The aim of the present study was to determine the prevalence and predictors of an abnormal echocardiogram in patients with Type 2 diabetes. Cardiac function and structure were rigorously assessed by comprehensive transthoracic echocardiographic techniques in 229 patients with Type 2 diabetes. Cardiovascular risk factors and diabetic complications were assessed, and predictors of an abnormal echocardiogram were identified using multivariate logistic regression analysis. An abnormal echocardiogram was present in 166 patients (72%). LVH (left ventricular hypertrophy) occurred in 116 patients (51%), and cardiac dysfunction was found in 146 patients (64%), of whom 109 had diastolic dysfunction alone and 37 had systolic±diastolic dysfunction. Independent predictors of an abnormal echocardiogram were obesity, age, the number of antihypertensive drugs used (all P<0.001) and creatinine clearance (P<0.05). The risk of an abnormal echocardiogram increased by 9% for each year over 50 years of age {OR (odds ratio), 1.09 [95% CI (confidence interval), 1.04–1.15]}, 3-fold if obesity was present [BMI (body mass index) >30; OR, 4.2 (95% CI, 1.9–9.0)] and by 80% for each antihypertensive agent used [OR, 1.8 (95% CI, 1.3–2.4) per agent]. In conclusion, an abnormal cardiac echocardiogram is common in patients with Type 2 diabetes. Importantly, although cardiac abnormalities can be predicted by traditional risk factors, such as age, obesity and renal function, the absence of micro- or macro-vascular complications does not predict a normal echocardiogram. We suggest that an echocardiogram identifies those with Type 2 diabetes at increased cardiovascular risk due to occult LVH and diastolic dysfunction, and this information may lead to more aggressive management of known risk factors in the clinic.

INTRODUCTION

Type 2 diabetes mellitus is a CHD (coronary heart disease) risk equivalent [1] associated with a significant cardiovascular disease burden [2,3], which includes a 2–3-fold increase in cardiac death [4] and an increased risk of CHF (congestive heart failure) [5]. Diabetes also directly contributes to the development of LVH [LV (left ventricular) hypertrophy] [6], which in itself is a predictor of cardiac morbidity and mortality. Systolic dysfunction is associated with macrovascular disease, but, early in the course of diabetes, cardiac fibrosis and hypertrophy [7,8] cause impaired relaxation and diastolic dysfunction [9].

Cardiac function can be assessed non-invasively using cardiac echocardiography. With the advent of the newer echocardiographic modalities of TDI (tissue Doppler imaging) and colour M-mode echocardiography measuring Vp (flow propagation velocity), it is now possible to more accurately classify diastolic dysfunction into progressively worsening grades that include the abnormal relaxation pattern (mild), the intermediate or pseudonormal pattern (moderate) and the restrictive physiology pattern (severe) [1015]. Conventional echocardiographic methods exhibit a U-shaped parabolic pattern when progressing from normal to worsening degrees of diastolic dysfunction, and thus the ‘pseudonormal’ pattern has not always been distinguished from the true ‘normal’ group. The distinction is important, as diastolic dysfunction has adverse prognostic implications [16,17] and is an independent risk factor for all-cause mortality [11].

Current guidelines for patients with diabetes recommend preventative strategies to minimize the risk of cardiovascular events [18,19]. To this end, the accurate and non-invasive identification of occult cardiac dysfunction should form a routine part of the management [20]. Screening via a 12-lead ECG lacks sufficient sensitivity and specificity for the detection of cardiac disease in asymptomatic patients with diabetes [21]. To date, it is unclear whether routine echocardiography in asymptomatic patients has any clinical utility in the general diabetic population seen in the outpatient setting. The present study examines the prevalence and clinical predictors of cardiac functional and structural abnormalities in a cross-sectional survey of 229 patients with Type 2 diabetes, using conventional Doppler echocardiography and the newer modalities of TDI and colour M-mode echocardiography.

MATERIALS AND METHODS

Study population

A total of 251 with Type 2 diabetes were studied prospectively. An echocardiogram is part of a routine complications surveillance programme at the Diabetic Clinic at Austin Health, Melbourne, Australia. As our primary referral base (80%) is from general practitioners, with only 20% referred from within the hospital, the cohort is representative of patients with Type 2 diabetes seen in the wider community. The research was carried out according to the Declaration of Helsinki (2000) of the World Medical Association, and was approved by the Human Research Ethics Committee at Austin Health. All subjects provided written consent.

Each patient completed a questionnaire, cross-checked by a review of the medical record, on cardiac risk, including a history of hypertension, presence of dyslipidaemia, duration of diabetes, and the presence of microvascular disease (nephropathy, retinopathy and neuropathy), cardiac disease [MI (myocardial infarction), angina, CHF, previous angioplasty and/or bypass surgery] and other macrovascular complications [stroke, TIA (transient ischaemic attack) and PVD (peripheral vascular disease)]. Systolic and diastolic BPs (blood pressures) were measured after 5 min of recumbency, using the appearance and disappearance of the Korotkoff sounds. Height and weight were measured for determination of BMI (body mass index), and BSA (body surface area) was measured using the formula (weight0.425)×(height0.725)×71.84/10000. Urine and plasma electrolytes were measured on a Hitachi 911 automatic analyser (Roche Diagnostics). Plasma high-sensitivity CRP (C-reactive protein) levels were measured using the SYNCHRON LX® system on a Beckman Coulter analyser. HbA1c (glycated haemoglobin) was measured by automated HPLC (Bio-Rad Laboratories). Fasting lipids were measured by enzymatic colorimetry. Total cholesterol was measured by enzymatic colorimetric methods, and LDL (low-density lipoprotein)-cholesterol was calculated using the Friedewald equation. Urinary albumin excretion was estimated by immunoturbidimetry (Dade-Behring) from a 24-h urine collection. The level of albuminuria was defined categorically according to standard guidelines. Creatinine clearance was determined using the Cockroft–Gault formula and was expressed per 1.73 m2 of BSA.

