PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Circulation. Author manuscript; available in PMC 2013 May 1.
Published in final edited form as:
PMCID: PMC3427730
NIHMSID: NIHMS373576

Multiple Biomarkers and Risk of Clinical and Subclinical Vascular Brain Injury: The Framingham Offspring Study

Abstract

Background

Several biomarkers have been individually associated with vascular brain injury but no prior study has explored the simultaneous association of a biologically plausible panel of biomarkers with the incidence of stroke/TIA, and the prevalence of subclinical brain injury.

Methods and Results

In 3127 stroke-free Framingham Offspring (59±10 yrs, 54%F), we related a panel of 8 biomarkers assessing inflammation(C-reactive protein[CRP]), hemostasis(D-dimer and plasminogen activator inhibitor-1), neurohormonal activity(aldosterone-to renin ratio, B-type natriuretic peptide[BNP] and N-terminal pro-atrial natriuretic peptides) and endothelial function (homocysteine and urinary albumin/creatinine ratio[UACR]) measured at the 6th examination(1995–98) to risk of incident stroke/TIA. In a subset of 1901 participants with available brain MRI (1999–2005), we further related these biomarkers to total cerebral brain volume (TCBV), covert brain infarcts (CBI), and large white matter hyperintensity volume(LWMHV).

During a median follow-up of 9.2 years, 130 participants experienced incident stroke/TIA. In multivariable analyses adjusted for stroke risk factors, the biomarker panel was associated with incident stroke/TIA and with TCBV (p<0.05 for both), but not with CBI or LWMHV (p >0.05). In backwards elimination analyses higher log-BNP (hazards ratio [HR] 1.39/SD, p=0.002) and log-UACR (HR1.31/SD, p=0.004) were associated with increased risk of stroke/TIA and improved risk prediction over using the Framingham stroke risk profile alone; using <5%, 5–15% or >15% 10-year risk categories the net reclassification index was 0.109;p=0.037). Higher CRP (β=−0.21/SD,p=0.008), D-dimer(β==−0.18/SD,p=0.041), tHcy(β=−0.21/SD,p=0.005), and UACR(β=−0.15/SD,p=0.042) were associated with lower TCBV.

Conclusions

In a middle-aged community sample, we identified multiple biomarkers that were associated with clinical and subclinical vascular brain injury and could improve risk stratification.

Keywords: biomarkers, epidemiology, magnetic resonance imaging, risk stratification, stroke prevention

Introduction

Stroke is the fourth leading cause of death and is a major cause of long-term disability in America. In spite of progress in acute treatment protocols, prevention remains the most effective approach to reduce the personal and public health impact of stroke.1 The advent of sensitive brain imaging techniques has established that the burden of cerebrovascular disease is far greater than that of overt clinical disease.2 In large population–based studies, the MRI findings of subclinical brain injury: presence of covert brain infarcts (CBI), larger white matter hyperintensity volumes (LWMHV), and smaller total cerebral brain volumes (TCBV) are common, with CBI prevalence ranging from 11% in younger subjects (<age 65 yrs) to 28% in older subjects.3,4 Although these subclinical findings do not cause abrupt clinical symptoms, they are often associated with subtle consequences such as cognitive decline5 and motor function impairment,6 and have been associated with an increased risk of subsequent stroke.7

Vascular risk factors have also been implicated in subclinical brain injury,3,8 but the underlying pathophysiological pathways remain poorly understood. Several individual circulating and urinary biomarkers923 have been identified that lie along biologically plausible pathways that could lead to vascular brain injury. However, to our knowledge, the relative contribution of these biomarkers from distinct biological pathways to an increased risk of both covert and overt vascular brain disease has not been explored. We hypothesized that select biomarkers will be associated with the presence of covert and/or with the incidence of overt vascular brain disease in a large community-based, middle-aged cohort. We tested this hypothesis by relating a panel of 8 biomarkers (inflammation (C-reactive protein[CRP]), hemostasis (D-dimer [DD] and plasminogen activator inhibitor-1[PAI-1]), neurohormonal activity (aldosterone-to renin ratio, B-type brain natriuretic peptide [BNP] and N-terminal pro-atrial natriuretic peptides [NT-ANP]) and endothelial function (homocysteine [tHcy] and urinary albumin/creatinine ratio [UACR])) to incident stroke/TIA prospectively, and to prevalent subclinical brain disease in a large middle-aged community-based sample using a conservative statistical approach that minimized multiple testing.

Methods

Study Sample

In 1971, the Framingham Offspring Study was initiated with enrollment of 5124 individuals who were either offspring of the Original Cohort participants or spouses of the offspring.24 These participants have been assessed every four years, are currently undergoing their 9th examination, and remain under surveillance for incident stroke and TIA. Of 3532 participants who attended the 6th examination cycle (1995–1998) at which biomarkers were measured, 3209 had biomarker information available, the remainder either did not undergo phlebotomy, or had inadequate sample volume. When compared to those without biomarkers data (n=323), the participants with available biomarker measures were younger, and less likely to have diabetes mellitus or prevalent cardiovascular disease. (Supplemental Table 1).

