| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles |
Service de Diabétologie et d’Hormonologie et Service de Radiologie, Hôpital Saint-Louis, 75475 Paris Cedex 10, France; and IMASSA, Centre d’Essais en Vol, Département de Physiologie Systémique, 91223 Bretigny sur Orge, France
Address all correspondence and requests for reprints to: Dr. Jean-François Gautier, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, 4212 North 16th Street, Phoenix, Arizona 85016.
Abstract
Accumulation of visceral adipose tissue is associated with metabolic complications such as noninsulin-dependent diabetes mellitus. The aim of this study was to evaluate the effect of abdominal adipose tissue on insulin sensitivity in subjects with noninsulin-dependent diabetes mellitus (NIDDM). Areas of abdominal fat were calculated from axial magnetic resonance images obtained at the level of the umbilicus in 21 men with NIDDM [age, 45.6 ± 8.3 (±SD) yr; body mass index, 29.3 ± 4.5 kg/m-2; total body fat (skinfold thickness), 26.8 ± 5.4%; waist to hip ratio, 0.97 ± 0.07; duration of diabetes, 59 ± 47 months; hemoglobin A1c, 8.1 ± 1.5%]. Insulin sensitivity was evaluated by an insulin tolerance test. The areas of deep abdominal fat and sc abdominal fat were, respectively, 135.3 ± 55.1 and 211.8 ± 99.1 cm2. The blood glucose disappearance rate was 2.11 ± 0.87%/min and was negatively related to deep abdominal fat (r = 0.72; P = 0.0025). In contrast, areas of sc abdominal fat, total body fat, body mass index, and waist to hip ratio were not related to the blood glucose disappearance rate. Plasma triglyceride concentrations averaged 1.8 ± 0.8 mmol/L and were positively related to deep abdominal fat (r = 0.69; P = 0.0018). We conclude that insulin sensitivity is strongly related to visceral adipose tissue accumulation in NIDDM.
SEVERAL prospective studies have shown that obesity is a risk factor for noninsulin-dependent diabetes mellitus (NIDDM). However, as Vague suggested in 1956 (1), recent studies based on the waist to hip ratio (WHR) indicate that the distribution of fat in the upper body area seems to be a greater risk factor for NIDDM and cardiovascular disease than obesity itself (2, 3, 4). The WHR is a simple anthropometric variable that provides an estimation of the proportion of abdominal or upper body fat. However, it does not distinguish accumulations of visceral adipose tissue (VAT), also called deep abdominal fat, from sc abdominal adipose tissue (SAT). Recently, computed tomography was used to measure the amounts of VAT and SAT. The results obtained with this technique indicated that several abnormalities of the plurimetabolic syndrome that predispose to NIDDM, such as insulin resistance, increased blood pressure, and dyslipidemia, were more pronounced in subjects with a type of fat distribution that reflected greater adiposity in the visceral abdominal regions (5, 6, 7). A prospective study conducted in Japanese-American men showed that in individuals who develop NIDDM, increased deep abdominal fat is present before the onset of NIDDM (8). More recently, magnetic resonance imaging (MRI) was used, as adipose tissue (AT) is characterized by a typical short longitudinal relaxation time (t1) compared to that in other tissues (9, 10). Moreover, this method does not involve x-ray exposure, and the duration for abdominal evaluation of fat is shorter than that using computed tomography. The reproducibility of MRI results is known to be lower than that of computed tomography, but could be greatly increased by the use of a reference tube to eliminate internal parameters (11). Measurements of abdominal fat distribution evaluated by computed tomography or MRI in Caucasians with type 2 diabetes mellitus suggest that NIDDM is associated with more accumulation of deep abdominal fat than that in nondiabetic subjects with similar body weight (12, 13). The aim of our study was to evaluate abdominal fat distribution in these patients and to investigate the relationship between the amount of VAT and insulin resistance. Whole body insulin sensitivity was assessed using the insulin tolerance test (ITT) according to the method of Bonora et al. (14).
