SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

The patterns of fat distribution and their relationship to severity of nonalcoholic fatty liver disease (NAFLD) are unknown. The objectives of this study were to define the fat distribution patterns and their relationship to histological severity and metabolic parameters in subjects with NAFLD. Anthropometric indices and total body fat were measured in 123 subjects. Fat distribution patterns were defined as: general, abdominal, limb, truncal, and dorsocervical lipohypertrophy (DCL) a novel finding in NAFLD. Eighty-one (66%) of the subjects were obese, and 94 (76%) had abdominal obesity. Thirty-five (28.5%) had DCL. Whereas body mass index (BMI) correlated best with the presence of diabetes (r = 0.22, P < 0.05), waist circumference (WC) correlated best with hypertension (r = 0.2, P < 0.05), hypertriglyceridemia (r = 0.37, P < 0.001), and insulin resistance (homeostasis model of assessment for insulin resistance [r = 0.68, P < 0.0001]). None of the patterns of fat distribution were significantly associated with severity of hepatic steatosis. Abdominal obesity (WC) correlated with inflammation (r = 0.2, P < 0.05) only. DCL correlated significantly with the severity of all histological parameters except steatosis. Whereas DCL was the single greatest contributor to the variability in severity of histological parameters, a model combining BMI, WC, and DCL showed the greatest contribution to the variability in severity of individual histological parameters. The addition of steatosis grade to the model significantly increased its contribution to the range of lobular inflammation. Conclusion: WC predicts metabolic risk profile with the most significance. However, DCL is most strongly associated with severity of steatohepatitis. WC and BMI added modestly to the contribution of DCL to severity of nonalcoholic steatohepatitis. (HEPATOLOGY 2007.)

Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver disease in North America and affects up to 30% of the general population.1 The clinical-histological spectrum of NAFLD includes nonalcoholic fatty liver (NAFL) to nonalcoholic steatohepatitis (NASH).2 NASH can progress to cirrhosis in 15%-20% of subjects.3

NAFLD is strongly associated with insulin resistance and the metabolic syndrome.4, 5 The current epidemic of the metabolic syndrome is closely linked to the prevalence of obesity.6 Although obesity is defined by body mass index (BMI), is a general measure of increased adipose tissue mass, and is linked to insulin resistance,7 several studies suggest that the distribution of adipose tissue may also be important in determining development of the metabolic syndrome.8, 9 Specifically, the presence of abdominal or visceral adiposity has been strongly implicated in development of the metabolic syndrome.10–13 Recently, there is growing awareness that the distribution of fat in other areas may also contribute to the metabolic risks associated with obesity.14 The potential metabolic effects of fat depots in various parts of the body are further underscored by the association of insulin resistance with fat redistribution syndromes in those with HIV infection.15, 16 We have recently observed that some subjects with NAFLD have a hump at the base of their neck similar to that seen in Cushing's syndrome (Fig. 1). There are no published data related to this finding or the distribution patterns of fat and their relationship to liver histology in NAFLD.

thumbnail image

Figure 1. (A) Enlargement of dorsocervical fat pad (dorsocervical lipohypertrophy) in a patient. (B) Close-up of dorsocervical lipohypertrophy.

Download figure to PowerPoint

The objectives of this study were (1) to define the distribution patterns of adipose tissue and frequency of dorsocervical lipohypertrophy (DCL) in subjects with NAFLD and (2) to evaluate whether abdominal obesity had a greater impact on the severity of NAFLD and associated metabolic and clinical parameters than other patterns of fat distribution.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

Study Population

Consecutive subjects seen in the NAFLD clinic at the Virginia Commonwealth University Medical Center between 1999 and 2001 were screened for this study. The principal inclusion criterion was the presence of biopsy-proven NAFLD within 6 months before entry into the study. Liver histology was analyzed in 2005 using the NASH clinical research network criteria.17 All subjects had at least grade 1 steatosis, and steatohepatitis was diagnosed by the additional presence of cytologic ballooning and inflammation with or without Mallory hyaline or pericellular fibrosis.18 The presence of NAFLD was established via liver biopsy demonstrating either a fatty liver or steatohepatitis and a history of either no alcohol consumption or consumption of <20 g/day on the average assessed clinically. All subjects also underwent baseline tests for hepatitis B and C, hemochromatosis (iron panel and gene tests for those with an iron saturation >50%), Wilson disease, autoimmune hepatitis, and α1-antitrypsin deficiency. Those who had evidence of any of these conditions were excluded. Also, subjects who were either pregnant or had been pregnant within the previous 6 months were excluded. Subjects who had undergone bariatric surgery were also excluded.

Measurement of Body Fat and Fat Distribution Indices

These were all measured by a single nutritionist (S.S.) with many years of experience in nutrition-related research in the general clinical research center of the author's institution. Triceps skinfold, biceps skinfold, subscapular skinfold, and suprailiac skinfold were measured using Lange calipers (Cambridge Scientific Industries, Cambridge, MD) on the right side of the body. Waist circumference (WC) and hip circumference were measured to the nearest 0.1 cm with a Gullick II tape measure (Country Technology, Inc., Gays Mills, WI). WC was measured at a level midway between the lowest rib margin and the iliac crest, while hip circumference was measured at the widest level over the greater trochanters as described previously19–22 (http://www.nhlbi.nih.gov/guidelines/obesity, http://www.who.int/chp/steps/Part3_Section4.pdf). In some obese individuals, it may be difficult to identify a waist narrowing around the midpoint between the ribcage and iliac crest. In such cases, WC was approximated at the lateral level of the 12th rib or lower floating rib.23–26 These are well established and validated methods for assessment of hip circumference and WC and are used in several large scale studies.27 Bioelectrical impedance was done using the Quantum II Bioelectrical Body Composition Analyzer (RJL Systems, Clinton Township, MI). Results were entered into a computer program (Cyprus 2.7, RJL Systems), and percentage body fat and body water were derived based on equations described by Kotler et al.28

Classification of Fat Distribution

Fat distribution was classified and defined as follows.

