLT levels (p = 0.2471; adj. p = 1.0). As with the vagal ratio measure, those in the Aβ+ group did not consistently show any change in RSA at any point during testing (pre- versus GMLT p = 0.6321, GMLT versus post- p = 0.7719, post- versus pre-GMLT p = 0.8510; adj. p = 1.0 for all). Further, the increased RSA scores during the cognitive stress condition (p = 0.0021; adj. p = 0.0104) were significantly larger for the Aβ–group compared to the Aβ+ group. The subsequent decrease in RSA for the Aβ+ group after the cognitive stress were significantly greater than for the Aβ–group prior to adjustment, but not after multiplicity adjustment (p = 0.0304; adj. p = 0.0913).

This study was approved by and complied with the regulations of Rhode Island Hospital’s Institutional Review Board, and all participants provided written informed consent.

RESULTS

There were no statistically significant differences between the two groups for any demographic or clinical outcome, with the exception of the expected elevation in cortical amyloid aggregation as demonstrated on florbetapir PET scan (Table 1). There were also no between-groups differences in at-rest vagal ratio or RSA, with respect to main effects for APOE ɛ4 carrier status or presence of hypercholesterolemia and hypertension (p > 0.05 for all analyses).

With respect to performance on the GMLT, no between-groups differences on the composite score [14], were found between those with Aβ+ PET scans versus those in the Aβ–healthy control group (Fig. 1).

However, subjects in the healthy control group (those with Aβ–PET scans) showed an increase in vagal ratio during the cognitive task condition, relative to either of the two at-rest conditions (Pre-GMLT: p = 0.0061; adj. p = 0.0304; Post-GMLT: p < 0.0001; adj. p = 0.0007), with post-GMLT vagal ratio not differing significantly from pre-GMLT levels (p = 0.2283; adj. p = 0.9185). In contrast, the preclinical AD (Aβ+) group did not consistently show any change in vagal ratio at any point during testing (pre- versus GMLT p = 0.1837, GMLT versus post- p = 0.2221, post- versus pre-GMLT p = 0.7883; adj. p = 0.9985 for all). Further, the increased vagal ratio during the cognitive stress condition (p = 0.0099; adj. p = 0.0304) and subsequent decreases (p = 0.0013; adj. p = 0.0007) for the Aβ–group differed significantly from the Aβ+ group (Fig. 2).

Similarly, to vagal ratio, the Aβ–subjects showed a significant increase in RSA during the cognitive task condition, relative to either of the two at-rest conditions (Pre-GMLT: p < 0.0001; adj. p < 0.0001; Post-GMLT: p≤0.0001; adj. p = 0.0003), with post-GMLT RSA not differing significantly from pre-GMLT levels (p = 0.2471; adj. p = 1.0). As with the vagal ratio measure, those in the Aβ+ group did not consistently show any change in RSA at any point during testing (pre- versus GMLT p = 0.6321, GMLT versus post- p = 0.7719, post- versus pre-GMLT p = 0.8510; adj. p = 1.0 for all). Further, the increased RSA scores during the cognitive stress condition (p = 0.0021; adj. p = 0.0104) were significantly larger for the Aβ–group compared to the Aβ+ group. The subsequent decrease in RSA for the Aβ+ group after the cognitive stress were significantly greater than for the Aβ–group prior to adjustment, but not after multiplicity adjustment (p = 0.0304; adj. p = 0.0913) (Fig. 3).

DISCUSSION

Participants in this preclinical AD (Aβ+) group demonstrated statistically significant attenuation in the reactivity of two separate measures (vagal ratio and RSA) of parasympathetic response to a mild cognitive stress condition. Not only was the magnitude of their response to the cognitive stressor less than those without elevated neocortical Aβ, but no significant differences found for these measures, within the Aβ+ group, across any of the three ECG time points. In order to further explore the relationship between neocortical Aβ aggregation and inefficient myocardial oxygen consumption in the absence of significant metabolic demands [8], it is essential to rely on metrics that accurately reflect the autonomic processes that control cardiac function. Cardiac autonomic dysfunction is prevalent among individuals with AD, and this generally reflects a decrease in parasympathetic activity [31, 32]. It has been previously shown that individuals with mild cognitive impairment demonstrate significant parasympathetic deficits [33]. The parasympathetic nervous system influences heart rate by the release of acetylcholine via the vagus nerve, which causes the heart rate activity to slowdown. This reduction in parasympathetic nervous system activity is often referred to as vagal withdrawal, and is indicative of impaired autonomic function [34].

