PAPER. KH Liu 1 *, YL Chan 1, WB Chan 2, WL Kong 1, MO Kong 1 and JCN Chan 2. Introduction

Transcription

1 (2003) 27, & 2003 Nature Publishing Group All rights reserved /03 $ PAPER Sonographic measurement of mesenteric fat thickness is a good correlate with cardiovascular risk factors: comparison with subcutaneous and preperitoneal fat thickness, magnetic resonance imaging and anthropometric indexes KH Liu 1 *, YL Chan 1, WB Chan 2, WL Kong 1, MO Kong 1 and JCN Chan 2 1 Department of Diagnostic Radiology and Organ Imaging, The Chinese University of Hong Kong, The Prince of Wales Hospital, Shatin, NT, Hong Kong SAR; and 2 Department of Medicine and Therapeutics, The Chinese University of Hong Kong, The Prince of Wales Hospital, Shatin, NT, Hong Kong SAR OBJECTIVE: Visceral fat, notably mesenteric fat, which is drained by the portal circulation, plays a critical role in the pathogenesis of metabolic syndrome through increased production of free fatty acids, cytokines and vasoactive peptides. We hypothesize that mesenteric fat thickness as measured by ultrasound scan could explain most of the obesity-related health risk. We explored the relationships between cardiovascular risk factors and abdominal fat as determined by sonographic measurements of thickness of mesenteric, preperitoneal and subcutaneous fat deposits, total abdominal and visceral fat measurement by magnetic resonance imaging (MRI) and anthropometric indexes. DESIGN: A cross-sectional study. SUBJECTS: Subjects included 18 healthy men and 19 women (age: y, BMI: kg/m 2 ). MEASUREMENTS: The maximum thickness of mesenteric, preperitoneal and subcutaneous fat was measured by abdominal ultrasound examination. MRI examinations of whole abdomen and pelvis were performed and the amount of total abdominal and visceral fat was quantified. The body mass index, waist circumference and waist hip ratio were recorded. Cardiovascular risk factors were assessed by physical examination and blood taking. RESULTS: Men had more adverse cardiovascular risk profile, higher visceral fat volume and thicker mesenteric fat deposits than women. Among all the investigated obesity indexes, the mesenteric fat thickness showed the highest correlations with total cholesterol, LDL-C, triglycerides, fasting plasma glucose, HbA 1c and systolic blood pressure in men, and with triglycerides and HbA 1c in women. On stepwise multiple regression analysis with different obesity indexes as independent variables, 30 65% of the variances of triglycerides, total cholesterol, LDL-C and HbA 1c in men, and triglycerides in women were explained by the mesenteric fat thickness. CONCLUSION: Compared with sonographic measurement of subcutaneous and preperitoneal fat thickness, MRI measurement of total abdominal and visceral fat and anthropometric indexes, sonographic measurement of mesenteric fat thickness showed better associations with some of the cardiovascular risk factors. It may potentially be a useful tool to evaluate regional distribution of obesity in the assessment of cardiovascular risk. (2003) 27, doi: /sj.ijo Keywords: mesenteric fat thickness; subcutaneous fat thickness; preperitoneal fat thickness; cardiovascular risk factors; magnetic resonance imaging; ultrasound *Correspondence: KH Liu. tongyc@netvigator.com Received 3 January 2003; revised 20 April 2003; accepted 9 May 2003 Introduction There are now clinical and experimental evidence showing that visceral fat plays a critical role in the pathogenesis of the metabolic syndrome. Compared to subcutaneous fat and other deposits of abdominal fat, visceral fat, notably

2 1268 mesenteric fat, is more sensitive to the lipolytic effects of catecholamines to release more free fatty acid (FFA) which can induce insulin resistance. 1 Besides, visceral adipocytes secrete many cytokines and vasoactive peptides which have direct effects on the vasculature to increase cardiovascular risk. 2 In support of these hypothesis, epidemiological studies have also demonstrated the associations of anthropometric indexes such as body mass index (BMI), waist circumference (WC) and waist hip ratio (WHR) with cardiovascular risk factors and diseases. 3 6 Due to the more powerful predictive role of central over general obesity on cardiovascular diseases, magnetic resonance imaging (MRI), which does not involve ionizing radiation, is increasingly being used to quantify intra-abdominal fat and to differentiate visceral from subcutaneous fat. 7,8 However, MRI examination is expensive and not easily available in most clinical settings. Ultrasound scan is a noninvasive method which can be used to image body fat without radiation exposure. To date, it has only been used to measure subcutaneous fat 9 11 and preperitoneal fat thickness. 12,13 However, these fat deposits are not usually regarded as visceral fat drained by the portal system. 13 By contrast, mesenteric fat can be easily recognized by ultrasound scan, especially in obese subjects with thickened mesenteries. 