The Expanding Problem of Regional Adiposity: Revisiting a 1985 Diabetes Classic by Ohlson et al.

Over 75 years ago, the French physician Vague first described the “android” and “gynoid” phenotypes of obesity, reflecting the differential patterns of body fat distribution commonly observed in males and females, respectively (1). Importantly, Vague observed that those with the android phenotype (subsequently referred to as “upper-body,” “central,” “truncal,” or “abdominal” obesity) were more likely to develop diabetes, atherosclerosis, and other disorders than those individuals with the gynoid phenotype (“lower-body,” “gluteo-femoral,” or “peripheral” obesity). In the 1980s, epidemiological studies confirmed and extended the strong association between excess abdominal adipose tissue (assessed as a high measured waist-to-hip ratio) and the incidence of hypertension, hypertriglyceridemia, hyperinsulinemia, and glucose intolerance (2–5). Among these was the work of Ohlson et al. (6) titled “The Influence of Body Fat Distribution on the Incidence of Diabetes Mellitus: 13.5 Years of Follow-up of the Participants in the Study of Men Born in 1913,” which was published in 1985 in Diabetes. We chose to highlight the work of Ohlson et al. in this Classics in Diabetes article as it is one of the first long-term longitudinal population-based studies revealing abdominal adiposity as a predictor of the future development of diabetes.

Ohlson et al. set out to investigate whether body fat distribution can predict the future incidence of diabetes. This team of investigators used data from a large Swedish study population established in 1963 by coauthors Tibblin and Wilhelmsen (7). From the Swedish population registry, all men born in 1913 (on a calendar date divisible by three) and living in Gothenburg were selected. A total of 972 men met these criteria; remarkably, 855 men (88%) agreed to undergo a vast array of anthropometric measurements. This study cohort, often referred to as the “Gothenburg Study of Men Born in 1913,” formed the basis of many notable epidemiological studies in the late 20th century (7).

In 1967, at age 54 years, subjects initially examined in 1963 were invited to a reexamination. A total of 792 men (representing 95% of those still alive) consented to additional studies. In 1967, each of the participants underwent measurements of weight, height, skinfold thickness, and waist and hip circumference. Follow-up data collected 13.5 years later included blood glucose drawn after a 6- to 8-h fast. The results are impressive and quite convincing of the importance of abdominal adipose tissue distribution, measured as the waist–to–hip circumference ratio, as a predictor for developing diabetes. Importantly, this correlation held up even when the confounding effect of BMI was considered (6). Previous cross-sectional clinical studies had demonstrated these findings, but this was arguably the first study to show a predictive nature of abdominal adipose tissue, particularly in a cohort with a BMI that was not in the range of obesity at the time of baseline examination.

The epidemiological studies from the 1980s, including the work of Ohlson et al., motivated a search for mechanisms linking abdominal adiposity to metabolic dysfunction. Measurements of waist circumference are a proxy for “abdominal” fat; however, the abdominal region possesses several anatomically distinct white adipose tissue (WAT) depots, including abdominal subcutaneous WAT and intra-abdominal WAT. The latter is composed of distinct visceral (mesenteric and omental) and retroperitoneal depots with unique properties. The development and implementation of imaging approaches in the 1990s, including DEXA scans, MRI, and computed tomography scans, allowed for more precise assessments of body composition and adipose tissue mass, with a focus on the precise adipose depots of interest (8). Investigators of many studies concluded that visceral adipose mass was the link to insulin resistance and other cardiometabolic abnormalities, both in lean individuals and individuals with obesity (9–12). The attractive hypothesis at the time was that released free fatty acids from visceral adipose tissue into portal circulation led to deleterious lipid accumulation in the liver and subsequent insulin resistance (13). Nevertheless, by the late 1990s and early 2000s, some notable studies challenged this widely held hypothesis that visceral WAT was the driver of insulin resistance. Jensen and colleagues provided compelling evidence that subcutaneous WAT is a larger contributor to systemic free fatty acids that eventually reach the skeletal muscle and pancreas, whereas visceral WAT could contribute to hepatic accumulation of lipids (14). Klein and colleagues demonstrated that increased visceral WAT mass per se is not associated with insulin resistance unless there is a concomitant increase in hepatic steatosis (15). Moreover, in notable studies investigators found abdominal subcutaneous WAT mass, rather than intra-abdominal WAT mass, to be the better predictor of insulin sensitivity in both lean individuals and those with obesity (16,–,18). The question of whether increased visceral WAT mass drives insulin resistance became a matter of debate by the early to mid-2000s and was the focus of fascinating point-counterpoint articles published in Diabetes Care in 2005 (19,– 21).

