Microalbuminuria and left ventricular function in type 2 diabetes: A quantitative assessment by stress echocardiography in the Myocardial Doppler in Diabetes (MYDID) Study III

Abstract

Background. Left ventricular (LV) function might be altered in type 2 diabetes (DM) and microalbuminuria (MA) may accentuate the abnormalities. We sought to investigate whether additional LV dysfunction could be unmasked using tissue Doppler (TVE)-enhanced dobutamine stress echocardiography (TVE-DSE) in patients with DM+MA. Methods. Twenty seven DM subjects with MA, (DM+MA), 31 DM subjects without MA (DM−MA), and 13 Controls were evaluated using TVE-DSE. LV basal peak systolic (PSV), early (E′) and late (A′) diastolic velocities (cm/sec) at rest and peak stress were post-processed. LV filling pressure was assessed using E/E′ratio. Results. PSV and E′velocity at peak stress in the respective three groups were 13.7±1.0, 10.1±1.1, 10.0±1.2 for PSV; and 10.0±1.6, 5.0±1.4, 4.8±1.4 for E′ (p <0.001 for controls vs. both groups). E/E′ at rest was 7.9±0.7 in the controls, 10.8±2.4 in DM−MA, and 11.0±2.2 in DM+MA (p <0.01 Controls vs. both the DM groups). Conclusions. Patients with DM+MA do not have additional LV regional systolic and diastolic dysfunctions compared with DM−MA, as revealed by TVE-DSE, when controlled for glycemia levels, lipids, and treatment strategies.

• Tissue velocity echocardiography
• type 2 diabetes
• microalbuminuria
• left ventricular function

The urinary excretion of small amounts of albumin in the urine could be a serious predictor of future events, such as elevation of systemic arterial pressure, cardiovascular disease, and progressive renal dysfunction 1. The incremental usefulness of multiple biomarkers from distinct biologic pathways for predicting the risk of cardiovascular events was evaluated recently. One of the biomarkers that most strongly predicted major cardiovascular events was urinary albumin-to-creatinine ratio (UACR). Since left ventricular (LV) structure and function might be altered in type 2 diabetes even in the absence of cardiovascular disease, presence of coexistent microalbuminuria in DM may be an early marker of or contribute to more severe LV dysfunctions. Previous studies from our group have suggested that uncomplicated DM may decrease LV myocardial functional reserve, which could be further compromised when associated with systemic arterial hypertension (HTN) or coronary artery disease (CAD) or their combination 2, 3. Also shown is the independent association in non-diabetics between MA and surface electrocardiogram (ECG) showing ischemic abnormalities, suggesting that MA has additional value to conventional risk indicators in predicting cardiovascular disease 4. The aim of the present study was to assess whether MA could cause additional myocardial dysfunction in patients with isolated DM.

Patients and methods

Population

Seventy one human subjects were evaluated for atypical chest pain. Subjects with only a normal dobutamine stress echo (DSE) were analyzed. Patients with known CAD, HTN, any structural heart disease, LV systolic dysfunction (LV ejection fraction-LVEF less than 50% at rest), abnormal regional LV wall motion at rest, primary myocardial disease, significant pericardial disease, significant valvular disease or arrhythmias were excluded from the present analysis. Subjects were grouped as controls (Control, n = 13), patients with DM without MA (DM − MA, n = 31), and DM with MA (DM + MA, n = 27). The duration of diabetes (years) was 5.6±4.7 in DM − MA and 8.2±5.1 in DM + MA (p = NS). Subjects in the control group were matched for age, LV size and mass, as well as for resting heart rate and other resting ECG parameters (Tables I–III). They did not have any associated diseases, had normal biochemical profile, were not on any medications and had a low pretest probability of CAD. Echocardiography was performed with a commercial Vivid 7 dimension™ equipment (General Electric, Vingmed, Horten, Norway) using an adult matrix probe for image acquisition. The images were acquired in parasternal long- and short axis as well as in apical 4, 3, and 2 chamber views.

Table I.  Demographic and biochemical data of the study subjects.

