Reply:

We read with interest the recent article on adiponectin in nonalcoholic steatohepatitis (NASH) by Hui et al.1 The authors reported that plasma adiponectin levels (1) are reduced in patients with NASH and (2) are inversely correlated with the histologically assessed degree of necroinflammation. These observations are highly relevant and bear therapeutic potential for the treatment of NASH; however, we feel that a potentially important point regarding the underlying basis of reduced adiponectin in NASH is underestimated in this report. In a number of studies, plasma adiponectin is negatively associated with body fat mass.2,3 Although body mass index and waist/hip ratio were given, Hui and colleagues did not provide a direct assessment of body fat mass. We measured plasma adiponectin levels using enzyme-linked immunosorbent assay (B-Bridge International, San Jose, CA4) in 34 patients with a histological diagnosis of NASH and in 23 controls (all values are given as the mean SD) matched for age (46.9 12.2 vs. 46.0 13.5 years), sex (female/male: 23:11 vs. 15:8), body mass index (29.8 5.2 vs. 28.1 3.6 kg/m2), and body fat mass (37.9% 10.0% vs. 37.4% 8.7%) as determined via bioelectrical impedance analysis.5 Although body fat mass was identical, the waist-to-hip ratio as measure of central obesity was significantly higher in the subjects with NASH compared with controls (0.95 0.09 vs. 0.88 0.13, respectively; P .05). In the study by Hui et al., the waist-to-hip ratio was also significantly higher in the patients presenting with NASH compared with the control group, despite having a similar body mass index. In our study population, plasma adiponectin was significantly decreased in patients with NASH compared with controls (6.0 2.7 vs. 10.7 5.1 g/mL; P .001), consistent with the report by Hui et al. Taken together, although the control group had exactly the same degree of obesity, the patients with NASH had a significantly altered body fat distribution toward central obesity and significantly lower plasma adiponectin levels. Data in the literature indicate that different fat stores might have different metabolic and inflammatory activity and that central obesity, as indicated by a high waist-to-hip ratio, is associated with unfavorable factors.6–8 Obesity is an established risk factor for the development of hepatic steatosis.9 Based on our data and the literature, we propose that the transition toward a hepatic inflammatory response and the development of NASH within a fatty liver are dependent on a shift in body fat distribution. Increasing visceral obesity results in (1) increased production of proinflammatory cytokines and adipokines such as leptin, tumor necrosis factor , and interleukin 68,10,11 and (2) decreased production of protective adipokines such as adiponectin.8,12 This abnormal balance might ultimately lead to the clinical and histopathological occurence of NASH. In conclusion, visceral obesity might be an important causative risk factor for NASH. Prospective multicenter studies with long-term follow-ups are necessary to further investigate the role of visceral obesity in the pathogenesis of NASH to identify patients at risk and thus provide early treatment for them.