OBJECTIVES:

Metformin proved useful in the treatment of nonalcoholic fatty liver disease (NAFLD), but its superiority over nutritional treatment and antioxidants has never been demonstrated. We aimed to compare the usefulness of metformin versus prescriptive diet or vitamin E.

METHODS:

In an open label, randomized trial, nondiabetic NAFLD patients were given metformin (2 g/day; n = 55) for 12 months. The control cases were given either vitamin E (800 IU/day; n = 28) or were treated by a prescriptive, weight-reducing diet (n = 27). Outcome measures were liver enzymes, insulin resistance (homeostasis model assessment), parameters of the metabolic syndrome, and histology.

RESULTS:

Aminotransferase levels improved in all groups, in association with weight loss. The effects in the metformin arm were larger (p < 0.0001), and alanine aminotransferase normalized in 56% of cases (odds ratio (OR) versus. controls, 3.11; 95% confidence interval (CI), 1.56–6.20; p = 0.0013). In multivariate analysis, metformin treatment was associated with higher rates of aminotransferase normalization, after correction for age, gender, basal aminotransferases, and change in body mass index (OR, 5.98; 95% CI, 2.05–17.45). Differences were maintained when the two control groups were separately analyzed. The distribution of positive criteria for the metabolic syndrome was reduced only in the metformin arm (p = 0.001, signed rank test). A control biopsy in 17 metformin-treated cases (14 nonresponders) showed a significant decrease in liver fat (p = 0.0004), necroinflammation, and fibrosis (p = 0.012 for both). No side effects were observed during metformin treatment.

CONCLUSIONS:

Metformin treatment is better than a prescriptive diet or vitamin E in the therapy of NAFLD patients receiving nutritional counseling. Limited histological data support an association between improved aminotransferases and biopsy findings, which require confirmation in a double-blind trial with appropriate statistical power based on liver histology.

INTRODUCTION

In the last few years, considerable attention has been paid to nonalcoholic fatty liver disease (NAFLD) as a condition accounting for a remarkable percentage of cirrhosis, terminal chronic liver failure, or liver transplantation. Approximately 10% of NAFLD cases have histological signs of nonalcoholic steatohepatitis (NASH), and one in three NASH cases progress to cirrhosis (¹). Accordingly, the treatment of NAFLD is now considered a relevant issue in clinical hepatology, and several therapeutic approaches have been tested in pilot, uncontrolled, or controlled studies.

The significant association of NASH with insulin resistance and the metabolic syndrome (^2,3,4) is the rational basis for treatment with weight-reducing programs and/or insulin-sensitizing agents. Nutritional treatment, either by dietary prescription or by nutritional counseling, appears to be effective in the short term (^5,6,7), and a moderate improvement in histology has also been reported (⁷). Nutritional treatment remains the background therapy in drug trials also, but systematic comparison with drugs on hard endpoints is lacking. Beneficial effects with insulin-sensitizing agents have been demonstrated in uncontrolled or pilot studies using both metformin (^8,9,10) and the peroxisome proliferators activator receptor- agonists (PPAR- troglitazone, pioglitazone, rosiglitazone) (^11,12,13), but the superiority of insulin-sensitizers versus the nutritional approach has never been demonstrated. PPAR- agonist treatment also improved histology in uncontrolled series (^11,13,14), whereas metformin effects on histology do not appear to be superior to nutritional counseling or weight-reducing diets (⁹). Metformin significantly reduces liver-cell necrosis (^8,9), but a recent study questioned the long-term effectiveness of the drug on aminotransferase levels (¹⁰). Metformin compares favorably with PPAR- on a cost-effectiveness basis, and the risk of lactic acidosis appears to be negligible in subjects with well-preserved liver cell function (^8,9).

Also, antioxidant agents have been proposed as a potentially effective treatment. Vitamin E is a potent antioxidant compound, which has been tested in pediatric NAFLD because of the absence of side effects. Conflicting results have been reported in clinical trials, both in children and in adults (^15,16,17), and the alleged superiority to placebo has been questioned (¹⁸).

Following a pilot report of metformin on NAFLD (⁸), a large, randomized, controlled study was set up in two Italian units, where metformin was compared with vitamin E supplements or a prescriptive weight-reducing program. We report the 1-yr results on biochemical parameters and features of the metabolic syndrome. Also, histology was considered as a major outcome, and liver biopsy was repeated at the end of the treatment in a limited number of cases.