Generalized Obesity.

Generalized obesity was defined by a BMI ≥30 kg/m2.29 Total body fat was also used as another measure of generalized adiposity.

Abdominal Adiposity.

WC and waist/hip ratio were used as markers of abdominal obesity.30 In addition, suprailiac skinfold thickness was used as a measure of abdominal subcutaneous fat.31 Based on WC, the presence of abdominal obesity was defined in 2 ways: (1) WC >102 cm for men and >88 cm for women (Adult Treatment Panel III criteria)32 and (2) WC greater than the 95% confidence limits for the mean value predicted from age, sex, race, and BMI.33 The latter method was used to evaluate if there was abdominal obesity out of proportion to the degree of generalized obesity.

Dorsocervical Lipohypertrophy.

It was observed that some patients with NAFLD had a clinically obvious “hump” at the base of their neck (Fig. 1). There is no absolute unequivocal quantitative method for the diagnosis of DCL, and virtually all studies of the subject have made a diagnosis based on visual observation of enlargement of the dorsal cervical fat pad.34, 35 Its absence or presence was denoted as 0 or 1, respectively.

Limb Adiposity.

Limb adiposity was measured as a continuous numerical variable based on the biceps and triceps skinfold thickness. Subjects in the fourth quartile were considered to have the greatest limb adiposity.

Truncal Adiposity.

The torso is defined as the general area between the neck and legs. The subscapular skinfold thickness and hip circumference were measured along with WC as measures of adiposity around the torso.

Assessment of Chronic Medical Conditions

Past and current medical problems were thoroughly assessed during patient interviews and clinical assessment. Primary chronic morbidities related to the metabolic syndrome included type II diabetes mellitus and hypertension. Disease status was established and confirmed using appropriate diagnostic tests when applicable, and its absence or presence was denoted as 0 or 1, respectively.

Assessment of Liver Histology

Liver biopsies from subjects with a complete data set (n = 123) were rescored at the time of analysis using the NASH clinical research network criteria specifically for the purpose of this study.17 This was done so that the data were comparable to other studies that are also likely to use this system of histological grading, which was developed around 2005. All biopsies were scored by a single investigator who had 20 years of experience in reading liver histology and was blinded to the clinical data. All biopsies were stained with hematoxylin-eosin as well as Masson's trichrome stain. Histological lesions were assessed for the following parameters as defined previously17: degree of steatosis, lobular inflammation, cytologic ballooning, pericellular fibrosis, Mallory hyaline, and NASH.

Statistical Analysis

The distribution of demographic, clinical, laboratory, anthropometric, and histological data was recorded, and intergroup comparisons were made with a t test (continuous variables), Kruskal-Wallis test (ordinal variables), and Fisher exact test (categorical variables) as appropriate. Spearman correlation coefficients were calculated to assess the relationships between anthropometric indices and (1) other anthropometric indices, (2) severity of liver histology, (3) presence of diseases associated with the metabolic syndrome and obesity, and (4) metabolic risks measured using various biochemical parameters. Differences in correlation coefficients were evaluated using Fisher's z transformation. Univariate ordinal regression analyses were performed to determine the percentage of variation in the severity of liver histology that was attributable to each pattern of fat distribution. The contribution of interactions between various patterns of fat distribution grade to the severity of histological parameters was assessed via multiple ordinal logistic regression. In the base model, one of the following parameters was used as the starting point for the model: BMI, total body fat, WC, DCL. The markers of specific patterns of fat distribution were added to the base model as independent variables. Finally, the steatosis grade was added to the model to evaluate the impact of interactions between fat distribution patterns and steatosis grade on other histological features of NAFLD. A P value of 0.05 was considered significant.

Results

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

Patient Characteristics (Table 1)

Data from a total of 123 consecutive subjects on whom a full data set was available were used for this study. The median age was 56 years and the majority of the subjects were Caucasian. As expected, males were both significantly taller and heavier than females, but their BMI was not significantly different from females. However, females had significantly higher total body fat (39.2% versus 27.9% [P < 0.003]) than males. Female subjects also had a significantly higher prevalence of diabetes (29/83 versus 5/40 [P < 0.009]). The median biopsy length was 2 cm (range: 1.2-3.5 cm). Although the prevalence of NASH was not significantly different between males and females (85% versus 80.7%), males had a significantly lower stage of disease (1.3 versus 1.9 [P < 0.03]). Twenty-two subjects had nonalcoholic fatty liver.

Table 1. Descriptive Analysis of Study Participants
 Overall (n = 123)Male (n = 40)Female (n = 83)
  • Values are expressed as the mean ± SD.

  • *

    Abdominal obesity is defined as waist circumference >95% confidence limits for a given BMI, sex, age, and race.

  • Dorso-cervical lipohypertrophy is defined as the presence of a “buffalo hump.”

  • ‡, §

    Limb adiposity is defined as the 4th quartile of triceps skinfold thickness.

  • d

    Hypercholesterolemia is present if total cholesterol is greater than 200 mg/dL based on the Adult Treatment Plan III and the American Heart Association guidelines.

  • P <0.005 when compared to the other sex group.

  • P <0.05 when compared to the other sex group.

