Abstract
Vascular risk factors (VRFs) have been associated with clinically diagnosed Alzheimer disease (AD), but few studies have examined the association between VRF and AD neuropathology (ADNP) in cognitively normal individuals. We used longitudinal data from the National Alzheimer’s Disease Center’s Uniform Data Set and Neuropathology Data Set to examine the association between VRF and ADNP (moderate to frequent neuritic plaques; Braak stage III–VI) in those with normal cognition. Our sample included 53 participants with ADNP and 140 without ADNP. Body mass index (BMI), resting heart rate (HR), and pulse pressure (PP) were measured at each visit; values were averaged across participant visits and examined annual change in BMI, PP, and HR. Hypertension, diabetes, and hypercholesterolemia were self-reported. In the multivariable logistic regression analyses, average BMI and HR were associated with lower odds of ADNP, and annual increases in HR and BMI were associated with higher odds of ADNP. A previously experienced decline in BMI or HR in late-life (therefore, currently low BMI and low HR) as well as a late-life increase in BMI and HR may indicate underlying AD pathology. Additional clinicopathological research is needed to elucidate the role of changes in late-life VRF and AD pathogenesis.
Keywords: Alzheimer disease, Asymptomatic, Body mass index, Heart rate, Neuropathology, Normal cognition, Vascular risk factor.
INTRODUCTION
Neuropathological examination remains the gold standard for the diagnosis of Alzheimer disease (AD) but AD pathophysiologic processes begin many years before symptom onset, a period known as preclinical AD (1). Depending on age and genetic status, 20%–40% of older adults may exhibit amyloid burden at death despite antemortem normal cognition (2, 3). However, amyloid plaque accumulation continues with increasing age and may be present in up to 100% of the oldest old (ie centenarians) (4, 5). The preclinical phase of the AD continuum represents a critical opportunity for therapeutic intervention; however, robust methods to detect AD-related pathophysiological changes during life must first be established. Cerebrospinal fluid (CSF) protein assays and neuroimaging (eg positron emission tomography [PET]) are currently the gold standards for the detection of AD neuropathology during life (1, 6). The high expense and invasive nature of CSF assays and neuroimaging limit their utility as frontline detection methods, however, resulting in a need for non-invasive and cost-efficient preclinical AD biomarkers (7).
Vascular risk factors (VRFs) may be non-specific biomarkers of AD that can be combined with more specific biomarkers and genotypes to aid in predicting AD dementia risk. The epidemiological literature shows an association between VRF and clinically diagnosed incident AD (8–11). VRF such as hypertension, type 2 diabetes mellitus (T2DM), obesity, pulse pressure (PP), and resting heart rate (HR) may be particularly useful for identifying individuals in the preclinical stages of AD. For example, hypertension, insulin resistance, and PP, have been linked with greater PET- or CSF-measured AD pathology (eg phosphorylated tau, amyloid-β [Aβ] burden) in cognitively healthy older adults (12–15). However, inferences regarding the role of VRF in the development of clinical AD are limited due to the unknown etiology of the dementing illness, which could include AD, cerebrovascular disease, other neurodegenerative diseases, or a combination (16).
Prospective studies with postmortem neuropathological validation are essential for understanding the role of VRF in the pathogenesis of AD neuropathology (ADNP). Although limited, some ex vivo evidence suggests that VRF (eg hypertension, obesity, T2DM) may contribute to ADNP, possibly due to their association with cerebrovascular disease (17–21). Cerebrovascular disease may contribute to blood-brain barrier dysfunction and paravascular drainage disruptions that promote Aβ deposition and neurodegeneration and interfere with Aβ clearance (22, 23). Nonetheless, the role of VRF as valid biomarkers of preclinical AD remains unclear due to the lack of studies, if any, that have examined VRF and ADNP in an autopsy sample of subjects with normal cognition prior to death. More broadly, the existing literature relies on static assessments of VRF, despite the sensitivity of long-term fluctuations in vascular function with brain pathology (24).
The objectives of the current study were to examine the cross-sectional and longitudinal relationship between several common VRF and markers of vascular function and ADNP in older adults with antemortem normal cognition from the National Alzheimer’s Coordinating Center (NACC) Uniform Data Set (UDS), and Neuropathology Data Set (NDS). This study targeted older adults with normal cognition and excluded those with mild cognitive impairment (MCI) or dementia in order to determine whether a distinct relationship between VRF and ADNP exists during the preclinical phase of AD.
MATERIALS AND METHODS
Participants
Longitudinal, observational study data from the NACC UDS and NDS were used to study participants at 28 past and present US Alzheimer’s Disease Centers (ADC) funded by the National Institute on Aging. ADCs have collected annual demographic, clinical, diagnostic, and neuropsychological data on UDS participants with normal cognition, MCI, and dementia since 2005. UDS participants are recruited from population-based samples, clinics, public recruitment efforts, participant referrals, and other ongoing studies. Because recruitment methods vary, UDS participants are best described as a clinical case series of patients from each ADC. Additional details about the UDS sample are found elsewhere (25). Data collected between September 2005 and December 2015 were included in this study.
Our sample was restricted to those who had a clinical diagnosis of normal cognition at their last UDS visit before death and underwent neuropathological examination, with data entered into the NACC NDS. We included participants with at least 1 UDS visit within 3 years of their death, and excluded those with no ADNP data or no data on blood pressure, height and weight, and resting HR.
Vascular Risk Factors
Height, weight, systolic (SBP) and diastolic blood pressure (DBP), and resting HR were recorded at each UDS visit. Body mass index (BMI) (kg/m2) was calculated using height and weight data; PP was calculated by subtracting DBP from SBP. We examined PP instead of SBP or DBP because elevated PP in late-life has been associated with AD in previous studies (10, 13) and is expected to rise incrementally with age (26). In contrast, evidence suggests that mid-life but not late-life high blood pressure is a risk factor for dementia (27). Moreover, the NACC data do not include mid-life blood pressure measures. For each participant, we calculated the average of these measures across their UDS visits, in order to provide a better representation of the subjects’ average BMI in late-life rather than a single measure, which can be easily influenced by age-related illness or terminal decline. In addition, at each visit, participants and their co-participants reported history (recent or remote) of or taking medications for hypertension, diabetes, and hypercholesterolemia, and we created dichotomous variables for each condition (eg yes: any past or present history of hypertension; no: never reported hypertension).