MATERIAL AND METHODS

Patients

The study was carried out in 110 patients with NAFLD, randomized in two units (Bologna and Turin, Italy) specifically involved in the diagnosis and care of NAFLD cases. Their clinical and laboratory values are reported in Table 1. The large majority of patients were males, in the age range 21–68 yr (median, 43). They were recruited among subjects referred to hospital for elevated alanine aminotransferases (ALT), exceeding 1.5 times normal values for 6 months or more. None had clinical or biochemical evidence of advanced liver disease or cirrhosis, but cirrhosis was detected at liver biopsy in four cases. Exclusion criteria were alcohol consumption exceeding 20 g/day, positive screening for B and C viral hepatitis, autoimmune phenomena indicating autoimmune hepatitis or celiac disease, presence of gene markers of familial hemochromatosis. Subjects with previously diagnosed diabetes (¹⁹) were also excluded, since the majority were pharmacologically treated at the time of observation, frequently with metformin. Patients with moderate newly discovered hyperglycemia in the diabetes range were enrolled (18% in the metformin arm, 11% in vitamin E, 11% in the diet group), considering that initial treatment with nutritional counseling is recommended by international agencies (¹⁹). Severe or morbid obesity (body mass index (BMI) 35 kg/m²) was an additional exclusion criterion, since different or more intensive therapeutic approaches may be immediately considered (²⁰). The 110 randomized patients are part of a larger group of 147 NAFLD cases who consecutively had a liver biopsy in the two recruiting units in an 18-month period. After exclusion of grade 2–3 obesity and type 2 diabetes, 129 cases were left, but 19 patients did not accept the randomization procedure.

Table 1 - Clinical and Laboratory Variables in NAFLD Subjects.

Full table

Investigations at entry included a thorough investigation of habitual diet by an experienced dietitian, an oral glucose tolerance test (OGTT), and a liver biopsy.

All subjects signed an informed consent for participation in the study, which was approved by the Ethical Committee of the University Hospital of Bologna.

Study Protocol

After basal assessment, all patients received a 2-h nutritional counseling by an experienced dietitian on the basis of the dietary investigation. Counseling was aimed at reducing saturated fats and refined sugars, toward a Step I American Heart Association diet (²¹). All subjects were also encouraged to walk or to jog at least 30 min daily. Subsequently, they were randomly assigned to treatment according to different protocols in the two participating units. The experimental treatment consisted of metformin, at a daily maximum dose of 2,000 mg/day. The dosage was progressively increased from 250 mg b.i.d to the maximum dose at 500 mg weekly intervals, to reduce gastrointestinal side effects. The control treatment in the Bologna unit consisted of vitamin E, at a daily dose of 400 IU b.i.d during the whole study period. In Turin, the control treatment consisted of a weight-reducing prescriptive diet to determine a calorie deficit of 500 kcal per day (determining a weight loss of approximately 500 g/wk). The diet was tailored on individual preferences. The randomization procedure was centralized in Bologna, and based on a random sequence. Sealed envelopes were used to conceal randomization in both centers.

The treatment period lasted 12 months, with biochemical and clinical control visits every 3 months. Both patients and investigators were not blind to treatment. Patients whose ALT were reduced below the upper normal limit were considered as responders. Compliance was tested by pill counts in subjects on metformin and vitamin E, whereas in subjects on prescriptive diet only weight loss was considered as a compliance marker.

The two control treatments were considered a priori equivalent, and the initial analysis was planned to compare metformin (all cases) with nonmetformin-treated patients. Additional analyses were planned to compare patients in the metformin arm treated in different units with the respective control arms, and to compare the effectiveness of the two control arms.

The primary outcome was ALT normalization. According to a previous study, ALT were expected to decrease down to normal values in 50% of cases under metformin (⁸). The results reported in the literature by the sole nutritional approach, with/without vitamin E, were extremely variable (^6,8,22), depending on basal BMI and on the amount of weight loss. We estimated an average 30% ALT normalization by the nutritional approach. Considering a type I error of 0.05 and a type II error of 0.20, 55 subjects per arm were needed to achieve statistical significance.

A second liver biopsy was also programmed at the end of the study, but was considered optional. The ethical committee raised concern for a control biopsy in subjects with slowly progressing liver disease, particularly in subjects treated by the sole nutritional approach or by vitamins. A second biopsy is therefore available only in 17 patients treated with metformin, and in no cases of the two control groups.

