Steatohepatitis/Metabolic Liver Disease
Article first published online: 8 AUG 2011
Copyright © 2011 American Association for the Study of Liver Diseases
Volume 54, Issue 3, pages 837–845, 2 September 2011
How to Cite
Lomonaco, R., Ortiz-Lopez, C., Orsak, B., Finch, J., Webb, A., Bril, F., Louden, C., Tio, F. and Cusi, K. (2011), Role of ethnicity in overweight and obese patients with nonalcoholic steatohepatitis. Hepatology, 54: 837–845. doi: 10.1002/hep.24483
Potential conflict of interest: Nothing to report.
Supported by the Burroughs Wellcome Fund (K. C.), the American Diabetes Association (1-08-CR-08 [K. C.]), and a VA Merit Award (1 I01 CX000167-01 [K. C.]). Additional support was received from the Veterans Affairs Medical Research Fund and by Award No. UL 1RR025767 from the National Center for Research Resources.
The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources of the National Institutes of Health.
- Issue published online: 25 AUG 2011
- Article first published online: 8 AUG 2011
- Accepted manuscript online: 14 JUN 2011 08:06AM EST
- Manuscript Accepted: 28 MAY 2011
- Manuscript Received: 29 APR 2011
The role of ethnicity in determining disease severity in nonalcoholic steatohepatitis (NASH) remains unclear. We recruited 152 patients with biopsy-proven NASH, 63% of whom were Hispanic and 37% of whom were Caucasian. Both groups were well matched for age, sex, and total body fat. We measured: (1) liver fat by magnetic resonance imaging and spectroscopy; (2) fasting plasma glucose, fasting plasma insulin (FPI), and free fatty acid (FFA) levels; (3) total body fat by dual energy x-ray absorptiometry (DXA); (4) liver and muscle insulin sensitivity (insulin clamp with 3-[3H] glucose); (5) insulin resistance at the level of the liver (fasting endogenous glucose production derived from 3-[3H] glucose infusion × FPI) and adipose tissue (fasting FFA × FPI). Liver fat was slightly, but not significantly, higher in Hispanic vs. Caucasian patients (27 ± 2% vs. 24 ± 2%, p = 0.16). However, this trend did not translate into worse liver steatosis, necroinflammation or fibrosis. Patients with NASH had severe hepatic, adipose tissue and muscle insulin resistance versus healthy subjects without NASH nonalcoholic fatty liver disease, but there were no differences between both ethnic groups on these parameters. However, Hispanics versus Caucasians with type 2 diabetes mellitus (T2DM) had a trend for worse hepatic/adipose tissue insulin resistance and fibrosis. Conclusion: When Hispanic and Caucasian patients with NASH are well matched for clinical parameters, particularly for adiposity, slightly higher liver fat content is not associated with worse hepatic insulin resistance or more severe NASH on histology. Hispanic ethnicity does not appear to be a major determinant of disease severity in NASH, although those with diabetes may be at greater risk of fibrosis. Given the higher risk of T2DM in Hispanics, long-term studies are needed to define their risk of disease progression. (HEPATOLOGY 2011;)
Nonalcoholic fatty liver disease (NAFLD) represents a broad spectrum of clinical and histopathological manifestations, ranging from mild hepatic steatosis through nonalcoholic steatohepatitis (NASH), to fibrosis and ultimately cirrhosis and hepatocellular carcinoma.1 NAFLD is frequently associated with obesity, insulin resistance, dyslipidemia, and type 2 diabetes mellitus (T2DM), all of which are components of the metabolic syndrome (MetS).2 However, not all individuals with MetS develop hepatic steatosis, nor do all individuals with hepatic steatosis develop NASH or cirrhosis.3 Thus, the factors leading to steatosis and steatohepatitis in humans remain poorly understood.