Neuropathology
Neuropathology exam data (NP) are collected using a standardized Neuropathology Form and Coding Guidebook on a subset of UDS participants who died and previously consented to autopsy. Alzheimer disease neuropathologic change (ADNC), according to the current National Institute of Aging-Alzheimer’s Association criteria, was added to NACC’s Neuropathology Form in January 2014. Due to the small sample size with ADNC data collected to date, we created a surrogate measure, ADNP, which was defined as the presence of moderate to frequent neuritic plaques and Braak stages III–VI. The absence of ADNP was defined as none to sparse neuritic plaques and Braak stages 0–II. Additionally, we used NP data to describe the presence of cerebrovascular disease (large artery infarct or lacune, hemorrhage or microhemorrhage, cortical microinfarct[s], moderate to severe cerebral amyloid angiopathy, moderate to severe atherosclerosis, moderate to severe arteriolosclerosis, hippocampal sclerosis, Lewy body disease, and frontotemporal lobar degeneration).
Demographic, Clinical, and Genetic Variables
UDS participants were evaluated approximately annually using a standardized clinical exam and data on demographics, health history, and medications were ascertained. For this study, the demographic and clinical characteristics included age at initial visit and at death (years), sex, education (years), non-white race, Clinical Dementia Rating sum of boxes score at last visit before death, and number of UDS visits completed. Race was self-reported by the participants and a dichotomous race variable (non-white or white) was created due to the small sample of non-white participants. Race was included in this study to describe the sample, but couldn’t be included in the analysis due to the small sample of non-white participants. Additionally, we described the sample by any self-reported history (recent or remote) of cardiovascular disease, thyroid disease, atrial fibrillation, or smoking. History of stroke was determined by any self-reported history, clinical diagnosis, or imaging evidence. We also assessed whether participants reported past or present use of anti-lipid or anti-hypertensive medications, and described the number of subjects with 1 or more APOE e4 alleles (12% were missing APOE genotype data).
Standard Protocol Approvals, Registrations, and Patient Consents
Research using the NACC database was approved by the University of Washington Institutional Review Board. Informed consent was obtained at the individual ADCs. The NACC data were de-identified.
Statistical Analysis
Descriptive statistics (mean and SD, frequency, and percent) were used to describe the sample demographic, clinical, and neuropathological characteristics, stratified by the presence or absence of ADNP (ADNP-positive or ADNP-negative). Differences were tested using Pearson chi-squared test (categorical variables) or the t-test or Wilcoxon-Mann-Whitney test (normal and non-normal continuous variables, respectively). The Fisher exact test was used when at least 1 comparison group included fewer than 5 participants.
Unadjusted and adjusted logistic regression models with generalized estimating equations examined the association between the VRF (BMI, PP, HR, history of hypertension, diabetes, or hypercholesterolemia) and ADNP. Generalized estimating equations accounted for clustering of the data and the first adjusted model controlled for age at death, sex, and education. The second adjusted model additionally controlled for the presence of at least 1 APOE e4 allele. Annual change in BMI, PP, and HR was assessed by creating interaction terms between these measures and time since the initial visit in years, and each interaction term was included in the multivariable model, first controlling for age at death, sex, and education, and then additionally controlling for presence of at least 1 APOE e4 allele. As a sensitivity analysis, all multivariable models were re-run additionally adjusting for history of stroke, smoking, atrial fibrillation, thyroid disease, and cardiovascular disease, with no significant change in our findings (data not shown).
RESULTS
The sample consisted of a total of 193 subjects, 53 ADNP-positive participants and 140 ADNP-negative participants (Fig. 1). When compared with the ADNP-negative participants, the ADNP-positive participants were older at death and more often had ≥1 APOE e4 allele (33% in ADNP-positive group; 7% in ADNP-negative group) (Table 1). Approximately 44% of the total sample was male, a small minority (8%) was non-white, and Clinical Dementia Rating sum of boxes scores were worse in the ADNP-positive than ADNP-negative group at the last visit before death. The mean number of visits completed was 3.7 visits for the ADNP-positive participants and 4.2 visits for the ADNP-negative participants.
FIGURE 1.
Sample size flow chart. AD, Alzheimer disease; ADNP, Alzheimer disease neuropathology; BMI, body mass index; HR, resting heart rate; PP, pulse pressure.
TABLE 1.
Baseline Characteristics of the Sample
| Characteristics | Total | ADNP+ | ADNP− | p Value |
|---|---|---|---|---|
| Number of subjects, n | 193 | 53 | 140 | NA |
| Age at initial visit (years), mean (SD) | 80.2 (8.5) | 83.5 (6.5) | 79.0 (8.9) | 0.001 |
| Age at death (years), mean (SD) | 84.6 (8.7) | 87.6 (6.6) | 83.5 (9.1) | 0.008 |
| Education (years)a, mean (SD) | 15.6 (2.6) | 15.5 (2.7) | 15.7 (2.6) | 0.58 |
| Male, n (%) | 84 (43.5%) | 23 (43.4%) | 61 (43.6%) | 0.99 |
| Non-white racea, n (%) | 16 (8.4%) | 1 (1.9%) | 15 (10.8%) | 0.07 |
| CDR-SB at last visit, mean (SD) | 0.12 (0.40) | 0.25 (0.62) | 0.07 (0.27) | 0.02 |
| ≥1 APOEe4 allelea, n (%) | 25 (14.0%) | 16 (32.7%) | 9 (6.9%) | <0.0001 |
| Number of UDS visits, mean (SD) | 4.1 (2.2) | 3.7 (2.0) | 4.2 (2.2) | 0.15 |
ADNP, Alzheimer disease neuropathology; CDR-SB, Clinical Dementia Rating Sum of Boxes; APOE, Apolipoprotein E; UDS, Uniform Data Set.
aNumber missing data: race (ADNP+, n = 1; ADNP−, n = 1); education (ADNP+, n = 1; ADNP-, n = 0); APOE genotype (ADNP+, n = 4; ADNP−, n = 10).
