BIOMARKERS OF LEFT VENTRICULAR HYPERTROPHY AND REMODELING IN BLACKS

Abstract

Left ventricular hypertrophy, a marker for adverse cardiovascular events, is more common in blacks than non-Hispanic whites. Mechanisms leading to left ventricular hypertrophy and mediating its clinical sequelae in blacks are not fully understood. We investigated the associations of 39 candidate biomarkers in distinct biological pathways with left ventricular mass and geometry in blacks. Participants included 1193 blacks (63 ± 9 years, 72% women, 78% hypertensive) belonging to hypertensive sibships. Left ventricular mass was measured by transthoracic echocardiography and indexed to height^2.7. Left ventricular geometry was categorized as: normal, concentric remodeling, concentric hypertrophy and eccentric hypertrophy. Generalized estimating equations were employed to assess associations of the 39 biomarkers with left ventricular mass index after adjustment for age, sex, and conventional risk factors. After adjustment for potential confounders, log-transformed levels of the following biomarkers were independently associated with left ventricular mass index: N-terminal pro-brain natriuretic peptide (β±SE= 0.07±0.01 pg/mL, P< 0.0001), mid-regional pro-atrial natriuretic peptide (β±SE= 0.08±0.02 pmol/L, P< 0.0001), mid-regional pro-adrenomedullin (β±SE= 0.09±0.03 nmol/L, P= 0.0006), C-terminal pro-endothelin (β±SE= 0.05±0.02 pmol/L, P=0.0009) and osteoprotegerin (β±SE=0.07±0.02 pg/mL, P=0.0005). The associations of these biomarkers with left ventricular mass index were mainly due to their association with eccentric hypertrophy. Higher circulating levels of natriuretic peptides, adrenomedullin, endothelin and osteoprotegerin were associated with increased left ventricular mass index, providing insights into the pathophysiology of left ventricular hypertrophy in blacks.

Introduction

The prevalence of hypertension among blacks in the United States is among the highest in the world, and it continues to increase.¹ Furthermore, hypertension-related mortality rates are three times higher in black (49%) compared to non-Hispanic white men (16%) and more then double in black (37%) compared to non-Hispanic white women (14%).¹ Hypertensive blacks are more likely to develop target-organ damage than their non-Hispanic white counterparts, including left ventricular hypertrophy (LVH),² renal failure ³ and stroke.⁴ The prevalence and severity of hypertension tends to be higher in blacks than non-Hispanic whites,¹ and LVH is more often present in blacks despite similar blood pressure levels and hypertension duration.² Further, LVH is an important prognostic factor in hypertensive individuals, and the association of LVH with mortality is even more pronounced in blacks than in other ethnic groups.⁵

Given the higher prevalence of LVH and the associated poor prognosis in blacks, identifying biological pathways that contribute to the pathogenesis and clinical sequelae of LVH in this ethnic group is important. Such knowledge may have significant clinical and therapeutic implications, including detection of individuals at higher risk. We hypothesized that biomarkers in pathways of inflammation, hemodynamic stress, lipoprotein metabolism, insulin resistance, and calcification are associated with LVH in blacks. We investigated the associations of 39 candidate biomarkers in these pathways with left ventricular (LV) mass and LV remodeling/ geometry in a cohort of blacks without a history of myocardial infarction or stroke.

Methods

Study participants

Study participants were part of the Genetic Epidemiology Network of Arteriopathy (GENOA) study, a community-based study aimed at identifying genetic variants influencing blood pressure (BP) levels and the development of target organ damage due to hypertension.⁶ Participants were ascertained from sibships in which at least two family members were diagnosed with hypertension before the age of 60 years. Between December 1^st, 2000, and October 31^st, 2004, 1324 black participants from Jackson, Mississippi completed the study protocol. We excluded 131 participants with history of myocardial infarction or stroke, and the remaining 1193 participants were included in the final analysis. The project was approved by the University of Mississippi’s Institutional Review Board and participants gave informed consent.

