Article Text

other Versions

Download PDFPDF
Myocardial steatosis, cardiac remodelling and fitness in insulin-sensitive and insulin-resistant obese women
  1. Wolfgang Utz1,
  2. Stefan Engeli2,
  3. Sven Haufe2,3,
  4. Petra Kast3,
  5. Mario Hermsdorf3,
  6. Susanne Wiesner3,
  7. Martin Pofahl1,
  8. Julius Traber1,
  9. Friedrich C Luft3,
  10. Michael Boschmann3,
  11. Jeanette Schulz-Menger1,
  12. Jens Jordan2
  1. 1Working Group Cardiac MR Medical Faculty of the Charité Campus Buch and HELIOS Klinikum Berlin Buch, Berlin, Germany
  2. 2Institute of Clinical Pharmacology, Hannover Medical School, Hannover, Germany
  3. 3Franz Volhard Clinical Research Center at the Experimental and Clinical Research Center, Charité and Max Delbrück Center for Molecular Medicine, Berlin, Germany
  1. Correspondence to Jens Jordan, Institute for Clinical Pharmacology, Hannover Medical School, Carl-Neuberg-Str. 1, 30625 Hannover, Germany; jordan.jens{at}


Background Obesity predisposes to heart failure and premature cardiovascular death, particularly in sedentary women. In animal models and in men with type 2 diabetes mellitus, impaired cardiac function is associated with myocardial triglyceride (MTG) accumulation. Lipotoxic injury from altered myocardial metabolism may be causative. Whether such association also exists in obese, non-diabetic women is unknown.

Objective To explore the relation between MTG content, cardiac remodelling and cardiorespiratory fitness in obese, insulin-sensitive and insulin-resistant non-diabetic women.

Design Cross-sectional investigation.

Setting Academic clinical research centre.

Patients 65 Overweight/obese and sedentary, but otherwise healthy women (body mass index 33±4 kg/m2; age 45±10 years).

Interventions None.

Main outcome measures Cardiac structure and function measured by cardiovascular magnetic resonance imaging and MTG content of the interventricular septum by 1H MR spectroscopy. Additional outcomes were cardiopulmonary fitness and insulin sensitivity during oral glucose tolerance testing.

Results Insulin resistance (composite insulin sensitivity index (C-ISI) <4.6) was present in 29 women. MTG content was higher (0.83±0.30 vs 0.61±0.23, p=0.002) and left ventricular diastolic (p<0.01), but not systolic function was reduced in women with insulin resistance compared with insulin-sensitive women. The remodelling index defined as left ventricular mass divided by end-diastolic volume was increased in women with impaired glucose tolerance (p=0.006). Furthermore, cardiopulmonary fitness was equal in both groups, but was inversely correlated with MTG (r=−0.28, p=0.02).

Conclusions In overweight and obese women, insulin resistance is associated with increased MTG content, cardiac remodelling and reduced diastolic function.

Clinical Trial Registration NCT00956566.

  • Cardiac remodelling
  • obesity
  • diastolic dysfunction
  • MRI
  • gender

Statistics from

Request Permissions

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.


Obesity increases the risk of left ventricular hypertrophy, diastolic dysfunction, atrial fibrillation,1 and is associated with heart failure and premature cardiovascular death in women.2–4 Intravascular volume expansion and increased cardiac output, arterial hypertension, insulin resistance and neurohumoral activation, all probably contribute to obesity-associated changes in cardiac structure and function.5 Altered cardiac metabolism and diminished efficiency are already evident in otherwise healthy obese subjects, with obese women being mainly affected.6 Previous studies demonstrated myocardial lipid accumulation in animal models of obesity and in non-ischaemic failing human hearts, particularly in obese diabetic subjects.7 Lipotoxic intermediates emerging from harmful fatty acid pathways after myocardial lipid overstorage contribute to the deterioration of cardiac function.8 1H magnetic resonance spectroscopy has been introduced as a method to quantify myocardial triglyceride (MTG) content in human subjects.9 10 In subsequent studies, MTG was found to be excessive in patients with type 2 diabetes or impaired glucose tolerance versus healthy controls and was associated with diastolic dysfunction.11 12 Furthermore, impaired longitudinal contractility, an early feature in diabetic heart disease, was present in diabetic men with excessive MTG.13 However, most of the evidence of the effects of MTG on cardiac function is provided from studies on diabetic men and differences between normal and impaired glucose tolerance in obese subjects were not significant. Results cannot simply be extrapolated to women because of gender differences in cardiovascular, neurohumoral and cardiac metabolic responses to adiposity.14 15 The same applies to the beneficial effects of regular exercise training on MTG, cardiac performance and physical fitness seen in men.16 Thus, the purpose of this study was to explore the relationship between MTG, cardiac remodelling and cardiopulmonary fitness and the impact of insulin resistance in otherwise healthy obese women.

