Upper-body obese women are resistant to postprandial stimulation of protein synthesis

Abstract

Background and Aims

Upper-body, i.e. visceral, obesity is associated with insulin resistance and impaired protein synthesis. It is unclear whether postprandial stimulation of protein synthesis is affected by body fat distribution. We investigated the postprandial protein anabolic response in a cohort of obese women.

Methods

Participants were studied after an overnight fast and after a mixed meal, grouped as upper-body obese (UBO, waist-to-hip ratio, WHR, >0.85, n=6) vs. lower-body obese (LBO, WHR<0.80, n=7). Lipid and carbohydrate metabolism were assessed by measurements of plasma free fatty acids (FFA), insulin and glucose concentrations, and calculation of the Quicki index from fasting glucose and insulin values. Different labels of stable isotopes of phenylalanine were administered intravenously and orally, and leg and whole-body protein breakdown and synthesis were calculated from phenylalanine/tyrosine isotopic enrichments in femoral arterial and venous blood, using equations for steady-state kinetics. Data are denoted as mean±SD.

Results

Age (38 vs. 40, p=0.549) and body-mass index (33.7±1.9 vs. 35.0±1.8, p=0.241) were similar in both groups. UBO subjects had more visceral fat (p=0.002) and higher fat-free body mass (FFM) (p=0.015). Plasma insulin concentrations were greater in UBO than LBO women (p=0.013), and UBO were less insulin sensitive (Quicki=0.32±0.01 vs. 0.36±0.02, p=0.005). Protein kinetics across the leg were not different between groups. Fasting whole body protein balance was similarly negative in both groups (UBO −6.5±2.4 vs. LBO −7.6±0.9 μmol/kgFFM/h, p=1.0). Postprandially, whole body protein balance became less positive in UBO than in LBO (14.8±3.7 vs. 20.2±3.7 μmol/kgFFM/h, p=0.017).

Conclusions

Whole-body protein balance following a meal is less positive in upper-body obese, insulin-resistant, women than in lower-body obese women.

Introduction

While mammal carbohydrate and lipid metabolism involves specialised molecules (glycogen, triacylglycerol) and tissues (adipose tissue) to store substrates and energy, no such storage exists for amino acids. Instead, tissues and organs undergo a continuous recycling of their functional and structural proteins, a process in which amino acids are exchanged within the cells through intracellular recycling and between cells via the circulation. To match the availability of amino acids with the requirements of the organism as a whole, as well as of the individual tissues and organs, a fine-tuned equilibrium of protein breakdown and synthesis must be maintained. Shifts in this equilibrium have been described in a variety of physiologic and pathologic conditions, e.g. growth, exercise, fasting and weight loss, aging, and cachexia. Some primary metabolic conditions such as glucose intolerance and type 2 diabetes have also been characterised in terms of protein metabolism. However, the protein metabolism of the most prevalent metabolic disorder, obesity, remains incompletely understood.

Alterations of protein metabolism in obesity are a subject of controversy (reviewed in (16)). Muscle protein metabolism has been studied in rodent models of obesity where decreased muscle protein content and protein synthesis were found (2, 9). In humans, leucine release from muscle was found to be lower in fasting obese women than in lean women (23). Whole-body protein turnover of obese subjects has been compared with lean subjects under either postabsorptive conditions or insulin stimulation and has been reported as either increased (14) or unchanged (6, 21). In one study of obese vs. non-obese males, postabsorptive protein metabolism was compared under hyperinsulinemic, hyperaminoacidemic, euglycemic conditions (11), where a reduced turnover of skeletal muscle protein, an impaired stimulatory effect of insulin and amino acid on mitochondrial protein synthesis, and an impaired effect of insulin and amino acids on whole-body protein breakdown were found in obese subjects. In obese females under hyperinsulinemic conditions, with amino acid plasma concentrations clamped at postabsorptive concentrations, a blunted whole-body anabolic response to insulin was observed and was ascribed to insulin resistance of protein synthesis (7). Finally, upper-body (viscerally) obese women are particularly prone to insulin resistance of whole body protein metabolism in euglycemic, hyperinsulinemic clamp studies (14). We therefore wanted to study the anabolic response to a more physiologic stimulus, in the form of feeding, in obese women.

