Cardiac and Carotid Markers Link With Accelerated Brain AtrophyHighlights
The AGES–Reykjavik Study (Age, Gene/Environment Susceptibility–Reykjavik)
Objective—Pathologies in the heart–brain axis might, independently or in combination, accelerate the process of brain parenchymal loss. We aimed to investigate the association of serum N-terminal brain natriuretic peptide (NT-proBNP), as a marker of cardiac dysfunction, and carotid intima media thickness (CIMT), as a marker of carotid atherosclerosis burden, with structural brain changes.
Approach and Results—In the longitudinal population-based AGES–Reykjavik study (Age, Gene/Environment Susceptibility–Reykjavik), we included 2430 subjects (mean age, 74.6 years; 41.4% men) with baseline data on NT-proBNP and CITM (assessed by ultrasound imaging). Participants underwent a high-resolution brain magnetic resonance imaging at baseline and 5 years later to assess total brain (TBV), gray matter, and white matter volumes. Each unit higher log-transformed NT-proBNP was associated with 3.6 mL (95% confidence interval [CI], −6.0 to −1.1) decline in TBV and 3.5 mL (95% CI, −5.7 to −1.3) decline in gray matter volume. Likewise, each millimeter higher CIMT was associated with 10.8 mL (95% CI, −17.3 to −4.2) decline in TBV and 8.6 mL (95% CI, −14.4 to −2.8) decline in gray matter volume. There was no association between NT-proBNP and CIMT and changes in white matter volume. Compared with participants with low NT-proBNP and CIMT, participants with both high NT-proBNP and CIMT had 3.8 mL (95% CI, −6.0 to −1.6) greater decline in their TBV and 4 mL (95% CI, −6.0 to −2.0) greater decline in GMW. These associations were independent of sociodemographic and cardiovascular factors.
Conclusions—Older subjects with both cardiac dysfunction and carotid atherosclerosis are at an increased risk for brain parenchymal loss. Accumulated pathologies in the heart–brain axis might accelerate brain atrophy.
Current evidence indicates that individuals with high burden of cardiovascular comorbidities, even independent of cerebrovascular diseases, run a greater risk of brain atrophy.1 The brain is a highly vascular organ and requires a constant and well-regulated levels of blood flow to maintain its structural integrity.2 It is well-established that the intact function of the heart and extracranial vessel is crucial for regulation of the cerebral circulation.3 Hence, accumulation of pathologies in the heart–brain axis can potentially increase the risk of brain parenchymal loss.
See accompanying editorial on page 2141
Previous patient-based studies have consistently reported that advanced cardiac and carotid pathologies are associated with accelerated brain structural changes.4 Patients with congestive heart failure5 and carotid stenosis6 more frequently develop brain parenchymal loss, and it has been shown that patients who develop cardiac arrest, after resuscitation, have an extensive reduction of their brain parenchymal volume in particular gray matter.7 Although it is widely accepted that cardiovascular pathologies in the heart–brain axis are relevant for brain health, up to now limited robust longitudinal data from general populations exists to substantiate this common preconception. Despite the evidence from patients’ populations, it remains to be known whether community-dwelling older subjects with less severe or subclinical cardiac impairment or carotid atherosclerosis are also at a greater risk for accelerated brain structural changes. Furthermore, it is unclear whether cumulative pathologies in the heart and carotid arteries result in higher degrees of brain parenchymal loss.
In this population-based study of older subjects, we aim to investigate whether higher levels of serum N-terminal brain natriuretic peptide (NT-proBNP), as a marker of left ventricular dysfunction,8,9 and common carotid intima media thickness (CIMT), reflecting atherosclerosis burden in the carotid artery,10 independently or in combination are linked with accelerated structural brain changes.
Materials and Methods
Materials and Methods are available in the online-only Data Supplement.
