Introduction

Obesity is known to be a major risk factor for insulin resistance and type 2 diabetes (1,2). Ferrannini et al. showed that 35% of individuals with a BMI between 30 and 35 kg/m² and 60% of those with a BMI >35 are insulin resistant (1). Consequently, these observations imply that 65% of the population with a BMI between 30 and 35 kg/m², and 40% of individuals with a BMI >35 kg/m² would be insulin sensitive.

The link between obesity and insulin resistance is not fully understood. Studies conducted over the past decades have clearly shown that body fat distribution, and especially visceral fat (VF) accumulation, play a major role in the etiology of insulin resistance and its associated metabolic abnormalities (3,4,5,6,7). However, the contribution of lean body mass (LBM; mainly composed of muscle mass) to the pathogenesis and development of the metabolic syndrome and insulin resistance must also be considered (8), as shown in other studies (9,10,11,12,13,14,15,16). Despite the fact that some studies conducted over the past decades have recognized that a high LBM is associated with high levels of insulin sensitivity (17,18,19,20,21), at least four studies also reported that in obese women, a higher LBM can be associated with metabolic alterations (9,16,22,23). First, in 1986, Krotkiewski and Björntorp (9) showed that obese women with less trunk fat and lower LBM were more insulin sensitive than those with a muscular body type and greater trunk fat accumulations. More recently, in studies using precise techniques to measure body composition (dual-energy X-ray absorptiometry), body fat distribution (computed tomography (CT)), and insulin sensitivity (euglycemic/hyperinsulinemic clamp), we showed that insulin-resistant obese postmenopausal (PM) women had significantly more VF and LBM than insulin-sensitive obese women (16,22). Our observations were confirmed by You et al., who also reported higher LBM in viscerally obese PM women displaying the metabolic syndrome (23) compared to those with a normal metabolic profile.

Overall, these four studies conducted in sedentary obese PM women showed that contrary to the general belief, higher levels of LBM are not systematically associated with better insulin sensitivity (9,16,22,23). In fact, these studies showed that trunk body fat (9), and more particularly VF levels (22,23), as well as higher LBM (22,23) were associated with metabolic disturbances. This study was thus conducted to investigate the potential association between VF levels and LBM as well as the contribution of LBM to insulin resistance and C-reactive protein (CRP) levels in patients with different levels of VF.

Methods and Procedures

Subjects

The baseline study population consisted of 132 nondiabetic PM women with measures of insulin sensitivity by euglycemic/hyperinsulinemic clamp. First, subjects were categorized as having either "higher" or "lower" VF (higher 172 cm² vs. lower 172 cm²) and lean BMI (LBMI; higher 15.8 kg LBM/m² vs. lower 15.8 kg LBM/m²) based on above or below the 50th percentile for each variable in our cohort. Four groups of subjects were then created (group 1: lower VF/lower LBMI; group 2: lower VF/higher LBMI; group 3: higher VF/lower LBMI; group 4: higher VF/higher LBMI). We used VF levels (5,6,7) and LBMI (16) because both variables have been independently associated to insulin resistance and the metabolic profile. Second, because (i) the amount of total fat mass (FM) has been associated with glucose homeostasis (24) and (ii) the population studied displayed a broad range of FM (26.8–49.8 kg), we excluded subjects with FM levels above and below an s.d. of 1.5 (n = 29). This strategy allowed us to control the effect of this potential confounder by having comparable amount of total FM among the four groups, while VF and LBM levels remained significantly different as designed (Table 1). Consequently, 103 subjects (% FM: 45.5 4.1% , ranging from 37.4 to 54.9% ) between 49 and 70 years old (58.0 4.9 years; mean s.d.) were used for data analyses.

Table 1 - Characteristics of subjects categorized based on higher vs. lower visceral fat (VF) accumulations and higher vs. lower lean BMI.

Full table (73K)

Inclusion criteria were: BMI 27 kg/m², no menstruation for >1 year, having a follicle stimulating hormone level >30 U/l, sedentary (<2 h per week of structured exercise), nonsmokers, low to moderate alcohol consumers (2 drinks/day), and not on hormone replacement therapy. All participants were apparently healthy and had no history or evidence on physical examination or laboratory testing of (i) cardiovascular disease, peripheral vascular disease, or stroke; (ii) diabetes (2 h standard 75-g oral glucose tolerance test (OGTT)); (iii) severe hypertension (resting blood pressure >170/100 mm Hg); (iv) orthopedic limitations; (v) body weight fluctuation >5 kg in the previous 6 months; (vi) uncontrolled thyroid or pituitary disease; and (vii) medication that could affect cardiovascular function or metabolism. All participants signed an informed consent document. This study was approved by the University of Montreal Medical Committee on Human Research.

Diet stabilization period

Before the study, subjects were asked to maintain normal activity level and eating habits for 4 weeks before the testing sequence and to maintain their body weight within a 2 kg range in order to reduce the acute effects of fluctuations in energy balance on outcome variables (25,26). If the subjects were unable to maintain their body weight, stabilization period was prolonged until their weight was stable for 4 consecutive weeks.

