Author + information
- Received June 11, 2013
- Revision received August 16, 2013
- Accepted August 26, 2013
- Published online December 24, 2013.
- Mikael K. Poulsen, MD, PhD∗∗ (, )
- Jordi S. Dahl, MD, PhD∗,
- Jan Erik Henriksen, MD, PhD†,
- Thomas M. Hey, MD∗,
- Poul Flemming Høilund-Carlsen, MD, DMSci‡,
- Henning Beck-Nielsen, MD, DMSci† and
- Jacob E. Møller, MD, PhD, DMSci∗
- ∗Department of Cardiology, Odense University Hospital, Odense, Denmark
- †Department of Endocrinology, Odense University Hospital, Odense, Denmark
- ‡Department of Nuclear Medicine, Odense University Hospital, Odense, Denmark
- ↵∗Reprint requests and correspondence:
Dr. Mikael Kjær Poulsen, Department of Cardiology, Odense University Hospital, 29 Sdr. Boulevard, Fyn 5000 Odense C, Denmark.
Objectives The study sought to determine the prognostic importance of left atrial (LA) dilation in patients with type 2 diabetes mellitus (T2DM) and no history of cardiovascular disease (CVD).
Background T2DM is associated with the development of CVD, and morphological changes in the heart may appear before symptoms arise.
Methods A total of 305 T2DM patients without known CVD referred to a diabetes clinic were included consecutively (age 58.6 ± 11.3 years, diabetes duration 2.0 [interquartile range: 0 to 6.0] years). Each patient underwent a comprehensive echocardiogram and a myocardial perfusion scintigraphy (MPS) at inclusion. Patients were divided according to left atrial volume index (LAVi) ≥32 ml/m2. Patients were followed for median of 5.6 (interquartile range: 5.1 to 6.1) years for the occurrence of major cardiac events and death.
Results LAVi ≥32 ml/m2 was found in 105 patients (34%). During follow-up, 60 patients (20%) experienced the composite endpoint, of whom 28 (9%) died. Patients with LAVi ≥32 ml/m2 had a significantly higher cardiac event rate and death rate (p < 0.001 and p = 0.002, respectively). Univariate predictors of the composite endpoint were age, hypertension, left ventricular diastolic function, E/e′septum-ratio and LAVi ≥32 ml/m2; however, myocardial ischemia on MPS was not a predictor. When adjusting for age and hypertension, only LAVi ≥32 ml/m2 was a predictor of the composite endpoint (hazard ratio: 1.82 [95% confidence interval: 1.08 to 3.07], p = 0.024).
Conclusions Increased LAVi was an independent and incremental predictor of cardiovascular morbidity and mortality in T2DM patients with no history of CVD. (Presence of Macrovascular Disease in Type 2 Diabetes Mellitus; NCT00298844).
The morbidity (1,2) and mortality (3) of type 2 diabetes mellitus (T2DM) is closely associated with development of cardiovascular disease (CVD). In T2DM, morphological changes suggestive of heart disease may appear before symptoms arise. The prevalence of left ventricular (LV) diastolic dysfunction has been shown to be increased among T2DM patients (4) and, further, to also predict outcome (5).
The evaluation of diastolic function requires multiple Doppler echocardiography variables (6). As these variables are influenced by pre- and after-load, they will reflect the instantaneous interaction of LV filling pressures and LV compliance (7). During ventricular diastole, left atrial (LA) and LV pressures will equalize. Factors causing LV filling pressure to rise will, therefore, also cause LA pressure overload and LA dilation (7,8). Thus, in the absence of LA volume overload, LA volume will reflect both the severity and duration of LV diastolic dysfunction (6,8,9). In agreement with this, LA size has proven to be a powerful predictor of outcome in several disease entities, including myocardial infarction, severe aortic valve stenosis, and heart failure (10–12). The importance of LA dilation in T2DM patients is, however, not known. On the basis of this, we hypothesized that left atrial volume index (LAVi) would predict long-term outcome in these patients.
The present single-center study has previously been described in detail (13). In brief, 431 consecutive T2DM patients without any medical history of CVD met the inclusion and exclusion criteria, of whom 126 declined participation; thus, 305 (71%) patients were enrolled after providing informed consent. The study was registered at www.clinicaltrials.gov (NCT00298844) and was approved by the local ethics committee.
Each patient underwent a single-day structured examination program including an echocardiogram, myocardial perfusion scintigraphy (MPS), and 1-point 51Cr-EDTA clearance procedure.
