0021-972X/90/7104-1028$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 71, No. 4 Printed in U.S.A.
Abdominal Fat Cell Lipolysis, Body Fat Distribution, and Metabolic Variables in Premenopausal Women* P. MAURIEGE, J. P. DESPRESf, M. MARCOTTE, M. FERLAND, A. TREMBLAY, A. NADEAU, S. MOORJANI, P. J. LUPIEN, G. THERIAULT, AND C. BOUCHARD Physical Activity Sciences Laboratory and the Department of Medicine, Laval University, Ste-Foy, Quebec, GlK 7P4 Canada
ABSTRACT. It is well established that abdominal obesity is related to numerous metabolic abnormalities and that this correlation represents a significant risk factor for coronary heart disease and related mortality. In the present study the relationships among the regional distribution of body fat, selected metabolic variables, and abdominal adipose cell lipolysis were investigated in 30 premenopausal women, 34 ± 8 yr (mean ± SD) of age, with body mass indices ranging from 17-45 kg/m2. Basal as well as epinephrine- and isoproterenol-stimulated lipolyses were positively correlated with fasting plasma insulin and triglyceride levels (0.48 < r < 0.64; 0.05 > P < 0.0005 and 0.46 < r < 0.60; 0.05 > P < 0.005, respectively) and with the insulin area measured during an oral glucose tolerance test (0.49 < r < 0.67; 0.005 > P < 0.0005). With the exception of epinephrine-stimulated lipolysis, these correlations remained significant when lipolysis
S
INCE the pioneering work of Vague (1, 2), several studies have been reported on the associations between adipose tissue distribution and metabolic disturbances (3-5). It is now well recognized that excessive abdominal adiposity is associated with hyperinsulinemia and insulin resistance (6) and diabetes mellitus (7). Abdominal fat accumulation is also accompanied by a higher incidence of hypertriglyceridemia and changes in plasma lipoprotein levels (8-11) as well as by a slightly increased susceptibility to hypertension (12-14). Prospective studies have shown that excess abdominal adipose tissue could be considered as a significant risk factor for cardiovascular disease and premature death (15-19). Although the associations between the regional distribution of adipose tissue and certain metabolic alterations observed in the obese state have been well documented, the mechanisms responsible for these associations are not well understood. A role for adipocyte lipolysis in the
Received December 18, 1989. * This work was supported by Health and Welfare Canada, FCARQuebec, the Quebec Heart Foundation, the Canadian Fitness and Lifestyle Research Institute, and the Fonds de la Recherche en Sante du Quebec (FRSQ). fFRSQ Scholar. To whom all correspondence and requests for reprints should be addressed.
was corrected for cell surface area. Basal and maximal epinephrine- and isoproterenol-induced lipolyses were also negatively related to plasma high density lipoprotein cholesterol (—0.52 < r < -0.36; 0.05 > P < 0.005). However, these relationships were no longer significant after control for fat cell surface. The associations between abdominal lipolysis and fat distribution did not remain significant when data were adjusted for total adiposity. Taken together, these results support the notion that variations in abdominal adipocyte lipolysis 1) depend more on total body fatness than on fat distribution, and 2) may be involved in the metabolic complications associated with abdominal obesity, particularly those pertaining to plasma insulin and triglyceride metabolism. (J Clin Endocrinol Metab 7 1 : 10281035, 1990)
etiology of the metabolic complications associated with abdominal obesity has already been suggested (20), but experimental evidence in support of this notion is lacking. Therefore, the aim of the present study was to test whether variations in abdominal fat cell lipolytic activity are associated with the metabolic alterations of abdominal obesity. The relationships among regional fat distribution, the metabolic profile, and abdominal fat cell lipolysis were investigated in a sample of 30 premenopausal women. Results emphasize the potential role of abdominal adipose tissue lipolysis in the etiology of some metabolic abnormalities observed in abdominal obesity.
Material and Methods Subjects Thirty premenopausal women gave their written consent to participate in this study, which received the approval of the Laval University Medical Ethics Committee. They were ascertained as sedentary before the experiment by means of a questionnaire. All subjects had a medical examination, and none of them had experienced recent illness or endocrine abnormalities. The metabolic profile was measured while subjects were in the early follicular phase of the menstrual cycle.