Peak VO ₂

Subjects performed a graded exercise test on an ergocycle Ergoline 900 (Bitz, Germany) to voluntary exhaustion. During the test, power output was increased by 25 W every 2 min. Peak VO₂ (l/min) was considered to be the highest value obtained during the test. Expired gas was analyzed during the exercise protocol using an Ergocard (software version 6; MediSoft, Dinant, Belgium) cardiopulmonary exercise test station. Standard 12-lead electrocardiograms were performed at the end of every 2-min stage. Three of the following criteria were required for a successful test: a respiratory exchange ratio >1.1; heart rate within 10 bpm of maximal predicted heart rate value (220 - age); volitional cessation of exercise by the subject and a plateau in oxygen consumption for 60 s. A test–retest reliability trial (n = 19) for VO₂ max performed in our laboratory showed an intraclass correlation coefficient of 0.96.

Anthropometry

Body weight was measured to the nearest 0.1 kg on a calibrated balance (Balance Industrielle Montreal, Montreal, Quebec, Canada) and subject's height was obtained using a standard stadiometer (Perspective Enterprises, Portage, MI). Determination of total FM, percentage of FM (% FM), total LBM, and bone mineral content were assessed using dual-energy X-ray absorptiometry using software version 6.10.019 (General Electric Lunar Prodigy, Madison, WI). During the procedure, subjects were asked to wear only a standard hospital gown while in the supine position (26,27). Calibration was executed daily with a standard phantom before each test and the intraclass coefficient correlation for test–retest for FM and LBM was 0.99 (n = 18).

The BMI was calculated first. Then, because the BMI is a nonspecific measure of fatness that does not discriminate lean body and fat tissues (28,29), we also calculated the FM index (FMI = FM (kg)/height (m²)) and the LBM index (LBMI = LBM (kg)/height (m²); without taking into account bone mass into the calculation) using data obtained from the dual-energy X-ray absorptiometry, as described previously (16). One advantage of using the FMI and the LBMI, as compared to the BMI alone, is that it better takes into account the relative effect of aging and individual variations on body fat and LBM (16,28,29,30,31,32). Furthermore, interindividual variations in both variables in absolute value fail to allow an appropriate comparison among subjects of different sizes (27,28,29,30). Finally, the use of LBMI as indices of muscle mass was also justified by the fact that LBM was significantly correlated with height (r = 0.55; P < 0.0001) in our cohort (data not shown).

CT

A GE LightSpeed 16 CT scanner (General Electric Medical Systems, Milwaukee, WI) was used to measure the visceral and the subcutaneous fat (ScF) area. Subjects were examined in the supine position with both arms stretched above their head. The position of the scan was established at the L4–L5 vertebral disc using a scout image of the body (5,22). VF was quantified by delineating the intra-abdominal cavity at the internal most aspect of the abdominal and oblique muscle walls surrounding the cavity and the posterior aspect of the vertebral body. The ScF area was quantified by highlighting fat located between the skin and the external most aspect of the abdominal muscle wall. Deep and superficial ScF areas were measured by delineating the subcutaneous fascia within the ScF and by computing areas of the layers of fat on each side of the fascia (33). The cross-sectional areas of fat were highlighted and computed using an attenuation range of - 190 to - 30 Hounsfield Units (HU).

Mid-thigh cross-sectional skeletal muscle and fat areas as well as muscle attenuation (MA) were measured (34). Areas of skeletal muscle, fat, and MA were calculated by delineating the regions of interest and then computing the surface areas using an attenuation range of - 190 to - 30 HU for fat, and 0–100 HU for skeletal muscle. Test–retest measures of the different body fat distribution indices on 10 CT scans yielded a mean absolute difference of 1% (22).

OGTT

Before the study, a 2-h 75-g OGTT was performed in the morning after a 12-h fast according to the guidelines of the American Diabetes Association (35). The aim of the OGTT was to identify undiagnosed diabetic patients, which was an exclusion criterion. Blood samples were collected through a venous catheter from an antecubital vein in vacutainer tubes containing EDTA (SST Gel and Clot Activator) at 0, 30, 60, 90, and 120 min. Plasma glucose was rapidly measured on the COBAS INTEGRA 400+ (Roche Diagnostic, Montreal, Canada), while insulin levels were determined in duplicate using a human insulin–specific radioimmunoassay (RIA kit; Linco Research, St Charles, MO).