Assessment of baseline characteristics

Information about the participants’ baseline characteristics, medical history, medications and demographics was obtained from a comprehensive questionnaire administered on the day of the study by trained staff, as previously described.⁷ Race was self-reported. Hypertension was defined as a systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg at the study visit, or report of a prior diagnosis of hypertension and current treatment with anti-hypertensive agents. Diabetes was considered present if a subject was being treated with insulin or oral agents or had a fasting glucose level ≥ 7.0 mmol/L (126 mg/dL). ‘Ever’ smoking was defined as having smoked more than 100 cigarettes in the past. Weight was measured in kilograms by an electronic scale, and height was measured in centimeters by a stadiometer.

Assessment of LVH

We utilized echocardiographically-derived left ventricular mass index (LVMi) as the measure of LVH. Two-dimensional and Doppler echocardiography was performed in the left lateral decubitus position on an Acuson 128XP/10c machine (Acuson, Malvern, PA) using 2.5-, 3.5-, and 5.0-MHz transducers, and LV wall thickness and dimensions were measured as previously described.⁸ Two-dimensional left atrial diameter and left ventricular ejection fraction were obtained from the parasternal long axis view. LV mass was derived from the simplified cubed equation formula ⁹ and indexed to height^2.7 to normalize heart size to body size (LVMi = left ventricular mass/height^2.7).¹⁰ Increased LVMi was defined as LVMi > 44g/m^2.7 in women and LVMi > 48g/m^2.7 in men.¹¹ Relative wall thickness (RWT) was calculated as the sum of left ventricular septal and posterior wall thickness at end-diastole divided by LV internal dimension at end-diastole. A RWT > 0.42 was used to define increased RWT in men and women.¹¹ Patterns of LV geometry were defined as: (1) normal: normal LVMi and normal RWT < 0.44; (2) concentric remodeling: normal LVMi and high RWT; (3) concentric hypertrophy: high LVMi and high RWT; and (4) eccentric hypertrophy: high LVMi and normal RWT.

Circulating biomarkers

Thirty nine candidate biomarkers representing pathways of inflammation, hemodynamic stress, lipoprotein metabolism, insulin resistance and vascular calcification were measured in plasma or serum.¹² The biomarkers and the corresponding abbreviations are listed in Table 1. Assays for the individual biomarkers, including precision, accuracy and stability, as well as methods for quality control have been previously described in detail.¹²

Table 1.

Circulating biomarkers, organized by etiologic pathway, and their abbreviations

Pathway	Biomarkers
Inflammation	C-reactive protein (CRP), Serum amyloid A (SAA), Intercellular adhesion molecule (ICAM), Vascular cell adhesion molecule (VCAM), Interleukin-6 (IL-6), Interleukin-8 (IL-8), Tumor necrosis factor receptor-1 (TNFRI), Tumor necrosis factor receptor-2 (TNFRII), Monocyte chemotactic protein-1 (MCP-1), E-selectin, P-selectin, Heat shock protein 27 (Hsp27), Myeloperoxydase (MPO), Receptor for advanced glycation end-products (RAGE), Matrix metalloproteinase-2 (MMP-2), Matrix metalloproteinase-9 (MMP-9), Tissue inhibitor of metalloproteinases-1 (TIMP-1), Tissue inhibitor of metalloproteinases-2 (TIMP-2).
Lipoprotein metabolism	Apolipoprotein A-I (ApoA-I), Apolipoprotein B (ApoB), Apolipoprotein C-III (ApoC-III), Apolipoprotein E (ApoE), Low-density lipoprotein particle size (LDL size), Lipoprotein (a) [Lp(a)], Oxidized low-density lipoprotein (Ox-LDL), Lipoprotein-associated phospholipase A2 mass (Lp-PLA2 mass), Lipoprotein-associated phospholipase A2 activity (Lp-PLA2 activity)
Hemodynamic stress	N-terminal pro-brain natriuretic peptide (NT-proBNP), Mid- regional pro-atrial natriuretic peptide (MR-proANP), C- terminal pro-arginine vasopressin (CT-proAVP), Mid- regional pro-adrenomedullin (MR-proADM), C-terminal pro-endothelin (CT-proET)
Calcification	Osteopontin (OPN), Osteoprotegerin (OPG), Osteonectin (ONN), Osteocalcin (OCN)
Insulin resistance	Leptin, Adiponectin, Resistin