Demographics   
 Age (yr)55.5 (median, 56.7)53.5 (median, 55.7)56.6 (median, 57.65)
 African American18/123 (14.6%)5/40 (12.5%)13/83 (15.7%)
 Caucasian95/123 (77.2%)31/40 (77.5%)64/83 (77.1%)
 Hispanic2/123 (1.6%)02/83 (2.4%)
 Others8/123 (6.5%)4/40 (10%)4/83 (4.8%)
Anthropometric indices   
 Height (cm)166.4 ± 12.1175.9 ± 10.3161.9 ± 10
 Weight (kg)96.1 ± 22.5107.9 ± 2690.6 ± 18
 BMI (kg/m2)34 ± 6.1534.9 ± 7.133.5 ± 5.7
 Waist circumference (cm)97.53 ± 27.76103.4 ± 26.796 ± 26.4
 Waist/hip ratio0.85 ± 0.230.91 ± 0.220.84 ± 0.2
 Triceps skinfold (mm)25.82 ± 10.625.5 ± 826.5 ± 11.4
 Biceps skinfold (mm)21.4 ± 10.219.6 ± 7.422.6 ± 11
 Suprailiac skinfold (mm)29.04 ± 11.627 ± 10.630.4 ± 11.7
 Subscapular skinfold (mm)27.24 ± 10.926.5 ± 10.227.9 ± 11
 Hip circumference (cm)106.7 ± 31.5106.9 ± 32.4107.9 ± 29.6
 Total body fat (%)35.4 ± 16.827.9 ± 1439.2 ± 16.4
Patterns of lipohypertrophy   
 General obesity (4th quartile BMI)38/123 (31%)12/40 (30%)26/83 (31.3%)
 Abdominal obesity*59/123 (48%)20/40 (50%)39/83 (47%)
 Dorso-cervical lipohypertrophy35/123 (28.5%)7/40 (17.5%)28/83 (33.7%)
 Limb adiposity33/123 (26.8%)7/40 (17.5%)26/83 (31.3%)
Disease status   
 Diabetes mellitus II34/123 (27.6%)5/40 (12.5%)29/83 (34.9%)
 Hypertension64/123 (52%)20/40 (50%)44/83 (53%)
 Hypercholesterolemia§30/123 (24.4%)10/40 (25%)20/83 (24.1%)
Liver histology   
 Steatosis2.2 ± 0.72.2 ± 0.72.2 ± 0.7
 Lobular inflammation1.2 ± 0.51.2 ± 0.51.2 ± 0.5
 Cytologic ballooning0.9 ± 0.60.8 ± 0.60.9 ± 0.6
 Pericellular fibrosis1.6 ± 1.41.3 ± 1.31.9 ± 1.4
 Mallory hyaline0.4 ± 0.60.4 ± 0.50.4 ± 0.6
 NASH present101/123 (82%)34/40 (85%)67/83 (80.7%)

Distribution of Findings in Subjects with Various Patterns of Fat Distribution (Table 2)

Table 2. Patterns of Lipohypertrophy with Distribution of Age, Sex, and Race: Comparison of Severity of Histologic Features, Metabolic Parameters, and Anthropometric Measurements
 General ObesityAbdominal ObesityDCLLimb Adiposity
1stQuartile BMI (n = 29)4thQuartile BMI (n = 38)1stQuartile TBF (n = 27)4thQuartile TBF (n = 53)Based on WC ValueBased on 95% CIAbsent (n = 88)Present (n = 35)Absent (n = 39)Present (n = 33)
Absent (n = 29)Present (n = 94)Absent (n = 64)Present (n = 59)
  • General obesity is defined as the 4th quartile of BMI or TBF and is compared with the 1st quartile of BMI or TBF. Abdominal obesity: (1) absolute WC >102 cm in males and >88 cm in females and (2) WC >95% CI for a given BMI, sex, age, and race. Limb adiposity is defined as the 4th quartile of triceps skinfold (present) and is compared with the 1st quartile (absent). HOMA-IR: mean value for the group as a whole (n = 123) is 4.3.

  • Abbreviations: AA, African American; BMI, body mass index; C, Caucasian; CB, cytologic ballooning; DCL, dorso-cervical lipohypertrophy; HOMA-IR, homeostasis model of assessment for insulin resistance; LI, lobular inflammation; MH, Mallory hyaline; NASH, nonalcoholic steatohepatitis; PF, pericellular fibrosis; TBF, total body fat; WC, waist circumference.

  • *

    P <0.05

  • **

    P <0.005

  • ***

    P <0.0005 when compared with the group without the same condition.

  • P <0.05

  • ††

    P <0.005

  • †††

    P <0.0005 when compared with the total group (n = 123).

Males, n (%)7 (24%)12 (31.6%)24 (88.9%)9 (17%)***15 (52%)25 (27%)21 (32.8%)19 (32.2%)33 (37.5%)7 (20%)16 (41%)7 (21.2%)
Females, n (%)22 (76%)26 (68.4%)3 (11.1%)44 (83%)***14 (48.3%)69 (73.4%)43 (67.2%)40 (67.8%)55 (62.5%)28 (80%)23 (59%)26 (78.8%)
AA, n (%)4 (14%)9 (24%)3 (11.1%)7 (13.2%)6 (21%)12 (12.8%)6 (9.3%)12 (20.4%)12 (13.6%)6 (17%)7 (18%)7 (21.2%)
C, n (%)23 (79.3%)23 (60.5%)22 (81.5%)30 (56.6%)*21 (72.4%)74 (78.8%)54 (84%)41 (69.5%)70 (80%)25 (72%)30 (77%)21 (63.6%)
Mean age54.6352.7754.1655.372.2556.657.1454.655.556.554.5554.07
Histology            
 Steatosis2.12.142.32.192.072.242.132.272.182.252.282.28
 LI1.031.241.071.161.071.21.11.251.11.3*1.051.3*
 CB0.91.080.70.880.710.94*0.840.980.81.15**0.81.1*
 PF1.591.811.01.84*1.111.83*1.581.751.32.5***††1.531.56
 MH0.380.460.330.470.390.440.50.360.30.6***,0.280.53
 NASH24 (83%)35 (92.1%)20 (74.1%)31 (58.5%)21 (72.4%)80 (85%)46 (72%)55 (93%)72 (82%)29 (83%)30 (77%)29 (87.9%)
Diseases            
 Diabetes4 (14%)13 (34.2%)2 (7.4%)15 (28.3%)*3 (10.3%)31 (33%)13 (20.3%)20 (34%)24 (27%)10 (28%)9 (23%)12 (36.4%)
 Hypertension13 (44.8%)19 (50%)13 (48.1%)18 (34%)10 (34.5%)54 (57.4%)29 (45.3%)34 (57.6%)48 (54%)16 (45%)20 (51.2%)18 (54.5%)
Indices/DCL            
 BMI27.2442.7***†††31.837.6***†††28.235.4***30.4737.5††2935*3140***†††
 TBF36.8847.1***†††26.152.4***†††34.942.3**38.2642.73242*3746***††
 DCL7 (24%)13 (34.2%)4 (14.8%)10 (18.7%)6 (21%)29 (31%)10 (15.6%)25 (42.3%)  10 (25.6%)12 (36.4%)
HOMA-IR3.85.08*4.74.73.84.44.334.254.24.54.345.02
Generalized Obesity.