Definitions

Hypertension was defined as present if patients were on antihypertensive medication, had a history of hypertension and/or had evidence of hypertension (BP >130/85 mmHg). Patients were considered to have microalbuminuria if two out of three consecutive urine samples had an AER (albumin excretion rate) of ≥20 μg/min but ≤200 μg/min, and macroalbuminuria if two out of three consecutive urine samples revealed an AER of >200 μg/min. Retinopathy was assessed after examination of a dilated fundus by an ophthalmologist or if the subject had a history of laser photocoagulation. Peripheral neuropathy was determined upon examination and chart review at the Diabetic Clinic. The presence of macrovascular disease was defined by a positive history of cardiac, cerebral or PVD.

Echocardiography

Transthoracic echocardiography examination was performed as described previously by our group [1315] using an Acuson Sequoia ultrasound system, with measurements made according to the guidelines from the American Society of Echocardiography. M-mode echocardiography [22] was used to measure LV EF (ejection fraction) (using the modified Simpson's rule) and LVMI (LV mass index). Systolic dysfunction was defined by evidence of regional wall motion abnormalities and/or an EF <50%. LVH was defined as LV wall thickness >1.1 cm and/or LVMI >100 g/m2 in women and 130 g/m2 in men as per Framingham criteria.

Both conventional and newer parameters of diastolic function were measured using M-mode, pulmonary venous flow and colour M-mode. Measurements included peak mitral E and A (early and late diastolic peak filling velocities respectively) waves, E/A ratio, mitral E wave DT (deceleration time), IVRT (isovolumic relaxation time) and mitral Adur (A wave duration). TDI was used to obtain E′ (TDI-defined peak early diastolic filling velocity) and to calculate the E/E′ ratio. This ratio correlates with LV filling pressures and helps differentiate normal (normal filling pressures) from pseudonormal (high filling pressures) diastolic filling patterns. To determine the effect of loading conditions in distinguishing normal from pseudonormal diastolic dysfunction, the Valsalva manoeuvre, TDI and Vp modalities were used. A positive Valsalva manoeuvre was designated if E/A reversal, DT prolongation and >25% decrease in E occurred during the strain phase of the Valsalva manoeuvre. Diastolic function was classified as normal or into the progressively worsening grades of abnormal relaxation, pseudonormal or restrictive physiology patterns [12,13]. For a diagnosis of pseudonormal diastolic dysfunction, evidence of increased LV filling pressures was needed with three out of four of the following criteria being met: E/E′ ratio >10, depressed Vp (<50 cm/s), pulmonary Adur longer than mitral Adur or pulmonary venous atrial reversal maximal velocity (PulAVmax) >0.35 m/s, and a positive Valsalva manoeuvre.

Statistical analysis

Continuous data are presented as means±S.E.M., unless otherwise specified. Dichotomous variables are expressed as percentages. One-way ANOVA with a Bonferroni post-hoc test was used to examine differences among the groups. An unpaired Student's t test (two-tailed) was used to assess differences between continuous variables in two groups or one-way ANOVA (three or more groups, where subgroups were compared using Fisher's protected least-significant difference post-hoc test). Differences in categorical variables were compared using χ2 analysis. Multivariate analysis utilized multiple regression and ANCOVA (analysis of covariance) to model the independent predictors of cardiac functional indices. Logistic regression was used to estimate ORs (odds ratios) for statistically independent risk factors of cardiac dysfunction. Statistical analyses were performed using Stata (StataCorp) and StatView (SAS Institute). A value of P<0.05 was considered statistically significant.

RESULTS

Patient characteristics

Investigations were performed in 251 patients with Type 2 diabetes. As 22 patients were excluded due to significant valvular heart disease, mechanical valve prostheses or non-diagnostic imaging, 229 patients remained in the final study group. The total group studied included 145 men and 84 women with a mean age of 62±1 years, and a median duration of diabetes of 10 years. Most patients had established microvascular complications of diabetes (64%), including retinopathy (29%), elevated albuminuria (AER >20 μg/min; 40%) or peripheral neuropathy (28%). At the time of review, over 78% of patients were hypertensive, 33% of patients had a history of ischaemic heart disease and 12% of patients had established macrovascular disease at other sites (peripheral or cerebrovascular disease), meaning that over one-third of patients had established macrovascular disease. Table 1 shows the clinical characteristics of patients with a completely normal cardiac echocardiogram compared with those with an abnormal echocardiogram (LVH alone, diastolic dysfunction and/or systolic dysfunction).

Table 1
Clinical characteristics of the patients with Type 2 diabetes based on the echocardiographic results

Values are means±S.E.M. or percentages. Results are shown for patients with Type 2 diabetes with a normal or abnormal echocardiogram, along with the patients with an abnormal echocardiogram subdivided according to the presence of LVH alone, diastolic dysfunction, or systolic with or without diastolic dysfunction. *P<0.05 compared with patients with a normal echocardiogram; †P<0.01 compared with diastolic dysfunction. HDL, high-density lipoprotein.