Experimental Subjects
Inclusion criteria required that subjects be middle-aged men with NIDDM at stable body weight for at least 3–4 months. A known duration of diabetes less than 10 yr was mandatory. Twenty-one patients were selected for the study. The study protocol was approved by the ethics committee of Saint-Louis Hospital. All patients gave their written informed consent. Most of them were slightly obese. Glycated hemoglobin was 8.1 ± 1.5% (mean ± SD). Patients were treated with hypoglycemic agents (n = 17) or by diet alone (n = 4). Hypoglycemic agents consisted of metformin alone for 14 patients and sulfonylurea alone for 3 patients. They did not receive hypotensive treatment, and blood pressure was recorded after 10 min in the supine position with a mercury sphygmomanometer. No patient had significant renal, hepatic, or cardiac disease, and none was using agents known to affect glucose metabolism, except hypoglycemic agents. They had no micro- or macrodiabetic angiopathy. Maximal electrocardiogram exercise testing was negative.
Food diaries were recorded daily for 1 week by the patients and were analyzed using dietetic tables (15). The level of physical activity was evaluated using Beacke’s questionnaire (16).
Materials and Methods
Exercise testing
The maximal oxygen uptake (
O2max) was
determined by a direct method using an open system to analyze expired
gases (Sensor Medics 2900, Sensor Medics Corp., Yorba Linda, CA).
Subjects exercised on a mechanically braked ergocycle (Monark Crescent,
Valberg, Sweden) according to an incremental protocol used for
O2max determination. The maximal exercise test began
by a 2-min warm-up at 30 watts, followed by the incremental test during
which power output increased by 30 watts every 2 min (17). A pedaling
frequency of 60 rpm was selected. The test protocol was adjusted to
ensure attainment of maximal effort within 15–20 min. Oxygen uptake,
per min ventilation, respiratory exchange ratio, heart rate, and power
(watts) were continuously monitored. O2max was
achieved when O2 demonstrated a plateau during an
increase in power output and/or the respiratory exchange ratio was
higher than 1.1, the blood lactate concentration was higher than 8
mmol/L, and the subject exhibited motor deficiencies and visual
tiredness (17). Because O2 did not reach a plateau
in most patients, the highest VO2 value obtained during
incremental exercise corresponded to a O2 peak
rather than O2max.
Resting metabolic rate
Subjects went to the laboratory at 0830 h in the
postabsorptive state after a 12-h fast. They sat quietly in a
semisupine position. After a 30-min rest period, baseline
postabsorptive metabolic rate was measured during 30 min by Sensor
Medics 2900’s dilution mode indirect calorimetry testing with a mask
(without a mouthpiece). Laboratory temperature was maintained at 24 C
throughout the study. The gas analyzers were recalibrated before each
test to correct for drift in the analyzers. During the postabsorptive
metabolic rate, subjects were allowed to read throughout the
measurement period. For each metabolic measurement, the respiratory
quotient (CO2/O2) was calculated,
and the result was converted to kilocalories using Weir’s equation
(18). For repeated measurements of the resting metabolic rate, we
reported that the mean difference was 3.7% (19).
Anthropometric variables
Height of the barefoot subject was measured to the nearest 0.1 cm. Body weight was measured on the same standard medical weight scale with an accuracy of ±100 g, and body mass index (BMI; kilograms per m2) was calculated. The waist circumference was measured as the narrowest circumference between the lower costal margin and the iliac crest in the standing position. The hip measurement was the maximum circumference at the level of the femoral trochanters.
Whole body fat was estimated using the skinfold thickness method developed by Durnin and Womersley (20). Briefly, skinfold measurements were taken from four sites (bicipital, tricipital, subscapular, and suprailiac) using a Harpenden skinfold calliper (Serita, East Rutherford, NJ). A minimum of two measurements were made at each skinfold site by the same highly experienced investigator. The values were averaged and used to estimate body fat percentage.