Among potential factors for the development of NASH, ethnicity is believed to be an independent risk factor for NASH that has recently received increasing attention.3-11 Several studies have suggested a significant variation in the risk for NAFLD and disease severity based on ethnicity, with Hispanics believed to be more and African Americans less predisposed to develop NAFLD compared with Caucasians.3-6, 8-10 However, these studies had limitations posed either by their retrospective nature or by the fact that groups were not carefully matched for major clinical variables—namely, Hispanics were usually more obese or had more diabetes or features of MetS (e.g., higher levels of triglycerides and lower levels of high-density lipoprotein cholesterol). Previous studies also have the shortcoming of having used surrogate markers of NASH (e.g., elevated aminotransferase levels or imaging)3, 4, 6, 11, 12 rather than a histological diagnosis when comparing both ethnic groups. In addition, none of the studies performed an assessment of hepatic, adipose tissue, or muscle insulin sensitivity using glucose turnover measurements when comparing Hispanic versus Caucasian subjects with NASH.
The aim of this study was to determine the role of ethnicity (Hispanic versus Caucasian) in the severity of NASH and whether differences could be explained by the degree of hepatic, adipose tissue, and muscle insulin resistance between ethnic groups.
Patients and Methods
A total of 152 overweight or obese patients were recruited from the general population of San Antonio, Texas. Patients with elevated liver aminotransferases and/or hepatic steatosis on magnetic resonance spectroscopy (MRS) were identified either from responses to local newspaper advertisements or from referrals from Hepatology clinics at the University of Texas Health Science Center at San Antonio, Texas (UTHSCSA) or the VA Medical Center. The study included 45 participants with biopsy-proven NASH previously reported.13 Ten healthy subjects without T2DM and without fatty liver by MRS served as controls for the metabolic studies. Participants were in good general health without evidence of any chronic diseases (other than NASH or T2DM) as determined by history, physical examination, routine blood chemistries, urinalysis, and electrocardiography. No subject was involved in strenuous physical activity and had a stable body weight (±2%) for at least 3 months before the study. Volunteers were excluded if they had a history of alcohol abuse (≥20 g/day); liver disease other than NASH (hepatitis B or C, autoimmune hepatitis, hemochromatosis, Wilson disease, drug-induced disease, other); type 1 diabetes; or a history of clinically significant renal, pulmonary, or heart disease (New York Heart Classification greater than grade II). The study was approved by the UTHSCSA Institutional Review Board, and informed written consent was obtained from each patient prior to participation.
All studies were performed at the Frederic C. Bartter Clinical Research Unit (except liver imaging, which was performed at the UTHSCSA Research Imaging Center, and liver biopsy, which was performed at the VA Radiology Department). Baseline metabolic measurements included: (1) fasting plasma glucose, A1c, lipid profile, liver function tests, insulin, free fatty acid (FFA); (2) whole body fat by dual energy x-ray absorptiometry (DXA); (3) hepatic fat content by MRS; (4) 75-g oral glucose tolerance test to establish the diagnosis of normal glucose tolerance or diabetes according to American Diabetes Association criteria14; (5) euglycemic hyperinsulinemic clamp with 3-[3H] glucose to measure endogenous (primarily hepatic) and total body (largely muscle) insulin sensitivity15; (6) liver biopsy for histology to confirm the diagnosis of NASH and establish the stage and grade of the disease.
Measurements of Total Body and Liver Fat Content.
Total body fat content was measured by DXA (Hologic Inc, Waltham, MA). For the measurement of hepatic fat content, localized 1H NMR spectra of the liver were acquired on a Siemens TIM TRIO 3.0T MRI whole body scanner as described.13 In brief, two areas of interest were taken using an echo time/repetition time/angle of 30 milliseconds/2,000 milliseconds/90 degrees, and two liver areas with a volume of 30 × 30 × 30 mm were used. A liver fat content of >5.5% was considered diagnostic of NAFLD.12
Euglycemic Hyperinsulinemic Clamp.