Methods

All subjects had a complete clinical, anthropometric, and laboratory investigation at the time of enrollment. Clinical data included age, sex, height, weight, and BMI. Height was measured at the nearest half centimeter, and weight was recorded at the nearest half kilogram. Three blood pressure readings were obtained at 1-min intervals, and the second and third systolic and diastolic pressure readings were averaged and used in the analyses.

Laboratory investigations included glucose and insulin levels, routine liver enzymes (alanine and aspartate aminotransferases (ALT and AST), -glutamyl transpeptidase (GGT), alkaline phosphatase), lipid profile (total cholesterol, HDL cholesterol, triglycerides), assessment of iron status (serum iron, transferrin saturation, ferritin). Fasting glucose and insulin levels were used to calculate insulin resistance according to the homeostasis model assessment technique (HOMA-R) (²³).

The presence of the metabolic syndrome was diagnosed according the Adult Treatment Panel III criteria (²⁴).

Liver biopsy was performed under ultrasound guidance using a Tru-Cut needle within 1 month from clinical and laboratory evaluation. A sample length of at least 15 mm was considered as an acceptable specimen; histology was scored according to Brunt et al. (²⁵), with minor modifications. Steatosis was observed in all the biopsies, and was graded 1–3, according to the percentage of fatty infiltration (1, 0–33; 2, 34–66; 3, 67–100). Necroinflammation was graded 0–3 (0—absent; 1—occasional ballooned hepatocytes and no or very mild inflammation; 2—ballooning of the hepatocytes and mild-to-moderate portal inflammation; 3—intra-acinar inflammation and moderate portal inflammation). Fibrosis was graded 0–4 (0—absent; 1—perisinusoidal/pericellular fibrosis; 2—periportal fibrosis; 3—bridging fibrosis, and 4—cirrhosis). NASH was diagnosed on the basis of the presence of fibrosis (grade 1 or more) or necroinflammation (grade 2 or more). Staging and grading were performed by the same experienced pathologist in the two different units, who also blindly reviewed all biopsies in cases where a follow-up histology was available. A NASH index was calculated as sum of scores of fat, necroinflammation, and fibrosis.

Lactic acid levels were tested during follow-up visits in subjects treated with metformin.

Statistical Analysis

The data were processed on a personal computer and analyzed using StatView 5.0™ (SAS Institute Inc., Cary, NC).

The analysis was carried out on an intention-to-treat basis. In the course of follow-up, two patients with newly diagnosed type 2 diabetes (one on vitamin E and one on diet) received metformin (up to 2 g/day) because of progressive hyperglycemia not responding to the sole nutritional approach. Five cases (three in the metformin arm, one in vitamin E, one in the diet group) were lost to follow-up between 3 and 6 months from enrollment. These cases were included in the analysis using the last-observation-carried-forward technique. The ALT values at 3 months were elevated in these cases. Repeated-measures analysis of variance or paired t-test, unpaired t-test, (2-tail), contingency test, and Fisher's exact test were used, whenever appropriate, to compare groups at baseline and at the end of the study period. Nonparametric methods were also used for nonnormally distributed values. All data in the text and in the tables are given as means SD or as median (range) or prevalence. Data were also expressed as odds ratio (OR) and 95% confidence interval (CI).

RESULTS

Primary Outcome

The groups randomized to metformin and to the sole nutritional approach with/without vitamin E were similar in clinical and laboratory variables (Table 1). Only the overall prevalence of males was significantly lower in the metformin group (73% vs 96%; p = 0.0036, Fisher's exact test). In addition histological parameters were similar (Table 2) as was the NASH index. The percentage of hepatocytes with fatty droplets was 39 24% in the metformin group and 34 20 in the control arm (p = 0.254; unpaired t-test). Criteria for NASH were present in 86% of cases treated with metformin and 69% of cases in the control arm.

Table 2 - Histological Lesions in NAFLD Groups: Data are Expressed as Percent of Cases.

Full table

During the study period, BMI levels decreased from 28.7 3.5 kg/m² to 27.4 3.4 in the metformin arm (p < 0.0001, paired t-test), and from 28.8 3.2 to 27.8 2.9 in the control arm (p = 0.0003). No differences were observed between the two groups when tested by repeated-measures analysis of variance (ANOVA) (p = 0.113). BMI showed a progressive decline throughout the study period in the metformin group, whereas in the control arm a lower BMI was observed after 6 months (27.5 2.9) because of a moderate weight regain at 1 yr.