Baseline mean age of the participants was 74.6 years and 41.4% were men. Median NT-proBNP level was 124.9 ng/L and mean value of CIMT was 1.0 mm. Average total brain tissue, gray matter, and white matter volumes were 1077.8, 686.7, and 391.1 mL, respectively (Table 1). Cross-sectional analyses showed that higher NT-proBNP was associated with lower total brain volume and grey matter volume (Table II in the online-only Data Supplement). There was no cross-sectional association between CIMT and brain volumes at baseline (Table III in the online-only Data Supplement).
In the longitudinal analyses, each unit higher log-transformed serum NT-proBNP was associated with 3.9 mL (95% confidence interval [CI], −6.0 to −1.8) decrease in total brain parenchymal volume (Table 2, model 1). Similarly, each unit higher log-transformed serum NT-proBNP was associated with 3.7 mL (95% CI, −5.6 to −1.8) decrease in gray matter volume. Although a trend was observed in the association between higher serum NT-proBNP and higher white matter volume loss, the associations did not reach statistical significance (P=0.09). Further adjustment of the analyses for sociodemographic and cardiovascular risk factors did not essentially change these associations (Table 2, model 2).
In the longitudinal analyses, each millimeter higher CIMT was associated with 13.3 mL (95% CI, −19.3 to −7.2) decrease in total brain parenchymal volume. Similarly, each millimeter higher CIMT was associated with 12.6 mL (95% CI, −18.1 to −7.0) decrease in gray matter volume (Table 3, model 1). There was no association between higher CIMT and decrease in white matter volume (P=0.84). Further adjustments for sociodemographic and cardiovascular risk factors only slightly changed the magnitude of the associations (Table 3, model 2).
Subjects with high NT-proBNP and high CIMT had the greatest total brain parenchymal and grey matter volume loss compared with the other groups (all P<0.05). For both total and gray matter volumes, the largest difference was observed between subjects with low NT-proBNP and CIMT and subjects with high NT-proBNP and CIMT (both P<0.001). In contrast, there was not such an association in relation to decline in white matter volume (Figure). Further adjustments of the analyses for sociodemographic and cardiovascular risk factors revealed similar associations (Table IV in the online-only Data Supplement). Compared with participants with low NT-proBNP and CIMT, participants with both high NT-proBNP and CIMT had 3.8 mL (95% CI, −6.0 to −1.6) higher decline in their TBV and 4 mL (95% CI, −6.0 to −2.0) higher decline in GMW. The interaction between NT-proBNP and CIMT in relation to brain volume loss was significant for total brain parenchymal volume (P=0.04) and suggestive for grey matter volume (P=0.15), indicating that subjects with both high NT-proBNP and CIMT had greatest decline in brain parenchymal volumes. In a sensitivity analysis, we excluded participants who had stroke and/or dementia (n=128). We observed similar findings after this exclusion (data not shown). In a series of extra analyses, we tested the associations using inverse probability weighting method. Repeating the analyses with this method revealed similar findings with larger effect estimates (Tables V through VII in the online-only Data Supplement).
In this prospective cohort study of community-dwelling older subjects, we showed that elevated NT-proBNP, as a marker of left ventricular dysfunction,8 and CIMT, as a measure of atherosclerosis burden at the carotid artery,10 are associated with steeper decline in the structural brain volumes. We found that combination of high NT-proBNP and CIMT is linked with greatest risk for accelerated structural brain changes.
Different lines of evidence from epidemiological and neuroimaging studies indicate that long-lasting exposure to cardiovascular risk factors, independent of cerebrovascular accidents, are associated with higher risk of brain parenchymal loss and atrophy.11–13 Although the exact mechanism behind this association is not fully known, an increasing body of evidence suggests that cerebral hypoperfusion is a key factor linking cardiovascular morbidities with accelerated brain structural changes.14 The brain is a demanding organ consuming ≈20% of total body oxygen and 25% of total body glucose. Therefore, integrity of the brain depends on adequate supply of oxygen and energy through blood flow.15 It is known that multiple systemic and cerebrovascular mechanisms act in concert to maintain cerebral blood flow in a stable range despite fluctuations in the systemic perfusion.16 However, in the presence of prolonged low systemic perfusion, because of pathologies in the heart and large vessels, these regulatory mechanisms might fail to safeguard the brain and eventually cerebral hypoperfusion occurs.