14 This study aimed to investigate the relationships between mesenteric fat thickness and cardiovascular risk factors. We also compared the relationships between the latter and various abdominal fat deposits measured by ultrasound scan, intra-abdominal fat by MRI and anthropometric indexes. Figure 1 Ultrasonogram of mesenteric leaves. The peritoneal surfaces are indicated by arrows and the maximum mesenteric fat thickness is measured with calipers. Subjects and methods In all, 37 healthy Chinese subjects were recruited (18 men and 19 women with a mean BMI of 24.9 kg/m 2, range: kg/m 2 ). All subjects had no known medical history and none was receiving medications. The study was approved by the Clinical Research Ethics Committee of the Chinese University of Hong Kong and all subjects gave informed written consent. Ultrasound examination All subjects underwent ultrasound examination of abdomen. An ATL HDI 5000 (CA, Bothell) with CL 4-7 MHz or CL 2-5 MHz curvilinear transducer was used for the study. A complete survey examination, with emphasis in the para-umbilical area, was performed in each subject with special attention to identify the mesenteric leaves (Figure 1). The mesenteric leaves appeared to be elongated structures with highly reflecting peritoneal surfaces. It had high-level echoes in the periphery and small vascular structures (1 2 mm in diameter) could be seen within it. 14 The mesenteric leaves were divided from each other by specular echoes corresponding to their peritoneal surfaces. The Figure 2 Ultrasonogram of preperitoneal fat. The thickness of preperitoneal fat is indicated by the white line, and the arrow indicates the subcutaneous layer. mesenteries demonstrated no peristalsis, in contrast to small bowel loops attached to the distal end. 14 Not all the mesenteries could be visualized on the scan as some of them might be obscured by bowel gas. When different mesenteric leaves were visualized, the maximum thickness was measured. Usually 6 10 measurements were made on each ultrasound examination, and the mean of the three thickest mesenteric leaves was used for the analysis. Preperitoneal fat thickness was measured with modified criteria of Suzuki et al. 13 It was scanned longitudinally with L5-12 MHz linear transducer placed perpendicular to the skin surface and scanned along the midline of the abdomen, between xiphoid process and umbilicus. The maximum thickness of preperitoneal fat (Figure 2) was measured three times and the mean value was taken for analysis. The

3 Figure 3 Ultrasonogram of abdominal subcutaneous fat. The thickness of subcutaneous fat is indicated by the white line. preperitoneal fat attached to the upper and lower edges of the liver was excluded in the previous study by Suzuki et al, 13 but was taken into account in this study because this would make the measurement more compatible with criteria of maximum thickness and more reproducible. The abdominal subcutaneous fat thickness was measured with the L5-12 MHz linear transducer transversely placed perpendicular to the skin in the midline of abdomen, between the xiphoid process and umbilicus. The maximum thickness of subcutaneous fat (Figure 3) was measured three times and the mean value was taken. All ultrasound measurements of preperitoneal and subcutaneous fat thickness were made at the expiratory phase of quiet respiration. Application of the transducer on the body surface was done without undue pressure that would alter the body layer contour and thickness. A typical ultrasound examination would take about 5 min. Reliability of sonographic measurements Of the 37 subjects, 17 were examined by two experienced sonographers who did not have knowledge of the other operator s scanning results. The measurement of mesenteric, preperitoneal and subcutaneous fat thickness was made by each sonographer with the above-mentioned procedure. The interoperator and intraoperator reliability was assessed. MRI examination All subjects underwent MRI examination for the quantification of intra-abdominal fat. MRI was performed with the subjects supine in a 1.5 T imager (Gyroscan ACS-NT, Philips Medical System, Best, The Netherlands) with a body coil, T1-weighted Turbo spin-echo sequence, 15 repetition timef250 ms, echo timef8 ms, flip angle 901, field of viewf45 cm. Breath-hold consecutive images with slice thickness 15 mm were obtained from the diaphragm to pubic symphysis. All subjects were asked to fast for 2 h before the examination to decrease bowel peristalsis. From each cross-sectional MRI image, the area of fat was quantified by measuring the pixels above cutoff points established as minima between peaks of fat and soft tissues on signal intensity histograms. 