¹Department of Medicine III, Okayama University Medical School, Okayama, Japan

²Faculty of Education, Okayama University, Okayama, Japan

³Okayama Southern Institute of Health, Okayama, Japan

⁴Department of Public Health, Okayama University Medical School, Okayama, Japan

Correspondence to: J Wada, Department of Medicine III, Okayama University Medical School, 2-5-1 Shikata-cho, Okayama 700-8558, Japan. E-mail: junwada@md.okayama-u.ac.jp

OBJECTIVE: To investigate whether the changes in IGF-I concentrations after weight reduction in Japanese overweight men are associated with changes in visceral and subcutaneous fat.

DESIGN: Cross-sectional and longitudinal clinical intervention study with exercise education.

SUBJECTS: One-hundred and twelve Japanese overweight men aged 30-59 y (body mass index (BMI) 28.4±2.5 kg/m²) and 33 normal-weight men aged 30-39 y (BMI 22.1±1.5 kg/m²) at baseline. From the participants, 56 randomly selected overweight men (BMI 28.8±2.8) were further enrolled into a 1 y exercise program.

MEASUREMENTS: Fat distribution was evaluated by visceral fat (V) and subcutaneous fat (S) areas measured with computed tomography scanning at umbilical levels, metabolic parameters and hormones including insulin, leptin and IGF-I at baseline and after 1 y.

RESULTS: In 112 overweight subjects at baseline, insulin (10.5±5.0 µU/ml) and leptin (6.4±3.7 ng/ml) significantly correlated with both V (r=0.260, P=0.0073; r=0.410, P<0.0001) and S areas (r=0.377, P<0.0001; r=0.613, P<0.0001), respectively. IGF-I (156.8±48.7 µU/ml) significantly and negatively correlated with V area (r=-0.242, P=0.0125) and age (r=-0.192, P=0.0480). In normal-weight men aged 30-39 y (n=33) and age-matched subjects (n=30) selected from the 112 overweight men, the serum IGF-I further tightly correlated with V area (r=-0.467, P<0.0001). Visceral fat area and age were independently related to serum IGF-I levels by multiple regression analysis. By intervention with exercise education, 56 overweight subjects showed an increase in daily steps (6224±2781 to 7898±4141 steps/day) and reduction of BMI (28.8±2.8 to 27.7±2.9). IGF-I significantly correlated with V area (r=-0.432, P=0.0009) but not with S area or BMI.

CONCLUSION: The present study indicated a negative correlation between IGF-I levels and visceral fat at baseline as well as an association between the reduction in visceral fat and increase in IGF-I levels after an exercise intervention.

International Journal of Obesity (2002) 26, 361-369. DOI: 10.1038/sj/ijo/0801899

insulin-like growth factor-I (IGF-I); visceral fat; exercise

Introduction

Excess of body weight increases the risk of death from any cause and from cardiovascular disease in adults between 30 and 74 y of age. The relative risk associated with greater body weight is higher among younger subjects.¹ In addition, the regional distribution of body fat is also recognized as a critical determinant for vascular complications. Accumulation of abdominal fat estimated from anthropometric variables, such as waist-to-hip ratio and waist circumference, is the major risk factor of cardiovascular diseases both in men and women.^2,3 It is now generally accepted that abdominal obesity is frequently associated with highly atherogenic metabolic complications, such as insulin resistance, hyperinsulinemia, hypertension, glucose intolerance and dyslipidemia.^4,5

In obesity, it has been reported that both basal and stimulated growth hormone (GH) secretion are reduced. Although GH stimulates insulin-like growth factor I (IGF-I) synthesis and release by the liver and other peripheral tissues, there are still controversies regarding IGF-I concentrations in obese subject, which have been reported to be normal,^6,7,8,9 reduced^{10,11,12,13,14,15,16} or high.^17,18,19 It is widely accepted that GH is a major determinant of serum IGF-I level; however, many factors which affect the IGF-I level have been reported. For example, IGF-I concentrations in obese subjects are negatively and independently associated with body mass index (BMI) as well as age.¹⁶ Furthermore, it has been shown that the elevated leptin in overnutrition, probably through increased insulin secretion, induces up-regulation of GH receptor and increases IGF-I secretion despite low GH secretion.^20,21

It is interesting to note that, in GH deficient obese adults, recombinant human GH (rhGH) treatment reduced visceral fat and metabolic abnormalities in patients with syndrome X.²² This report suggested that abnormality of the GH/IGF-I axis plays a pivotal role in the pathogenesis of abdominal obesity and its complications. To investigate such a role, we studied healthy overweight men aged 30-59 y who volunteered for the study. We measured IGF-I concentrations and visceral and subcutaneous fat areas with computed tomography (CT), and compared with hormonal variables, and clinical parameters at baseline and after a 1 y exercise intervention.