We considered two study samples for the present investigation (Supplemental Figure 1). Among the 3209 participants with biomarker information, after excluding those with prevalent stroke (n=72) and those without subsequent stroke/TIA follow up information (n=10); 3127 participants (59±10 yrs, 54%F) were eligible for investigation of the association between the panel of biomarkers and incident stroke/TIA (Study Sample 1). All participants who attended the 7th examination cycle were invited to undergo brain MRI (1999–2005). Of the 3209 who had biomarker information from the 6th Offspring examination; 33 died prior to the start of the MRI study, 63 were excluded for prevalent dementia and 10 for prevalent stroke at the time of MRI; 1173 persons either declined MRI or had a contra-indication. After additionally excluding persons with another neurological illness (such as multiple sclerosis or brain tumor) that could affect MRI measurements, persons with a serum creatinine >2.0 mg/dL and those with missing covariate data (n=29), 1901 participants (58±10yrs, 54%F) with both biomarkers and brain MRI measures available were eligible for investigation of the association between the panel of biomarkers and subclinical brain disease (Study Sample 2). Baseline characteristics comparing persons who were in our MRI study sample with those who had biomarker data but were not in our brain MRI analysis sample are shown in Supplemental Table 2.

Biomarkers measurements

We assessed 8 independent biomarkers from biologically plausible pathways implicated in the pathogenesis of overt and covert vascular brain injury. Plasma concentrations of each biomarker were routinely measured on previously frozen (stored) blood samples drawn in the fasting state from persons who attended the 6th examination cycle and urinary albumin-to-creatinine ratio was assessed in stored urine samples. All measurements were performed using previously reported methods specific for each biomarker;25 high-sensitivity CRP levels were determined by Dade Behring BN100 nephelometer; plasma BNP and NT-ANP were measured by high-sensitivity immunoradiometric assays; D-dimer and plasma PAI-1 antigen were determined using a commercially available ELISA; serum aldosterone was measured from extracted and fractionated serum using a radioimmunoassay; direct plasma renin was measured by immunochemiluminometric assay; and total plasma homocysteine was measured using high-performance liquid chromatography with fluorometric detection.23 The average interassay coefficients of variation (CV) for each biomarker were as follows: CRP (2.2%); D-dimer (11.7%); PAI-1 (7.7%); aldosterone (4.0% for high concentrations, 9.8% for low concentrations); renin (2.0% for high concentrations, 10.0% for low concentrations); BNP (12.2%); NT-ANP (12.7%); and tHcy (9%). UACR (mg/g) was assessed on a morning urine specimen as a ratio between urinary albumin measured using immunoturbidimetry and urinary creatinine measured using a modified Jaffe method. The average interassay CVs for these two UACR assays were 7.2% and 2.3%, respectively.

Although the biomarkers were not measured at the same time as the initial brain MRI, we chose the 6th examination cycle as our baseline since (i) the entire panel of biomarkers was available at that examination, (ii) this design increases the follow-up period available to study the risk of incident stroke/TIA, and (iii) MRI changes accumulate over time, hence prior rather than concurrent levels of biomarkers are more likely to be related to MRI markers of subclinical disease.23

Outcomes

Stroke and TIA (as a composite clinical outcome)

Our stroke surveillance and protocol for determining the diagnosis and type of stroke have been previously published, all clinical events were ascertained by 2 or more neurologists at a consensus review.26 Stroke was defined as an acute onset focal neurological deficit of presumed or definite vascular etiology, persisting for more than 24 hours. A TIA was defined clinically as a focal neurological deficit lasting for ≤24 hours. We also utilized the data on clinical features, imaging studies and other laboratory criteria, noninvasive vascular studies, cardiac evaluations for a source of embolus and, when available, information from autopsy studies.

Brain Imaging Measures (as a subclinical outcome)

MRI acquisition, measurement techniques and inter-rater reliability have been described previously.27 Briefly, the images were analyzed by operators blinded to the participant’s demographics, exposure to stroke risk factors, and plasma biomarker levels. Brain volume was determined by manual delineation of the intracranial vault to determine the total cranial volume (TCV) followed by subsequent mathematical modeling to determine total brain parenchymal volume (TCB). Total cerebral brain volume (TCBV) was then computed as a ratio of TCB/TCV; representing a measure of brain parenchymal volume corrected for differences in head size. The volume of abnormal white matter hyperintensity was determined according to previously published methods.27 The participants were categorized as having large WMH volumes (LWMHV) if the log-WMH volume/TCV was more than 1 SD above the age-adjusted mean in this cohort.The presence or absence of covert brain infarcts (CBI) was determined manually by the operator, based on the size (>3 mm), location and imaging characteristics of the lesion.

Definition of Covariates

Previously described and validated components of the Framingham Stroke Risk Profile (FSRP) were used as baseline covariates.26 These include age, sex, systolic blood pressure, antihypertensive therapy, diabetes, smoking status, history of cardiovascular disease(CVD), and the presence of atrial fibrillation(AF).