1028 Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
ABDOMINAL FAT CELL LIPOLYSIS AND METABOLISM Body composition and fat distribution The body mass index was obtained from height and weight measured without shoes. Body composition was assessed by several methods. Subcutaneous skinfold thicknesses (biceps, triceps, subscapular, abdominal, suprailiac, thigh, and medial calf) were measured, using a Harpenden skinfold caliper (21). The value for a given site was the mean of three measurements. Suprailiac, subscapular, and abdomen skinfolds were summed up to represent the trunk sc fat, whereas the sum of biceps, triceps, thigh, and medial calf skinfold data represented the extremity sc fat. The sum of the seven skinfolds was considered an indicator of total sc fat. Subcutaneous regional fat distribution was estimated by the trunk to extremity (T/E) skinfolds ratio as well as the waist to hip circumferences ratio (WHR). The body density was measured by the hydrostatic weighing technique (22), as previously described (23). Percent body fat was derived from body density (24). Residual lung volume was measured by the helium dilution method (25). Fat mass was calculated as total body weight minus fat-free mass. Computed axial tomography was performed on a Siemens Somatom DRH scanner (Erlangen, West Germany) (26), as previously described (27). A radiograph of the skeleton established the position of the abdominal scan (between L4 and L5) to the nearest millimeter. Total and deep abdominal fat areas were calculated by delineating these areas with a graph pen and then computing the adipose tissue surfaces using an attenuation range of -30 to -190 HU (26, 28). Deep abdominal fat area was measured by drawing a line within the muscle wall surrounding the abdominal cavity. The abdominal sc fat area was calculated by subtracting the amount of deep fat from total fat areas. Adipose tissue morphology and Upolysis After an overnight fast, participants were subjected to a biopsy of sc fat in the abdominal region. After skin anesthesia with 1% xylocaine, a 1-cm incision was made, and adipose tissue was surgically removed. Adipocytes were isolated by collagenase digestion (29), as previously described (30). Briefly, adipose tissue was minced in fresh Krebs-Ringer bicarbonate buffer at pH 7.4, supplemented with glucose (5 mM), BSA (4%), and collagenase (1 mg/mL). Adipocytes were then incubated under an atmosphere of 95% O2 and 5% CO2 in a water bath at 37 C, shaking at 60 cycles/min for 45 min. Fat cell diameters were determined using a Leitz microscope equipped with a graduated ocular (Rockleigh, NJ). Mean fat cell diameter was assessed from the measurement of at least 500 cells, and the density of triolein was used to transform adipose cell volume into fat cell weight. Basal, epinephrine-maximal (10~5 M), and isoproterenol-maximal (10~5 M) lipolyses were assessed as previously described (30). After a 2-h incubation period, incubation was stopped by transferring the vial content in polystyrene tubes on ice, floating adipocytes were then aspirated, and the glycerol content of the infranatant was enzymatically assessed (31). Oral glucose tolerance test (OGTT) A 75-g OGTT was performed in the morning after an overnight fast. Blood samples were collected through a catheter
Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
1029
from an antecubital vein at -15, 0, 15, 30, 45, 60, 90, 120, 150, and 180 min. Plasma glucose was enzymatically measured (32), whereas the plasma insulin concentration was determined by RIA with polyethylene glycol separation (33). The postglucose plasma insulin and glucose areas under the curves were also calculated. Plasma lipids and lipoproteins Blood samples were obtained in the morning after a 12-h fast. Cholesterol and triglyceride levels were determined in plasma and lipoprotein fractions, using an Auto-Analyzer II (Technicon Instruments Corp., Tarrytown, NY) (34), after extraction with isopropanol and treatment with zeolite according to the Technicon AA-II procedure (35). Plasma very low density lipoproteins (VLDL; density 1.006 g/mL) with heparin and MnCl2 (37). Statistical analyses Data reported in the tables are expressed as the mean and SD. Relationships between two variables were evaluated by the Pearson product-moment correlation coefficient. Results Tables 1 and 2 present the physical and metabolic
characteristics of the sample of 30 premenopausal women. The body fatness indices indicated variations from normal to massive adiposity. Accordingly, there was considerable individual variation in the metabolic profile of the subjects. Basal, epinephrine-induced, and isoproterenol-induced maximal lipolyses were all significantly higher in obese than in lean women when results were expressed either per cell number or per cell surface area (Fig. 1). The maximum lipolytic effect of isoproterenol (a pure /3adrenergic agonist) was much greater (1.7 and 1.4 times TABLE 1. Descriptive characteristics and body fat distribution variables of the 30 subjects
Age (yr) BW (kg) BMI (kg/m2) Body fat (%) WHR Abdominal fat areas (L4L5) Total (cm2) sc (cm2) Deep (cm2) Abdominal fat cell wt (ng lipid/cell) BMI, Body mass index.
Mean ± SD
Range
34 ± 8 74 ±20 29 ± 8 39 ±10 0.8 ± 0.05
25-49 45-120 17-45 20-55 0.7-0.9
499 ± 264 92 ± 4 8 406 ± 222 0.59 ± 0.29
100-962 28-182 65-812 0.12-1.23
MAURIEGE ET AL.