15 Fat within bone marrow and paraspinal muscles were excluded from the measurement. Subcutaneous fat was defined as fat superficial to the abdominal and back muscles. Visceral fat was defined as the intra-abdominal fat bound by parietal peritoneum or transversalis fascia. Total intra-abdominal fat represented the summation of both subcutaneous and visceral fat. 15 The volume of total visceral fat in cm 3 was calculated by: (summation of area of visceral fat in all slices in cm 2 ) 1.5 cm. 16 The volume of total intra-abdominal fat in cm 3 was calculated by: (summation of area of total intraabdominal fat in cm 2 in all slices) 1.5 cm. 16 The total number of slices included for quantification ranged from 21 to 28, depending on the body build of the subjects. Clinical assessments All subjects, wearing light clothing and no shoes, were weighed and their standing height was measured. BMI was calculated by body weight in kilogram divided by (body height in metre 2 ). WC was measured at the narrowest circumference between xiphisternum and umbilicus. Hip circumference (HC) was measured at the maximum circumference at the level of femoral trochanter. 17 WHR was calculated by WC divided by HC. After sitting for at least 10 min, blood pressure was measured using a standard mercury sphygmomanometer. The Korotkoff sound V was taken as the diastolic blood pressure. 18 All subjects had at least 12 h of fasting before blood taking for measurement of serum total cholesterol, high-density lipoprotein cholesterol (HDL-C), triglycerides, low-density lipoprotein cholesterol (LDL-C), plasma glucose, insulin and glycated haemoglobin (HbA 1c ). Insulin resistance was estimated using the homeostasis model assessment (HOMA) equation, 15 which simplifies to fasting plasma insulin glucose product divided by Statistical analysis Data are expressed as means7s.d. and nonparametric test was used for comparing all the mean values. A two-tailed P-value o0.05 was considered significant. Intraclass correlation was used to assess the reliability of sonographic measurements. Male and female subjects were separately analysed. Fasting plasma glucose, HbA 1c, systolic blood pressure in men were logarithmically transformed due to skewed distribution. Partial correlation analysis controlling 1269

4 1270 for age was used to examine the relationships among different variables in the entire sample. The stepwise multiple regression analysis was used to examine the variance in cardiovascular risk factors explained independently by each of the obesity measurement indexes. In each analysis, the three sonographic measurements of fat deposit thickness, three anthropometric indexes and two MRI-fat volume measurements were used as independent variables and one of the cardiovascular risk factors was used as the dependent variable. Statistical analysis was performed with the Statistical Package for the Social Sciences (SPSS) for windows, version 9.0. Results There were no significant differences in age and BMI between men and women, but men had higher WC and WHR than women. Men also had more adverse cardiovascular risk profile as shown by higher BP, HbA 1C, plasma insulin, HOMA, triglycerides and LDL-C, but lower HDL-C than women. The mean mesenteric fat was thicker in men than women, but there were no differences in the subcutaneous and preperitoneal fat thickness between both the groups. There was no difference in total intra-abdominal fat between men and women, but men had higher visceral fat volume than women (Table 1). The sonographic measurements of different fat deposits showed good interoperator and intraoperator reliability, with intraclass correlation coefficients of mesenteric fat thickness ranging from 0.89 to 0.97 (Table 2). The correlation coefficients between mesenteric fat thickness and all cardiovascular risk factors were higher in men than women. The mesenteric fat thickness also showed higher correlations with most of these risk factors than preperitoneal and subcutaneous fat thickness (Table 3). The total visceral fat volume determined by MRI showed higher correlation with most cardiovascular risk factors than the total abdominal fat volume (Table 3). Among three anthropometric indexes, WC was generally a better correlate with cardiovascular risk factors in men, while BMI was a better correlate in women. In men, among all the investigated obesity indexes (including ultrasound, MRI and anthropometric measurements), the mesenteric fat thickness showed the highest correlations with most of the lipid and glycaemic indexes, BMI showed the highest correlations with HDL-C, plasma insulin and diastolic blood pressure, while WC correlated best with HOMA value. In women, the mesenteric fat thickness showed highest correlations with triglycerides and HbA 1C, while the total visceral fat determined by MRI showed the best correlation with other cardiovascular risk factors. As in men, BMI showed the highest correlation with plasma insulin concentration in women. Using stepwise multiple regression analysis with different obesity indexes as independent variables (Table 4), mesenteric fat thickness