Statistical methods

The mean levels of the significant biomarkers across the four patterns of LV geometry were compared with one-way ANOVA. Biomarkers levels and LVMi were log-transformed to reduce skewness. We employed linear regression to assess the associations of each of the 39 biomarkers with LVMi after adjustment for age and sex (Model 1); after further adjustment for potential confounders (hypertension, diabetes, weight, smoking, total and HDL cholesterol, estimated glomerular filtration rate (eGFR), and use of aspirin, statins, estrogens, and anti-hypertensives) (Model 2); and additional adjustment for left ventricular ejection fraction and left atrial diameter (Model 3). Regression analyses were performed using generalized estimating equations to account for potential correlation among siblings in the sample. Model selection was performed using a ‘forward’ approach with criteria of P≤0.10 to enter and P≤0.05 to stay in the model. We then created a separate multivariable model including all the significant biomarkers in the same model, while adjusting for the above covariates. Similar analyses were also performed in the subset of hypertensive individuals (n = 936). In order to reduce the possibility of type I error caused by multiple testing, we employed the Bonferroni correction, and a P-value ≤ 0.001 (0.05 ÷ 39, since we analyzed 39 different biomarkers) was considered to be statistically significant.

In addition, multivariate logistic regression was employed to assess the associations of each significant biomarker with the presence of LVH and each of the three patterns of abnormal LV geometry, after adjustment for covariates in Models 2 and 3 above. For these analyses, adjusted odds ratios (OR) and 95% confidence intervals (95% CI) for 1 logarithmic unit increase in biomarker levels were calculated. A P-value ≤ 0.01 was considered statistically significant (0.05 ÷ 5, since only five biomarkers were assessed at this level) for the logistic regression analyses.

To assess the discrimination power of various statistical models for predicting presence of LVH and each abnormal LV geometry pattern, we calculated the c-statistic for models that included: (i) age and sex, (ii) age, sex and significant biomarkers, (iii) age, sex and conventional risk factors (hypertension, diabetes, systolic and diastolic BP, weight, smoking, total and HDL cholesterol, eGFR), and (iv) age, sex, conventional risk factors and significant biomarkers. The incremental utility of adding the significant biomarkers to conventional risk factors for predicting presence of LVH was evaluated with the net reclassification index as previously described.¹³

Statistical analyses were performed with SAS vs. 9.0 (SAS Institute Inc., Cary NC).

Results

The baseline characteristics of the study participants are described in Table 2. The distributions of the 39 candidate biomarkers are summarized in Supplemental Table S1 (please see http://hyper.ahajournals.org).

Table 2.

Baseline characteristics of the participants (n = 1193)