The ADNP-positive group had a significantly lower average BMI, lower average HR, and less cardiovascular disease compared with the ADNP-negative group (Table 2). BMI was lower in the ADNP-positive group compared with the ADNP-negative group regardless of the time point examined (initial visit, last visit, average over visits). The 2 groups did not differ in their average values of SBP or DBP over visits, and did not differ on use of anti-hypertensive (eg beta blockers) or anti-lipid medications except that the ADNP-positive group less often took diuretics. Microinfarcts and amyloid angiopathy were more common in the ADNP-positive group than the ADNP-negative group, with no other differences in neuropathology between the 2 groups (Table 3). The average time between the last clinical visit and autopsy was 12 months for the ADNP-positive group and 11 months for the ADNP-negative group.
TABLE 2.
Clinical Characteristics of the Sample
| Characteristicsa | ADNP+ | ADNP− | p Value |
|---|---|---|---|
| At initial visit, mean (SD) | |||
| BMI (kg/m2) | 24.4 (4.1) | 27.3 (5.7) | 0.001 |
| PP (mmHg) | 66.2 (18.4) | 65.0 (17.1) | 0.68 |
| HR (bpm) | 68.6 (9.3) | 71.1 (12.2) | 0.18 |
| At visit closest to deathb, mean (SD) | |||
| BMI (kg/m2) | 23.9 (5.3) | 26.4 (6.7) | 0.03 |
| PP (mmHg) | 61.6 (13.8) | 58.6 (15.7) | 0.24 |
| HR (bpm) | 71.6 (11.6) | 73.4 (13.5) | 0.46 |
| Average over visitsc, mean (SD) | |||
| BMI (kg/m2) | 23.9 (4.2) | 26.8 (5.8) | 0.001 |
| PP (mmHg) | 65.5 (14.7) | 62.0 (12.9) | 0.11 |
| HR (bpm) | 69.0 (7.8) | 72.3 (10.7) | 0.04 |
| SBP (mmHg) | 136.1 (18.6) | 133.7 (16.9) | 0.41 |
| DBP (mmHg) | 70.6 (9.1) | 71.8 (9.2) | 0.42 |
| Hypertensiond, n (%) | 46 (86.8%) | 118 (84.3%) | 0.66 |
| Diabetesd, n (%) | 4 (7.6%) | 26 (18.6%) | 0.07 |
| Hypercholesterolemiad, n (%) | 30 (56.6%) | 96 (68.6%) | 0.12 |
| Strokee, n (%) | 4 (7.6%) | 14 (10.0%) | 0.78 |
| Cardiovascular diseasef, n (%) | 20 (37.7%) | 76 (54.3%) | 0.04 |
| Thyroid diseasef, n (%) | 13 (24.5%) | 45 (32.4%) | 0.29 |
| Atrial fibrillationf, n (%) | 10 (18.9%) | 35 (25.2%) | 0.36 |
| Any smoking history, n (%) | 24 (45.3%) | 79 (56.4%) | 0.17 |
| Medication useg, n (%) | |||
| Beta blocker | 32 (60.4%) | 71 (50.7%) | 0.23 |
| Diuretic | 17 (32.1%) | 70 (50.0%) | 0.03 |
| Calcium channel blocker | 17 (32.1%) | 43 (30.7%) | 0.86 |
| Ace inhibitor | 19 (35.9%) | 56 (40.0%) | 0.60 |
| Angiotensin II inhibitor | 9 (17.0%) | 31 (22.1%) | 0.43 |
| Vasodilator | 2 (3.8%) | 11 (7.9%) | 0.52 |
| Anti-lipid | 26 (49.1%) | 75 (53.6%) | 0.58 |
ADNP, Alzheimer disease neuropathology (−, negative; +, positive); bpm = beats per minute.
aMissing data: atrial fibrillation (ADNP+, n = 0; ADNP−, n = 1); thyroid disease (ADNP+, n = 0; ADNP−, n = 1).
bAmong subjects with at least 2 visits.
cIndividual-level values were averaged over visits.
dSelf-reported as active/remote or took medication for the condition at any point.
eAny self-reported history, clinical diagnosis, or imaging evidence of stroke.
fReported as active/remote at any UDS follow-up.
gReported using medication at any visit; a subject could be counted in ≥ 1 drug category.
TABLE 3.
Neuropathology at Autopsy
| Characteristica | ADNP+ | ADNP− | p Value |
|---|---|---|---|
| Months between last visit and autopsy, mean (SD) | 12.0 (7.7) | 11.3 (7.7) | 0.51 |
| Cerebrovascular disease, n (%) | |||
| Infarct/lacune(s) | 15 (28.3%) | 38 (27.3%) | 0.89 |
| Hemorrhage/microhemorrhage(s) | 3 (5.7%) | 11 (7.9%) | 0.76 |
| Microinfarct(s) | 20 (37.7%) | 28 (20.0%) | 0.01 |
| Cerebral amyloid angiopathy | 16 (31.4%) | 10 (7.5%) | <0.001 |
| Arteriolosclerosis | 32 (82.1%) | 97 (77.6%) | 0.66 |
| Atherosclerosis | 45 (84.9%) | 116 (82.9%) | 0.73 |
| Hippocampal sclerosis, n (%) | 1 (1.9%) | 2 (1.5%) | 0.99 |
| Lewy body disease, n (%) | 10 (18.9%) | 19 (13.9%) | 0.39 |
| FTLD, n (%) | 2 (5.3%) | 10 (8.9%) | 0.73 |
ADNP, Alzheimer disease neuropathology; bpm, beats per minute; FTLD, frontotemporal lobar degeneration.
aMissing data: infarct/lacune (ADNP+: n = 0; ADNP−: n = 1); hemorrhage/microhemorrhage (ADNP+: n = 0; ADNP−: n = 1); amyloid angiopathy (ADNP+: n = 2; ADNP−: n = 2); arteriolosclerosis (ADNP+: n = 14; ADNP−: n = 15); hippocampal sclerosis (ADNP+: n = 0; ADNP−: n = 5); Lewy body disease (ADNP+: n = 0; ADNP−: n = 3); FTLD (ADNP+: n = 15; ADNP−: n = 27).
In the unadjusted analysis, higher average BMI, higher average HR, and history of diabetes were associated with a lower odd of ADNP, whereas average PP, history of hypertension, and history of hypercholesterolemia were not associated with ADNP (Table 4). In the adjusted models controlling for age at death, sex, and education, higher average BMI and higher average HR were associated with a lower odds of ADNP, and no other VRF were associated with ADNP (Table 5). The results were similar after additionally controlling for the presence of at least 1 APOE e4 allele (Table 5).