The ratio between low and high frequency components of the HRV (the LF/HF ratio, or vagal ratio) has long been relied on as a measure of sympatho-vagal balance [35 –37]. Healthy individuals under mental stress while completing a cognitive task showed a reduced HF HRV component compared to a control group who were monitored while awake and at rest [38, 39]. The clinical reliability of this vagal ratio as marker of autonomic activity has been nonetheless controversial [40, 41]. Therefore, we have chosen to supplement this measure of autonomic cardiac function in preclinical AD by computing the RSA, which considers the variation in heart rate during the breathing cycle and is commonly relied on as a measure of parasympathetic nervous system activity. Typically, both vagal ratio and RSA decrease with normal aging due to decline in parasympathetic tone [9, 10].

Disruption of cholinergic neurotransmission with low-dose scopolamine has been previously shown to exacerbate subtle cognitive deficits that are otherwise not (yet) detectable for subjects in the Aβ+ group [14], whereas standard baseline testing with the GMLT does not reveal any group differences. We believe that the subjects enrolled in the Aβ+ group, for this study, are in the preclinical stage of the disease and that it might be too early to observe clear impairments in learning efficiency and working memory/executive function unless such defects are unmasked by intentionally down-regulating cholinergic tone [14].

Participants in this Aβ+ group, however, did not consistently show changes in vagal ratio or RSA at any point during the experiment, failing to demonstrate the expected response to modest stress elicited during cognitive task performance. Both measures were expected to show modest increases, reflecting heightened autonomic arousal, during completion of a neuropsychological test, even after adjusting for age effects [42]. Enhanced vagal ratio is characterized by greater vagal reactivity and faster vagal recovery from psychological stressors, and it has been linked with greater ANS flexibility and an improved response selection in the face of stress [43, 44] and return to homeostasis [45]. Conversely, in a study of healthy male sailors who completed a fitness training program, those with higher vagal ratio scores demonstrated improved performance on measures of working memory, reaction time, and attentional controls [46]. In our study, the Aβ–group demonstrated an expected increase in vagal ratio in response to the cognitive stressor, whereas the Aβ+ group demonstrated no change in vagal ratio across all three testing conditions.

The same pattern of results was observed for the related RSA measure, with participants in the Aβ–group showing an expected increase in RSA during GMLT performance. RSA is a composite of integrated respiratory and cardiovascular responses [47] that are responsive not only to metabolic demands but also to levels of alertness and, in humans at least, different types of emotion, mental activity and arousal metabolism [48]. Typically, expression of RSA decreases with age [49]; however, adults in excellent cardiovascular health, such as athletes, are likely to have a higher RSA [50]. Higher RSA is also usually associated with fewer errors in a complex memory task, on measures of attentional control and response selection and on cognitive function tasks that are associated with the integrity of the ACC [51]. Aβ+ participants showed decreased RSA compared to Aβ–during the cognitive test, and this pattern was present in participants with cognition impairments or with diabetes and cardiovascular disease [10]. Cardiac autonomic dysfunction is prevalent among individuals with AD, as reflected by exacerbated sympathetic nervous activity and decreased parasympathetic nervous activity [31 , 52–54]. The results of the current study suggest that for individuals identified as falling within the preclinical stage of AD (our Aβ+ group), there is a muted response on two related indices of phasic vagal cardiac control, RSA and vagal ratio, when presented with a modest cognitive stressor. Both RSA and vagal ratio are directly modulated by both muscarinic and nicotinic cholinergic autonomic neurotransmission, and the earliest stages of AD are marked, in part, by altered function of the basal forebrain cholinergic system, with eventual degenerative changes including neuronal loss [54]. We have recently reported that a downregulation of central cholinergic neurotransmission appears to be one of the earliest neuropathological changes in preclinical AD [14] and we have also found that individuals with evidence of both decreased central cholinergic tone and amyloid aggregation within the ACC region show evidence of increased resting cardiac workload [8]. The aggregation of Aβ plaques in the neocortex, within this specific region of interest that is part of the central cholinergic system, appears to be directly associated with higher cognitive impairment as well as myocardial oxygen consumption [8]. These results add to a growing body of literature that suggests a direct link between Aβ aggregation, basal forebrain cholinergic damage, and impaired cardiovascular function, even in the preclinical stage ofAD [1].

