Point-Counterpoint SEE MILES AND JENSEN (P. 2326)

Point: Visceral Adiposity Is Causally Related to

he relationship between and neous or total mass was sig- adipose tissue pool into a retroperitoneal insulin resistance, while well recog- nificantly correlated in multivariate adipose tissue compartment that presum- T nized for many years, has nonethe- analyses with insulin resistance, type 2 di- ably drains into the systemic circulation less been confusing since not all obese abetes, and cardiovascular events (9– and an intraperitoneal compartment that individuals have insulin resistance (1) 11,13,14,16,19,20). The hypothesis that drains into the portal circulation (23). and because insulin resistance occurs in visceral adiposity and not total adiposity Other investigators have separated the ab- individuals who have BMIs that are within was the cause of the components and clin- dominal subcutaneous adipose tissue into the normal or mildly catego- ical consequences of the metabolic syn- deep and superficial compartments and ries (2). drome has been challenged by several found that the superficial compartment Early attempts to understand the re- studies that found either that abdominal had no relationship with insulin resis- lationships between obesity, type 2 diabe- subcutaneous adipose tissue mass and not tance, whereas the deep compartment tes, and focused on visceral adipose tissue mass was indepen- was as highly correlated with insulin re- the waist-to-hip ratio as a means of distin- dently correlated with insulin resistance sistance as the visceral compartment (17). guishing those individuals who were at or that they were both equally correlated While the results of the studies from increased risk from those who were not (8,12,17,18,21). Studies using isotopic various investigators differ, there is re- (3,4). A high waist-to-hip ratio is a surro- techniques have calculated that the ma- markable consistency of results from the gate for masculine distribution of obesity jority of circulating free fatty acids are de- same investigator. The studies of Fuji- (central obesity). Cross-sectional studies rived from peripheral adipose tissue (22) moto and colleagues (13,16) in Japanese by Kissebah et al. (5) and Krotkiewski et and that if free fatty acids are postulated to Americans over the last 10 years show al. (6) in the 1980s demonstrated that hy- be responsible for the insulin resistance that visceral adiposity is associated with pertension, hypertriglyceridemia, hyper- associated with obesity, then total adipos- and predicts the development of type 2 insulinemia, and glucose intolerance ity rather than visceral adiposity should , coronary heart disease, and hy- were increased in subjects with a high be responsible for insulin resistance. pertension. The numerous studies ema- waist-to-hip ratio. Long-term longitudi- Several lines of evidence support the nating from the laboratories of Despre´s nal population-based studies of men hypothesis that visceral adipose tissue and Bouchard (9,10) since 1990 have re- (13.5 years) and women (12 years) in and not subcutaneous adipose tissue is producibly shown that visceral adiposity Gothenburg, Sweden, showed that the the major contributor in causing insulin is the component of adiposity that is the waist-to-hip ratio was a predictor of the resistance and the . independent risk factor for insulin resis- future development of diabetes, myocar- They are discussed in the below sections. tance, glucose intolerance, and cardiovas- dial infarction, angina pectoris, stroke, cular risk factors. Abate and Garg and and independent of BMI (3,4). Correlation between adipose tissue their associates (8,23) have devised a pro- Technology developed in the 1990s, depots and insulin resistance cedure for measuring regional adiposity including computer tomography scans Numerous investigators have examined by multiple MRIs and consistently find and MRI, made it possible to precisely the relationship between insulin resis- that it is abdominal subcutaneous adipose measure specific adipose tissue depots tance, as measured by the euglycemic- tissue that is associated with insulin resis- such as total body adipose tissue mass, hyperinsulinemic clamp, and various