Subjects and methods


We recruited 65 overweight/obese otherwise healthy women through advertisements in local newspapers. Women aged 18–65 years, with body mass index (BMI) >27 kg/m2, no regular medication except hormonal contraceptives (a total of five women) or hormone replacement therapy and who claimed to have <2 h physical activity a week were eligible. Major exclusion criteria were diabetes mellitus, vascular or cardiac disease, pregnancy, lactation period, presence of metallic implants, or claustrophobia. Our institutional review board approved the study and written informed consent was obtained from subjects before entry.


Subjects visited the laboratory on two separate occasions for anthropometric, metabolic and cardiovascular evaluations and exercise testing. After an overnight fast, we determined body weight, body height, waist circumference and hip circumference in a standardised fashion. We then obtained fasting venous blood samples and this was followed by an oral glucose tolerance test (75 g glucose/500 ml). After another overnight fast, subjects underwent MR studies in the early morning. Then, we assessed physical fitness by spiroergometry testing. In addition, subjects underwent 24 h blood pressure monitoring (Spacelab 90207, Spacelabs Healthcare, Feucht, Germany).

Cardiac MRI and MR spectroscopy

For all measurements, a clinical 1.5 T MR scanner (Sonata and Avanto, Siemens Medical Solutions AG, Erlangen, Germany) was used. MR imaging and spectroscopy parameters were identical in both scanners. We determined cardiac morphology and function and quantified MTG content. After initial anatomical scout images had been obtained, we performed high temporal resolution cine imaging with a retrospective gated, balanced steady-state free precession sequence (TR 16.3 ms, TE 1.15 ms, 64 phases, matrix 208×256, FOV 325×400 mm2, in plane resolution 1.6×1.6 mm2). We acquired a stack of contiguous short-axis slices (slice thickness 7 mm, interslice gap 3 mm) during repetitive breath-holds in end-expiration. We quantified left ventricular structure and ejection fraction by manually drawing endocardial and epicardial contours in end-diastole and end-systole using dedicated software (MASS7.1, Medis AG, Leiden, Netherlands). Left ventricular mass and end-diastolic volume were indexed for height and the left ventricular remodelling index was calculated as left ventricular mass divided by end-diastolic volume.17 18 For assessment of left ventricular diastolic function, we acquired high-temporal resolution cine images in horizontal and vertical long axes. We followed endocardial contours in all diastolic phases and calculated left ventricular volumes using a biplane model.19 Peak filling rates in the early (PFRE) and in the atrial (PFRA) filling phase were derived from resulting left ventricular volume–time curves using Origin 8.0 (OriginLab Corporation, Northampton, USA).20 Furthermore, we tracked left ventricular length (ie, distance between apex and mitral annular plane) over diastolic phases in the horizontal long-axis cine image. Peak lengthening velocity (PLV) in the early filling phase was determined from the length–time curve in analogy to PFRE. The inter- and intraobserver variabilities, determined in 10 subjects, were 10.1% (CI −37.1% to 16.9%) and 7.4% (CI −5.6% to 19.4%) for maximum diastolic filling rates (PFR) and −1.3% (CI −25.5% to 23.0%) and −0.2% (CI −12.7% to 12.3%) for peak mitral annular velocity (PLV).