Footnotes

ACKNOWLEDGMENTS

This article was first-authored by CYS in partial fulfillment of her Ph.D. dissertation project. These results were first presented as a poster presentation at the 2016 annual meeting of the American Neurological Association (Baltimore, MD, October 2016) published on Annals of Neurology 141st Annual Meeting American Neurological Association; Abstract M122: S156, where it received a poster presentation award from the ANA.

Authors’ disclosures available online ().

References

Santos

, Snyder

, Wu

, Zhang

, Echeverria

, Alber

(2017) Pathophysiologic relationship between Alzheimer’s disease, cerebrovascular disease and cardiovascular risk: A review and synthesis. Alzheimers Dement (Amst) 7, 69–87.

Snowdon

, Greiner

, Mortimer

, Riley

, Greiner

, Markesberry

(1997) Brain infarction and clinical expression of Alzheimer’s disease. The Nun Study. JAMA 277, 813–817.

Esiri

, Nagy

, Smith

, Barnetson

, Smith

(1999) Cerebrovascular disease and threshold for dementia in the early stages of Alzheimer’s disease. Lancet 354, 919–920.

Helzner

, Luchsinger

, Scarmeas

, Cosentino

, Brickman

, Glymour

, Stern

(2009) Contribution of vascular risk factors to the progression in Alzheimer disease. Arch Neurol 66, 343–348.

Silvestrini

, Pasqualetti

, Baruffaldi

, Bartolini

, Handouk

, Matteis

, Moffa

, Provinciali

, Vernieri

(2006) Cerebrovascular reactivity and cognitive decline in patients with Alzheimer disease. Stroke 37, 1010–1015.

Zlokovic

(2011) Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nat Rev Neurosci 12, 723–738.

Iadecola

(2004) Neurovascular regulation in the normal brain and in Alzheimer’s disease. Nat Rev Neurosci 5, 347–360.

Santos

, Lim

, Wu

, Polynice

, Schindler

, Maruff

, Snyder

(2016) Resting-State Cardiac Workload is related to both increased neocortical aggregation of β-Amyloid and relative impairment in spatial working memory in Pre-clinical Alzheimer’s disease. J Alzheimers Dis 50, 127–131.

Fukusaki

, Kawakubo

, Yamamoto

(2000) Assessment of the primary effect of aging on heart rate variability in humans. Clin Auton Res 10, 123–130.

10.

Masi

, Hawkley

, Rickett

, Cacioppo

(2007) Respiratory sinus arrhythmia and diseases of aging: Obesity, diabetes mellitus, and hypertension. Biol Physiol 74, 212–223.

11.

Snyder

, Lim

, Schindler

, Ott

, Salloway

, Daiello

, Getter

, Gordon

, Maruff

(2014) Microdosing of scopolamine as a “cognitive stress test”: Rationale and test of a very low dose in an at-risk cohort of older adults. Alzheimers Dement 10, 262–267.

12.

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

, 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.

13.

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.

14.

Lim

, Maruff

, Schindler

, Ott

, Salloway

, Yoo

, Noto

, Santos

, Snyder

(2015) Disruption of cholinergic neurotransmission exarcerbates Aβ-related cognitive impairment in preclinical Alzheimer’s disease. Neurobiol Aging 36, 2709–2715.

15.

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.

16.

Rowe

, Bourgeat

, Ellis

, Brown

, Lim

, Mulligan

, Jones

, Maruff

, Woodward

, Price

, Robins

, Tochon-Danguy

, O’Keefe

, Pike

, Yates

, Szoeke

, Salvado

, Macaulay

, O’Meara

, Head

, Cobiac

, Savage

, Martins

, Masters

, Ames

, Villemagne

(2013) Predicting Alzheimer disease withβ-amyloid imaging: Results from the Australian Imaging, Biomarkers, and Lifestyle Study of Ageing. Ann of Neurol 74, 905–913.

17.

Petersen

, Aisen

, Beckett

, Donohue

, Gamst

, Harvey

, Jack

Jr , Jagust

, Shaw

, Toga

, Trojanowski

, Weiner

(2011) Alzheimer’s Disease Neuroimaging Initiative (ADNI): Clinical characterization. Neurology 74, 201–209.