Insulin sensitivity measurement during the clamp

Patients underwent a 3-h euglycemic/hyperinsulinemic clamp after the weight stabilization period. The test began at 0730 AM after a 10-h overnight fast following the procedure described by DeFronzo et al. (36). An antecubital vein was cannulated for infusion of 20% dextrose and insulin (Actrapid; Novo-Nordisk, Canada). The other arm was cannulated for sampling of blood. Three basal sample of plasma glucose and insulin were taken over 30 min. Then, a primed-constant insulin infusion was started at the rate of 75 mU/m²/min for 180 min. Plasma glucose was measured every 10 min with a glucose analyzer (Beckman Instruments, Fullerton, CA) and maintained at fasting level with a variable infusion rate of 20% dextrose. Blood was drawn every 10 min during the last 30-min euglycemic/hyperinsulinemic clamp to determine plasma glucose and insulin. The mean rate of glucose disposal (exogenous dextrose infusion) during the last 30 min of the clamp was considered as the insulin sensitivity index or "M" value.

Blood samples

After an overnight fast (12 h), venous blood samples were collected for the measurement of plasma concentrations of total cholesterol, high-density lipoprotein cholesterol (HDL-chol), low-density lipoprotein cholesterol (LDL-chol), and triglycerides. Plasma was analyzed on the day of collection. Analyses were done on the COBAS INTEGRA 400 analyzer (Roche Diagnostics, Montreal, Canada) for total cholesterol, HDL-chol, and triglycerides. Total cholesterol, HDL-chol, and triglycerides levels were used in the Friedewald formula (37) to calculate LDL-chol concentrations. Serum high-sensitivity CRP (hsCRP) concentrations were assessed by immunonephelometry on IMMAGE analyzer (Beckman Coulter, Villepinte, France).

Identifying subjects with the metabolic syndrome

Subjects were characterized as having or not the metabolic syndrome based on the National Cholesterol Education Program's Adult Treatment Panel III report (ATP III) (38). ATP III–defined metabolic syndrome required at least three of the following criteria: elevated waist circumference (>88 cm in women), hypertriglyceridemia (1.69 mmol/l), low HDL-chol (<1.30 mmol/l in women), high blood pressure (130/85 mm Hg or pharmacological treatment for hypertension), and elevated fasting plasma glucose levels (6.1 mmol/l).

Statistical analyses

Data in tables are presented as mean values s.d. ANOVA revealed that age was significantly different among groups (Table 1). Then, we used Pearson's correlations to see if age was associated with variables of interest in our cohort. Our results showed that age was significantly correlated with peak VO₂ (P < 0.005), VF (P < 0.05), MA (P < 0.005), and resting systolic blood pressure (P < 0.05) (data not shown).

Consequently, ANOVA were performed to compare groups for most of the variables of interest, while analysis of covariance (ANCOVA) adjusting for age were used to compare groups for peak VO₂, VF accumulations, MA, and resting systolic blood pressure. The Tukey–Kramer test was used for posteriori group comparisons when a main model effect was noted (Tables 1 and 2). Pearson's correlations were used to quantify the associations between various body composition and metabolic variables (Table 3), while a partial correlation adjusted for age, FM, and relative peak VO₂ was used to quantify the association between LBMI and VF accumulations (Figure 1). 2 2 ANOVA were used to identify the independent contributions of LBMI and VF as well as potential interactions between both the variables on dependent variables of interest (Table 4). Finally, the frequency of the metabolic syndrome among groups was compared using the ²-test (Table 2). A P value of <0.05 was considered statistically significant.

Figure 1.

Partial correlations between the lean BMI and visceral fat levels after adjustment for age, fat mass, and relative peak VO₂. LBM, lean body mass.

Full figure and legend (18K)

Table 2 - Metabolic characteristics of subjects categorized based on higher vs. lower visceral fat (VF) accumulations and higher vs. lower lean BMI (LBMI).

Full table (67K)

Table 3 - Correlations adjusted for age between body composition and metabolic variables of interest.

Full table (46K)

Table 4 - Interaction between visceral fat (VF) levels and lean BMI on variables of the metabolic profile.

Full table (40K)

Results

Data distributions indicate that our cohort of sedentary PM women displayed a broad range of age (49–70 years old), percentage of FM (37.4–54.9% ), and VF levels (84–346 cm²) (data not shown). Overall, our results revealed that peak VO₂ ranged between 11.6 and 25.3 ml O₂/kg/min (data not shown), which is considered "poor" based on the American College of Sports Medicine (39).

Groups comparisons

In order to better understand the impact of VF levels and LBMI on the metabolic profile, subjects were divided into four groups. As designed, no significant difference between higher LBMI/higher VF and lower LBMI/higher VF groups, as well as between higher LBMI/lower VF and lower LBMI/lower VF groups, were observed for VF levels (Table 1). As expected, both groups displaying lower levels of VF were significantly different from those with higher VF levels. Similar results were also observed when comparing groups based on LBMI. Our data also showed that groups were similar for peak VO₂, FM, FMI, abdominal ScF, and MA. Subjects displaying the higher LBMI/higher VF phenotype also had higher values for waist circumference than the other three groups, as well as higher values for total body weight and BMI than the groups with lower LBMI (all P < 0.0001).