The earlier studies of regional adiposity focused heavily on the importance of WAT mass and location. The relationship between adipose mass and insulin sensitivity was of course never clear-cut. At one end of the extreme, the condition of lipodystrophy, where adipocytes either cannot be produced or cannot store lipid, leads to hepatic and skeletal muscle lipotoxicity and the subsequent development of insulin resistance (22). The other end of the extreme is perhaps exemplified by genetic mouse models of insulin-sensitive obesity generated by Scherer and colleagues (23,24). In one model, adipocyte-specific overexpression of the mitochondrial protein mitoNEET, in a leptin-deficient genetic background (ob/ob), results in massive adipose expansion (24). These transgenic mice reach upward of 130 g—which was approximately three times the mass of the diabetic ob/ob controls—but maintain remarkable insulin sensitivity with no signs of hepatic steatosis. The increase in adiposity in these animals is driven by the expansion of subcutaneous WAT, highlighting the potential protective effect of preferentially expanding subcutaneous depots in the setting of caloric excess. Today, we recognize that a missing variable in many of the earlier discussions of adiposity was the health, or quality, of adipose tissue. The expansion of adipose tissue coincides with qualitative and quantitative changes in the cellular composition and structure of the tissue, a process broadly referred to as “tissue remodeling.” Pathologic WAT remodeling, often observed in individuals with obesity and insulin resistance, is characterized by the presence of large hypertrophic adipocytes, proinflammatory immune cell infiltration, and fibrosis, leading ultimately to a loss of adipocyte function (25). Healthy WAT remodeling and expansion, often observed in “metabolically healthy” obesity (and in the mitoNEET transgenic mouse), is characterized by the presence of smaller, more numerous adipocytes, a relatively lower degree of inflammation and fibrosis, and preserved adipocyte function (e.g., adiponectin production) (26). In this regard, the ability to promote adipocyte differentiation in the setting of overnutrition represents a protective mechanism to ensure healthy adipose expansion, preserve adipocyte function, and prevent ectopic lipid accumulation. These concepts are supported by several engineered animal models as well as by clinical observations (26,–,28). Notably, individuals or rodents treated with the thiazolidinedione class of glucose-lowering drugs often experience weight gain, associated with a healthy expansion and remodeling of subcutaneous WAT (28,29). An increase in adiposity is probably undesirable to most patients; however, the healthy remodeling of expanded subcutaneous WAT adipose tissue likely contributes to systemic insulin sensitization. As such, adipose tissue quality may be as important as adipose tissue mass and location as a predictor and determinant of metabolic health.

Defining the factors that control region-specific adipose expansion remains a high priority in the field today. WAT distribution and plasticity are strikingly sexually dimorphic (as initially pointed out by Vague). Estrogens and androgens are undoubtedly key regulators of this patterning and likely relate to the age-related changes in body fat distribution that are observed (30,31). The mechanisms by which sex hormones shape adipose plasticity are still not entirely understood and require much more attention. It is very likely that genetic variance also plays an important role in determining body fat distribution and plasticity. This is highlighted most directly by conditions of familial partial lipodystrophy, where select adipose depots are affected (22). Furthermore, rare disorders such as multiple symmetric lipomatosis, or “Madelung disease,” further highlight how distinct regulatory mechanisms control regional adiposity. Individuals with multiple symmetric lipomatosis present with metabolic syndrome, coinciding with symmetric overgrowth of select subcutaneous adipose tissues, often in the neck, torso, and abdominal areas and in the upper and lower limbs (32). Genome-wide association studies revealed a strong association of genetic variance, waist-to-hip ratio, and insulin resistance (33,34). Lotta et al. (35) identified dozens of independent loci for which genetic variance is associated with insulin resistance and reduced peripheral adiposity, further suggesting that the inability to expand subcutaneous WAT may be a more important determinant of insulin resistance in obesity than accumulation of visceral WAT per se.

Since the studies of the 1980s, much attention has been placed on the importance of visceral versus subcutaneous WAT in the context of obesity and cardiometabolic disease. Today, the problem of regional adiposity extends beyond these specific depots. In adults, the abundance of brown adipose depots, independent of BMI, inversely correlates with risk for developing various cardiovascular disorders (36). The epicardial adipose depot is likewise linked to several cardiovascular diseases and is rapidly gaining attention as a unique fat depot located between the myocardium and the epicardium (37). Regional adiposity also impacts other chronic disorders. For instance, the accumulation of adipocytes in the bone marrow can negatively impact osteogenic and hematopoietic processes (38). Moreover, the rapid expansion of mesenteric fat (“creeping fat”) occurring in association with Crohn disease is initially adaptive but may ultimately be pathologic and contribute to the inflammation and scarring observed in the intestine (39,40). Importantly, a connection between visceral adiposity and the pathogenesis and prognosis of various cancers is evident, yet still poorly understood (41).

The arrival of incretin-based therapies now provides effective tools for weight management and weight loss in those already affected by obesity and metabolic syndrome (42); however, pharmacological approaches to alter adipose tissue quantity and/or quality in a region-specific manner have not been devised. To this end, many important and fundamental questions regarding the biology of anatomically distinct adipose depots still need to be resolved. What are the developmental origins (i.e., lineage) of these distinct adipose tissues? What are the unique signals that drive their development, expansion, and remodeling, under physiological and pathophysiological conditions? The emergence of single-cell and spatial transcriptomics is beginning to define the depot-specific properties and heterogeneity of adipocytes, their precursor cells, and supportive cells of their microenvironment (43). Addressing these fundamental biological questions may ultimately aid in identifying those at risk for developing chronic disorders and/or help conceptualize new therapeutic strategies to target regional adiposity to improve health and prevent disease.

Funding. R.K.G. is supported in part by National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, grants R01 DK104789, R01 DK119163, and RC2 DK118620.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.