Prevalence of systolic and diastolic dysfunction in patients with type 1 diabetes without known heart disease: the Thousand & 1 Study

Diabetologia (2014) 57:672 680 DOI 10.1007/s00125-014-3164-5 ARTICLE Prevalence of systolic and diastolic dysfunction in patients with type 1 diabetes without known heart disease: the Thousand & 1 Study Magnus T. Jensen & Peter Sogaard & Henrik U. Andersen & Jan Bech & Thomas F. Hansen & Søren Galatius & Peter G. Jørgensen & Tor Biering-Sørensen & Rasmus Møgelvang & Peter Rossing & Jan S. Jensen Received: 24 November 2013 /Accepted: 20 December 2013 /Published online: 22 January 2014 # Springer-Verlag Berlin Heidelberg 2014 Abstract Aims/hypothesis Heart failure is one of the leading causes of mortality in type 1 diabetes. Early identification is vitally important. We sought to determine the prevalence and clinical characteristics associated with subclinical impaired systolic and diastolic function in type 1 diabetes patients without known heart disease. Methods In this cross-sectional examination of 1,093 type 1 diabetes patients without known heart disease, randomly selected from the Steno Diabetes Center, complete clinical and echocardiographic examinations were performed and analysed in uni- and multivariable regression models. Results The mean (SD) age was 49.6 (15)years, 53% of participants were men, and the mean duration of diabetes was 25.5 Electronic supplementary material The online version of this article (doi:10.1007/s00125-014-3164-5) contains peer-reviewed but unedited supplementary material, which is available to authorised users. M. T. Jensen (*): P. Sogaard : J. Bech : T. F. Hansen : S. Galatius : P. G. Jørgensen: T. Biering-Sørensen : R. Møgelvang : J. S. Jensen Department of Cardiology, Copenhagen University Hospital Gentofte, Niels Andersens Vej 65, 2900 Hellerup, Denmark e-mail: magnustjensen@gmail.com M. T. Jensen: H. U. Andersen : P. Rossing Steno Diabetes Center, Gentofte, Denmark P. G. Jørgensen: J. S. Jensen Institute of Clinical Medicine, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark P. Rossing Novo Nordisk Foundation Center for Basic Metabolic Research, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark P. Rossing Faculty of Health, University of Aarhus, Aarhus, Denmark (15)years. Overall, 15.5% (n=169) of participants had grossly abnormal systolic or diastolic function, including 1.7% with left ventricular ejection fraction (LVEF)<45% and 14.4% with evidence of long-standing diastolic dysfunction. In univariable models, clinical characteristics associated with abnormal myocardial function were: age (per 10 years), OR (95% CI) 2.1 (1.8, 2.4); diabetes duration (per 10 years), 1.7 (1.4, 1.9); systolic BP 140 mmhg, 2.7 (1.9, 3.8); diastolic BP 90 mmhg, 1.8 (1.0, 3.1); estimated (e)gfr<60 ml min 1 1.73 m 2, 3.8 (2.5, 5.9); microalbuminuria, 2.0 (1.3, 3.0); macroalbuminuria, 5.9 (3.8, 9.3); proliferative retinopathy, 3.6 (2.3, 5.8); blindness, 10.1 (3.2, 31.6); and peripheral neuropathy, 3.8 (2.7, 5.3). In multivariable models only age (2.1 [1.7, 2.5]), female sex, (1.9 [1.2, 2.8]) and macroalbuminuria (5.2 [2.9, 10.3]) remained significantly associated with subclinical grossly abnormal myocardial function. Conclusions/interpretation Subclinical myocardial dysfunction is a common finding in type 1 diabetes patients without known heart disease. Type 1 diabetes patients with albuminuria are at greatly increased risk of having subclinical abnormal myocardial function compared with patients without albuminuria. Echocardiography may be particularly warranted in patients with albuminuria. Keywords Albuminuria. Diabetes. Diastolic dysfunction. Echocardiography. Heart disease. Heart failure. Prevention. Systolic dysfunction. Type 1 diabetes Abbreviations A ACE-I ASE ATII-A CVD Mitral atrial inflow velocity Angiotensin-converting enzyme inhibitor American Society of Echocardiography Angiotensin II receptor antagonist Cardiovascular disease

Diabetologia (2014) 57:672 680 673 E Mitral early inflow velocity e Early diastolic tissue Doppler velocity EAE European Association of Echocardiography E/e Estimated left ventricular filling pressure egfr Estimated GFR FLEMENGHO Flemish study on Environment, Genes and Health Outcomes ICD Implantable cardioverter defibrillator IQR Interquartile range LVEF Left ventricular ejection fraction uacr Urinary albumin/creatinine ratio uaer Urinary AER Introduction Cardiovascular disease (CVD) is the most common cause of death in diabetes. In patients with type 1 diabetes, the risk of death due to CVD is increased between six and 12 times compared with the general population [1] and the risk for CVD in type 1 diabetes patients under age 40 is increased up to 40-fold [2]. A register-based study has shown that the incidence of heart failure in type 1 diabetes patients between 41 and 45 years of age was equal to that in individuals