Variables	Controls Group 1	DM − MA Group 2	DM + MA Group 3	p
Male/Female	8/5	20/11	14/13	–
Age (years)	49.7±5.4	49.2±6.3	49.4±6.6	NS
Height, cm	168.4±5.8	164.9±8.5	165±7.9	NS
Weight, kg	66.2±8.2	70.3±11.1	68.4±8.0	NS
Hemoglobin, g/L	142±17	137±13	137±18	NS
Glycosylated Hemoglobin, %	–	8.1±1.7	8.3±1.6	NS
Fasting Plasma glucose, mmol/l	4.6±1.4	9.1±3.6	9.7±3.0	0.001*
Serum Creatinine, mmol/l	70.7±8.8	79.5±8.8	97.2±17.6	0.001^†
Microalbuminuria, µg/mg	1.2±4.0	7.2±4.0	127.5±41.6	0.001^†
Serum Cholesterol, mmol/l	3.2±0.6	4.6±0.6	4.9±1.0	0.001*
Serum Triglycerides, mmol/l	1.1±0.1	1.8±0.8	1.7±0.5	0.001*
Serum HDL, mmol/l	1.4±0.1	1.1±0.1	1.0±0.1	0.001*
Serum LDL, mmol/l	1.5±0.5	2.7±0.6	2.7±0.8	0.001*

DM − MA: Diabetics without microalbuminuria, DM + MA: Diabetics with microalbuminuria, *p < 0.001 Gr1 vs. Gr 2, Gr 3; ^†p < 0.001 Gr3 vs. Gr 1, Gr 2; NS: non significant (p > 0.05).

Table II.  Conventional echocardiographic data.

Variables	Controls Group 1	DM − MA Group 2	DM + MA Group 3	p
LA size, mm	28.3±2.7	31.2±3.5	32.7±3.0	0.05 ^§
LV size, mm	41±3.2	44±3.6	43±3.9	NS
LV mass, g	141±37	144±42	146±40	NS
LVEF, %	68.6±5.3	67.2±4.4	66.0±3.2	NS
E-wave, cm/sec	74.7±4.6	65.4±9.7	67.6±6.6	0.05^§
A-wave, cm/sec	60.4±4.8	71.7±10.7	64.5±6.5	0.001^¶
E/A ratio	1.2±0.1	0.9±0.1	1.0±0.1	0.001*
Tei index	0.40±0.1	0.53±0.1	0.52±0.1	0.001*

E-wave: E-wave velocity of the mitral valve inflow; A-wave: A-wave velocity of the mitral valve inflow; LA: Left Atrium; LV: Left ventricular; LVEDD: Left ventricular end-diastolic diameter; LVEF: Left ventricular ejection fraction; IVCT: Isovolumic contraction time; IVRT: Isovolumic relaxation time; ET: Ejection time. ^§p < 0.05 Gr1 vs. Gr 2, Gr 3; ^¶p < 0.001 Gr2 vs. Gr1, Gr3; *p < 0.001 Gr1 vs. Gr 2, Gr 3; ^†p < 0.001 Gr3 vs. Gr 1, Gr 2; NS: non significant (p > 0.05).

Table III.  Electrocardiogram findings.

Variables	Controls Group 1	DM − MA Group 2	DM + MA Group 3	p
HR, bpm	77.0±11.5	82.2±13.1	86.8±14.1	NS
PR interval, ms	150.6±23.1	141.8±21.0	144.8±16.1	NS
QRS duration, ms	84.1±10.5	80.5±9.7	93.0±33.3	NS
QTc, ms	399.6±2.1	408.7±14.8	420.6±27.7	NS

HR: Heart rate; QTc: corrected QT interval. NS: non significant (p > 0.05).

Ethical clearance

The Institutional review board of the BMJ Heart Centre at Bangalore, India, approved the study protocol. All study subjects gave informed consent.

Conventional Echocardiography Protocol

LV and left atrium (LA) dimensions along with LV mass and were measured by M-mode according to the American Society of Echocardiography guidelines. LVEF was calculated by using the modified Simpson's method. Mitral inflow velocities of early E-wave and late A-wave velocities and E/A ratio were measured by conventional pulsed wave Doppler, by positioning the sample volume at the level of the tips of mitral leaflets in the apical 4-chamber view. Tei index by conventional Doppler from the LV outflow tract was also measured. Tei index was defined as the sum of isovolumic contraction time and isovolumic relaxation time divided by ejection time.

