Commentary on: Increase in visceral fat per se does not induce insulin resistance in the canine model

Increased intra-abdominal fat accumulation is an important determinant of glucose tolerance and blood triglycerides and cholesterol concentrations (1). The observation that visceral and subcutaneous fat have different histological, biochemical, and functional characteristics, including fat cell size, free-fatty acid (FFA) release, and secretion of inflammatory factors, has led to the “portal hypothesis” (2). This hypothesis states that an overflow of FFA and inflammatory factors toward liver impairs insulin action and ultimately glucose and lipid homeostasis.

In this issue, Castro et al. (3) determined the effect of increasing canine omental fat through pharmacological sympathectomy on hepatic and peripheral insulin sensitivity. Dogs were investigated before and after 2 weeks of denervation and compared with a sham group. As anticipated, Castro et al. (3) detected a 25% increase in visceral fat mass in treated versus sham dogs. However, hepatic and peripheral insulin sensitivity remained similar, which led to the conclusion that visceral fat per se does not lead to insulin resistance.

On the one hand, this study supports earlier cross-sectional and intervention studies in humans. For instance, in one study, insulin sensitivity was similar in individuals with contrasting visceral fat mass but similar hepatic fat content. However, impaired insulin sensitivity was detected in individuals with high versus low hepatic fat content but similar visceral fat mass (4). Further evidence came from humans undergoing surgical removal of omental fat who later showed no meaningful improvement in glucose homeostasis (5). Taken together, these findings are consistent with an early study in humans showing that less than one-fifth of portal vein FFA comes from visceral fat, and only ∼10% of the total FFA in peripheral blood circulation is derived from liver (6). On the other hand, the study by Castro et al. (3) may not be suitable to test the portal hypothesis. Basically, as a consequence of the pharmacological sympathectomy, in vivo visceral fat cell lipolysis, FFA release, and FFA flux to the liver would be expected to decrease when compared with sham animals. Unfortunately, portal blood concentrations of FFA and inflammatory factors were not determined in this study. Furthermore, 2 weeks may not be a sufficient time period to influence hepatic fat content and inflammatory status.

An alternative experimental scenario to test the portal hypothesis can be performed in animals with hepatic steatosis undergoing omental fat sympathetic denervation. If the portal hypothesis plays a role, one would anticipate an enhancement of hepatic and systemic metabolism. Last, a causal relationship between the size of the visceral fat depot and metabolic control may not exist. Eventually, enlarged visceral fat and increased cardiovascular and metabolic risk could be both a consequence of an underlying, not yet determined common driving factor.