Cystatin C and cardiac hypertrophy in primary hypertension

Abstract

Introduction. Cystatin C is a marker of kidney function and a predictor of cardiovascular morbidity and mortality. It is unknown whether this protein may be related to the cardiac involvement that is common among patients with essential hypertension. Patients and methods. We evaluated the relationship between serum cystatin C, serum creatinine, estimated glomerular filtration rate and cardiac structure assessed by echocardiography, in a group of 49 non-diabetic patients with primary hypertension and normal serum creatinine. Results. Mean cystatin C levels were 0.74 ± 0.15 mg/l. Age, body mass index, triglycerides and creatinine, estimated glomerular filtration rate and left ventricular mass index were independently associated with cystatin C levels. Seventy three per cent of patients had cardiac hypertrophy. The prevalence of left ventricular hypertrophy was higher in patients who had cystatin C levels above the 70th percentile (0.79 mg/dl) than patients below this percentile (93.3% vs 66.7%, respectively, p = 0.04). Serum cystatin C (β = 0.48, p = 0.009), but not serum creatinine nor estimated glomerular filtration rate, was independently related to left ventricular mass index in a logistic regression analysis. Conclusion. Cystatin C is closely related to left ventricular mass in hypertensive patients, and could be a marker for cardiac hypertrophy in these patients.

Key Words::

• Cardiac hypertrophy
• cystatin C
• essential hypertension
• kidney function

Introduction

Cystatin C is a low molecular weight protein, produced constantly by nucleated cells. It is filtered through the glomerulus, and reabsorbed and broken down by tubular cells, without undergoing tubular secretion (1). Cystatin C is considered an endogenous marker of renal function, and has been shown to be more sensitive than creatinine in detecting slight to moderate decreases in the glomerular filtration rate (GFR), in theory partly because of its relative lack of dependence on demographic factors and muscle mass (2–4). However, recent studies have shown that cystatin C levels may also be affected by various clinical and demographic factors (5,6).

Epidemiology studies have shown that cystatin C is better than creatinine or the GFR based on creatinine-derived equations at predicting cardiovascular morbidity and mortality (7–10), particularly in elderly patients (9–11) and in general populations with no known chronic kidney disease (12). It has therefore been suggested that the relationship between cystatin C and cardiovascular risk may be independent of renal function (13), and that cystatin C may be a marker of cardiac involvement in patients with hypertension (14).

The aim of our study was to evaluate cystatin C as a predictor of cardiac hypertrophy in essential arterial hypertension, and to compare the predictive value of cystatin C with that of creatinine levels and the estimated GFR in the hypertensive population.

Patients and methods

Study population

We performed a cross-sectional study of 49 consecutive attending outpatients recruited from the Nephrology Service of the Joan XXIII University Hospital in Tarragona, Spain. All patients were non-diabetic, aged between 30 and 60 years, and had essential hypertension with normal serum creatinine (creatinine <1.3 mg/dl).

All patients gave informed consent and the study was approved by the Ethics Committee of Joan XXIII University Hospital.

Methods

All patients underwent the following procedures:

(i) Measurement of demographic, clinical and therapeutic variables, including body mass index and clinic blood pressure measured according to standard procedure, using an automated oscillometric monitor (Omron 705 CP, Healthcare GmbH, Hamburg, Germany). Patients’ blood pressure was considered controlled if their blood pressure was < 140/90 mmHg.

(ii) Conventional analytical tests, including serum glucose, total cholesterol, high-density lipoprotein (HDL) and low-density lipoprotein (LDL)-cholesterol, triglyceride, albumin, haemoglobin and fibrinogen.

(iii) Measurement of urinary albumin excretion over 24 h using an immunonephelometric assay (Behring Institute). Microalbuminuria was defined as urinary albumin of 30–300 mg/day.

(iv) Measurement of renal function parameters: serum creatinine levels (modified Jaffé color-imetric method) and estimated GFR using the abbreviated formula (four variables) derived from serum creatinine values (15).

(v) Measurement of serum cystatin C levels, using an immunonephelometric assay with intensifying particles (N Latex Cystatin C, Dade Behring BN) in a nephelometer (Dade Behring BN II) (16). The reference values were 0.53–0.95 mg/l. The sensitivity of the assay was 0.05 mg/l.; the intra- and interassay coefficients of variation were less than 3.1% and 3.5%, respectively.