TABLE 4.
Unadjusted Association Between VRFs and AD Neuropathology
| Characteristic | Model | OR | 95% CI | p Value |
|---|---|---|---|---|
| Average over visitsa | ||||
| BMI (kg/m2) | 1 | 0.92 | 0.88–0.96 | <0.001 |
| PP (mmHg) | 2 | 1.01 | 0.99–1.03 | 0.17 |
| HR (bpm) | 3 | 0.97 | 0.95–1.00 | 0.04 |
| Hypertensionb | 4 | 1.14 | 0.54–2.41 | 0.74 |
| Diabetesb | 5 | 0.38 | 0.19–0.77 | 0.007 |
| Hypercholesterolemiab | 6 | 0.70 | 0.35–1.39 | 0.31 |
OR, odds ratio; CI, confidence interval; bpm, beats per minute.
aIndividual-level values were averaged over visits.
bReported active/inactive condition or taking medication for the condition at any UDS follow-up.
TABLE 5.
Adjusted Association Between VRFs and AD Neuropathology
| Characteristic |
Multivariable model 1a |
Multivariable model 2b |
||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p Value | OR | 95% CI | p Value | |
| Average over visits | ||||||
| BMI (kg/m2) | 0.92 | 0.88–0.97 | 0.001 | 0.94 | 0.89–1.00 | 0.03 |
| PP (mmHg) | 1.01 | 0.99–1.04 | 0.40 | 1.01 | 0.99–1.04 | 0.31 |
| HR (bpm) | 0.97 | 0.94–1.00 | 0.04 | 0.97 | 0.93–1.00 | 0.05 |
| Hypertensionc | 0.92 | 0.37–2.27 | 0.85 | 0.72 | 0.29–1.79 | 0.48 |
| Diabetesc | 0.51 | 0.21–1.23 | 0.13 | 0.62 | 0.31–1.26 | 0.19 |
| Hypercholesterolemiac | 0.75 | 0.34–1.68 | 0.48 | 0.80 | 0.41–1.57 | 0.52 |
OR, odds ratio; CI, confidence interval; bpm, beats per minute.
aModel includes BMI, PP, HR, history of hypertension, diabetes, and hypercholesterolemia, age at death, sex, education.
bSame covariates as Model 1 except additionally controlling presence of at least 1 APOE e4 allele.
cSubject reported active/inactive history or taking medications for this condition at any UDS visit.
In the multivariable models examining annual change in BMI, PP, and HR (Table 6), annual increases in HR and BMI, but not in PP, were associated with an increased odd of ADNP after controlling for age at death, sex, education, and presence of at least 1 APOE e4 allele. However, the magnitude of the association was very small. In these models, having a higher BMI or higher HR at the initial visit was associated with a lower odd of ADNP, and history of hypertension, diabetes, and hypercholesterolemia was not associated with ADNP (Table 6).
TABLE 6.
Annual Change in VRFs and AD Neuropathology
|
Multivariable model 1a,b |
Multivariable model 2b,c |
|||||
|---|---|---|---|---|---|---|
| Variable | OR | 95% CI | p Value | OR | 95% CI | p Value |
| BMI at initial visit (kg/m2) | 0.98 | 0.97–0.99 | 0.006 | 0.995 | 0.991–0.999 | 0.02 |
| PP at initial visit (mmHg) | 1.000 | 0.999–1.002 | 0.58 | 1.000 | 1.000–1.001 | 0.32 |
| HR at initial visit (bpm) | 0.996 | 0.994–0.999 | <0.001 | 0.999 | 0.999–1.000 | 0.002 |
| Hypertensiond | 1.24 | 0.44–3.49 | 0.69 | 0.94 | 0.33–2.67 | 0.91 |
| Diabetesc | 0.47 | 0.15–1.47 | 0.19 | 0.64 | 0.20–2.03 | 0.45 |
| Hypercholesterolemiad | 0.74 | 0.35–1.55 | 0.42 | 0.78 | 0.35–1.74 | 0.54 |
| Time from initial visit (years) | 0.91 | 0.86–0.95 | <0.001 | 0.98 | 0.97–0.99 | <0.001 |
| Annual change in BMI (BMI × time) | 1.001 | 1.000–1.003 | 0.01 | 1.000 | 1.000–1.001 | 0.008 |
| Annual change in PP (PP × time) | 1.0001 | 0.9998–1.0004 | 0.59 | 1.0000 | 0.9999–1.0001 | 0.82 |
| Annual change in HR (HR × time) | 1.001 | 1.000–1.001 | <0.001 | 1.0001 | 1.0001–1.0002 | 0.002 |
BMI, body mass index; bpm, beats per minute; OR, odds ratio; CI, confidence interval.
aModel includes BMI, PP, HR, history of hypertension, diabetes, and hypercholesterolemia, age at death, sex, education.
bResults are reported to the third or fourth decimal place as needed for clarity of interpretation.
cSame covariates as Model 1 except additionally controlling for presence of at least 1 APOE e4 allele.
dSubjects reported active/recent or remote/inactive history or taking medication for the condition at any UDS visit.
DISCUSSION
ADNP detection prior to symptom onset is critical for timely intervention of disease modifying therapies. Based on epidemiological associations, some have posited a role for VRF in the pathogenesis of AD (whether it be additive or mechanistic is unclear) and in vivo research has found VRF associated with CSF and PET markers of β-amyloid (12–15). However, few studies, if any, have investigated the relationship between VRF and asymptomatic AD using autopsy data. The current study examined VRF and ADNP using a relatively large autopsy sample of NACC participants who were cognitively normal before death. We found that higher average BMI in late-life predicted lower odds of ADNP but a rise in BMI over time predicted higher odds of ADNP. Higher average HR in late-life was associated with lower ADNP burden, whereas annual increases in resting HR predicted higher odds of ADNP. There were no associations with PP, history of diabetes, hypertension, or hypercholesterolemia.