Measures of the metabolic profile are presented in Table 2 and Figures 2 and 3. Plasma triglycerides, total cholesterol, LDL-chol, and resting systolic and diastolic blood pressure were similar across groups. However, subjects displaying the higher LBMI/higher VF phenotype presented a deteriorated metabolic profile as shown by higher plasma insulin and glucose levels at fasting state (Table 2), higher insulin levels during the OGTT (Figure 2), lower glucose disposal during the euglycemic/hyperinsulinemic clamp (Figure 3), and higher plasma hsCRP levels. Although significant with the ANOVA, but not with the posteriori comparison, the higher LBMI/higher VF group also had lower HDL-chol values. Finally, both groups with higher VF levels had higher cholesterol/HDL-chol ratios than the higher LBMI/lower VF group.

Figure 2.

Mean values for glucose and insulin responses during the oral glucose tolerance test. LBM, lean body mass; LBMI, lean BMI (LBM (kg)/height (m²)); VF, visceral fat; open square, lower VF/lower LBMI; open circle, higher VF/lower LBMI; closed square, lower VF/higher LBMI; closed circle, higher VF/higher LBMI. ANOVA were performed to compare groups and the Tukey–Kramer test was used for posteriori comparisons. ^*Higher VF/higher LBMI (closed circles) significantly different from lower VF/lower LBMI (open squares) and lower VF/higher LBMI (closed squares) groups; ^**higher VF/higher LBMI (closed circles) significantly different from lower VF/lower LBMI (open square) group; ^†higher VF/higher LBMI (closed circles) significantly different from the three other groups.

Full figure and legend (33K)

Figure 3.

Total and relative glucose disposal in obese postmenopausal women with either low or high VF and low or high LBMI. LBMI, lean BMI (LBM (kg)/height (m²)), VF, visceral fat. ANOVA were performed to compare groups and the Tukey–Kramer test was used for posteriori comparisons. ^*P < 0.005.

Full figure and legend (30K)

The ² analysis revealed a close, but not significant, difference between the four groups for the prevalence of the metabolic syndrome (P = 0.07). However, complementary analyses revealed that both groups displaying higher levels of VF had a higher prevalence of the metabolic syndrome than the low VF groups (39% vs. 20% ; P < 0.05) (data not shown). Similarly, the prevalence of the metabolic syndrome was higher in women with higher LBMI compared to the low LBMI groups (41% vs. 21% ; P < 0.05) (data not shown).

Associations between measures of body composition and the metabolic profile

As shown by the Pearson's correlations, VF levels were positively and significantly associated with LBM, LBMI, total FM, FMI, and abdominal ScF (P values between 0.005 and 0.0001), while LBMI was also significantly correlated with total FM, FMI, and abdominal ScF (P values between 0.05 and 0.0001) (Table 3). Further analyses revealed that VF levels were still significantly correlated with LBMI when controlled for total FM, peak VO₂, and age (r = 0.39, P < 0.0001) (Figure 1).

Pearson's correlations also indicated that all body composition measures were significantly associated with insulin levels at fasting state and at 2 h during the OGTT, as well as with absolute glucose disposal and hsCRP levels (P values between 0.05 and 0.0001) (Table 3). LBMI and FMI were better correlated with hsCRP than absolute measures of LBM (r = 0.21 vs. r = 0.46) and FM (r = 0.31 vs. r = 0.43), respectively. Furthermore, LBMI was correlated with relative glucose disposal (r = - 0.42, P < 0.01) and hsCRP levels (r = 0.50, P < 0.0005) when VF was taken into account (data not shown). Finally, VF was the only body composition measure associated with the cholesterol-to-HDL-chol ratio (r = 0.29, P < 0.005) in our cohort. The association between both variables improved when a partial correlation adjusted for age, total FM, and peak VO₂ was used (r = 0.38, P < 0.0001; data not shown). Overall, similar results were also obtained when the subject displaying a LBMI of 24 kg/m² was excluded from statistical analyses (Figure 1, data not shown).

Independent and interaction effects of VF levels and LBMI on the metabolic profile

Results from the 2 2 ANOVA are presented in Table 4. Only metabolic variables presented in Table 3 were used for analyses. The LBMI was independently associated with plasma hsCRP levels, fasting insulin, fasting glucose, 2-h insulin levels during the OGTT, and relative glucose disposal. VF levels were independently associated with the Chol-to-HDL-chol ratio, plasma hsCRP levels, fasting insulin, 2-h insulin during the OGTT and measures of glucose disposal. Our results showed a significant interaction effect between LBMI and VF levels on insulin response during the OGTT as well as on total and relative glucose disposal.