Aminotransferase levels decreased in all groups. Metformin treatment produced a prompt decrease of ALT levels at 3 months, which further decreased in the course of follow-up (Fig. 1), whereas ALT improvement was either lower or less sustained in both control groups (repeated-measures ANOVA; p < 0.0001). Thirty-one patients had their ALT levels within the normal range (<40 U/l) in the treatment arm versus 12 in the control arm (p = 0.0006, Logrank Mantel-Cox).

Figure 1.

Time course of alanine aminostranferases (upper panel) in subjects treated with metformin and in the control groups. Data are expressed as mean 2SE. Black columns represent cases treated with metformin, open columns are patients randomized to vitamin E, gray columns are patients randomized to prescriptive diet. The percentage of cases with normal ALT during follow-up is presented on the lower panel (p values: time treatment ANOVA, upper panel; logrank Mantel-Cox, lower panel).

Full figure and legend (32K)

Linear regression showed a significant association between percent change in BMI and AST reduction (r = 0.325; p = 0.0008). However, the association was only present in the weight-reducing treated group (r = 0.511; p = 0.013), but neither in the metformin (p = 0.093) nor in the vitamin E groups (p = 0.306). Similarly, the association was significant between change in BMI and ALT reduction (r = 0.337, p = 0.0005), and was maintained in the diet and vitamin E groups (p = 0.040 and p = 0.002, respectively), but not in the metformin arm (p = 0.215).

In multivariate analysis, after correction for age, gender, and basal ALT, response to treatment, defined as normal ALT at the end of the study period, was significantly associated with percent change in BMI (OR, 1.22; 95% CI, 1.10–1.37, p = 0.0002) and with metformin treatment (OR, 5.98; 95% CI, 2.05–17.45; p = 0.0011). Similar changes were observed in AST levels.

At baseline, the overall prevalence of the metabolic syndrome was 41% (37% in the metformin group, 50% in vitamin E arm, and 38% in the prescriptive diet group; p = 0.507).

Fasting glucose, insulin, and HOMA values showed a significant trend in favor of metformin (Fig. 2). It reduced the average fasting glucose by nearly 10 mg/dl, without inducing hypoglycemia, versus 5 mg/dl in the controls (p = 0.125). It also reduced fasting insulin by 4.5 U/ml versus 1.5 (p = 0.029), and produced a maximum decrease of HOMA of 1.5 units versus 0.5 (p = 0.0002).

Figure 2.

Time course of fasting glucose, insulin, and HOMA insulin resistance during the study period (mean 2SE). Black columns represent metformin, open columns represent vitamin E, gray columns are cases treated by prescriptive diet (p values: time treatment ANOVA).

Full figure and legend (68K)

Both HDL-cholesterol and triglyceride levels improved during the study period, and waist circumference and arterial pressure had a favorable trend. However, changes were not significantly different between groups. The distribution of positive criteria for the metabolic syndrome decreased significantly in the metformin arm (p = 0.001; Wilcoxon signed rank test), and nonsignificantly in the control arms. The overall prevalence of the metabolic syndrome at the end of the study period was 25% (metformin, 18%; p vs baseline, 0.065; vitamin E, 40%; p = 0.687; and prescriptive diet, 25%; p = 0.250).

Metformin versus Prescriptive Diet (n = 26 and n = 27)

Differences between metformin and sole dietary prescription were tested in the subset treated in Turin. Only a few patients reached the expected weight-loss target. BMI decreased by nearly 2 kg/m² in both groups (p = 0.558), but more slowly in the metformin arm. Changes in ALT levels were nonsignificantly in favor of metformin (p = 0.205), whereas changes in fasting glucose (p = 0.046), in insulin (p = 0.023), and HOMA (p = 0.0002) were significant. The percent of cases with normal ALT at the end of follow-up was 69% (metformin) versus 31% (diet) (p = 0.006; Fischer's exact test), and metformin significantly increased the chances of normal ALT at the end of the study period (OR, 2.81; 95% CI, 1.17–6.75; p = 0.021).

Metformin versus Vitamin E (n = 29 and n = 28)

The analysis of the subset treated in Bologna showed significant difference in favor of metformin. Only a few patients reached the expected weight loss target. Forty-four percent of cases in the metformin arm normalized ALT levels versus 14% on vitamin E (p = 0.019; Fisher's exact test). The beneficial effects of metformin were supported by Cox proportional hazard model (OR, 3.59; 95% CI, 1.17–11.06; p = 0.026). All metabolic parameters showed a greater improvement with metformin treatment.