was the major explanatory variable for increased triglycerides, total cholesterol, LDL-C and HbA 1c in men, and increased triglycerides in women. Subcutaneous fat thickness was the major explanatory variable for reduced Table 1 Clinical and biochemical characteristics and measurement of various fat deposits of the study subjects Subjects characteristics Male Female P-value N F Age (y) BMI (kg/m 2 ) WC (cm) WHR Systolic blood pressure (mmhg) Diastolic blood pressure (mmhg) HbA 1C (%) Fasting plasma glucose (mmol/l) Plasma insulin concentration (mu/ml) HOMA (mu/ml 1 mmol) Total cholesterol (mmol/l) HDL-cholesterol (mmol/l) Triglycerides (mmol/l) LDL-cholesterol (mmol/l) MRI visceral fat volume (cm 3 ) MRI total intra-abdominal fat (cm 3 ) U/S Mesenteric fat thickness (cm) U/S Preperitoneal fat thickness (cm) U/S Subcutaneous fat thickness (cm) Mean7s.d. U/S: ultrasound scan, MRI: magnetic resonance imaging, HDL-C: high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, HbA 1c : glycated haemoglobin.

5 Table 2 deposits Reliability Reliability indexes of sonographic measurement of different fat Mesenteric fat thickness Preperitoneal fat thickness Subcutaneous fat thickness Interoperator ICC % CI s.e.m cm 0.21 cm 0.23 Intraoperator ICC % CI s.e.m cm 0.1 cm 0.1 cm ICC¼ Intraclass correlation coefficients; CI ¼ confidence intervals; s.e.m. ¼ standard error of measurement. HDL-C in men. Most of the variance in other cardiovascular risk factors was explained by central adiposity as reflected by total abdominal and visceral fat determined by MRI, WC and WHR. However, the variances of plasma insulin concentration in men and women and HbA 1c in women were explained by BMI reflecting general obesity. Discussion In agreement with other workers, 16,19,20 we have confirmed that men had more visceral fat, worse lipid and glycaemic control and higher blood pressure than women. Due to this gender difference, the results were analysed in men and women separately. This is the first study that examined the usefulness of sonographic measurement of mesenteric fat thickness in correlating with cardiovascular risk factors. High resolution real-time ultrasound allows good delineation of normal abdominal anatomy including the recognition of normal mesenteric leaves, especially when they are thick and contain fat. 14 In two separate studies that examined the sonographic appearances of small bowel mesentery in 30 healthy subjects, the thickness of the mesentery ranged from 7 to 12 mm 14 and from 5 to 14 mm, 21 irrespective of age or body habitus of the subjects. The mean mesenteric thickness in our subjects was 8.6 mm, ranging from 3.4 to 18 mm. We also showed that the measurement of mesenteric fat thickness had good intraoperator and interoperator reliability. Ultrasound is the only imaging method that can visualize individual mesenteric leaves. By contrast, MRI can only demonstrate the whole bulk of mesenteries, but cannot reliably differentiate omental, pre- and retroperitoneal from mesenteric adipose tissues. There is now growing evidence showing that abnormal fat metabolism may be the culprit of insulin resistance syndrome and type II diabetes. Visceral fat, notably mesenteric fat, drained by the portal vein, is metabolically more active than subcutaneous fat. Visceral adipocytes are more sensitive to the lipolytic effects of catecholamines and are more resistant to the antilipolytic effects of insulin, leading to increased FFA production. The latter may lead to reduced fat oxidation and ectopic fat deposition in the liver and muscle which worsens insulin resistance by reducing peripheral glucose uptake. 1,22 Besides, visceral adipocytes are known to produce a large number of cytokines and vasoactive peptides including interleukin-6, tumour necrosis factor a, angiotensin II, plasminogen activator inhibitor-i, etc all of which can increase cardiovascular risks. 2 In support of these hypotheses, we demonstrated the strong relationships between mesenteric fat and multiple cardiovascular risk factors compared to preperitoneal and subcutaneous fat deposits. Of note, mesenteric fat thickness in men showed better correlations with systolic blood pressure, most lipid and glycaemic indexes than anthropometric and MRI measurements. In women, mesenteric fat measurement also showed 1271 Table 3 Age adjusted correlation coefficients of various obesity indexes with cardiovascular risk factors Mesenteric fat thickness (U/S) Preperitoneal fat thickness (U/S) Subcutaneous fat thickness (U/S) Total abdominal fat (MRI) Total visceral fat (MRI) BMI (kg/m 2 ) WC WHR M F M F M F M F M F M F M F M F Cholesterol HDL-C LDL-C Triglycerides HOMA FPG HbA 1c Insulin conc Systolic BP Diastolic BP M: male, F: female, U/S: ultrasonography, MRI: magnetic resonance imaging, HDL-C: high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, FPG: fasting plasma glucose; HbA 1c : glycated haemoglobin, BP: blood pressure.