The mean BMI of the patient cohort was 34 kg/m2; the total body fat was 35.4%. Subjects in the 4th quartile for BMI were not significantly different from those in the 1st quartile with respect to sex or race distribution and the severity of histological changes. However, they had a significantly greater insulin resistance index as measured by the homeostasis model of assessment for insulin resistance (HOMA-IR) when compared with subjects in the 1st quartile and the group as a whole (5.08 versus 3.8 versus 4.3 [P < 0.02]). On the other hand, when total body fat was used as a measure of general obesity, those in the 4th quartile were more likely to be female, Hispanic, and diabetic. They also had a significantly greater fibrosis stage compared with those in the 1st quartile (1.84 versus 1.0 [P < 0.01; Kruskal-Wallis test]).

Abdominal Obesity.

A total of 94 subjects met the criteria for abdominal obesity as defined by their absolute WC using the Adult Treatment Panel III criteria.32 Interestingly, 59 of these 94 subjects had a WC greater than the 95% confidence limits for that predicted on the basis of age, sex, race, and BMI.33 This indicated that although many subjects had abdominal obesity, this was commensurate with BMI, age, sex, and race in 35 (37%) subjects, while the other 59 (63%) subjects had an enlargement of abdominal girth that was beyond that predicted simply on the basis of their height and weight. Those with abdominal obesity defined by absolute WC had a significantly higher grade of cytologic ballooning (0.94 versus 0.71 [P < 0.04]) and fibrosis stage (1.83 versus 1.1 [P < 0.01]). On the other hand, those with abdominal obesity defined by WC values greater than the 95% confidence limits of that predicted by BMI, age, sex, and race did not have significantly greater severity of any histological parameter compared with those without abdominal obesity.

Dorsocervical Lipohypertrophy.

A total of 35 subjects had DCL as defined by a dorsocervical hump. These subjects were more likely to be female, but this finding was not statistically significant. Subjects with DCL were significantly heavier and had a greater total body fat content compared with those without it (BMI, 35 versus 29 [P < 0.05]; total body fat, 42% versus 32% [P < 0.02]). Of note, subjects with DCL had significantly greater inflammation, cytologic ballooning, Mallory hyaline, and fibrosis stage compared with those without DCL.

Limb Adiposity.

Subjects in the 4th quartile had significantly greater inflammation and cytologic ballooning compared with those in the 1st quartile and exhibited a trend toward greater fibrosis. There were no instances of limb or generalized lipoatrophy in this cohort.

Relationship Between Various Patterns of Fat Distribution (Table 3).

To further analyze whether these patterns usually coexisted or occurred independent of each other, the Spearman correlation coefficients between markers of specific patterns of adiposity were calculated. Interestingly, although BMI and total body fat correlated significantly, the correlation was modest (r = 0.34, P < 0.001). Although BMI correlated strongly with markers of abdominal obesity (WC: r = 0.76, P < 0.0001), the correlation of total body fat with these markers was relatively modest (WC: r = 0.13, P value not significant). Both BMI and total body fat correlated significantly with markers of limb adiposity and DCL. WC also correlated with markers of limb lipohypertrophy (triceps skinfold: r = 0.31, P < 0.001) and truncal lipohypertrophy (subscapular skinfold: r = 0.55, P < 0.0001; hip circumference: r = 0.64, P < 0.0001). Whereas WC correlated with DCL (r = 0.23, P < 0.05), the suprailiac skinfold, a measure of abdominal subcutaneous fat,31 did not. These data indicate that increasing generalized obesity is associated with increased abdominal obesity. Also, increasing BMI and abdominal girth are both associated with increased likelihood of having DCL; of the two, abdominal obesity correlated better with DCL.

Table 3. Relationship Between Various Patterns of Fat Distribution: A Correlation Matrix of Anthropometric Indices
  General ObesityAbdominal ObesityDCLLimb AdiposityTruncal Obesity
BMITBFWCSISFWHRTSFBSFSSSFHC
  • Values are expressed as Spearman correlation coefficient (r).

  • Abbreviations: BMI, body mass index; BSF, biceps skinfold; DCL, dorso-cervical lipohypertrophy; HC, hip circumference; SISF, suprailiac skinfold; SSSF, subscapular skinfold; TBF, total body fat; TSF, triceps skinfold; WC, waist circumference; WHR, waist/hip ratio.

  • *

    P <0.05.

  • **

    P <0.001.

  • ***

    P <0.0001.

General obesity           
 BMI10.34**0.76***0.55***0.140.19*0.51***0.54***0.60***0.76***
 TBF0.34**10.130.26**-0.150.19*0.28*0.38***0.34**0.28*
Abdominal obesity           
 WC0.75***0.1310.43***0.50***0.23*0.31**0.40***0.55***0.64***
 SISF0.55***0.26**0.43***10.040.160.49***0.53***0.63***0.44***
 WHR0.14-0.150.43***0.051−0.07−0.14−0.060.1−0.16
 DCL0.19*0.19*0.23*0.16-0.0710.120.24*0.25*0.25*
Limb adiposity           
 TSF0.52***0.28*0.34**0.50***−0.140.1210.72***0.51***0.54***
 BSF0.54***0.38***0.40***0.53***−0.050.24*0.72***10.56***0.57***
Truncal obesity           
 SSSF0.60***0.34**0.56***0.64***0.10.25*0.51***0.56***10.57***
 HC0.69***0.28*0.64***0.43***−0.170.25*0.54***0.57***0.57***1
Relationship of Fat Distribution Patterns to Other Features of the Metabolic Syndrome (Table 4).