To assess longitudinal contractility—a marker of subclinical early dysfunction in diabetic heart disease—systolic fractional longitudinal shortening was calculated as difference between left ventricular end-diastolic and end-systolic length normalised to end-diastolic length.21 To quantify MTG content, a 6–8×20×25 mm3 voxel was positioned in the interventricular septum (figure 1). We applied a cardiac and respiratory gated 1H single voxel spectroscopy sequence (spin-echo: TR according to respiratory cycle (>5 s), TE 30 ms) to acquire spectra at end-systole and in end-expiration.22 Lipid signals were taken from water-suppressed spectra (96 averages) and water signals from unsuppressed spectra (four averages). Areas under water and lipid peaks were quantified using standard line-fitting procedures (Siemens Syngo Spectroscopy) and MTG content was expressed as fat-to-water ratio (%). The percentage coefficient of variation of line fitting was −0.04% (CI −0.39% to 0.32%) for the unsuppressed water signal and 0.08% (CI −1.12% to 1.29%) for myocardial lipids. The interstudy reproducibility of MTG quantification was 0.01% (CI −0.05% to 0.06%).

Figure 1

Left side: Voxel position (white square) in the myocardial septum used for 1H MR lipid spectroscopy. Right side: Respective MR spectrum with water suppression. Arrow points at MR signal from triglyceride methylene group.

Incremental exercise test

Subjects underwent stepwise incremental exercise testing on a bicycle ergometer (VIAsprint 150P, Ergoline, Bitz, Germany) until volitional exhaustion. Exercise was performed in a temperature-controlled room (21–22°C) approximately 2 h after subjects had ingested a standardised breakfast (containing ∼520 kcal: 24% fat, 68% carbohydrate, 8% protein). Alcohol and caffeine were not permitted for 48 h before the exercise test. After 3 min in the seated position, resting measurements were recorded. Exercise was then started at a workload of 25 W. Workload was increased every 2 min by 25 W until the subjects could not maintain the requested 60 rpm pedal frequency. Using an open spirometric system (Vmax Spectra Model 229D analyzer, SensorMedics, Yorba Linda, USA), the time course of oxygen uptake and carbon dioxide production was recorded breath-by-breath and averaged in 10 s intervals. Heart rate was recorded by an electrocardiogram (GE Medical Systems Inc, Milwaukee, Wisconsin, USA) throughout the exercise test. We assumed that subjects had reached maximal oxygen uptake (VO2max) when at least two of the following criteria were met: (a) respiratory exchange ratio >1.10; (b) increase in oxygen uptake during the last 60 s of the test of <100 ml/min and (c) heart rate within 10 beats/min of the predicted maximum heart rate.23 To consider the individual differences in body weight, oxygen uptake was expressed as kilograms of body weight (VO2: ml/min/kg).

Biochemical and metabolic analyses

We estimated insulin sensitivity by oral glucose tolerance testing. During an oral glucose load (75 g glucose/500 ml), we obtained blood samples at baseline and 15, 30, 45, 60, 90 and 120 min after glucose ingestion to measure glucose and insulin. Insulin sensitivity was calculated by the composite insulin-sensitivity index (C-ISI) as described before.24 A cut-off value of 4.6 was used to identify insulin-resistant (IR) and insulin-sensitive (IS) subjects.25 A homoeostasis model assessment index (HOMA-IR) was calculated to estimate insulin resistance from fasting insulin and glucose concentrations (insulin (μU/ml) × glucose (mmol/l)/22.5).26 Glucose, insulin and blood lipids were measured by standard laboratory procedures in a certified clinical chemistry laboratory.


We first tested data for normal distribution and variance homogeneity. A Student t test was used to compare parameters between IS and IR subjects. We applied univariate regression analysis to detect relations of cardiopulmonary fitness and MTG with other parameters. Stepwise backward multivariate regression analysis was used to identify independent predictors of MTG. All statistical analyses were performed with SPSS version 18 (SPSS, Inc). Significance was accepted at p<0.05. Values are given as mean ± SD.


Subject characteristics

Anthropometric and metabolic measurements are given in table 1. All data were normally distributed. Six women were hyperlipidaemic with triglycerides >2.28 mmol/l and/or cholesterol >6.2 mmol/l.27 Twenty-nine women had resistance to insulin and 36 sensitivity to insulin during oral glucose testing. IR and IS women differed in body mass index and circulating triglyceride levels, but not in ambulatory blood pressure.

