Intra-abdominal
Fat Is Associated KS.
With Decreased Young Men
Insulin
Sensitivity
in Healthy
Park, B.D. Rhee, K-U. Lee, S.Y. Kim, H.K. Lee, C.-S. Koh, and H.K. Min
To distinguish the relative role of intra-abdominal and subcutaneous abdominal fat in metabolic aberrations in upper body fat localization, we measured the relationship between regional fat distribution and insulin sensitivity in nine young men (28.8 ? 0.7 years; body mass index [BMI], 24.7 + 1.3 kg/m’). Regional fat distribution was measured by anthropometric measurement and computed tomography (CT). Insulin sensitivity was measured by euglycemic hyperinsulinemic glucose clamp. Insulin sensitivity, expressed as the ratio of rate of glucose utilization to the mean plasma insulin concentration during the second hour of glucose clamp (M/I) was negatively correlated with BMI (r = -.81, P < .OOl), waist to hip girth ratio (WHR) (r = -.80, P < .Ol), subcutaneous abdominal fat area (r = -.80, P < .OOl), and intra-abdominal fat area (r = -.88, P < .Ol). Stepwise forward regression analysis showed that in addition to BMI, intra-abdominal fat area was a significant correlate of decrease in insulin sensitivity. These findings suggest that intra-abdominal fat play an important role in decreasing insulin sensitivity, even in healthy young men. Copyright 0 199 1 by W. B. Saunders Company
R
EGIONAL FAT DISTRIBUTION has been shown to be associated with metabolic abnormalities and disease states, beyond what is explainable by general adiposity.‘.’ Thus, distribution of excessive fat on the upper part of body or abdomen, as measured by the waist to hip girth ratio (WHR), is associated with reduced glucose tolerance, insulin resistance, and the risk for diabetes.‘~’ Although the precise mechanism of these metabolic aberrations in upper body obesity is not clear, several studies”-” suggested that an increased production of free fatty acids (FFA) from the enlarged abdominal adipose tissues might be of importance. Abdominal adipose tissues comprise subcutaneous abdominal fat and intra-abdominal fat, which are different in size, metabolic activity, and drainage to circulation. Thus, it would be of interest to distinguish the relative role of intra-abdominal and subcutaneous abdominal fat in metabolic aberrations in upper body obesity. Computed tomography (CT) provides a convenient means to measure the amount of subcutaneous and intraabdominal fat.‘.’ Several studies’~’ reported that intraabdominal fat determined by CT was associated with insulin resistance on oral glucose tolerance test (OGTT) in middle- or old-aged subjects. The aim of this study was to determine whether this association is observable in healthy young men. We used the euglycemic glucose clamp technique to assess insulin sensitivity more accurately. MATERIALS
AND METHODS
Subjects
Nine young, healthy male medical doctors (28.6 + 0.7 years; body
mass
index
[BMI],
24.7 2 1.3 kg/m*) volunteered
for the
study. All participants were not taking any medication, nor had any medical problems known to affect glucose metabolism. They had no family history of diabetes mellitus. The experimental protocol was approved by the Human Experimental Committee of the Department of Internal Medicine at the Seoul National University Hospital. South Korea. Characteristics of the subjects are shown in Table 1.
Body Fat Distribution Measurements for regional fat distribution included WHR and subcutaneous abdominal and intra-abdominal fat areas. Waist circumference was measured at umbilicus level and hip circumference was measured at maximal hip circumference. WHR was calculated from these measurements. Intra-abdominal and subcutaneous abdominal fat areas were measured in each subject with CT at umbilicus level on supine position using a General Electric CT scanner (Milwaukee, WI). Fat areas were estimated by a procedure described by Tokunaga et al.”