Abnormal echocardiogram with
Normal echocardiogramAbnormal echocardiogramLVH aloneDiastolic dysfunctionSystolic±diastolic dysfunction
n 63 166 20 109 37 
Age (years) 55±1 65±2* 61±3 65±1* 65±2* 
Male (%) 62 63 75 57 73 
Duration (years) 8±1 12±1* 10±2 13±1* 10±1† 
Obesity (%) 38 58* 50 57 * 62* 
HbA1c (%) 8.0±0.2 7.6±0.1 7.5±0.2 7.7±0.1 7.5±0.1 
Total cholesterol (mmol/l) 5.0±0.1 4.7±0.1* 4.6±0.2 4.8±0.1 4.5±0.1† 
LDL-cholesterol (mmol/l) 2.9±0.1 2.6±0.1* 2.5±0.2 2.7±0.1 2.5±0.1 † 
HDL-cholesterol (mmol/l) 1.1±0.0 1.1±0.0 1.2±0.1 1.2±0.0 1.1±0.1* 
Triacylglycerols (mmol/l) 2.2±0.1 2.1±0.2 2.0±0.3 2.0±0.2 2.2±0.2 
CRP (mg/l) 6.3±0.9 9.0±2.2 3.6±1.0 10.6±2.9 9.9±6.0 
Hypertension (%) 60 85* 68 86* 92* 
Systolic BP (mmHg) 134±2 144±2* 140±3 146±2* 140±2*† 
Diastolic BP (mmHg) 74±1 76±1* 78±2 76±1 78±2* 
Antihypertensive medication (n0.9±0.1 2.1±0.1* 1.7±0.3 2.0±0.1* 2.4±0.2* 
Microvascular disease (%) 54 77* 80* 78* 89*† 
 Renal impairment (%)a 28* 15* 29* 33* 
 Elevated albuminuria (%) 33 43 45 40 51 
 Retinopathy (%) 13 35* 35* 40* 22† 
 Peripheral neuropathy (%) 19 32* 25 36* 24 
Macrovascular disease (%) 19 40* 25 33* 65*† 
 Myocardial infarction (%) 16* 11* 43*† 
 Coronary artery bypass graft (%) 13* 11* 30*† 
 Percutaneous revascularization (%) 11 22*† 
 Other macrovascular disease (%)b 11 22* 19 25* 2* 
Abnormal echocardiogram with
Normal echocardiogramAbnormal echocardiogramLVH aloneDiastolic dysfunctionSystolic±diastolic dysfunction
n 63 166 20 109 37 
Age (years) 55±1 65±2* 61±3 65±1* 65±2* 
Male (%) 62 63 75 57 73 
Duration (years) 8±1 12±1* 10±2 13±1* 10±1† 
Obesity (%) 38 58* 50 57 * 62* 
HbA1c (%) 8.0±0.2 7.6±0.1 7.5±0.2 7.7±0.1 7.5±0.1 
Total cholesterol (mmol/l) 5.0±0.1 4.7±0.1* 4.6±0.2 4.8±0.1 4.5±0.1† 
LDL-cholesterol (mmol/l) 2.9±0.1 2.6±0.1* 2.5±0.2 2.7±0.1 2.5±0.1 † 
HDL-cholesterol (mmol/l) 1.1±0.0 1.1±0.0 1.2±0.1 1.2±0.0 1.1±0.1* 
Triacylglycerols (mmol/l) 2.2±0.1 2.1±0.2 2.0±0.3 2.0±0.2 2.2±0.2 
CRP (mg/l) 6.3±0.9 9.0±2.2 3.6±1.0 10.6±2.9 9.9±6.0 
Hypertension (%) 60 85* 68 86* 92* 
Systolic BP (mmHg) 134±2 144±2* 140±3 146±2* 140±2*† 
Diastolic BP (mmHg) 74±1 76±1* 78±2 76±1 78±2* 
Antihypertensive medication (n0.9±0.1 2.1±0.1* 1.7±0.3 2.0±0.1* 2.4±0.2* 
Microvascular disease (%) 54 77* 80* 78* 89*† 
 Renal impairment (%)a 28* 15* 29* 33* 
 Elevated albuminuria (%) 33 43 45 40 51 
 Retinopathy (%) 13 35* 35* 40* 22† 
 Peripheral neuropathy (%) 19 32* 25 36* 24 
Macrovascular disease (%) 19 40* 25 33* 65*† 
 Myocardial infarction (%) 16* 11* 43*† 
 Coronary artery bypass graft (%) 13* 11* 30*† 
 Percutaneous revascularization (%) 11 22*† 
 Other macrovascular disease (%)b 11 22* 19 25* 2* 
a

GFR (glomerular filtration rate) <60 ml·min−1·1.73 m−2

b

Stroke, transient ischaemic attack, PVD, amputation and peripheral vascular angioplasty.

Echocardiography

A normal echocardiogram was present in 28% of patients (n=63), and abnormalities were detected in 72% (n=166) (Table 2). A total of 20 in patients had LVH alone (9%), 109 subjects had diastolic dysfunction alone (48%), 11 had systolic dysfunction (5%) and 26 had both systolic and diastolic dysfunction (11%). Subjects with diastolic dysfunction only were classified into those with an abnormal relaxation (n=51) or a pseudonormal pattern (n=57). There was one patient with a restrictive physiology pattern (results not shown).