To investigate the effect of body fat distribution on muscle and whole-body protein metabolism, we studied upper-body obese vs. lower-body obese women after an overnight fast and after a mixed meal. Since upper-body obese women are known to be more insulin resistant than lower-body obese, this allows studying the effects of insulin resistance independent of total adiposity. Fatty acid and glucose metabolism were characterised by measurements of plasma glucose, insulin and FFA, and steady-state protein kinetics were determined using different phenylalanine tracers by the intravenous and oral route.

Methods

Subjects and body composition

A subgroup of the women recruited for another study (12) was investigated for protein metabolism in simultaneous experiments and is described in this report. The study was approved by the institutional review board and subjects were enrolled after their informed consent had been obtained. To be eligible, subjects were required to be non-smokers, premenopausal, and not acutely ill, to have a BMI between 30 and 37 and a non-diabetic fasting blood glucose value, and not take medications know to affect protein or fatty acid metabolism. Body fat, fat-free mass and leg fat mass were measured by dual energy x-ray absorptiometry as previously described (15) and visceral and abdominal subcutaneous fat area by single-slice computed tomography. Subjects were recruited into groups of upper-body obesity (UBO) with a waist-to-hip ratio (WHR) greater than 0.85 or lower-body obesity (LBO) with WHR <0.80.

Experimental protocol

Starting 10 days before the study, all subjects were given a standardised diet at the clinical research center, providing 20% protein, 40% fat and 40% carbohydrate by energy content and adjusted in dose to maintain a stable body weight. Subjects were admitted to the research facility the evening before the study and a 18 gauge indwelling catheter was placed in a forearm vein and infused with 0.45% NaCl at 20 ml/h. After an overnight fast, resting energy expenditure was measured by indirect calorimetry (DeltaTrac Metabolic Cart, Sensormedics, Yorba Linda, California, USA). Infusion and sampling catheters were placed in the right femoral artery and right femoral vein. Leg plasma flow was measured by indocyanine green clearance as described (12). A primed continuous intravenous infusion of either L-(ring²H₅)-phenylalanine or L-¹⁵N-phenylalanine was begun after baseline blood samples and continued throughout the 8 h study period. The primer bolus consisted of 0.75 mg × kg FFM ⁻¹ phenylalanine and the continuous infusion rate was set at 0.75 mg × kg FFM ⁻¹ × h ⁻¹ phenylalanine. Experimental meals were administered after a 3 h run-in period of intravenous tracer infusion. Meals consisted of a liquid feeding formula (Ensure Plus, Ross Products Division, Abbott Laboratories, Columbus, Ohio, United States) containing 53% carbohydrate, 32% fat and 15% protein by energy content and were dispensed individually to provide a total of 40% of measured resting daily energy expenditure. Meal aliquots were given at 20 min intervals over 5 h with priming (i.e. doubled) doses for the first hour, followed by equal portions for the remainder of the study period. The formulation was combined with either L-(ring²H₅)-phenylalanine or L-¹⁵N-phenylalanine in a dose resulting in an hourly intake of 1.5 mg phenylalanine × kg FFM ⁻¹ during the one hour priming period and 0.75 mg phenylalanine × kg FFM ⁻¹ during steady-state feeding. Arterial and venous blood samples were collected at baseline, at 15 minute intervals during the last 45 minutes of the 3 h run-in period, and again at 15 minute intervals during the last 45 minutes of the feeding period. Half of the UBO and half of LBO subjects received the L-(ring²H₅)-phenylalanine intravenously and the L-¹⁵N-phenylalanine orally whereas the other half received the reverse. Because no differences were observed between the two approaches, the results were pooled.

