We determined the prevalence of echocardiographic abnormalities, including LV hypertrophy, RA and/or LA dilatation, and LV systolic and diastolic dysfunction, respectively, and their relation to P-NT-proBNP levels or significant CAD in asymptomatic type 2 diabetic patients with microalbuminuria that received intensive multifactorial treatment. In these patients, P-NT-proBNP levels were not associated with echocardiograpic abnormalities. LV hypertrophy was present in 24% when indexed to height2.7 and LV diastolic dysfunction was observed in 55% of patients. Finally, LV mass index to height2.7 and LV hypertrophy but not LV diastolic dysfunction were strongly and independently associated with significant CAD.
Plasma NT-proBNP and echocardiographic abnormalities
We have previously demonstrated a strong and independent risk for CV mortality with elevated P-NT-proBNP in type 2 diabetic patients , and in the present study we therefore investigated if echocardiographic abnormalities, that holds important prognostic value, were more common in patients with increased P-NT-proBNP levels . Our subjects with increased P-NT-proBNP levels did not have more LV hypertrophy, manifest atrial dilatation and LV systolic or diastolic dysfunction but we demonstrated that left and right atrial volume indexes within the normal limits were independently higher in patients with increased NT-proBNP. In animal models, BNP gene expression reflect atrial and ventricular pressures  In elderly human subjects LA > 32 ml/m2 was associated with an increased risk of heart failure . However, this cut-off represents 2 SDs from the normal mean LA volume of 22 ± 5 ml/m2 and it is therefore difficult to translate our of finding of 2.2 ml/m2 difference between the two groups. In contrast, the LA dilates in response to the chronic increased LV and LA filling pressures and therefore as such is a sensitive cumulative marker of diastolic dysfunction. Accordingly, a small structural change of LA volume could therefore reflect long-term risk (NT-proBNP) rather than the functional diastolic echocardiographic parameters, which are more subject to high fluctuation with preload and other short term conditions. This could also explain why NT-proBNP was not related to diastolic dysfunction as described in the present study. This is in line with some studies , however other studies have shown that BNP is an early determinant of both diastolic and systolic dysfunction. Specifically, in asymptomatic newly diagnosed diabetic patients, it has been demonstrated that BNP is associated with the presence of diastolic dysfunction . Accordingly, the use of BNP in the screening for preclinical moderate-severe diastolic dysfunction in diabetic patients has been suggested . Finally, BNP is increased in subjects with systolic dysfunction in the general population . The prognostic value of right atrial volume index is even more scarcely described but the right atrial association with NT-proBNP has previously been reported in other subjects . The association in the latter study, was particular important when pulmonary hypertension occurred as a result of chronic obstructive pulmonary disease . Therefore a pulmonal component of poor prognosis might be revealed in this study, however the true meaning of NT-proBNP, left and right atrial volumes will be examined in our planned follow-up of these patients. Other limitations may explain why patients with increased NT-proBNP levels did not have more major echocardiographic abnormalities in the present study. First, patients in the present study were without prior heart disease and all patients had normal P-creatinine where as subjects in our earlier study with higher prevalence of echocardiographic abnormalities included all type 2 diabetic patients followed in an out-patient clinic (4). Second, the increasingly intensive treatment used today in our patients is likely to have influenced the results compared to the earlier study , since almost all of our patients received statins, aspirin, and RAAS blockade. Noteworthy, in the Steno-2 study, the P-NT-proBNP levels increased over time in all patients, but was less in the intensive therapy group . Furthermore, as shown in Table 2, some modifiable risk factors targeted by the multifactorial treatment were lower in patients with high NT-proBNP (cholesterol, bloodpressure but not haemoglobinA1c), where as the non-modifiable risk factors such as diabetes duration, plasma creatinine and signs of peripheral artery disease were higher in patients with high NT-proBNP. Along this line and as discussed below, the low prevalences of echocardiographic moderate-severe abnormalities in this study might also dilute the putative associations with NT-proBNP levels. Third, other factors such as cardiac autonomic neuropathy, is likely to influence CV mortality. Of note, we have previously reported the correlations between NT-proBNP and significant CAD and/or atherosclerosis in different territories. Along this line, NT-proBNP levels might be a more sensitive marker of early asymptomatic prognostic ventricular disease than the echocardiographic examination.
