Cheung and co-authors have addressed the impact of fat distribution on the severity of nonalcoholic fatty liver disease (NAFLD).1 They observed a prevalence of dorsocervical lipohypertrophy as high as 28.5%, with dorsocervical lipohypertrophy being the single greatest contributor to the variability in severity of histological parameters. Abdominal obesity as measured by the waist circumference correlates weakly with the severity of hepatic inflammation (Spearman correlation coefficient 0.20), and patients with waist circumference in the highest quartile have the highest rate of portal fibrosis (see table 2 in Cheung et al.1).
Dorsocervical lipohypertrophy is one of the most representative signs of Cushing's syndrome, suggestive of an impairment of the hypothalamic-pituitary-adrenal axis which might occur in NAFLD as well as in several other conditions of disturbed carbohydrate metabolism (that is, insulin resistance and obesity2). Measurements of cortisol in the morning and after the dexamethasone suppression test in 4 patients with dorsocervical lipohypertrophy are not reliable to verify an impairment of the hypothalamic-pituitary-adrenal axis.3–6 What we would expect to be increased in these patients is the free cortisol which is biologically active.3 Free cortisol is the amount of hormone unbound to the cortisol-binding globulin; thus, it increases as cortisol-binding globulin levels decrease. Cortisol-binding globulin might be reduced in NAFLD, because its levels are low in obese patients with insulin resistance or type 2 diabetes.7, 8 In the series of Cheung et al.1, subjects with dorsocervical lipohypertrophy are those with the highest body mass index, highest fat mass, and the lowest insulin sensitivity. In these subjects, increased cortisol disposal may favor locally the accumulation of fat at atypical anatomical sites (“the buffalo hump” and the liver). To strengthen this speculation, in a condition of free fatty acid (FFA) overflow, and NAFLD is such a case, cortisol-binding globulin levels are further decreased, because FFAs displace cortisol from its binding globulin.8
As the relation among body circumferences, liver histology, and cardiovascular risk factors is concerned, apart from that of waist circumference, the relation of hip circumference to the severity of fatty liver disease may be of interest, if there is any relation. Tables 4 and 5 in Cheung et al.1 show these correlations, but in both sexes together. Waist and hip circumferences measure different aspects of body composition and have independent and often opposite effects on cardiovascular disease.9 A narrow waist coupled with a large hip may protect against cardiovascular disease.10 Women have a greater relative risk of cardiovascular diseases at lower waist circumference than men do.9 Thus, there may be value in separately analyzing circumferences in men and women, and, in the latter group, to distinguish fertile and menopausal women. Estrogens may protect women against NAFLD by promoting gluteal depots and contrasting visceral fat accumulation, similar to their protection against insulin resistance and cardiovascular risk,10 and their role in determining the different rates of NAFLD in childhood and in adulthood.
Re-examination of the results reported by Cheung et al.1 might suggest a role for cortisol metabolism in the development of NAFLD and progression to nonalcoholic steatohepatitis in patients with dorsocervical lipohypertrophy. Insight into the impact of the sex hormone milieu and body composition on NAFLD might also be gained.