Objectives The question of whether diabetes mellitus (DM) duration correlates with the severity of dysfunction has not been well studied. We hypothesised that the longer the duration of DM the worse the severity of left ventricular (LV) diastolic dysfunction and increased risk of cardiovascular disease (CVD).
Methods We reviewed 547 diabetic patients between January 2005 and April 2010. Finally, 92 consecutive patients who presented with type 2 DM and who underwent echocardiographic assessment were enrolled according to the selection criteria. In all patients, ischaemic heart disease and heart failure were excluded.
Results Diastolic parameters were significantly worsened with increasing duration of DM (p<0.05). In the ≥7 years DM duration group (n=50), the E/Ea ratio increased significantly and the Ea/Aa ratio decreased significantly, compared with those in the <7 years DM duration group (n=42). CVD developed in 28 patients (30.4%) during the follow-up period. However, the duration of DM showed less statistical correlation with the incidence of CVD (p=0.188) and other LV diastolic function indices did not differ significantly between groups with or without CVD.
Conclusions Alteration of diastolic function induced by DM worsens with increasing duration of DM. DM duration on echocardiographic evaluation time did not differ significantly between the CVD incident and the non-CVD incident groups. The rate of CVD development was not significantly different if the duration of DM was more than 7 years. Therefore, active medical care including echocardiography should be undertaken to prevent CVD from the point of diagnosis of type 2 DM.
- Diabetes mellitus
- diastolic function
- diabetes & endocrinology
Statistics from Altmetric.com
Diabetes mellitus (DM) is one of the most common endocrine disorders and the incidence of DM is continuously increasing. Reports from the Framingham study and other studies have established DM as a strong risk factor for cardiovascular morbidity and mortality.1–3 DM has direct effects on the heart, independent of obstructive disease and hypertension (HTN), and causes development of a specific cardiomyopathy that results in impaired left ventricular (LV) diastolic filling. Abnormalities in LV diastolic filling often precede systolic dysfunction in various disease states, such as coronary artery disease and cardiac hypertrophy. Several studies have demonstrated deterioration in LV diastolic properties in diabetic patients by Doppler echocardiography.4–7 However, the correlation of DM duration with severity of dysfunction and onset of ischaemic heart disease has not yet been identified. The aim of our study was to determine the preclinical effects of diabetes duration on the severity of LV dysfunction using echocardiography and the risk of cardiovascular disease (CVD). We hypothesised that the longer the DM duration the worse the severity of diastolic dysfunction and increased risk of CVD.
All consecutive type 2 diabetic patients were identified using an electronic coding system (547 patients) between January 2005 and April 2010 at our institution. From these, 455 patients were excluded due to (1) an inexact record of DM diagnosis date; (2) Doppler echocardiographic assessment not done at the time of diagnosis of DM; (3) all types of cardiovascular disease prior to echocardiogram; (4) a short follow-up period after echocardiographic assessment (<2 years); (4) an incomplete record of echocardiographic result; (5) being younger than 30 years. Finally, 92 patients (36 men, 56 women; age, 59±11 years; range, 32–82) were recruited into the study. We used a 7-year (median value) cut-point for DM duration in order to divide patients: <7 years DM duration group (<7-year group, n=42) and ≥7 years DM duration group (≥7-year group, n=50). And for CVD development after echocardiographic evaluation, we divided patients into the CVD incident group (CVD group, n=28) and the non-CVD incident group (non-CVD group, n=64). Diagnosis of CVD, such as angina and myocardial infarction, was done based on clinical reports and all study subjects were on medication for diabetes. The institutional review board of Gachon University Gil Hospital approved this study protocol.
A standard two-dimensional (2D) and a tissue Doppler echocardiographic examination were performed on all subjects using a 3.5 MHz transducer (Vivid 7 Dimension ultrasound equipment; General Electric, Horten, Norway) and the echocardiographic examiners were blinded to the patients' information. Measurements were made according to the American Society of Echocardiography guidelines.8 LV ejection fraction was calculated by the Teicholz method. Left atrial volume was determined according to the biplane area–length method from the apical four-chamber and two-chamber views and was indexed by body surface area (BSA). LV mass was calculated using the corrected ASE cube formula and indexed for BSA to obtain the LV mass index. LV hypertrophy (LVH) was defined as LV mass index >95 g/m2 in female patients and >115 g/m2 in male patients. Pulsed wave Doppler of transmitral LV inflow was performed in the apical four-chamber view, with the sample volume placed at the level of the mitral valve tips and Doppler variables were analysed during three consecutive beats. The following measurements of global LV diastolic function were determined using tissue Doppler imaging: peak early (E) and late (A) diastolic mitral flow velocity and their ratio E/A, deceleration time (DT) of the E wave, and early (Ea) and late (Aa) diastolic mitral annular velocity.