Patients were admitted to the research unit at 6:30 AM after a 12-hour overnight fast, and the study was performed as reported by our group.16 In brief, upon arrival at the unit, a polyethylene catheter was inserted into an antecubital vein for infusion of all test substances. A second catheter was inserted retrogradely into an ipsilateral wrist vein on the dorsum of the hand for collection of arterialized blood sampling, and the hand was kept in a heated box at 65°C. A primed (25 μCi × [fasting glucose/100])–continuous (0.25 μCi/minute) infusion of 3-[3H] glucose (DuPont-NEN, Boston, MA) was initiated and continued until the end of the study. During the last 30 minutes of the basal equilibration period (150-180 minutes), plasma samples were taken at 5- to 10-minute intervals for determination of plasma glucose, insulin concentrations, and 3-[3H] glucose-specific activity. After the basal equilibration period, insulin was administered as a primed-continuous infusion at 10 μU/m2·minute for 120 minutes to assess suppression of endogenous (hepatic) glucose production, followed by another 2 hours at an infusion rate of 80 μU/m2·minute for 120 minutes to assess whole body insulin stimulated glucose disposal (Rd). The plasma glucose level was measured every 5 minutes after start of insulin, and a variable infusion of 20% glucose was adjusted based on the negative feedback principle to maintain the plasma glucose concentration at ≈90-100 mg/dL with a coefficient of variation <5%. Plasma samples were collected every 5-10 minutes for determination of plasma glucose, insulin, and FFA concentrations and 3-[3H] glucose-specific activity.
An ultrasound-guided liver biopsy was performed in patients with elevated liver aminotransferase levels when all other causes of liver disease were ruled out, or normal liver aminotransferase levels with NAFLD by MRS and well-known risk factors for NASH such as T2DM, MetS, and/or insulin resistance as established during a euglycemic insulin clamp. Histopathologic characteristics for the diagnosis of NASH were determined using standard criteria.17 Biopsies were evaluated by an experienced pathologist that was unaware of the patients' identity or clinical information with a good-to-excellent intraobserver agreement between readings (weighted kappa coefficient 0.84 for steatosis, 0.69 for necroinflammation, and 0.82 for fibrosis).13 Only 7% of patients had a biopsy length less than 1 centimeter, which is within the 14% reported by other groups.7
Plasma glucose was measured by the glucose oxidase method (Analox Glucose Analyzer; Analox Instruments, Lunenburg, MA), plasma insulin by radioimmunoassay, and plasma FFA by standard colorimetric methods. A1c level was measured using high-performance liquid chromatography (TOSOH G-7). The 3-[3H] glucose-specific activity was measured on barium hydroxide/zinc sulfate–deproteinized plasma samples as described.16, 18
Both endogenous (hepatic) glucose production (EGP) and insulin stimulated glucose disposal (Rd) were calculated as reported by our group.16, 18 We also calculated an index of hepatic insulin resistance (HIRi) and of adipose tissue insulin resistance (Adipo-IRi) as described.19-23 The rationale for both indexes is based on the linear relationship between the rise in the fasting plasma insulin (FPI) level and the decline in the rate of basal (fasting) EGP in healthy subjects. The higher the rate of EGP and the level of FPI, the greater the severity of hepatic insulin resistance. Therefore, a hepatic insulin resistance index was calculated as the product of fasting EGP and FPI concentration (HIRi = EGP × FPI [mg·kg−1·minute−1·μU/mL]). Insulin is also a strong inhibitor of lipolysis, and a similar relationship exists in healthy subjects between the FPI concentration and fasting plasma FFA levels.23 Thus, an index of adipocyte insulin resistance was calculated as the product of the fasting plasma FFA and insulin concentration (Adipo-IRi = FFA × FPI [mmol/L·μU/mL]). Experimental validation for both indexes has been reported by our group.19-23
All values are reported as the mean ± SEM for continuous variables and the number (percent) for categorical variables. Comparison of ethnic groups was performed using analysis of variance or Kruskal-Wallis for continuous variables or Pearson's chi-square or Fisher's exact test for categorical variables. Adjusted P values were calculated using fixed effect models. Statistical significance was set at P < 0.05. All statistical calculations were performed using SAS version 9.2 software (SAS, Cary, NC).