1030 TABLE 2. Metabolic profile of the subjects
Insulin (pmol/L) Glucose (mmol/L) Insulin area Glucose area TG (mmol/L) VLDL chol (mmol/L) LDL chol (mmol/L) HDL chol (mmol/L) HDL chol/LDL chol
Mean ± SD
Range
71 ± 5 3 4.8 ± 0.5 31.7 ± 18.6 1.15 ± 0.21 1.2 ± 0.6 0.4 ± 0.2 3.5 ± 0.9 1.2 ± 0.3 0.4 ± 0.2
18-204 4-6.4 6.9-77.6 0.7-1.7 0.3-2.5 0.05-0.88 1.5-5.8 0.8-2.0 0.2-1.4
Insulin and glucose areas are, respectively, expressed in (picomoles • L-M80 min"1) x 10~3 and millimoles • L-M80 min"1) x 10"3. TG, Triglycerides; chol, cholesterol.
higher for obese and lean women, respectively) than that of epinephrine (a mixed a2- and /3-agonist). Stimulation of lipolysis by isoproterenol and epinephrine was maximal at 10~5 M in both groups, with a more pronounced effect in adipocytes from obese than lean women. Table 3 shows the associations between body fatness measurements and the lipolytic response. Increases in body fat were associated with elevations in the rate of basal lipolysis expressed on a per cell basis (r = 0.55; P < 0.005). Basal lipolysis increased with the amount of sc trunk fat (r = 0.63; P < 0.0005) and the waist circumference (r = 0.51; P < 0.05). Significant correlations were also observed with both the T/E ratio (r = 0.44; P < 0.05) and the WHR (r = 0.41; P < 0.005). All of these correlations remained significant when lipolysis was corrected for cell surface area, with the exception of the WHR. The maximal epinephrine- and isoproterenol-
JCE & M • 1990 Vol71«No4
induced lipolyses, when expressed on a per cell basis, were strongly and positively associated with the percentage of body fat (r = 0.65 and r = 0.77, respectively; P < 0.0005). Similarly, these lipolytic responses showed significant positive correlations with the trunk fat (0.65 < r < 0.79; P < 0.0005) as well as with the waist circumference (0.59 < r < 0.68; 0.0005 < P < 0.005). Moreover, the T/E ratio displayed lower but significant associations with lipolytic activity stimulated by catecholamines (0.50 < r < 0.56; 0.0005
FIG. 1. Basal, epinephrine (Epi.)-stimulated and isoproterenol (Iso.)-stimulated maximal lipolyses in lean (•; n = 14) and obese (D; n = 16) women. Lipolytic activities are expressed either in micromoles of glycerol per 106 cells/2 h (A) or nanomoles of glycerol per 108 cells/ Mm2-2 h (B). Values are the mean ± SD.
u o
20
o QH
H3
10
BASAL
Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
EPI. ISO. -5 10 M
BASAL
EPI. ISO. -5 10 M
ABDOMINAL FAT CELL LIPOLYSIS AND METABOLISM TABLE 3. Correlation coefficients between the abdominal lipolytic response and anthropometric indices as well as abdominal scan areas determined by CT
TABLE 4. Correlation coefficients between the abdominal lipolytic response and metabolic variables Basal
Basal 6
Epinephrine 2
6
2
/10 cell /Mm /10 cell /urn
Isoproterenol 6
/10 cell /Mm
FCW
0.39" 0.48" 0.406 0.38" 0.32 0.28
0.65c 0.65c 0.516 0.59° 0.50° 0.59c
0.376 0.396 0.386 0.36" 0.33 0.26
0.77c 0.79c 0.55° 0.68c 0.56° 0.76c
0.406 0.43" 0.44" 0.37* 0.416 0.29
Abdominal fat areas sc Deep
0.50° 0.50°
0.35 0.35
0.57° 0.57°
0.32 0.31
0.67c 0.66c
0.30 0.29
Deep/total
-0.11
-0.06
-0.17
-0.15
-0.14
Epinephrine 6
/10 6 cell
Isoproterenol 2
/10 cell
/Mm
0.48" 0.37" 0.49° 0.36 0.51° -0.36* -0.39 b
0.25 0.15 0.23 0.31 0.29 -0.17 -0.24
2
0.55° 0.63c 0.44" 0.516 0.41° 0.49°
Body fat Trunk fat T/E ratio Waist WHR
1031
-0.03
FCW, Fat cell weight; Basal, basal lipolysis; Epinephrine, epinephrine-maximal lipolysis; Isoproterenol, isoproterenol-maximal lipolysis. " P < 0.005. * P < 0.05. c P < 0.0005.