Variable	Mean	± SD
	n	%
Age, years	63.26	9.29
Men, n (%)	335	28.1%
Body mass index, kg/m2	31.60	6.77
Hypertension, n (%)	936	78.5%
Systolic blood pressure, mmHg	138	20
Diastolic blood pressure, mmHg	79	10
Diabetes, n (%)	341	28.6%
History of smoking (past or current), n (%)	462	38.7%
Total cholesterol, mmol/L*	5.23	1.06
HDL cholesterol, mmol/L*	1.49	0.47
Estimated glomerular filtration rate (mL/s/1.73 m2) †	1.27	0.33
Statin use, n (%)	205	17.2%
ACEi/ARB use, n (%)	446	37.4%
Beta blocker use, n (%)	177	14.8%
Calcium channel blocker use, n (%)	331	27.8%
Diuretic use, n (%)	536	44.9%
Aspirin use (%)	365	30.6%
Echocardiographic variables:
Ventricular septal thickness, mm	8.9	1.3
Posterior wall thickness, mm	7.4	1.3
Left ventricular end-diastolic diameter, mm	51.6	5.0
Left ventricular end-systolic diameter, mm	34.7	5.5
Relative wall thickness, mm	0.32	0.05
Increased relative wall thickness present, n (%)	33	2.8%
Left ventricular mass, g (median and IQR)	148.1	126.7–181.4
Left ventricular mass index, g/m2.7 (median and IQR)	37.4	31.9–45.0
Left ventricular hypertrophy present, n (%)	288	24.8%
Patterns of left ventricular geometry:
Normal, n (%)	856	73.7%
Concentric remodeling, n (%)	18	1.5%
Concentric hypertrophy, n (%)	22	1.9%
Eccentric hypertrophy, n (%)	266	22.9%

ACEi: angiotensin converting enzyme inhibitor. ARB: angiotensin receptor blocker.

eGFR: estimated glomerular filtration rate. HDL: high-density lipoprotein. SD: standard deviation

^*To convert total and HDL cholesterol to mg/dL, multiple the value by 38.67

^†To convert estimated glomerular filtration rate to ml/min/1.73m², divide the value by 0.01667.

After adjustment for age and sex, higher levels of the following biomarkers were associated with greater LVMi: CRP, IL-6, E-selectin, leptin, resistin, NT-proBNP, MR-proANP, CT-proAVP, MR-proADM, CT-proET and OPG (Table 3). After further adjustment for potential confounders, higher NT- proBNP, MR-proANP, MR-proADM, CT-proET and OPG remained independently associated with greater LVMi (Table 3). The associations of NT-proBNP, CT-proET and OPG with LVMi were independent of left ventricular ejection fraction and left atrial size (Table 3). Lastly, in the multivariable model that included all the significant biomarkers, higher NT-proBNP (β±SE: 0.06 ± 0.01, P <0.0001) and OPG (β±SE: 0.06 ± 0.02, P = 0.0013) remained independently associated with greater LVMi. When hypertensive participants were analyzed separately (n = 936), similar results were observed (Supplemental Table S2, please see http://hyper.ahajournals.org).

Table 3.

Multivariable linear regression to predict log left ventricular mass index (1 g/m^2.7) in the entire cohort (n = 1193)

	Model 1*		Model 2†		Model 3‡
Biomarker	β ± SE	P-value	β ± SE	P-value	β ± SE	P-value
Log CRP (mg/L)	0.04 ± 0.006	<0.0001	0.02 ± 0.006	0.005	0.02 ± 0.006	0.01
Log IL-6 (pg/mL)	0.05 ± 0.01	< 0.0001	0.02 ± 0.01	0.09	0.02 ± 0.01	0.13
Log E-selectin (ng/mL)	0.09 ± 0.02	0.0005	0.04 ± 0.02	0.08	0.03 ± 0.02	0.18
Log Leptin (ng/mL)	0.07 ± 0.01	< 0.0001	−0.01 ± 0.01	0.28	−0.01 ± 0.01	0.53
Log Resistin (ng/mL)	0.06 ± 0.01	< 0.0001	0.03 ± 0.01	0.05	0.02 ± 0.01	0.07
Log NT-proBNP (pg/mL)	0.08 ± 0.01	< 0.0001	0.07 ± 0.01	< 0.0001	0.05 ± 0.009	<0.0001
Log MR-proANP (pmol/L)	0.07 ± 0.02	< 0.0001	0.08 ± 0.02	< 0.0001	0.04 ± 0.01	0.006
Log CT-proAVP (pmol/L)	0.04 ± 0.01	0.0002	0.02 ± 0.01	0.14	0.01 ± 0.01	0.20
Log MR-proADM (nmol/L)	0.17 ± 0.02	< 0.0001	0.09 ± 0.03	0.0006	0.06 ± 0.02	0.007
Log CT-proET (pmol/L)	0.08 ± 0.02	< 0.0001	0.05 ± 0.02	0.0009	0.05 ± 0.02	0.001
Log OPG (pg/ml)	0.09 ± 0.02	< 0.0001	0.07 ± 0.02	0.0005	0.06 ± 0.02	0.0009