Lower average BMI was associated with higher odds of ADNP in this autopsy sample of older adults with antemortem normal cognition. Previous studies have shown that higher midlife BMI increases risk for the clinical diagnosis of AD (28–30), but this effect is diminished or even reversed in late-life (29–33). A similar pattern is evident in the ex vivo literature. A higher midlife BMI has been associated with greater ADNP burden in an autopsy sample of 191 subjects with varying levels of cognition (18). Consistent with our findings, a previous study found that lower BMI proximate to death predicted higher ADNP burden in participants with and without dementia (34). Although higher BMI in late-life can confer risk for ADNP in the presence of cardiovascular disease, previous autopsy samples and our sample, in particular, included relatively healthy and normal weight older adults (according to standard BMI classification). As such, a higher BMI may serve as a marker of overall general health and survival, and may even be neuroprotective (35). For example, in a longitudinal population-based study of community-dwelling older adults, being underweight at age 65 predicted worse outcomes (defined by future and active life expectancy, years of healthy life) relative to normal weight individuals. In contrast, being overweight and obese at age 65 was not associated with worse outcomes compared to being normal weight and occasionally predicted better outcomes (36). Indeed, decreased BMI at older ages can reflect poorer health and is a typical sequela of AD that can occur up to 10 years before the onset of AD symptoms, possibly due to ADNP changes in brain structures (eg medial temporal lobe, olfactory bulb) that may modulate weight control, appetite, olfaction, and taste (37–40). Therefore, lower BMI in late-life may serve as an indicator of underlying ADNP changes occurring before cognition is affected.
Additionally, we found that an increase in late-life BMI over follow-up was associated with slightly higher odds of ADNP. A previous longitudinal study found increases in BMI predicted greater dementia risk (41); however, the current study is the first to examine the association between changes in late-life BMI over time and ADNP in those with antemortem normal cognition. The mechanisms underlying increases in BMI and ADNP are unclear, but may involve cardiovascular risk and related risk factors for ADNP that can accompany increases in BMI. Inflammation and hormones (eg leptin) that are independently associated with adiposity may also play a role. Notably, we also observed a trivial overall decline in BMI (on average) from the initial visit to the visit closest to death, and thus our findings may be capturing the subset of individuals with increasing BMI over time. These individuals may have been the least likely to have ADNP that disturbed brain regions regulating weight, as is typically observed in the setting of clinically diagnosed AD. Indeed, it is possible that there may be significant individual differences and subtypes in weight change in AD that is dependent on the pathological variants of the disease. Some AD patients may lose weight due to ADNP disruptions in brain regions that modulate weight control, whereas other patients may gain weight with the progression of ADNP. For example, there is a subgroup of those with ADNP who experience neuropathological and behavioral changes consistent with Klüver-Bucy syndrome, including hyperorality, which can be associated with increased eating and weight gain (42, 43). Longitudinal studies examining BMI across the adult lifespan are needed to improve understanding on the relationship between weight change and preclinical ADNP.
Higher average HR was associated with lower odds of ADNP; and annual increases in HR corresponded to greater ADNP burden. Resting HR has clinical utility in cardiovascular disease populations and is associated with cognitive decline in post-stroke patients and individuals with high cardiovascular risk (44, 45). Limited research has examined resting HR as it relates to aging and neurodegenerative disease although 1 study found higher resting HR to predict worse functional status in older adults independent of cardiovascular disease (46). Regardless, resting HR is easily influenced by an array of external (eg caffeine) and internal (eg anxiety) stimuli, limiting its reliability, particularly in cross-sectional assessments. In terms of the link between annual increases in resting HR and higher ADNP, the directionality of this relationship is unclear and it is possible that increased resting HR is secondary to a change in central and autonomic control over time that is influenced by brain regions damaged by age- or AD-related pathological changes. For example, the insular cortex is believed to play a role in the regulation of the cardiovascular system (47), and this brain region can be affected by ADNP before symptom onset (48, 49). Alternatively, increased HR may contribute to the development of ADNP. Increased HR can be a sign of cardiac problems, predicts incident cardiovascular events (50), and is not typically affected by aging. Therefore, some of the subjects in the sample may have experienced deterioration in overall cardiovascular health over time, potentially leading to cerebrovascular changes, a proposed core feature of AD (51). Increases in resting HR may also be capturing changes in HR variability (52), which may be sensitive to AD detection and severity (11). Overall, this is the first study to report an association between resting HR and ADNP over time but the pathophysiology underpinning this relationship is not known. Further research is needed to replicate our results and clarify the potential biological mechanism(s).
Late-life PP, diabetes, and hypercholesterolemia did not predict ADNP in this study. The older age of the current sample may again partially explain these findings because the presence of these factors during midlife appears to represent the critical risk window for developing AD, and this risk is attenuated or reversed in late-life possibly due to survival bias (21, 53–60) and the increased prevalence of these conditions with aging. When examined as a continuous variable, PP has previously been found to be unrelated to the clinical diagnosis of AD in a sample of ≥75 year olds, and other studies have also failed to find an association between antemortem PP and ADNP severity (10, 20). Interestingly, PP has been linked with in vivo CSF tau and Aβ biomarkers, but only among those < 80-years old, and not among those ≥80 years old (14). The pathological processes associated with brain aging in the ≥80-year olds may have attenuated the effects of high PP in predicting underlying ADNP. The early development of ADNP in brain vasomotor centers may also lower blood pressure to decrease risk for cardiovascular disease (61); this might translate to a null relationship with ADNP burden. Lastly, the presence of T2DM has also been shown to lack association with ADNP in neuropathologically confirmed cases of AD, but T2DM exacerbated the effects of APOE on ADNP (19). It is possible that the relationship between VRF and ADNP in this sample may also be moderated by factors such as genotype status, though the overall lack of participants with at least 1 APOE e4 allele precluded formal analytic test of this possibility.