analysis, and the relations of lipolytic activity to the residual anthropometric variables (corrected for the percent fat or fat mass) were also investigated. Among all of the variables tested, the trunk sc fat (corrected for fat mass) was the only variable that remained significantly correlated with either basal or isoproterenol-induced lipolytic activity (0.37 < r < 0.47; P < 0.05), whereas it was no longer associated with epinephrine-stimulated lipolysis. All of the other anthropometric indicators of body fat and fat distribution (WHR, waist circumference, CT fat areas, T/E ratio, and fat cell weight corrected for fat mass) did not remain significantly associated with abdominal adipose cell lipolysis. Finally, when corrected for percentage of body fat, no relationship was observed between body fat distribution variables and the abdominal fat cell lipolytic activity. These results suggest that variations in abdominal adipose cell lipolysis are related to body fatness rather than to the regional distribution of body fat. Table 4 presents correlation coefficients for fat cell lipolysis and metabolic variables. Firstly, there was no relationship between lipolysis, either basal or stimulated by adrenergic agents, with the plasma glucose area measured during an OGTT. Secondly, the fasting plasma insulin and glucose levels were positively correlated with lipolysis, expressed on a per cell basis. After correction for fat cell surface area, low but significant relationships were still observed between fasting insulin levels and either basal lipolysis (r = 0.44; P < 0.05) or maximal lipolysis induced by isoproterenol (r = 0.41; P < 0.005). Significant positive relationships were also noted between the plasma insulin area under the curve measured during the OGTT and the lipolytic response (0.49 < r < 0.67; 0.0005 < P < 0.005). After correction for the adi-
Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
Insulin Glucose Insulin area Glucose area TG (log,,,) HDLchol HDL/LDL
0.55" 0.31° 0.57° 0.14 0.46° -0.38 6 -0.29
0.44" 0.18 0.436 0.09 0.34" -0.28 -0.16
/10 6 cell
/Mm2
0.64c 0.406 0.67c 0.23 0.60c -0.52° -0.45fe
0.41° 0.05 0.38* 0.10 0.366 -0.31 -0.23
Basal, Basal lipolysis; Epinephrine, epinephrine-maximal lipolysis; Isoproterenol, isoproterenol-maximal lipolysis; TG, triglyceride; chol, cholesterol. a
P< 0.005.
b
P < 0.05. P < 0.0005.
c
pocyte surface area, both basal and maximal isoproterenol-stimulated lipolyses remained significantly associated with the insulin area. Moreover, the insulinogenic index (assessed by the ratio of insulin area/glucose area) was not correlated with the adipose cell lipolytic response (data not shown). Plasma triglyceride levels were also significantly and positively related to lipolysis when expressed on a per cell basis (Table 4). The associations between basal and isoproterenol-stimulated lipolytic activities and plasma triglyceride levels decreased considerably, but still remained significant after correction for fat cell surface (0.34 < r < 0.36; P < 0.05); this was not the case for maximal epinephrine-stimulated lipolysis (r = 0.29). Finally, as shown in Table 4, plasma HDL cholesterol was negatively correlated with the maximal isoproterenol lipolytic response (r = —0.52; P < 0.005) as well as with the basal and maximal epinephrineinduced lipolyses (r = —0.38 and r = -0.36, respectively; P < 0.05), but these associations were no longer significant when lipolysis was corrected for fat cell surface. The HDL cholesterol/LDL cholesterol ratio also displayed inverse relationships with isoproterenol- and epinephrine-stimulated lipolyses (r = —0.45 and r = —0.39 respectively; P < 0.05). However, these correlations did not remain significant after control for fat cell surface. Discussion The aim of the present study was to investigate the potential role of abdominal adipose cell lipolysis in the metabolic complications generally associated with abdominal obesity. Although it has become increasingly evident that excess abdominal fat is an important factor involved in the development of metabolic disturbances (3-5, 15-19, 38), the mechanisms responsible for the association between abdominal fat and metabolism are not fully understood. The increase in stimulated adipocyte lipolysis noted
1032
MAURIEGE ET AL.
with obesity suggests an elevated metabolic activity of abdominal adipose tissue. Fat cells are obviously larger in obese than in nonobese subjects (39), and lipolysis stimulated by catecholamines has been reported to be enhanced in large human adipocytes (40). Furthermore, cells from all depots in obese subjects are more lipolytically responsive than those from lean individuals, and lipolysis is more elevated in obese patients, particularly in individuals with abdominal obesity (41). In vitro, basal and catecholamine-stimulated lipolyses determined in isolated abdominal fat cells increased with the level of obesity, as shown by others (42). It seems reasonable to assume that fat cell size is a major correlate of adipose tissue metabolism and, thus, of hormone responsiveness (43). Our findings are in accordance with those of a previous study in which basal and catecholamine-stimulated lipolyses were positively correlated with fat cell size when lipolysis was expressed per cell number (44). It has already been suggested that the expression of lipolysis on a per cell basis can be misleading at times, as it does not discriminate between the effects due to changes in cell size per se and those caused by alterations related to the intrinsic properties of the cell (45). Our results suggest that the overall increase in lipolytic activity in obesity can be explained mostly by the higher fat cell size found in obese individuals (46). The mechanisms involved in the relationships between fat cell size and adipose tissue metabolism are still poorly understood. Nonetheless, it is attractive to consider that an increase in /3-adrenoceptor number as well as cAMP levels (44, 47) and/or a decrease in the a2-adrenergic component (45) may contribute to the faster rate of catecholamine-stimulated lipolysis in large compared to small adipocytes. Further studies are needed to test such a hypothesis, but it has been suggested that the a2-/(3adrenergic receptor status could be of importance (48). In the present study we noted that the physiological agonist epinephrine by itself exerted a lipolytic action. The absence of adenosine deaminase in the incubation medium could explain such an observation. Indeed, in the presence of adenosine deaminase, epinephrine inhibits lipolysis at lower concentrations, which may be of physiological relevance (i.e.
Vol. 71, No. 4 Printed in U.S.A.