AT measurement by MRI
MRI examinations were performed on a 0.5-Tesla superconducting
MR system (MRMax, General Electric, Milwaukee, WI). Subjects were
positioned in the magnet in a supine feet-first position. A T1-weighted
multislice spin echo sequence was obtained using a repetition time of
500 ms and an echo time of 25 ms. These parameters provide optimal
contrast between AT and lean tissue. The MR sequence included three
7-mm-thick axial images with a 3-mm gap and was performed in the
abdomen at the level of the umbilicus. The measurement of AT surfaces
was performed on large regions of interest (ROIs) drawn manually on
each slice (Fig. 1
). One ROI included the
largest cross-sectional area of SAT, and the other one included the
largest cross-sectional area of VAT. Segmentation of AT within ROIs was
performed as previously described (11, 19). Briefly, unlike images from
computed tomography, no standard scale exists for magnetic resonance
images. Consequently, pixel intensity value for a given tissue may not
be consistent from slice to slice or from one individual to the next.
To set a common dynamic range for all images, a 1.5-cm diameter
reference tube was filled to capacity with a solution of
gadolinium-DOTA (0.5 mmol/L) and placed on either side of the subjects
during data acquisition. The reference tube serves as an internal
marker to determine the threshold of pixels corresponding to AT. The
analysis of a sample of typical images and their respective gray level
histograms allowed us to determine the optimal threshold of AT as
0.7-fold that of the reference tube. Above this value, pixels were
considered as bearing witness to AT. The next step involved
highlighting and counting the AT pixels in response to the selected
threshold, and multiplying by the pixel surface to calculate the
surface of AT. This measurement was performed on the three slices and
then averaged. The reproducibility of the measurements was tested on a
sample of 10 patients. Intra- and interobserver variations did not
exceed 2%. Because we used a reference tube, the pixel intensity
threshold of AT was the same in all patients. The intra- and
interobserver variations depended only on the drawing of ROIs.
|
An iv ITT was performed after an overnight fast, at least 48 h after the withdrawal of metformin. The ITT consisted of a bolus iv injection of regular insulin (0.1 U/kg BW) (14). Blood samples were collected 1 min before and 3, 6, 9, 12, and 15 min after insulin injection. The constant rate plasma glucose disappearance (Kitt) was calculated from the formula 0.693/t1/2. The plasma glucose t1/2 was calculated from the slope of least square analysis of the plasma glucose concentrations from 3–15 min after iv insulin injection, when the plasma glucose concentration declined linearly. Fasting blood glucose levels were assayed in neutralized blood samples using a glucose oxidase enzymatic method (21), whereas plasma insulin concentrations were measured by RIA using a commercial reagent (INSIK-5, Sorin Biomedica). Glycated hemoglobin levels (normal value, <6%) were determined by high performance liquid chromatography. Fasting serum lipids were measured enzymatically, and fasting plasma free fatty acids (FFA) concentrations were assayed according to the method of Ho (22).
Statistical analysis
All data are given as the mean ± SD. All statistical analyses were performed using StatView II software (Abacus Concept, Berkeley, CA). The relations between variables were analyzed by both simple and multiple regression with stepwise variable selection. P < 0.05 was selected to indicate significant values.
Results
Subjects were 45.6 ± 8.3 yr old. They had a BMI of 29.3
± 4.5 kg/m2 and a body fat of 26.8 ± 5.4%. The WHR
averaged 0.97 ± 0.07. Systolic and diastolic blood pressure
values were near normal (133 ± 14 and 84 ± 8 mm Hg,
respectively). The O2 peak was 22.7 ± 4.0
mL/kg·min. Resting energy expenditure averaged 1603 ± 336
Cal/day. The biological characteristics of the study subjects are
summarized in Table 1.
|
Correlation between insulin sensitivity and anthropometric parameters
Table 2 shows the correlation
coefficients between anthropometric measures and metabolic indexes.