The patient characteristics of the study population are summarized in Table 1. The Hispanic and Caucasian groups were closely matched, with no significant differences regarding age, sex, prevalence of MetS, body mass index (BMI) or total body fat, A1c, aminotransferase levels, or lipid panel, with the exception that Hispanics had a higher prevalence of T2DM (62% versus 41%, P = 0.02) and a higher fasting plasma insulin concentration (18 ± 1 versus 14 ± 2 μU/mL, P = 0.05). The proportion of patients with NASH and normal liver aminotransferase levels was similar in both groups (aspartate aminotransferase [AST], 54% versus 60%; alanine aminotransferase [ALT], 25% versus 32%; P = not significant). Of note, the percent liver fat measured by MRS was slightly but not significantly higher in overweight and obese Hispanic versus Caucasian patients (27 ± 2% versus 24 ± 2%, respectively; P = 0.16). This remained true even after adjusting for total body fat, diabetes, and MetS. A group of healthy subjects without NAFLD or T2DM was also studied as a control for the parameters related to insulin sensitivity in liver, adipose tissue, and muscle (age, 43 ± 3 years; BMI, 29 ± 2 kg/m2; total body fat by DXA, 29 ± 2%; fasting plasma glucose, 98 ± 9 mg/dL; A1c, 5.4 ± 0.3%; fasting plasma insulin, 3 ± 1 μU/L; fasting plasma FFA, 456 ± 79 μmol/L).
|Hispanics (n = 96)||Caucasians (n = 56)||P*|
|Age, years||50 ± 1||53 ± 1||0.07|
|Sex, %, M/F||62/38||62/38||0.86|
|Total body fat, %||34 ± 1||36 ± 1||0.18|
|BMI, kg/m2||34 ± 1||34 ± 1||0.86|
|Liver fat by MRS, %||27 ± 2||24 ± 2||0.16|
|Presence of MetS, %†||93||91||0.76|
|Presence of T2DM, %||62||41||0.02|
|Patients with diabetes||6.9 ± 1.1||6.8 ± 0.9||0.80|
|Patients without diabetes||5.8 ± 0.9||5.6 ± 0.6||0.51|
|ALT, IU/L||65 ± 3||59 ± 4||0.12|
|AST, IU/L||47 ± 2||43 ± 3||0.13|
|Triglycerides, mg/dL||193 ± 15||175 ± 16||0.94|
|LDL-C, mg/dL‡||113 ± 4||112 ± 5||0.91|
|HDL-C, mg/dL||39 ± 1||37 ± 1||0.33|
|Fasting glucose, mg/dL||119 ± 3||123 ± 5||0.91|
|Fasting insulin, μU/L||18 ± 1||14 ± 2||0.05|
|Fasting FFA, μmol/L||611 ± 26||614 ± 35||0.55|
Role of Ethnicity in Liver Histology and the Severity of NASH in Overweight or Obese Patients.
We compared the role of ethnicity (Hispanic versus Caucasian) in the histological features of NASH (Fig. 1, Table 2). There were no significant differences in the mean scores for steatosis (Fig. 1A), ballooning necrosis (Fig. 1B), lobular inflammation (Fig. 1C), or fibrosis stage (Fig. 1D). The trend toward worse fibrosis among Hispanic patients compared with Caucasian patients was entirely driven by patients with T2DM, the fibrosis stage being identical among nondiabetics (0.9 ± 0.2 versus 1.0 ± 0.2, respectively; P = 0.59). The histological findings were also similar with further analysis using the breakdown described in Table 2, where it can be appreciated that steatosis and lobular inflammation had a similar proportion of patients with grades 1, 2, or 3 as well as for fibrosis stages 0-4. There were no significant differences between both ethnic groups when we integrated the above results as the NAFLD activity score or NAS (the sum of steatosis, lobular inflammation and ballooning necrosis: 4.7 ± 0.1 versus 4.9 ± 0.2; P = 0.4).