Abdominal Fat Cell Lipolysis, Body Fat Distribution, and Metabolic Variables in Premenopausal Women* P. MAURIEGE, J. P. DESPRESf, M. MARCOTTE, M. FERLAND, A. TREMBLAY, A. NADEAU, S. MOORJANI, P. J. LUPIEN, G. THERIAULT, AND C. BOUCHARD Physical Activity Sciences Laboratory and the Department of Medicine, Laval University, Ste-Foy, Quebec, GlK 7P4 Canada
ABSTRACT. It is well established that abdominal obesity is related to numerous metabolic abnormalities and that this correlation represents a significant risk factor for coronary heart disease and related mortality. In the present study the relationships among the regional distribution of body fat, selected metabolic variables, and abdominal adipose cell lipolysis were investigated in 30 premenopausal women, 34 ± 8 yr (mean ± SD) of age, with body mass indices ranging from 17-45 kg/m2. Basal as well as epinephrine- and isoproterenol-stimulated lipolyses were positively correlated with fasting plasma insulin and triglyceride levels (0.48 < r < 0.64; 0.05 > P < 0.0005 and 0.46 < r < 0.60; 0.05 > P < 0.005, respectively) and with the insulin area measured during an oral glucose tolerance test (0.49 < r < 0.67; 0.005 > P < 0.0005). With the exception of epinephrine-stimulated lipolysis, these correlations remained significant when lipolysis
S
INCE the pioneering work of Vague (1, 2), several studies have been reported on the associations between adipose tissue distribution and metabolic disturbances (3-5). It is now well recognized that excessive abdominal adiposity is associated with hyperinsulinemia and insulin resistance (6) and diabetes mellitus (7). Abdominal fat accumulation is also accompanied by a higher incidence of hypertriglyceridemia and changes in plasma lipoprotein levels (8-11) as well as by a slightly increased susceptibility to hypertension (12-14). Prospective studies have shown that excess abdominal adipose tissue could be considered as a significant risk factor for cardiovascular disease and premature death (15-19). Although the associations between the regional distribution of adipose tissue and certain metabolic alterations observed in the obese state have been well documented, the mechanisms responsible for these associations are not well understood. A role for adipocyte lipolysis in the
Received December 18, 1989. * This work was supported by Health and Welfare Canada, FCARQuebec, the Quebec Heart Foundation, the Canadian Fitness and Lifestyle Research Institute, and the Fonds de la Recherche en Sante du Quebec (FRSQ). fFRSQ Scholar. To whom all correspondence and requests for reprints should be addressed.
was corrected for cell surface area. Basal and maximal epinephrine- and isoproterenol-induced lipolyses were also negatively related to plasma high density lipoprotein cholesterol (—0.52 < r < -0.36; 0.05 > P < 0.005). However, these relationships were no longer significant after control for fat cell surface. The associations between abdominal lipolysis and fat distribution did not remain significant when data were adjusted for total adiposity. Taken together, these results support the notion that variations in abdominal adipocyte lipolysis 1) depend more on total body fatness than on fat distribution, and 2) may be involved in the metabolic complications associated with abdominal obesity, particularly those pertaining to plasma insulin and triglyceride metabolism. (J Clin Endocrinol Metab 7 1 : 10281035, 1990)
etiology of the metabolic complications associated with abdominal obesity has already been suggested (20), but experimental evidence in support of this notion is lacking. Therefore, the aim of the present study was to test whether variations in abdominal fat cell lipolytic activity are associated with the metabolic alterations of abdominal obesity. The relationships among regional fat distribution, the metabolic profile, and abdominal fat cell lipolysis were investigated in a sample of 30 premenopausal women. Results emphasize the potential role of abdominal adipose tissue lipolysis in the etiology of some metabolic abnormalities observed in abdominal obesity.
Material and Methods Subjects Thirty premenopausal women gave their written consent to participate in this study, which received the approval of the Laval University Medical Ethics Committee. They were ascertained as sedentary before the experiment by means of a questionnaire. All subjects had a medical examination, and none of them had experienced recent illness or endocrine abnormalities. The metabolic profile was measured while subjects were in the early follicular phase of the menstrual cycle.