from the general population aged 55 64 years [3]. While most of the knowledge about heart disease in type 1 diabetes comes from register-based or smaller clinical studies [4 7], very few studies have assessed heart disease in type 1 diabetes in a large-scale clinical setting. Diabetes is a systemic disease that affects not only the vasculature but also the myocardial tissue, i.e. diabetic cardiomyopathy [8]. A feature of the diabetic cardiomyopathy is stiffening of the myocardial tissue, leading to diastolic dysfunction, increased preload and increasing filling pressures [9]. Diastolic dysfunction is associated with poor prognosis [10] and can progress to systolic dysfunction [11]. In order to understand and recognise diabetic heart disease and to identify type 1 diabetes patients at particular risk clinical studies of myocardial function are much needed. In the present study we investigated myocardial function and haemodynamic variables, and their association with clinical characteristics in 1,093 randomly selected ambulatory type 1 diabetes patients without known heart disease. Methods Steno Diabetes Center and Department of Cardiology The Steno Diabetes Center is an integrated part of the Danish public healthcare system and 90% of the patients are referred from the Copenhagen Capital Region. The total number of patients followed at Steno Diabetes Center is around 6,000, of whom 3,500 patients have type 1 diabetes and the remaining patients have type 2 diabetes. Type 1 diabetes patients are followed for life, with ambulatory visits every third month at the outpatient clinic at Steno Diabetes Center. The Department of Cardiology, Copenhagen University Hospital (Gentofte, Denmark) is an invasive centre with a core echocardiography laboratory performing more than 6,400 echocardiographic examinations annually. Study population The Thousand & 1 Study is a cohort study of patients with type 1 diabetes without known heart disease. Invitation, screening and inclusion of patients started on 1 April 2010 and inclusion was completed on 1 April 2012. Patients were eligible if they were: 18 years or older; attending the outpatient clinic at the Steno Diabetes Center; diagnosed with type 1 diabetes; without known heart disease; and willing to participate. Known heart disease was any known heart failure, coronary artery disease, including previous myocardial infarction, stable angina, previous percutaneous coronary intervention, or coronary artery bypass surgery, atrial fibrillation or atrial flutter, left bundle branch block, congenital heart disease, pacemaker or implantable cardioverter defibrillator (ICD) implantation, which were all exclusion criteria. No financial compensationwasofferedtothepatients for participation. Invitations to participate in the study were sent to 1,998 patients randomly selected from the Steno Diabetes Center (Fig. 1). Patients were invited if they had undergone the biennial status examination within 6 months. During the status examination patients undergo testing for albuminuria, peripheral neuropathy and retinopathy, extensive biochemical testing and thorough clinical history and examination; the status examination is therefore a more extensive clinical assessment than the usual routine examination conducted with 3 4 month intervals. This method of selecting patients ensured inclusion of a random sample of the population and updated clinical information. Patients were screened by telephone or email for history of heart disease and willingness to participate. Patients who fulfilled the inclusion criteria were examined at the Department of Cardiology. During the screening process 115 patients (5.8%) were excluded because of known heart disease and 790 patients (39.5%) were non-responders or not interested in participating, leaving a total of 1,093 patients (54.7%) to be included in the Thousand & 1 Study (Fig. 1). The basic characteristics for patients included and not included are shown in Table 1. Patients in the not included, known heart disease category were older, had longer diabetes duration, higher HbA 1c and more complications in the form of kidney disease, retinopathy and peripheral neuropathy. The included patients differed from the not included, nonresponders, not interested in participation in terms of age, where the included patients were 2 years older, with slightly higher BMI, lower HbA 1c and lower prevalence of