The BMI, but not total body fat, correlated significantly with the risk of diabetes and HOMA-IR values. However, it did not correlate significantly with any other feature of the metabolic syndrome. The presence of DCL was significantly associated with diabetes. Abdominal obesity correlated significantly with HOMA-IR (WC: r = 0.68, P < 0.0001) as expected.36, 37 HOMA-IR was also associated significantly with other markers of truncal obesity (subscapular skinfold: r = 0.52, P < 0.001, and hip circumference: r = 0.39, P < 0.001), as well as DCL (r = 0.33, P < 0.001).38, 39 WC also correlated significantly with hypertension and best correlated with hypertriglyceridemia. These data confirm the contribution of abdominal obesity to the metabolic and other clinical abnormalities associated with the metabolic syndrome.

Table 4. Spearman Correlation Analysis of Anthropometric Indices with Metabolic Risks and Diseases
  High-Density LipoproteinLow-Density LipoproteinTriglycerideTotal CholesterolHemoblobin A1cHOMA-IRDiabetes MellitusHypertension
  • Values are expressed as Spearman correlation coefficient (r).

  • Abbreviations: BMI, body mass index; BSF, biceps skinfold; DCL, dorso-cervical lipohypertrophy; HC, hip circumference; HOMA-IR, homeostasis model of assessment for insulin resistance; SISF, suprailiac skinfold; SSSF, subscapular skinfold; TBF, total body fat; TSF, triceps skinfold; WC, waist circumference; WHR, waist/hip ratio.

  • *

    P <0.05.

  • **

    P <0.001.

  • ***

    P <0.0001.

General obesity         
 BMI−0.036−0.090.19−0.170.29**0.5***0.22*0.16
 TBF0.20.180.070.060.160.090.150.09
Abdominal obesity         
 WC−0.2−0.060.37**−0.130.24*0.68***0.10.20*
 SISF0.080.010.29*0.050.180.34***0.24*0.14
 WHR−0.3*0.150.35**0.06−0.020.3**−0.020.08
 DCL0.14−0.1−0.040.030.26*0.33**0.25*0.16
Limb adiposity         
 TSF−0.03−0.0070.110.0070.090.2*0.130.10*
 BSF−0.007−0.170.16−0.00070.160.170.200.23*
Truncal adiposity         
 SSSF−0.080.0010.30*−0.050.41**0.52**0.24*0.20*
 HC0.08−0.140.07−0.180.23*0.39**0.130.12
Contribution of Fat Distribution Patterns to the Severity of Histologic Changes in the Liver (Tables 5 and 6).

Initially, the relationship between the severity of individual histological parameters and the specific measures of the various patterns of fat distribution was assessed (Table 5). Interestingly, the severity of generalized obesity was not significantly associated with the severity of steatosis. It was, however, associated with severity of lobular inflammation and pericellular fibrosis. Abdominal obesity (WC) correlated with the severity of inflammation (r = 0.2, P < 0.05) but not fibrosis stage. DCL correlated significantly with the severity of lobular inflammation, cytologic ballooning, Mallory hyaline, and pericellular fibrosis. Triceps and biceps skinfold thickness also correlated with the severity of inflammation and cytologic ballooning.

Table 5. Spearman Correlation Analysis of Anthropometric Indices with Liver Histology
  SteatosisLobular InflammationCytologic BallooningPericellular FibrosisMallory Hyaline
  • Values are expressed as Spearman correlation coefficient (r).

  • Abbreviations: BMI, body mass index; BSF, biceps skinfold; DCL, dorso-cervical lipohypertrophy; HC, hip circumference; SISF, suprailiac skinfold; SSSF, subscapular skinfold; TBF, total body fat; TSF, triceps skinfold; WC, waist circumference; WHR, waist/hip ratio.

  • *

    P < 0.05.

  • **, *

    P < 0.005.

General obesity      
 BMI0.030.20*0.160.140.06
 TBF0.050.090.10.20*0.03
Abdominal obesity      
 WC0.120.20*0.140.080.03
 SISF−0.070.070.22*0.20*−0.0019
 WHR0.090.120.11−0.03−0.058
 DCL0.060.20*0.25**0.33**0.24*
Limb adiposity      
 TSF0.010.22*0.22*0.080.1
 BSF0.00020.20*0.23*0.160.20*
Truncal adiposity      
 SSSF0.0020.17*0.160.17−0.019
 HC0.020.120.050.080.04
Table 6. Univariate and Multivariate Ordinal Regression Analysis of Liver Histology with Patterns of Lipohypertrophy
Independent VariablesDependent Variables
Steatosis (R2)Lobular Inflammation (R2)Cytologic Ballooning (R2)Fibrosis (R2)Mallory Hyaline (R2)
  • Values are expressed as regression coefficient R2 (expressed in % for percent variability).

  • Abbreviations: BMI, body mass index; BSF, biceps skinfold; DCL, dorso-cervical lipohypertrophy; TBF, total body fat; TSF, triceps skinfold; WC, waist circumference.

  • *

    P < 0.05.

  • **

    P < 0.005.

  • ***

    P < 0.0005.

