Left Ventricular Diastolic Dysfunction in Diabetes

Background: Left ventricular diastolic dysfunction (LVDD) is a prevalent feature of cardiomyopathy in diabetes. It is also a significant predictor of cardiovascular disease in general at risk populations. B- type Natriuretic Peptide (BNP) is a well established marker of diastolic function in the general population but emerging data indicate a natriuretic peptide (NP) handicap in patients with diabetes which may compromise its use in this population. Aims: To describe the prevalence, progression and importance of LVDD in diabetes, to determine the accuracy of NP in predicting LVDD and its progression and to determine the usefulness of LVDD progression in cardiovascular risk prediction. Methods: Patients enrolled in the STOP-HF service between Jan 2011 and Dec 2013 with diabetes were included and compared with a non-diabetic cohort. Using transthoracic echocardiography (TTE), we measured LAVI in addition to standard Doppler-echocardiographic parameters (including E/E’ and left ventricular mass index (LVMI)). TTE was performed at baseline and at a predefined follow up visit (2 to 5 years post baseline). A significant increase in LAVI (LAVI progression) was defined as a change of > 3.5 ml/m2. BNP was measured at baseline and follow up visits. Hospitalisation for major adverse cardiovascular events (MACE) was identified from hospital discharges at least 2 years after the baseline visit. Multivariable-adjusted logistic regression was used to identify factors associated with LAVI progression and hospitalisation for MACE. Results: There were 1238 patients enrolled in the service. Four hundred and forty seven (36.1%) of these were diabetic (mean age 64.7 years). The mean LAVI was 26.1 ml/m2 in the diabetic group and 26.8 ml/m2 in the non diabetic group (p = 0.5). LAVI progression was similar in both groups (31.6% DM vs 29.4% non DM, p = 0.44). Use of RAAS modifying therapy was higher in the diabetic group (56.3% vs 43.1%, p < 0.001). MACE occurred in 7.2% of the diabetic population and 3.2% of the non diabetic population (p = 0.001). MACE occurred in 6.4% of those with LAVI progression and in 3.7% of those without LAVI progression (p = 0.047). Baseline BNP was predictive of LAVI progression in the non diabetic group (OR 2.00, CI 1.04-3.86) but not in the diabetic group (OR 2.16, CI 0.86-5.42). Progression of BNP to an abnormal level at follow up was predictive of LAVI progression in the non diabetic group (OR 2.86, CI 1.62-5.07) but not in the diabetic group (OR 1.41, CI 0.64-3.09). Baseline BNP and progression of BNP to an abnormal level at follow up was predictive of MACE hospitalisation (OR 3.6, CI 1.54-8.41 and OR 3.93, CI 1.83-8.42 respectively). LAVI progression was associated with MACE on univariate analysis (OR 1.79, CI 1.00, 3.20) but not on multivariate analysis (OR 1.32, CI 0.71-2.44). Conclusions: Preclinical diastolic dysfunction is prevalent among patients with diabetes, progresses over time in a significant proportion and is potentially linked to clinical events. While NP identified the non-diabetic at risk group, it did not predict risk in those with diabetes suggesting a handicap in NP function. This leaves an opportunity to intervene in the natural history of diabetic cardiomyopathy by modulating the NP system.