18.

Corkin

(1981) Acetylcholine, aging and Alzheimer’s disease: Implications for treatment. Trends Neurosci 4, 287–290.

19.

Brown

, Chisholm

, Owens

, Pimlott

, Patterson

, Wyper

(2003) Acetylcholine muscarinic receptors and response to anti-cholinesterase therapy in patients with Alzheimer’s disease. Eur J Nucl Med Mol Imaging 30, 296–300.

20.

Rosenberg

, Wong

, Edell

, Ross

, Joshi

, Brašić

, Zhou

, Raymont

, Kumar

, Ravert

, Dannals

, Pontecorvo

, Skovronsky

, Lyketsos

(2013) Cognition and amyloid loadin Alzheimer disease imaged with florbetapir F 18(AV-45) positronemission tomography. Am J Geriatr Psychiatry 21, 272–278.

21.

Saint-Aubert

, Barbeau

, Péran

, Nemmi

, Vervueren

, Mirabel

, Payoux

, Hitzel

, Bonneville

, Gramada

, Tafani

, Vincent

, Puel

, Dechaumont

, Chollet

, Pariente

(2013) Cortical florbetapir-PET amyloid load in prodromal Alzheimer’sdisease patients. Eur J Nucl Med Mol Imaging 3, 43.

22.

Clark

, Schneider

, Bedell

, Beach

, Bilker

, Mintun

, Pontecorvo

, Hefti

, Carpenter

, Flitter

, Krautkramer

, Kung

, Coleman

, Doraiswamy

, Fleisher

, Sabbagh

, Sadowsky

, Reiman

, Zehntner

, Skovronsky

, AV45-A07 Study, Group (2011) Use of florbetapir-PET for imaging beta-amyloid pathology. JAMA 305, 275–283.

23.

Goleta

(2014) BioPac Inc. AcqKnowledge^® 4.4.2Software Guide, For Life Science Research Applications DataAcquisition and Analysis with BIOPAC Hardware Systems, pp. 389.

24.

(1996) Heart rate variability: Standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation 93, 1043–1065.

25.

(2016) Application Note 246: R-R Interval Processing Using BIOPAC’s HRV Algorithm Implementation, Appendix F, BioPac Inc., Goleta, CA, pp. 16–17.

26.

Grossman

, van Beek

, Wientjes

(1990) A comparison of three quantification methods for estimation of respiratory sinus arrhythmia. Psychophysiology 27, 702–714.

27.

Pietrzak

, Cohen

, Snyder

(2007) Spatial learning efficiency and error monitoring in normal aging: An investigation using a novel hidden maze learning test. Arch Clin Neuropsychol 22, 235–245.

28.

Lim

, Prang

, Cysique

, Pietrzak

, Snyder

, Maruff

(2009) A method for cross-cultural adaptation of a verbal memory assessment. Behav Res Methods 41, 1190–1200.

29.

Liang

K-Y

, Zeger

(1986) Longitudinal data analysis using generalized linear models. Biometrika 73, 13–22.

30.

White

(1982) Maximum likelihood estimation of misspecified models. Econometrica 50, 1–25.

31.

Toledo

MAV

, Junqueira

Jr (2008) Cardiac sympathovagal modulation evaluated by short-term heart interval variability is subtly impaired in Alzheimer’ disease. Geriatr Gerontol Int 8, 109–118.

32.

Toledo

MAV

, Junqueira

(2010) Cardiac autonomic modulation and cognitive status in Alzheimer’s disease. Clin Auton Res 20, 11–17.

33.

Collins

, Dillon

, Finucane

, Lawlor

, Kenny

(2012) Parasympathetic autonomic dysfunction is common in mild cognitive impairment. Neurobiol Aging 33, 2324–2333.

34.

Ravenswaaij-Arts

CMA

, Kollee

LAA

, Hopman

JCW

, Stoelinga

GBA

, Geijn

(1993) Heart Rate Variability. Ann Intern Med 118, 436–447.

35.

Pagani

, Lombardi

, Guzzetti

, Rimoldi

, Furlan

, Pizzinelli

, Sandrone

, Malfatto

, Dell’Orto

, Piccaluga

(1986) Powerspectral analysis of heart rate and arterial pressurevariabilities as a marker of sympatho-vagal interactions in manand conscious dog. Circ Res 59, 178–193.