Echocardiographic abnormalities and intensive multifactorial treatment
Individual and multifactorial intensive interventions including polypharmacologic therapy according to the Steno-2 study is the current standard of care in our institution and is known to reduce CV morbidity and mortality . Noteworthy, an audit compared the levels of treatment targets before and after the Steno-2 study (2002 and 2009) and demonstrated that the Steno-2 study results of lowering haemoglobinA1c, lipids and blood pressure were successfully implemented into clinical practise at Steno Diabetes Center . Almost all patients in the current study received statins, aspirin and RAAS blockading agents, and 33%, 28%, and 17% received 2, 3, and ≥ 4 antihypertensive drugs on top of RAAS blockade. Indeed, this intensive treatment yielded mean total cholesterol levels of 3.9 mM, arterial blood pressures of 130/75 mmHg, and haemoglobinA1c levels of 7.9%, respectively, which may explain why most of the conventional CV risk factors as well as levels of albuminuria were not associated with echocardiographic variables in our study (Table 2).
The reported prevalence of LV hypertrophy depends on classification criteria and population characteristics, ranging from 3% in the normotensive general population to almost 75% in hypertensive patients. The LV hypertrophy diagnosis varies with the methods, formulas, indexing and cut-off values that are applied. In a recent relatively large echocardiographic study with use of equivalent methods and classification schemes as used in our current study, moderate-severe LV hypertrophy (g/m2) was present in 13% of 305 type 2 diabetic patients . That study included asymptomatic patients referred to a diabetes clinic for the first time with less than 5 years of diabetes duration and without prior CV disease. In comparison, moderate-severe LV hypertrophy (g/m2) was only seen in 4% of our cases despite that our patients had microalbuminuria and frequently demonstrated other signs of microvascular disease (retinopathy and neuropathy) with average diabetes duration of 13 years. In a previous study with use of different echocardiographic methods and cut-off values, we reported that LV hypertrophy (g/m2) was present in 75% or 51% of type 2 diabetic patients with or without diabetic nephropathy, respectively, and in 9% of controls . In another study that included 426 normoalbuminuric type 2 diabetic patients without cardiovascular disease and not taking antihypertensive treatment, we also showed that LV hypertrophy (g/m2.7) was present in 43% of subjects . Had these previously used criteria been used in the present study we would have demonstrated LV hypertrophy in 21% (g/m2) or 46% (g/m2.7) of patients, respectively (data not shown). The similar prevalence of LV hypertrophy in patients with normoalbuminuria as observed in our previous study (5) and the higher prevalence of LV hypertrophy in newly referred diabetic patients , compared to our complicated patients with microalbuminuria suggests that the prevalence of LV hypertrophy has been reduced by multifactorial treatment. This is in line with our recent finding that the prevalence of carotid intima-media thickness was indeed lower than expected in the present cohort . The beneficial effect of antihypertensive agents on LV hypertrophy in diabetics is well-established, although the regression of LV hypertrophy in these patients appears to be less than in subjects without diabetes . LV hypertrophy and even increased LV mass within normal values were associated with significant CAD, and this observation may contribute to the previously described higher CV mortality in diabetic patients with residual LV hypertrophy despite treatment [9, 32]. Whether additional treatment aimed at reduction of LV hypertrophy in diabetes patients on intensive multifactorial treatment (with or without CAD) can ameliorate prognosis remains to be determined.
Our finding of more than 50% of patients with LV diastolic dysfunction in the present cohort was not unexpected, since LV diastolic dysfunction is considered to be resistant to medical treatment and prevalent in diabetic patients [5, 16]. The pathogenesis of LV diastolic function is not known in detail, but a contribution of increased afterload, e.g. due to augmented arterial stiffness and microvascular disease, reduced myocardial perfusion, and increased myocardial stiffness associated with extracellular matrix alterations, fibrosis and metabolic derangements have been suggested . Interestingly, LV diastolic dysfunction was not related to LV mass or carotid compliance in patients with or without significant CAD, but a strong association with glycaemic control (HbA1c levels) was found. This finding is in accordance with the hypothesis that hyperglycaemia can induce increased myocardial stiffness, e.g. by extracellular advanced glycation products leading to myocardial fibrosis . No patients had restrictive LV diastolic dysfunction and only 4 patients had E/e' > 15 which would indicate severely increased LV filling pressures and is considered to be pathognomonic for diastolic heart failure in the presence of heart failure symptoms and normal LV systolic function . The prevalence of moderate-severe LA dilatation was also low and particular severely abnormal LA volume ( ≥ 40 ml/m2) was only observed in 5 patients. A recent study reported LA distension ( > 32 ml/m2) in almost one third of patients with a lower CV risk profile compared to our study cohort . These observations add to the notion that intensive multifactorial treatment may reduce the detrimental cardiac consequences of diabetes and hereby potentially delay clinical heart failure and reduce CV mortality. However, as stated above the true clinical impact of the presence of echocardiographic abnormalities in our present cohort of aggressively treated patients will be examined by planned prospective follow-up.