There was a high significant inverse correlation between
Kitt and VAT, as depicted in Fig. 2. A negative correlation was seen
between Kitt and plasma triglyceride concentrations (Fig. 2). Conversely, Kitt was not associated with
areas of SAT, total body fat, BMI, or WHR. No relationship was seen
between Kitt and fasting insulin concentrations. Figure 2
also shows the lack of correlation between Kitt and SAT
(r = 0.17). The stepwise multiple regression with Kitt
as the dependent variable and anthropometric measures, MRI parameters,
biochemical analyses, and VO2peak as independent variables
yielded only one significant variable (VAT), which accounted for about
52% of the variance in Kitt (r = 0.72).
|
|
As shown in Table 2
, areas of VAT were correlated only with
Kitt and fasting plasma triglyceride levels. The stepwise
multiple regression excluded plasma triglyceride concentrations as a
significant variable. WHR were weakly associated with areas of SAT, but
not with areas of VAT. Areas of SAT were also related to BMI (r =
0.87; P = 0.001) and to whole body fat (r = 0.66;
P = 0.0026)
Relationship of insulin sensitivity and VAT to physical fitness and cardiovascular parameters
Simple regression analysis did not show any relationship between
O2peak and Kitt (r = 0.14) or areas
of VAT (r = 0.13). Resting energy expenditure values were also not
related to these parameters. No correlation was seen between insulin
sensitivity and diastolic or systolic blood pressure (r = 0.15 and
r = 0.23, respectively). Furthermore, blood pressure was not
correlated with areas of VAT. No relationship was observed between
insulin sensitivity and physical exercise level or diet
composition.
Discussion
Accumulation of deep abdominal fat is associated with a multitude of metabolic complications, such as glucose intolerance, hyperinsulinemia, changes in plasma lipid concentrations, and increased blood pressure. Indeed, it is a stronger predictor of cardiovascular disease and other metabolic complications of obesity, such as NIDDM, than is overall fitness (7, 8). It was previously suggested in the U.S. that overweight Caucasian patients with NIDDM have more VAT accumulation than nondiabetic obese patients with similar BMI (13). Moreover, lean Japanese-America subjects have more intraabdominal adiposity than control subjects (12). Thus, the present clinical investigation was aimed at investigating abdominal fat distribution in NIDDM patients living in Europe and the relationship between insulin sensitivity and the amount of VAT. Thanks to the use of a reference tube and the reliability of MRI, the method we used to estimate abdominal fat distribution was close to computed tomography. Our MRI protocol determined abdominal fat distribution at only one level. However, Ross et al. (23) showed that the results obtained by this method correlate well with total intraabdominal fat. Insulin sensitivity was determined by the short ITT, a simple test widely used in clinical investigations. A good correlation had been demonstrated between the blood glucose disappearance rate (Kitt) and the amount of glucose uptake obtained from the euglycemic hyperinsulinic clamp (14, 24, 25). The reproducibility of this test seems to be satisfactory, as the intratest coefficient of variation varied from 6–14% (14, 24, 25, 26) (Gautier, J.-F., personal data). Contrary to the euglycemic hyperinsulinic clamp, the ITT is an indirect method that estimates the whole body insulin sensitivity without distinguishing the effect of insulin on muscle from that on liver. However, one can speculate that an iv bolus of 0.1 IU/kg leads to such supraphysiological plasma insulin concentrations that it rapidly inhibits hepatic glucose production (27).
Our finding shows clearly that insulin sensitivity is inversely associated with the amount of VAT in NIDDM patients. The higher the amount of VAT, the lower the insulin sensitivity. Several studies found such an association between glucose metabolism and VAT in nondiabetic populations or in the prediabetic state (5, 6, 7, 8, 28, 29). Our results are also consistent with the study by Banerji et al. (30) carried out in African-American NIDDM men with approximately the same range of weight as our patients, in which a strong correlation was seen between insulin-mediated glucose uptake and VAT. Moreover, they found that insulin sensitivity is inversely related to serum triglyceride levels, as has been described in other, but nondiabetic, populations. However, in the study of Abate et al. (31), no correlation was seen between insulin sensitivity and intraperitoneal AT in Caucasian subjects with NIDDM, whereas these parameters were correlated in the control group. Our patients were younger than those of Abate’s study and were approximately the same age as his control group. This age difference may partly explain the discrepancy between both studies. In addition, in Abate’s study, insulin sensitivity has been adjusted for lean body mass. We did not report our data per kg lean body mass because of the limitations of the method using skinfold thickness in overweight subjects.