Subjects and methods

Subjects

Japanese overweight men (n=112), aged 30-59 y (44.9±7.4) and 30 normal-weight men, aged 30-39 y (34.7±2.5), were enrolled into this study with written informed consent. Overweight was diagnosed according to the criteria of WHO²³ and the average BMI of overweight and normal-weight subjects was 28.4±2.5 (25.1-39.9) and 22.1±1.5 (18.3-24.1), respectively. All subjects did not receive any medications for diabetes, hypertension and/or dyslipidemia throughout the observation period.

In a first cross-sectional analysis, we used baseline data on 112 overweight subjects, aged 30-59 y, and investigated the relationships between body fat distribution, IGF-I, IGFBP-3, leptin and insulin concentrations. From these participants, we selected overweight subjects aged 30-39 y (average 35.5±2.7 y, n=30) to compare with the age-matched normal-weight Japanese healthy volunteers (average 34.7±2.5 y, n=33).

In a second longitudinal analysis, we used follow-up data from 56 subjects (45.7±7.2 y) randomly selected from the 112 overweight participants, who met the following criteria: (1) no electrocardiogram changes in response to exercise; (2) a repeat CT scan at follow-up; (3) a follow-up duration of 1 y; (4) joining exercise program at Okayama Southern Institute of Health. They visited the Okayama Southern Institute of Health weekly and were monitored for 1 y. Daily steps were measured by pedometer (WZ100A, SEIKO Corporation, Japan) and the 7 days average was monitored every week throughout follow-up period. They were instructed to check daily steps every day and increase daily walking by at least 1000 steps compared to the daily walking at baseline. In addition, trained nutritionists determined total energy intake using food diaries before and after the 1 y follow-up. The aim of this study was to examine the IGF-I response to the reduction of visceral adipose tissues after regular exercise.

Blood pressure measurement, blood sampling and hormone assays

Blood pressure of each participant was measured twice after resting for at least 15 min in a sitting position. We measured overnight fasting serum levels of total cholesterol²⁴ and high density lipoprotein (HDL) cholesterol,²⁵ triglycerides (L Type Wako Triglyceride⋅H, Wako Chemical, Osaka, Japan), uric acid, liver enzymes (AST, ALT, -GTP), total insulin-like growth factor-I (IGF-I), IGF binding protein-3 (IGFBP-3), insulin, leptin and plasma glucose. Homeostasis model assessment of insulin resistance (HOMA-IR=(fasting insulin (µU/ml)´fasting glucose (mmol/l)/22.5) was calculated as an indicator of insulin resistance.²⁶ Insulin levels were measured by immunoradiometric assay (IRMA) using INSULIN RIABEADÒII (DAINABOT, Tokyo, Japan). Serum leptin was measured by human leptin radioimmunoassay (RIA) kit (Linco Research, St Charles, MO, USA). Total IGF-I was determined with IGF-I (somatomedin-C) IRMA 'Daiichi' (Daiichi Isotope Research, Tokyo, Japan). IGFBP-3 was measured by radioimmunoassay (RIA; Bioclone Australlia Pty Ltd, NSW, Australia).

Visceral and subcutaneous fat areas

The intra-abdominal visceral fat and the subcutaneous fat areas were measured by CT scan at the umbilical levels.²⁷ CT films were converted into digital images and both visceral and subcutaneous fat areas were measured with image analysis software OPTIMAS version 6.5 (Media Cybernetics, Silver Spring, MD, USA). The intraperitoneal area with same density as the subcutaneous fat layer was defined as visceral fat area.