Materials, samples and analytic techniques

L-(ring²H₅)-phenylalanine and L-¹⁵N-phenylalanine were obtained from Cambridge Isotope Laboratories (Andover, Massachusetts, Unites States), were tested for chemical and isotopic purity, prepared as aqueous solutions and tested for sterility and non-pyrogenicity. Indocyanine green (Cardio-Green, Becton-Dickinson, Cockeysville, Maryland, United States) was used for measurement of leg plasma flow. Arterial and venous blood samples were collected from catheters in the right femoral artery and vein. Blood plasma was immediately separated by centrifugation and stored at −80°C until analysis. Plasma free fatty acid concentrations were determined by HPLC using [²H₃₁]palmitate as internal standard, plasma glucose concentration with a glucose analyzer (Beckman Instruments, Fullerton, California, United States), and plasma insulin concentrations by radioimmunoassay, as described previously (12). Plasma amino acid concentrations were determined from arterial blood as described (20) using a HPLC system (HP 1090 series 2 HPLC with 1046 fluorescence detector) with precolumn O-phthalaldehyde derivatization. ²H5-phenylalanine, ¹⁵N-phenylalanine, ²H4-tyrosine and ¹⁵N-tyrosine were measured as t-butyldimethylsilyl ether derivatives by gas chromatography-mass spectrometry as described (20).

Calculations

Isotopic equilibrium of the tracers was confirmed by regression analysis (slope not significantly different from 0). Mean values of isotopic enrichment were used to calculate steady-state amino acid kinetics. All whole body kinetics are reported as per kg FFM.

To calculate whole body phenylalanine kinetics in the steady state, the model described in (25) was applied. Whole body flux for phenylalanine was calculated by using the following equation:

$${\mathrm{Q}}_{\mathrm{p}}=\mathrm{I}(({\mathrm{E}}_{\mathrm{i}}/{\mathrm{E}}_{\mathrm{A}})-1),$$

in which I is the infusion rate of the tracer, E_i the enrichment in the infusate and E_A the enrichment of the tracer in arterial plasma at isotopic equilibrium given as APE. Whole body conversion of phenylalanine to tyrosine (Q_PT) was calculated as follows:

$${\mathrm{Q}}_{\text{PT}}=(0.73\times {\mathrm{Q}}_{\mathrm{P}})\times {\mathrm{Q}}_{\mathrm{P}}/(({\text{EPhe}}_{\mathrm{A}}/{\text{ETyr}}_{\mathrm{A}}-1)\times ({\mathrm{I}}_{\mathrm{P}}+{\mathrm{Q}}_{\mathrm{P}})),$$

in which ETyr_A and EPhe_A are isotopic enrichment of tyrosine and phenylalanine respectively ([²H₄]Tyrosine is used when [²H₅]Phenylalanine was infused and [¹⁴N]Tyrosine when [¹⁴N]phenylalanine was infused) and I_P the infusion rate of phenylalanine. Whole body incorporation of phenylalanine into protein was calculated by subtracting Q_PT from Q_P and is also referred to as rate of disappearance (Rd), whereas protein breakdown is represented as the Q_P and is also referred to as rate of appearance (Ra). Protein balance was calculated by subtracting Ra from Rd.

To obtain the endogenous rate of appearance, which represents whole body protein breakdown, endogenous Ra was assumed to equal Q_p in the fasting state, and calculated as

$$\text{Endo}\text{Ra}\text{Phe}={\mathrm{Q}}_{\mathrm{p}}-\text{dietary}\text{Ra}$$

in the postprandial state. Splanchnic metabolism of phenylalanine must be accounted for as a certain amount of the dietary phenylalanine will be utilised in the splanchnic area and not appear in the central plasma pool. To correct for this, another phenylalanine tracer was added to the food to measure this splanchnic extraction. Splanchnic metabolism of phenylalanine was calculated as described earlier (3). In short, first the rate of appearance of the dietary phenylalanine tracer is calculated:

$${\text{Ra}}_{(\text{dietary}\text{Phe}\text{tracer})}={\mathrm{I}}_{\text{iv}}/({\text{EPhe}}_{\mathrm{A},\text{iv}}/{\text{EPhe}}_{\mathrm{A},\text{or}}),$$

in which EPhe_A,iv and EPhe_A,or are the arterial enrichments of the intravenous and dietary administered phenylalanine tracers respectively. This rate of appearance of the dietary tracer is then used to calculate the rate of appearance of dietary unlabeled phenylalanine:

$$\text{Dietary}\text{Ra}\text{Phe}=\text{Ra}(\text{dietary}\text{Phe}\text{tracer})/{\text{EPhe}}_{\text{or}},$$

in which EPhe_or is the enrichment of the tracer in the food. The splanchnic uptake is then the total dietary intake of phenylalanine minus the dietary Ra Phe.

To calculate regional muscle amino acid kinetics, equations described previously were used (26). The rate of appearance of phenylalanine in leg tissue (R_a Phe; representing protein breakdown rates) was calculated as follows:

$${\mathrm{R}}_{\mathrm{a}}\text{Phe}=({\text{EPhe}}_{\mathrm{A}}/{\text{EPhe}}_{\mathrm{V}}-1)\times {\text{Phe}}_{\mathrm{A}}\times F,$$

in which EPhe_A and EPhe_V represent phenylalanine enrichments in the artery and the femoral vein respectively, Phe_A is the arterial phenylalanine concentration and F is plasma flow. The rate of disappearance of phenylalanine (R_d Phe) can be calculated by adding the rate of appearance and the net balance of phenylalanine:

$${\mathrm{R}}_{\mathrm{d}}\text{Phe}={\mathrm{R}}_{\mathrm{a}}\text{Phe}+({\text{Phe}}_{\mathrm{A}}-{\text{Phe}}_{\mathrm{V}})\times F$$

Since phenylalanine is not metabolized in muscle tissue the rate of disappearance represents protein synthesis in the leg.

The Quicki index, a measure of insulin sensitivity (17), was calculated using the formula:

$$\text{Quicki}=1/({log}_{10}\text{FPI}+{log}_{10}\text{FPG}),$$

where FPI is the fasting plasma insulin concentration expressed in microunits/mL and FPG is the fasting plasma glucose concentration expressed in mg/dL (Note The formula originally (17) refers to whole blood rather than plasma glucose).

Statistical analysis

Data were analysed using Statistica, version 10 (StatSoft Scandinavia AB, Uppsala, Sweden). Within-group comparisons between fasting and postprandial timepoints and between-group comparisons were analysed by repeated-measures ANOVA, and where appropriate with post-hoc analysis of ANOVA results using a Bonferroni correction. Anthropometric data and between-group comparisons for fasting-postprandial change were analysed by t-test for independent samples. Between-group comparisons for Quicki index were analysed by Mann-Whitney U test, as the method of calculating Quicki index (addition of values derived from measurements in differing units) yields ordinal rather than strictly quantitative values. Values are presented as mean ± standard deviation (SD) unless indicated otherwise.

Results

Subjects

Fasting and postprandial plasma glucose values of 12.3 and 17.0 mmol/L respectively were found in one UBO participant despite non-diabetic fasting blood glucose on the screening evaluation. Because this was consistent with a type 2 diabetes metabolic profile, we did not include data from this volunteer in the analysis, leaving n=6 in the UBO group and n=7 in the LBO group for the final analysis. The subjects’ anthropometric data are provided in Table 1. Groups were well matched for age, weight, BMI, amount of abdominal subcutaneous fat, leg fat-free mass and leg fat mass. The UBO women had higher fat-free body mass, more visceral fat, and a lower fraction of body fat than the LBO women.

Table 1.