Table 2
Echocardiographic findings in patients with Type 2 diabetes with a normal or abnormal echocardiogram

Values are means±S.E.M. The results from the single patient with a restrictive pattern is not shown in this table. *P<0.05, **P<0.01 and ***P<0.001 compared with patients with a normal echocardiogram.

Abnormal echocardiogram with
Diastolic dysfunction
ParameterNormal echocardiogramLVH aloneTotalAbnormal relaxationPseudonormalSystolic±diastolic dysfunction
n 63 20 109 51 57 37 
Structural       
 LVMI (g/m290±3.0 123±7*** 113±3*** 115±5*** 112±5*** 141±9‡ 
 LVH 0% 100%*** 68%*** 70%*** 66%*** 62%*** 
 LA area (cm219.7±0.4 21.4±0.9 22.9±0.5*** 21.9±0.7*** 23.6±0.6*** 26.1±1.1*** 
Functional       
 LVEF (%) 60±1 60±2 59±1 59±1 59±1 48±3*** 
S′ medial (cm/s) 11.1±0.3 11.1±0.3 10.5±0.2 11.5±0.5 10.3±0.2* 10.1±0.4** 
E/A ratio 1.08±0.06 0.94±0.05* 0.84±0.03*** 0.77±0.06*** 0.90±0.03*** 0.92±0.06** 
Vp (cm/s) 66±2 65±3 53±2*** 54±3** 54±2*** 48±3*** 
E′ medial (cm/s) 13.7±0.5 11.7±0.5** 10.9±0.3*** 11.0±0.6*** 10.8±0.5*** 11.1±0.8*** 
E/E′ medial 5.9±0.2 6.9±0.5 7.7±0.3*** 6.9±0.4 8.2±0.3*** 8.7±0.82*** 
E/V1.24±0.04 1.27±0.04 1.52±0.05*** 1.35±0.06** 1.64±0.06*** 1.95±0.06*** 
Abnormal echocardiogram with
Diastolic dysfunction
ParameterNormal echocardiogramLVH aloneTotalAbnormal relaxationPseudonormalSystolic±diastolic dysfunction
n 63 20 109 51 57 37 
Structural       
 LVMI (g/m290±3.0 123±7*** 113±3*** 115±5*** 112±5*** 141±9‡ 
 LVH 0% 100%*** 68%*** 70%*** 66%*** 62%*** 
 LA area (cm219.7±0.4 21.4±0.9 22.9±0.5*** 21.9±0.7*** 23.6±0.6*** 26.1±1.1*** 
Functional       
 LVEF (%) 60±1 60±2 59±1 59±1 59±1 48±3*** 
S′ medial (cm/s) 11.1±0.3 11.1±0.3 10.5±0.2 11.5±0.5 10.3±0.2* 10.1±0.4** 
E/A ratio 1.08±0.06 0.94±0.05* 0.84±0.03*** 0.77±0.06*** 0.90±0.03*** 0.92±0.06** 
Vp (cm/s) 66±2 65±3 53±2*** 54±3** 54±2*** 48±3*** 
E′ medial (cm/s) 13.7±0.5 11.7±0.5** 10.9±0.3*** 11.0±0.6*** 10.8±0.5*** 11.1±0.8*** 
E/E′ medial 5.9±0.2 6.9±0.5 7.7±0.3*** 6.9±0.4 8.2±0.3*** 8.7±0.82*** 
E/V1.24±0.04 1.27±0.04 1.52±0.05*** 1.35±0.06** 1.64±0.06*** 1.95±0.06*** 

Structural abnormalities

Patients with an abnormal echocardiogram had higher rates of LVH, with raised LVMI (P<0.001) and significantly lower E′. Those with any LV dysfunction (systolic±diastolic dysfunction) had higher LA (left atrial) areas (P<0.001) than those with normal functional parameters (normal and LVH alone) (Table 2).

Functional abnormalities

By definition, subjects with systolic dysfunction (n=37) had lower EFs (P<0.001) than those with a normal echocardiogram. An abnormal echocardiogram was associated with lower E/A ratios (P<0.05) and, as expected, the conventional Doppler (DT and IVRT) results exhibited a parabolic pattern (Table 2). By contrast, the newer modalities of TDI and colour M-mode exhibited linear properties. Thus subjects with LV dysfunction (systolic±diastolic dysfunction) had higher E/E′ and E/Vp ratios (P<0.001; Table 1) and lower S′ (TDI-derived systolic myocardial velocity) and Vp (P<0.001) compared with normal, confirming the utility of the newer modalities in the differentiation of normal from pseudonormal diastolic function.

Clinical predictors of cardiac dysfunction on echocardiography

On univariate analysis, subjects with an abnormal echocardiogram were approx. 10 years older, had a longer duration of diabetes and were more likely to be obese (Table 1, and Figure 1 and Figure 2). They also had poorly controlled hypertension with a higher BP, despite the use of more antihypertensive agents, and established micro- and macro-vascular complications (P<0.01) (Table 1 and Figure 1). Total cholesterol and LDL-cholesterol were significantly lower in patients with an abnormal echocardiogram (P<0.05), reflecting the greater use of statins (P<0.05). However, neither glycaemic control nor the mode of hypoglycaemic therapy were significantly different between groups. In addition, there were no differences in gender, smoking status, systemic inflammation (as measured by CRP) or a family history of CHD.