674 Diabetologia (2014) 57:672 680 Steno Diabetes Center Patients with type 1 diabetes from the Copenhagen Capital Region n=3,500 Invitation Patients with type 1 diabetes from the outpatient clinic at Steno Diabetes Center randomly selected and invited to participate in the Thousand & 1 Study n=1,998 Not included Known heart disease n=115 Not included Non-responders or not interested in participation n=790 Echocardiography Echocardiography was performed with a General Electric, Vivid 7 Dimension imaging system device (GE Vingmed Ultrasound, Horten, Norway) with a 3.5 MHz transducer, in accordance with the recommendations from the European Association of Echocardiography (EAE)/American Society of Echocardiography (ASE) [12]. Echocardiographic examinations were read and analysed using General Electric EchoPAC software (BT11). Three consecutive heart cycles were recorded. LVEF was determined by Simpson s biplane method. Left atrial volume was determined by the recommended biplane area length method, which is (8/3π) (A 1 A 2 /L), where A 1 and A 2 are areas from planimetry of the four chamber and two chamber views, and L is the shortest atrial length of the two projections, and indexed for body surface area. Left ventricular (LV) mass was determined by the linear method (0.8 [1.04{LV diameter + posterior wall thickness + septal wall thickness} 3 {LV diameter} 3 ]+0.6 g) and indexed for body surface area. Pulsed-wave Doppler was performed in the apical four-chamber view with the sample volume placed between the mitral leaflet tips to obtain diastolic mitral early (E) and atrial (A) inflow velocities and deceleration time of the E velocity wave. Pulsed-wave early diastolic tissue Doppler velocities (e ) were determined from the apical four-chamber view at the lateral region of the mitral annulus [13]. Included in the Thousand & 1 Study n=1,093 Fig. 1 Flow chart showing the design of the Thousand & 1 Study albuminuria (30.9% vs 37.9%). However, there were no differences in sex, diabetes duration, estimated (e)gfr, degree of retinopathy or peripheral neuropathy. Thus, the current study population of patients included in the Thousand & 1 Study was similar or with slightly better controlled diabetes compared with patients who were non-responders or not interested in participation. However, there was no evidence of any major selection bias. The study was performed in accordance with the second Helsinki declaration, approved by the regional ethics committee (H-3-2009-139), and the Danish Data Protection Agency (00934-Geh-2010-003). All participants gave written informed consent. Study visit Prior to examination all patients received study information, signed the consent form and filled out a questionnaire with information about lifestyle factors, including smoking, exercise, alcohol consumption, cardiorespiratory symptoms and use of medication. Electrocardiogram (Cardiosoft version 6.61, GE Healthcare) and BP at rest were recorded in a supine position. Systolic and diastolic myocardial impairment Systolic dysfunction was determined as LVEF <45%. As the classification of diastolic dysfunction is inherently difficult [14] we classified patients as having diastolic dysfunction if there was evidence of long-standing elevated LV filling pressure defined as E/e 12 or E/e 8 to 12 and left atrial volume >34 ml/m 2. We defined pathological myocardial function as LVEF <45% and/or if there was evidence of diastolic dysfunction. We also graded the patients in groups of diastolic dysfunction according to the guidelines from the EAE/ASE [13], though these guidelines do not take into account the normal age-related decrease in tissue Doppler velocities. Patients were categorised as normal if e 10 cm/s; impaired relaxation (mild diastolic dysfunction) if e <10 cm/s and E/A velocity ratio <0.8; pseudo normal filling (moderate diastolic dysfunction) if e <10 cm/s and E/A velocity ratio 0.8 2 orif e <10 cm/s, E/A velocity ratio 2 and E/e <12; or restrictive filling (severe diastolic dysfunction) if e <10 cm/s, E/A velocity ratio 2 and E/e 12. ECGs, BP measurements and informed consent were obtained by Magnus T. Jensen. The echocardiographic examinations were performed, read and analysed by Magnus T. Jensen; all examinations were validated with a second opinion from another experienced imaging cardiologist. Inter- and intraobserver variability for Simpson s biplane LVEF was assessed with 25 randomly selected patients and showed good agreement, with no systematic bias; the limits of agreement were (mean of difference [± 2SD]) 0.6% ( 9.0, 7.0) for

Diabetologia (2014) 57:672 680 675 Table 1 Patients invited to participate in the Thousand & 1 Study (included and not included) Variable Included Not included p value Non-responders, not interested in participation Known heart disease Included vs non-responders, not interested in participation Included vs known heart disease Number of patients 1,093 790 115 Age (years) a 50.5 (14.5) 48.5 (16.9) 63.6 (11.1) 0.007 <0.001 Male sex (%) 52.6 55.6 54.8 0.20 0.66 Diabetes duration (years) a 27.2 (16.1) 27.1 (18.2) 38.5 (13.9) 0.85 <0.001 BMI (kg/m 2 ) 25.5 (3.9) 25.1 (4.3) 26.1 (4.0) 0.008 0.47 HbA 1c <0.001 <0.001 (%) 8.2 (1.3) 8.5 (1.5) 8.7 (1.6) (mmol/mol) 66 (14) 69 (17) 72 (17) egfr (ml min 1 1.73 m 2 ) a 87 (22) 87 (27) 68 (27) 0.90 <0.001 Albuminuria 0.004 <0.001 Normo- (%) 