Generalized obesity     
 BMI0.01%3.2%*1.3%0.8%0.35%
 BMI, steatosis12.4%*2.5%0.8%2.9%
 TBF0.01%1%1.6%3.3%**0.40%
 TBF, steatosis10.6%**3.0%3.5%4.4%
Abdominal obesity     
 WC0.4%3.6%*0.9%0.3%0.03%
 WC, BMI0.6%4.3%1.2%0.7%0.9%
 WC, BMI, DCL0.7%5.9%4.8%5.7%4.7%
 WC, BMI, steatosis13.4%**2.7%0.9%3.8%
DCL     
 DCL1%4.1%3.6%*4.5%**3.5%*
 DCL, BMI0.07%4.7%4.2%*4.9%***3.6%*
 DCL, BMI, steatosis13.8%**5.2%5.2%**5.7%
 DCL, WC0.5%5.3%3.8%*4.6%***3.7%**
 DCL, WC, steatosis14%**5.6%*5.0%**6.0%*
 DCL, WC, BMI, steatosis14.7%**6.1%6.2%**6.9%*
Limb adiposity     
 TSF0.04%3.4%*2.9%*0.2%0.7%
 BSF0.06%1.8%2.9%*0.7%2.6%*

The proportion of the uncertainty attributed to the model fit (r2) for specific markers of adiposity to predict the severity of individual histological parameters by ordinal logistic regression are tabulated in Table 6. None of the indices contributed susbtantially to the grade of steatosis. Of the patterns of fat distribution, DCL contributed to the greatest degree to variability in severity of histological parameters. When BMI or WC were added to DCL in the model as independent variables, they only contributed marginally to the variability of severity of lobular inflammation, cytologic ballooning, and fibrosis. Conversely, the addition of DCL to the base model of either BMI or WC increased the contribution of the model to the variability in severity of these histological parameters.

In a final analysis, steatosis grade was added to the fat distribution pattern in the model. When steatosis grade was added to WC and BMI, it added significantly to the contribution of these parameters to the variability in lobular inflammation but not cytologic ballooning or fibrosis. A similar effect was seen when steatosis grade was added to DCL and BMI.

Discussion

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References

The aim of this study was to define the fat distribution patterns in subjects with NAFLD and to determine the relationship between such patterns and the histological parameters of NAFLD as well as features of the metabolic syndrome. This study demonstrates that, in addition to the well-described relationship between generalized obesity and abdominal obesity, many subjects have DCL, which correlates with the histological severity of NAFLD. It must be noted that these data may be population-specific and are potentially subject to ascertainment bias, because they were derived from a group of North American subjects with known fatty liver disease who were referred to a specialty clinic in a tertiary care institution. Another potential source of bias is the variability in severity of histological findings in liver biopsies due to its inherent sampling variability. In this study, the median biopsy length was approximately 2 cm, which should keep this source of variability within acceptable limits.

Generalized obesity is a well-known risk factor for NAFLD.40, 41 This study shows that while subjects with the highest BMI or total body fat had higher fibrosis scores compared with the group as a whole (Table 2), the correlation coefficients between these parameters and the severity of individual histological features were generally low (Table 5). Additionally, although the sex distribution in the highest quartile for BMI was similar to that for the other groups, subjects with the highest total body fat were more likely to be female, Hispanic, and diabetic. Females are known to have higher body fat, especially subcutaneous fat, compared with males.42 The relationship between body fat and presence of diabetes suggests that total body fat is a key contributor to the overall insulin-resistant state. The lack of such a relationship with BMI suggests that some of the subjects with the highest BMI may not have had the highest total body fat content. Also, the nonsignificant differences in HOMA-IR between those with 1st versus 4th quartiles of total body fat may reflect pancreatic beta cell failure, as corroborated by higher prevalence of diabetes in the 4th quartile.

This study confirms the importance of abdominal obesity for the development of insulin resistance, diabetes, hypertension, and hypertriglyceridemia.36 Interestingly, WC and suprailiac skinfold thickness, a marker of abdominal subcutaneous fat, also correlated with the severity of inflammation and cytologic ballooning, respectively (Table 5). This corroborates recent evidence that subcutaneous fat adds to total body fat and abdominal obesity as a risk factor for the features of the metabolic syndrome.14

The presence and severity of the metabolic syndrome has been shown to be directly related to the severity of liver histology in subjects with NAFLD.43–45 However, abdominal obesity had only a modest contribution to the severity of liver histology. It is likely that while the severity of the insulin-resistant state, which is related to abdominal obesity, drives the development of the features of the metabolic syndrome (for example, NAFLD), there are other factors (for example, genetics) that may determine the severity of the histological changes.

A novel finding is the description of DCL in patients with NAFLD. It is physically similar to the cervical hump seen in subjects with Cushing's syndrome; while the early morning cortisol levels and dexamethasone suppression test were normal in 4 subjects who were tested (unpublished data), a systematic study of cortisol metabolism will be required to address the potential role of hyperregulated or dysregulated cortisol secretion in the genesis of DCL. DCL was most commonly seen in females and in subjects with the highest total body fat, WC, or subscapular fat.

Of the patterns of fat distribution, DCL correlated best and contributed most substantially to the severity of inflammation, ballooning, Mallory hyaline, and fibrosis. The additional risks for worsening histology posed by the presence of DCL on top of generalized obesity and/or abdominal obesity were further borne out by multiple regression analyses (Table 6). DCL was associated with greater insulin resistance. The biologic basis for increased insulin resistance and severity of liver disease in subjects with DCL now warrant further investigation.

It is also noteworthy that addition of steatosis grade to BMI, WC, or DCL increased the contribution of the model to the variability in lobular inflammation. Several lipid classes are known to promote inflammation.46, 47 It is likely that lipidomic approaches will show if the severity of steatosis is associated with proinflammatory lipids and/or activation of inflammatory signaling. On the other hand, none of the fat distribution patterns correlated with the severity of steatosis. The current study group all had hepatic steatosis and therefore could not be used to evaluate the relationship between body fat and presence of steatosis. Therefore, these data simply indicate a lack of relationship between fat content and its distribution with severity of hepatic steatosis rather than a lack of relationship between adiposity and presence of steatosis. The association of steatosis with severity of obesity is indeed well documented.48, 49

In conclusion, 31% of the subjects were obese. Seventy-six percent of the subjects had abdominal obesity based on Adult Treatment Panel III criteria; 37% of these had WC commensurate with their age, sex, race, and BMI. Twenty-eight percent of the subjects were found to have DCL. Although none of the anthropometric indices correlated with the severity of steatosis, generalized obesity correlated with the severity of fibrosis. DCL correlated best and significantly with severity of all histological parameters except steatosis. Steatosis grade added to the impact of BMI, WC, and DCL on lobular inflammation. These data demonstrate the presence of distinct fat distribution patterns superimposed on generalized obesity in patients with NAFLD and that these patterns are associated with the severity of the histological changes.

