Association of Ideal Cardiovascular Health With Vascular Brain Injury and Incident Dementia
Background and Purpose—The American Heart Association developed the ideal cardiovascular health (CVH) index as a simple tool to promote CVH; yet, its association with brain atrophy and dementia remains unexamined.
Methods—Our aim was to investigate the prospective association of ideal CVH with vascular brain injury, including the 10-year risks of incident stroke and dementia, as well as cognitive decline and brain atrophy on magnetic resonance imaging, measured for ≈7 years. We studied 2750 stroke- and dementia-free Framingham Heart Study Offspring cohort participants (mean age, 62±9 years; 45% men). Ideal CVH was quantified on a 7-point scale with 1 point awarded for each of the following: nonsmoking status, ideal body mass index, regular physical activity, healthy diet, as well as optimum blood pressure, cholesterol, and fasting blood glucose. Both recent (baseline) and remote (6.9 years earlier) ideal CVH scores were examined.
Results—Recent ideal CVH was associated with stroke (hazard ratio, 0.80; 95% confidence interval, 0.67–0.95), vascular dementia (hazard ratio, 0.49; 95% confidence interval, 0.30–0.81), frontal brain atrophy (P=0.003), and cognitive decline on tasks measuring visual memory and reasoning (P<0.05). In addition to predicting stroke, vascular dementia, whole-brain atrophy, and cognitive decline, remote ideal CVH was associated with the incidence of all-cause dementia (hazard ratio, 0.80; 95% confidence interval, 0.67–0.97) and Alzheimer disease (hazard ratio, 0.79; 95% confidence interval, 0.64–0.98).
Conclusions—Adherence to the American Heart Association’s ideal CVH factors and behaviors, particularly in midlife, may protect against cerebrovascular disease and dementia.
The global prevalence of dementia is expected to double nearly every 20 years, reaching 115 million by 2050.1 In addition to the toll taken on individuals and families, the healthcare costs relating to dementia have the potential to devastate economies, with yearly costs estimated to exceed 1 trillion by 2050 in the United States alone.2 Even in the absence of overt dementia, cognitive impairment and cerebrovascular disease are common among the elderly.3,4 Considering the rapid increase in life expectancies, there is a pressing need to protect the brain from disease.
Modifiable vascular risk factors are established predictors of stroke and cerebrovascular disease.5 There is now growing consensus that vascular risk factors, such as physical inactivity, smoking, hypertension, and obesity, are also associated with the risk of cognitive decline and dementia, including its most common form Alzheimer disease (AD).6 Recent estimates suggest that a third of all AD cases might be because of modifiable risk factors.6 Therefore, further advocating for vascular health may, in turn, promote healthy brain aging, helping to ease the burden of AD and all forms of vascular cognitive impairment.
The American Heart Association (AHA) has proposed and advocated for the use of a simple ideal cardiovascular health (CVH) metric.7 The metric pools together 7 CVH behaviors and factors, assigning 1 point for each factor at its ideal status, with higher scores predictive of less heart disease8,9 and stroke.9,10 Given that the brain is vulnerable to damage in the face of vascular risk factors,11 adhering to these simple ideal CVH guidelines may confer protection against vascular brain injury and consequently cognitive decline and dementia. However, to our knowledge, the ideal CVH score is yet to be examined with respect to incident dementia. Developing a single set of guidelines to protect against heart disease, stroke, and dementia may have considerable public health implications. The aim of this study was to examine whether greater adherence to AHA ideal CVH guidelines was associated with a lower risk of incident stroke, incident dementia, cognitive decline, and brain atrophy among participants of the Framingham Heart Study Offspring cohort.
Materials and Methods
The community-based Framingham Heart Study Offspring cohort commenced in 1971 with the recruitment of 5124 individuals.12 Since then the Offspring cohort has been examined 9 times, with the latest examination cycle ending in 2014. The study was originally established to define risk factors for heart disease and stroke, but more recently, a major aim of the study has been to define risk factors for neurodegenerative disease.13
As depicted in the Figure, we created overlapping samples enabling investigation of the 10-year risk of incident stroke and dementia (including dementia subtype), beginning from examination cycle 7 (1998–2001) onward. For the analysis of incident stroke, we excluded participants under the age of 45 years, those with prevalent stroke at examination 7, and those with no follow-up for stroke leaving a sample of 2631. For the analysis of incident dementia, we excluded participants under the age of 60 years, those with prevalent dementia at examination 7, and those with no follow-up for dementia leaving 1364 participants. We also created samples to examine the association of ideal CVH with the progression of cognitive decline (n=1597), as well as brain atrophy and white-matter injury on magnetic resonance imaging (MRI) of the brain (n=1287). Changes in the neuropsychological and MRI outcomes were examined over ≈7 years from examination 7 onward. All participants provided written informed consent, and the study was approved by the Institutional Review Board and Boston University Medical Center.