|1 (5%-33%)||25 (26%)||21 (38%)|
|2 (>33%-66%)||37 (39%)||17 (30%)|
|3 (>66%)||34 (35%)||18 (32%)|
|0 (None)||5 (5%)||1 (2%)|
|1 (Few ballooning cells)||71 (74%)||40 (71%)|
|2 (Many ballooning cells)||20 (21%)||15 (27%)|
|0 (No foci)||0 (0%)||1 (2%)|
|1 (<2 foci per 200× field)||35 (37%)||23 (41%)|
|2 (2-4 foci per 200× field)||60 (62%)||32 (57%)|
|3 (>4 foci per 200× field)||1 (1%)||0 (0%)|
|0 (None)||26 (27%)||19 (34%)|
|1 (Perisinusoidal or periportal)||38 (40%)||25 (44%)|
|2 (Perisinusoidal and portal or periportal)||14 (15%)||6 (11%)|
|3 (Bridging fibrosis)||14 (15%)||5 (9%)|
|4 (Cirrhosis)||3 (3%)||1 (2%)|
Role of Ethnicity in Liver, Adipose Tissue, and Muscle Insulin Sensitivity in Overweight or Obese Patients with NASH.
We examined the effect of insulin resistance across different target tissues in both ethnic groups. Figure 2A represents the HIRi, a validated index of hepatic insulin sensitivity in the fasting state,21, 24 as the product of the fasting EGP (largely hepatic) times the plasma insulin concentration. Patients with NASH had severe hepatic insulin resistance compared with healthy controls without NAFLD, either measured as the HIRi (both groups together versus controls 26.3 ± 2.1 versus 8.4 ± 0.6 mg·kg−1·minute−1·μU/mL; P < 0.01) (Fig. 2A) or the suppression of EGP (hepatic) by low-dose insulin infusion during the euglycemic insulin clamp (both groups together versus controls −41 ± 2% versus −59 ± 6%; P < 0.01) (Fig. 2B).
The HIRi was not different between Hispanic and Caucasian patients (26.8 ± 2.7 versus 25.3 ± 4.0 mg·kg−1·minute−1·μU/mL, respectively; P = 0.76) (Fig. 2A). Consistent with the above findings, suppression of EGP by low-dose insulin infusion during the euglycemic insulin clamp was also similar among Hispanics versus Caucasians (−39 ± 3% versus −46 ± 4%, respectively; P = 0.13) (Fig. 2B).
Because of the important role of adipose tissue insulin resistance in the pathogenesis of NASH,25, 26 we examined its role by using the validated adipose tissue insulin resistance index or Adipo-IRi21, 24 derived from the product of the fasting plasma FFA and insulin concentration (Fig. 3). Patients with NASH had severe insulin resistance at the level of adipose tissue, with the Adipo-IRi being four- to five-fold higher (worse) than in healthy controls without fatty liver (9.7 ± 0.6 versus 2.1 ± 0.3 mmol/L·μU/mL; P = 0.004). In Fig. 3A, it can also be appreciated that although there was a trend toward worse insulin resistance in Hispanics compared with their Caucasian counterparts, both ethnic groups had a similar decrease in adipose tissue insulin sensitivity overall (10.5 ± 0.8 versus 8.2 ± 1.1 mmol/L·μU/mL, respectively; P = 0.09). We also examined directly the suppression of plasma FFA concentration by way of low-dose insulin infusion (Fig. 3B). Consistent with the Adipo-IRi results, patients with NASH demonstrated again a diminished adipose tissue response to insulin compared with control subjects without a fatty liver (−44 ± 2% versus −74 ± 6%, respectively; P < 0.0001). However, we noted no differences when both ethnic groups were compared (Fig. 3B).
Figure 4 examines insulin-stimulated muscle glucose disposal (Rd) during the high-dose euglycemic insulin clamp. As with insulin resistance at the level of the liver and adipose tissue, patients with NASH were very insulin resistant compared with controls without NAFLD (5.7 ± 0.3 versus 14.3 ± 0.8 mg·kgLBM−1·minute−1, P < 0.0001). However, there were no significant differences between Hispanic and Caucasian patients (5.7 ± 0.4 versus 5.7 ± 0.5 mg·kgLBM−1·minute−1; P = 0.64).