(vi) Measurement of inflammatory markers: high-sensitivity interleukin-6 (IL-6) and high-sensitivity C-reactive protein (HS-CRP). IL-6 was measured using the non-competitive ELISA method (sandwich enzyme-linked immunosorbent assay). The R&D Systems Human IL6 Quantikine HS (High-Sensitivity) kit was used. The sensitivity of the method was 0.016 pg/ml, and the intra- and interassay coefficients of variation were less than 7.8% and 9.6%, respectively. HS-CRP was measured in serum samples using an immunonephelometric assay (N High-Sensitivity CRP) in a nephelometer (Dade Behring BN II). Sensitivity was 0.175 mg/l. The intra- and interassay coefficients of variation were less than 4.4% and 5.7%, respectively.

(vii) Echocardiographic studies: standard two-dimensional and two-dimensionally guided M-mode echocardiography were performed using a Diasonic 800 (Vingmed Sound, Horten, Norway), with a 3.5-MHz transducer. All studies were performed and analysed by the same experience echocardiographer (A.B.). Two-dimensionally guided M-mode echocardiograms were obtained following the American Society of Echocardiography guidelines (17,18). Measurements included left ventricular internal diameter at end-diastole (LVIDD) and end-systole (LVISD), left ventricular septal thickness (IVS) and posterior wall thickness at end-diastole (PWT). Two-dimensional studies were performed from the parasternal long-axis and short-axis views, apical four-chamber and apical long-axis two-chamber views, and subcostal views.

Left ventricular mass (LVM) was calculated according to the following formula:

LVM (g) = 0.8 × 1.04 ((LVIDD + IVS + PWT)³ − LVIDD³)+0.6.

Relative wall thickness (RWT) was calculated from the following formula:

RWT = 2 × PWT/LVIDD.

LVM index was obtained by correcting for height^2.7, the power that linearizes the relation between LVM and body height (19). Left ventricular hypertrophy (LVH) was defined as LVM index greater than ≥49.2 g/m^2.7 in men, and ≥46.7 g/m^2.7 in women (20).

Criteria for reproductibility of our echocardiographic measurements have been reported elsewhere (21).

Statistical analysis

Statistical analysis was carried out using the SPSS/ PC+ (IMC, Chicago, IL), version 11.5. Values were expressed as mean±standard deviation for normally distributed variables.

HS-CRP and IL-6 levels did not show normal distribution; therefore, log-transformed values were used in the analysis.

The univariate relation between variables was assessed by means of Pearson or Spearman correlation coefficients. Multiple linear regression was applied to analyse the relationship between variables and the confidence interval was 95%. Values of p less than 0.05 were considered statistically significant.

Results

Demographic, clinical and analytical information is shown in Table I. The mean age of patients was 50 years. All were treated with anti-hypertensive drugs, with an average of 2.5 drugs. Ninety-eight per cent of patients were treated with angiotensin-converting enzyme inhibitors or angiotensin receptor antagonists, and only 24 patients (49%) showed satisfactory control of their blood pressure. The overall prevalence of microalbuminuria was 18% (n=9). Mean estimated GFR was 79.3±15.6 ml/min/1.73 m², and only three patients had values lower than 60 ml/min/1.73 m²: 59, 58.7 and 57.3 ml/min/1.73 m², respectively.

Table I. Demographic, clinical and echocardiographic data of patients.

Clinical characteristics
Age	50.3±8.1
Gender (male/female)	37/12
Body mass index (kg/m^2)	28.8±4.2
Systolic blood pressure (mmHg)	139.3±20.4
Diastolic blood pressure (mmHg)	86±11.8
Pulse pressure (mmHg)	53.3±13.7
Laboratory fndings
Serum creatinine (mg/dl)	1.0±0.1
Estimated GFR (ml/min/1.73 m^2)	79.3±15.6
Serum cystatin C (mg/l)	0.74±0.1
Urinary albumin excretion (mg/day)	10.6 (0.75–152)
Total serum cholesterol (mg/dl)	196±28.6
HDL cholesterol (mg/dl)	49.8±11.5
LDL cholesterol (mg/dl)	119.8±23.5
Serum triglycerides (mg/dl)	113.7±54.2
Serum glucose (mg/dl)	94.2±12
Haemoglobin (g/dl)	14.5±1.1
Fibrinogen (mg/dl)	373.7±98.9
Serum albumin (g/l)	43.7±3.2
C-reactive protein (mg/l)	1.4 (0.26–34.7)
Interleukin-6 (pg/ml)	1.3 (0.39–6.57)
Echocardiographic data
LVIDD (mm)	31.1±0.59
LVISD (mm)	49±0.53
IVS (mm)	12±0.18
PWT (mm)	12±0.20
RWT	0.49±0.08
LVM index (g/m^2.7)	57.3±16.2

GFR, glomerular filtration rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LVIDD, left ventricular internal diameter at end-diastole; LVISD, left ventricular internal diameter at end-systole; IVS, left ventricular septal thickness; PWT, posterior wall thickness at end-diastole; RWT, relative wall thickness; LVM, left ventricular mass.