Statistical Analysis

All biomarkers that displayed skewed distribution were log-transformed to normalize their distributions. An age-adjusted Spearman correlation coefficient was calculated for each pair of biomarkers. The biomarker panel was related to the risk of incident stroke/TIA using Cox proportional hazards techniques and to subclinical MRI measures using linear (for continuous measures) or logistic (for binary measures) regression analyses. We used a 2-step analytical approach to minimize false positives due to multiple statistical testing. First, for each outcome, we used multivariable regression analysis to perform a global test of significance to examine whether the entire set of biomarkers was related to a given outcome. In this step, we tested whether the set of 8 biomarkers (denoted m1-m8 below) was associated with stroke/TIA risk using a 8 degree of freedom likelihood ratio test (LRT) of the null hypothesis H0: βm1 = βm2…=βm8 = 0. The LRT chi-squared statistic was obtained by comparing likelihoods from two models: (i) with covariates only and (ii) with covariates plus 8 biomarkers. Subsequent analyses were performed only if the global P value was <0.05. If the biomarker panel was associated with the outcome of interest, then as a second step, we performed a stepwise multivariable regression with backward elimination to select a final parsimonious set of informative biomarkers associated with the outcome of interest using a p <0.05 for retention in the final model. For clinical stroke/TIA we also assessed the incremental utility of the identified individual biomarkers in predicting stroke/TIA risk by calculating net reclassification improvement (NRI) using a more conventional logistic approach and a novel approach that allows for differences in survival times.28,29 We used clinically meaningful risk categories of <5%, 5–15% and >15% 10-year stroke risks,26,28 defined a priori as risks experienced by the average person, at the baseline age of the current sample, in the presence of zero (low risk), one or two(intermediate risk), or more than two (high risk) of the stroke risk factors included in the Framingham stroke risk profile.26

All analyses were adjusted for age and sex, baseline cardiovascular covariates and serum creatinine. For the MRI analyses, we additionally adjusted for the time interval between baseline examination and time of MRI acquisition. Finally we compared the magnitude of the association of specific biomarkers with TCBV to the impact of chronological age by dividing the regression coefficient for the biomarker by the regression coefficient for age, estimated using linear regression models.

Results

Study Samples

The baseline clinical characteristics and biomarkers levels are provided in Table 1 for the two study samples. The median follow-up period for stroke/TIA was 9.2 years (range 0.5–11.9 years), and the median time interval between biomarker and subclinical brain MRI measurements was 3.5 years (range 1.3–10.2 years). Pairwise age-adjusted Spearman correlations among selected biomarkers as well as of selected biomarkers with clinical covariatesare presented in Table 2. The highest age-adjusted correlations were observed for BNP and NT-ANP (r=0.60, p<0.001), CRP and D-dimer (r=0.33, p<0.001), and CRP and PAI-1(r=0. 29, p<0.001).

Table 1
Baseline Characteristics of the Study Samples (1 and 2) used to Estimate Association of Multiple Biomarkers with Risk of Incident Stroke or Transient Ischemic Attack (TIA) and with Brain MRI Measures.
Table 2
Age-Adjusted Pairwise Correlation among Biomarkers and Covariates

Association of Biomarker Panel and Stroke/TIA (Table 3)

We observed 130 incident stroke/TIA events in the 3127 participants; of these 77 were strokes and 53 were TIA events. In multivariable analyses adjusted for traditional stroke risk factors, the biomarker panel was associated with increased risk of stroke/TIA (global P=0.001). Upon backwards eliminations, only two biomarkers, BNP (p=0.002) and UACR (p=0.004), were retained in the final model. Higher BNP and UACR were each associated with an increased risk of stroke/TIA (HR 1.39 per 1-SD increment in log-BNP, p=0.002 and HR 1.31/SD increment in log-UACR, p=0.004). Adding these biomarker data to risk assessment models based on traditional stroke risk factors alone (as identified by the Framingham stroke risk profile) resulted in improved risk prediction. When the risk was reclassified into clinically meaningful risk categories of <5%, 5–15% or >15% 10-year risks, we observed a modest improvement in the discrimination of events with an NRI=0.109 [0.008–0.211]; p=0.037 using a conventional logistic model and an NRI=0.109 [95% CI: 0.008–0.211] allowing for differences in survival time(details in Supplemental Tables 3 and 4 and in Figure 1).On secondary analyses, we found a similar association of higher BNP and UACR with an increased risk of stroke alone (HR 1.28 per SD increment in log-BNP, p=0.04 and HR 1.34 per SD increment in log UACR, p=0.009).