Vitamin E versus Prescriptive Diet (n = 28 and n = 27)

The comparison between the two control arms failed to demonstrate any time-course difference in metabolic parameters and ALT levels, despite a moderately higher BMI loss with prescriptive diet (1.8 kg/m² vs 1.0; repeated measures ANOVA, p = 0.038). Also, the number of cases with normal liver enzymes at the end of follow-up was not different (31% with prescriptive diet vs 14%; p = 0.205, Fisher's exact test).

Follow-Up Histology

In 17 metformin-treated cases, a second liver biopsy was performed at the end of follow-up. On average, these cases (88% males, aged 23–60 yr) were similar to the whole group treated with metformin in biochemical and histological parameters (Table 3). Although a general improvement in aminotransferase levels was also observed in these patients, the main reason for acceptance of the second biopsy was failure of ALT normalization; only three cases had ALT within normal limits at the end of follow-up.

Table 3 - Clinical and Laboratory Variables in Metformin-Treated 17 NAFLD Subjects Who Had a Second Liver Biopsy at the End of Follow-Up (Means SD).

Full table

In all cases, the histological diagnosis at baseline was NASH, with severe necroinflammation and fibrosis; 3 patients had periportal fibrosis, 13 had bridging fibrosis, and one had cirrhosis (Fig. 3).

Figure 3.

Prevalence of histological lesions at baseline (left columns) and at end-of-study biopsy (right columns) in 17 metformin-treated patients (p values, paired sign test of histological grading and staging). Legend: white areas, absent; diagonal hatched area, grade 1 (mild steatosis or necroinflammation, pericellular/perisinusoidal fibrosis); gray areas, grade 2 (moderate steatosis or necroinflammation, periportal fibrosis); vertically hatched areas, grade 3 (severe steatosis or necroinflammation, bridging fibrosis); black areas, grade 4 (cirrhosis).

Full figure and legend (33K)

The second biopsy, blindly revised by the pathologist, showed a remarkable reduction in percent fat and a remarkable improvement in the distribution of necroinflammation (p = 0.012) and fibrosis (p = 0.012). In particular, either necroinflammation or fibrosis improved by at least one class in 10 cases, deteriorated in one patient, and were unchanged in 6 cases.

There was no correlation between histological changes and pretreatment parameters. Also the improvement in NASH index did not correlate with the change in ALT levels or the percent change BMI (r_s = 0.025 and r_s = 0.206, respectively). Fat infiltration (p = 0.016, paired-sign test) and fibrosis (p = 0.012) also improved in the 14 subjects whose ALT levels did not normalize, whereas necroinflammation was insignificantly reduced (p = 0.070).

Safety Monitoring and Compliance

No patients on metformin or vitamin E stopped treatment because of side effects. Time to reach the maximum metformin dose ranged from 3 to 5 wk. Compliance to treatment always exceeded 90% of prescribed dose.

Lactate levels were within 20 mg/dl in all control visits, except in a patient with recently onset diabetes, where lactic acid levels as high as 24 mg/dl were measured in the course of the treatment. In this patient, the metformin dose was limited to 1,000 mg/day, and no improvement in liver aminotransferases were observed.

DISCUSSION

The study demonstrates that long-term metformin treatment significantly reduces average ALT levels and increases the chances to have ALT within the normal ranges in NAFLD patients in comparison to control treatment (either vitamin E supplementation or a prescriptive diet inducing weight loss). Also grading and staging of histological lesions improve, not only as effects of decreased BMI, suggesting a specific action of the drug on hepatic cytolysis.

The study design is unconventional, and, although randomized in nature, it more closely resembles an observational study. Control treatment reflects current options in most gastroenterological centers, which expands the applicability of results. It was set up to validate a previous pilot study suggesting a beneficial effect of metformin on biochemical parameters (⁸) as well as on ultrasonographic variables (²⁶). Since that report has been published, a few studies confirmed the beneficial effects of metformin. Uygun et al. (⁹) showed that metformin (850 mg b.i.d for 6 months) is more effective than a lipid and calorie-restricted diet in improving ALT, which returned to normal in 59% of cases versus 37% in controls. Although more patients in the metformin group showed a histological improvement compared with dietary treatment, no differences in necroinflammation and fibrosis were demonstrated. The study had a limited sample size, with a considerable risk of type II error.