69.1 62.1 37.2 Micro- (%) 21.0 24.0 43.4 Macro- (%) 9.9 13.9 19.5 Retinopathy 0.70 <0.001 Nil (%) 36.4 35.0 18.3 Simple (%) 46.9 47.6 37.4 Proliferative (%) 15.5 15.5 40.0 Blind (%) 1.2 1.8 4.4 Peripheral neuropathy (%) 36.9 39.5 77.4 0.25 <0.001 Values are shown as mean (SD) a 1 April 2012 set as index date intraobserver variability, and 0.6% ( 11.0, 10.0) for interobserver variability (Bland Altman plots not shown) [15]. Retinopathy status was assessed from retinal photographs taken through dilated pupils and graded as nil, simple, proliferative or blind based on the worst eye. All patients attending the Steno Diabetes Center have regular ophthalmological examinations, and information on retinopathy for the present study was drawn from the electronic patient file. Peripheral neuropathy is assessed as part of the routine visits and measured as vibration threshold with biothesiometry. In the present study, a value >25 V was considered to indicate peripheral neuropathy. Biochemistry Information about biochemistry, such as HbA 1c, p-creatinine and albuminuric status was collected from the electronic patient files at Steno Diabetes Center from the ambulatory visit closest to study inclusion, which was maximally ± 4 months from inclusion. When performing, reading and analysing the echocardiographic examinations the investigators were blinded to information on biochemistry, including albuminuria, retinopathy and other complications. Urinary albumin/creatinine ratio (uacr) was measured in morning spot urine samples by immunoassay and patients were categorised as normoalbuminuric if uacr was below 3.4 mg/mmol creatinine. Elevated values >3.4 were confirmed in 24 h sterile urine collections (urinary [u]aer). Patients were categorised as microalbuminuric if the uaer, in two out of three consecutive measurements, was between 30 and 300 mg/24 h, and macroalbuminuric if uaer>300 mg/24 h. If the uaer subsequently was reduced with renoprotective treatment, patients were not reclassified. HbA 1c was measured by high-performance liquid chromatography (normal range: 21 46 mmol/mol, [4.1 6.4%]; Variant, Bio-Rad Laboratories, Munich, Germany) and serum creatinine concentration by an enzymatic method (Hitachi 912, Roche Diagnostics, Mannheim, Germany). The egfr values were calculated by the Modification of Diet in Renal Disease (MDRD) method [16]. Statistical analysis All analyses were performed with STATA 12.1 (StataCorp LP, College Station, TX, USA). Categorical variables were analysed with the χ 2 test and continuous variables with Student s t test or ANOVA. Continuous variables are reported as means (SD); in the case of non-normal distribution the data are reported as median (interquartile range [IQR]). Uni- and multivariable logistic regression models were performed to test the associations between clinical characteristics

676 Diabetologia (2014) 57:672 680 and echocardiographic findings. A p value <0.05 was considered statistically significant. Results Clinical characteristics Baseline clinical and echocardiographic characteristics of the 1,093 patients included in the study are shown in Table 2. The mean age was 49.6 years, and 53% of the population were men. The mean HbA 1c was 8.2%, mean BMI was 25.5 kg/m 2 and the mean duration of diabetes was 25.5 years. Systolic dysfunction and long-standing diastolic dysfunction An LVEF<45% was found in 1.7% of the population. The clinical characteristics associated with impaired myocardial function are shown in Table 2. Patients with systolic dysfunction were older, more likely to be men, to have higher diastolic BP and have macroalbuminuria. As shown in Table 3, the systolic dysfunction was accompanied by markedly decreased e velocity, altered Doppler profile and increased LV mass. A total of six patients (33.3%) with systolic dysfunction were also categorised as having long-standing diastolic dysfunction. Evidence of long-standing diastolic dysfunction was found in 14.4% of the population (Table 2)andwasassociatedwith Table 2 Demographics Variable All patients Normal function vs myocardial dysfunction Systolic dysfunction Long-standing diastolic dysfunction LVEF 45% LVEF<45% p value Normal Impaired p value Number (%) 1,093 (100) 1,075 (98.4) 18 (1.7) NA 936 (85.6) 157 (14.4) NA Age (years) 49.6 (15) 49.5 (15) 58.7 (13) 0.007 47.7 (14) 61.1 (11.8) <0.001 Male sex (%) 52.7 52.2 77.8 0.03 54.1 43.3 0.012 Diabetes duration (years) 25.5 (15) 25.6 (15) 25.5 (20) 0.98 24.1 (14) 35.7 (15) <0.001 BMI (kg/m 2 ) 25.5 (3.9) 25.5 (3.9) 25.5 (4.3) 0.93 25.5 (4) 25.9 (4) 0.19 HbA 1c 0.34 0.79 (%) 8.2 (1.3) 8.2 (1.3) 8.5 (1.6) 8.2 (1.3) 8.2 (1.1) (mmol/mol) 66 (14) 66 (14) 69 (4) 66 (14) 66 (11) Ever smoker (%) 56.2 55.9 72.2 0.17 54.5 66.2 0.006 Systolic BP (mmhg) 133 (16) 133 (16) 138 (17) 0.25 132 (16) 142 (18) <0.001 Diastolic BP (mmhg) 74 (10) 73 (10) 80 (12) 0.006 74 (10) 73 (11) 0.22 Family