1028 Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
ABDOMINAL FAT CELL LIPOLYSIS AND METABOLISM Body composition and fat distribution The body mass index was obtained from height and weight measured without shoes. Body composition was assessed by several methods. Subcutaneous skinfold thicknesses (biceps, triceps, subscapular, abdominal, suprailiac, thigh, and medial calf) were measured, using a Harpenden skinfold caliper (21). The value for a given site was the mean of three measurements. Suprailiac, subscapular, and abdomen skinfolds were summed up to represent the trunk sc fat, whereas the sum of biceps, triceps, thigh, and medial calf skinfold data represented the extremity sc fat. The sum of the seven skinfolds was considered an indicator of total sc fat. Subcutaneous regional fat distribution was estimated by the trunk to extremity (T/E) skinfolds ratio as well as the waist to hip circumferences ratio (WHR). The body density was measured by the hydrostatic weighing technique (22), as previously described (23). Percent body fat was derived from body density (24). Residual lung volume was measured by the helium dilution method (25). Fat mass was calculated as total body weight minus fat-free mass. Computed axial tomography was performed on a Siemens Somatom DRH scanner (Erlangen, West Germany) (26), as previously described (27). A radiograph of the skeleton established the position of the abdominal scan (between L4 and L5) to the nearest millimeter. Total and deep abdominal fat areas were calculated by delineating these areas with a graph pen and then computing the adipose tissue surfaces using an attenuation range of -30 to -190 HU (26, 28). Deep abdominal fat area was measured by drawing a line within the muscle wall surrounding the abdominal cavity. The abdominal sc fat area was calculated by subtracting the amount of deep fat from total fat areas. Adipose tissue morphology and Upolysis After an overnight fast, participants were subjected to a biopsy of sc fat in the abdominal region. After skin anesthesia with 1% xylocaine, a 1-cm incision was made, and adipose tissue was surgically removed. Adipocytes were isolated by collagenase digestion (29), as previously described (30). Briefly, adipose tissue was minced in fresh Krebs-Ringer bicarbonate buffer at pH 7.4, supplemented with glucose (5 mM), BSA (4%), and collagenase (1 mg/mL). Adipocytes were then incubated under an atmosphere of 95% O2 and 5% CO2 in a water bath at 37 C, shaking at 60 cycles/min for 45 min. Fat cell diameters were determined using a Leitz microscope equipped with a graduated ocular (Rockleigh, NJ). Mean fat cell diameter was assessed from the measurement of at least 500 cells, and the density of triolein was used to transform adipose cell volume into fat cell weight. Basal, epinephrine-maximal (10~5 M), and isoproterenol-maximal (10~5 M) lipolyses were assessed as previously described (30). After a 2-h incubation period, incubation was stopped by transferring the vial content in polystyrene tubes on ice, floating adipocytes were then aspirated, and the glycerol content of the infranatant was enzymatically assessed (31). Oral glucose tolerance test (OGTT) A 75-g OGTT was performed in the morning after an overnight fast. Blood samples were collected through a catheter
Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
1029
from an antecubital vein at -15, 0, 15, 30, 45, 60, 90, 120, 150, and 180 min. Plasma glucose was enzymatically measured (32), whereas the plasma insulin concentration was determined by RIA with polyethylene glycol separation (33). The postglucose plasma insulin and glucose areas under the curves were also calculated. Plasma lipids and lipoproteins Blood samples were obtained in the morning after a 12-h fast. Cholesterol and triglyceride levels were determined in plasma and lipoprotein fractions, using an Auto-Analyzer II (Technicon Instruments Corp., Tarrytown, NY) (34), after extraction with isopropanol and treatment with zeolite according to the Technicon AA-II procedure (35). Plasma very low density lipoproteins (VLDL; density 1.006 g/mL) with heparin and MnCl2 (37). Statistical analyses Data reported in the tables are expressed as the mean and SD. Relationships between two variables were evaluated by the Pearson product-moment correlation coefficient. Results Tables 1 and 2 present the physical and metabolic
characteristics of the sample of 30 premenopausal women. The body fatness indices indicated variations from normal to massive adiposity. Accordingly, there was considerable individual variation in the metabolic profile of the subjects. Basal, epinephrine-induced, and isoproterenol-induced maximal lipolyses were all significantly higher in obese than in lean women when results were expressed either per cell number or per cell surface area (Fig. 1). The maximum lipolytic effect of isoproterenol (a pure /3adrenergic agonist) was much greater (1.7 and 1.4 times TABLE 1. Descriptive characteristics and body fat distribution variables of the 30 subjects
Age (yr) BW (kg) BMI (kg/m2) Body fat (%) WHR Abdominal fat areas (L4L5) Total (cm2) sc (cm2) Deep (cm2) Abdominal fat cell wt (ng lipid/cell) BMI, Body mass index.
Mean ± SD
Range
34 ± 8 74 ±20 29 ± 8 39 ±10 0.8 ± 0.05
25-49 45-120 17-45 20-55 0.7-0.9
499 ± 264 92 ± 4 8 406 ± 222 0.59 ± 0.29
100-962 28-182 65-812 0.12-1.23
MAURIEGE ET AL.