There are limitations to the current findings. The assessment of hypertension, diabetes, and hypercholesterolemia using participant and informant self-report could be confounded by subjectivity, memory lapses, etc. More reliable assessments of VRF (eg insulin sensitivity) and examination of fluctuation (eg blood pressure variability) should be the target of future studies examining VRF and ADNP. BMI is a practical method for assessing adiposity, but it is a coarse index and lacks accuracy in older adults due to age-related losses in lean body mass and increased adiposity without weight gain (62). Future work that uses other anthropometric measures of adiposity (eg waist circumference) that are more reliable in older adult cohorts is needed to validate our findings, particularly in terms of the association between weight change and ADNP. The current sample consisted of normal weight, relatively healthy older adults who agreed to brain donation, representing a highly selective sample that limits the external validity of our findings. Postmortem studies are methodologically problematic because of selection bias (63) and survival bias. For example, the relationship between BMI and ADNP may be distinct in overweight/obese populations, due in part to a higher prevalence of cardiovascular disease and associated VRF (eg diabetes), but these groups may be less likely to be in our sample due to higher risk for premature mortality. Some research suggests that aggregation of VRF into a single measure may be more clinically meaningful in terms of risk for dementia (8). Additional studies are needed to determine the specific VRF associated with asymptomatic AD, which could be incorporated into an aggregate measure that more strongly predicts asymptomatic AD than the single measures. In addition, microinfarcts and amyloid angiopathy were more frequent in the ADNP-positive than ADNP-negative group. Future work by our group will use advanced statistical modeling to examine cerebrovascular disease as a potential mediator between VRF and ADNP. Finally, the current sample included participants with only normal cognition and excluded those with MCI and dementia in order to investigate whether previously found associations between VRF and ADNP extend to the preclinical stage of AD. As such, MCI and dementia participants were not the target of this research study, and the relationship between VRF and ADNP is likely distinct across the AD continuum due to variability in ADNP burden.
Late-life BMI and resting HR were associated with ADNP in this autopsy sample with antemortem normal cognition. Future clinicopathological research is needed to examine VRF across the lifespan, using robust measures of vascular function and longitudinal biomarkers of ADNP. These future studies will help determine whether epidemiological associations between VRF and AD are due to reverse causality, in which ADNP causes changes in the regions of the brain responsible for HR and weight, or whether VRF are directly responsible for the development of ADNP.
ACKNOWLEDGEMENTS
The authors thankfully acknowledge the patients and families enrolled at the ADCs who contributed data to the UDS and the faculty and staff of the ADCs who conducted the evaluations and collected the data used in these analyses. The authors would also like to thank the NIA, which provided support for the ADCs and NACC, as well as NACC staff (Duane Beekly, George Thomas, Mark Bollenbeck, Janene Hubbard, Mary Jacka, Joylee Wu, Elizabeth Robichaud, Nicole Barlow, Simone Wilk, Merilee Teylan, Kristen Schwabe-Fry, and Margaret Dean) who help in programming of the data submission systems, data management, research coordination, and administration, and without whom this research would not be possible.
Funding: The National Alzheimer’s Coordinating Center is funded by NIH U01 AG016976. NACC data are contributed by the NIA funded ADCs: P30 AG019610 (PI Eric Reiman, MD), P30 AG013846 (PI Neil Kowall, MD), P50 AG008702 (PI Scott Small, MD), P50 AG025688 (PI Allan Levey, MD, PhD), P30 AG010133 (PI Andrew Saykin, PsyD), P50 AG005146 (PI Marilyn Albert, PhD), P50 AG005134 (PI Bradley Hyman, MD, PhD), P50 AG016574 (PI Ronald Petersen, MD, PhD), P50 AG005138 (PI Mary Sano, PhD), P30 AG008051 (PI Steven Ferris, PhD), P30 AG013854 (PI M. Marsel Mesulam, MD), P30 AG008017 (PI Jeffrey Kaye, MD), P30 AG010161 (PI David Bennett, MD), P30 AG010129 (PI Charles DeCarli, MD), P50 AG016573 (PI Frank LaFerla, PhD), P50 AG016570 (PI David Teplow, PhD), P50 AG005131 (PI Douglas Galasko, MD), P50 AG023501 (PI Bruce Miller, MD), P30 AG035982 (PI Russell Swerdlow, MD), P30 AG028383 (PI Linda Van Eldik, PhD), P30 AG010124 (PI John Trojanowski, MD, PhD), P50 AG005133 (PI Oscar Lopez, MD), P50 AG005142 (PI Helena Chui, MD), P30 AG012300 (PI Roger Rosenberg, MD), P50 AG005136 (PI Thomas Montine, MD, PhD), P50 AG033514 (PI Sanjay Asthana, MD, FRCP), and P50 AG005681 (PI John Morris, MD).
Disclosures: Lilah Besser, Liliana Ramirez Gomez, Andrew Zhou, Ann C. McKee, and Walter Kukull report no disclosures. Michael L. Alosco is supported by an NIA post-doctoral fellowship under award number T32-AG06697. Robert A. Stern reports receiving research support from Avid Radiopharmaceuticals and serves as a paid consultant to Avanir Pharmaceuticals, Amarantus Bioscience, and Biogen. He also receives royalties from Psychological Assessment Resources, Inc., for published neuropsychological tests. John Gunstad is supported in part by a grant from the National Institutes of Health (R01AA022336). Helena Chui is supported by P50 AG05142. Julie Schneider is a consultant for Navidea biopharmaceuticals, and AVID Radiopharmaceuticals, and is supported by P30AG010161.