Anthropometric data of lower-body obese vs. upper-body obese women fed a mixed meal

	LBO (n=7)	UBO (n=6)	p
Age [yrs]	38±5	40±6	0.549
Weight [kg]	90.5±9.0	95.2±8.5	0.356
Body FFM [kg]	43.6±3.1	49.9±4.8	0.015
BMI [kg/m2]	33.7±1.9	35.0±1.8	0.241
Body fat [%]	52.0±2.8	46.3±4.8	0.021
Visceral fat [cm2]	108±22	181±45	0.002
Abdominal subcutaneous fat [cm2]	296±85	333±55	0.378
Leg fat mass [kg] (n=5 in UBO)	16.7±4.5	13.8±2.0	0.212
Leg FFM [kg] (n=5 in UBO)	14.7±1.2	15.5±2.1	0.453
WHR	0.76±0.02	0.90±0.07

Data denote mean±SD. P values by T-test for independent samples. P not indicated for WHR which is the grouping variable. Leg fat mass and leg fat free mass: data missing for one subject in UBO

BMI, body-mass index; FFM, fat-free mass; LBO, lower-body-obese; UBO, upper-body obese; WHR, waist-to-hip-ratio.

Response to feeding

Results for metabolic changes in response to feeding are shown in Table 2 and Figure 1. Regarding muscle protein metabolism, the initially negative leg NB Phe returned to zero. At the whole-body level, endogenous R_a Phe decreased, Q_PT was unchanged, R_d Phe increased, and protein balance switched from negative to positive in both groups. Regarding metabolites and insulin, plasma concentrations of phenylalanine and leucine increased postprandially, as did plasma glucose. Plasma insulin increased postprandially by more than fivefold, and plasma FFA decreased significantly.

Table 2.

Protein kinetics, plasma metabolites, and insulin sensitivity in lower-body obese vs. upper-body obese women fed a mixed meal

	Fasting			Postprandial			p (ANOVA)
	LBO	UBO	p between groups	LBO	UBO	p between groups	group	timepoint	group × timepoint
Leg protein metabolism
Leg Ra Phe [μmol/kg leg FFM/min]	0.31±0.13	0.26±0.05		0.24±0.08	0.20±0.06		0.301	0.059	0.796
Leg Rd Phe [μmol/kg leg FFM/min]	0.21±0.11	0.20±0.04		0.24±0.08	0.20±0.08		0.640	0.569	0.655
Leg NB Phe [μmol/kg leg FFM/min]	−0.11±0.04	−0.06±0.04		−0.00±0.05	0.00±0.03		0.156	<0.001	0.240
Whole body protein metabolism
Endo Ra Phe [μmol/kg FFM/h]	61.6±11.3	52.2±8.2		46.8±12.1	41.8±7.5		0.219	<0.001	0.059
QPT [μmol/kg FFM/h]	7.6±0.9	6.5±2.4		9.5±2.4	7.7±1.3		0.124	0.012	0.486
Rd Phe [μmol/kg FFM/h]	54.0±11.2	45.7±6.1		67.0±12.7	56.5±9.8		0.130	<0.001	0.291
NB [μmol/kg FFM/h]	−7.6±0.9	−6.5±2.4	1.000	20.2±3.7	14.8±3.7	0.017	0.077	<0.001	0.018
Metabolites & insulin sensitivity
Plasma Leu [μmol/L]	104.0±26.5	103.4±15.2		119.4±24.5	125.1±24.0		0.834	0.003	0.542
Plasma Phe [μmol/L]	48.6±6.0	49.4±2.0		66.9±7.1	64.0±3.9		0.718	<0.001	0.060
Plasma glucose [mmol/L]	5.1±0.3	5.3±0.3		6.7±0.4	6.5±0.6		0.940	<0.001	0.193
Plasma insulin [pmol/L]	46±13	95±26		290±107	493±171		0.013	<0.001	0.052
Relative change plasma insulin [%]				529±146	436±152	0.290
Quicki index	0.36±0.02	0.32±0.01	0.005
Plasma FFA [μmol/L]	629±92	777±103		74±36	252±116		0.003	<0.001	0.605
Relative change plasma FFA [%]				−88±5	−68±13	0.002

Data denote mean±SD. LBO n=7, UBO n=6, UBO n=5 for leg data. P values by repeated measures ANOVA; for between-group comparisons by post-hoc analysis of ANOVA results with Bonferroni correction; for relative change in plasma insulin and plasma FFA by t-test for independent samples; for Quicki index by Mann-Whitney U test.