36.

Pagani

, Lombardi

, Guzzetti

, Sandrone

, Rimoldi

, Malfatto

, Cerutti

, Malliani

(1984) Power spectral density of heartrate variability as an index of symptho-vagal interactions innormal and hypertensive subjects. J Hypertens 2, 383–385.

37.

Malliani

, Pagani

, Lombardi

, Cerutti

(1991) Cardiovascular neural regulation explored in the frequency domain. Circulation 84, 482–492.

38.

Hjortskov

, Rissen

, Blangsted

, Fallentin

, Lundberg

, Søgaard

(2004) The effect of mental stress on heart rate variability and blood pressure during computer work. Eur J Appl Physiol 92, 84–89.

39.

Moses

, Luecken

, Eason

(2007) Measuring task-related changes in heart rate variability. Conf Proc IEEE Eng Med Biol Soc 2007, 644–647.

40.

Milicevic

(2005) Low to high frequency ratio of heart rate variability spectra fails to describe sympatho-vagal balance in cardiac patients. Coll Antropol 29, 295.

41.

Billman

(2013) The LF/HF ratio does not accurately measurecardiac sympatho-vagal balance. Front Physiol 4, 26.

42.

Mathewson

, Dywan

, Snyder

, Tays

, Segalowitz

(2011) Autonomica regulation and maze-learning performance in older andyounger adults. Biol Psychol 88, 20–27.

43.

Sloan

, Bagiella

, Shapiro

, Kuhl

, Chernikhova

, Berg

, Myers

(2001) Hostility, gender, and cardiac autonomic control. Psychosom Med 63, 434–440.

44.

Thayer

, Fischer

(2009) Heart rate variability, overnighturinary norepinephrine and C-reactive protein: Evidence for thecholinergic anti-inflammatoryathway in healthy human adults. J Intern Med 265, 439–447.

45.

Mezzacappa

, Kelsey

, Katkin

, Sloan

(2001) Vagal rebound and recovery from psychological stress. Psychosom Med 63, 650–657.

46.

Hansen

, Johnsen

, Sollers

III , Stenvik

, Thayer

(2004) Heart rate variability and its relation to prefrontalcognitive function: The effects of training and detraining. Eur J Appl Physiol 93, 263–272.

47.

Lane

, Adcock

, Burnett

(1992) Respiratory sinus arrhythmia and cardiovascular responses to stress. Psychophysiology 29, 461–470.

48.

Grossman

, Taylor

(2007) Toward understanding respiratory sinus arrhythmia: Relations to cardiac vagal tone, evolution and biobehavioral functions. Biol Physiol 74, 263–285.

49.

Hrushesky

, Fader

, Schmitt

, Gilbertsen

(1984) The respiratory sinus arrhythmia: A measure of cardiac age. Science 224, 1001–1004.

50.

Hirsch

, Bishop

(1981) Respiratory sinus arrhythmia in humans: How breathing pattern modulates heart rate. Am J Physiol 241, H620–H629.

51.

Dywan

, Mathewson

, Choma

, Rosenfeld

, Segalowitz

(2008) Autonomic and electrophysiologicalcorrelates of emotional intensity in older and younger adults. Psychophysiology 45, 389–397.

52.

Giubilei

, Strano

, Imbimbo

, Tisei

, Calcagnini

, Lino

, Frontoni

, Santini

, Fieschi

(1998) Cardiac autonomic dysfunction in patients with Alzheimer’s disease: Possible pathogenetic mechanisms. Alzheimer Dis Assoc Disord 12, 356–361.

53.

Allan

(2007) Orthostatic hypotension and autonomic dysfunction. In Parkinson’s Dementia, Ballard

, ed. Carma Publishing, Delray Beach, FL, pp. 99–110.

54.

Zulli

(2008) Increased prevalence of silent myocardial ischaemia and severe ventricular arrhythmias in untreated patients with Alzheimer’s disease and mild cognitive imparment without overt coronary artery disease. Clin Neurol Neurosurg 110, 791–796.

55.

Strong

, Huang

, Chung

, Hale

, Burke

(1991) Degeneration of the cholinergic innervation of the locus coerelusin Alzheimer’s disease. Brain Res 542, 23–28.