Discussion

As anticipated, this study confirms the commonly accepted observation that the amount of VF is associated with disturbances in the metabolic profile. However, glucose homeostasis alterations and elevated CRP levels were only observed in women displaying the higher VF/higher LBMI phenotype. Interestingly, except for the total cholesterol-to-HDL-chol ratio, women with the higher VF/lower LBMI phenotype displayed a metabolic profile similar to those in lower VF groups. This result is of great interest considering the fact they had VF levels ranging between 186 and 278 cm² (data not shown), which is higher than the previously proposed thresholds of 130 cm² (40) and 163 cm² (41), over which metabolic alterations are significantly increased in women. Our observation is also very interesting because both groups with higher VF levels presented similar accumulations of visceral adipose tissue. Taken together, this is the first study showing that (i) LBM, as expressed by the LBMI, interacts with VF levels on insulin resistance in obese individuals, and (ii) LBMI is an independent correlate of CRP levels when VF is taken into account.

Previous work by our laboratory as well as those done by others showed that insulin-resistant obese PM women have higher levels of VF and higher amounts of LBM (9,22,23). However, it is unclear why LBM was not associated with metabolic abnormalities in these previous studies when VF was taken into account. Consequently, we can first hypothesize that absolute LBM may not be the appropriate way to express and quantify the contribution of muscle mass to the metabolic profile given its significant association with height, as shown in our cohort. In other words, the use of nonadjusted LBM for height does not allow for an adequate discrimination and comparison of individuals of different sizes and shapes (28,29,30,31), such as those with a muscular phenotype in this study. Our results regarding CRP levels seem to be in agreement with this concept because LBMI was better correlated than LBM (r = 0.46 vs. r = 0.21, respectively). Second, it can be hypothesized that VF levels and LBM are, to some extent, related. Results from this study are also in agreement with this hypothesis. Similar to results from You et al. (23), we found a positive and significant correlation between VF levels and LBMI after adjusting for potential confounding variables such as age, total FM, and fitness levels. Although cross-sectional designs preclude conclusions regarding causal relationships, results from other studies support our observations on the concomitant association between high LBM and android fat accumulations with altered glucose homeostasis (9,22,23), we found a positive and significant).

Interestingly, our results with overweight and obese PM women do not necessarily support the idea that higher LBM is associated with better glucose disposal (17,18,19,20,21). In fact, as shown in the higher VF/higher LBMI group, the combination of higher levels of VF and a muscular phenotype (or high LBMI) appears to be more detrimental to glucose disposal than each component taken separately. Furthermore, our results do not seem to support the notion fully that an increased level of obesity (as expressed by the percentage of FM or total FM) is associated with higher muscle mass due to the extra weight carried daily (42,43,44,45), because both lower VF and higher VF groups had similar amounts of total FM despite significant differences in total LBM. Our observations regarding the latter point are strengthened by studies that reported the existence of a subgroup of obese older individuals with low levels of muscle mass, known as sarcopenic-obese individuals (46,47,48). The hypothesis of higher physical activity levels in high LBMI groups does not seem to be applicable since the four groups in this study were all sedentary and had similar fitness levels. It is also important to note that other physiological variables measured were not helpful in understanding the unfavorable metabolic profile in women with the higher VF/higher LBMI phenotype. We considered intramuscular fat contents (as expressed by MA) (49,50,51), abdominal ScF accumulations (52,53,54), and deep abdominal ScF accumulations (55), as these variables have been shown to be related to variations in insulin sensitivity. However, this study failed to find associations between these components and the metabolic profile in our cohort.

Our observations must also be interpreted with caution since studies have shown that resistance training programs aimed to increase muscle mass also improved insulin sensitivity (17,56,57,58). Our study design does not allow us to give a definitive explanation of the link between LBMI and the metabolic profile because of the cross-sectional approach used as well as the absence of measures of muscle characteristics such as fiber types, capillarity, enzymes. Furthermore, we believe that it is necessary to distinguish results from studies done in sedentary subjects biologically predisposed to have a muscular body type (9,22,23 and this study) from those where increases in muscle mass are induced by a training intervention. Exercise training has been associated with various changes in muscle quality (muscle fiber size and distribution, lipid content, capillary density, oxidative capacity, etc.) and with increased glucose consumption via increases in O₂ consumption (59,60), which is different from measures obtained in individuals with a sedentary lifestyle.

This study has some limitations. First, our study was composed of 103 nondiabetic sedentary overweight and obese PM women. Furthermore, we excluded 29 subjects with FM levels above and below an s.d. of 1.5, based on the average FM calculated in our cohort. Therefore, our findings are limited to this population. Second, the cross-sectional approach used does not allow us to draw any conclusions regarding causal associations between glucose disposal, VF levels, and LBMI. Finally, measures of fiber areas and capillaries, proportion of muscle types, biochemical properties of skeletal muscle and hormonal profile (androgens, growth hormones, etc.) would have given us a better general picture of the factors associated with glucose disposal.

Despite these limitations, this study is strengthened by the study design and a well-characterized cohort. First, we used the best available techniques for the measurement of body composition, body fat distribution, and glucose disposal. Second, we used a 1-month weight stabilization period prior to testing to minimize the impact of body weight fluctuations on the metabolic profile (26). Third, despite the fact that the study group was composed only of sedentary overweight and obese PM women, we had a broad range of age, body composition, and body fat distribution, which improved our ability to identify potential contributors associated with glucose homeostasis. All in all, we consider that the methodology used strengthens our conclusions.