Effect of Ethnicity on the Severity of NASH and Insulin Sensitivity in Overweight or Obese Patients with T2DM.
When we analyzed only the subgroup of patients with T2DM (60 Hispanic patients and 23 Caucasian patients), we found no significant differences between ethnic groups in the mean scores of steatosis, lobular inflammation, and ballooning necrosis. However, there was a trend toward worse fibrosis among Hispanic versus Caucasian patients with diabetes (1.5 ± 0.1 versus 1.0 ± 0.2, P = 0.052).
As shown in Table 3, Hispanic versus Caucasian patients with NASH and T2DM had similar degrees of insulin resistance at all levels examined (liver, adipose tissue, and skeletal muscle), although there was a trend for the HIRi and the Adipo-IRi to be slightly worse in Hispanic patients.
|HIRi, mg·kg−1·minute−1·μU/mL||31.8 ± 4.2||25.4 ± 6.2||0.39|
|Suppression of hepatic glucose production by insulin, %||−36 ± 3||−45 ± 5||0.12|
|Adipo-IRi, mmol/L·μU/mL||9.8 ± 1.4||6.4 ± 2.1||0.18|
|Suppression of FFA by insulin, %||−35 ± 3||−35 ± 5||0.92|
|Muscle insulin sensitivity, mg·kg−1·minute−1||4.7 ± 0.4||5.7 ± 0.6||0.53|
The aim of this study was to identify whether Hispanic compared with Caucasian inidividuals are at greater risk of more severe NASH. We felt this issue to be clinically relevant because Hispanics are an increasing segment of the United States population, and they cluster metabolic risk factors that may promote the development of hepatic steatosis such as obesity, T2DM and MetS.27 Indeed, prior studies have supported this notion3, 10, 28, 29 and there have been reports suggesting that Hispanics may have a disproportionally high prevalence of NAFLD-related cirrhosis.30 Unfortunately, careful metabolic and histological studies have been lacking. This study aims to fill this knowledge gap by becoming the first comprehensive comparison of NASH and associated metabolic factors in Hispanic versus Caucasian individuals. In contrast with previous reports,3-5 in the present study Hispanics and Caucasians were closely matched for all relevant variables, both clinically (BMI, total body fat, and prevalence of MetS) and biochemically (similar degree of glycemic control in diabetics and proportion of patients with elevated plasma liver aminotransferases, lipids, and FFA concentrations). We also took special care to assess the degree of hepatic steatosis, not only by histology, but also by the gold standard MRS imaging technique,12 and assessed key metabolic parameters using state-of-the-art glucose turnover measurements. Taken together, this study design provided the optimal conditions to address the issue as to whether patients of Hispanic ancestry are at greater risk of developing more severe disease than Caucasians.
Consistent with previous studies, Hispanics showed a trend toward slightly higher (although not significant) hepatic fat content by MRS (27 ± 2% versus 24 ± 2%; P = 0.16). Of note, the notion that Hispanics have higher liver fat carries on from the initial 2004 report by the Dallas Heart Study3 in which Hispanic women (but not men) compared with Caucasian women had nearly a two-fold higher prevalence of NAFLD (45% versus 24%). This was confirmed in a more recent report from this group5 and in Hispanics as a group by other investigators.6, 10 However, in these reports, Hispanic groups were usually enriched with unfavorable metabolic factors by having more obesity, T2DM, and insulin resistance compared with their Caucasian counterparts, making the final interpretation of the risk for steatosis difficult. Instead, when Hispanic and Caucasian subjects are well matched for obesity, as in the current study (even as Hispanics had a higher rate of diabetes), differences in hepatic steatosis by MRS are minimal and not overall significant. More importantly, there was no difference in the severity of NASH by histology. Few studies have analyzed the severity of histological disease in subjects of Hispanic versus Caucasian ancestry, but overall the results have been inconsistent. Hispanics either have had more ballooning and Mallory bodies4 and a stronger association with definitive NASH if the NAFLD activity score is ≥5,4, 31 or on the contrary, less advanced fibrosis.7 In part, the inconsistencies could be related to the small proportion of Hispanics included in these cohorts (≈12%-14%). Some studies have compared extremely obese subjects (BMI ≥45 kg/m2)9, 32 but not the more commonly observed overweight or mildly obese subject with NASH as reported here. Thus, our observation of similar histology in both ethnic groups is a departure from currently held beliefs but nevertheless highlights the importance of controlling obesity and associated unfavorable metabolic factors in the Hispanic population for the prevention of steatosis and NASH.