Primary Exposure—The Ideal CVH Score
Ideal CVH scores were calculated according to AHA guidelines7 by summing together 7 dichotomous variables with 0 reflecting poor and 7 ideal CVH status. One point was awarded for each of the following: current self-reported nonsmoker, body mass indexes of <25 and >18.5 kg/m2, adequate physical activity, a healthy diet, untreated total cholesterol <200 mg/dL, untreated resting blood pressure <120/<80 mm Hg, and fasting blood glucose <100 mg/dL. All factors were obtained from the Framingham Heart Study clinic except diet and physical activity, which were measured with the use of a validated food frequency and physical activity index questionnaire, respectively. Blood pressures were measured twice by a physician, with the average systolic and diastolic values used in the ideal CVH score. Complete details of how ideal CVH scores were calculated, including quantification of diet and physical activity, are available in Table I in the online-only Data Supplement. Given that certain vascular risk factors have an age-dependent association with dementia and cognitive function,14 we calculated both recent (examination 7; 1998–2001) and remote (examination 5; 1991–1995) ideal CVH scores.
Assessment of Incident Stroke and Dementia
We examined whether recent and remote ideal CVH scores were associated with the 10-year risk of stroke, all-cause dementia, clinically characterized AD, and vascular dementia (VaD), beginning from examination 7. Follow-up occurred for a maximum of 10 years from the baseline examination to the time of incident event. Follow-up time for participants with no incident events was censored at the last time they were known to be event free, again through to a maximum of 10 years. There were 27 944 person-years for stroke and 11 386 person-years for dementia.
We assessed the incidence of stroke by monitoring hospital admissions in Framingham and by reviewing all available medical records and results. Stroke was defined as focal neurological symptoms of rapid onset and presumed vascular origin, lasting >24 hours or resulting in death within 24 hours. A diagnosis of dementia was made in accordance with the Diagnostic and Statistical Manual of Mental Disorders, 4th edition.15 The type of dementia was also recorded (details on the assessment of incident stroke and incident dementia are available in the Methods section in the online-only Data Supplement).
Assessment of Cognitive Decline
Cognitive decline was assessed as annualized change over a mean follow-up of 7 years, beginning from examination 7. We examined a battery of validated neuropsychological tests sensitive to vascular brain injury, as well as 1 verbal memory task, sensitive to deficits typically observed in AD.16 Our battery included Visual Reproductions Delayed and Logical Memory Delayed from the Wechsler Memory Scale, Similarities from the Wechsler Adult Intelligence Scale, Trail Making A and B, as well as a global measure of cognitive decline derived from principal component analysis (Methods section and Table II in the online-only Data Supplement). All cognitive tests were treated such that higher scores reflected less decline in performance.
Assessment of Brain Atrophy and White-Matter Injury
We examined annualized change in brain atrophy and white-matter injury on MRI also for a mean follow-up of 7 years from examination 7 onward. We examined measures of total brain volume, frontal brain volume, lateral ventricular volume, and white-matter hyperintensity volume. Total brain volume was calculated as the total brain parenchymal volume. Lateral ventricular volumes were calculated by analyzing central cerebrospinal fluid spaces, excluding the temporal horn. We used a Siemens 1T or 1.5T field strength machine with a T2-weighted double spin-echo coronal imaging sequence in contiguous slices of 4 mm. Full details of the imaging methodology are published elsewhere.17 Analysis of MRI images was completed by a neurologist (C.D.) who was blind to subject demographics and ideal CVH scores.
Results were analyzed using SAS Software (SAS Institute, Cary, NC). The assumption of proportionality of hazards was examined and confirmed by including terms for interaction with time. We used multivariable Cox proportional hazards regression models to examine the association between ideal CVH and incident stroke, all-cause dementia, AD, and VaD. For these analyses, hazard ratios are presented accompanied by 95% confidence intervals. The associations between the ideal CVH and the annualized change in each MRI and neuropsychological outcome were examined using linear regression. All models were adjusted for age at examination 7 and sex. Models involving dementia and AD were additionally adjusted for education. Models involving the MRI outcomes were additionally adjusted for age squared and baseline (examination 7) scores on the respective measure, whereas neuropsychological outcomes were further adjusted for baseline scores and education. All analyses were performed separately using recent (examination 7) and remote (examination 5) ideal CVH scores as the exposure. Results were considered statistically significant if P<0.05.