The heart is the generator of the cerebral perfusion pressure and intra- and extracranial vessels modify the cerebral perfusion as they deliver the blood to the brain.17 Given the increasing prevalence of pathologies affecting the heart and cerebropetal arteries in the elderly, these pathologies can be expected to contribute to changes in the brain as people age. Although previous studies showed that individual pathologies at the levels of heart or carotid arteries are independently associated with adverse brain outcomes,18 the current study shows that accumulated pathologies in the heart and carotid arteries might reinforce each other and put older subjects at an extra risk for accelerated brain parenchymal loss. We observed that impaired cardiac function and carotid atherosclerosis are mainly linked to gray matter volume decline rather than to white matter volume decline. The brain receives ≈15% of cardiac output, and it is well-known that more than two thirds of total cerebral blood flow is directed toward gray matter because of its high metabolic demand.19 Therefore, it is expected that gray matter, more than white matter, becomes vulnerable to pathologies in the heart–brain axis that jeopardize blood flow to the brain.
Adequate and constant cerebral perfusion is safeguarded by the so-called cerebral autoregulation. Cerebral autoregulation encompasses a series of complex physiological functions to match the fluctuating systemic supply of blood flow with the high energy demands of the brain.20 Although such a regulatory mechanism acts quickly in the face of immediate changes in the systemic perfusion pressure, it is yet to be determined whether cerebral autoregulation can also handle the long-lasting hypoperfusion state because of established pathologies in the heart–brain axis.21 Previous studies have shown that cardiovascular risk factors are associated with lower cerebral blood flow.22 Hence, chronic cerebral hypoperfusion might be a key mechanism behind the link between cardiac and carotids pathologies and accelerated brain parenchymal loss. In addition to the plausible roles of hypoperfusion in this link, it needs to be acknowledged that impaired cardiac function and advanced atherosclerosis might also reflect a global systemic vascular condition that affects not only the large vessels but also small vessels.23 In addition, other pathways such proinflammatory state and activation of the renin–angiotensin system might play contribute in this link. Both high systemic inflammation and activation of renin–angiotensin system have been explained in association with higher cardiovascular burden and atherosclerosis, which can independently affect the brain and speed up the pace of brain parenchymal loss.24,25
The growing numbers of older people and increasing prevalence of age-related disorders of the brain such as dementia warrant further studies to identify novel strategies to decelerate pace of brain ageing.26,27 Although the pivotal roles of the heart and extracranial arteries in providing adequate oxygen, glucose, and nutrient for the brain are well-established, limited evidence exists on the contribution of the heart–brain axis in the development and progression of abnormal brain ageing in older adults. Findings of this study might highlight that older adults with higher loads of multiple pathologies in the heart–brain axis should be considered as high-risk groups for accelerated brain aging. Although previous reports indicate that in normal aging subjects experience 0.32% annual decline in their total brain volumes,28 in our study, subjects with both high NT-proBNP and CIMT had about 1% annual brain parenchymal loss. The magnitude of this change becomes even clearer when it is compared with annual brain parenchymal loss of patients with Alzheimer disease, which is 2%.29 Collectively, our results warrant closer collaboration between cardiologists and neurologists in identification of older patients at highest risk for accelerated brain ageing. Once clinicians detect pathologies at the cardiac level should search for concomitant abnormalities in the other parts of the heart–brain axis and vice versa.
This study has many strengths and limitations. Availability of repeated neuroimaging data and information on the cardiovascular status of >2000 community-dwelling older men and women are among the major strengths of this study. As a limitation, we could only include subjects who survived up to the follow-up session, which might limit generalizability of our results. Nonetheless, as the inverse probability weighting analyses suggested, it is expected that our findings might even underestimate the magnitude of the associations because survivors are generally healthier. It needs to be pointed out that despite all the adjustments for the conventional cardiovascular risk factors, we cannot exclude possible roles of unmeasured confounders on the associations. In this study, we used the mean of far- and near-wall CIMT to assess the degree of carotid atherosclerosis. Although this method has been validated before, it is generally recognized that the far-wall CIMT might better reflect the true thickness of the carotid wall. We assessed the left ventricular function using serum NT-proBNP. Although it has been shown that NT-proBNP is a reliable marker for left ventricular dysfunction,9 additional studies are needed to confirm the link between impaired cardiac function and brain parenchymal loss using imaging techniques such as echocardiography and cardiac magnetic resonance imaging.