6 1272 Table 4 Variances of various cardiovascular risk factors explained by different measurements of body fat using stepwise multiple regression analysis Men Women Dependent variables Predictor variables R 2 Predictor variables R 2 Total cholesterol Mesenteric fat thickness (U/S) a 0.64 Total visceral fat (MRI) 0.39 Total abdominal fat (MRI) b 0.78 HDL-C Subcutaneous fat thickness (U/S) 0.24 F F LDL-C Mesenteric fat thickness (U/S) a 0.66 Total visceral fat (MRI) 0.48 Total abdominal fat (MRI) b 0.81 Triglycerides Mesenteric fat thickness (U/S) 0.55 Mesenteric fat thickness (U/S) 0.32 HOMA WC 0.70 Total abdominal fat (MRI) 0.56 FPG WHR 0.37 F F HBA 1c Mesenteric fat thickness (U/S) 0.57 BMI 0.51 Insulin conc BMI 0.46 BMI 0.49 Systolic BP WC 0.38 Total visceral fat (MRI) 0.23 Diastolic BP Total visceral fat (MRI) 0.63 Total visceral fat (MRI) 0.26 R 2 : total variances explained by the variable(s). a The first predictor variable that entered the regression analysis. b The second predictor variable that entered the regression analysis. U/S: ultrasonography, MRI: magnetic resonance imaging, HDL-C: high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, FPG: fasting plasma glucose; HbA 1c : glycated haemoglobin, BP: blood pressure. the best correlations with triglycerides and HbA 1C. On stepwise multiple regression analysis, mesenteric fat thickness explained 30 65% of the variances of triglycerides, total cholesterol, LDL-C and HbA 1c in men, and triglycerides in women. These strong relationships are likely to be due to unique capacity of ultrasound scan to measure portal adipose tissue deposit, while visceral fat determined by MRI included both portal (mesenteric) and nonportal (preperitoneal and retroperitoneal) adipose tissues. Not unexpectedly, BMI, WC and WHR which include both abdominal nonportal and subcutaneous adipose tissues had relatively weaker associations with these cardiovascular risk factors. Given the relatively cheap, noninvasive, technically less demanding nature of ultrasound scan together with its good reproducibility, measurement of mesenteric fat thickness may potentially become a useful imaging tool in obesity research. Visceral fat usually shows greater responses than subcutaneous fat to interventional therapy such as changes in caloric intake or physical exercise. 23 Hence, serial measurement of mesenteric fat thickness may also be used to evaluate the efficacy of various treatment programmes. Nonetheless, findings from this pilot study will need to be confirmed in prospective studies involving larger population with different demographic features and cardiovascular risk profiles. In conclusion, we have demonstrated the gender differences in cardiovascular risk profiles and fat deposits. We also confirmed the reliability of using ultrasound scan to measure mesenteric fat thickness. In men and to a lesser extent in women, mesenteric fat thickness was the main explanatory variable for most of the cardiovascular risk factors compared to the thickness of other fat deposits, anthropometric indexes and intraabdominal fat determined by MRI. Acknowledgements Special thanks are extended to Dr Yang WT for her expert advice in the initial design of this research study. We also thank Ms Delanda Wong, Yee Mui Lee and Cherry Chiu for their assistance in blood taking. References 1 Bjorntorp P. Portal adipose tissue as a generator of risk factors for cardiovascular disease and diabetes. Arteriosclerosis 1990; 10: Ahima RS, Jeffery SF. Adipose tissue as an endocrine organ. Trends Endocrinol Metab 2000; 11: Kannel WB, Cupples LA, Ramaswami R, Stokes J, Kreger BE, Higgins M. Regional obesity and risk of cardiovascular disease; the Framingham Study. J Clin Epidemiol 1991; 44: Bjorntorp