Figure 1
Kaplan-Meier Curves of the Cumulative Probability of Incident Stroke within Risk Categories Defined Based only on Clinical Stroke Risk Factors (Reduced Models) or on both Clinical Risk Factors and Biomarker Levels (Full Models). Risk categories were classified ...
Table 3
Effect of Multiple Biomarkers on Incident TIA/Stroke Risk

Association of Biomarker Panel and Subclinical Brain MRI Measures (Table 4)

We further examined the association between the same biomarker panel and three measures of subclinical brain injury (TCBV, CBI and LWMHV). Mean TCBV and LWMHV (these are expressed as ratios over intracranial volume) were 79.4 and −3.0, respectively, and the prevalence of CBI was 11% (Table 1). In multivariable analyses, the biomarker panel was associated with TCBV (global P<0.001), but not with CBI or z-LWMHV (global P>0.05). Using backward elimination models, adjusted for traditional stroke risk factors, time to MRI acquisition and serum creatinine levels, four biomarkers were significantly associated with lower TCBV: higher CRP (β=−0.21/SD, p=0.008), D-dimer (β=−0.18/SD, p=0.041), tHcy (β=−0.21/SD, p=0.005), and log-UACR (β=−0.15/SD, p=0.042). The impact of a single SD unit increase in plasma CRP or tHcy levels on TCBV was exactly equivalent to one year of brain aging, whereas one SD unit increases in D-Dimer and log-UACR levels corresponded to 9 and 7 months, respectively, of brain aging.

Table 4
Effect of Multiple Biomarkers on brain MRI measures

Discussion

In our community-based, stroke-free middle-aged sample, we investigated a panel of eight biologically plausible biomarkers for association with clinical and subclinical vascular brain injury. In recent years, a large number of biomarkers have been individually associated with the risk of cerebrovascular disease. These biomarker levels are often correlated or interact with each other in complex ways reflecting the actions of interrelated biological pathways. It is essential to identify a parsimonious set of maximally informative markers and to assess their relative contributions to risk prognostication if we are to improve upon clinical risk prediction scores in a practical and cost-effective manner. We observed that the overall biomarker panel was associated with risk of stroke/TIA and with smaller brain volumes. We further identified that higher BNP levels and albuminuria were associated with the risk of initial incident stroke/TIA and addition of these biomarker data resulted in a modest improvement in categorizing stroke risk among participants. Albuminuria (but not BNP) was also associated with smaller total brain volume, as were higher CRP, D-dimer and tHcy levels.

To our knowledge, this is the first study to relate a large biomarker panel to both clinical and subclinical vascular brain injury. Our findings concur with prior observations that abnormalities in multiple biological pathways affect the risk of incident cardiovascular events (including stroke/TIA) and death.25,30 They confirm and extend to a middle-aged, community sample, prior reports describing associations of several of these individual biomarkers with the risk of stroke/TIA or of subclinical brain disease, respectively.

Biological pathways underlying Stroke/TIA Risk

In previous reports we had observed that elevated plasma CRP9 and tHcy levels23 were individually associated with an increased risk of stroke/TIA and that elevated UACR was associated with a higher risk of all CVD events.31 Using a multiple biomarker approach to evaluate the risk of all initial CVD events (including stroke) in offspring, we had observed that BNP and UACR were associated with the risk of a major vascular event, but that study was underpowered to examine the outcome of stroke/TIA.25 Our present data extend those observations by demonstrating associations of BNP and UACR with the risk of incident stroke/TIA using a longer follow-up period and including a larger number of incident stroke/TIA events. We did not confirm in the present investigation our previously reported observation of an association of CRP and tHcy with incident stroke/TIA, most likely because the present investigation evaluated a much younger sample that was associated with lower mean values of these particular biomarkers, and hence was underpowered due to a smaller number of incident events. Interestingly, the effect size of the association between CRP and stroke/TIA in the present study was similar to what we had observed in the prior study but it did not reach statistical significance. In addition, in prior reports we have also observed that the mean tHcy levels were lower in the Offspring at the 6th Offspring examination (after dietary folate fortification was initiated) compared to the 5th Offspring examination, which was prefortification.32 Thus, lower levels of tHcy among participants included in the present study may have contributed to the lack of the association between tHcy levels and incident stroke/TIA.

BNP and Stroke/TIA

BNP is a natriuretic peptide with diuretic and vasodilatory activities released by cardiomyocytes during hemodynamic stress and cleared by the kidneys.33 Elevated plasma levels of BNP have been observed in older patients, perhaps due to myocardial fibrosis and cardiac hypertrophy and/or secondary to renal dysfunction in otherwise healthy elderly persons, but also in persons with essential hypertension, left ventricular hypertrophy (LVH), diastolic dysfunction, AF and renal failure.33 As a biological marker, BNP is used for diagnostic and prognostic purposes in patients with chronic congestive heart failure (CHF), and has been evaluated for the risk stratification of individuals with an acute coronary syndrome (ACS).34 Most of the cardiac conditions mentioned above have been associated with the risk of stroke/TIA,30 but an independent link between higher BNP levels and stroke/TIA risk observed in our middle-aged population after adjusting for blood pressure, cardiac and renal disease suggests there may be additional pathophysiological mechanisms. For example, an increased level of physical activity (which reduces stroke risk) may decrease BNP levels.35

Albuminuria and Stroke/TIA

Albuminuria has been associated with the risk of CVD events and the presence of subclinical vascular disease in other organs,25,36 but its role in cerebrovascular disease is less clear, and only a few cohort studies have evaluated the association of albuminuria or microalbuminuria (MA) with stroke as a primary end point.1820 Our findings concur with these prior reports. While the underlying mechanism for this association is unclear, albuminuria is a known marker of glomerular endothelial dysfunction, has been associated with diffuse endothelial dysfunction in other vascular beds,37,38 and is a marker of cardiovascular target organ damage.