Endo R_a, endogenous rate of appearance; FFA, free fatty acids; FFM, fat-free mass; LBO, lower-body obese; Leu, leucine;Phe, phenylalanine; Q_PT, rate of conversion of phenylalanine to tyrosine; R_a, rate of appearance; R_d, rate of disappearance; UBO, upper-body obese.

Figure 1.

Whole-body protein kinetics in upper-body-obese and lower-body obese women before/after a mixed meal. * denotes p <0.05 for between-groups comparison

Upper-body vs. lower-body obese

Results of the group comparison are shown in Table 2. There were no between-group differences in plasma glucose; in the UBO group plasma insulin concentration was higher and Quicki index was lower, indicating lower insulin sensitivity in the fasting state. No group differences were seen in muscle protein kinetics. On the whole-body level, protein balance in the fasting state was similarly negative in both groups, but postprandial protein balance was less positive in the UBO group. Plasma FFA was higher and the postprandial drop in FFA was smaller in the UBO group.

Discussion

We studied the effect of feeding on the protein metabolism of obese women, calculating leg and whole-body protein kinetics using intravenous and oral phenylalanine tracers. We found a postprandial suppression of whole body protein breakdown and stimulation of protein synthesis resulting in positive protein balance. Upper-body obese individuals had a less positive postprandial whole-body protein balance than lower-body obese.

Our study participants were grouped into LBO (WHR<0.80) and UBO (WHR>0.85) and several findings confirm that the criterion of WHR >0.85 is appropriate to select a group of viscerally obese (13), insulin resistant individuals. At lower FFM and body fat by percentage, UBO subjects had approximately 1.7 times more visceral fat than LBO subjects. No group difference was seen in fasting or postprandial plasma glucose concentrations, but the UBO group had higher plasma insulin concentrations and a significantly lower Quicki index, indicating lower insulin sensitivity.

In response to a mixed meal, we observed slight increases in plasma glucose concentrations, moderate increases of plasma phenylalanine and leucine concentrations and a more than fivefold increase of plasma insulin. With respect to muscle protein metabolism, a change from net negative balance to zero balance was found. On the whole-body level, protein balance shifted from negative to positive, Endo R_a Phe decreased, and R_d Phe increased. These postprandial changes correspond to a physiologic response to food intake, representing stimulated insulin release and a reversal of negative protein balance resulting from a decline in protein catabolism and enhancement of anabolism. Postprandial protein metabolism in obesity has not been studied in much detail. Using whole body leucine kinetics in grossly obese (mean BMI=45.1) women, whole body leucine flux, oxidation and protein synthesis normalised to FFM have been found to be unchanged postprandially (6); types of body fat distribution were not reported. The response to feeding has been also studied in a murine model of diet-induced obesity (1), where the stimulation of skeletal muscle protein synthesis in response to enteral feeding was found to be impaired, while basal protein synthesis was not affected.

There were no between-group differences of whole-body protein kinetics in the fasting state, but postprandial protein balance was less positive in the UBO group. Our data do not allow us to locate the anabolic defect, as no between-group differences in leg protein kinetics were observed in the small sample available. Possible mechanisms can be conjectured based on the available knowledge on the impact of amino acids, insulin and FFA on protein metabolism. Alterations of branched-chain amino acid plasma concentrations may affect protein metabolism. BCAA, specifically leucine, stimulate protein synthesis and inhibit protein degradation (19), but may also induce insulin resistance (28), and elevated plasma BCAA concentrations in obesity correlate with insulin resistance (22). Our data are limited to measurements of leucine plasma concentrations, in which no between-groups differences were found, which prima facie argues against an implication of altered BCAA signalling in the anabolic defect. However, interactions between BCAA, insulin and other hormones and metabolites are complex and a comprehensive assessment would require more extensive metabolic profiling than was available in our study.