Prevalence of cardiac abnormalities on an echocardiogram stratified according to the duration of diabetes (A), number of antihypertensive agents (B), AER (C) and creatinine clearance (D)

Figure 1
Prevalence of cardiac abnormalities on an echocardiogram stratified according to the duration of diabetes (A), number of antihypertensive agents (B), AER (C) and creatinine clearance (D)

Open bars, normal echocardiogram; stippled bars, LVH only; grey shading, diastolic dysfunction alone; black shading, systolic dysfunction only; and hatched bars, systolic±diastolic dysfunction.

Figure 1
Prevalence of cardiac abnormalities on an echocardiogram stratified according to the duration of diabetes (A), number of antihypertensive agents (B), AER (C) and creatinine clearance (D)

Open bars, normal echocardiogram; stippled bars, LVH only; grey shading, diastolic dysfunction alone; black shading, systolic dysfunction only; and hatched bars, systolic±diastolic dysfunction.

Age of subject and BMI according to the presence of cardiac abnormalities on an echocardiogram

Figure 2
Age of subject and BMI according to the presence of cardiac abnormalities on an echocardiogram

(□) Patients with a normal echocardiogram, and (▲) patients with an abnormal echocardiogram; 80% of abnormal studies occurred in those over 50 years of age with a BMI >30.

Figure 2
Age of subject and BMI according to the presence of cardiac abnormalities on an echocardiogram

(□) Patients with a normal echocardiogram, and (▲) patients with an abnormal echocardiogram; 80% of abnormal studies occurred in those over 50 years of age with a BMI >30.

Subjects with systolic dysfunction were twice as likely to have established macrovascular disease than those with diastolic dysfunction. Over 70% of patients with systolic dysfunction had an established cardiac history compared with less than 30% of patients with diastolic dysfunction (P<0.01) and 21% of subjects with a normal echocardiogram (P<0.01 compared with systolic dysfunction; P=0.2 compared with diastolic dysfunction). Microvascular disease was also more common in systolic dysfunction than those with isolated diastolic disease (Table 1).

Patients with LVH but no dysfunction were clinically different from those with a completely normal echocardiogram. In particular, these patients were more likely to have microvascular complications (P=0.03), and were twice as likely to be receiving insulin as part of their hypoglycaemic regimen (P=0.04).

Independent predictors

On multivariate logistic regression analysis, independent predictors of an abnormal echocardiogram were obesity, age, the number of antihypertensive medications used (all P<0.001) and creatinine clearance (P<0.05). The overall model fit was good (Hosmer–Lemeshow statistic, 8.401; P=0.395). However, the four variables accounted for only 28% of the variance in the population studied. Notably, on multivariate analysis, there was no independent association between the presence of micro- and macro-vascular complications, including retinopathy and vascular disease, or a history of cardiac disease, meaning that normal cardiac function cannot be assumed in patients without established complications. By contrast, the presence of systolic dysfunction was reliably predicted by a history of cardiac disease on multivariate analysis (P<0.001).

The strongest predictor of an abnormal echocardiogram was age (Figure 2). For each year over 50 years of age, the risk of finding an abnormal echocardiogram was increased by 9% {OR, 1.09 [95% CI (confidence interval), 1.04–1.15]}. For patients over 70 years of age, only 11% had a normal echocardiographic study compared with 68% of patients less than 50 years of age (P<0.001). This effect was statistically independent of the longer duration of diabetes seen in older patients. The presence of obesity was also a strong risk factor for an abnormal echocardiogram, with an over 3-fold increase in risk if BMI was in the obese range [BMI >30; OR, 4.2 (95% CI, 1.9–9.0)]. In addition, the presence of hypertension was strongly associated with abnormal findings on echocardiography. Furthermore, the greater the number of agents required to control BP levels, the more likely that an abnormal echocardiogram was present [OR, 1.8 (95% CI, 1.3–2.4) per agent]. Notably, this effect was independent of both renal function and the duration of diabetes.

DISCUSSION

Cardiac dysfunction is common in patients with Type 2 diabetes. Using a combination of myocardial TDI and colour M-mode Vp, as well as conventional Doppler profiles, cardiac abnormalities were present in 72% of adults with Type 2 diabetes; up to 35% of these abnormalities would not have been diagnosed without rigorous echocardiographic assessment. The independent predictors of an abnormal echocardiogram were obesity, age, the number of antihypertensive agents used and creatinine clearance. Glycaemic control, dyslipidaemia, smoking status and a family history of CHD did not predict an abnormal echocardiogram. Most importantly, we found no independent association of an abnormal echocardiogram with clinical evidence of micro- and macro-vascular complications. This result indicates that normal cardiac findings cannot be assumed in patients without established diabetic complications.

LVH

As in the Framingham Heart Study [6], LVH was present in 70% (116/166) of patients with an abnormal echocardiogram, many of whom also had hypertension, diabetic complications and diastolic dysfunction. LVH was present in 21% (20 out of 94 patients) of those with normal diastolic function and in 65% (88 out of 135 patients) with diastolic dysfunction. Epidemiological studies show the prevalence of LVH is 6% in normal diastolic function and increases to 29% in those with diastolic dysfunction, but these studies contain few subjects with diabetes (45 out of 1274 patients, or 3.5%) [23]. Our present results confirm that diabetes accelerates the development of LVH [24] and the subsequent development of diastolic dysfunction.