tance. Goodpaster et al. (12) reported in abdominal subcutaneous adipose tissue adipose tissue depots, such as total adi- 1997 that subcutaneous abdominal fat mass, visceral adipose tissue mass, and posity, visceral adipose tissue, abdominal predicted insulin sensitivity indepen- hepatic and intramuscular triglyceride subcutaneous adipose tissue, and in- dently of visceral fat. However, when they content (7,8). Utilizing those techniques, tramyocellular triglyceride. The tech- subsequently examined the effects of many studies have examined the relation- niques used to estimate the various pools on regional fat distribution ship between total adipose tissue, abdom- have varied considerably. The most com- and insulin sensitivity in obesity, they inal subcutaneous adipose tissue, and monly used technique is a computed to- found that the reduction in visceral adi- visceral adipose tissue mass and insulin mography scan or a magnetic resonance posity was the only adipose tissue param- resistance, the components of the meta- imaging (MRI) done at one level, such as eter that predicted the improvement in bolic syndrome, and the development of L4-5. The area of the visceral adipose tis- insulin sensitivity (15). Banerji and col- or clinical cardiovascular sue is then compared with the area of the leagues (11,14), using multiple abdomi- events (9–21). subcutaneous adipose tissue. The most nal computed tomography slices to Despite the much smaller size of the rigorous technology involves scanning measure visceral and abdominal subcuta- visceral adipose tissue depot compared the abdomen with multiple slices using a neous adipose tissue volumes and eugly- with the total subcutaneous adipose tis- computer program to calculate the total cemic-hyperinsulinemic clamps to sue depot or total adiposity, many inves- volume or mass of the visceral and ab- measure insulin sensitivity, have found tigations demonstrated that the visceral dominal subcutaneous depots. Some in- visceral adiposity to be independently as- adipose tissue mass and not the subcuta- vestigators have separated the visceral sociated with insulin resistance in type 2

2322 DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 Lebovitz and Banerji diabetic African-American men and It is thought to be due to developmental or any of the lipid or inflammatory com- women and in nondiabetic Southeast- abnormalities in the hypothalamus. Of ponents of the metabolic syndrome. Asian men. Visceral obesity has been re- particular interest has been the observa- ported to be an important correlate of sex tions that they don’t seem to have many of Peroxisome proliferator–activated differences in cardiovascular disease risk the obesity-related metabolic abnormali- receptor ␥ agonists increase (24). ties. Goldstone et al. (27) compared total abdominal subcutaneous adipose In obese children and adolescents and regional adipose tissue depots in pa- tissue mass but decrease insulin with impaired glucose tolerance, in- tients with the Prader-Willi syndrome resistance and improve most of the tramyocellular and visceral lipid deposits with comparable obese individuals and components of the metabolic are increased and the abdominal subcuta- nonobese control subjects. BMI (36.6 Ϯ syndrome neous adipose tissue depot is decreased, 9.9 kg/m2), total adipose tissue volume Pioglitazone, rosiglitazone, and troglita- as indicated by MRI. These abnormalities (55.7 Ϯ 25.8 l), percent body fat (46.9 Ϯ zone have been shown to increase total are associated with the development of 6.9), and abdominal subcutaneous adi- body fat and abdominal subcutaneous ad- severe peripheral insulin resistance (25). pose tissue as percent of total body fat ipose tissue while having no effect on or (25.9 Ϯ 3.2) were the same in the Prader- slightly decreasing visceral adipose tissue. Absence of insulin resistance in Willi syndrome patients as in the obese Despite this increase in overall obesity, overweight and obese individuals subjects. In contrast, their visceral adi- the thiazolidinediones improve insulin with lower body obesity pose tissue mass (5.7 Ϯ 1.3%) was the sensitivity and improve all of the compo- Individuals with lower body obesity have same as the nonobese individuals (5.7 Ϯ nents of the metabolic syndrome (29). an increase in total adiposity that is 1.7%) and ϳ60% that of the obese indi- Plasma free fatty acid levels decrease predominately in subcutaneous adipose viduals (9.1 Ϯ 1.6%). The values of met- ϳ20–25%, and hepatic fat decreases 25– tissue; visceral adipose tissue is not signif- abolic factors that characterize the obese 50%. icantly affected. phenotype, such as plasma insulin levels, This sequence of events suggests that In near-normoglycemic African- insulin sensitivity, and plasma triglycer- insulin resistance and the components of American men with the same BMI (ϳ26.5 ide levels, were proportional to visceral the metabolic syndrome are the conse- kg/m2), insulin resistance is present in adipose tissue mass and not to total fat quence of the metabolic effects of the those with increased visceral adipose tis- mass. In the Prader-Willi syndrome, the products being released from the adipose sue mass (4.71 Ϯ 1.85 l) but not in those increase in total fat mass does not cause tissue rather than an effect of the absolute with normal visceral adipose tissue mass insulin resistance and the metabolic mass of the tissue. Adipose tissue releases (2.49 Ϯ 2.3 l) (26). syndrome. free fatty acids and cytokines (e.g., tumor Lemieux et al. (10) followed a cohort necrosis factor ␣) and modulates the se- of women for 7 years to assess changes in Liposuction removal of abdominal cretion of a large number of metabolically total obesity, regional adiposity, and mea- subcutaneous adipose tissue in active adipokines (e.g., adiponectin). It is sures of glucose and insulin homeostasis. obese nondiabetic and type 2 the net effect of these factors that appears They compared two subgroups of women diabetic subjects to be responsible for the metabolic abnor- with similar mean increases in body fat If the metabolic effects of obesity are due malities of obesity. mass but with either small or large in- to total body fat and contributed to by the creases in visceral fat mass. The greatest abdominal subcutaneous adipose tissue, Reflections deterioration of glucose tolerance and in- the removal of a large mass of abdominal We have presented evidence that we be- creases in insulin secretion occurred in subcutaneous adipose tissue should re- lieve supports the thesis that increases in those with the large increase in visceral sult in some amelioration of insulin resis- visceral adipose tissue are the conduit by adiposity. In contrast, when they com- tance and components of the metabolic which obesity leads to insulin resistance pared two subgroups with similar in- syndrome. Klein et al. (28) studied 15 and the metabolic syndrome. Further val- creases in visceral fat mass but with small obese patients (8 nondiabetic control idation requires many additional pieces of or large changes in body fat mass, they subjects and 7 type 2 diabetic patients) the puzzle to be unraveled and explained. found no difference in the changes in glu- before and 10–12 weeks after liposuction Visceral adipose tissue is biochemically cose tolerance and insulin secretion be- of abdominal subcutaneous adipose tis- quite distinct from subcutaneous adipose tween the groups. They concluded that sue. The type 2 diabetic patients had a tissue, as has been reviewed by several visceral fat mass and not total fat mass was mean baseline BMI of 39.9 Ϯ 5.6 kg/m2 authors (30), and it is unclear as to how the determining factor in regulating glu- and had 10.5 Ϯ 3.3 kg fat removed, the large number of adipokines, cyto- cose homeostasis. which was 28% of the depot. The nondi- kines, and enzymes as diverse as adiponec- abetic subjects had a baseline BMI of tin and 11␤-hydroxysteroid dehydrogenase Adipose tissue composition and 35.1 Ϯ 2.4 kg/m2 and had 9.1 Ϯ 3.7 kg type 1 are regulated and integrated. Sev- metabolic status in patients with the fat removed, which was 44% of the sub- eral recent studies involving nondiabetic Prader-Willi syndrome cutaneous abdominal adipose tissue. Asian Indians suggest that regional adi- The Prader-Willi syndrome is a genetic There was no loss of visceral adipose tis- pose tissue metabolic activity may be disorder characterized by hyperphagia, sue. In both groups, the removal of the quantitatively different depending on the marked obesity from childhood, mental abdominal subcutaneous fat had no effect ethnic background of the individual. For retardation, short stature from growth on insulin resistance, plasma glucose or the same fat mass and regional distribu- hormone deficiency, and hypogonadism. insulin levels, plasma adiponectin levels, tion, Asian Indians have more insulin re-

DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 2323 Point-Counterpoint sistance, higher plasma free fatty acids, Evans PJ, Hartz AJ, Kalkhoff RK, Adams Metab 278:E941–E948, 2000 higher C-reactive protein and plasmino- PW: Relation of body fat distribution to 18. Raji A, Seely EW, Arky RA, Simonson DC: gen activator inhibitor-1, and lower metabolic complications of obesity. J Clin Body fat distribution and insulin resis- plasma adiponectin than Caucasians (31). Endocrinol Metab 54:254–260, 1982 tance in healthy Asian Indians and Cau- The relationship between visceral adipose 6. Krotkiewski M, Bjorntorp P, Sjostrom I, casians. J Clin Endocrinol Metab 86:5366– Smith W: Impact of obesity on metabo- 5371, 2001 tissue and abdominal subcutaneous tis- lism in men and women: importance of 19. Rendell M, Hulthen UL, Tornquist C, sue may be more closely linked than has regional adipose tissue. J Clin Invest 72: Groop L, Mattiasson I: Relationship be- been thought. There are a few studies sug- 1150–1162, 1983 tween abdominal fat compartments and gesting that increases in visceral adipose 7. Kvist H, Sjostrom L, Tylen U: Adipose tis- glucose and lipid metabolism in early tissue mass or activity may alter subcuta- sue volume determinations in women by postmenopausal women. J Clin Endocrinol neous adipose cell lipolysis. computed tomography: technical consid- Metab 86:744–749, 2001 It has become increasingly evident erations. Int J Obes 10:53–67, 1986 20. Ross R, Freeman J, Hudson R, Janssen I: that the effects of visceral adiposity must 8. Abate N, Garg A, Peshock RM, Stray-Gun- Abdominal obesity, muscle composition, be mediated by multiple factors. Free fatty dersen J, Grundy SM: Relationship of gen- and insulin resistance in premenopausal acid release from visceral adipose tissue in eralized and regional adiposity to insulin women. J Clin Endocrinol Metab 87:5044– sensitivity in men. J Clin Invest 96:88–98, 5051, 2002 obese individuals has been reported to ac- ϳ 1995 21. Miyazaki Y, Glass L, Triplitt C, Wajcberg count for 20% of that delivered to the 9. Pouliot MC, Despre´s J-P, Nadeau A, E, Mandarino LJ, DeFronzo RA: Abdom- liver, although in an individual it can be as Moorjani S, Prud’Homme D, Lupien PJ, inal fat distribution and peripheral and high as 50% (22). As Klein (32) has com- Tremblay A, Bouchard C: Visceral obesity hepatic insulin resistance in type 2 diabe- mented in a recent editorial, it is unlikely in men: associations with glucose toler- tes mellitus. Am J Physiol Endocrinol Metab that the contribution that visceral fat ance, plasma insulin, and lipoprotein lev- 283:E1135–E1143, 2002 makes to the free fatty acid flux in the els. Diabetes 41:826–834, 1992 22. Nielsen S, Guo Z-K, Johnson M, Hensrud circulation is its major mechanism in 10. Lemieux S, Tremblay A, Prud-Homme D, DD, Jensen MD: Splanchnic lipolysis in causing insulin resistance. Bouchard C, Nadeau A, Despre´s J-P: Sev- obesity. J Clin Invest 113:1582–1588, Adiponectin and other newly discov- en-year changes in body fat and visceral 2004 adipose tissue in women: associations 23. Garg A: Regional adiposity and insulin re- ered adipokines are likely to contribute. with indexes of plasma glucose-insulin sistance. J Clin Endocrinol Metab 89:4206– homeostasis. Diabetes Care 19:983–991, 4210, 2004 HAROLD E. LEBOVITZ, MD 1996 24. Lemieux S, Despre´s J-P, Moorjani S, MARY ANN BANERJI, MD, FACP 11. Banerji MA, Lebowitz J, Chaiken RL, Gor- Nadeau A, Theriault G, Prud’homme D, don D, Kral JG, Lebovitz HE: Relationship Tremblay A, Bouchard C, Lupien PJ: Are From the Division of Endocrinology, Department of of visceral adipose tissue and glucose dis- gender differences in cardiovascular dis- Medicine, State University of New York Health Sci- posal is independent of sex in black ease risk factors explained by the