Table 1

Participant characteristics and metabolic parameters

Cardiac structure, function and fat

Table 2 gives an overview of cardiac structure and function measurements as well as MTG content in all women, women with IS or IR. HOMA-IR was positively associated with MTG (r=0.27, p=0.03). Insulin sensitivity (C-ISI) and BMI tended to correlate with MTG in univariate regression analysis (r=0.22, p=0.08; r=0.23, p=0.07, respectively). MTG content was higher in women with IR than in those with IS (0.83±0.30 vs 0.61±0.23, p=0.002). On multivariate regression analysis (including VO2max, BMI, age, serum triglycerides and insulin resistance (HOMA-IR)) insulin resistance together with VO2max remained as independent predictors of MTG, but not BMI (table 3). Left ventricular mass and volume was similar in both groups. However, the left ventricular remodelling index was increased in women with IR (0.59±0.08 vs 0.54±0.08, p=0.006) without correlation to BMI (p=0.4). Furthermore, all diastolic function indices were impaired in the IR group, including parameters of left ventricular filling (PFRE, PFRE/EDV, PFRE/PFRA) and mitral annular velocity (PLV). Correlation of diastolic function (PFRE) with MTG was of borderline statistical significance (r=−0.24, p=0.052). In contrast, systolic function characterised by ejection fraction or longitudinal contractility did not differ significantly between groups (p>0.4).

Table 2

MR study parameters

Table 3

Multivariate backward regression analyses with myocardial triglycerides as dependent variable

Cardiorespiratory fitness and cardiac fat

Cardiorespiratory fitness expressed as VO2max was not significantly different between groups (p=0.4). However, cardiorespiratory fitness inversely correlated with MTG (r=−0.28, p=0.02) in univariate regression analysis (figure 2). The slope of the relationship was −4.0 ml/min/kg (CI −10.0 to −0.1) in women with IS and −4.4 ml/min/kg (CI −9.3 to 0.6) in women with IR. On multivariate regression analysis, BMI and MTG were the only determinants of VO2max (β=−0.28 for BMI; β=−0.21 for MTG). Thus, only a minor part of cardiorespiratory fitness variability in obese women is explained by MTG.

Figure 2

Myocardial triglyceride (MTG) correlates with cardiorespiratory fitness in bivariate correlation analysis (r=−0.28, p=0.02). Filled circles/solid line indicate insulin-resistant subjects, empty circles/dotted line indicate insulin-sensitive subjects.


The main finding of our study is that IR obese non-diabetic women show significantly increased MTG levels compared with obese IS women. BMI was not independently associated with MTG in multivariate analyses. The MTG increase is paralleled by a significant deterioration of CMR-derived left ventricular diastolic function, whereas systolic left ventricular function parameters are well maintained. Furthermore structural LV adaptation was more pronounced in IR obese women, reflected by an increased LV remodelling index. Finally, reduced cardiopulmonary fitness was weakly but significantly correlated with MTG. Thus, our study identified two risk factors for excessive MTG in women—namely, insulin resistance and reduced cardiopulmonary fitness, which are amenable to lifestyle interventions.

Animal studies support the idea that MTG deposition may adversely affect cardiac function. Cardiac steatosis induced by cardiomyocyte-specific overexpression of the acyl CoA synthetase transgene on an α-myosin heavy chain gene promoter in non-obese mice induced dilated lipotoxic cardiomyopathy and premature death.28 Cardiac-specific overexpression of fatty acid transport protein 1 led to increased myocardial free fatty acid uptake, deposition and metabolism. Three-month-old mice showed impaired LV filling and biatrial enlargement with preserved systolic function.29 In part preventable or reversible non-oxidative fatty acid metabolites including ceramides have been identified as key mediators of lipotoxic myocardial disease. Apparently, cytotoxic intermediates of these metabolites elicit mitochondrial dysfunction, cell damage, oxidative stress and lipoapoptosis.8 In the non-ischaemic failing heart, excessive MTG storage in obese/diabetic patients was associated with gene expression changes typical for cardiac dysfunction.7