Biological pathways underlying Subclinical Vascular Brain Disease

In prior reports we had demonstrated that the components of the FSRP and selected inflammatory markers are related to LWMH8 and to brain atrophy,11 respectively. We have also previously shown that elevated plasma tHcy levels are associated with an increased risk of CBI and smaller brain volumes.23 Our present findings, however, identify additional biological pathways that also may be involved in mediating brain atrophy.

Inflammation and Brain Volumes

Links between inflammation and brain atrophy, and between lower TCBV and poorer cognitive performance have been previously reported.11 However, the findings of an association between higher CRP levels and smaller brain volumes are novel. CRP is a marker of a chronic low-grade inflammation and a marker of endothelial dysfunction. As a part of an innate-immune system response, CRP deposits have been detected in brain micorovessels.39 Several cohort studies have observed increased CRP levels prior to clinical onset of Alzheimer’s disease (AD) and/or vascular dementia but brain MRI measurements were not described in these reports.40,41 Our failure to find an independent association between CRP and TCBV in an earlier report could reflect the smaller sample size in that investigation and the use of a later baseline examination cycle closer to the date of MRI in that study because inflammation likely affects brain volumes over months to years. Further, it is possible that an association with CRP was obscured by a stronger association with a correlated inflammatory marker (IL6), which was included in the previous report11 but not in the current study (IL6 levels are not available at the 6th examination which served as the baseline). Since CRP is a commonly measured biomarker, we believe it could be a valuable contributor to a future multimarker panel for determining risk of cerebrovascular injury.

Hemostasis and Brain Volumes

D-dimer (DD) as a marker of fibrinolytic turnover is present in very small amounts in a healthy population. Its role in activation of coagulation in clinical and subclinical brain infarcts remains controversial.13 Prior observations have suggested an association between hypercoagulability and brain atrophy in the elderly.12 In the same report increased levels of DD have been associated with white matter hyperintensities and Binswanger’s Disease (BD).12 It is plausible that endothelial dysfunction and/or inflammation, might trigger an alteration in the hemostatic pathways within the brain microvasculature/parenchyma and accelerate microvascular ischemia. However, further experimental and human studies are required to define these interactions.

Endothelial Function and Brain Volumes

Our previous observation that higher plasma tHcy levels predispose to smaller brain volumes23 has been confirmed in this report. Homocysteine is an independent risk factor for atherosclerosis and cardiovascular disease. As an excitatory neurotransmitter, it binds to the NMDA receptors and leads to oxidative stress, endothelial dysfunction and inflammation and neuronal injury.42 Potential mechanisms through which tHcy could result in brain atrophy have been briefly discussed in a prior publication.23

Albuminuria and Brain Volumes

Albuminuria is not only an indicator of renal microvascular disease, but may reflect diffuse microvascular disease, endothelial dysfunction and vascular damage (the Steno hypothesis).37,38 In a case-control study of patients with HTN, but no DM, microalbuminuria was independently associated with a higher prevalence of silent lacunar infarcts and greater carotid intima-media thickness.43 As confirmed in our report, albuminuria has been also associated with stroke and brain atrophy.1821 However, the mechanisms underlying this association remains unclear. It has been suggested that with aging, endothelial dysfunction, intimal fibromuscular hyperplasia and lipohyalinosis in cerebral microvasculature might develop in parallel with vascular changes in the renal microvascular system.21

The absence of an association between the biomarker panel and CBI and/or WMH does not necessarily mean that such associations do not exist. It is plausible that, with the relatively younger age of our sample, we had limited statistical power to detect such associations.

Strengths and Limitations

The strengths of our study are the inclusion of a stroke-free, middle-aged study sample, their longitudinal follow-up with careful prospective surveillance of clinical events, availability of a range of biomarkers measured by previously validated methods, availability of baseline values for traditional stroke risk factors and availability of volumetric brain MRI measurements performed by radiologists blinded to clinical events and biomarker information. The limitations of our study are the predominantly white study sample which limits the generalizability of our findings to other ethnicities/races. At present, we have only a single-occasion measurement for each biomarker and single brain MRI measurements. Further, the subsetof participants who underwent brain MRI and also had biomarker data had a lower prevalence of vascular risk factors and was healthier thanthe rest of the Framingham Offspring; this could potentially bias our results. Finally we assessed the incremental prognostic value of BNP and UACR in predicting risk of stroke/TIA using the NRI and clinically reasonable cut-points, but recognize that changing the cut-points used to define low, intermediate and high-risk categories would vary the NRI computed.

Conclusions

In summary, alterations in several different biological pathways may underlie the pathophysiology of clinical and subclinical brain disease. Using a conservative multimarker approach we identified two biomarkers associated with the risk of incident stroke and four biomarkers associated with a smaller brain volume in a middle-age, stroke-free sample. Urinary albuminuria, a comparatively less well studied biomarker of cerebrovascular risk was associated with both outcomes. However, both BNP and UACR increased the risk for stroke/TIA, and offered modest improvements in accuracy of risk classification. These findings should be taken into consideration when building risk stratification models and may also point to novel, potentially synergistic preventive and therapeutic measures. Finally, the performance of such multimarker strategies in risk stratification for clinical and subclinical vascular disease requires further investigation, particularly in the current ‘Omics’ era as the number of putative biomarkers expands at an exponential rate providing both an opportunity and a challenge to develop improved, cost-effective risk prediction models.