level of ence Center at Brooklyn, Brooklyn, New York. NIDDM subjects. Am J Physiol 273:E425– visceral adipose tissue? Diabetologia Address correspondence and reprint requests to E432, 1997 37:757–764, 1994 Harold E. Lebovitz, 416 Henderson Ave., Staten Is- 12. Goodpaster BH, Thaete FL, Simoneau 25. Weiss R, Dufour S, Taksali SE, Tambor- land, NY 10310. E-mail: [email protected]. © 2005 by the American Diabetes Association. J-A, Kelley DE: Subcutaneous abdominal lane WV, Petersen KF, Bonadonna RC, fat and thigh muscle composition predict Boselli L, Barbetta G, Allen K, Rife F, Sa- insulin sensitivity independently of vis- voye M, Dziura J, Sherwin R, Shulman GI, ●●●●●●●●●●●●●●●●●●●●●●● ceral fat. Diabetes 46:1579–1585, 1997 Caprio S: in obese youth: a References 13. Fujimoto WY, Leonetti DL, Bergstrom syndrome of impaired glucose tolerance, 1. Ferrannini E, Natali A, Bell P, Cavallo- RW, Newell-Morris L, Boyko EJ, Shofer severe insulin resistance, and altered Perin P, Lalic N, Mingrone G: Insulin re- JB, Chen K-W, Wahl PW: Visceral adipos- myocellular and abdominal fat partition- sistance and hypersecretion in obesity: ity and incident coronary heart disease in ing. Lancet 362:951–957, 2003 European Group for the Study of Insulin Japanese-American men. Diabetes Care 26. Banerji MA, Chaiken RL, Gordon D, Kral Resistance (EGIR). J Clin Invest 100: 22:1808–1812, 1999 JG, Lebovitz HE: Does intra-abdominal ad- 1166–1173, 1997 14. Banerji MA, Faridi N, Atluri R, Chaiken ipose tissue in black men determine 2. Banerji MA, Lebovitz HE: Insulin-sensi- RL, Lebovitz HE: Body composition, vis- whether NIDDM is insulin-resistant or in- tive and insulin-resistant variants in ceral fat, leptin and insulin resistance in sulin-sensitive? Diabetes 44:141–146, 1995 NIDDM. Diabetes 38:784–792, 1989 Asian men. J Clin Endocrinol Metab 84: 27. Goldstone AP, Thomas EL, Brynes AE, 3. Lapidus L, Bengtsson C, Larsson B, Pen- 137–144, 1999 Bell JD, Frost G, Saeed N, Hajnal JV, nert K, Rybo E, Sjostrom L: Distribution 15. Goodpaster BH, Kelley DE, Wing RR, Howard JK, Holland A, Bloom SR: Vis- of adipose tissue and risk of cardiovascu- Meier A, Thaete FL: Effects of weight loss ceral adipose tissue and metabolic com- lar disease and death: a 12 year follow-up on regional fat distribution and insulin plications of obesity are reduced in of participants in the population study of sensitivity in obese. Diabetes 48:839– Prader-Willi syndrome female adults: ev- women in Gothenburg, Sweden. Br Med J 847, 1999 idence for novel influences on body fat (Clin Res Ed) 289:1257–1261, 1984 16. Boyko EJ, Leonetti DL, Fujimoto WY, distribution. J Clin Endocrinol Metab 86: 4. Ohlson LO, Larsson B, Svardsudd K, We- Newell-Morris L: Visceral adiposity and 4330–4338, 2001 lin L, Eriksson H, Wilhelmsen L, Bjorn- risk of type 2 diabetes. Diabetes Care 23: 28. Klein S, Fontana L, Young VL, Coggan torp P, Tibblin G: The influence of body 465–471, 2000 AR, Kilo C, Patterson BW, Mohammed fat distribution on the incidence of dia- 17. Kelley DE, Thaete FL, Troost F, Huwe T, BS: Absence of an effect of liposuction on betes mellitus. Diabetes 34:1055–1058, Goodpaster BH: Subdivisions of subcuta- insulin action and risk factors for coro- 1985 neous abdominal adipose tissue and insu- nary heart disease. N Engl J Med 350: 5. Kissebah AH, Vydelingum N, Murray R, lin resistance Am J Physiol Endocrinol 2549–2557, 2004

2324 DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 Lebovitz and Banerji

29. Miyazaki Y, Mahankali A, Matsuda M, 2791, 2002 lin resistance and insulin resistance in Mahankali S, Hardies J, Cusi K, Man- 30. Montague CT, O’Rahilly S: The perils of nondiabetic Asian Indian men. J Clin En- darino LJ, DeFronzo RA: Effect of piogli- portliness: causes and consequences of vis- docrinol Metab 89:2750–2755, 2004 tazone on abdominal fat distribution and ceral adiposity. Diabetes 49:883–888, 2000 32. Klein S: The case of visceral fat: argument insulin sensitivity in type 2 diabetic pa- 31. Abate N, Chandalia M, Snell PG, Grundy for the defense (Editorial). J Clin Invest tients. J Clin Endocrinol Metab 87:2784– SM: Adipose tissue metabolites and insu- 113:1530–1531, 2004

DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 2325