Statistical analysis

All data are expressed as mean±standard deviation (s.d.) values. Differences between groups were examined for statistical significance using the Student's t-test (for paired or unpaired data). The correlations between fat distribution (S and V areas) and the quantitative variables, ie age, BMI, fasting plasma glucose, insulin, HOMA-IR, leptin, IGF-I and IGFBP-3 were studied by linear regression analysis using Pearson coefficient. In addition, IGF-I and the variables, ie S and V areas, age, BMI, fasting plasma glucose, insulin, HOMA-IR, leptin and IGFBP-3, were also examined. To further determine the variables independently associated with IGF-I levels, a stepwise multiple regression analysis was performed. All parameters were used as independent variables. P-values less than 0.05 were considered statistically significant. The data were analyzed with the Statistical Package for the Social Science Software Release 6.x.

Results

Baseline clinical characteristics of overweight subjects

The clinical characteristics of 112 overweight subjects are shown in Table 1. The visceral fat (V) area was 105.5±46.9 cm² and the subcutaneous fat (S) area was 149.3±52.4 cm² in overweight subjects. The presence of systolic blood pressure>140 mmHg and/or diastolic pressure >90 mmHg was noted in 40 overweight men.

Simple correlations of fat distribution with age, BMI, insulin, HOMA-IR, leptin, IGF-I, and IGFBP-3

The fasting serum insulin levels positively correlated with both S area (r=0.377, P<0.0001) and V area (r=0.260, P=0.007). HOMA-IR was found to positively correlate with S area (r=0.378, P<0.0001), but not with V area (r=0.177, P=0.0705) (Figure 1, Table 2). The fasting serum leptin concentration showed positive correlations with both S area (r=0.613, P<0.0001) and V area (r=0.410, P<0.0001; Figure 2, Table 2). IGF-I levels were negatively correlated with V area (r=-0.242, P<0.0125); however, significant correlation with S area was not found (Figure 3, Table 2). The serum IGFBP-3 did not reveal significant correlation with either S or V areas. The fasting serum leptin concentration significantly correlated with the fasting serum insulin (r=0.484, P<0.0001) and HOMA-IR (r=0.453, P<0.0001), however it was not related to the serum IGF-I concentration.

Visceral fat area is correlated with serum IGF-I by linear regression analysis

Serum IGF-I levels were significantly and negatively related with age (r=-0.192, P=0.0480) and V area (r=-0.242, P=0.0125; Table 2). Since serum IGF-I levels gradually decrease with age, we investigated the relationship between the serum IGF-I and V area among age-matched normal-weight subjects (average 35.5±2.7 y, n=33) and overweight subjects (average 34.7±2.5 y, n=30). The clinical characteristics of normal-weight and overweight subjects are shown in Table 3. The serum IGF-I levels of the overweight subjects were significantly lower than the normal-weight subjects; in turn the IGFBP-3 levels of the overweight subjects were significantly higher than the normal-weight subjects. In age-matched overweight and normal-weight subjects, the serum IGF-I negatively correlated with V area (r=-0.467, P<0.0001). Although serum IGF-I in overweight subjects was significantly lower compared with normal-weight subjects, linear analysis revealed that IGF-I did not significantly correlate with BMI (Figure 4).

Visceral fat area and age are independently related to serum IGF-I by multiple regression analysis

The linear regression analyses were followed by a multiple regression analysis using IGF-I as the dependent variables to further analyze the significant predictors (Table 4). In model 1, all 112 Japanese overweight men were analyzed. Age, V area and IGFBP-3 were independently related to serum IGF-I levels. All other variables including BMI, S area, FPG, insulin, HOMA-IR and leptin did not enter the equation at significant levels. In model 2, 30 age-matched Japanese overweight men and 33 normal-weight men are analyzed. Age and V area were independently related to serum IGF-I levels. All other variables did not enter the equation.

Reduction of V area is associated with elevation of serum IGF-I level

At follow-up after 1 y of increased daily walking, 56 obese subjects repeated the same examinations as baseline. After exercise education by instructors, daily steps significantly increased at 4 months and maintained until end of the observation period (Figure 5). Their BMI decreased from 28.8±2.8 to 27.7±2.9 (kg/m²), V area from 109.8±53.7 to 92.1±48.9, and S area 152.3±50.5 to 130.4±52.5 (cm²). Total energy intake also significantly decreased from 2023±474 kcal to 1897±415 kcal. V area significantly correlated with IGF-I (r=-0.432, P=0.0009), while both S area and BMI did not show significant relations with IGF-I (r=-0.210, P=0.1199; r=-0.164, P=0.2281, respectively; Figure 6).