Dobutamine Stress Echocardiography (DSE)

DSE was performed using a standard 3-min stage protocol (10–40 µg/kg/min). Patients who failed to achieve target heart rate were given atropine in increments of 0.3 mg up to a maximum of 1.8 mg. End points of DSE were the achievement of 85% of maximum heart rate, wherein all patients achieved this, and the patients who developed wall motion abnormalities of any grade were excluded from this study. Rate pressure products were identical and the mean value exceeded 21 000 in the three groups at peak dobutamine stress with no evidence of wall motion abnormalities at peak stress. Standard grey scale with superimposed color Doppler images was acquired in apical 4, 3 and 2 chamber projections. For visual analysis, the left ventricle was divided into 16 segments, according to the recommendations of the American Society of Echocardiography. The wall motion of each segment during DSE was scored as follows: hyperkinetic, normal, hypokinetic, akinetic, or dyskinetic. A test was considered normal in the absence of a new-onset wall motion abnormality in at least two consecutive segments. A test was considered eligible for the analysis if at least 12 of 16 segments were interpretable. The echocardiograms were analyzed by one independent investigator. A second interpretation was done by an independent investigator wherever there was ambiguity and consensus opinion prevailed.

Tissue Velocity Echocardiography (TVE)

TVE enhanced LV apical images were obtained at an average frame rate of 178 frames per second and were digitally stored. Cine loops containing three consecutive heart cycles were analyzed off-line at rest and during peak dobutamine stress. A sample volume was placed at the basal segments of each LV wall (septum, lateral, inferior and anterior) to measure regional function. Peak systolic velocity (PSV), early (E′) and late (A′) diastolic velocities (cm/s) were measured at rest and during peak stress. LV myocardial function was assessed by taking the average of the LV regional measurements.

LV filling pressure was calculated by measuring E′ by pulse Doppler at the septal annulus and this was divided by the transmitral inflow E velocity to obtain the E/E′ratio.

Biochemical methods

Urinary albumin was measured by a morning spot collection on two different days. The UACR was measured by modified Jaffe's method using the immunoturbidimetry method. The UACR measured in a spot urine sample is highly correlated with 24-hour urine albumin. MA was defined by urine albumin: creatinine ratio as more than 30 and less than 300 µg albumin/mg creatinine as described earlier 5. All other biochemical analyses were done prior to DSE, using a standard spectrophotometer.

Electrocardiography (ECG)

ECG was performed using commercial GE MAC “series” equipment that were enabled to show automated values, and subsequently used for analysis after manual corrections were made whenever needed.

Statistical methods

Data are expressed as mean±SD. One-way ANOVA followed by post hoc Scheffe's tests were performed to compare the differences between the groups. A PC-based version of Statistica™ version 6.0 (Statsoft, Tulsa, OK, USA) was used for data analysis. A p-value of < 0.05 was considered statistically significant.

Results

Table I shows patient demography and biochemical values. Age, height and weight did not vary significantly among the groups. Male and female subjects were equally distributed except in the DM − MA group. The DM − MA group had eight patients with non-proliferative diabetic retinopathy (NPDR) and two patients had neuropathy by microfilament testing, while in the DM + MA group there were 11 patients with NPDR and three with neuropathy. The medication profile varied marginally between the two groups, while in the DM − MA group eight patients were on angiotensin converting enzyme inhibitors (ACEI) and 14 on atorvastatin, DM + MA group had 15 on ACEI or angiotensin receptor blockers and 18 were on atorvastatin. There was no significant difference in haemoglobin and glycosylated haemoglobin levels. The fasting plasma glucose did not differ significantly among the diabetic population, but was significantly higher than the controls. Serum creatinine levels were significantly higher in the DM + MA group in comparison with other groups, as was the MA levels. Lipid levels did not vary among the diabetic groups, but were higher when compared with the controls.

Analysis of conventional echocardiographic parameters (Table II) showed that there was no significant difference in LV dimension, LV mass and LV ejection fraction, which were all normal, but LA size in DM + MA group was significantly greater than the Controls, though it was well within normal limits in all the groups. Early mitral inflow (E) velocity was significantly higher in normal subjects, while A velocity (atrial systolic velocity) was higher and consequently also E/A ratio was significantly lower in the DM − MA group when compared with the controls. Tei index in DM − MA and DM + MA patients was significantly higher when compared to controls.