Beyond the small sample size, one possible explanation for the lack of between-groups differences of skeletal muscle protein metabolism may lie in the time schedule of measurements. Protein synthesis rates may be responsive to changes of blood AA concentrations rather than absolute values, and have in skeletal muscle been shown to return to baseline within two hours after initiation of a primed infusion of mixed AA (5). The protein component in the feeding formula consists of milk protein, known to undergo rapid digestion and absorption. For example, an oral bolus of casein results in a rapid rise of plasma AA and insulin concentrations, peaking within less than 60 minutes postprandially (18). In our protocol, feeding of repeated small boluses during five hours was designed to yield quasi-steady state plasma AA concentrations, and by taking measurements at the end of this period we may have missed the peak effect on skeletal muscle protein synthesis.

Another conceivable mechanism for the observed between-group differences of whole-body protein metabolism may be related to an alteration of insulin action. Insulin has anti-proteolytic effects on the whole-body level, on splanchnic, and on skeletal muscle protein breakdown (24), as well as protein synthesis stimulating properties. Insulin-mediated vascular effects are necessary for the protein anabolic effect of insulin (27) and these mechanisms are defective in the skeletal muscle of obese subjects (8), which might affect their regional protein metabolism. We found no between-group difference in fasting or postprandial muscle protein metabolism, despite clear evidence of different insulin sensitivity. This may represent a compensatory effect of hyperinsulinemia. Plasma insulin was higher in UBO than in LBO, but plasma glucose concentrations was similar in both groups, indicating that hyperinsulinemia was adequate to maintain glucose homeostasis in the presence of insulin resistance. While dietary protein in a mixed meal does not independently stimulate insulin release (29), insulin-dependent protein anabolism may be stimulated as a collateral effect of plasma insulin concentrations reacting to carbohydrate content in a mixed meal. Thus in our UBO subjects, compensatory hyperinsulinemia may have been sufficient to outweigh a defect of insulin-dependent protein anabolism in skeletal muscle.

Finally, an inhibitory effect of FFA on protein anabolism may contribute to the lower protein balance observed in UBO subjects. FFA modulate protein synthesis and breakdown in skeletal muscle as well as on the whole-body level (10). Elevated plasma FFA concentrations are markers of pathological glucose tolerance (30) and may be causally linked with insulin resistance in obesity (4). We found elevated plasma FFA concentrations and a smaller postprandial drop in the UBO group (Table 2). These findings could suggest that an insufficient postprandial suppression of lipolysis in upper-body obesity may contribute to impaired postprandial protein anabolism.

We recognize limitations to the interpretation of our findings. The small sample size, particularly with respect to muscle protein kinetics in the UBO group, and use of conservative statistical tests are likely to have affected the power to detect within-group or between-group differences. Furthermore, while upper-body obese women are a metabolically well-defined insulin-resistant population, it is unclear whether our findings can be generalized to other populations with similar conditions.

To summarise, we found that the postprandial stimulation of whole body protein anabolism was less robust in insulin-resistant, upper-body obese women than in lower-body obese women. The protein anabolic defect in upper-body obese women was seen at plasma FFA and insulin concentrations that were higher than in lower-body obese women, while plasma leucine concentrations were identical. This pattern is suggestive of a state of anabolic resistance that may be mediated by effects of FFA.

Abbreviations

AA

amino acid(s)

APE

atom percent excess

BCAA

branched-chain amino acid(s)

BMI

body mass index

FFA

free fatty acids

FFM

fat-free mass

HPLC

high-performance liquid chromatography

LBO

lower-body obese

NB

net balance

QPT

whole-body conversion of phenylalanine to tyrosine

Ra

rate of appearance

Rd

rate of disappearance

UBO

upper-body obese

WHR

waist-to-hip ratio