1030 TABLE 2. Metabolic profile of the subjects
Insulin (pmol/L) Glucose (mmol/L) Insulin area Glucose area TG (mmol/L) VLDL chol (mmol/L) LDL chol (mmol/L) HDL chol (mmol/L) HDL chol/LDL chol
Mean ± SD
Range
71 ± 5 3 4.8 ± 0.5 31.7 ± 18.6 1.15 ± 0.21 1.2 ± 0.6 0.4 ± 0.2 3.5 ± 0.9 1.2 ± 0.3 0.4 ± 0.2
18-204 4-6.4 6.9-77.6 0.7-1.7 0.3-2.5 0.05-0.88 1.5-5.8 0.8-2.0 0.2-1.4
Insulin and glucose areas are, respectively, expressed in (picomoles • L-M80 min"1) x 10~3 and millimoles • L-M80 min"1) x 10"3. TG, Triglycerides; chol, cholesterol.
higher for obese and lean women, respectively) than that of epinephrine (a mixed a2- and /3-agonist). Stimulation of lipolysis by isoproterenol and epinephrine was maximal at 10~5 M in both groups, with a more pronounced effect in adipocytes from obese than lean women. Table 3 shows the associations between body fatness measurements and the lipolytic response. Increases in body fat were associated with elevations in the rate of basal lipolysis expressed on a per cell basis (r = 0.55; P < 0.005). Basal lipolysis increased with the amount of sc trunk fat (r = 0.63; P < 0.0005) and the waist circumference (r = 0.51; P < 0.05). Significant correlations were also observed with both the T/E ratio (r = 0.44; P < 0.05) and the WHR (r = 0.41; P < 0.005). All of these correlations remained significant when lipolysis was corrected for cell surface area, with the exception of the WHR. The maximal epinephrine- and isoproterenol-
JCE & M • 1990 Vol71«No4
induced lipolyses, when expressed on a per cell basis, were strongly and positively associated with the percentage of body fat (r = 0.65 and r = 0.77, respectively; P < 0.0005). Similarly, these lipolytic responses showed significant positive correlations with the trunk fat (0.65 < r < 0.79; P < 0.0005) as well as with the waist circumference (0.59 < r < 0.68; 0.0005 < P < 0.005). Moreover, the T/E ratio displayed lower but significant associations with lipolytic activity stimulated by catecholamines (0.50 < r < 0.56; 0.0005
FIG. 1. Basal, epinephrine (Epi.)-stimulated and isoproterenol (Iso.)-stimulated maximal lipolyses in lean (•; n = 14) and obese (D; n = 16) women. Lipolytic activities are expressed either in micromoles of glycerol per 106 cells/2 h (A) or nanomoles of glycerol per 108 cells/ Mm2-2 h (B). Values are the mean ± SD.
u o
20
o QH
H3
10
BASAL
Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
EPI. ISO. -5 10 M
BASAL
EPI. ISO. -5 10 M
ABDOMINAL FAT CELL LIPOLYSIS AND METABOLISM TABLE 3. Correlation coefficients between the abdominal lipolytic response and anthropometric indices as well as abdominal scan areas determined by CT
TABLE 4. Correlation coefficients between the abdominal lipolytic response and metabolic variables Basal
Basal 6
Epinephrine 2
6
2
/10 cell /Mm /10 cell /urn
Isoproterenol 6
/10 cell /Mm
FCW
0.39" 0.48" 0.406 0.38" 0.32 0.28
0.65c 0.65c 0.516 0.59° 0.50° 0.59c
0.376 0.396 0.386 0.36" 0.33 0.26
0.77c 0.79c 0.55° 0.68c 0.56° 0.76c
0.406 0.43" 0.44" 0.37* 0.416 0.29
Abdominal fat areas sc Deep
0.50° 0.50°
0.35 0.35
0.57° 0.57°
0.32 0.31
0.67c 0.66c
0.30 0.29
Deep/total
-0.11
-0.06
-0.17
-0.15
-0.14
Epinephrine 6
/10 6 cell
Isoproterenol 2
/10 cell
/Mm
0.48" 0.37" 0.49° 0.36 0.51° -0.36* -0.39 b
0.25 0.15 0.23 0.31 0.29 -0.17 -0.24
2
0.55° 0.63c 0.44" 0.516 0.41° 0.49°
Body fat Trunk fat T/E ratio Waist WHR
1031
-0.03
FCW, Fat cell weight; Basal, basal lipolysis; Epinephrine, epinephrine-maximal lipolysis; Isoproterenol, isoproterenol-maximal lipolysis. " P < 0.005. * P < 0.05. c P < 0.0005.