^*Model 1 was adjusted for age and sex.

^†Model 2 was adjusted for age, sex, hypertension, diabetes, smoking history, weight, total cholesterol, HDL cholesterol, estimated glomerular filtration rate, and use of aspirin, statins, estrogens, beta-blockers, calcium channel blockers, diuretics, and angiotensin converting enzyme inhibitors/ angiotensin receptor blockers

^‡Model 3 was adjusted for the same variables as Model 2, plus left ventricular ejection fraction and left atrial size A P-value ≤ 0.001 was considered statistically significant.

The levels of all five biomarkers associated with LVMi were higher among participants with concentric hypertrophy than in participants with other LV geometric patterns (Supplemental Figure S1, please see http://hyper.ahajournals.org. P-value for trend <0.0001 for each of the five biomarkers).

In multivariable logistic regression analyses, a 1 unit increase in log NT-pro BNP (OR: 1.63, 95% CI: 1.31, 2.02. P<0.0001), log MR-proANP (OR: 1.52, 95% CI: 1.13, 2.06. P=0.006), log MR-proADM (OR: 1.86, 95% CI: 1.13, 3.04. P=0.01), and log CT-proET (OR: 1.72, 95% CI: 1.19, 2.51. P=0.004) were each significantly associated with greater odds of having LVH. After further adjustment for left ventricular ejection fraction and left atrial size, NT-proBNP remained significantly associated with LVH (OR: 1.36, 95% CI: 1.09, 1.70. P=0.007).

Further, log NT-proBNP (OR: 1.52, 95% CI: 1.23, 1.88. P=0.0001), log MR-proANP (OR: 1.46, 95% CI: 1.09, 1.97. p=0.01) and log CT-proET (OR: 1.71, 95% CI: 1.19, 2.47. P=0.004) were also associated with the presence of eccentric hypertrophy after adjustment for age, sex and other confounders, but not after adjustment for left atrial size and ejection fraction. None of the biomarkers were independently associated with concentric remodeling or concentric hypertrophy, although the number of participants in these groups was low.

The discriminatory power of various models to predict LVH is summarized in Table 4. Adding the five significant biomarkers to age, sex and conventional risk factors improved the prediction of LVH and each abnormal left ventricular geometry pattern. The net reclassification index for the multivariable models that included the five significant biomarkers was 10.5% (P=0.007).

Table 4.

C-statistics of the receiver-operating characteristic curves for prediction of left ventricular hypertrophy (left ventricular mass index > 44 g/m^2.7 in women and > 48 g/m^2.7 in men) and the patterns of abnormal left ventricular geometry.

		Abnormal left ventricular geometry patterns
Predictors	LVH	Concentric remodeling	Concentric hypertrophy	Eccentric hypertrophy
Age and sex	0.64	0.74	0.76	0.63
Age, sex and conventional risk factors*	0.75	0.84	0.89	0.75
Age, sex, conventional risk factors*, NT-proBNP, MR-proANP, MR-proADM, CT-proET and OPG	0.77	0.89	0.90	0.77

^*Conventional risk factors included in the models were: history of hypertension, diabetes and smoking, weight, total and high-density lipoprotein cholesterol, estimated glomerular filtration rate, systolic and diastolic blood pressure.