Short Commentary – Clinical Prospective

Numerous biomarkers, representing various biological pathways, have been independently implicated in cerebrovascular disease, but their relative contribution to the prediction of clinical cerebrovascular disease is uncertain. In a middle-aged community sample, we used a multimarker approach to identify plausible biomarkers associated with clinical and subclinical vascular brain injury. This approach also permitted us to assess the contribution of each biomarker to clinical risk reclassification after accounting for standard vascular risk factors that comprise the Framingham stroke risk profile. In analyses adjusted for traditional risk factors, we observed an association between an aggregate panel of biomarkers including markers of inflammation (C-reactive protein[CRP]), hemostasis (D-dimer and plasminogen activator inhibitor-1), neurohormonal activity (aldosterone-to renin ratio, B-type natriuretic peptide[BNP] and N-terminal pro-atrial natriuretic peptides) and endothelial function (plasma homocysteine[tHcy] and urinary albumin/creatinine ratio[UACR]) and the risk of stroke/TIA as well as with MRI brain volumes. In further analyses, we identified that higher BNP levels and albuminuria were associated with the risk of stroke/TIA and addition of these biomarker data provided incremental information over clinical risk factors predicting stroke. We further indentified that albuminuria (but not BNP) was also associated with smaller brain volume, as were higher CRP, D-dimer and tHcy levels. Our findings are consistent with prior observations that alterations in several different biological pathways may underlie the pathophysiology of clinical and subclinical brain disease. However, whether the knowledge of BNP and albuminuria may be used to target individuals at risk of stroke for more intense primary prevention or monitoring requires further investigation.

Supplementary Material

01

Acknowledgments

Funding Sources: This work was supported by the National Institutes of Health/National Heart, Lung, and Blood Institute Contract N01-HC-25195, the National Institute on Aging (R01 AG16495; AG08122; AG033193, 031287, 033040, P30AG013846, P30 AG 10129) the National Institute of Neurological Disorders and Stroke (R01 NS17950), the National Institute of Diabetes and Digestive and Kidney Diseases (K24 DK080140), by a clinical research grant from the American Diabetes Association and by Roche Diagnostics who donated assay reagents for measurement of urinary albumin and creatinine. Dr. Debette was partially supported by an award grant from the Bettencourt-Schueller Foundation. The content is solely the responsibility of the authors and does not necessarily represent the official views of NINDS, NHLBI, NIA or NIH.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of Interest Disclosures: None