The characteristics of the cohorts are summarized in Table 1. Across the samples, the mean ideal CVH score ranged from 3.1 to 3.3, with <1% of the sample having either a maximum or minimum ideal CVH score. The elapsed time between the quantification of recent and remote ideal CVH scores was 6.9 years (SD, 0.9).
The number of events and characteristics of the outcome measures can be seen in Table 2. Of those participants studied for incident dementia, 25 (3%) had prevalent stroke at baseline. Of the 87 subjects who developed stroke during the study, 7 (8%) subjects also developed incident dementia during follow-up. Four (5%) of these cases were consistent with clinical AD, and 5 (6%) were consistent with VaD. The mean time from baseline to incident all-cause dementia onset was 5.6 years (SD, 2.7 years).
Ideal CVH and the 10-Year Risk of Stroke and Dementia
Higher recent ideal CVH scores were associated with a lower 10-year risk of incident stroke and VaD (Table 3). There was no association between recent ideal CVH and incident all-cause dementia or clinically apparent AD. Higher remote ideal CVH was associated with a lower 10-year risk of incident stroke, all-cause dementia, AD, and VaD.
Ideal CVH and the Risk of Subclinical Brain Injury and Cognitive Decline
Both higher recent and remote ideal CVH scores were associated with less cognitive decline on the tasks of Visual Reproductions Delayed and Similarities (Table 4), indicating less decline in the domains of visual memory, reasoning, and verbal comprehension. Higher remote ideal CVH was also associated with less global cognitive decline as measured by the principal component score. Higher recent ideal CVH was associated with less frontal brain atrophy, whereas higher remote ideal CVH was associated with less whole-brain atrophy.
An additional 14-point ideal CVH score was created using 3 levels for each variable (0=poor status, 1=intermediate status, and 2=ideal status). Analyses were repeated using both the recent and the remote 14-point ideal CVH score. Results were similar to the 7-point ideal CVH scores with the exception that the remote 14-point ideal CVH score was associated neither with the incidence of clinical AD nor with VAD (Tables III and IV in the online-only Data Supplement).
There is a pressing need to develop effective strategies to prevent stroke and dementia. This community-based prospective cohort study revealed that adherence to the AHA’s ideal CVH guidelines was associated with a lower risk of vascular brain injury, including a lower risk of stroke, dementia, cognitive decline, and brain atrophy. To our knowledge, this is the first study to demonstrate an association between the ideal CVH score and the risk of incident dementia. These data support the hypothesis that maintaining CVH protects against all forms of vascular brain injury, including AD. The concept of ideal CVH should, thus, be further promoted to protect the brain, as well as the heart, from vascular risk factors.
The present results are consistent with that of the northern Manhattan9 and Kailuan10 studies both of which found that ideal CVH was associated with a lower risk of stroke. Similarly, consistent with our findings, both the Coronary Artery Risk Development in Young Adults10 and the Maine-Syracuse Longitudinal18 studies reported that ideal CVH was associated with superior neuropsychological performance, across multiple cognitive domains. The present study extends this previous body of research by demonstrating that higher ideal CVH was associated with a vascular pattern of subclinical brain injury and cognitive decline. Although ideal CVH was not associated with white-matter hyperintensity volume in the present study, we have previously found total brain volume to be more sensitive to vascular risk factors in the Framingham Offspring Cohort,19,20 perhaps because white-matter hyperintensities are an uncommon outcome on T2-weighted images in young adults.21 A principal finding of the current study was that remote ideal CVH scores (reflecting midlife CVH) were associated with the future risk of dementia, including AD, VaD, and all-cause dementia.
Many experts conceptualize AD as stemming from the abnormal accumulation of amyloid-β within the brain22; yet, drug discovery has targeted this single mechanism with disappointing results.23 An alternative approach is to conceptualize dementia as a syndrome involving the convergence of multiple pathways and causes.24 Autopsy studies suggest that cerebrovascular disease often coexists with the neurodegenerative hallmarks of AD even when individuals are diagnosed with 1 form of dementia during life.25 There is emerging evidence to suggest that, although the worldwide number of dementia cases is increasing, the age-specific risk may be declining in high-income countries.24,26 This decline seems to coincide with higher educational attainment and improved management of cardiovascular risk factors.24,26 Such findings complement the present results by suggesting that further promoting CVH may help ease the growing burden of dementia.