Cystatin C, demographic and clinical factors (Table II)

Table II. Univariate correlates of cystatin C and clinical, laboratory and echocardiographic data.

	Correlation coefficient	p
Clinical characteristics
Age	0.439	0.002
Body mass index	0.430	0.002
Systolic blood pressure	0.067	ns
Diastolic blood pressure	0.040	ns
Laboratory findings
Serum creatinine	0.312	0.03
Estimated GFR	−0.381	0.008
Log urinary albumin excretion	0.093	ns
Total serum cholesterol	0.035	ns
HDL cholesterol	−0.181	ns
LDL cholesterol	−0.118	ns
Serum triglycerides	0.415	0.004
Serum glucose	−0.075	ns
Serum albumin	−0.379	0.01
Haemoglobin	−0.165	ns
Fibrinogen	0.218	ns
Log C-reactive protein	0.31	0.02
Log interleukin-6	0.154	ns
Echocardiographic data
LVIDD	0.02	ns
LVISD	0.06	ns
IVS	0.37	0.009
PWT	0.31	0.03
RWT	0.25	0.08
LVM index	0.43	0.002

GFR, glomerular filtration rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein; LVIDD, left ventricular internal diameter at end-diastole; LVISD, left ventricular internal diameter at end-systole; IVS, left ventricular septal thickness; PWT, posterior wall thickness at end-diastole; RWT, relative wall thickness; LVM, left ventricular mass.

Cystatin C levels were 0.74±0.15 mg/1. Thirty-one per cent of patients had values >0.79 mg/1 (the 70th percentile of our sample). Cystatin C was significantly correlated with age, body mass index, serum triglyceride levels, albumin (negatively), log CRP, creatinine, estimated GFR and LVM index. The predicting factors independent of cystatin C levels were age, triglycerides, estimated GFR and LVM index (Table III).

Table III. Multiple linear regression results for cystatin C levels.

Variable	Beta coefficient	p
Serum triglycerides	0.241	0.045
Log C-reactive protein	0.133	ns
Body mass index	0.173	ns
Age	0.300	0.013
LVM index	0.262	0.022
GFR estimated	−0.265	0.025

R²=0.572. LVM, left ventricular mass; GFR: glomerular filtration rate.

Cystatin C and cardiac mass

Echocardiographic results are reported in Table I. Thirty-six patients (73%) had LVH. The prevalence of cardiac hypertrophy was greater in patients with cystatin C values > 0.79 mg/dl (70th percentile of the serum cystatin levels) than in patients with lower cystatin C values (93.3% vs 66.7%, respectively, p=0.04). Cystatin C was significantly correlated with left ventricular posterior wall thickness, interventricular septum thickness and LVM index (Table II). No correlation was observed between serum creatinine levels or estimated GFR and LVM index. When multiple linear regression analysis was performed to determine which variables were independently related to LVM index, only cystatin C levels, after adjustment for age, sex, body mass index, systolic blood pressure and estimated GFR, were found to be significantly related (Table IV).

Table IV. Multiple lineal regression results for left ventricular mass index.

Variable	Beta coefficient	p
Body mass index	−0.027	ns
Age	0.050	ns
Gender	−0.192	ns
Systolic blood pressure	0.375	0.06
GFR estimated	0.164	ns
Cystatin C levels	0.487	0.009

R²=0.298. GFR, glomerular filtration rate.

Discussion

In the present study, a significant proportion of our hypertensive patients with normal serum creatinine showed mild renal abnormalities, i.e. microalbuminuria and a slight reduction in estimated GFR. In these patients, the prevalence of cardiac hypertrophy was greater in patients with high cystatin C levels, and unlike serum creatinine or estimated GFR, cystatin C was independently correlated with LVM, suggesting that cystatin C may be a marker of cardiac hypertrophy in this population. In our study, overall LVH prevalence was very high, probably related to a selection bias in a referral centre.