history of CVD (%) 38.7 38.6 44.4 0.25 39.2 35.7 0.40 egfr (ml min 1 1.73 m 2 ) 87 (22) 87 (22) 92 (29) 0.36 90 (21) 73 (25) <0.001 Albuminuria 0.19 <0.001 Normo- (%) 69.1 69.7 61.1 73.4 46.5 Micro- (%) 21.0 20.8 16.7 19.7 27.4 Macro- (%) 9.9 9.5 22.2 6.9 26.1 Retinopathy 0.084 <0.001 Nil (%) 36.4 36.1 55.6 38.8 21.9 Simple (%) 46.9 47.1 33.3 47.8 41.3 Proliferative (%) 15.5 15.7 5.6 12.7 32.3 Blind (%) 1.2 1.1 5.6 0.6 4.5 Peripheral neuropathy (%) 36.9 36.6 55.6 0.10 32.3 64.3 <0.001 Medication Statin (%) 43.5 43.4 50.0 0.57 39.7 65.6 <0.001 Beta blocker (%) 4.6 4.6 5.6 0.84 2.8 15.3 <0.001 ACE-I/ATII-A (%) 46.0 44.7 66.7 0.076 40.9 76.4 <0.001 Calcium antagonist (%) 18.9 18.8 27.8 0.34 15.1 42.0 <0.001 Diuretic (%) 26.1 25.9 38.9 0.21 22.4 47.8 <0.001 Insulin (%) 100 NA NA Values are mean (SD) unless otherwise stated NA, not applicable

Diabetologia (2014) 57:672 680 677 Table 3 Echocardiographic characteristics All patients Normal vs myocardial dysfunction Systolic dysfunction Long-standing diastolic dysfunction Variable LVEF 45% LVEF<45% p value Normal Impaired p value Number (%) 1,093 (100) 1,075 (98.4) 18 (1.7) NA 936 (85.6) 157 (14.4) NA Echocardiography LVEF (%) 58 (5) 58 (5) 41 (3) NA 58 (5) 57 (6) 0.31 Dimensions LV internal diameter (cm) 4.5 (0.5) 4.5 (0.4) 4.8 (0.6) 0.08 4.6 (0.5) 4.4 (0.5) <0.001 Septal wall (cm) 0.9 (0.2) 0.9 (0.2) 1.0 (0.2) 0.01 0.9 (0.2) 1.0 (0.2) <0.001 Posterior (cm) 0.9 (0.2) 0.9 (0.2) 1.0 (0.1) 0.057 0.9 (0.1) 1.0 (0.2) <0.001 E peak (m/s) 0.9 (0.2) 0.9 (0.2) 0.6 (0.2) <0.001 0.8 (0.2) 1.0 (0.2) <0.001 A peak (m/s) 0.7 (0.2) 0.7 (0.2) 0.8 (0.3) 0.30 0.7 (0.2) 1.0 (0.2) <0.001 E/A ratio 1.26 (0.5) 1.26 (0.5) 0.8 (0.3) <0.001 1.29 (0.5) 1.06 (0.4) <0.001 E deceleration time (ms) 195 (44) 195 (44) 199 (52) 0.75 195 (41) 204 (57) 0.007 e, lateral (cm/s) 12 (4) 13 (4) 7 (3) 0.001 12 (4) 8 (3) <0.001 E/e, median (IQR) 6.9 (3.3) 6.9 (3.3) 7.4 (4.9) 0.14 6.5 (2.5) 12.1 (4.6) <0.001 LA volume index (ml/m 2 ) 30 (7) 30 (6) 30 (9) 0.076 29 (6) 35 (8) <0.001 LV mass index (g/m 2 ) 71 (15) 71 (15) 85 (11) <0.001 70 (14) 79 (17) <0.001 Values are mean (SD) unless otherwise stated NA, not applicable age (electronic supplementary material [ESM] Table 1). Using the EAE/ASE grade of diastolic dysfunction, 30.7% were considered to have diastolic dysfunction, including 9.2% with mild impairment, 21.3% with moderate dysfunction and 0.2% with severe dysfunction. As shown in Table 2, compared with patients without longstanding diastolic dysfunction patients were older, more likely to be women, more likely to have ever smoked, to have higher systolic BP and more complications in the form of albuminuria, retinopathy, peripheral neuropathy, and were more likely to be receiving cardio- and renoprotective medication. There was no difference in LVEF between patients with and without long-standing diastolic dysfunction. However, all other echocardiographic characteristics were significantly different, including Doppler velocities, ventricular dimensions and peak E deceleration time. Clinical characteristics and myocardial dysfunction The clinical characteristics associated with a pathological echocardiogram were examined in uni- and multivariable models. A pathological echocardiogram was defined as LVEF <45% and/or long-standing diastolic dysfunction and included 15.5% (n=169) of the population. Univariable The clinical characteristics are shown in Table 4. Patients with a pathological echocardiogram were more likely to be older, have longer diabetes duration, hypertension, decreased kidney function, albuminuria, retinopathy, neuropathy and to receive medication. In these univariable analyses patients with decreased egfr were almost four times more likely to have subclinical myocardial impairment than patients with egfr>60 ml min 1 1.73 m 2. For patients with macroalbuminuria the risk was increased sixfold, for blind patients it was increased tenfold and for patients with peripheral neuropathy the risk of significant myocardial dysfunction was increased almost fourfold. Multivariable In a multivariable model including all clinical characteristics the following remained significant: age per 10 year increase (OR [95% CI] 2.1 [1.7, 2.5], p<0.001); female sex (1.9 [1.2, 2.8], p=0.003) and macroalbuminuria (5.2 [2.9, 10.3], p<0.001). Systolic BP and BMI were borderline significant (1.4 [0.9, 2.1], p=0.08, and 1.04 [1.0, 1.1], p= 0.08); no other characteristics reached statistical significance (all p>0.1). When including medication in the full model, patients who received beta blockers were 2.3 times ([1.1, 4.5], p=0.019) as likely to have subclinical myocardial impairment and patients receiving angiotensin-converting enzyme inhibitor (ACE-I)/angiotensin II receptor antagonist (AII-A) 1.6 times ([1.0, 2.5], p=0.057) as likely, while there was no association in patients receiving calcium antagonists or statins (p>0.1).