Euglycernic Glucose Clamp Studies Euglycemic subject after scribed.14
glucose clamp studies were performed in each a 12- to 14-hour overnight fast as previously de-
Primed (25 PCi) and continuous (0.25 p,Ci/min) infusion of tritiated glucose ([3--‘HI glucose, Amersham. Buckinghamshire. England; specific activity, 5.0 Ciimmol) was maintained throughout the studies. After a 2-hour period to attain an isotopic steady-state, the subjects received a primed continuous insulin infusion (Velosulin, Nordisk, Coppenhagen. Denmark) at 1.0 mU/kglmin for 2 hours by an IMED pump (San Diego, CA) in conjunction with the infusion of somatostatin (300 kg/h; Bachem, Torrance, CA). Concurrent with insulin infusion, blood glucose levels were clamped at 4.44 mmol/L by determining the blood glucose concentrations every 5 minutes and periodically adjusting the rate of administration of 20% glucose solution based on negative feedback princip1e.l’
Isotopic Determination of Glucose Turnover From the Department of Internal Medicine, College of Medicine. Seoul National University, Inje University, and University of Ulsan, Seoul, Korea. Address reprint requests to H.K. Min, MD, Department of Internal Medicine, Seoul National Universiry Hospital, #28 Yunkun-Dong, Chongno-ku, Seoul 110-744, Korea. Copyright 0 1991 by W.B. Saunders Company 0026-049519114006-0008$03.00/O 600
Following the administration of insulin, non-steady-state rates of glucose turnover were determined using the equation of Steele” in the derivative form for non-steady-state condition. Since endogenous glucose production determined by subtracting the amount of exogenously infused glucose from the isotopically determined total glucose appearance was negative in all subjects during the second hour clamp, peripheral glucose utilization was determined from the infusion rate of glucose during the second hour of clamp. Metabolism,Vol40,No
6 (June), 1991:~~600-603
INTRA-ABDOMINAL
FAT AND INSULIN SENSITIVITY
Table 1. Clinical Characteristics
of Subjects (n = 9) Mean k
Age (vr)
601
SEM
28.6 + 0.7
Range 24-32
Weight (kg)
73.7 + 4.5
55-93
Height (cm)
172.5 + 1.7
164-181
BMI (kg/m2)
24.7 2 1.3
18.2-30.0
Fasting plasma glucose (mmol/L)
4.7 2 0.1
4.2-5.3
Fasting plasma insulin (mu/L)
7.4 z 0.7
4.6-l 1.1
Fasting plasma C-peptide (ng/mL)
1.6 2 0.2
0.7-2.4
Insulin sensitivity (M/I) was expressed as the ratio of rate of glucose utilization (M) at the second hour of clamp to the mean plasma insulin level (I) of the corresponding period. Analytical Procedures Blood samples for the determination of blood glucose concentrations were collected every 5 minutes and measured immediately by glucose oxidase method using a YSI glucose analyzer (Yellow Springs Instrument, Yellow Springs, OH). Blood samples for glucose specific activity were drawn every 10 to 15 minutes into EDTA-containing tubes and centrifuged. The separation and processing of plasma and measurement of glucose specific activity were performed as previously described.” Blood samples for insulin assays were collected every 20 minutes in a prechilled tubes containing EDTA and aprotinin (a trypsin inhibitor, 500 U/mL), immediately centrifuged, and stored at -70°C until analyzed. Plasma insulin concentrations were measured by radioimmunoassay using commercial kits from Dainabot, Tokyo, Japan. Fasting plasma C-peptide concentrations were measured by radioimmunoassay using commercial kits from Daiichi, Tokyo, Japan. StatisticalAnalysis All data are presented as mean 2 SEM. The relationship between variables was evaluated using correlation and regression procedures as implemented by the Statistical Package for the Social Sciences. Stepwise forward regression analysis was used to determine the relative importance of the indices of regional fat distribution on insulin sensitivity. RESULTS
Body Fat Distribution
The correlations among BMI and indices of regional fat distribution in our subjects are shown in Table 2. In these subjects, BMI was correlated positively with WHR (r = .89, P < .Ol), subcutaneous abdominal fat area (r = .96, P < .OOl), and intra-abdominal fat area (I = .74, P < .0.5).