References

  1. Top of page
  2. Abstract
  3. Patients and Methods
  4. Results
  5. Discussion
  6. References
  • 1
    Browning JD, Szczepaniak LS, Dobbins R, Nuremberg P, Horton JD, Cohen JC, et al. Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity. HEPATOLOGY 2004; 40: 13871395.
  • 2
    Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic experiences with a hitherto unnamed disease. Mayo Clin Proc 1980; 55: 434438.
  • 3
    Ekstedt M, Franzen LE, Mathiesen UL, Thorelius L, Holmqvist M, Bodemar G, et al. Long-term follow-up of patients with NAFLD and elevated liver enzymes. HEPATOLOGY 2006; 44: 865873.
  • 4
    Sanyal AJ. AGA technical review on nonalcoholic fatty liver disease. Gastroenterology 2002; 123: 17051725.
  • 5
    Marchesini G, Bugianesi E, Forlani G, Cerrelli F, Lenzi M, Manini R, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. HEPATOLOGY 2003; 37: 917923.
  • 6
    Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA 2002; 287: 356359.
  • 7
    James WP. Overweight and obesity. In: EzzatiA, LopezHD, RodgersHR, MurrayCJL, eds. WHO Comparative Quantification of Health Risks. Volume 1. Geneva, Switzerland: World Health Organization, 2003: 497596.
  • 8
    Bjorntorp P. Metabolic implications of body fat distribution. Diabetes Care 1991; 14: 11321143.
  • 9
    Goodpaster BH, Krishnaswami S, Harris TB, Katsiaras A, Kritchevsky SB, Simonsick EM, et al. Obesity, regional body fat distribution, and the metabolic syndrome in older men and women. Arch Intern Med 2005; 165: 777783.
  • 10
    Kaye SA, Folsom AR, Sprafka JM, Prineas RJ, Wallace RB. Increased incidence of diabetes mellitus in relation to abdominal adiposity in older women. J Clin Epidemiol 1991; 44: 329334.
  • 11
    Golay A, Chen N, Chen YD, Hollenbeck C, Reaven GM. Effect of central obesity on regulation of carbohydrate metabolism in obese patients with varying degrees of glucose tolerance. J Clin Endocrinol Metab 1990; 71: 12991304.
  • 12
    Lundgren H, Bengtsson C, Blohme G, Lapidus L, Sjostrom L. Adiposity and adipose tissue distribution in relation to incidence of diabetes in women: results from a prospective population study in Gothenburg, Sweden. Int J Obes 1989; 13: 413423.
  • 13
    Despres JP. Abdominal obesity as important component of insulin-resistance syndrome. Nutrition 1993; 9: 452459.
  • 14
    Vega GL, Adams-Huet B, Peshock R, Willett D, Shah B, Grundy SM. Influence of body fat content and distribution on variation in metabolic risk. J Clin Endocrinol Metab 2006; 91: 44594466.
  • 15
    Hadigan C, Miller K, Corcoran C, Anderson E, Basgoz N, Grinspoon S. Fasting hyperinsulinemia and changes in regional body composition in human immunodeficiency virus-infected women. J Clin Endocrinol Metab 1999; 84: 19321937.
  • 16
    Hadigan C, Corcoran C, Stanley T, Piecuch S, Klibanski A, Grinspoon S. Fasting hyperinsulinemia in human immunodeficiency virus-infected men: relationship to body composition, gonadal function, and protease inhibitor use. J Clin Endocrinol Metab 2000; 85: 3541.
  • 17
    Kleiner DE, Brunt EM, Van Natta M, Behling C, Contos MJ, Cummings OW, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. HEPATOLOGY 2005; 41: 13131321.
  • 18
    Brunt EM, Janney CG, Di Bisceqlie AM, Neuschwander-Tetri BA, Bacon BR. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol 1999; 94: 24672474.
    Direct Link:
  • 19
    Visscher TL, Seidell JC, Molarius A, van der Kuip D, Hofman A, Witteman JC. A comparison of body mass index, waist-hip ratio and waist circumference as predictors of all-cause mortality among the elderly: the Rotterdam study. Int J Obes Relat Metab Disord 2001; 25: 17301735.
  • 20
    Lohman TG. Anthropometric Standardization Reference Manual. Human Kinetics Books (Champaign, IL), 1988: 4445.
  • 21
    Taylor RW, Jones IE, Williams SM, Goulding A. Evaluation of waist circumference, waist-to-hip ratio, and the conicity index as screening tools for high trunk fat mass, as measured by dual-energy X-ray absorptiometry, in children aged 3-19 y. Am J Clin Nutr 2000; 72: 490495.
  • 22
    Wang Z, Hoy WE. Waist circumference, body mass index, hip circumference and waist-to-hip ratio as predictors of cardiovascular disease in Aboriginal people. Eur J Clin Nutr 2004; 58: 888893.
  • 23
    Wang Y, Rimm EB, Stampfer MJ, Willett WC, Hu FB. Comparison of abdominal adiposity and overall obesity in predicting risk of type 2 diabetes among men. Am J Clin Nutr 2005; 81: 555563.
  • 24
    Bruner B, Chad K, Chizen D. Effects of exercise and nutritional counseling in women with polycystic ovary syndrome. Appl Physiol Nutr Metab 2006; 31: 384391.
  • 25
    Dobbelsteyn CJ, Joffres MR, MacLean DR, Flowerdew G. A comparative evaluation of waist circumference, waist-to-hip ratio and body mass index as indicators of cardiovascular risk factors. The Canadian Heart Health Surveys. Int J Obes Relat Metab Disord 2001; 25: 652661.
  • 26