We examined carefully whether Hispanics had worse insulin resistance at the level of the liver, adipose tissue, and skeletal muscle. Of note, patients with NASH were very insulin resistant in all target tissues, and plasma FFA was significantly higher compared with healthy control subjects (612 ± 21 versus 456 ± 79 μmol/L; P < 0.05). This finding supports an important role of elevated rates of lipolysis/plasma FFA concentration and lipotoxicity in the pathogenesis of NASH.26, 33, 34 Of note, we did not observe any significant difference in either hepatic or adipose tissue insulin resistance among ethnic groups using two different approaches, the fasting EGP and plasma FFA levels in relation to the ambient fasting plasma insulin concentration, respectively, or in the direct response to a 2-hour suppression by low-dose insulin infusion during the euglycemic insulin clamp studies. Therefore, it is likely that clinically relevant differences do not exist between both ethnic groups when well matched for adiposity, even as there was a small trend toward worse hepatic (and even adipose tissue) insulin resistance in Hispanics. A few studies have compared Hispanic subjects with other ethnic groups using the homeostatic model assessment (HOMA) (fasting plasma glucose × insulin concentration), which is a useful epidemiological tool but a rather crude indicator of hepatic insulin resistance in patients with NAFLD.3, 5, 35 Reports indicate either similar35 or worse3, 5 insulin resistance by HOMA in Hispanics, but again, Hispanics usually had a worse metabolic profile in these studies, as discussed earlier. Only one study by Gastaldelli and colleagues24 compared both ethnic groups in 232 patients with or without T2DM (although they did not assess for NAFLD or NASH by MRS) using in-depth metabolic studies. Consistent with the current report, neither liver nor adipose tissue insulin resistance was different when Hispanics were compared with Caucasians, although muscle insulin-stimulated glucose disposal was somewhat lower in Hispanics.
Both groups had well matched aminotransferase levels and the same proportion of patients with increased liver aminotransferase levels. As in previous reports,3, 5, 7, 13 liver aminotransferase levels were not sensitive enough to assist in the detection of NAFLD, as the majority of patients with NASH had normal liver enzymes. Moreover, we found no correlation between plasma AST/ALT and insulin resistance at any level (i.e., hepatic, adipose tissue, or muscle). In addition, the correlation of AST/ALT with liver fat content by MRS and the overall NAS was weak and not significant, questioning the clinical value as a test for the screening of follow-up of NASH patients.