P. The association between obesity, adipose tissue distribution and disease. Acta Med Scand Suppl 1988; 723: Masuda T, Imai K, Komiya S. Relationship of anthropometric indices of body fat to cardiovascular risk in Japanese women. Ann Physiol Anthropol 1993; 12: Woo J, Ho Sc, Chan SG, Sham A, Yuen YK, Masarei JL. Lipid profiles in the Chinese elderlies: comparison with younger age groups and relationship with some cardiovascular risk factors and presence of disease. Cardiology 1993; 83: Abate N, Gray A, Coleman R, Grundy SM, Peshock RM. Prediction of total subcutaneous abdominal, intraperitoneal, and retroperitoneal adipose tissue masses in men by a single axial magnetic resonance imaging slice. Am J Clin Nutr 1997; 65:

7 8 Gray DS, Fujioka K, Colletti PM, Kim H, Devine W, Guyegkeng T. Magnetic-resonance imaging used for determining fat distribution in obesity and diabetes. Am J Clin Nutr 1991; 54: Ramirez ME. Measurement of subcutaneous adipose tissue using ultrasound images. Am J Phys Anthropol 1992; 89: Orphanidou C, McCargar L, Birmingham CL, Mathieson J, Goldner E. Accuracy of subcutaneous fat measurement: comparison of skinfold calipers, ultrasound, and computed tomography. J Am Diet Assoc 1994; 94: Fanelli MT, Kuczmarski RJ, Hirsch M. Estimation of body fat from ultrasound measures of subcutaneous fat and circumferences in obese women. Int J Obes Relat Metab Disord 1988; 12: Tadokoro N, Murano S, Nishide T, Suzuki R, Wantanabe S, Murayama H, Morisaki N, Saito Y. Preperitoneal fat thickness determined by ultrasonography is correlated with coronary stenosis and lipid disorders in non-obese male subjects. Int J Obes Relat Metab Disord 2000; 24: Suzuki R, Wantanabe S, Hirai Y, Akiyama K, Nishide T, Matsushima Y, Murayama H, Ohshima H, Shinomiya M, Shirai K, Saito Y, Yoshida S, Saisho H, Ohto M. Abdominal wall fat index, estimated by ultrasonography, for assessment of the ratio of visceral fat to subcutaneous fat in the abdomen. Am JMed1993; 95: Derchi LE, Solbiati L, Rizzatto G, Pra LD. Normal anatomy and pathologic changes of the small bowel mesentery: US appearances. Radiology 1987; 164: Anderson PJ, Chan JCN, Chan YL, Tomlinson B, Young RP, Lee ZSK, Lee KK, Metreweli C, Cockram CS, Critchley JA. Visceral fat and cardiovascular risk factors in Chinese NIDDM patients. Diabetes Care 1997; 20: Ross R, Shaw KD, Rissanen J, Martel Y, Guise JD, Avruch L. Sex differences in lean and adipose tissue distribution by magnetic resonance imaging: anthropometric relationships. Am J Clin Nutr 1994; 59: Dowse GK, Zimmet PZ, Garceboo H, George K, Alberti MM, Tuomilehto J, Finch CF, Chitson P, Tulsida H. Abdominal obesity and physical inactivity as risk factors for NIDDM and impaired glucose tolerance in Indians, Greole, and Chinese Mauritians. Diabetes Care 1991; 14: Mathews DR, Hoskers JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and b-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: Bertrais S, Balkau B, Vol S, Forhan A, Calvet C, Marre M, Eschwega E. Relationships between abdominal body fat distribution and cardiovascular risk factors: an explanation for women s healthier cardiovascular risk profile. The D.E.S.I.R. study. Int J Obes Relat Metab Disord 1999; 23: Reckelhoff JF. Gender differences in the regulation of blood pressure. Hypertension 2001; 37: Jain R, Sawhney S, Bhargava DK, Berry M. Diagnosis of abdominal tuberculosis: sonographic findings in patients with early disease. AJR 1995; 165: Bjorntorp P. Metabolic implications of body fat distribution. Diabetes Care 1991; 14: Matsuzawa Y, Funahashi T, Nakamura T, Shimomura I, Arita Y. Visceral obesity: from genes to diseases. In: Guy-Grand B, Ailhaud G. (eds). Progress in obesity research, vol. 8. John Libbey & company: London 1999, pp