“Buffalo” hump in nonalcoholic fatty liver disease

Abstract

I read with interest the recent article by Cheung et al.1 regarding the impact of fat distribution, as estimated by several anthropometric parameters, on the severity of nonalcoholic fatty liver disease (NAFLD) and metabolic syndrome in obese NAFLD patients. Notably, they found that dorsocervical lipohypertrophy (“buffalo” hump) was the single greatest predictor of the histological severity of NAFLD, whereas waist circumference was the greatest predictor of the severity of metabolic syndrome.1 Dorsocervical lipohypertrophy (DCL) was diagnosed only clinically but no quantitative measure (for example, ultrasound imaging) was used for diagnosing the “hump” at the base of the neck. Obviously, this is a major limitation of the study. Additionally, it is conceivable that the differential impact of DCL and waist circumference on the severity of NAFLD histology and metabolic syndrome is simply due to the fact that they are only indirect and crude indices of abdominal visceral fat distribution.2 A direct and more accurate measure of abdominal visceral fat, obtained by computed tomography or magnetic resonance imaging,2 would be useful to explore this issue adequately. However, the authors have interestingly shown that DCL, which is commonly seen in Cushing's syndrome, is a highly prevalent condition (28.5%) in obese NAFLD patients,1 thus suggesting an underlying subclinical hypercortisolism in NAFLD. Regarding this, recent studies have consistently shown a subclinical hypothalamo-pituitary-adrenal (HPA) axis dysfunction in NAFLD. Westerbacka et al.3 reported that obese men with higher liver fat content (as evaluated by magnetic resonance imaging spectroscopy) have a selective increase in 24-hour urinary excretion of 5-beta-reduced cortisol metabolites and a lower ratio of urinary cortisol/cortisone metabolites than men with lower liver fat content. Saruc et al.4 found that postmenopausal women with NAFLD, as assessed by ultrasound, have higher plasma dehydroepiandrosterone levels than controls. Others have shown that obese individuals5 and overweight type 2 diabetic patients6 with ultrasonographically diagnosed NAFLD have higher 24-hour urinary free cortisol excretion and postdexamethasone suppression cortisol concentrations than nonsteatotic controls. More importantly, we have recently demonstrated that overweight patients with biopsy-proven NAFLD have a subclinical, chronic HPA axis activation (as reflected by the increased, albeit still normal, 24-hour urinary free cortisol values and the blunted suppression of cortisol levels by 1.0 mg-dexamethasone with normal serum corticosteroid-binding globulin concentrations) that is correlated to the histopathological severity of NAFLD (that is, necro-inflammatory grade and fibrosis stage) independently of multiple potential confounders, including insulin resistance and metabolic syndrome components.7 Overall, therefore, the evidence from these studies,3-7 although only correlative in nature, suggests the hypothesis that overweight/obese NAFLD patients have a subtle HPA axis dysfunction that is strongly correlated with the severity of NAFLD histology, leading to subclinical cortisol overproduction that might be implicated in the development/progression of NAFLD. Future prospective and interventional (by using a cortisol inhibitor or specific 11beta-hydroxysteroid dehydrogenase-type 1 inhibitors) studies are obviously needed to validate this hypothesis and to better understand the underlying biological mechanisms linking HPA axis dysfunction to the development of NAFLD. Giovanni Targher M.D.*, * Section of Endocrinology, Department of Biomedical and Surgical Sciences, University Hospital, Verona, Italy.