The echocardiograms were obtained on a Vivid 5 GE medical ultrasound machine with a 2.5 MHz transducer (GE Medical System, Horten, Norway). Blinded analysis of images was done using EchoPac 7.0, PC-08 (GE Medical System, Horten, Norway). For all Doppler and tissue Doppler recordings, a horizontal sweep of 100 mm/s was used, and the average of 5 consecutive beats was measured and averaged.
LV Systolic Function
LV Diastolic Function
Mitral inflow was assessed with pulsed-wave Doppler obtained with the transducer in the apical 4-chamber view, with the Doppler beam aligned perpendicular to the plane of the mitral annulus. The Doppler sample volume was placed between the tips of the mitral leaflets during diastole. Pulsed-wave tissue Doppler was used at the septal and lateral mitral annulus to measure the mitral plane movement. LV filling was divided into 4 distinct filling patterns (normal, impaired relaxation, pseudonormal, and restrictive filling) on the basis of a combination of mitral inflow and tissue Doppler measurements of the mitral plane movement (6).
LAVi was estimated by the biplane area-length method, using measurements at the apical 4- and 2-chamber views at end-systole and indexed by body surface area (14). LAVi ≥32 ml/m2 was considered abnormally increased. The absolute interobserver agreement between 2 observers obtained in 20 randomly-selected study patients on LV diastolic filling pattern and LAVi over or below 32 ml/m2 was 95% (kappa = 0.91) and 85% (kappa = 0.70), respectively.
MPS examinations were performed using electrocardiogram-gated single photon emission computed tomography with 99mtechnetium sestamibi. Whenever a potential stress-induced perfusion defect was observed, an additional rest study was carried out to evaluate the degree of reversibility. A semiquantitative visual interpretation was made by 2 observers using a 20-segment model. A summed stress score (SSS) was calculated. Myocardial ischemia was defined as a regional perfusion abnormality with a total SSS ≥4 and at least 1 segment with an SSS ≥2. The diagnostic accuracy has been reported previously (15).
The primary endpoint was a composite endpoint of death from all causes or the appearance of the first-occurred cardiovascular event during follow-up. Major adverse cardiovascular events (MACE) were nonfatal myocardial infarction, unstable angina, acute heart failure, ischemic stroke, and coronary or peripheral revascularization using the criteria from the DIAD (Detection of Ischemia in Asymptomatic Diabetics) trial (16). The secondary endpoint was death from all causes.
Overall mortality and MACE were recorded from the Danish Personnel Register and from discharge notes available in the Danish Admission Registry on October 1, 2012. In case of ambiguous information, original patient charts were reviewed.
Continuous variables are presented as mean ± SD and variables with a non-Gaussian distribution as median (interquartile range). Categorical variables are presented as numbers and percentages. The Student t test and Mann-Whitney 2-sample tests were used to test for between-group differences in independent continuous parametric and nonparametric variables. The chi-square test was used to test for differences between categorical variables.
Overall mortality and event rates were calculated using the product limit method and presented in Kaplan-Meier plots. Event rates were compared using the log-rank test. Further, estimation of hazard ratios was performed using Cox proportional hazards models. Comparison of each model’s predictive capability was performed by comparing the C-statistic derived from the area under the receiver-operating characteristic curves using the generalized U statistic. A p value <0.05 was considered significant using STATA/SE 12.0 (StataCorp, College Station, Texas).
The study included 305 consecutive T2DM patients who were followed for a median 5.6 (5.1 to 6.1) years, with complete follow-up data for all patients. Patient and echocardiographic characteristics are presented in Tables 1 and 2⇓⇓.
During follow-up, 60 patients (20%) experienced the composite endpoint, and 41 (15%) experienced a MACE. At follow-up, patients with LAVi ≥32 ml/m2 had significant higher event rates (p < 0.001) (Fig. 1). Finally, during follow-up 28 patients (9%) died (Table 3), and further patients with LAVi ≥32 ml/m2 had lower overall survival (p = 0.002) (Fig. 1). Finally, the 5-year event rates increased with increasing LAVi (Fig. 2). Univariate and multivariate predictors of the composite endpoint are presented in Table 4. When adjusting for age and hypertension, only LAVi ≥32 ml/m2 was a predictor of the composite endpoint (Table 4). Finally, the incremental value of LAVi assessed in 4 modeling steps is shown in Figure 3. The addition of LAVi ≥32 ml/m2 increased the area under the receiver-operating characteristic curve in a model containing diabetes duration, hypertension, and age (p = 0.02). Myocardial ischemia demonstrated on MPS was not a univariate predictor of the composite endpoint. When adjusting for myocardial ischemia in a multivariate Cox regression analysis, LAVi ≥32 ml/m2 was still a predictor of the composite endpoint (hazard ratio: 2.46 [95% confidence interval: 1.47 to 4.14], p = 0.001).