Table III shows the parameters measured in ECG. There was no significant difference in the heart rate as well as PR interval, QRS duration and QTc interval in any of the groups. None of these patients had ST-T abnormalities

Table IV shows the TVE findings of regional function where velocities were measured at rest and peak dobutamine stress. Average PSV at rest was the same in all the groups, but at peak stress it was significantly greater in the controls when compared with the diabetic subjects, but did not differ between the DM groups. Average E′ velocity was significantly higher in the controls at rest as well as during peak stress, when compared with diabetics. Average A′ velocity did not vary either at rest or peak stress. Average E′/A′ ratio was significantly higher in controls when compared to diabetics.

Table IV.  Tissue velocity enhanced Dobutamine stress echocardiography data.

Variables	Stages	Controls Group 1	DM − MA Group 2	DM + MA Group 3	p
PSV, cm/sec	Rest	5.6±0.4	5.1±0.8	5.1±0.8	NS
	Peak stress	13.7±1.0	10.1±1.1	10.0±1.2	0.001*
E′ wave, cm/sec	Rest	9.3±0.6	6.2±1.2	6.3±1.2	0.001*
	Peak stress	10.0±1.6	5.0±1.4	4.8±1.4	0.001*
A′ wave, cm/sec	Rest	6.4±0.7	6.6±1.1	6.5±1.2	NS
	Peak stress	8.1±0.9	9.0±1.6	9.0±1.7	NS
E′/A′ ratio	Rest	1.4±0.2	0.9±0.3	1.0±0.3	0.001*
	Peak stress	1.2±0.2	0.5±0.1	0.5±0.1	0.001*
E/E′	Rest	7.9±0.7	10.8±2.4	11.0±2.2	0.001*

PSV: peak systolic velocity; E′-wave: myocardial velocity during the rapid filling period; A′-wave: myocardial velocity during the atrial contraction period; (Velocities are expressed as average of four left ventricular bases). *p < 0.001 Gr 1 vs. Gr 2, Gr 3; NS: non significant (p > 0.05).

LV filling pressure, estimated as E/E′ratio, was significantly higher in the DM groups compared with the controls.

Discussion

The role of MA and its relationship with cardiovascular disease is a question that has not been clearly answered 6. However, the importance of MA as a strong predictor of cardiovascular risk in diabetes has been clearly stated 7, though the effects of MA on left ventricular systolic and diastolic functions have not been studied using advanced echocardiographic methods such as tissue velocity echocardiography. The published data using conventional resting echocardiography 8 showing diastolic dysfunctions in DM on the basis of altered transmitral and pulmonary venous flows, though rather convincing, needs to be further validated by more robust methods because these transmitral and pulmonary velocities are load dependent. Moreover, the assessment of left ventricular diastolic dysfunctions using these parameters may not only be inadequate, but could also be less informative because of the complex mechanism of diastole in different disease states 9, 10.

Tissue velocity echocardiography on the other hand is less load independent at least within physiological limits and has been studied extensively in various clinical situations for more accurate assessment of systolic and diastolic functions of the heart at rest as well as during dobutamine stress 11, 12. In the present study the diabetic subjects irrespective of the presence of MA had diminished left ventricular functional reserve assessed by quantified stress echocardiography compared with the Controls while no differences were observed between the two diabetic populations when matched for age, gender, glycemia and lipid status. The Tei index, a marker of global myocardial performance 13, was also higher in the diabetic population compared with the Controls.

The diabetic subjects however did not show any difference in regional LV systolic function at rest in comparison with the Controls emphasizing the fact that their left ventricular geometry was not altered as evidenced by the left ventricular ejection fraction (Table II). However, the diabetic subjects had evidence of diastolic dysfunction at rest as measured by both conventional Doppler (transmitral E/A ratio, pulmonary venous systolic/diastolic ratio) and tissue Doppler (E′ velocity and E′/A′ ratio). The latter variables persisted also at peak stress. The findings of our study support those of others where evidence of diastolic dysfunction using conventional Doppler techniques has been identified 14. Although limited by the shortcomings of the conventional methods, as explained earlier, the tissue Doppler data on diastolic dysfunction may provide supportive and more objective evidence of disturbed left ventricular diastolic functions in DM. Whether this by itself constitute what some investigators claim as “diabetic cardiomyopathy” 15 remains to be seen, though evidence of collagen deposition in the diabetic myocardium has been proposed to be responsible for the so called “diabetic heart muscle disease” 16.

Can the E/E′ratio along with standard Doppler variables further strengthen the diastolic functional assessments?