References

1. Heart Disease and Stroke Statistics Update 2012: A report from AHA Statistics Committee and Statistics Subcommittee. Circulation. 2012;125:e2–e220. [PubMed]
2. Longstreth WT., Jr Brain Vascular Disease Overt and Covert. Stroke. 2005;36:2062–2063. [PubMed]
3. Das RR, Seshadri S, Beiser AS, Kelly-Hayes M, Au R, Himali JJ, Kase CS, Benjamin EJ, Polak JF, O'Donnell CJ, Yoshita M, D'Agostino RB, Sr, DeCarli C, Wolf PA. Prevalence and correlates of silent cerebral infarcts in the Framingham offspring study. Stroke. 2008;39:2929–2935. [PMC free article] [PubMed]
4. Bryan RN, Wells SW, Miller TJ, Elster AD, Jungreis CA, Poirier VC, Lind BK, Manolio TA. Infarct like lesions in the brain: prevalence and anatomic characteristics at MR imaging of the elderly: data from the Cardiovascular Health Study. Radiology. 1997;202:47–54. [PubMed]
5. Seshadri S, Wolf PA, Beiser A, Elias MF, Au R, Kase CS, D'Agostino RB, DeCarli C. Stroke risk profile, brain volume, and cognitive function: the Framingham Offspring Study. Neurology. 2004;63:1591–1599. [PubMed]
6. Starr JM, Leaper SA, Murray AD, Lemmon HA, Staff RT, Deary IJ, Whalley LJ. Brain white matter lesions detected by magnetic resonance imaging are associated with balance and gait speed. J NeurolNeurosurg Psychiatry. 2003;74:94–98. [PMC free article] [PubMed]
7. Vermeer SE, Hollander M, van Dijk EJ, Hofman A, Koudstaal PJ, Breteler MM. Rotterdam Scan Study. Silent brain infarcts and white matter lesions increase stroke risk in the general population: the Rotterdam Scan Study. Stroke. 2003;34:1126–1129. [PubMed]
8. Jeerakathil T, Wolf PA, Beiser A, Massaro J, Seshadri S, D'Agostino RB, DeCarli C. Stroke risk profile predicts white matter hyperintensity volume: the Framingham Study. Stroke. 2004;35:1857–1861. [PubMed]
9. Rost NS, Wolf PA, Kase CS, Kelly-Hayes M, Silbershatz H, Massaro JM, D'Agostino RB, Franzblau C, Wilson PW. Plasma concentration of C-reactive protein and risk of ischemic stroke and transient ischemic attack: the Framingham study. Stroke. 2001;32:2575–2579. [PubMed]
10. Fornage M, Chiang YA, O'Meara ES, Psaty BM, Reiner AP, Siscovick DS, Tracy RP, Longstreth WT., Jr Biomarkers of Inflammation and MRI-Defined Small Vessel Disease of the Brain: The Cardiovascular Health Study. Stroke. 2008;39:1952–1959. [PMC free article] [PubMed]
11. Jefferson AL, Massaro JM, Wolf PA, Seshadri S, Au R, Vasan RS, Larson MG, Meigs JB, Keaney JF, Jr, Lipinska I, Kathiresan S, Benjamin EJ, DeCarli C. Inflammatory biomarkers are associated with total brain volume: the Framingham Heart Study. Neurology. 2007;68:1032–1038. [PMC free article] [PubMed]
12. Tomimoto H, Akiguchi I, Ohtani R, Yagi H, Kanda M, Shibasaki H, Yamamoto Y. The coagulation-fibrinolysis system in patients with leukoaraiosis and Binswanger disease. Arch Neurol. 2001;58:1620–1625. [PubMed]
13. Gottesman RF, Cummiskey C, Chambless L, Wu KK, Aleksic N, Folsom AR, Sharrett AR. Hemostatic factors and subclinical brain infarction in a community-based sample: the ARIC study. Cerebrovasc Dis. 2009;28:589–594. [PMC free article] [PubMed]
14. Marcheselli S, Micieli G. Renin-angiotensin system and stroke. Neurol Sci. 2008;29(Suppl 2):S277–S278. [PubMed]
15. Shibazaki K, Kimura K, Iguchi Y, Okada Y, Inoue T. Plasma brain natriuretic peptide can be a biological marker to distinguish cardioembolic stroke from other stroke types in acute ischemic stroke. Intern Med. 2009;48:259–264. [PubMed]
16. Wang TJ, Larson MG, Levy D, Benjamin EJ, Leip EP, Omland T, Wolf PA, Vasan RS. Plasma natriuretic peptide levels and the risk of cardiovascular events and death. N Engl J Med. 2004;350:655–663. [PubMed]
17. Takahashi T, Nakamura M, Onoda T, Ohsawa M, Tanno K, Itai K, Sakata K, Sakuma M, Tanaka F, Makita S, Yoshida Y, Ogawa A, Kawamura K, Okayama A. Predictive value of plasma B-type natriuretic peptide for ischemic stroke: a community-based longitudinal study. Atherosclerosis. 2009;207:298–303. [PubMed]
18. Madison JR, Spies C, Schatz IJ, Masaki K, Chen R, Yano K, Curb JD. Proteinuria and risk for stroke and coronary heart disease during 27 years of follow-up: the Honolulu Heart Program. Arch Intern Med. 2006;24:166, 884–889. [PubMed]
19. Yuyun MF, Khaw KT, Luben R, Welch A, Bingham S, Day NE, Wareham NJ. Microalbuminuria and stroke in a British population: the European Prospective Investigation into Cancer in Norfolk (EPIC-Norfolk) population study. J Intern Med. 2004;255:247–256. [PubMed]
20. Klausen KP, Scharling H, Jensen JS. Very low level of microalbuminuria is associated with increased risk of death in subjects with cardiovascular or cerebrovascular diseases. J Intern Med. 2006;260:231–237. [PubMed]
21. Knopman DS, Mosley TH, Jr, Bailey KR, Jack CR, Jr, Schwartz GL, Turner ST. Associations of microalbuminuria with brain atrophy and white matter hyperintensities in hypertensive sibships. J Neurol Sci. 2008;271:53–60. [PMC free article] [PubMed]