LVH: Left ventricular hypertrophy

Discussion

The main finding of our study is that higher levels of vasoactive peptides (NT-proBNP, MR-proANP, MR-proADM, CT-proET) and OPG were independently associated with greater LVMi in blacks. Addition of these five biomarkers to age, sex and conventional risk factors improved the ability to discriminate between those with and without LVH. The associations of these biomarkers with LVMi appeared to be driven by their associations with eccentric hypertrophy. The absence of an independent association of the biomarkers with concentric hypertrophy may be due to the small number of participants (22 out of 1193) who had this left ventricular geometric pattern. To the best of our knowledge, our study is the first to employ a multi-marker approach in the study of LVH and adverse LV remodeling in blacks. Our findings are relevant for understanding potential pathophysiologic mechanisms associated with LVH and adverse LV remodeling in this ethnic group.

Hemodynamic stress and left ventricular hypertrophy

Higher circulating levels of four markers of hemodynamic stress, NT-proBNP, MR-proANP, MR-proADM and CT-proET were independently associated with greater LVMi and presence of LVH in blacks. Furthermore, NT-proBNP, MR-proANP and CT-proET were associated with the presence of eccentric hypertrophy. Although the associations of brain natriuretic peptide (BNP) and its precursors with LVMi and LVH have been previously reported, our study is the first to focus on a black cohort. BNP, atrial natriuretic peptide (ANP) or related precursors appear to be stronger predictors of LVH in higher-risk populations ^14–17 than in the general population.^{18, 19} Similar to results from our study, levels of BNP were also higher among individuals with concentric hypertrophy in the Framingham Offspring Study. ²⁰ In contrast, in studies of elderly men (age ≥70), N-terminal ANP ²¹ and BNP ²² levels were the highest among those with eccentric hypertrophy. The clinical utility of BNP and NT-proBNP in hypertensives appears to go beyond prediction of LVH, as these markers are also associated with an adverse prognosis independent of LVH, ²³ highlighting the potential use of the natriuretic peptides for identifying blacks at higher risk of adverse cardiovascular events.

The natriuretic peptides offset the deleterious effects of volume overload by improving myocardial relaxation and counter-acting the sodium restriction and vasoconstriction promoted by the renin-angiotensin-aldosterone system.²⁴ In the setting of hypertensive heart disease, increased wall stress due to myocardial hypertrophy and pressure overload and elevated LV filling pressures may lead to ventricular and atrial production of ANP and BNP, thereby promoting vasodilation, diuresis and natriuresis to offset the deleterious effects of LVH and its hemodynamic consequences.²⁵

Contemporary studies in patients with heart failure suggest that the assays commonly used to quantify circulating BNP and NT-proBNP also capture forms of the peptide that are not biologically active,²⁶ introducing the concept of BNP dysfunction/ resistance. We can speculate that reduced biological activity of the natriuretic peptides in pathological conditions such as heart failure and LVH may lead to decrease in the compensatory natriuresis and vasodilation, resulting in increased production of NT-proBNP and MR-proANP. However, additional studies are required to prove this hypothesis.

Two previous studies found higher adrenomedullin levels to be associated with LV mass in humans, ^{27, 28} but this is the first report of associations of adrenomedullin with patterns of LV remodeling. Adrenomedullin, a peptide abundantly expressed in cardiovascular tissues, ²⁹ reduces peripheral arterial resistance and BP. ³⁰ In animal models, adrenomedullin levels increased with LV mass after aortic banding,³¹ while in salt-sensitive hypertensive rats, adrenomedullin gene delivery resulted in sustained decreases in BP and LV mass.³² Elevated adrenomedullin levels may therefore represent a compensatory mechanism aimed at lowering BP and preventing LV remodeling in the setting of hypertension. The potential of adrenomedullin as therapy for hypertension and LVH needs further investigation.