678 Diabetologia (2014) 57:672 680 Table 4 Clinical characteristics associated with a pathological echocardiogram Association with an abnormal echocardiogram OR (95% CI) Discussion p value Age, per 10 year increase 2.1 (1.8, 2.4) <0.001 Male sex 0.8 (0.6, 1.1) 0.10 Diabetes duration, per 10 year increase 1.7 (1.4, 1.9) <0.001 BMI (kg/m 2 ), per unit increase 1.03 (1.0, 1.1) 0.18 HbA 1c (%), per unit increase 1.0 (0.9, 1.3) 0.92 Ever smoker 1.8 (1.2, 2.5) 0.001 Systolic BP 140 mmhg 2.7 (1.9, 3.8) <0.001 Diastolic BP 90 mmhg 1.8 (1.0, 3.1) 0.04 Family history of CVD 0.9 (0.6, 1.2) 0.45 egfr<60 ml min 1 1.73 m 2 3.8 (2.5, 5.9) <0.001 Albuminuria Micro- vs normo- 2.0 (1.3, 3.0) 0.001 Macro- vs normo- 5.9 (3.8, 9.3) <0.001 Retinopathy Simple vs nil 1.4 (0.9, 2.0) 0.15 Proliferative vs nil 3.6 (2.3, 5.8) <0.001 Blind vs nil 10.1 (3.2, <0.001 31.6) Peripheral neuropathy (%) 3.8 (2.7, 5.3) <0.001 Medication Beta blocker 6.2 (3.5, 11.2) <0.001 ACE-I/ATII-A 4.4 (3.0, 6.4) <0.001 Calcium antagonist 4.1 (2.8, 5.8) <0.001 Diuretic 3.2 (2.3, 4.5) <0.001 Even though CVD is the most common complication of type 1 diabetes, echocardiography is not routinely performed in patients with type 1 diabetes. The main question is how to identify the patients at risk for future cardiovascular events, i.e. which patients and clinical characteristics should prompt further cardiac evaluation. This is what we sought to determine in the present study comprising a representative sample of 1,093 type 1 diabetes patients aged 18 to 85 years without known heart disease. Of these patients, we found that around 2% had LVEF<45%. Second, more than 14% had evidence of long-standing diastolic dysfunction and more than 30% were determined to have diastolic dysfunction using the EAE/ASE guidelines. Third, we identified the clinical characteristic associated with particular risk of having subclinical myocardial function: besides age and sex, macroalbuminuria was, independently of all other risk factors, associated with a fivefold likelihood of pathologically impaired myocardial function. Micro- and macroalbuminuria are well-known risk factors for CVD and mortality in type 1 diabetes [17]. Here, we find measurable subclinical changes in myocardial function that not only contribute to the mechanical understanding of the increase in incidence of heart failure in type 1 diabetes [3], but also suggest which patients should receive particular attention. Routine cardiac evaluation may be warranted in at-risk patients in order to detect subclinical dysfunction and prevent manifest disease. In the present population, the prevalence of undiagnosed systolic dysfunction is around 2%; this is comparable with the prevalence of LVEF 40% in an older (by 13.2 years) North American general population as reported by Redfield et al [18], who studied 2,042 individuals with a mean age of 62.8 years. In the Copenhagen City Heart Study [19] the prevalence of LVEF<50% in 1,036 individuals from the background population with a mean age of 59.8 years was reported to be only 1.1%, and in a British echocardiographic study [20] of 3,960 individuals from the background population with a mean age 61 years, the prevalence of LVEF<40% was reported to be 1.8%. Thus, the prevalence of systolic dysfunction in the current population of 1,093 patients with type 1 diabetes is similar to the reported prevalence of systolic dysfunction in various general populations. However, the important difference is that these type 1 diabetes patients are 10 13 years younger and without known heart disease. In a recent register-based study of 20,000 Swedish patients with type 1 diabetes the incidence of heart failure during 9 years of follow-up was similar to patients 10 15 years older from the background population [3]. Our results not only agree with these findings but provide the echocardiographic and haemodynamic characteristics of myocardial performance in type 1 diabetes. Diastolic dysfunction was very common in these patients with type 1 diabetes. Increased LV filling pressure, which is a strong indicator of long-standing diastolic dysfunction, is associated with a poor prognosis [21] and was common in this population of type 1 diabetes patients. Redfield et al reported a prevalence of diastolic dysfunction, using criteria similar to those in the EAE/ASE guidelines, that was comparable with the prevalence in the present study (28.1% vs 30.8%). However, in the present study patients were more than 13 years younger and, in contrast to the study by Redfield et al, moderate diastolic dysfunction was more prevalent than mild dysfunction. This is interesting, as moderate diastolic dysfunction is considered to have a poor prognosis [10]. Further evidence of altered diastolic function in our population of type 1 diabetes patients can be seen when comparing the E/A ratio with a healthy reference of 239 participants from the Belgian Flemish study on Environment, Genes and Health Outcomes (FLEMENGHO) study [22]. In age-stratified analyses of the E/A ratio the current type 1 diabetes population had an E/A ratio that was equal to a healthy population older by 10 15 years: the mean (SD) E/A ratio for the healthy reference in the FLEMENGHO study was 1.67 (0.41) for age 30 39 years compared with 1.30 (0.35) in the current study (p<0.001; Student s t test); 1.47 (0.34) vs 1.10 (0.32) for age 40 49 years