The association between ideal CVH and stroke was similar for both recent and remote variables. In contrast, remote ideal CVH was a better predictor of all-cause dementia, AD, and global cognitive decline when compared with recent ideal CVH. This discrepancy may reflect the fact that a stroke is an acute event that can develop quickly in response to poor CVH, whereas clinical dementia develops insidiously over decades. The calculation of remote ideal CVH captures a longer length of exposure, allowing more time for dementia to develop in those with poor CVH. Our findings are in line with other studies, suggesting that vascular risk factors are most strongly associated with the risk of later-life cognitive impairment and dementia when measured closer to midlife.14
The focus of the ideal CVH score is on modifiable health behavior and factors.7 Assuming causality, this implies that, with education, guidance, and motivation, individuals can increase their ideal CVH score and potentially decrease their risk of vascular brain injury and dementia. In particular, middle-aged adults should be informed of the importance of adhering to ideal CVH guidelines. It is necessary to consider that the 7 ideal CVH variables are not mutually exclusive. Advocating for regular physical activity, quitting or never smoking, and adhering to a healthy diet are the likely cornerstone to maintaining many of the CVH factors at their ideal statuses, such as normal body weight, blood pressure, cholesterol, and blood glucose.
Limitations of the present study include the fact that ethnic minorities were not well represented in our cohort limiting the generalizability of our findings to these populations. Second, diet and physical activity were measured using self-report techniques, which may be subject to recall bias or social desirability. Scores for diet and physical activity were also calculated using a formula that, although consistent with other publications,8 varied from the exact AHA definition. Finally, we did not examine the individual components of the ideal CVH score given the potential overlap with other Framingham Heart Study publications, which have previously reported on the association of single vascular risk factors with neurological outcomes.
Preventing dementia is one of the most profound challenges facing our aging society. Cardiovascular disease risk factors seem to increase the risk of dementia suggesting that maintaining optimal CVH may be useful in lowering dementia risk.11,25 Thus, there is cause for optimism, with emerging evidence suggesting that society can both grow older and lower dementia burden.27 In line with this notion, our data suggest that adhering to CVH guidelines protects against all forms of vascular brain injury, lessening the burden of cognitive decline, stroke, brain atrophy, and dementia, including AD. Further promoting ideal CVH, particularly to middle-aged adults, may improve neurological outcomes for our aging citizens.
Sources of Funding
Dr Pase was funded by an Australian National Health and Medical Research Council Early Career Fellowship (APP1089698), and his work on stroke was funded by a Rebecca L Cooper Medical Research Foundation grant. The Framingham Heart Study was supported by the National Heart, Lung and Blood Institute Framingham Heart Study (contracts, N01-HC-25195 and HHSN268201500001I) and grants from the National Institute of Neurological Disorders and Stroke (NS17950) and the National Institute of Aging (AG08122, AG033193, and AG049607). Dr DeCarli directs the UC Davis Alzheimer’s Disease Center with funding from the National Institutes of Health (P30 AG010182).
Dr DeCarli is a consultant to Novartis on a clinical trial of LCZ696 for heart failure. The other authors report no conflicts.
Presented in part at the International Stroke Conference, Los Angeles, CA, February 17–19, 2016, and the 68th American Academy of Neurology Annual Meeting, Vancouver, Canada, April 15–21, 2016.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.115.012608/-/DC1.
- Received December 30, 2015.
- Revision received February 25, 2016.
- Accepted February 26, 2016.
- © 2016 American Heart Association, Inc.
- Wolf PA,
- D’Agostino RB,
- Belanger AJ,
- Kannel WB
- Lloyd-Jones DM,
- Hong Y,
- Labarthe D,
- Mozaffarian D,
- Appel LJ,
- Van Horn L,
- et al
- Xanthakis V,
- Enserro DM,
- Murabito JM,
- Polak JF,
- Wollert KC,
- Januzzi JL,
- et al
- Dong C,
- Rundek T,
- Wright CB,
- Anwar Z,
- Elkind MS,
- Sacco RL
- Zhang Q,
- Zhou Y,
- Gao X,
- Wang C,
- Zhang S,
- Wang A,
- et al
- Gorelick PB,
- Scuteri A,
- Black SE,
- Decarli C,
- Greenberg SM,
- Iadecola C,
- et al
- 15.↵American Psychatric Association. Diagnostic and Statistical Manual of Mental Disorders 4th ed. Text Revision. 4th ed. Washington DC: American Psychatric Assocaition; 2000.
- Graham NL,
- Emery T,
- Hodges JR
- Seshadri S,
- Wolf PA,
- Beiser AS,
- Selhub J,
- Au R,
- Jacques PF,
- et al