Discussion

The evaluation of total IGF-I levels in obese subjects has yielded conflicting results, with normal,^6,7,8,9 low^{10,11,12,13,14,15,16} or high^17,18,19 serum concentrations. Studies on selected obese patients have shown that low IGF-I levels are observed in those with abdominal obesity associated with higher serum concentrations of non-esterified fatty acids, and negative correlations were found between IGF-I levels and the amount of visceral fat.^12,13 The studies of larger numbers of obese and non-obese subjects indicated that IGF-I concentrations were negatively correlated with age and BMI; however, they were not associated with fat distribution, ie waist-hip ratio.¹⁵ It is now generally accepted that the total IGF-I concentrations are normal or slightly reduced within the normal range in obesity despite the marked impairment of both basal and stimulated GH secretion.²⁹ No significant relation between fat distribution and IGF-I levels in the studies with large numbers of subjects may be due to inaccurate assessment of fat distribution such as waist-hip ratio. Thus we performed CT scans on a large number of Japanese overweight and normal-weight subjects and investigated the relation between abdominal fat accumulation and IGF-I levels. Since the obesity-related complications such as diabetes and atherosclerotic vascular disease also influenced the serum IGF-I levels in previous studies, only overweight Japanese men who were not taking any medications for diabetes, hypertension or other vascular diseases were enrolled into the study.

The main finding in the study is the inverse relationship between total IGF-I levels and visceral fat area in Japanese overweight men (n=112). By evaluating the age-matched normal-weight and overweight subjects aged 30-39 y (n=63), IGF-I levels are further closely associated with V area; however, no significant relation with BMI or S area is observed. In addition, visceral fat area and age are independently related to serum IGF-I by multiple regression analysis. This evidence suggests that a reduced IGF-I serum level is tightly related to the accumulation of abdominal fat. Furthermore, the longitudinal follow-up study also supported this notion, since V area was significantly correlated with IGF-I after the 1 y exercise program.

Recent work has suggested that overnutrition causes hyperinsulinemia, which increases GH-receptor and IGF-I secretion despite low GH secretion.³⁰ In addition, a decreased binding of IGF-I to its receptor has been reported,¹⁹ and thus obesity is characterized by increased sensitivity to GH as well as resistance to IGF-I.²⁹ Hypoactivity of IGF-I, as observed in patients with GH deficiency, causes metabolic alterations such as insulin resistance, hypertriglyceridemia and low HDL-cholesterol associated with accumulation of visceral adipose tissue. Furthermore, rhGH therapy restores biochemical, metabolic alterations and even reduces the visceral adipose tissue mass.²² The GH/IGF-I system plays a predominant role in muscle strength and has favorable effects on body composition³¹ and GH has direct lipolytic effects and indirect anabolic and growth promoting effects mediated by IGF-I. Such observations suggested that the GH/IGF-I axis plays a pivotal role in metabolic syndrome X and vascular complications observed in obese subjects.

Recently, it has been reported that measurement of the free fraction of IGF-I demonstrates higher levels in obese patients compared with lean subjects despite a normal or reduced level of total IGF-I.^32,33,34 This is probably explained by a reduced level of IGFBP-1 and IGFBP-2,^32,33,34,35 whose expression is inhibited by insulin, that is generally increased in obese subjects. In contrast, a high or normal serum level of IGFBP-3 has been reported in obese subjects.^34,35 Since GH and insulin were shown to positively control IGFBP-3, an elevated level of IGFBP-3 may reflect increased sensitivity to GH and increased serum level of insulin in obese subjects. Leptin is another regulator of the GH-IGF-I axis. In rodents, leptin stimulated GH release³⁶ and in human elevated serum leptin in hypopituitary patients was reduced by GH treatment.^37,38 Thus an elevated leptin level and low GH level would fit the concept of a leptin resistance in obese subjects.²⁹

The estimation of visceral adipose fat, ie V area, is well correlated with atherogenic risk factors such as hypertension, reduced HDL and elevated triglyceride levels. The measurement of V area seems to be clinically useful because it is tightly linked to atherogenic risk factors. Exercise was associated with a decrease in abdominal fat mass and an increase in IGF-I levels, and thus it serves as a valuable follow-up marker in the course of obesity treatment. It also suggested that GH-IGF-I may be involved in abdominal obesity and the possibility is raised of using rhGH or IGF-I for the reduction of abdominal fat accumulation in addition to exercise and diet therapy.