Our results in women extend earlier findings on interactions between insulin sensitivity, MTG content and cardiac function, which were mainly based on findings in diabetic subjects and men. In IR states, MTG content was increased compared with that of lean healthy control subjects and predictive of impaired diastolic function.11 12 Increased MTG levels in diabetic men were associated with impaired biventricular longitudinal contractility.13 However, these studies only included men to avoid gender and sex hormone level dependent influences on lipid metabolism and MTG storage13 or did not perform a gender-specific analysis.12 In our study, insulin resistance was associated with increased MTG content and reduced diastolic function in obese women. The observation is in contrast to earlier findings in a gender-mixed cohort, where MTG and cardiac diastolic function were similar in mildly obese subjects with normal or impaired glucose tolerance.12 Diastolic peak filling rates (from cine or phase contrast imaging) have been used in previous CMR studies to correlate diastolic function and MTG. In addition, early diastolic mitral annular velocity (PLV), was impaired in the IR group in our study. Similarly, earlier studies showed a gradual decrease in related tissue Doppler echo indices with increasing BMI and concomitant glucose intolerance.30 However, with univariate regression analysis, the association between diastolic function and MTG in our study was weaker than in patients with type 2 diabetes mellitus.11 The discrepancy may be explained by the smaller range of MTG values in our study, the less pronounced insulin resistance and the impact of sex hormone status on cardiac lipid metabolism and diastolic function.

Reduced cardiorespiratory fitness predisposes to cardiovascular disease and predicts premature death.31 32 We observed a negative correlation between cardiorespiratory fitness and MTG in untrained overweight/obese women. However, the correlation was relatively weak and MTG content explained no more than 10% of the variability in cardiorespiratory fitness. In our data, MTG and VO2max were mutually interrelated. Thus, the question as to whether increased MTG is a cause or consequence of reduced cardiopulmonary fitness cannot be resolved.

In small study samples both metabolic interventions and regular exercise induced significant changes in MTG content with concomitant impact on cardiac function. In obese patients with type 2 diabetes mellitus long-term caloric restriction lowered MTG content and improved diastolic function, whereas in lean healthy men short-term fasting increased MTG and worsened diastolic function.33 34 Furthermore, regular exercise training without weight loss in obese non-diabetic men lowered MTG while improving cardiac function.16 Our study suggests that improvement in physical fitness may have a similar beneficial effect on MTG in women. However, the role of MTG as a surrogate marker for interventions aiming at improvement of cardiac metabolism needs to be established in future studies.


A potential limitation of our study was that we did not include lean women as control group. Thus we cannot clarify whether MTG values in the IS group differ from MTG values of lean women. An earlier study failed to detect significant differences in mildly obese subjects of both sexes with normal glucose tolerance compared with lean normal controls.12 However, MTG in our women with IS were similar to those reported for healthy overweight men.11 A second limitation of our study is that we studied women aged 23–63 years. For practical reasons, it was impossible to study all premenopausal women during the same time of the menstrual cycle. The sex hormone status was not assessed and the confounding influence of sex hormones on cardiac lipid metabolism was not controlled or investigated in detail.


Our study is the first to specifically investigate the relation between cardiac remodelling, MTGs and insulin sensitivity in obese women. The combination of obesity with insulin resistance and poor cardiopulmonary fitness predisposes to MTG accumulation and diastolic dysfunction. However, a large part of the variability in MTG cannot be explained by traditional cardiovascular and metabolic risk factors. Given the importance of MTG accumulation in the pathogenesis of experimental heart failure models, the mechanisms regulating MTG in patients deserve to be studied in more detail.



  • Funding The study was supported by the Federal Ministry of Education and Research (BMBF-0313868). The work was also supported in part by the Commission of the European Communities (Collaborative Project ADAPT, Contract No. HEALTH-F2-2008-201100) and the German Obesity Network of Competence (Collaborative Project ADIPOSETARGET, 01 Gl0830). The study was part of a joint project between metanomics GmbH (Berlin, Germany), Charité - University Medical School and Hannover Medical School.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the institutional review board.

  • Provenance and peer review Not commissioned; externally peer reviewed.