Cardiac dysfunction

In the present study, we found a surprisingly low rate of impaired systolic function (16%), despite high rates of diastolic abnormalities (59%) and CHD. It is unlikely that the results reflect a sampling bias, as echocardiograms were not ordered according to clinical indication. However, in keeping with the increased mortality rate in patients with diabetes after MI, it is possible that the occurrence of coronary ischaemia in a cardiac milieu of fibrosis, hypertrophy and diastolic dysfunction is more likely to result in death and, consequently, less patients with systolic dysfunction.

In the present study, the rate of diastolic dysfunction was 59%, which is higher than in a population-based survey that reported a prevalence of 28% [11]. Most studies in diabetes have involved relatively small numbers of patients without cardiac risk factors (other than diabetes itself) or complications. Furthermore, many studies were performed before the newer echocardiographic techniques were available and, therefore, under-diagnosed diastolic abnormalities. More rigorous assessment of cardiac function to unmask the pseudonormal pattern results in approx. 25% more cases of diastolic dysfunction.

Predictors of cardiac abnormalities in diabetes

The independent predictors of an abnormal heart echocardiogram were obesity, age, the number of antihypertensive agents used and creatinine clearance. Age was the strongest predictor of an abnormal echocardiogram, with the risk increasing by 9% for each year over 50 years of age. Although abnormal relaxation does occur as part of normal aging, pseudonormal diastolic filling is not part of physiological aging and is therefore abnormal at any age [9]. Obesity was also a strong risk factor for an abnormal echocardiogram, with a 3-fold increase in prevalence if patients were obese. A graded increase in the risk of heart failure is observed across categories of BMI [25]. Given that patients with diabetes already have a 2–3-fold increase in the risk for systolic CHF, the combination of diabetes and increased BMI may contribute to the over-representation of patients with diabetes in heart failure databases.

Hypertension was the strongest reversible risk factor for cardiac abnormalities, and the independent association with the number of antihypertensive drugs used supports an association with the severity of hypertension. Over 80% of patients were on medication with known cardiac benefits, and most patients were treated with multiple antihypertensive agents (average of approx. 2 agents per person), including ACE (angiotensin-converting enzyme) inhibitors or ARBs (angiotensin receptor blockers) (70% of patients), diuretics (36%), calcium channel blockers (32%) or β-blockers (15%). However, although 86% of those with diastolic dysfunction were hypertensive, 60% of patients with a normal echocardiogram were also hypertensive. This suggests the occurrence of diastolic dysfunction is not simply an outcome of increased BP. Certainly, both myocardial histological studies and tissue Doppler studies have shown that diabetes itself results in myocardial architectural changes (or ‘diabetic cardiomyopathy’), and that hypertension accelerates these processes rather than causing them [26,27].

Hypertension is more common in patients with diabetes, and especially in those with diabetic nephropathy. We found that failing renal function assessed by creatinine clearance was strongly associated with abnormal echocardiographic findings. By contrast with the Strong Heart Study in diabetes, which found urinary albumin excretion was independently associated with cardiac dysfunction [28], the presence or severity of albuminuria was not an important predictor in our present study after correction for renal impairment. This difference may be partly explained by the fact that no adjustment for creatinine clearance was performed in the Strong Heart Study [28]. Our present results would support the assertion that creatinine clearance is a better predictor than albuminuria for cardiac dysfunction, although both are clearly manifestations of the same process (i.e. diabetic nephropathy).

Perspectives

Non-invasive and early identification of patients with diabetes at increased cardiovascular risk is needed. In the present study of asymptomatic subjects with diabetes, >70% of patients had an abnormal echocardiogram. Although the presence of impaired systolic function was reliably predicted by a past history of cardiac disease, normal cardiac function cannot be assumed in those without established diabetic complications. The importance of diagnosing diastolic dysfunction has only recently been recognized, and its presence provides important prognostic information [11,17,29]. Whether the abnormalities we observed in the present study will have prognostic significance is the subject of on-going research in our centre.

Our present findings reinforce the need to aggressively treat hypertension at the earliest possible opportunity in patients with diabetes. The benefit of BP lowering in diabetics with ECG-defined LVH is clear [30] and, although the optimal treatment of diastolic dysfunction in diabetes is unknown, improved BP control will improve diastolic parameters in hypertensive subjects [31]. Guidelines set arbitrary BP levels at which treatment should be initiated, but a recent study suggests the need to lower BP in diabetes, whatever the initial BP [32]. In the ADVANCE trial, administration of a fixed combination of perindopril and indapamide to patients with Type 2 diabetes, irrespective of baseline BP or the use of other antihypertensives, including renin–angiotensin system blockade, reduced BP (5.6/2.2 mmHg) and the risk of major vascular deaths [32].

Early detection and treatment of cardiac disease is paramount in improving health outcomes in Type 2 diabetes. Our present results suggest an echocardiogram may serve to more precisely identify those subjects with diabetes at increased cardiovascular risk due not only to impaired systolic function, but also due to occult LVH and diastolic dysfunction. If such information were available to the physician, it may lead to more aggressive management of known risk factors, such as hypertension in the clinic.