It is well known that increased cystatin C levels are associated with a higher probability of cardiovascular complications in patients with or without coronary disease and patients with acute coronary syndromes (9,22–24). Furthermore, cystatin C levels are better than serum creatinine levels as a predictive factor of cardiovascular diseases, cardiac insufficiency and mortality (9–11). This fact is more pronounced in patients with an estimated GFR above 60 ml/min; cystatin C therefore maintains a marked association with cardiovascular risk in subjects with GFRs above 60 ml/min, which is not observed with creatinine levels (12). Thus, it is seems that cardiovascular risk is not completely captured by the standard measurements of renal function used in clinical practice, and that cystatin C may be a better predictor of cardiovascular risk.

One issue that has not been studied in depth is whether cystatin C may also be associated with the structural and functional cardiac anomalies present in hypertensive patients with normal renal function and in the initial stages of chronic kidney disease (25). In a study carried out in young patients with chronic kidney disease stage 2–4, cystatin C, unlike creati-nine and GFR, was independently related to diastolic dysfunction (26). In the Heart and Soul Study carried out in patients with coronary disease but without clinical cardiac insufficiency, cystatin C levels were related to LVH and diastolic dysfunction. This relationship was stronger than that observed with estimated GFR (27). Watanabe et al. (14) show that in patients with essential hypertension, cystatin C is an early marker of the severity of damage to target organs related to hypertension, such as LVM index, carotid intima media thickness and microalbuminuria. In this latter study, cystatin C was directly associated with mean 24-h systolic blood pressure, suggesting that the relationship between cystatin C and LVM index may have been dependent on patients’ systolic blood pressure.

The results of our study suggest that cystatin C may be a marker of the development of cardiac hypertrophy in the hypertensive population, unlike other markers of renal function such as creatinine or estimated GFR. It is not clear whether the association between cystatin C and LVM reflects a causal relationship, or whether cystatin C might be a marker of underlying cardiac hypertrophy related to the initial impairment of renal function. Cystatin C may identify a preclinical state of renal dysfunction in the hypertensive population, which would not be detected with serum creatinine levels or estimated GFR, and which would reinforce the well-known association between chronic kidney disease and heart disease. LVH is present in 30–50% of patients with chronic kidney disease stage 1–2 (28,29), and in 62% of patients with essential hypertension and kidney damage, defined as microalbuminuria and/or estimated GFR <60 ml/min (30). Cystatin C might therefore detect cardiac hypertrophy early in essential hypertension. Results published recently by the investigators in the MESA study (31) would seem to reinforce this hypothesis. The MESA study demonstrates that baseline cystatin C levels in patients with normal blood pressure, with no cardiovascular disease or renal damage, are predictors of subsequent development of hypertension. All this suggests that differences in renal function captured by cystatin C levels rather than conventional methods, would be associated with the development of hypertension, and would imply initial renal damage in the pathogenesis of essential hypertension.

Finally, high cystatin C levels might also reflect longer duration or increased severity of established risk factors for the development of cardiac hypertrophy, such as the duration and severity of hypertension (32). In fact, the Heart and Soul Study shows that systolic blood pressure and pulse pressure are associated with cystatin C levels, even in subjects with apparently normal renal function according to creatinine levels (33). Our study did not reveal any correlation between blood pressure and cystatin C levels, and so we do not believe that the relationship between cystatin C and ventricular mass was related to blood pressure. Neither did we observe any relationship between cystatin C, urinary albumin excretion and markers of inflammation such as C-reactive protein and IL-6, all factors that may be involved in the development of cardiac hypertrophy (34–36).

Study limitations

Limitations to this study should be noted. First, the study population was a relatively small sample of non-diabetic patients selected consecutively from a university outpatient clinic, and may not be representative of the hypertensive population in general. Second, the size of our study cohort might underestimate the relationship between cystatin C, LVM and other variables, including serum creatinine and estimated GFR. Third, the cross-sectional design of our analysis does not allow us to determine the direction of the association between cystatin C and left ventricular hypertrophy, or determine causality. Future studies on larger patient samples are indicated to confirm our data.

In summary, our findings suggest that cystatin C could be a more predictable marker of cardiac hypertrophy than serum creatinine and creatinine-based GFR, in hypertensive patients with normal serum creatinine.

Conflict of interest: None.