The present prospective observational follow-up study of T2DM patients without overt CVD at inclusion demonstrated that a dilated LA was associated with increased morbidity and mortality independent of other important risk factors. Thus, a dilated LA proved in T2DM to be a morphological change associated with increased cardiovascular risk.
As the LA in absence of LA volume overload will dilate secondary to increased LA afterload, increased LV filling pressure and decreased LV chamber compliance are believed to be important predictors of LA dilation in the present population. Although the present study does not offer insight into the mechanism causing LA pressure overload, it seems plausible that there exists a link between T2DM and diastolic dysfunction rather than assuming that this is a coincidental association. A decrease in LV compliance in T2DM patients may be caused by microvascular disease, altered myocardial metabolism, and structural changes in the myocardium with increased fibrosis. Some evidence is emerging on the role of myocardial lipotoxic injury from lipid oversupply. Visceral adipose tissue insulin resistance leads to increased myocardial fatty acid delivery and uptake in the myocardium, with associated myocardial triglyceride accumulation (17,18). It has been suggested that accumulation of fatty acid intermediates is associated with mitochondrial dysfunction, leading to cell damage, replacement with fibrosis, and decreased myocardial compliance (18).
In the present study, increased LAVi was a predictor of death and MACE, even after adjustment for age and hypertension. Further, we could demonstrate that LAVi predicted cardiovascular morbidity and mortality independently of myocardial ischemia. This is in agreement with the DIAD trial, where MPS-detected myocardial ischemia proved not to be a predictor of outcome in T2DM patients without CVD during a follow-up of 4.8 years (16). There are several possible explanations why increased LAVi can predict long-term outcome. First, LA volume reflects both the severity and duration of increased LA and LV filling pressure (7,8), whereas Doppler assessment of LV diastolic function only reflects the beat-to-beat interaction of LV filling pressures and LV compliance (7). Therefore, the LAVi may be a better prognostic parameter as compared with assessment of LV diastolic function. Second, abnormal diastolic function is characterized by a disproportional increase of LV filling pressure to maintain cardiac output especially during exercise (19), and these hemodynamic abnormalities caused by diastolic dysfunction are believed to drive the progression from an asymptomatic stage to symptomatic heart failure. Thus, the LAVi may be considered a morphological link between the asymptomatic and symptomatic stage.
Despite great effort, almost one-third of eligible patients refused to participate in the study. Although no statistical differences were present between the study group and the nonparticipants with regard to age and diabetes duration (data not shown), selection bias cannot be ruled out.
In T2DM patients without known CVD, increased LAVi was associated with worse overall survival and increased risk of MACE including myocardial infarction and ischemic stroke. Further, the prognostic value of LAVi persisted after adjusting for age and hypertension. Therefore, measurement of LAVi could emerge as a simple and important tool for risk stratification in T2DM patients.
The study was funded by the Danish Cardiovascular Research Academy (DaCRA), the Danish Diabetes Association, and the Danish Heart Foundation. Dr. Beck-Nielsen has received a research grant from Novo Nordisk. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiovascular disease
- left atrial volume index
- left ventricular
- major cardiovascular event(s)
- myocardial perfusion scintigraphy
- summed stress score
- type 2 diabetes mellitus
- Received June 11, 2013.
- Revision received August 16, 2013.
- Accepted August 26, 2013.
- 2013 American College of Cardiology Foundation
- Adler A.I.,
- Stevens R.J.,
- Neil A.,
- Stratton I.M.,
- Boulton A.J.,
- Holman R.R.
- Burchfiel C.M.,
- Curb J.D.,
- Rodriguez B.L.,
- Abbott R.D.,
- Chiu D.,
- Yano K.
- Moller J.E.,
- Pellikka P.A.,
- Hillis G.S.,
- Oh J.K.
- Hurrell D.G.,
- Nishimura R.A.,
- Ilstrup D.M.,
- Appleton C.P.
- Moller J.E.,
- Hillis G.S.,
- Oh J.K.,
- et al.
- Dahl J.S.,
- Videbaek L.,
- Poulsen M.K.,
- et al.
- Johansen A.,
- Gaster A.L.,
- Veje A.,
- Thayssen P.,
- Haghfelt T.,
- Holund-Carlsen P.F.
- Andersen M.J.,
- Ersboll M.,
- Bro-Jeppesen J.,
- et al.