Abbreviations

     
  • A

    late diastolic peak filling velocity

  •  
  • Adur

    A wave duration

  •  
  • AER

    albumin excretion rate

  •  
  • BMI

    body mass index

  •  
  • BSA

    body surface area

  •  
  • BP

    blood pressure

  •  
  • CHD

    coronary heart disease

  •  
  • CHF

    congestive heart failure

  •  
  • CI

    confidence interval

  •  
  • CRP

    C-reactive protein

  •  
  • DT

    deceleration time

  •  
  • E

    early diastolic peak filling velocity

  •  
  • EF

    ejection fraction

  •  
  • HbA1c

    glycated haemoglobin

  •  
  • IVRT

    isovolumic relaxation time

  •  
  • LA

    left atrial

  •  
  • LDL

    low-density lipoprotein

  •  
  • LV

    left ventricular

  •  
  • LVH

    LV hypertrophy

  •  
  • LVMI

    LV mass index

  •  
  • MI

    myocardial infarction

  •  
  • OR

    odds ratio

  •  
  • PVD

    peripheral vascular disease

  •  
  • TDI

    tissue Doppler imaging E′, TDI-derived peak early diastolic filling velocity

  •  
  • S

    TDI-derived systolic myocardial velocity

  •  
  • Vp

    flow propagation velocity

This work was supported by educational grants from the National Health and Medical Research Council of Australia, Servier Laboratories, Australia, and a Pfizer CardioVascular Lipid award to P.M.S.

References

References
1
Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults
Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III)
JAMA, J. Am. Med. Assoc.
2001
, vol. 
285
 (pg. 
2486
-
2497
)
2
Fox
 
C. S.
Coady
 
S.
Sorlie
 
P. D.
, et al 
Increasing cardiovascular disease burden due to diabetes mellitus: the Framingham Heart Study
Circulation
2007
, vol. 
115
 (pg. 
1544
-
1550
)
3
Mulnier
 
H. E.
Seaman
 
H. E.
Raleigh
 
V. S.
Soedamah-Muthu
 
S. S.
Colhoun
 
H. M.
Lawrenson
 
R. A.
 
Mortality in people with type 2 diabetes in the UK
Diabetic Med.
2006
, vol. 
23
 (pg. 
516
-
521
)
4
Stamler
 
J.
Vaccaro
 
O.
Neaton
 
J. D.
Wentworth
 
D.
 
Diabetes, other risk factors and 12 year cardiovascular mortality for men in the multiple risk factor intervention trial (MRFIT)
Diabetes Care
1993
, vol. 
16
 (pg. 
434
-
444
)
5
Nichols
 
G. A.
Gullion
 
C. M.
Koro
 
C. E.
Ephross
 
S. A.
Brown
 
J. B.
 
The incidence of congestive heart failure in type 2 diabetes: an update
Diabetes Care
2004
, vol. 
27
 (pg. 
1879
-
1884
)
6
Galderisi
 
M.
Anderson
 
K. M.
Wilson
 
P. W.
Levy
 
D.
 
Echocardiographic evidence for the existence of a distinct diabetic cardiomyopathy (the Framingham Heart Study)
Am. J. Cardiol.
1991
, vol. 
68
 (pg. 
85
-
89
)
7
Rubler
 
S.
Dlugash
 
J.
Yuceoglu
 
Y. Z.
Kumral
 
T.
Branwood
 
A. W.
Grishman
 
A.
 
New type of cardiomyopathy associated with diabetic glomerulosclerosis
Am. J. Cardiol.
1972
, vol. 
30
 (pg. 
595
-
602
)
8
Candido
 
R.
Forbes
 
J. M.
Thomas
 
M. C.
, et al 
A breaker of advanced glycation end products attenuates diabetes-induced myocardial structural changes
Circ. Res.
2003
, vol. 
92
 (pg. 
785
-
792
)
9
Zile
 
M. R.
Brutsaert
 
D. L.
 
New concepts in diastolic dysfunction and diastolic heart failure: Part II: causal mechanisms and treatment
Circulation
2002
, vol. 
105
 (pg. 
1503
-
1508
)
10
Garcia
 
M. J.
Palac
 
R. T.
Malenka
 
D. J.
Terrell
 
P.
Plehn
 
J. F.
 
Color M-mode Doppler flow propagation velocity is a relatively preload-independent index of left ventricular filling
J. Am. Soc. Echocardiogr.
1999
, vol. 
12
 (pg. 
129
-
137.
)
11
Redfield
 
M. M.
Jacobsen
 
S. J.
Burnett
 
J. C.
Mahoney
 
D. W.
Bailey
 
K. R.
Rodeheffer
 
R. J.
 
Burden of systolic and diastolic ventricular dysfunction in the community: appreciating the scope of the heart failure epidemic
JAMA, J. Am. Med. Assoc.
2003
, vol. 
289
 (pg. 
194
-
202
)
12
Khouri
 
S. J.
Maly
 
G. T.
Suh
 
D. D.
Walsh
 
T. E.
 
A practical approach to the echocardiographic evaluation of diastolic function
J. Am. Soc. Echocardiogr.
2004
, vol. 
17
 (pg. 
290
-
297
)
13
Srivastava
 
P.
Burrell
 
L. M.
Calafiore
 
P.
 