REFERENCES
- 1.Sperling RA, Aisen PS, Beckett LA, et al. 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 2011; 7:280–92 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Arriagada PV, Marzloff K, Hyman BT. Distribution of Alzheimer-type pathologic changes in nondemented elderly individuals matches the pattern in Alzheimer’s disease. Neurology 1992;42:1681–8 [DOI] [PubMed] [Google Scholar]
- 3.Morris JC, Storandt M, McKeel DW, et al. Cerebral amyloid deposition and diffuse plaques in “normal” aging: Evidence for presymptomatic and very mild Alzheimer’s disease. Neurology 1996;46:707–19 [DOI] [PubMed] [Google Scholar]
- 4.Delaère P, He Y, Fayet G, et al. Beta A4 deposits are constant in the brain of the oldest old: an immunocytochemical study of 20 French centenarians. Neurobiol Aging 1993;14:191–5 [DOI] [PubMed] [Google Scholar]
- 5.Hauw JJ, Vignolo P, Duyckaerts C, et al. Neuropathological study of 12 centenarians: the incidence of Alzheimer type senile dementia is not particularly increased in this group of very old patients. Rev Neurol (Paris) 1986;142:107–15 [PubMed] [Google Scholar]
- 6.Fiandaca MS, Mapstone ME, Cheema AK, et al. The critical need for defining preclinical biomarkers in Alzheimer’s disease. Alzheimers Dement 2014;10(3 Suppl):S196–212 [DOI] [PubMed] [Google Scholar]
- 7.Laske C, Sohrabi HR, Frost SM, et al. Innovative diagnostic tools for early detection of Alzheimer’s disease. Alzheimers Dement 2015;11:561–78 [DOI] [PubMed] [Google Scholar]
- 8.Luchsinger JA, Reitz C, Honig LS, et al. Aggregation of vascular risk factors and risk of incident Alzheimer disease. Neurology 2005;65:545–51 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Whitmer RA, Gunderson EP, Quesenberry CP, et al. Body mass index in midlife and risk of Alzheimer disease and vascular dementia. Curr Alzheimer Res 2007;4:103–9 [DOI] [PubMed] [Google Scholar]
- 10.Qiu C, Winblad B, Viitanen M, et al. Pulse pressure and risk of Alzheimer disease in persons aged 75 years and older: a community-based, longitudinal study. Stroke 2003;34:594–9 [DOI] [PubMed] [Google Scholar]
- 11.Zulli R, Nicosia F, Borroni B, et al. QT dispersion and heart rate variability abnormalities in Alzheimer’s disease and in mild cognitive impairment. J Am Geriatr Soc 2005;53:2135–9 [DOI] [PubMed] [Google Scholar]
- 12.Starks EJ, Patrick O’Grady J, Hoscheidt SM, et al. Insulin resistance is associated with higher cerebrospinal fluid tau levels in asymptomatic APOEε4 carriers. J Alzheimers Dis 2015;46:525–33 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Nation DA, Edland SD, Bondi MW, et al. Pulse pressure is associated with Alzheimer biomarkers in cognitively normal older adults. Neurology 2013;81:2024–7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Nation DA, Edmonds EC, Bangen KJ, et al. Pulse pressure in relation to tau-mediated neurodegeneration, cerebral amyloidosis, and progression to dementia in very old adults. JAMA Neurol 2015;72:546–53 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Reed BR, Marchant NL, Jagust WJ, et al. Coronary risk correlates with cerebral amyloid deposition. Neurobiol Aging 2012;33:1979–87 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Knopman DS, Roberts R. Vascular risk factors: imaging and neuropathologic correlates. J Alzheimers Dis 2010;20:699–709 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hoffman LB, Schmeidler J, Lesser GT, et al. Less Alzheimer disease neuropathology in medicated hypertensive than nonhypertensive persons. Neurology 2009;72:1720–6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chuang YF, An Y, Bilgel M, et al. Midlife adiposity predicts earlier onset of Alzheimer’s dementia, neuropathology and presymptomatic cerebral amyloid accumulation. Mol Psych 2016;21:910–5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Malek-Ahmadi M, Beach T, Obradov A, et al. Increased Alzheimer’s disease neuropathology is associated with type 2 diabetes and ApoE ϵ.4 carrier status. Curr Alzheimer Res 2013;10:654–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Nation DA, Delano-Wood L, Bangen KJ, et al. Antemortem pulse pressure elevation predicts cerebrovascular disease in autopsy-confirmed Alzheimer’s disease. J Alzheimers Dis 2012;30:595–603 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Petrovitch H, White LR, Izmirilian G, et al. Midlife blood pressure and neuritic plaques, neurofibrillary tangles, and brain weight at death: the HAAS. Honolulu-Asia aging Study. Neurobiol Aging 2000;21:57–62 [DOI] [PubMed] [Google Scholar]
- 22.Zlokovic BV. Neurovascular pathways to neurodegeneration in Alzheimer’s disease and other disorders. Nat Rev Neurosci 2011;12:723–38 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Weller RO, Boche D, Nicoll JA. Microvasculature changes and cerebral amyloid angiopathy in Alzheimer’s disease and their potential impact on therapy. Acta Neuropathol 2009;118:87–102 [DOI] [PubMed] [Google Scholar]
- 24.Brickman AM, Reitz C, Luchsinger JA, et al. Long-term blood pressure fluctuation and cerebrovascular disease in an elderly cohort. Arch Neurol 2010;67:564–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Morris JC, Weintraub S, Chui HC, et al. The Uniform Data Set (UDS): clinical and cognitive variables and descriptive data from Alzheimer Disease Centers. Alzheimer Dis Assoc Disord 2006;20:210–6 [DOI] [PubMed] [Google Scholar]
- 26.Franklin SS, Gustin W, Wong ND, et al. Hemodynamic patterns of age-related changes in blood pressure. The Framingham Heart Study. Circulation 1997;96:308–15 [DOI] [PubMed] [Google Scholar]
- 27.Qiu C. Preventing Alzheimer’s disease by targeting vascular risk factors: hope and gap. J Alzheimers Dis 2012;32:721–31 [DOI] [PubMed] [Google Scholar]
- 28.Whitmer RA, Gunderson EP, Barrett-Connor E, et al. Obesity in middle age and future risk of dementia: a 27 year longitudinal population based study. Bmj 2005;330:1360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Tolppanen AM, Ngandu T, Kåreholt I, et al. Midlife and late-life body mass index and late-life dementia: results from a prospective population-based cohort. J Alzheimers Dis 2014;38:201–9 [DOI] [PubMed] [Google Scholar]
- 30.Anstey KJ, Cherbuin N, Budge M, et al. Body mass index in midlife and late-life as a risk factor for dementia: a meta-analysis of prospective studies. Obes Rev 2011;12:e426–37 [DOI] [PubMed] [Google Scholar]