analysis, and the relations of lipolytic activity to the residual anthropometric variables (corrected for the percent fat or fat mass) were also investigated. Among all of the variables tested, the trunk sc fat (corrected for fat mass) was the only variable that remained significantly correlated with either basal or isoproterenol-induced lipolytic activity (0.37 < r < 0.47; P < 0.05), whereas it was no longer associated with epinephrine-stimulated lipolysis. All of the other anthropometric indicators of body fat and fat distribution (WHR, waist circumference, CT fat areas, T/E ratio, and fat cell weight corrected for fat mass) did not remain significantly associated with abdominal adipose cell lipolysis. Finally, when corrected for percentage of body fat, no relationship was observed between body fat distribution variables and the abdominal fat cell lipolytic activity. These results suggest that variations in abdominal adipose cell lipolysis are related to body fatness rather than to the regional distribution of body fat. Table 4 presents correlation coefficients for fat cell lipolysis and metabolic variables. Firstly, there was no relationship between lipolysis, either basal or stimulated by adrenergic agents, with the plasma glucose area measured during an OGTT. Secondly, the fasting plasma insulin and glucose levels were positively correlated with lipolysis, expressed on a per cell basis. After correction for fat cell surface area, low but significant relationships were still observed between fasting insulin levels and either basal lipolysis (r = 0.44; P < 0.05) or maximal lipolysis induced by isoproterenol (r = 0.41; P < 0.005). Significant positive relationships were also noted between the plasma insulin area under the curve measured during the OGTT and the lipolytic response (0.49 < r < 0.67; 0.0005 < P < 0.005). After correction for the adi-
Downloaded from https://academic.oup.com/jcem/article-abstract/71/4/1028/2652831 by East Carolina University user on 05 April 2018
Insulin Glucose Insulin area Glucose area TG (log,,,) HDLchol HDL/LDL
0.55" 0.31° 0.57° 0.14 0.46° -0.38 6 -0.29
0.44" 0.18 0.436 0.09 0.34" -0.28 -0.16
/10 6 cell
/Mm2
0.64c 0.406 0.67c 0.23 0.60c -0.52° -0.45fe
0.41° 0.05 0.38* 0.10 0.366 -0.31 -0.23
Basal, Basal lipolysis; Epinephrine, epinephrine-maximal lipolysis; Isoproterenol, isoproterenol-maximal lipolysis; TG, triglyceride; chol, cholesterol. a
P< 0.005.
b
P < 0.05. P < 0.0005.
c
pocyte surface area, both basal and maximal isoproterenol-stimulated lipolyses remained significantly associated with the insulin area. Moreover, the insulinogenic index (assessed by the ratio of insulin area/glucose area) was not correlated with the adipose cell lipolytic response (data not shown). Plasma triglyceride levels were also significantly and positively related to lipolysis when expressed on a per cell basis (Table 4). The associations between basal and isoproterenol-stimulated lipolytic activities and plasma triglyceride levels decreased considerably, but still remained significant after correction for fat cell surface (0.34 < r < 0.36; P < 0.05); this was not the case for maximal epinephrine-stimulated lipolysis (r = 0.29). Finally, as shown in Table 4, plasma HDL cholesterol was negatively correlated with the maximal isoproterenol lipolytic response (r = —0.52; P < 0.005) as well as with the basal and maximal epinephrineinduced lipolyses (r = —0.38 and r = -0.36, respectively; P < 0.05), but these associations were no longer significant when lipolysis was corrected for fat cell surface. The HDL cholesterol/LDL cholesterol ratio also displayed inverse relationships with isoproterenol- and epinephrine-stimulated lipolyses (r = —0.45 and r = —0.39 respectively; P < 0.05). However, these correlations did not remain significant after control for fat cell surface. Discussion The aim of the present study was to investigate the potential role of abdominal adipose cell lipolysis in the metabolic complications generally associated with abdominal obesity. Although it has become increasingly evident that excess abdominal fat is an important factor involved in the development of metabolic disturbances (3-5, 15-19, 38), the mechanisms responsible for the association between abdominal fat and metabolism are not fully understood. The increase in stimulated adipocyte lipolysis noted
1032
MAURIEGE ET AL.
with obesity suggests an elevated metabolic activity of abdominal adipose tissue. Fat cells are obviously larger in obese than in nonobese subjects (39), and lipolysis stimulated by catecholamines has been reported to be enhanced in large human adipocytes (40). Furthermore, cells from all depots in obese subjects are more lipolytically responsive than those from lean individuals, and lipolysis is more elevated in obese patients, particularly in individuals with abdominal obesity (41). In vitro, basal and catecholamine-stimulated lipolyses determined in isolated abdominal fat cells increased with the level of obesity, as shown by others (42). It seems reasonable to assume that fat cell size is a major correlate of adipose tissue metabolism and, thus, of hormone responsiveness (43). Our findings are in accordance with those of a previous study in which basal and catecholamine-stimulated lipolyses were positively correlated with fat cell size when lipolysis was expressed per cell number (44). It has already been suggested that the expression of lipolysis on a per cell basis can be misleading at times, as it does not discriminate between the effects due to changes in cell size per se and those caused by alterations related to the intrinsic properties of the cell (45). Our results suggest that the overall increase in lipolytic activity in obesity can be explained mostly by the higher fat cell size found in obese individuals (46). The mechanisms involved in the relationships between fat cell size and adipose tissue metabolism are still poorly understood. Nonetheless, it is attractive to consider that an increase in /3-adrenoceptor number as well as cAMP levels (44, 47) and/or a decrease in the a2-adrenergic component (45) may contribute to the faster rate of catecholamine-stimulated lipolysis in large compared to small adipocytes. Further studies are needed to test such a hypothesis, but it has been suggested that the a2-/(3adrenergic receptor status could be of importance (48). In the present study we noted that the physiological agonist epinephrine by itself exerted a lipolytic action. The absence of adenosine deaminase in the incubation medium could explain such an observation. Indeed, in the presence of adenosine deaminase, epinephrine inhibits lipolysis at lower concentrations, which may be of physiological relevance (i.e.