Lateral versus medial mitral annular tissue doppler in the echocardiographic assessment of diastolic function and filling pressures: which should we use?
Eur. J. Echocardiogr.
2005
, vol. 
6
 (pg. 
97
-
106
)
14
Srivastava
 
P. M.
Thomas
 
M. C.
Calafiore
 
P.
Macisaac
 
R. J.
Jerums
 
G.
Burrell
 
L. M.
 
Diastolic dysfunction is associated with anaemia in patients with Type II diabetes
Clin. Sci.
2006
, vol. 
110
 (pg. 
109
-
116
)
15
Srivastava
 
P. M.
Calafiore
 
P.
MacIsaac
 
R. J.
Hare
 
D. L.
Jerums
 
G.
Burrell
 
L. M.
 
Thiazolidinediones and congestive heart failure–exacerbation or new onset of left ventricular dysfunction?
Diabetic Med.
2004
, vol. 
21
 (pg. 
945
-
950
)
16
Wang
 
M.
Yip
 
G. W.
Wang
 
A. Y.
, et al 
Tissue Doppler imaging provides incremental prognostic value in patients with systemic hypertension and left ventricular hypertrophy
J. Hypertens.
2005
, vol. 
23
 (pg. 
183
-
191
)
17
Sanderson
 
J. E.
Wang
 
M.
Yu
 
C. M.
 
Tissue Doppler imaging for predicting outcome in patients with cardiovascular disease
Curr. Opin. Cardiol.
2004
, vol. 
19
 (pg. 
458
-
463
)
18
American Diabetes Association
Standards of medical care in diabetes–2007
Diabetes Care
2007
, vol. 
30
 (pg. 
S4
-
S41
)
19
De Backer
 
G.
Ambrosioni
 
E.
Borch-Johnsen
 
K.
, et al 
European Guidelines on Cardiovascular Disease Prevention in Clinical Practice. Third Joint Task Force of European and other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of eight societies and by invited experts)
Atherosclerosis
2004
, vol. 
173
 (pg. 
381
-
391
)
20
Candido
 
R.
Srivastava
 
P.
Cooper
 
M. E.
Burrell
 
L. M.
 
Diabetes mellitus: a cardiovascular disease
Curr. Opin. Invest. Drugs
2003
, vol. 
4
 (pg. 
1088
-
1094
)
21
Struthers
 
A. D.
Morris
 
A. D.
 
Screening for and treating left-ventricular abnormalities in diabetes mellitus: a new way of reducing cardiac deaths
Lancet
2002
, vol. 
359
 (pg. 
1430
-
1432
)
22
Schiller
 
N. B.
 
Two-dimensional echocardiographic determination of left ventricular volume, systolic function, and mass. Summary and discussion of the 1989 recommendations of the American Society of Echocardiography
Circulation
1991
, vol. 
84
 
Suppl. 3
(pg. 
I280
-
I287
)
23
Fischer
 
M.
Baessler
 
A.
Hense
 
H. W.
, et al 
Prevalence of left ventricular diastolic dysfunction in the community: results from a Doppler echocardiographic-based survey of a population sample
Eur. Heart J.
2003
, vol. 
24
 (pg. 
320
-
328
)
24
Grossman
 
E.
Shemesh
 
J.
Shamiss
 
A.
Thaler
 
M.
Carroll
 
J.
Rosenthal
 
T.
 
Left ventricular mass in diabetes-hypertension
Arch. Intern. Med.
1992
, vol. 
152
 (pg. 
1001
-
1004.
)
25
Kenchaiah
 
S.
Evans
 
J. C.
Levy
 
D.
, et al 
Obesity and the risk of heart failure
N. Engl. J. Med.
2002
, vol. 
347
 (pg. 
305
-
313
)
26
Frustaci
 
A.
Kajstura
 
J.
Chimenti
 
C.
, et al 
Myocardial cell death in human diabetes
Circ. Res.
2000
, vol. 
87
 (pg. 
1123
-
1132.
)
27
Kosmala
 
W.
Kucharski
 
W.
Przewlocka-Kosmala
 
M.
Mazurek
 
W.
 
Comparison of left ventricular function by tissue Doppler imaging in patients with diabetes mellitus without systemic hypertension versus diabetes mellitus with systemic hypertension
Am. J. Cardiol.
2004
, vol. 
94
 (pg. 
395
-
399
)
28
Liu
 
J. E.
Robbins
 
D. C.
Palmieri
 
V.
, et al 
Association of albuminuria with systolic and diastolic left ventricular dysfunction in type 2 diabetes: the Strong Heart Study
J. Am. Coll. Cardiol.
2003
, vol. 
41
 (pg. 
2022
-
2028
)
29
Whalley
 
G. A.
Doughty
 
R. N.
Gamble
 
G. D.
, et al 
Pseudonormal mitral filling pattern predicts hospital re-admission in patients with congestive heart failure
J. Am. Coll. Cardiol.
2002
, vol. 
39
 (pg. 
1787
-
1795.
)
30
Lindholm
 
L. H.
Ibsen
 
H.
Dahlof
 
B.
, et al 
Cardiovascular morbidity and mortality in patients with diabetes in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol
Lancet
2002
, vol. 
359
 (pg. 
1004
-
1010
)
31
Solomon
 
S.
Janardhanan
 
R.
Verma
 
A.
, et al 
Effect of angiotensin receptor blockade and antihypertensive drugs on diastolic function in patients with hypertension and diastolic dysfunction: a randomised trial
Lancet
2007
, vol. 
369
 (pg. 
2079
-
2087
)
32
Patel
 
A.
ADVANCE Collaborative Group
 
Effects of a fixed combination of perindopril and indapamide on macrovascular and microvascular outcomes in patients with type 2 diabetes (the ADVANCE trial): a randomised controlled trial
Lancet
2007
, vol. 
370
 (pg. 
829
-
840
)