More recently, metformin (20 mg/kg per day for 48 wk) was reported to produce only a transient improvement in liver aminotransferases, with a gradual rise to pretreatment values between 6 months and 1 yr of treatment (¹⁰). Among patients with posttreatment liver biopsy, an improvement in necroinflammation and fibrosis was observed in 20 and 10%, respectively. Our results contrasted with those of prior studies. The improvement of liver aminotransferases was sustained, and there was indeed a progressive increase in the number of cases with normal ALT in the course of the study period.

The percentage of ALT normalization in our study is similar to figures reported with thiazolidinedione treatment (^11,12,13). The advantage is that body weight and BMI decrease with metformin, contrary to what may occur in a few patients treated with thiazolidinediones.

The beneficial effects of metformin were not limited to decreased body weight. Weight loss was observed also in the control groups, and BMI showed the highest decrement with the prescriptive diet. However, as frequently observed, these changes were not maintained, and weight regain is usually followed by a new increase in liver aminotransferases (⁷).

The study failed to demonstrate any additional advantage of long-term, high-dose vitamin E administration over weight loss on biochemical indexes. -Tocopherol has been extensively used in NAFLD to reduce lipid peroxidation, with inconsistent results (^15,16,17,27), and with histological improvements not superior to placebo (¹⁸). The modest improvement achieved in our vitamin E-treated patients is probably more the effect of the modest weight loss and changed lifestyles, as a consequence of nutritional counseling, than a beneficial effect of the antioxidant drug.

We were unable to obtain a systematic histological control during follow-up. Reasons were partly due to the ethical committee, who opted for an optional second biopsy, and to physicians caring for patients who remained in the control arm, who did not support the request of the enrolling centers to have a histological control. Nonetheless, the number of patients who had a second biopsy on metformin is larger than those reported in two previous studies, where only 13 and 10 cases had a histological control at the end of the study (^9,10). In our series, the beneficial effects on histology were observed also in subjects with incomplete improvement in biochemical indexes. A lack of association between biochemistry and histology has been reported in NAFLD; severe liver disease may be present also in subjects with normal liver aminotransferases (²⁸). However, the use of use of enzyme levels as surrogate markers is supported by the association of higher aminotransferases with more severe histological findings in cross-sectional studies (⁴), and by the finding that reduced aminotransferases are associated with improved histology in intervention studies (^7,12,13,29). The failure to correlate systematically biochemical improvement with improved histology has been attributed to time delay; it is possible that reduced, but persistently abnormal liver enzymes signal a remarkable histological improvement.

Also, liver fat decreased at biopsy. Metformin reversed fatty liver in leptin-deficient ob/ob mice (³⁰), and in humans with NAFLD metformin decreased liver volume (⁸) and reduced liver fat at histology (^9,10). On the contrary, in diabetic patients, no systematic improvement of liver fat, measured by proton spectroscopy, has been recently observed following a 16-wk treatment with metformin (2 g/day) (³¹), despite a significant reduction of glucose and insulin levels, reduced body weight, and improved insulin sensitivity. Liver fat in our patients was more abundant (nearly 40%) than in the diabetic patients studied by Tiikkainen et al. (15%) (³¹), and the experimental treatment was three times longer, possibly accounting for the reported differences.

Both metformin and nutritional counseling were probably effective via reduced insulin resistance. Metformin acts primarily on hepatic insulin sensitivity (³¹), but also whole-body glucose disposal increases (⁸). Similarly, decreased body weight increases insulin sensitivity. The HOMA method confirmed a superiority of metformin treatment over weight loss in NAFLD cases; drug treatment is also applicable in the small subset of cases (approximately 20% in the present series), with normal body weight, where weight loss is not a target.

Finally, metformin was better than control treatments in the overall parameters of the metabolic syndrome, whose presence identifies persons at high risk of cardiovascular mortality (^32,33). Increased body fat is probably the key factor responsible for the metabolic abnormalities (^34,35); it may be treated by the sole dietary approach, but metformin treatment was more effective in maintaining weight loss, exerting additional beneficial effects on metabolic parameters (³⁶), whose results are difficult to predict in the long term.

In summary, the study supports the use of metformin in NAFLD subjects, coupled with nutritional counseling, in an attempt to prevent disease progression. This conclusion, based on surrogate markers, is supported by limited histological data, which require confirmation in a double-blind trial with appropriate statistical power based on liver histology.