Acknowledgements

This study was approved by Ministry of Health, Labour and Welfare, Japan. This research was supported by Health Science Research Grants for 'Research on Health Services' from the Ministry of Health, Labour and Welfare, Japan. It was partly supported by Uehara Memorial Foundation, The Naito Foundation, ONO Medical Foundation to J Wada and Uehara Memorial Foundation to H Makino.

1 Stevens J, Cai J, Pamuk ER, Williamson DF, Thun MJ, Wood JL. The effect of age on the association between body-mass index and mortality. New Engl J Med 1998; 338: 1-7. MEDLINE

2 Kannel WB, Cupples LA, Ramaswami R, Stokes J III, Kreger BE, Higgins M. Regional obesity and risk of cardiovascular disease; Framingham study. J Clin Epidemiol 1991; 44: 183-190. MEDLINE

3 Rexrode KM, Carey VJ, Hennekens CH, Walters EE, Colditz GA, Stampfer MJ, Willett WC, Manson JE. Abdominal adiposity and coronary heart disease in women. JAMA 1998; 280: 1843-1848. MEDLINE

4 Poiliot MC, Despres JP, Nadeeau A, Moorjani S, Prud'Homme D, Lupien PJ, Tremblay A, Bouchard C. Visceral obesity in men. Associations with glucose tolerance, plasma insulin, and lipoprotein levels. Diabetes 1992; 41: 826-834. MEDLINE

5 Boyko EJ, Leonetti DL, Fujimoto WY, Newell-Morris L. Visceral adiposity and risk of type 2 diabetes. Diabetes Care 2000; 23: 465-471. MEDLINE

6 Caufriez A, Goldstein J, Lebrun P, Herchuelz A, Furlanetto R, Copinschi G. Relations between immunoreactive somatomedin C, insulin and T3 patterns during fasting in obese subjects. Clin Endocrinol 1984; 20: 65-70.

7 Nam SY, Kim KR, Song YD, Lim SK, Lee HC, Huh KB. GH-binding protein in obese men with varying glucose tolerance: relationship to body fat distribution, insulin secretion and the GH-IGF-I axis. Eur J Endcrinol 1999; 140: 159-163.

8 Frystyk J, Skjærbæk, Vestbo E, Fisker S, Ørskov H. Circulating levels of free insulin-like growth factors in obese subjects: the impact of type 2 diabetes. Diabetes Metab Res Rev 1999; 15: 314-322. Article MEDLINE

9 Nyomba BLG, Johnson M, Berard L, Murphy LJ. Relationship between serum leptin and the insulin-like growth factor-I system in humans. Metabolism 1999; 48: 840-844. MEDLINE

10 Copeland KC, Colletti RB, Devlin JT, McAuliffe TL. The relationship between insulin-like growth factor-I, adiposity, and aging. Metabolism 1990; 41: 106-112.

11 Yamamoto H, Kato Y. Relationship between plasma insulin-like growth factor I (IGF-I) levels and body mass index (BMI) in adults. Endcrinol J 1993; 40: 41-45.

12 Marin P, Kvist H, Lindstedt G, Sjöstrom L, Björntörp P. Low concentrations of insulin-like growth factor I in abdominal obesity. Int J Obes Relat Metab Disord 1993; 17: 83-89. MEDLINE

13 Rasmussen MH, Frystyk J, Andersen T, Breum L, Christiansen, Hilsted J. The impact of obesity, fat distribution, and energy restriction on insulin-like growth factor-I (IGF-I). IGF-binding protein-3, insulin, and growth hormone. Metabolism 1994; 43: 315-319. MEDLINE

14 Rasmussen MH, Hvidberg A, Juul A, Main KM, Gotfredsen A, Skakkebae NE, Hilsted J. Massive weight loss restores 24 hour growth hormone release profiles and serum insulin-like growth factor-I levels in obese subjects. J Clin Endocrinol Metab 1995; 80: 1407-1415. MEDLINE

15 Iranmanesh A, South S, liem AY, Clemmons D, Thorner MO, Weltman A, Veldhuis JD. Unequal impact of age, percentage body fat, and serum testosterone concentrations on somatotropic, IGF-I, and IGF-binding protein responses to a three-day intravenous growth hormone-releasing hormone pulsatile infusion in men. Eur J Endcrinol 1998; 139: 59-71.