Diabetologia (2014) 57:672 680 679 (p<0.001) and 1.18 (0.21) vs 1.00 (0.30) for age 50 59 years (p<0.001). A comparison of the mean E/A ratios for both populations, who have similar mean ages, reveals that the difference is highly significant (1.48 [0.46] vs 1.26 [0.46], p<0.001). These findings suggest that myocardial function is affected much earlier in type 1 diabetes patients compared with healthy controls. Diastolic dysfunction is a common finding in diabetes populations [23]. Diastolic dysfunction may therefore be the first sign of diabetic cardiomyopathy [8, 24, 25]. As the LV myocardium becomes stiffer, possibly because of fibrosis, changes in calcium metabolism or hypertrophy [11], the inflow pattern of blood from the left atrium changes, with a shift towards less diastolic early ventricular filling and increased importance of left atrial function. As the left ventricle becomes even stiffer and the filling pressure increases, the inflow pattern changes towards a shorter filling time, higher inflow velocities and an enlarged atrium. The results are that patients with diabetes develop an initial decrease in E/A ratio earlier than those without diabetes. As the myocardium becomes stiffer yet, with a resulting decrease in the elastic recoil, e will decrease [26] and E/e will increase [9, 27]. As diastolic dysfunction often precedes systolic dysfunction, diastolic dysfunction may therefore not only be an indicator of diabetic cardiomyopathy but mayalsoidentifypatientsatparticular risk of developing heart failure and increased mortality [23]. As diabetic cardiomyopathy gives rise to measurable changes in myocardial function, we can, and perhaps to a greater extent should, perform echocardiography in patients at increased risk. Potential study limitations should be considered. As all patients were ambulatory and physically capable of visiting the Department of Cardiology for examination there may have been a selection bias toward including healthier patients with fewer complications than in the general type 1 diabetes population without known heart disease. However, as shown in Table 1, patients who were non-responders or not interested in participation were quite similar to the patients included. There was a small difference in prevalence of macroalbuminuria, but on the other hand they tended to be around 2 years younger. In any case, if there has indeed been a selection bias towards a healthier population this would draw the findings towards the null hypothesis and could therefore not explain the findings; the true prevalence of cardiac dysfunction may therefore have been underestimated in the present study. In addition, patients were included across the entire adult age span ( 18 years) and any potential biases from selective age criteria have therefore been avoided. Second, inclusion of a control population could potentially have strengthened the findings. Third, the echocardiographic measurements and definition of systolic and diastolic dysfunction may differ from those used in other studies. However, we used Simpson s biplane method of estimating LVEF. This method is not flawless, but is generally well understood. The interand intraobserver variability showed good agreement and reproducibility. Also, the definition and classification of diastolic dysfunction may vary across different studies and different echocardiography readers [14]. However, we are reporting prevalence using both the EAE/ASE grading guidelines and increased LV filling pressure as evidence of long-standing diastolic dysfunction using a combination of left atrial volume and E/e. This has two purposes. First, using elevated filling pressure as a combination of E/e and left atrial size is a very simple and reproducible approach. Second, whereas diastolic dysfunction to a certain degree is a normal and physiological feature of increasing age grossly increased LV filling pressure is not [18, 23]. Last, the population in the Thousand & 1 Study may not be representative of all type 1 diabetes patients without known heart disease, as the majority of patients were of Scandinavian descent. In conclusion, in this study of 1,093 patients with type 1 diabetes and no known heart disease the most significant and consistent clinical predictor of subclinical myocardial impairment was albuminuria. Overall, we find that myocardial performance decreased earlier than would be expected when comparing with background populations. Compared with other echocardiographic studies, the prevalence of cardiac dysfunction and haemodynamic changes are comparable with a background population 10 13 years older. Diabetic cardiomyopathy gives rise to measurable changes in myocardial function which can be detected by echocardiography. Early diagnosis of heart failure before manifest disease develops could possibly improve by performing echocardiography in patients with type 1 diabetes and particularly those with albuminuria. Acknowledgements We are indebted to the staff and patients of the Steno Diabetes Center for their participation and contribution to the Thousand & 1 Study. Funding The support for the Thousand & 1 Study has been provided by the European Foundation for the Study of Diabetes/Pfizer European Programme 2010 for Research into Cardiovascular Risk Reduction in Patients with Diabetes; and The Danish Heart Foundation (number12-04- R90-A3840-22725). Additional funding has been received from the Torben & Alice Frimodts Foundation, Carl & Ellen Hertz Legat til Dansk Læge-og Naturvidenskab and the Beckett Foundation. Duality of interest The authors declare that there is no duality of interest associated with this manuscript. Contribution statement MTJ was involved in the design of the study, the acquisition, analysis and interpretation of data and drafting the manuscript and gave final approval of the version to be published. PS, PR and JSJ were involved in the conception and design of the study and the acquisition and interpretation of data, revised the manuscript critically for important intellectual content and gave final approval of the version to be published. JSJ is responsible for the integrity of the work as a whole. HUA, JB and TFH made substantial contributions to the acquisition of data, revised the manuscript critically for important intellectual content and gave final approval of the version to be published. SG, PGJ, TB-S and RM made substantial contributions to the interpretation of data, revised the manuscript critically for important intellectual content and gave final approval of the version to be published.