We previously reported CT-proET to be independently associated with LVMi, ⁸ and similar findings were also observed by Hua et al.³³ In contrast, no significant association was present in the Losartan Intervention for Endpoint Reduction (LIFE) sub study.³⁴ Endothelin, an endothelium-derived peptide, is one of the most potent vasoconstrictors known, and its production is stimulated by angiotensin II in vivo.³⁵ Potential mechanisms linking CT-proET to LVH have been reviewed.⁸ It is possible that the association of CT-proET with LVMi may be in part due to its adverse hemodynamic effects ³⁵ and also a direct role in myocyte hypertrophy.³⁶ Whether endothelin receptor blockade attenuates or prevents development of LVH in hypertensives needs further investigation.

In summary, elevations of NT-proBNP, MR-proANP and MR-proADM levels may represent compensatory mechanisms aimed at lowering BP and pre-load and decreasing left ventricular mass, while elevated CT-proET may promote development of hypertension and LVH. It is possible that the associations found may be due to a hemodynamic effect of LVH rather than being specific to blacks. The cross-sectional nature of our study does not allow us to make inferences regarding the causality or temporality of the associations.

Osteoprotegerin and left ventricular hypertrophy

OPG was independently associated with LVMi and the presence of LVH in blacks. Although OPG levels were highest among individuals with concentric hypertrophy, OPG was not independently associated with concentric LVH, likely due to the low number of participants with concentric LVH (22 out of 119). We have previously reported OPG to be independently associated with LVMi in 898 hypertensive blacks.³⁷ Omland and colleagues studied 2715 subjects (mean age 44 years) with a low prevalence of hypertension and found that OPG was independently associated with LVMi in men, but not in women. Similarly, in 100 patients with coronary artery disease, OPG was correlated with increased LVMi, but not after adjustment for age.³⁸ OPG has been shown to be associated with cardiovascular diseases,^{39, 40} however our observation of an independent association of OPG with LVMi and eccentric hypertrophy is novel. Potential mechanisms linking the OPG/RANKL system to LVH have been reviewed.³⁷ The results from our study motivate further investigation to elucidate the pathophysiologic basis of the association of OPG with LVH in blacks and the potential use of OPG as a maker of adverse cardiovascular outcomes.

Strengths and limitations

We investigated the associations of a panel of multiple biomarkers in several candidate biologic pathways with LVMi in a relatively large sample of blacks. This ethnic group is at higher risk for hypertension, LVH and adverse cardiovascular events, and has been under-represented in research studies. Additional studies will be needed to confirm whether these associations are present in other ethnic groups. The cross-sectional associations observed in the present study do not allow us to establish causality or temporality. NT-proBNP and MR-proANP are known to be produced by the ventricles whereas factors leading to higher MR-proADM, CT-proET and OPG levels need to be established. Given the small number of participants with concentric remodeling and concentric hypertrophy, our study may have been underpowered to detect significant associations between biomarkers and these left ventricular remodeling patterns.

Perspectives.

We have demonstrated that higher circulating levels of NT-proBNP, MR-proANP, MR-proADM, CT-proET and OPG are independently associated with greater LVMi in a large cohort of blacks enriched for hypertension, mainly due to their association with eccentric LVH. Furthermore, adding these biomarkers to age, sex and conventional risk factors improved the discrimination between those with and without LVH, highlighting their potential clinical value. The results of our study are relevant for better understanding the pathophysiology of LVH and its clinical sequelae in blacks, and motivate further investigation to disentangle the roles of natriuretic peptides, adrenomedullin, endothelin and OPG on cardiac hypertrophy. Furthermore, since LVH is associated with poor prognosis in hypertensives, these biomarkers may aid in identification of individuals at higher risk of adverse cardiovascular events. Additional studies are necessary to establish whether the natriuretic peptide, adrenomedullin, endothelin and the OPG/ RANKL systems are targets for drug therapy to prevent/reverse LVH.