Standard statistical software (SPSS V.12, SPSS Inc, Chicago, Illinois, USA) was used for data analysis. Continuous variables, expressed as mean±SD, were compared using Student t test for independent groups. Simple linear regression test was used for assessment of the association of DM duration with the other variables. Differences between the groups were evaluated by independent t test, χ2 test and Fisher's exact test. A p value <0.05 was considered statistically significant.
Clinical characteristics of the subjects
Among the 92 patients enrolled in the study, 36 (39.1%) patients were male. CVD had developed in 28 (30.4%) patients and 53 (57.6%) patients had a diagnosis of HTN. Age was higher in the ≥7-year group compared to the <7-year group (62±11 vs 56±10 years, p=0.008). Height, weight, body mass index, BSA and the rate of HTN were similar in both groups (p>0.05). Detailed data are shown in table 1.
Effect of duration of DM on echocardiographic parameters
Using simple linear regression, a significant association was observed between the E/Ea ratio and duration of DM; for 1 year after onset of DM, E/Ea increased by 0.17 (r=0.22, p=0.039). In addition, Ea/Aa decreased by 0.01 for every 1-year increase in the duration of DM (r=−0.314, p=0.012). The Ea value decreased by 0.001 for every 1-year increase in the duration of DM (r=−0.25, p=0.02). An increase in the duration of DM was associated with other LV diastolic function indices: E/A ratio and DT; however, this association was not significant (E/A, r=−0.20, p=0.063; DT, r=0.17, p=0.109). The correlation of duration of DM with Ea/Aa and E/Ea is shown in figure 1. The results of the conventional Doppler echocardiographic parameters are summarised in table 2. There were significant univariate associations with E/Ea, which was positively associated with DM duration (r=0.219, p=0.039) and HTN (r=0.230, p=0.030). When these associated factors were entered a forward stepwise regression model, the results showed the associations of DM duration (r=0.204, p=0.049) and HTN (r=0.215, p=0.039) to be independent predictor of Ea (r2=0.13). There are significant univariate associations with Ea, which was negatively associated with age (r=−0.263, p=0.013), HTN (r=−0.261, p=0.013) and DM duration (r=−0.247, p=0.020). When all these associated factors were entered in a forward stepwise regression model, the results showed only the associations of HTN (r=−0.243, p=0.017) to be independent predictor of Ea (r2=0.16).
In the ≥7-year group, mean A velocity was higher (0.84±0.15 m/s vs 0.73±0.16 m/s, p=0.001) and mean Ea velocity was lower (5.63±1.61 cm/s vs 6.47±1.95 cm/s, p=0.028) than in the <7-year group. And the ≥7-year group had lower Ea/Aa ratio (0.59±0.13 vs 0.71±0.27, p=0.023); however, the E/Ea ratio (12.5±4.5 vs 10.2±3.5, p=0.01) was higher than in the <7-year group. Other parameters did not differ between the two groups.
Effect of duration of DM on CVD incidence
The percentage of women was significantly higher in the group of longer duration of DM, that is, 67.1% in the ≥5-year group and 70% in the ≥6-year group. CVD incidence after echocardiographic evaluation was analysed. There were 16 cases of CVD incidence in the ≥7-year group (32.0%) and 12 cases of CVD incidence in the <7-year group (28.6%). However, CVD incidence rate did not differ significantly between the two groups (p=0.82) (table 1). Also, the duration of DM and the percentage of HTN cases did not differ significantly between the CVD group and the non-CVD group; however, the CVD group had a tendency of a longer duration of DM and a higher rate of HTN. And there was no significant difference in echocardiographic parameters between the CVD group and the non-CVD group (table 3).
The main clinical findings of the our study are as follows: (1) there was a significant correlation between the duration of DM and the LV diastolic function, (2) the E/Ea and Ea/Aa ratios were the parameters that directly changed according to the duration of DM, (3) the duration of DM had less association with CVD incidence and (4) there was no reliable echocardiographic parameter for the prediction of CVD incidence.
Several studies have shown that the E/A and E/Ea ratios and DT are well correlated with LV diastolic function and demonstrated characteristic echocardiographic findings in diabetic patients.9–11 However, there is a lack of studies on the association between duration of DM and alteration of diastolic dysfunction. Our study showed that E/Ea and Ea/Aa ratios as indices of diastolic function are significantly correlated with the duration of DM. The E/Ea ratio increased and Ea/Aa ratio decreased according to the increasing duration of DM. This finding suggests that glycosylation of protein cross-linking and excessive fibrosis is the cause of this dysfunction in diabetic cardiomyopathy. Association between duration of DM and E/Ea is consistent with a recent study.12 However, our result is not consistent with those of other studies showing the significance of the E/A ratio and DT value for evaluation of diabetic diastolic dysfunction.9 10 This discrepancy may be associated with the study group and differences in age. The ageing process and gender differences should be considered in making interpretations of diastolic function.13 The E/A ratio and DT were not compared between patients with DM and normal controls, but were compared between the ≥7-year group and the <7-year group in this study. However, the power of the E/A ratio and DT value with regard to the effect of DM duration on the heart cannot be overlooked.