This study shows that high serum NT-proBNP, as a marker for impaired cardiac function, and intima media thickness reflecting atherosclerosis burden in carotid arteries are linked with accelerated structural brain changes. Combination of high NT-proBNP and high CIMT was associated with greatest brain parenchymal loss. This study suggests that pathologies at the levels of the heart and carotid arteries act in concert and, instead of discipline-specific evaluations by cardiologists and neurologists, older subjects might further benefit from a comprehensive evaluation of the heart–brain axis in relation to adverse brain outcomes. Findings of this study call for further research on the cumulative effects of cardiac and carotid abnormalities on the brain parenchymal loss. Our data in a general population of older subjects need to be replicated in middle-aged and younger adults when interventions might have a bigger impact on brain health in the subsequent years. Furthermore, the exact mechanisms behind the observed associations need to be further explored by experiments on animal models and interventional approaches targeting multiple components of the heart–brain axis to better understand the interplays of various cardiovascular mechanisms that affect brain structural integrity.
Sources of Funding
This study was funded by the National Heart, Lung, and Blood Institute Intramural Research Program (Z01 HL004607-08 CE), the National Institute on Aging Intramural Research Program (N01-AG-12100), Hjartavernd (the Icelandic Heart Association), and the Althingi (the Icelandic Parliament). The study was approved by the Icelandic National Bioethics Committee (VSN: 00–063) and the Medstar Research Institute (project 2003–145). Behnam Sabayan is partly supported by a grant from Internationale Stichting Alzheimer Onderzoek.
The online-only Data Supplement is available with this article at http://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.116.308018/-/DC1.
- Nonstandard Abbreviations and Acronyms
- N-terminal brain natriuretic peptide
- confidence interval
- carotid intima media thickness
- Received May 31, 2016.
- Accepted August 24, 2016.
- © 2016 American Heart Association, Inc.
- Raichle ME,
- Gusnard DA
- Marshall RS,
- Lazar RM
- Bay M,
- Kirk V,
- Parner J,
- Hassager C,
- Nielsen H,
- Krogsgaard K,
- Trawinski J,
- Boesgaard S,
- Aldershvile J
- Macheret F,
- Boerrigter G,
- McKie P,
- Costello-Boerrigter L,
- Lahr B,
- Heublein D,
- Sandberg S,
- Ikeda Y,
- Cataliotti A,
- Bailey K,
- Rodeheffer R,
- Burnett JC Jr.
- van Buchem MA,
- Biessels GJ,
- Brunner la Rocca HP,
- de Craen AJ,
- van der Flier WM,
- Ikram MA,
- Kappelle LJ,
- Koudstaal PJ,
- Mooijaart SP,
- Niessen W,
- van Oostenbrugge R,
- de Roos A,
- van Rossum AC,
- Daemen MJ
- van Beek AH,
- Claassen JA,
- Rikkert MG,
- Jansen RW
- Pase MP,
- Grima NA,
- Stough CK,
- Scholey A,
- Pipingas A
- Nitkunan A,
- Lanfranconi S,
- Charlton RA,
- Barrick TR,
- Markus HS
- Trollor JN,
- Valenzuela MJ
Markers of cardiac impairment and carotid atherosclerosis are associated with accelerated brain parenchymal loss.
The combination of pathologies at the levels of the heart and carotid arteries poses an accentuated risk for brain atrophy.
Older subjects can further benefit from a comprehensive evaluation of the heart–brain axis, instead of a discipline-specific approach, in relation to adverse brain outcomes.