22. Bostom AG, Rosenberg IH, Silbershatz H, Jacques PF, Selhub J, D'Agostino RB, Wilson PW, Wolf PA. Nonfasting plasma total homocysteine levels and stroke incidence in elderly persons: The Framingham Study. Ann Intern Med. 1999;131:352–255. [PubMed]
23. Seshadri S, Wolf PA, Beiser AS, Selhub J, Au R, Jacques PF, Yoshita M, Rosenberg IH, D'Agostino RB, DeCarli C. Association of plasma total homocysteine levels with subclinical brain injury: cerebral volumes, white matter hyperintensity, and silent brain infarcts at volumetric magnetic resonance imaging in the Framingham Offspring Study. Arch Neurol. 2008;65:642–649. [PMC free article] [PubMed]
24. Kannel WB, Feinleib M, McNamara PM, Garrison RJ, Castelli WP. An investigation of coronary heart disease in families. The Framingham offspring study. Am J Epidemiol. 1979;110:281–290. [PubMed]
25. Wang TJ, Gona P, Larson MG, Tofler GH, Levy D, Newton-Cheh C, Jacques PF, Rifai N, Selhub J, Robins SJ, Benjamin EJ, D'Agostino RB, Vasan RS. Multiple biomarkers for the prediction of first major cardiovascular events and death. N Engl J Med. 2006;355:2631–2639. [PubMed]
26. Wolf PA, D'Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: a risk profile from the Framingham Study. Stroke. 1991;22:312–318. [PubMed]
27. DeCarli C, Massaro J, Harvey D, Hald J, Tullberg M, Au R, Beiser A, D'Agostino R, Wolf PA. Measures of brain morphology and infarction in the Framingham heart study: establishing what is normal. Neurobiol Aging. 2005;26:491–510. [PubMed]
28. Pencina MJ, D'Agostino RB, Sr, D'Agostino RB, Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med. 2008;27:157–172. [PubMed]
29. Pencina MJ, D'Agostino RB, Sr, Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med. 2011;30:11–21. [PMC free article] [PubMed]
30. Melander O, Newton-Cheh C, Almgren P, Hedblad B, Berglund G, Engström G, Persson M, Smith JG, Magnusson M, Christensson A, Struck J, Morgenthaler NG, Bergmann A, Pencina MJ, Wang TJ. Novel and conventional biomarkers for prediction of incident cardiovascular events in the community. JAMA. 2009;302:49–55. [PMC free article] [PubMed]
31. Arnlöv J, Evans JC, Meigs JB, Wang TJ, Fox CS, Levy D, Benjamin EJ, D'Agostino RB, Vasan RS. Low-grade albuminuria and incidence of cardiovascular disease events in nonhypertensive and nondiabetic individuals: the Framingham Heart Study. Circulation. 2005;112:969–975. [PubMed]
32. Jacques PF, Selhub J, Bostom AG, Wilson PW, Rosenberg IH. The effect of folic acid fortification on plasma folate and total homocysteine concentrations. N Engl J Med. 1999;340:1449–1454. [PubMed]
33. Sagnella GA. Practical implications of current natriuretic peptide research. J Renin Angiotensin Aldosterone Syst. 2000;1:304–315. [PubMed]
34. Di Angelantonio E, Chowdhury R, Sarwar N, Ray KK, Gobin R, Saleheen D, Thompson A, Gudnason V, Sattar N, Danesh J. B-type natriuretic peptides and cardiovascular risk: systematic review and meta-analysis of 40 prospective studies. Circulation. 2009;120:2177–2187. [PubMed]
35. Passino C, Severino S, Poletti R, Piepoli MF, Mammini C, Clerico A, Gabutti A, Nassi G, Emdin M. Aerobic training decreases B-type natriuretic peptide expression and adrenergic activation in patients with heart failure. J Am Coll Cardiol. 2006;47:1835–1839. [PubMed]
36. Mykkänen L, Zaccaro DJ, O'Leary DH, Howard G, Robbins DC, Haffner SM. Microalbuminuria and carotid artery intima-media thickness in nondiabetic and NIDDM subjects. The Insulin Resistance Atherosclerosis Study (IRAS) Stroke. 1997;28:1710–1716. [PubMed]
37. Clausen P, Jensen JS, Jensen G, Borch-Johnsen K, Feldt-Rasmussen B. Elevated urinary albumin excretion is associated with impaired arterial dilatory capacity in clinically healthy subjects. Circulation. 2001;103:1869–1874. [PubMed]
38. Deckert T, Feldt-Rasmussen B, Borch-Johnsen K, Jensen T, Kofoed-Enevoldsen A. Albuminuria reflects widespread vascular damage. The Steno hypothesis. Diabetologia. 1989;32:219–226. [PubMed]
39. Schmidt R, Schmidt H, Pichler M, Enzinger C, Petrovic K, Niederkorn K, Horner S, Ropele S, Watzinger N, Schumacher M, Berghold A, Kostner GM, Fazekas F. C-reactive protein, carotid atherosclerosis, and cerebral small-vessel disease: results of the Austrian Stroke Prevention Study. Stroke. 2006;37:2910–2916. [PubMed]
40. Engelhart MJ, Geerlings MI, Meijer J, Kiliaan A, Ruitenberg A, van Swieten JC, Stijnen T, Hofman A, Witteman JC, Breteler MM. Inflammatory proteins in plasma and the risk of dementia: the Rotterdam study. Arch Neurol. 2004;61:668–672. [PubMed]
41. Schmidt R, Schmidt H, Curb D, Masaki K, White LR, Launer LJ. Early inflammation and dementia: a 25-year follow-up of the Honolulu-Asia Aging Study. Ann Neurol. 2002;52:168–174. [PubMed]
42. McCully KS. Chemical pathology of homocysteine. IV. Excitotoxicity, oxidative stress, endothelial dysfunction, and inflammation. Ann Clin Lab Sci. 2009;39:219–322. [PubMed]
43. Ravera M, Ratto E, Vettoretti S, Viazzi F, Leoncini G, Parodi D, Tomolillo C, Del Sette M, Maviglio N, Deferrari G, Pontremoli R. Microalbuminuria and subclinical cerebrovascular damage in essential hypertension. J Nephrol. 2002;15:519–524. [PubMed]