Euglycemic Glucose Clamp Studies
The mean plasma insulin concentration was 89.9 ? 7.7 mU/L (66.2 to 117.8) and insulin sensitivity (M/I) was 0.10 2 0.02 mg/kg/min/mU/L (0.03 to 0.17) during the second hour of glucose clamp. M/I was negatively correlated with BMI (I = .91, P < .OOl), WHR (r = -.80, P < .Ol), subcutaneous abdominal fat area (r = -.90, P < .OOl), and intra-abdominal fat area (r = - .88, P < .Ol; Table 3). Since BMI and indices of regional fat distribution were closely intercorrelated in these subjects (Table 2), stepwise forward regression analysis was performed to determine the relative importance of these indices on decreases in M/I. The variables included were BMI, WHR, subcutaneous abdominal fat area, intraabdominal fat area, and M/I. Of these variables, BMI and intra-abdominal fat area were independently correlated with decreased M/I (Table 4). DISCUSSION
The present data show that intra-abdominal fat is associated with a decrease in insulin sensitivity, which is independent of the effect of BMI in a group of healthy young men. These results are consistent with those of Sparrow et al’ showing that in a sample of middle- or old-aged men (41 to 76 years), the amount of intra-abdominal fat determined by CT was a significant correlate of plasma glucose levels measured 2 hours after oral glucose administration. Fujioka et al.” also reported that in 15 men, visceral fat to subcutaneous fat ratio (VSR) measured by CT was correlated with the plasma glucose area under the curve after an OGTT, although subjects in the high VSR group were older than those in the lower VSR group. However, these studies included various age groups, mainly middle- or old-aged, as well as some diabetics. It is well known that regional fat distribution’7.‘H and insulin sensitivity’9.2”may be affected by aging. The inclusion of diabetic subjects with exhausted p-cell activity also could mislead the results. Thus, it would be desirable to select rather homogenous subjects in respect to age and metabolic disease in assessing the relationship between fat distribution and insulin sensitivity. In our study, we studied healthy young men who were 24 to 32 years of age and had no diabetes or family history of diabetes and also found that intra-abdominal fat is an important correlate to decreased insulin sensitivity. Our results showed that the correlation between WHR and M/I became insignificant in stepwise regression analy-
Table 2. Correlation Coefficients Between BMI and Indices of Regional Fat Distribution Abdominal BMI
BMI
1.oo
WHR
0.89t
WHR
Subcutaneous
Table 3. Correlation Coefficients Between Insulin Sensitivity (M/I) Fat Area
*P
Fat Is Associated KS.
With Decreased Young Men
Insulin
Sensitivity
in Healthy
Park, B.D. Rhee, K-U. Lee, S.Y. Kim, H.K. Lee, C.-S. Koh, and H.K. Min
To distinguish the relative role of intra-abdominal and subcutaneous abdominal fat in metabolic aberrations in upper body fat localization, we measured the relationship between regional fat distribution and insulin sensitivity in nine young men (28.8 ? 0.7 years; body mass index [BMI], 24.7 + 1.3 kg/m’). Regional fat distribution was measured by anthropometric measurement and computed tomography (CT). Insulin sensitivity was measured by euglycemic hyperinsulinemic glucose clamp. Insulin sensitivity, expressed as the ratio of rate of glucose utilization to the mean plasma insulin concentration during the second hour of glucose clamp (M/I) was negatively correlated with BMI (r = -.81, P < .OOl), waist to hip girth ratio (WHR) (r = -.80, P < .Ol), subcutaneous abdominal fat area (r = -.80, P < .OOl), and intra-abdominal fat area (r = -.88, P < .Ol). Stepwise forward regression analysis showed that in addition to BMI, intra-abdominal fat area was a significant correlate of decrease in insulin sensitivity. These findings suggest that intra-abdominal fat play an important role in decreasing insulin sensitivity, even in healthy young men. Copyright 0 199 1 by W. B. Saunders Company
R
EGIONAL FAT DISTRIBUTION has been shown to be associated with metabolic abnormalities and disease states, beyond what is explainable by general adiposity.‘.’ Thus, distribution of excessive fat on the upper part of body or abdomen, as measured by the waist to hip girth ratio (WHR), is associated with reduced glucose tolerance, insulin resistance, and the risk for diabetes.‘~’ Although the precise mechanism of these metabolic aberrations in upper body obesity is not clear, several studies”-” suggested that an increased production of free fatty acids (FFA) from the enlarged abdominal adipose tissues might be of importance. Abdominal adipose tissues comprise subcutaneous abdominal fat and intra-abdominal fat, which are different in size, metabolic activity, and drainage to circulation. Thus, it would be of interest to distinguish the relative role of intra-abdominal and subcutaneous abdominal fat in metabolic aberrations in upper body obesity. Computed tomography (CT) provides a convenient means to measure the amount of subcutaneous and intraabdominal fat.‘.’ Several studies’~’ reported that intraabdominal fat determined by CT was associated with insulin resistance on oral glucose tolerance test (OGTT) in middle- or old-aged subjects. The aim of this study was to determine whether this association is observable in healthy young men. We used the euglycemic glucose clamp technique to assess insulin sensitivity more accurately. MATERIALS
AND METHODS
Subjects
Nine young, healthy male medical doctors (28.6 + 0.7 years; body
mass
index
[BMI],
24.7 2 1.3 kg/m*) volunteered
for the
study. All participants were not taking any medication, nor had any medical problems known to affect glucose metabolism. They had no family history of diabetes mellitus. The experimental protocol was approved by the Human Experimental Committee of the Department of Internal Medicine at the Seoul National University Hospital. South Korea. Characteristics of the subjects are shown in Table 1.