One of the consequences of congestive heart failure with both systolic and diastolic dysfunction is the elevated left ventricular end diastolic pressure that used to be measured invasively until the universal availability of standard Doppler echocardiography that allows registration of transmitral velocities as well as pulmonary venous velocities. After the advent of tissue Doppler echocardiography, the ratio of early transmitral E velocity to regional early (E′) diastolic velocity measured either at the septal or lateral mitral annulus (E/E′) has now been increasingly used as a surrogate and non-invasive marker of elevated LV filling pressure both for diagnostic 17 and prognostic purposes, obviously in conjunction with the standard Doppler variables for more accurate assessment of left ventricular diastolic dysfunction/failure 18. In the current study the E/E′ ratio was found to be significantly higher in the diabetic population with and without MA compared with the Controls (Table III). The finding in conjunction with the conventional and tissue Doppler data most probably support the existence of altered left ventricular diastolic function in type 2 diabetes.

Does diminished functional reserve to dobutamine stress rule out “isolated” diastolic dysfunction?

One of the controversial issues of contemporary cardiology is whether isolated diastolic heart failure (or heart failure with preserved systolic function) exists or not. Even though the current data show evidence of resting diastolic dysfunction in DM in presence of normal left ventricular ejection fraction (Table II), diminished inotropic reserve during dobutamine stress most probably unmasks disturbances of longitudinal myocardial systolic functions that are not easily identified by 2-dimensional echocardiographic measurements of left ventricular ejection fraction. This is most probably the reason why large-scale epidemiological studies have reported high prevalence of “isolated” diastolic dysfunctions/failures (also called heart failure with normal ejection fraction) in the communities 19.

Diabetic cardiomyopathy: Need to look beyond the myocardium?

Although diastolic dysfunctions measured by conventional and tissue Doppler are present at rest, it is unlikely that they exist in isolated forms. The results suggest that some forms of subclinical disease are most probably present in the diabetic myocardium that is not recognized by 2-dimensional echocardiographic methods of estimation of left ventricular ejection fraction. Application of a combined conventional and tissue Doppler echocardiographic approach is therefore necessary to understand the complex and oftentimes hazardous natural history of progression of type 2 diabetes and its rather inevitable complications involving multiple organ systems. The combined approach might be useful in defining the myocardial pathology in diabetes more rationally because the differential diagnosis of decreased myocardial velocities does not only include “diabetic heart muscle disease” or “diabetic cardiomyopathy”, it might also indicate increased afterload because of the altered elastic properties of the aorta associated with macrovascular complications of type 2 diabetes 20. Actually, in a study from our group we have shown that TVE may identify improvement of longitudinal function following afterload reduction by the angiotensin receptor blocker, valsartan 21.

Clinical implications of the study

The lack of significant difference in DM subjects with or without MA could be attributed to the relatively shorter duration of diabetes and absence of co-morbidities. In our previous studies we have shown how co-morbid conditions negatively influence the myocardial velocity response in subjects with DM 3, 4. Since MA is a form of incipient diabetic nephropathy where glomerular filtration rate remains normal, we presume that the global burden of the disease was not heavy enough in the DM subjects to have more negative velocity response even in the presence of MA. We however do not think that MA is a purely benign condition and hence should be treated according to current practices. In fact it has been shown in a study that a reduced glomerular filtration rate (GFR) in asymptomatic diabetic patients was associated with a two-fold increase in cardiac events and reduced GFR independent of albuminuria was a significant predictor of cardiac events 22. Another factor may explain the data in the current study: patients in the DM group have been treated with angiotensin-converting enzyme inhibitors and the lipid-lowering agent, atorvastatin. Since both the drugs have so called “pleotropic” effects and since angiotensin receptor blocker improves myocardial systolic velocities 21, the longitudinal myocardial functions were most probably preserved in the DM groups.

Conclusion

Patients with DM + MA do not have additional LV regional systolic and diastolic dysfunctions compared with DM − MA, and subjects with type 2 diabetes irrespective of the presence of microalbuminuria have diminished myocardial functional reserve unmasked by quantitative stress echocardiography.

Acknowledgements

The study was supported by generous grants from the Swedish Cardiac Society, the Swedish Heart Lung Foundation, and the Karolinska Institute of Sweden. Dr. S. K. Saha is the recipient of the grant from the Swedish Cardiac Society. Dr. S. C. Govind has been a recipient of a guest scholarship from the Swedish Institute. We are thankful to Ms Jolanta Selin and Lars-Göran Persson for secretarial assistance.