We attempted to find an association between the duration of DM and CVD incidence. However, our study determined that there was less correlation between the duration of DM and CVD. Both the CVD group and the non-CVD group had a DM duration of >7 years, and more than 50% had HTN. Therefore, DM and HTN may already have induced changes in the myocardium and cardiovascular system. Moreover, all study patients underwent full conventional treatment and incidence of CVD was small. All of these causes may explain the similarity of LV systolic and diastolic functional parameters. CVD incidence is likely to be independent of the duration of DM, or not only the duration of DM; however, other factors, such as glycaemic control from the onset of the disease and comorbidities might also be of importance.14–16 Despite the absence of a significant relationships between CVD incidence and duration of DM, the impact of duration of DM on myocardial function is very important.
Longer duration of DM causes diastolic dysfunction that is independent of HTN. Our results showed that diastolic parameters differed significantly between the ≥7-year group and the <7-year group; however, systolic function and incidence of HTN were similar. Since diabetic cardiomyopathy has a high prevalence in patients with diabetes and is rarely clinically apparent unless associated with HTN and myocardial ischaemia,17 18 screening for its presence at the early stage of dysfunction is needed in order to prevent severe progression. Echocardiography can be a screening test for early detection of LV dysfunction in diabetic patients. Many factors have not yet been identified. Several studies are needed for the determination of the exact time for screening, and for demonstration of the interaction between duration of DM and other CVD risk factors. A prospective model with an accurate duration of DM, extent of diabetic abnormalities and parameters of health state may be required.
Our study has several limitations. This study was retrospective and may have selection biases of subjects. A small sample size and a short observation period yield results that lack significant clinical impact. And at the point of excluding patients with CVD prior to echocardiographic assessment and detection of CVD incidence, we only evaluated the clinical records with regard to CVD by cardiovascular related symptoms and presence of CVD diagnosis. At the point of screening, patients with subclinical cardiovascular change may be included. Coronary angiography, treadmill exercise test or stress echocardiography will be very helpful for exclusion of CVD. Large clinical trials using consistent echocardiographic assessment and other cardiovascular evaluation methods should be conducted. In this study, the association with glycated haemoglobin (HbA1C) was not analysed because small numbers of patients underwent testing for HbA1C at the time of echocardiographic examination. This is a limitation of a retrospective study. However, many previous studies have shown that diastolic dysfunction has a negative correlation with HbA1C.9 19
Despite several limitations such as small sample size and retrospective data analysis, the results highlight the importance of early active medical care including echocardiography in patients with DM because diastolic function decreases with a longer duration of DM and the CVD group had a tendency of a longer duration of DM and a higher rate of HTN.
An alteration of diastolic function induced by worsens with increasing duration of DM. The E/Ea and Ea/Aa ratios may play an important role in LV dysfunction observed in diabetic patients. Diastolic dysfunction may be related to the development of CVD; however, in order to prove it, a prospective study with a large-scale study group is warranted. DM duration on echocardiographic evaluation time did not differ significantly between the CVD and the non-CVD groups. CVD development was independent of the duration of DM after 7 years duration of DM, although a prospective large-scale study is needed to confirm this result. Therefore, we should undertake active medical care including echocardiography to prevent CVD from the point of diagnosis of type 2 DM.
There was a significant correlation between duration of diabetes mellitus (DM) and left ventricular (LV) diastolic function.
The E/Ea and Ea/Aa ratios were the parameters directly changing with the duration of DM.
The Duration of DM was poorly associated with CVD incidence.
We should undertake active medical care including echocardiography to prevent CVD from the point of diagnosis of type 2 DM.
Current research questions
A prospective model with accurate duration of DM, extent of diabetic abnormalities and parameters of health state may be required.
Large clinical trials using consistent echocardiographic assessment and other cardiovascular evaluation methods should be conducted.
The association between the level of glycaemic control and cardiovascular disease should be studied.
▶ Astorri E, Fiorina P, Contini GA, et al. Isolated and preclinical impairment of left ventricular filling in insulin-dependent and non-insulin-dependent diabetic patients. Clin Cardiol 1997;20:536–40.
▶ From AM, Scott CG, Chen HH. Changes in diastolic dysfunction in diabetes mellitus over time. Am J Cardiol. 2009 May 15;103:1463–6.
▶ Charvát J, Chlumský J, Sváb P, et al. The impact of early diastolic myocardial relaxation on the other parameters of diastolic function and association of tissue Doppler parameters of diastolic function with diabetes compensation and duration in type 2 diabetic patients. J Int Med Res. 2010;38:127–33.
Funding This study was supported by an intramural grant from Gachon University Gil Hospital.
Competing interests None.
Ethics approval This study was approved by the institutional review board of Gachon University Gil Hospital.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.