In conclusion, our data support the notion that an excess VF level is associated with alterations in glucose homeostasis and CRP levels in overweight and obese PM women. This study revealed, for the first time, that LBMI is an independent correlate of glucose homeostasis and CRP levels when VF levels are taken into account. We also showed that LBMI is a significant correlate of VF accumulations. Finally, our data indicate that the contribution of higher VF levels to insulin resistance seems to be modulated by LBM because the combination of higher VF levels with a muscular phenotype (or high LBMI), rather than VF alone, is associated with insulin resistance in our cohort of sedentary PM women.

This work was supported by the Fonds de la Recherche en Santé du Québec (M.B. and R.R.L.), the Canadian Institute of Health Research (CIHR) (OHN—63279 and MOP—62976), and the Fondation du Centre Hospitalier Universitaire de Montréal—Starting grant (R.R.-L. #8200). É.D. was supported by the CIHR/Merck-Frosst New Investigator Award, the Canadian Foundation for Innovation New Opportunities Award, and the Early Research Award of Ontario.

Disclosure

The authors declared no conflict of interest.

References

REFERENCES

1. Ferrannini E, Natali A, Bell P et al. IR and hypersecretion in obesity. European Group for the Study of IR (EGIR). J Clin Invest 1997;100:1166–1173. | Article | PubMed | ISI | ChemPort |
2. Klein S, Sheard NF, Pi-Sunyer X et al. Weight management through lifestyle modification for the prevention and management of type 2 diabetes: rationale and strategies. A statement of the American Diabetes Association, the North American Association for the Study of Obesity, and the American Society for Clinical Nutrition. Am J Clin Nutr 2004;80:257–263. | PubMed | ChemPort |
3. Kelley DE, Thaete FL, Troost F, Huwe T, Goodpaster BH. Subdivisions of subcutaneous abdominal adipose tissue and IR. Am J Physiol Endocrinol Metab 2000;278:E941–E948. | PubMed | ISI | ChemPort |
4. Goodpaster BH, Thaete FL, Simoneau JA, Kelley DE. Subcutaneous abdominal fat and thigh muscle composition predict insulin sensitivity independently of visceral fat. Diabetes 1997;46:1579–1585. | Article | PubMed | ISI | ChemPort |
5. Brochu M, Starling RD, Tchernof A et al. Visceral adipose tissue is an independent correlate of glucose disposal in older obese postmenopausal women. J Clin Endocrinol Metab 2000;85:2378–2384. | Article | PubMed | ISI | ChemPort |
6. Després JP. The IR-dyslipidemic syndrome of visceral obesity: effect on patients' risk. Obes Res 1998;6:8S–17S. | PubMed |
7. Ross R, Freeman J, Hudson R, Janssen I. Abdominal obesity, muscle composition, and IR in premenopausal women. J Clin Endocrinol Metab 2000;87:5044–5051. | Article | ChemPort |
8. Krotkiewski M. Role of muscle morphology in the development of IR and metabolic syndrome. Presse Med 1994;23:1393–1399. | PubMed | ChemPort |
9. Krotkiewski M, Björntorp P. Muscle tissue in obesity with different distribution of adipose tissue. Effects of physical training. Int J Obes 1986;10:331–341. | PubMed | ChemPort |
10. Holmang A, Brzezinska Z, Björntorp P. Effects of hyperinsulinemia on muscle fiber composition and capitalization in rats. Diabetes 1993;42:1073–1081. | Article | PubMed | ChemPort |
11. Yki-Jarvinen H, Koivisto VA. Effects of body composition on insulin sensitivity. Diabetes 1983;32:965–969. | Article | PubMed | ChemPort |
12. Marin P, Andersson B, Krotkiewski M, Björntorp P. Muscle fiber composition and capillary density in women and men with NIDDM. Diabetes Care 1994;17:382–386. | Article | PubMed | ChemPort |
13. Hedman A, Berglund L, Essen-Gustavsson B, Reneland R, Lithell H. Relationships between muscle morphology and insulin sensitivity are improved after adjustment for intra-individual variability in 70-year-old men. Acta Physiol Scand 2000;169:125–132. | Article | PubMed | ChemPort |
14. Gravholt CH, Nyholm B, Saltin B, Schmitz O, Christiansen JS. Muscle fiber composition and capillary density in Turner syndrome: evidence of increased muscle fiber size related to IR. Diabetes Care 2001;24:1668–1673. | Article | PubMed | ChemPort |
15. Larsson H, Daugaard JR, Kiens B, Richter EA, Ahren B. Muscle fiber characteristics in postmenopausal women with normal or impaired glucose tolerance. Diabetes Care 1999;22:1330–1338. | Article | PubMed | ChemPort |