References

1. Neuschwander-Tetri, BA, Caldwell, SH. Nonalcoholic steatohepatitis: Summary of an AASLD single topic conference. Hepatology 2003;37: 1202–1219. | Article | PubMed | ISI |
2. Marchesini, G, Brizi, M, Bianchi, G, et al. Nonalcoholic fatty liver disease: A feature of the metabolic syndrome. Diabetes 2001;50: 1844–1850. | Article | PubMed | ISI | ChemPort |
3. Sanyal, AJ, Campbell-Sargent, C, Mirshahi, F, et al. Nonalcoholic steatohepatitis: Association of insulin resistance and mitochondrial abnormalities. Gastroenterology 2001;120: 1183–1192. | Article | PubMed | ISI | ChemPort |
4. Marchesini, G, Bugianesi, E, Forlani, G, et al. Nonalcoholic fatty liver, steatohepatitis, and the metabolic syndrome. Hepatology 2003;37: 917–923. | Article | PubMed |
5. Andersen, T, Gluud, C, Franzmann, MB, et al. Hepatic effects of dietary weight loss in morbidly obese subjects. J Hepatol 1991;12: 224–229. | Article | PubMed | ISI | ChemPort |
6. Ueno, T, Sugawara, H, Sujaku, K, et al. Therapeutic effects of restricted diet and exercise in obese patients with fatty liver. J Hepatol 1997;27: 103–107. | Article | PubMed | ISI | ChemPort |
7. Hickman, IE, Jonsson, JR, Prins, JB, et al. Modest weight loss and physical activity in overweight patients with chronic liver disease results in sustained improvements in alanine aminotransferase, fasting insulin, and quality of life. Gut 2004;53: 413–419. | Article | PubMed | ChemPort |
8. Marchesini, G, Brizi, M, Bianchi, G, et al. Metformin in non-alcoholic steatohepatitis. Lancet 2001;358: 893–894. | Article | PubMed | ChemPort |
9. Uygun, A, Kadayifci, A, Isik, AT, et al. Metformin in the treatment of patients with non-alcoholic steatohepatitis. Aliment Pharmacol Ther 2004;19: 537–544. | Article | PubMed | ISI | ChemPort |
10. Nair, S, Diehl, AM, Wiseman, M, et al. Metformin in the treatment of non-alcoholic steatohepatitis: A pilot open label trial. Aliment Pharmacol Ther 2004;20: 23–28. | Article | PubMed | ChemPort |
11. Caldwell, SH, Hespenheide, EE, Redick, JA, et al. A pilot study of a thiazolidinedione, troglitazone, in nonalcoholic steatohepatitis. Am J Gastroenterol 2001;96: 519–525. | Article | PubMed | ISI | ChemPort |
12. Neuschwander-Tetri, BA, Brunt, EM, Wehmeier, KR, et al. Improved nonalcoholic steatohepatitis after 48 weeks of treatment with the PPAR-gamma ligand rosiglitazone. Hepatology 2003;38: 1008–1017. | Article | PubMed | ISI | ChemPort |
13. Promrat, K, Lutchman, G, Uwaifo, GI, et al. A pilot study of pioglitazone treatment for nonalcoholic steatohepatitis. Hepatology 2004;39: 188–196. | Article | PubMed | ISI | ChemPort |
14. Neuschwander-Tetri, BA, Brunt, EM, Wehmeier, KR, et al. Interim results of a pilot study demonstrating the early effects of the PPAR-gamma ligand rosiglitazone on insulin sensitivity, aminotransferases, hepatic steatosis and body weight in patients with non-alcoholic steatohepatitis. J Hepatol 2003;38: 434–440.
15. Lavine, JE. Vitamin E treatment of nonalcoholic steatohepatitis in children: A pilot study. J Pediatr 2000;136: 734–738. | Article | PubMed | ChemPort |
16. Harrison, SA, Torgerson, S, Hayashi, P, et al. Vitamin E and vitamin C treatment improves fibrosis in patients with nonalcoholic steatohepatitis. Am J Gastroenterol 2003;98: 2485–2490. | Article | PubMed | ISI | ChemPort |
17. Vajro, P, Mandato, C, Franzese, A, et al. Vitamin E treatment in pediatric obesity-related liver disease: A randomized study. J Pediatr Gastroenterol Nutr 2004;38: 48–55.
18. Adams, LA, Angulo, P. Vitamins E and C for the treatment of NASH: Duplication of results but lack of demonstration of efficacy. Am J Gastroenterol 2003;98: 2348–2350. | Article | PubMed | ChemPort |