- 31.Hughes TF, Borenstein AR, Schofield E, et al. Association between late-life body mass index and dementia: The Kame Project. Neurology 2009;72:1741–6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Fitzpatrick AL, Kuller LH, Lopez OL, et al. Midlife and late-life obesity and the risk of dementia: cardiovascular health study. Arch Neurol 2009;66:336–42 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Barrett-Connor E, Edelstein SL, Corey-Bloom J, et al. Weight loss precedes dementia in community-dwelling older adults. J Amer Geriat Soc 1996;44:1147–52 [DOI] [PubMed] [Google Scholar]
- 34.Buchman AS, Schneider JA, Wilson RS, et al. Body mass index in older persons is associated with Alzheimer disease pathology. Neurology 2006;67:1949–54 [DOI] [PubMed] [Google Scholar]
- 35.Gustafson DR, Luchsinger JA. High adiposity: risk factor for dementia and Alzheimer’s disease? Alzheimers Res Ther 2013;5:57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Diehr P, O’Meara ES, Fitzpatrick A, et al. Weight, mortality, years of healthy life, and active life expectancy in older adults. J Am Geriatr Soc 2008;56:76–83 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Knopman DS, Edland SD, Cha RH, et al. Incident dementia in women is preceded by weight loss by at least a decade. Neurology 2007;69:739–46 [DOI] [PubMed] [Google Scholar]
- 38.Grundman M, Corey-Bloom J, Jernigan T, et al. Low body weight in Alzheimer’s disease is associated with mesial temporal cortex atrophy. Neurology 1996;46:1585–91 [DOI] [PubMed] [Google Scholar]
- 39.Buchman AS, Wilson RS, Bienias JL, et al. Change in body mass index and risk of incident Alzheimer disease. Neurology 2005;65:892–7 [DOI] [PubMed] [Google Scholar]
- 40.Ulrich J. Alzheimer changes in nondemented patients younger than sixty-five: possible early stages of Alzheimer’s disease and senile dementia of Alzheimer type. Ann Neurol 1985;17:273–7 [DOI] [PubMed] [Google Scholar]
- 41.Ye BS, Jang EY, Kim SY, et al. Unstable body mass index and progression to probable Alzheimer’s disease dementia in patients with amnestic mild cognitive impairment. J Alzheimers Dis 2015;49:483–91 [DOI] [PubMed] [Google Scholar]
- 42.Kile SJ, Ellis WG, Olichney JM, et al. Alzheimer abnormalities of the amygdala with Klüver-Bucy syndrome symptoms: an amygdaloid variant of Alzheimer disease. Arch Neurol 2009;66:125–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Förstl H, Burns A, Levy R, et al. Neuropathological correlates of behavioural disturbance in confirmed Alzheimer’s disease. Br J Psychiatry 1993;163:364–8 [DOI] [PubMed] [Google Scholar]
- 44.Böhm M, Cotton D, Foster L, et al. Impact of resting heart rate on mortality, disability and cognitive decline in patients after ischaemic stroke. Eur Heart J 2012;33:2804–12 [DOI] [PubMed] [Google Scholar]
- 45.Böhm M, Schumacher H, Leong D, et al. Systolic blood pressure variation and mean heart rate is associated with cognitive dysfunction in patients with high cardiovascular risk. Hypertension 2015;65:651–61 [DOI] [PubMed] [Google Scholar]
- 46.Ogliari G, Mahinrad S, Stott DJ, et al. Resting heart rate, heart rate variability and functional decline in old age. Cmaj 2015;187:E442–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Nagai M, Hoshide S, Kario K. The insular cortex and cardiovascular system: a new insight into the brain-heart axis. J Am Soc Hypertens 2010;4:174–82 [DOI] [PubMed] [Google Scholar]
- 48.Collins O, Dillon S, Finucane C, et al. Parasympathetic autonomic dysfunction is common in mild cognitive impairment. Neurobiol Aging 2012;33:2324–33 [DOI] [PubMed] [Google Scholar]
- 49.Royall DR. Insular Alzheimer disease pathology and the psychometric correlates of mortality. Cleve Clin J Med 2008;75(Suppl 2):S97–9 [DOI] [PubMed] [Google Scholar]
- 50.Fox K, Borer JS, Camm AJ, et al. Resting heart rate in cardiovascular disease. J Am Coll Cardiol 2007;50:823–30 [DOI] [PubMed] [Google Scholar]
- 51.Lee S, Viqar F, Zimmerman ME, et al. White matter hyperintensities are a core feature of Alzheimer’s disease: Evidence from the dominantly inherited Alzheimer network. Ann Neurol 2016;79:929–39 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Hart J. Association between heart rate variability and manual pulse rate. J Can Chiropr Assoc 2013;57:243–50 [PMC free article] [PubMed] [Google Scholar]
- 53.Launer LJ, Ross GW, Petrovitch H, et al. Midlife blood pressure and dementia: the Honolulu-Asia aging study. Neurobiol Aging 2000;21:49–55 [DOI] [PubMed] [Google Scholar]
- 54.Skoog I, Lernfelt B, Landahl S, et al. 15-year longitudinal study of blood pressure and dementia. Lancet 1996;347:1141–5 [DOI] [PubMed] [Google Scholar]
- 55.Mielke MM, Zandi PP, Shao H, et al. The 32-year relationship between cholesterol and dementia from midlife to late life. Neurology 2010;75:1888–95 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Meng XF, Yu JT, Wang HF, et al. Midlife vascular risk factors and the risk of Alzheimer’s disease: a systematic review and meta-analysis. J Alzheimers Dis 2014;42:1295–310 [DOI] [PubMed] [Google Scholar]
- 57.White H, Pieper C, Schmader K. The association of weight change in Alzheimer’s disease with severity of disease and mortality: a longitudinal analysis. J Amer Geriat Soc 1998;46:1223–7 [DOI] [PubMed] [Google Scholar]
- 58.Irina A, Seppo H, Arto M, et al. Beta-amyloid load is not influenced by the severity of cardiovascular disease in aged and demented patients. Stroke 1999;30:613–8 [DOI] [PubMed] [Google Scholar]
- 59.Rosano C, Newman AB. Cardiovascular disease and risk of Alzheimer’s disease. Neurol Res 2006;28:612–20 [DOI] [PubMed] [Google Scholar]
- 60.Bergmann C, Sano M. Cardiac risk factors and potential treatments in Alzheimer’s disease. Neurol Res 2006;28:595–604 [DOI] [PubMed] [Google Scholar]
- 61.Burke WJ, Coronado PG, Schmitt CA, et al. Blood pressure regulation in Alzheimer’s disease. J Auton Nerv Syst 1994;48:65–71 [DOI] [PubMed] [Google Scholar]
- 62.Stevens J, Cai J. Juhaeri, et al. Consequences of the use of different measures of effect to determine the impact of age on the association between obesity and mortality. Am J Epidemiol 1999;150:399–407 [DOI] [PubMed] [Google Scholar]
- 63.Chui HC, Zheng L, Reed BR, et al. Vascular risk factors and Alzheimer’s disease: are these risk factors for plaques and tangles or for concomitant vascular pathology that increases the likelihood of dementia? An evidence-based review. Alzheimers Res Ther 2012;4:1. [DOI] [PMC free article] [PubMed] [Google Scholar]