16. Karelis AD, Faraj M, Bastard JP et al. The metabolically healthy but obese individual presents a favorable inflammation profile. J Clin Endocrinol Metab 2005;90:4145–4150. | Article | PubMed | ISI | ChemPort |
17. Eriksson JG. Exercise and the treatment of type 2 diabetes mellitus. An update. Sports Med 1999;27:381–391. | Article | PubMed | ChemPort |
18. Nam SY, Kim KR, Cha BS et al. Low-dose growth hormone treatment combined with diet restriction decreases IR by reducing visceral fat and increasing muscle mass in obese type 2 diabetic patients. Int J Obes Relat Metab Disord 2001;25:1101–1107. | Article | PubMed | ChemPort |
19. Koivisto VA, Yki-Jarvinen H, DeFronzo RA. Physical training and insulin sensitivity. Diabetes Metab Rev 1986;1:445–481. | PubMed | ChemPort |
20. Stone MH, Fleck SJ, Triplett NT, Kraemer WJ. Health- and performance-related potential of resistance training. Sports Med 1991;11:210–231. | Article | PubMed | ChemPort |
21. He J, Watkins S, Kelley DE. Skeletal muscle lipid content and oxidative enzyme activity in relation to muscle fiber type in type 2 diabetes and obesity. Diabetes 2001;50:817–823. | Article | PubMed | ISI | ChemPort |
22. Brochu M, Sites CK, Tchernof A et al. What are the physical characteristics associated with a normal metabolic profile despite a high level of obesity in postmenopausal women? J Clin Endocrinol Metab 2001;86:1020–1025. | Article | PubMed | ISI | ChemPort |
23. You T, Ryan AS, Nicklas BJ. The metabolic syndrome in obese postmenopausal women: relationship to body composition, visceral fat, and inflammation. J Clin Endocrinol Metab 2004;89:5517–5522. | Article | PubMed | ISI | ChemPort |
24. Virtanen KA, Iozzo P, Hällsten K et al. Increased fat mass compensates for insulin resistance in abdominal obesity and type 2 diabetes: a positron-emitting tomography study. Diabetes 2005;54:2720–2726. | Article | PubMed | ChemPort |
25. Weinsier GL, Nagy TR, Hunter GR et al. Do adaptive changes in metabolic rate favor weight regain in weight-reduced individuals? An examination of the set-point theory. Am J Clin Nutr 2000;71:1088–1094.
26. Brochu M, Tchernof A, Turner AN, Ades PA, Poehlman ET. Is there a threshold of visceral fat loss that improves the metabolic profile in obese postmenopausal women? Metabolism 2003;52:599–604. | Article | PubMed | ChemPort |
27. Brochu M, Starling RD, Ades PA, Poehlman ET. Are aerobically fit older individuals more physically active in their free-living time? A doubly labeled water approach. J Clin Endocrinol Metab 1999;84:3872–3876. | Article | PubMed | ChemPort |
28. Schutz Y, Kyle UU, Pichard C. Fat-free mass index and fat mass index percentiles in Caucasians aged 18-98 y. Int J Obes Relat Metab Disord 2002;26:953–960. | Article | ISI | ChemPort |
29. Kyle UG, Schutz Y, Dupertuis YM, Pichard C. Body composition interpretation. Contributions of the fat-free mass index and the body fat mass index. Nutrition 2003;19:597–604. | Article | PubMed | ISI |
30. Ramsay SE, Whincup PH, Shaper AG, Wannamethee SG. The relations of body composition and adiposity measures to ill health and physical disability in elderly men. Am J Epidemiol 2006;164:459–469. | Article | PubMed |
31. Kyle UG, Genton L, Gremion G, Slosman DO, Pichard C. Aging, physical activity and height-normalized body composition parameters. Clin Nutr 2004;23:79–88. | Article | PubMed |
32. Demerath EW, Schubert CM, Maynard LM et al. Do changes in body mass index percentile reflect changes in body composition in children? Data from the Fels Longitudinal Study. Pediatrics 2006;117:e487–e495. | Article | PubMed | ISI |
33. Misra A, Garg A, Abate N et al. Relationship of anterior and posterior subcutaneous abdominal fat to insulin sensitivity in nondiabetic men. Obes Res 1997;5:93–99. | PubMed | ChemPort |
34. Simoneau JA, Colberg SR, Thaete FL, Kelley DE. Skeletal muscle glycolytic and oxidative enzyme capacities are determinants of insulin sensitivity and muscle composition in obese women. FASEB J 1995;9:273–278. | PubMed | ISI | ChemPort |
35. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2004;27:S5–S10. | PubMed |
36. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979;237:E214–E223. | PubMed | ISI | ChemPort |
37. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without the use of the preparative ultracentrifuge. Clin Chem 1972;18:499–507. | PubMed | ISI | ChemPort |
38. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation 2002;106:3143–3421.
39. American College of Sports Medicine. ACSM's Resource Manual for Guidelines for Exercise Testing and Prescription, 2nd edn. Lea & Febiger, Philadelphia, 1993.
40. Després JP, Lamarche B. Low-intensity endurance exercise training, plasma lipoproteins and the risk of coronary heart disease. J Intern Med 1994;236:7–22. | PubMed |
41. Nicklas BJ, Penninx BW, Ryan AS et al. Visceral adipose tissue cutoffs associated with metabolic risk factors for coronary heart disease in women. Diabetes Care 2003;26:1413–1420. | Article | PubMed |
42. Forbes GB, Welle SL. Lean body mass in obesity. Int J Obes 1983;7:99–107. | PubMed | ISI | ChemPort |
43. Lafortuna CL, Maffiuletti NA, Agosti F, Sartorio A. Gender variations of body composition, muscle strength and power output in morbid obesity. Int J Obes (Lond) 2005;29:833–841. | Article | PubMed | ChemPort |
44. Womersley J, Durnin JV, Boddy K, Mahaffy M. Influence of muscular development, obesity, and age on the fat-free mass of adults. J Appl Physiol 1976;41:223–229. | PubMed | ChemPort |
45. Kelley DE, Slasky BS, Janosky J. Skeletal muscle density: effects of obesity and non-insulin-dependent diabetes mellitus. Am J Clin Nutr 1991;54:509–515. | PubMed | ISI | ChemPort |
46. Roubenoff R. Sarcopenic obesity: the confluence of two epidemics. Obes Res 2004;12:887–888. | PubMed | ISI |
47. Baumgartner RN, Wayne SJ, Waters DL et al. Sarcopenic obesity predicts instrumental activities of daily living disability in the elderly. Obes Res 2004;12:1995–2004. | PubMed |
48. Cesari M, Kritchevsky SB, Baumgartner RN et al. Sarcopenia, obesity, and inflammation—results from the Trial of Angiotensin Converting Enzyme Inhibition and Novel Cardiovascular Risk Factors study. Am J Clin Nutr 2005;82:428–434. | PubMed | ChemPort |
49. Goodpaster BH, Thaete FL, Simoneau JA, Kelley DE. Subcutaneous abdominal fat and thigh muscle composition predict insulin sensitivity independently of visceral fat. Diabetes 1997;46:1579–1585. | Article | PubMed | ISI | ChemPort |
50. Komiya H, Mori Y, Yokose T et al. Effect of intramuscular fat difference on glucose and insulin reaction in oral glucose tolerance test. J Atheroscler Thromb 2006;13:136–142. | PubMed | ChemPort |
51. Torriani M, Hadigan C, Jensen ME, Grinspoon S. Psoas muscle attenuation measurement with computed tomography indicates intramuscular fat accumulation in patients with the HIV-lipodystrophy syndrome. J Appl Physiol 2003;95:1005–1010. | PubMed |
52. Abate N, Garg A, Peshock RM, Stray-Gunderson J, Grundy SM. Relationship of generalized and original adiposity to insulin sensitivity in men. J Clin Invest 1995;96:88–98. | Article | PubMed | ISI | ChemPort |
53. Abate N, Garg A, Peshock RM et al. Relationship of generalized and original adiposity to insulin sensitivity in men with NIDDM. Diabetes 1996;45:1684–1693. | Article | PubMed | ChemPort |
54. Goodpaster BH, Thaete FL, Simoneau JA, Kelley DE. Subcutaneous abdominal fat and thigh muscle composition predict insulin sensitivity independently of visceral fat. Diabetes 1997;46:1579–1585. | Article | PubMed | ISI | ChemPort |
55. Kelley DE, Thaete FL, Troost F, Huwe T, Goodpaster BH. Subdivision of subcutaneous abdominal adipose tissue and insulin resistance. Am J Pysiol Endocrinol Metab 2000;278:E941–E948. | ChemPort |
56. Andersen JL, Schjerling P, Andersen LL, Dela F. Resistance training and insulin action in humans: effects of de-training. J Physiol 2003;551:1049–1058. | Article | PubMed | ChemPort |
57. Dionne IJ, Mélançon MO, Brochu M, Ades PA, Poelhman ET. Age-related differences in metabolic adaptations following resistance training in women. Exp Gerontol 2004;39:133–138. | Article | PubMed | ChemPort |
58. Ishii T, Yamakita T, Sato T, Tanaka S, Fujii S. Resistance training improves insulin sensitivity in NIDDM subjects without altering maximal oxygen uptake. Diabetes Care 1998;21:1353–1355. | Article | PubMed | ChemPort |
59. Henriksen EJ. Invited review: effects of acute exercise and exercise training on IR. J Appl Physiol 2002;93:788–796. | PubMed | ISI | ChemPort |
60. Ivy JL. Role of exercise training in the prevention and treatment of IR and non-insulin-dependent diabetes mellitus. Sports Med 1997;24:321–336. | PubMed | ChemPort |