19. Alberti, KG, Zimmet, PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: Diagnosis and classification of diabetes mellitus. Provisional report of a WHO consultation. Diabet Med 1998;15: 539–553. | Article | PubMed | ISI | ChemPort |
20. World Health Organization. Preventing and managing the global epidemic: Report of a WHO consultation. WHO technical report series. World Health Organization: Geneva, 2000: 1–253.
21. Kris-Etherton, P, Eckel, RH, Howard, BV, et al. AHA science advisory: Lyon diet heart study. Benefits of a Mediterranean-style, national cholesterol education program/American heart association step I dietary pattern on cardiovascular disease. Circulation 2001;103: 1823–1825. | PubMed | ISI | ChemPort |
22. Knobler, H, Schattner, A, Zhornicki, T, et al. Fatty liver—an additional and treatable feature of the insulin resistance syndrome. Q J Med 1999;92: 73–79. | ISI | ChemPort |
23. Matthews, DR, Hosker, JP, Rudenski, AS, et al. Homeostasis model assessment: Insulin resistance and beta-cell function from plasma fasting glucose and insulin concentrations in man. Diabetologia 1985;28: 412–419. | Article | PubMed | ISI | ChemPort |
24. Expert 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. | Article | PubMed | ISI |
25. Brunt, EM, Janney, CG, Di Bisceglie, AM, et al. Nonalcoholic steatohepatitis: A proposal for grading and staging the histological lesions. Am J Gastroenterol 1999;94: 2467–2474. | Article | PubMed | ISI | ChemPort |
26. Magalotti, D, Marchesini, G, Ramilli, S, et al. Splanchnic haemodynamics in non-alcoholic fatty liver disease: Effect of a dietary/pharmacological treatment. A pilot study. Dig Liver Dis 2004;36: 406–411.
27. Hasegawa, T, Yoneda, M, Nakamura, K, et al. Plasma transforming growth factor-beta1 level and efficacy of alpha-tocopherol in patients with non-alcoholic steatohepatitis: A pilot study. Aliment Pharmacol Ther 2001;15: 1667–1672. | Article | PubMed | ISI | ChemPort |
28. Mofrad, P, Contos, MJ, Haque, M, et al. Clinical and histologic spectrum of nonalcoholic fatty liver disease associated with normal ALT values. Hepatology 2003;37: 1286–1292. | Article | PubMed | ISI |
29. Dixon, JB, Bhathal, PS, Hughes, NR, et al. Nonalcoholic fatty liver disease: Improvement in liver histological analysis with weight loss. Hepatology 2004;39: 1647–1654. | Article | PubMed |
30. Lin, HZ, Yang, SQ, Chuckaree, C, et al. Metformin reverses fatty liver disease in obese, leptin-deficient mice. Nat Med 2000;6: 998–1003. | Article | PubMed | ISI | ChemPort |
31. Tiikkainen, M, Hakkinen, AM, Korsheninnikova, E, et al. Effects of rosiglitazone and metformin on liver fat content, hepatic insulin resistance, insulin clearance, and gene expression in adipose tissue in patients with type 2 diabetes. Diabetes 2004;53: 2169–2176. | Article | PubMed | ISI | ChemPort |
32. Isomaa, B, Almgren, P, Tuomi, T, et al. Cardiovascular morbidity and mortality associated with the metabolic syndrome. Diabetes Care 2001;24: 683–689. | Article | PubMed | ISI | ChemPort |
33. Hu, G, Qiao, Q, Tuomilehto, J, et al. Prevalence of the metabolic syndrome and its relation to all-cause and cardiovascular mortality in nondiabetic European men and women. Arch Intern Med 2004;164: 1066–1076. | Article | PubMed |
34. Maison, P, Byrne, CD, Hales, CN, et al. Do different dimensions of the metabolic syndrome change together over time? Evidence supporting obesity as the central feature. Diabetes Care 2001;24: 1758–1763. | Article | PubMed | ISI | ChemPort |
35. Anderson, PJ, Critchley, JA, Chan, JC, et al. Factor analysis of the metabolic syndrome: Obesity vs insulin resistance as the central abnormality. Int J Obes Relat Metab Disord 2001;25: 1782–1788. | Article | PubMed | ChemPort |
36. Mather, KJ, Verma, S, Anderson, TJ. Improved endothelial function with metformin in type 2 diabetes mellitus. J Am Coll Cardiol 2001;37: 1344–1350. | Article | PubMed | ISI | ChemPort |