Body Fat Distribution Measurements for regional fat distribution included WHR and subcutaneous abdominal and intra-abdominal fat areas. Waist circumference was measured at umbilicus level and hip circumference was measured at maximal hip circumference. WHR was calculated from these measurements. Intra-abdominal and subcutaneous abdominal fat areas were measured in each subject with CT at umbilicus level on supine position using a General Electric CT scanner (Milwaukee, WI). Fat areas were estimated by a procedure described by Tokunaga et al.”
Euglycernic Glucose Clamp Studies Euglycemic subject after scribed.14
glucose clamp studies were performed in each a 12- to 14-hour overnight fast as previously de-
Primed (25 PCi) and continuous (0.25 p,Ci/min) infusion of tritiated glucose ([3--‘HI glucose, Amersham. Buckinghamshire. England; specific activity, 5.0 Ciimmol) was maintained throughout the studies. After a 2-hour period to attain an isotopic steady-state, the subjects received a primed continuous insulin infusion (Velosulin, Nordisk, Coppenhagen. Denmark) at 1.0 mU/kglmin for 2 hours by an IMED pump (San Diego, CA) in conjunction with the infusion of somatostatin (300 kg/h; Bachem, Torrance, CA). Concurrent with insulin infusion, blood glucose levels were clamped at 4.44 mmol/L by determining the blood glucose concentrations every 5 minutes and periodically adjusting the rate of administration of 20% glucose solution based on negative feedback princip1e.l’
Isotopic Determination of Glucose Turnover From the Department of Internal Medicine, College of Medicine. Seoul National University, Inje University, and University of Ulsan, Seoul, Korea. Address reprint requests to H.K. Min, MD, Department of Internal Medicine, Seoul National Universiry Hospital, #28 Yunkun-Dong, Chongno-ku, Seoul 110-744, Korea. Copyright 0 1991 by W.B. Saunders Company 0026-049519114006-0008$03.00/O 600
Following the administration of insulin, non-steady-state rates of glucose turnover were determined using the equation of Steele” in the derivative form for non-steady-state condition. Since endogenous glucose production determined by subtracting the amount of exogenously infused glucose from the isotopically determined total glucose appearance was negative in all subjects during the second hour clamp, peripheral glucose utilization was determined from the infusion rate of glucose during the second hour of clamp. Metabolism,Vol40,No
6 (June), 1991:~~600-603
INTRA-ABDOMINAL
FAT AND INSULIN SENSITIVITY
Table 1. Clinical Characteristics
of Subjects (n = 9) Mean k
Age (vr)
601
SEM
28.6 + 0.7
Range 24-32
Weight (kg)
73.7 + 4.5
55-93
Height (cm)
172.5 + 1.7
164-181
BMI (kg/m2)
24.7 2 1.3
18.2-30.0
Fasting plasma glucose (mmol/L)
4.7 2 0.1
4.2-5.3
Fasting plasma insulin (mu/L)
7.4 z 0.7
4.6-l 1.1
Fasting plasma C-peptide (ng/mL)
1.6 2 0.2
0.7-2.4
Insulin sensitivity (M/I) was expressed as the ratio of rate of glucose utilization (M) at the second hour of clamp to the mean plasma insulin level (I) of the corresponding period. Analytical Procedures Blood samples for the determination of blood glucose concentrations were collected every 5 minutes and measured immediately by glucose oxidase method using a YSI glucose analyzer (Yellow Springs Instrument, Yellow Springs, OH). Blood samples for glucose specific activity were drawn every 10 to 15 minutes into EDTA-containing tubes and centrifuged. The separation and processing of plasma and measurement of glucose specific activity were performed as previously described.” Blood samples for insulin assays were collected every 20 minutes in a prechilled tubes containing EDTA and aprotinin (a trypsin