This study does have some limitations. First, African Americans were not included. This is because the ethnic mix of the San Antonio area is 61% Hispanic, 30% Caucasian and only 9% African American and other ethnicities. Therefore, we were unable to study enough African American patients to make a conclusive assessment relative to Hispanics and Caucasians. It must also be kept in mind that the Hispanic group was largely composed of patients of Mexican American ancestry, hence the findings may not apply to other Hispanic populations. Future work will clarify the impact of other ethnic backgrounds in NASH. Second, although there was no major insulin sensitivity nor were there histological differences between patients of Hispanic and Caucasian ancestry, there was a trend among Hispanics to have worse hepatic insulin resistance and for diabetics to have more severe liver fibrosis on histology. This deserves further evaluation in relation to potential pharmacological treatment response, particularly insulin sensitizers such as pioglitazone. More pronounced insulin resistance or fibrosis in Hispanics may highlight a group in need of treatment at an earlier stage, either because they may be more likely to respond to thiazolidinedione therapy or be at greater risk of liver fibrosis. Pioglitazone has proven to reverse hepatic insulin resistance in previous studies from our laboratory13 and others.36 Unfortunately, the relatively small number of subjects studied in our previous proof-of-concept study (n = 55 of which 50% were Hispanic)13 and the few Hispanics enrolled in the PIVENS study (37/247 [15%])36 do not allow firm conclusions at this time. Ongoing studies will likely shed light on this issue in the future.37, 38 Third, we only reported on overweight and obese subjects with NAFLD, though they represent the vast majority of patients affected. Whether the current findings apply to lean subjects remains to be studied. Finally, the current results should not be erroneously interpreted to mean that Hispanics are not at an increased risk of more severe NASH in the long term. Risk factors such as obesity, T2DM, and MetS occur more frequently and at an earlier age in Hispanics compared with other ethnic groups,27 and the cross-sectional nature of this study tells us nothing about the natural history of the disease and their lifetime risk of severe disease. In addition to environmental factors, it will also be of interest to examine genetic determinants of hepatic steatosis and NASH, such as PNPLA3 I148M and other polymorphisms,39, 40 which appear to be more common in Hispanics.39, 41 However, it should be noted that these polymorphisms explained only a minority of the heritability of steatosis in a large meta-analysis of genome-wide association studies.40 Moreover, the PNPLA3 I148M is not associated with insulin resistance,42-44 and the liver steatosis differences between carriers and noncarriers of the polymorphism is only ≈4%-5% in absolute terms (8%-14% versus 4%-10%, respectively, when liver fat is measured by MRS).39, 41, 43 Of note, this was the range of hepatic steatosis difference observed between Hispanics and Caucasians in the present study (27% versus 24%, respectively). Again, much more work in this area is needed to place the current findings in their true perspective.
In conclusion, the current study demonstrates that Hispanics and Caucasians have similar insulin resistance and severity of NASH when matched for major clinical variables, in particular for total body fat, and that reported differences were more likely a reflection of a more unfavorable metabolic risk of Hispanics compared with Caucasians. However, a reduction in vigilance in the Hispanic population with NASH is not the take-home message given their usual worse metabolic profile. Only longitudinal studies may fully establish the natural history of the disease in this ethnic group. Further work is needed to fully understand the role of hepatic steatosis in individuals of Hispanic ethnicity in relation to the natural history of the disease, its long-term effect on the risk of cirrhosis, and the response to pharmacological treatment of this population.
We thank our study volunteers, the Clinical Translational Science Award (CTSA) nursing staff (in particular, Norma Diaz and Rose Kaminski-Graham), and the nutrition and laboratory staff for assistance in performing the described studies.
- 27Unmet needs in Hispanic patients with type 2 diabetes mellitus. Am J Med 2011. In press., .
- 28Diabetes—United States, 2004 and 2008. MMWR Surveill Summ 2011; 60( Suppl.): 90-93., , .
- 29Obesity—United States, 1988-2008. MMWR Surveill Summ 2011; 60( Suppl.): 73-77..
- 30Ethnic differences in the prevalence of cryptogenic cirrhosis. Am J Gastroenterol 2004; 99: 292-298., , , .Direct Link:
- 34Dysfunctional adipose tissue, lipotoxicity and NAFLD. Gatroenterology. In press..
- 37Assessment of ethnic and metabolic factors in the development of NASH in a predominantly Hispanic population: the University of Texas at San Antonio NASH Trial [Abstract]. HEPATOLOGY 2010; 52( Suppl.): 653A., , , , , , et al.
- 38Department of Veteran Affairs. Prevalence of non-alcoholic fatty liver disease (NAFLD) in Hispanics with diabetes mellitus type 2 (T2DM) and role of treatment (VA NASH). http://clinicaltrials.gov/ct2/show/NCT01002547. Accessed April 28, 2011.
- 45Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. 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: 2486-2497.