16 Maccario M, Ramunni J, Oleandri SE, Procopio M, Grottoli S, Rossetto R, Savio P, Aimaretti G, Camanni F, Ghigo E. Relationships between IGF-I and age, gender, body mass, fat distribution, metabolic and hormonal variables in obese patients. Int J Obes Relat Metab Disord 1999; 23: 612-618. MEDLINE

17 Van Vliet G, Bosson D, Rummens E, Robyn C, Wolter R. Evidence against growth hormone-releasing factor deficiency in children with idiopathic obesity. Acta Endcrinol 1986; 279: 403-410.

18 Loche S, Cappa M, Borrelli P, Faedda A, Crinó A, Cella SG, Corda R, Müller EE, Pintor C. Reduced growth hormone response to growth hormone-releasing hormone in children with simple obesity: evidence for somatomedin-C mediated inhibition. Clin Endocrinol 1987; 27: 145-153.

19 Hochberg Z, Hertz P, Colin V, Ish-Shalom S, Yeshurun D, Youdim MBH, Amit T. The distal axis of growth hormone (GH) in nutritional disorders: GH-binding protein, insulin-like growth factor (IGF-I), and IGF-I receptors in obesity and anorexia nervosa. Metabolism 1992; 41: 106-112. MEDLINE

20 Yasunaga T, Furukawa S, Katsumata N, Horikawa R, Tanaka T, Tanae A, Hibi I. Nutrition related hormonal changes in obese children. Endocr J 1998; 45: 221-227. MEDLINE

21 Fernandez-Real JM, Granada ML, Ruzafa A, Casamitjana R, Ricart W. Insulin sensitivity and secretion influence the relationship between growth hormone-binding-protein and leptin. Clin Endcrinol 2000; 52: 159-164.

22 Johannsson G, Marin P, Lonn L, Ottosonn M, Stenlof K, Björntorp P, Sjöstrom L, Bengtsson B. Growth hormone treatment of abdominally obese men reduces abdominal fat mass, improves glucose and lipoprotein metabolism, and reduces diastolic blood pressure. J Clin Endocrinol Metab 1997; 82: 727-734. MEDLINE

23 . NHBI obesity education initiative expert panel on the identification, evaluation, and treatment of overweight and obesity in adults. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults¾the evidence report. Obes Res 1998; 6: 51S-63S.

24 Richimond W. Preparation and properties of a cholesterol oxidase from Nocardia sp. and its application to enzymatic assay of total cholesterol in serum. Clin Chem 1973; 19: 1350-1356. MEDLINE

25 Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. High density lipoprotein as a protective factor against coronary heart disease. The Framingham study. Am J Med 1977; 62: 707-714. MEDLINE

26 Haffner SM, Gonzalez C, Miettinen H, Kennedy E, Stern MP. A prospective analysis of the HOMA model: the Mexico City Diabetes Study. Diabetes Care 1996; 19: 1138-1141. MEDLINE

27 Tokunaga K, Matsuzawa Y, Ishikawa K, Tarui S. A novel technique for the determination of body fat by computed tomography. Int J Obes Relat Metab Disord 1983; 75: 437-445.

28 Scacchi M, Pincelli AI, Cavagnini F. Growth hormone in obesity. Int J Obes Relat Metab Disord 1999; 23: 260-271. MEDLINE

29 Lanzi R, Luzi L, Caumo A, Andreotti AC, Manzoni MF, Malighetti ME, Sereni LP, Pontiroli AE. Elevated insulin levels contribute to the reduced growth hormone (GH) response to GH-releasing hormone in obese subjects. Metabolism 1999; 9: 1152-1156.

30 Salomon F, Cuneo RC, Hesp R, Sonksen PH. The effects of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. New Engl J Med 1989; 321: 1797-1803. MEDLINE

31 Frystyk J, Vestbo E, Skjærbæk, Mogensen CE, Ørskov H. Free insulin-like growth factors in human obesity. Metabolism 1995; 44: (Suppl 10) 37-44. MEDLINE

32 Nam SY, Lee EJ, Kim KR, Cha BS, Song YD, Lim SK, Lee HC, Huh KB. Effect of obesity on total and free insulin-like growth factor (IGF-I), and their relationship to IGF-binding protein (BP)-1, IGFBP-2, IGFBP-3, insulin, and growth hormone. Int J Obes Relat Metab Disord 1997; 21: 355-359. MEDLINE