inhibitor, 500 U/mL), immediately centrifuged, and stored at -70°C until analyzed. Plasma insulin concentrations were measured by radioimmunoassay using commercial kits from Dainabot, Tokyo, Japan. Fasting plasma C-peptide concentrations were measured by radioimmunoassay using commercial kits from Daiichi, Tokyo, Japan. StatisticalAnalysis All data are presented as mean 2 SEM. The relationship between variables was evaluated using correlation and regression procedures as implemented by the Statistical Package for the Social Sciences. Stepwise forward regression analysis was used to determine the relative importance of the indices of regional fat distribution on insulin sensitivity. RESULTS
Body Fat Distribution
The correlations among BMI and indices of regional fat distribution in our subjects are shown in Table 2. In these subjects, BMI was correlated positively with WHR (r = .89, P < .Ol), subcutaneous abdominal fat area (r = .96, P < .OOl), and intra-abdominal fat area (I = .74, P < .0.5).
Euglycemic Glucose Clamp Studies
The mean plasma insulin concentration was 89.9 ? 7.7 mU/L (66.2 to 117.8) and insulin sensitivity (M/I) was 0.10 2 0.02 mg/kg/min/mU/L (0.03 to 0.17) during the second hour of glucose clamp. M/I was negatively correlated with BMI (I = .91, P < .OOl), WHR (r = -.80, P < .Ol), subcutaneous abdominal fat area (r = -.90, P < .OOl), and intra-abdominal fat area (r = - .88, P < .Ol; Table 3). Since BMI and indices of regional fat distribution were closely intercorrelated in these subjects (Table 2), stepwise forward regression analysis was performed to determine the relative importance of these indices on decreases in M/I. The variables included were BMI, WHR, subcutaneous abdominal fat area, intraabdominal fat area, and M/I. Of these variables, BMI and intra-abdominal fat area were independently correlated with decreased M/I (Table 4). DISCUSSION
The present data show that intra-abdominal fat is associated with a decrease in insulin sensitivity, which is independent of the effect of BMI in a group of healthy young men. These results are consistent with those of Sparrow et al’ showing that in a sample of middle- or old-aged men (41 to 76 years), the amount of intra-abdominal fat determined by CT was a significant correlate of plasma glucose levels measured 2 hours after oral glucose administration. Fujioka et al.” also reported that in 15 men, visceral fat to subcutaneous fat ratio (VSR) measured by CT was correlated with the plasma glucose area under the curve after an OGTT, although subjects in the high VSR group were older than those in the lower VSR group. However, these studies included various age groups, mainly middle- or old-aged, as well as some diabetics. It is well known that regional fat distribution’7.‘H and insulin sensitivity’9.2”may be affected by aging. The inclusion of diabetic subjects with exhausted p-cell activity also could mislead the results. Thus, it would be desirable to select rather homogenous subjects in respect to age and metabolic disease in assessing the relationship between fat distribution and insulin sensitivity. In our study, we studied healthy young men who were 24 to 32 years of age and had no diabetes or family history of diabetes and also found that intra-abdominal fat is an important correlate to decreased insulin sensitivity. Our results showed that the correlation between WHR and M/I became insignificant in stepwise regression analy-
Table 2. Correlation Coefficients Between BMI and Indices of Regional Fat Distribution Abdominal BMI
BMI
1.oo
WHR
0.89t
WHR
Subcutaneous
Table 3. Correlation Coefficients Between Insulin Sensitivity (M/I) Fat Area
*P