    Esmaillzadeh A, Mirmiran P, Azizi F. Clustering of metabolic abnormalities in adolescents with the hypertriglyceridemic waist phenotype. Am J Clin Nutr 2006; 83: 3646 [quiz: 183-184].
  • 27
    Snijder MB, Dekker JM, Visser M, Bouter LM, Stehouwer CD, Kostense PJ, et al. Associations of hip and thigh circumferences independent of waist circumference with the incidence of type 2 diabetes: the Hoorn Study. Am J Clin Nutr 2003; 77: 11921197.
  • 28
    Kotler DP, Burastero S, Wang J, Pierson RN Jr. Prediction of body cell mass, fat-free mass, and total body water with bioelectrical impedance analysis: effects of race, sex, and disease. Am J Clin Nutr 1996; 64(3 Suppl ): 489S497S.
  • 29
    WHO Technical Consultation on Obesity (Report Series 894). Geneva, Switzerland: World Health Organization, 2000.
  • 30
    Logue E, Smucker WD, Bourguet CC. Identification of obesity: waistlines or weight? Nutrition, Exercise, and Obesity Research Group. J Fam Pract 1995; 41: 357363.
  • 31
    Borkan GA, Gerzof SG, Robbins AH, Hults DE, Silbert CK, Silbert JE. Assessment of abdominal fat content by computed tomography. Am J Clin Nutr 1982; 36: 172177.
  • 32
    Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA 2001; 285: 24862497.
  • 33
    Okosun IS, Tedders SH, Choi S, Dever GE. Abdominal adiposity values associated with established body mass indexes in white, black and hispanic Americans. A study from the Third National Health and Nutrition Examination Survey. Int J Obes Relat Metab Disord 2000; 24: 12791285.
  • 34
    Mallon PW, Wand H, Law M, Miller J, Cooper DA, Carr A; HIV Lipodystrophy Case Definition Study; Australian Lipodystrophy Prevalence Survey Investigators. Buffalo hump seen in HIV-associated lipodystrophy is associated with hyperinsulinemia but not dyslipidemia. J Acquir Immune Defic Syndr 2005; 38: 156162.
  • 35
    Lo JC, Mulligan K, Tai VW, Algren H, Schambelan M. “Buffalo hump” in men with HIV-1 infection. Lancet 1998; 351: 867870.
  • 36
    Bari MR, Ostgren CJ, Rastam L, Lindblad U. Abdominal obesity and insulin resistance in patients with type 2 diabetes in a Swedish community. Skaraborg Hypertension and Diabetes Project. Scand J Prim Health Care 2006; 24: 211217.
  • 37
    Kuo CS, Hwu CM, Chiang SC, Hsiao LC, Weih MJ, Kao WY, et al. Waist circumference predicts insulin resistance in offspring of diabetic patients. Diabetes Nutr Metab 2002; 15: 101108.
  • 38
    Misra A, Vikram NK, Arva S, Pandey RM, Dhingra V, Chatterjee A, et al. High prevalence of insulin resistance in postpubertal Asian Indian children is associated with adverse truncal body fat patterning, abdominal adiposity and excess body fat. Int J Obes Relat Metab Disord 2004; 28: 12171226.
  • 39
    Srinivasan SR, Frontini MG, Berenson GS. Longitudinal changes in risk variables of insulin resistance syndrome from childhood to young adulthood in offspring of parents with type 2 diabetes: the Bogalusa Heart Study. Metabolism 2003; 52: 443450 [discussion: 451-453].
  • 40
    Lonardo A. Fatty liver and nonalcoholic steatohepatitis. Where do we stand and where are we going? Dig Dis 1999; 17: 8089.
  • 41
    Bellentani S, Saccoccio G, Masutti F, Croce LS, Brandi G, Sasso F, et al. Prevalence of and risk factors for hepatic steatosis in Northern Italy. Ann Intern Med 2000; 132: 112117.
  • 42
    Minocci A, Savia G, Lucantoni R, Berselli ME, Tagliaferri M, Calo G, et al. Leptin plasma concentrations are dependent on body fat distribution in obese patients. Int J Obes Relat Metab Disord 2000; 24: 11391144.
  • 43
    Wolf AM, Busch B, Kuhlmann HW, Beisiegel U. Histological changes in the liver of morbidly obese patients: correlation with metabolic parameters. Obes Surg 2005; 15: 228237.
  • 44
    Ryan MC, Wilson AM, Slavin J, Best JD, Jenkins AJ, Desmond PV. Associations between liver histology and severity of the metabolic syndrome in subjects with nonalcoholic fatty liver disease. Diabetes Care 2005; 28: 12221224.
  • 45
    Kang H, Greenson JK, Omo JT, Chao C, Peterman D, Anderson L, et al. Metabolic syndrome is associated with greater histologic severity, higher carbohydrate, and lower fat diet in patients with NAFLD. Am J Gastroenterol 2006; 101: 22472253.
    Direct Link:
  • 46
    Yamakawa T, Eguchi S, Yamakawa Y, Motley ED, Numaguchi K, Utsunomiya H, et al. Lysophosphatidylcholine stimulates MAP kinase activity in rat vascular smooth muscle cells. Hypertension 1998; 31: 248253.
  • 47
    Fang X, Gibson S, Flowers M, Furui T, Bast RC Jr, Mills GB. Lysophosphatidylcholine stimulates activator protein 1 and the c-Jun N-terminal kinase activity. J Biol Chem 1997; 272: 1368313689.
  • 48
    Dixon JB, Bhathal PS, O'Brien PE. Nonalcoholic fatty liver disease: predictors of nonalcoholic steatohepatitis and liver fibrosis in the severely obese. Gastroenterology 2001; 121: 91100.
  • 49
    Luyckx FH, Desaive C, Thiry A, Dewe W, Scheen AJ, Gielen JE, et al. Liver abnormalities in severely obese subjects: effect of drastic weight loss after gastroplasty. Int J Obes Relat Metab Disord 1998; 22: 222226.