33 Argente J, Caballo N, Barrios V, Pozo J, Muóoz MT, Chowen JA, Hernandez M. Multiple endocrine abnormalities of the growth hormone and insulin-like growth factor axis in prepubertal children with exogenous obesity: effect of short- and long-term weight reduction. J Clin Endocrinol Metab 1997; 82: 2076-2083. MEDLINE

34 Weaver JU, Holly JM, Kopelman PG, Noonan K, Giadom CG, White N, Virdee S, Wass JA. Decreased sex hormone binding globulin (SHBG) and insulin-like growth factor binding protein (IGFBP-1) in extreme obesity. Clin Endocrinol 1990; 33: 415-422.

35 Carro E, Senaris R, Considine RV, Casanueva FF, Dieguez C. Regulation of in vivo growth hormone secretion by leptin. Endocrinology 1997; 138: 2203-2206. MEDLINE

36 Florkowski CM, Collier GR, Zimmet PZ, Livesey JH, Espiner EA, Donald RA. Low-dose growth hormone replacement lowers plasma leptin and fat stores without affecting body mass index in adults with growth hormone deficiency. Clin Endocrinol 1996; 47: 191-198.

37 Nystrom F, Ekman B, Osterlund M, Lindstrom T, Ohman KP, Arnqvist HJ. Serum leptin concentrations in a normal population and in GH deficiency: negative correlation with testosterone in men and effects of GH treatment. Clin Endocrinol 1997; 47: 191-198.

38 Schwander JC, Hauri C, Zapf J, Froesch ER. Synthesis and secretion of insulin-like growth factor and its binding protein by the perfused rat liver: dependence on growth hormone status. Endocrinology 1983; 113: 297-305. MEDLINE

Figure 1 Correlation between insulin resistance and fat distribution in overweight subjects (n=112). The fasting serum IRI levels significantly correlates with subcutaneous fat (S) area (r=0.377, P<0.0001) and visceral fat (V) area (r=0.260, P=0.0073). HOMA-IR (homeostasis model assessment of insulin resistance) significantly correlates with S area (r=0.378, P<0.0001); however, significant correlation between HOMA-IR and V area is not observed (r=0.177, P=0.0705).

Figure 2 Correlation between leptin and fat distribution in overweight subjects (n=112). The fasting serum leptin significantly correlates with both subcutaneous fat (S) area (r=0.613, P<0.0001) and visceral fat area (r=0.410, P<0.0001).

Figure 3 Correlation between IGF-I and fat distribution in overweight subjects (n=112). The serum IGF-I level negatively correlates with visceral (V) area (r=-0.242, P=0.0125); however, it is not related to subcutaneous (S) area (r=-0.042, P=0.6858).

Figure 4 Serum level of IGF-I and fat distribution in age-matched overweight (n=30) and normal-weight subjects (n=33). The serum IGF-I significantly correlates with visceral fat (V) area (r=-0.467, P=0.0002), while serum IGF-I levels are not related to BMI.

Figure 5 Time course change of BMI and daily steps in Japanese overweight men with exercise education program (n=56). Arrows indicate the initiation point of the exercise education program (0 month). Data are means±standard error; * P<0.05 vs 0 month.

Figure 6 Correlations among BMI, S area, V area, and IGF-I. Overweight subjects (n=56) were enrolled into exercise education, and their BMI was decreased from 28.8±2.8 to 27.7±2.9 kg/m², V area from 109.8±53.7 to 92.1±48.9 cm² and S area 152.3±50.5 to 130.4±52.5 cm². V area significantly correlated with IGF-I (r=-0.432, P=0.0009), while both S area and BMI did not show significant relationships has with IGF-I (r=-0.210, P=0.1196, r=-0.164, P=0.2269, respectively).

Table 1 Clinical characteristics of Japanese overweight subjects (n=112)

Table 2 Simple correlations of body fat distribution with metabolic and hormonal parameters of Japanese overweight subjects

Table 3 Clinical characteristics in age-matched normal-weight and overweight subjects and simple correlations of visceral fat (V), subcutaneous fat (S) area, and BMI with metabolic and hormonal parameters

Table 4 Multiple linear regression analysis of overweight and normal-weight Japanese subjects using IGF-I as dependent variables