680 Diabetologia (2014) 57:672 680 References 1. Soedamah-Muthu SS, Fuller JH, Mulnier HE et al (2006) High risk of cardiovascular disease in patients with type 1 diabetes in the U.K.: a cohort study using the general practice research database. Diabetes Care 29:798 804 2. Laing SP, Swerdlow AJ, Slater SD et al (2003) Mortality from heart disease in a cohort of 23,000 patients with insulin-treated diabetes. Diabetologia 46:760 765 3. Lind M, Bounias I, Olsson M et al (2011) Glycaemic control and incidence of heart failure in 20,985 patients with type 1 diabetes: an observational study. Lancet 378:140 146 4. Mogensen UM, Jensen T, Køber L et al (2012) Cardiovascular autonomic neuropathy and subclinical cardiovascular disease in normoalbuminuric type 1 diabetic patients. Diabetes 61:1822 1830 5. Kim WY, Astrup AS, Stuber M et al (2007) Subclinical coronary and aortic atherosclerosis detected by magnetic resonance imaging in type 1 diabetes with and without diabetic nephropathy. Circulation 115:228 235 6. Sato A, Tarnow L, Parving HH (1999) Prevalence of left ventricular hypertrophy in Type I diabetic patients with diabetic nephropathy. Diabetologia 42:76 80 7. Sato A, Tarnow L, Parving HH (1998) Increased left ventricular mass in normotensive type 1 diabetic patients with diabetic nephropathy. Diabetes Care 21:1534 1539 8. Rubler S, Dlugash J, Yuceoglu YZ et al (1972) New type of cardiomyopathy associated with diabetic glomerulosclerosis. Am J Cardiol 30:595 602 9. Yazici D, Yavuz DG, Toprak A et al (2013) Impaired diastolic function and elevated Nt-proBNP levels in type 1 diabetic patients without overt cardiovascular disease. Acta Diabetol 50:155 161 10. Halley CM, Houghtaling PL, Khalil MK et al (2011) Mortality rate in patients with diastolic dysfunction and normal systolic function. Arch Intern Med 171:1082 1087 11. Fang ZY, Prins JB, Marwick TH (2004) Diabetic cardiomyopathy: evidence, mechanisms, and therapeutic implications. Endocr Rev 25: 543 567 12. Lang RM, Bierig M, Devereux RB et al (2005) Recommendations for chamber quantification: a report from the American Society of Echocardiography s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 18:1440 1463 13. Nagueh SF, Appleton CP, Gillebert TC et al (2009) Recommendations for the evaluation of left ventricular diastolic function by echocardiography. Eur J Echocardiogr 10:165 193 14. Unzek S, Popovic ZB, Marwick TH, Diastolic Guidelines Concordance Investigators (2011) Effect of recommendations on interobserver consistency of diastolic function evaluation. JACC Cardiovasc Imaging 4:460 467 15. Mantha S, Roizen MF, Fleisher LA et al (2000) Comparing methods of clinical measurement: reporting standards for Bland and Altman analysis. Anesth Analg 90:593 602 16. Levey AS, Bosch JP, Lewis JB et al (1999) A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 130:461 470 17. Andersen AR, Christiansen JS, Andersen JK et al (1983) Diabetic nephropathy in type 1 (insulin-dependent) diabetes: an epidemiological study. Diabetologia 25:496 501 18. Redfield MM, Jacobsen SJ, Burnett JC Jr et al (2003) Burden of systolic and diastolic ventricular dysfunction in the community: appreciating the scope of the heart failure epidemic. JAMA 289: 194 202 19. Mogelvang R, Sogaard P, Pedersen SA et al (2009) Cardiac dysfunction assessed by echocardiographic tissue Doppler imaging is an independent predictor of mortality in the general population. Circulation 119:2679 2685 20. Davies M, Hobbs F, Davis R et al (2001) Prevalence of leftventricular systolic dysfunction and heart failure in the Echocardiographic Heart of England Screening study: a population based study. Lancet 358:439 444 21. Hillis GS, Møller JE, Pellikka PA et al (2004) Noninvasive estimation of left ventricular filling pressure by E/e is a powerful predictor of survival after acute myocardial infarction. J Am Coll Cardiol 43:360 367 22. Kuznetsova T, Herbots L, López B et al (2009) Prevalence of left ventricular diastolic dysfunction in a general population. Circ Heart Fail 2:105 112 23. From AM, Scott CG, Chen HH (2010) The development of heart failure in patients with diabetes mellitus and pre-clinical diastolic dysfunction a population-based study. J Am Coll Cardiol 55:300 305 24. Von Bibra H, St John Sutton M (2010) Diastolic dysfunction in diabetes and the metabolic syndrome: promising potential for diagnosis and prognosis. Diabetologia 53:1033 1045 25. Di Bonito P, Moio N, Cavuto L et al (2005) Early detection of diabetic cardiomyopathy: usefulness of tissue Doppler imaging. Diabet Med J Br Diabet Assoc 22:1720 1725 26. Salem M, El Behery S, Adly A et al (2009) Early predictors of myocardial disease in children and adolescents with type 1 diabetes mellitus. Pediatr Diabetes 10:513 521 27. Gul K, Celebi AS, Kacmaz F et al (2009) Tissue Doppler imaging must be performed to detect early left ventricular dysfunction in patients with type 1 diabetes mellitus. Eur J Echocardiogr 10:841 846