Although to date no study has successfully identified the mechanisms of the link between insulin resistance and VAT, it has been suggested that the high lipolytic response of VAT to catecholamines (32, 33) exposes the liver to high FFA concentrations via the portal circulation, thereby leading to insulin resistance (34). Indeed, the link between FFA and hepatic glucose production is now well documented (35), and it has been shown that FFAs inhibit insulin-stimulated whole body glucose uptake and utilization in patients with NIDDM (36).
Another point raised by our study is that WHR was weakly associated with BMI, but was not related to areas of VAT or to insulin sensitivity. Banerji et al. (30) and Ross et al. (10) found a weak relation between WHR and VAT accumulation. One factor that may influence the relationship between VAT and WHR is the level of waist circumference used to calculate WHR. Thus, Peiris et al. (37) observed a significant correlation between WHR and areas of VAT when WHR was calculated using a waist circumference obtained at the level of minimum girth instead of the umbilicus. Another factor that may contribute to the discrepancy is that WHR is normally obtained with the subject in a standing position, whereas computed tomography or MRI are obtained in supine subject. Indeed, the effect of gravity may change the position of abdominal mass. However, Ross et al. (10) did not see an improvement in the relation between VAT and WHR when WHR was determined in supine position. Banerji et al. (30) found a better correlation between VAT and waist circumference than between VAT and WHR in black NIDDM subjects. Waist circumference has been shown to be a good predictor of cardiovascular risk factors (38, 39). Moreover, recent epidemiological studies showed that waist circumference is more strongly related to insulin sensitivity than is WHR itself in a population-based sample of young healthy Caucasian subjects (40) and in African-American and Caucasian men and women (41). The Manitoba heart study (42), a cross-sectional study including 2792 adults, indicates that BMI could be a better predictor than WHR of the metabolic effects of central obesity. Finally, according to some studies, WHR is a poor predictor of change in visceral fat after dietary restriction (43) or a physical exercise program (19).
In conclusion, insulin resistance is associated with visceral fat accumulation in NIDDM Caucasian subjects. In this population, WHR is not a predictor of VAT or insulin resistance. The close link between VAT and insulin resistance, a major characteristic of NIDDM, should encourage us to make visceral fat depot reduction a more important goal than reducing total body weight itself.
|
We are grateful to Françoise Lienhard, Maryse Souillard, and Elisabeth Treillard for their skillful technical help.
Footnotes
1 This work was supported by the Caisse Nationale de Prévoyance,
the Caisse Nationale d’Assurance Maladie des Professions
Indépendantes, the Caisse Maladie Régionale des Professions
Industrielles et Commerciales d’Ile de France, and the Assistance
Publique des Hôpitaux de Paris. 
Received May 27, 1997.
Revised October 17, 1997.
Accepted December 15, 1997.
References
This article has been cited by other articles:
 |
|
 |
|
  K. Gross, I. Karagiannides, T. Thomou, H. W. Koon, C. Bowe, H. Kim, N. Giorgadze, T. Tchkonia, T. Pirtskhalava, J. L. Kirkland, et al. Substance P promotes expansion of human mesenteric preadipocytes through proliferative and antiapoptotic pathways Am J Physiol Gastrointest Liver Physiol, May 1, 2009; 296(5): G1012 - G1019. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Somm, V. M. Schwitzgebel, D. M. Vauthay, E. J. Camm, C. Y. Chen, J.-P. Giacobino, S. V. Sizonenko, M. L. Aubert, and P. S. Huppi Prenatal Nicotine Exposure Alters Early Pancreatic Islet and Adipose Tissue Development with Consequences on the Control of Body Weight and Glucose Metabolism Later in Life Endocrinology, December 1, 2008; 149(12): 6289 - 6299. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Ray MD MSc, A. P. Mohllajee MPH, R. M. van Dam PhD, and K. B. Michels ScD PhD Breast size and risk of type 2 diabetes mellitus Can. Med. Assoc. J., January 29, 2008; 178(3): 289 - 295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Azuma, L. K. Heilbronn, J. B. Albu, S. R. Smith, E. Ravussin, D. E. Kelley, and and the Look AHEAD Adipose Research Group Adipose tissue distribution in relation to insulin resistance in type 2 diabetes mellitus Am J Physiol Endocrinol Metab, July 1, 2007; 293(1): E435 - E442. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Virtanen, P. Iozzo, K. Hallsten, R. Huupponen, R. Parkkola, T. Janatuinen, F. Lonnqvist, T. Viljanen, T. Ronnemaa, P. Lonnroth, et al. Increased Fat Mass Compensates for Insulin Resistance in Abdominal Obesity and Type 2 Diabetes: A Positron-Emitting Tomography Study Diabetes, September 1, 2005; 54(9): 2720 - 2726. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Giannopoulou, L. L. Ploutz-Snyder, R. Carhart, R. S. Weinstock, B. Fernhall, S. Goulopoulou, and J. A. Kanaley Exercise Is Required for Visceral Fat Loss in Postmenopausal Women with Type 2 Diabetes J. Clin. Endocrinol. Metab., March 1, 2005; 90(3): 1511 - 1518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. S. Fox, N. L. Heard-Costa, P. W.F. Wilson, D. Levy, R. B. D'Agostino Sr., and L. D. Atwood Genome-Wide Linkage to Chromosome 6 for Waist Circumference in the Framingham Heart Study Diabetes, May 1, 2004; 53(5): 1399 - 1402. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Mauvais-Jarvis, E. Sobngwi, R. Porcher, J.-P. Riveline, J.-P. Kevorkian, C. Vaisse, G. Charpentier, P.-J. Guillausseau, P. Vexiau, and J.-F. Gautier Ketosis-Prone Type 2 Diabetes in Patients of Sub-Saharan African Origin: Clinical Pathophysiology and Natural History of {beta}-Cell Dysfunction and Insulin Resistance Diabetes, March 1, 2004; 53(3): 645 - 653. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Q. Purnell, R. K. Dev, M. W. Steffes, P. A. Cleary, J. P. Palmer, I. B. Hirsch, J. E. Hokanson, and J. D. Brunzell Relationship of Family History of Type 2 Diabetes, Hypoglycemia, and Autoantibodies to Weight Gain and Lipids With Intensive and Conventional Therapy in the Diabetes Control and Complications Trial Diabetes, October 1, 2003; 52(10): 2623 - 2629. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Miyazaki, L. Glass, C. Triplitt, E. Wajcberg, L. J. Mandarino, and R. A. DeFronzo Abdominal fat distribution and peripheral and hepatic insulin resistance in type 2 diabetes mellitus Am J Physiol Endocrinol Metab, December 1, 2002; 283(6): E1135 - E1143. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Arfai, P. D. Pitukcheewanont, M. I. Goran, C. J. Tavare, L. Heller, and V. Gilsanz Bone, Muscle, and Fat: Sex-related Differences in Prepubertal Children Radiology, August 1, 2002; 224(2): 338 - 344. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Verma, C. M. Boney, R. Tucker, and B. R. Vohr Insulin Resistance Syndrome in Women with Prior History of Gestational Diabetes Mellitus J. Clin. Endocrinol. Metab., July 1, 2002; 87(7): 3227 - 3235. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. de Souza, M. Eckhardt, K. Gagen, M. Dong, W. Chen, D. Laurent, and B. F. Burkey Effects of Pioglitazone on Adipose Tissue Remodeling Within the Setting of Obesity and Insulin Resistance Diabetes, August 1, 2001; 50(8): 1863 - 1871. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Bando, Y. Ushiogi, K. Okafuji, D. Toya, N. Tanaka, and M. Fujisawa The Relationship of Fasting Plasma Glucose Values and Other Variables to 2-h Postload Plasma Glucose in Japanese Subjects Diabetes Care, July 1, 2001; 24(7): 1156 - 1160. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Hadigan, C. Corcoran, T. Stanley, S. Piecuch, A. Klibanski, and S. Grinspoon Fasting Hyperinsulinemia in Human Immunodeficiency Virus-Infected Men: Relationship to Body Composition, Gonadal Function, and Protease Inhibitor Use J. Clin. Endocrinol. Metab., January 1, 2000; 85(1): 35 - 41. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
x 174 DNA standards
(