 |
|
|
| |
Patterns of fat distribution and its relationship to liver histology in nonalcoholic fatty liver disease (NAFLD)
|
| |
| |
Reported by Jules Levin
DDW, May 24, 2007, Wash DC
A. Kapoor1; P. Puri1; O. Cheung1; M. J. Contos1; V. A. Luketic1; M. L. Shiffman1; R. Sterling1; R. T. Stravitz1; A. J. Sanyal1
1. Internal Medicine, Virginia Commonwealth University, Richmond, VA, USA.
Background: Lipohypertrophy (LH) and lipoatrophy are both associated with NAFLD. Other than the association with abdominal obesity, little is known about the patterns of LH in "garden variety" NAFLD. It is also unknown if increased waist circumference (WC) in this population reflects generalized obesity or LH.
Specific Aims: To define the (1) patterns of LH in NAFLD and their prevalence, and (2) relationship of specific patterns of LH with histologic severity of NAFLD.
Methods: Consecutive subjects with biopsy-proven NAFLD were enrolled within 6 months of their biopsy. Body fat was estimated by bioimpedance. Liver histology was scored by the NASH CRN criteria. Fat distribution was evaluated as follows: Generalized obesity: body mass index (BMI), total body fat (TBF), Abdominal LH: WC, waist-hip ratio (WHR), suprailiac skinfold (SIS), limb LH:(triceps and biceps skinfold (TSF and BSF)), dorso-cervical hump (DCH): present vs absent. Abdominal LH was considered to be present if the WC was greater than the 95% confidence limits for a given BMI, age, gender and race. Spearmans coefficient and stepwise logistic regression were used to determine the relationship of specific LH patterns with other anthropometric indices and the impact on liver histology.
Results: 123 (F= 83, M= 40, 76% obese) subjects were studied. The mean (± S.D.) TBF was 40 ± 10 %. The mean WC and WHR were 104 ± 12 cm and 0.91 ± 0.08 respectively. 39/83 females and 20/40 males had abdominal LH. The mean TSF and BSF were 24 ± 12 mm and 20 ± 10 mm. DCH was present in 7/40 males and 28/83 females (p=n.s). WC correlated with BMI, TSF, BSF, and SIS (p< 0.01 for all). TSF correlated with age, BSF, SIS, BMI and WC (p< 0.001 for all). DCH correlated with hip, TBF and mid arm circumference (p< 0.05 for all). Neither BMI nor any of the LH patterns correlated significantly with severity of steatosis. Only WC (r=0.2, p<0.02), TSF (r=0.2, p<0.03) and DCH (r=0.2, p<0.02) correlated significantly with lobular inflammation. Cytologic ballooning was significantly associated with age, TSF (r=0.2, p<0.01) and DCH (r=0.25, p<0.004). Pericellular fibrosis stage was associated with DCH (r=0.3, p< 0.0002) and TBF (r=0.2, p<0.02). Abdominal LH did not correlate with the severity of any of the histologic features of NAFLD. By logistic regression, DCH and TSF were the only significant predictors of the grade and stage of NAFLD.
Conclusions: DCH is common in subjects with NAFLD. DCH and limb LH, defined by TSF, are significant predictors of liver histology in NAFLD. The contribution of WC to liver histology can be explained by the increase in WC in proportion to BMI rather than LH.
|
| |
|
|
|
|
|
