Share this article on:

Pharmacological Interventions for Nonalcoholic Fatty Liver Disease in Adults and in Children: A Systematic Review

Socha, Piotr*; Horvath, Andrea; Vajro, Pietro; Dziechciarz, Piotr; Dhawan, Anil§; Szajewska, Hania

Journal of Pediatric Gastroenterology and Nutrition: May 2009 - Volume 48 - Issue 5 - p 587–596
doi: 10.1097/MPG.0b013e31818e04d1
Original Articles: Hepatology and Nutrition

Background: Uncertainty exists regarding the treatment of patients with nonalcoholic fatty liver disease (NAFLD) who are unable to lose weight and/or change lifestyle. The present study assesses the effectiveness and safety of pharmacological and dietary supplement interventions for NAFLD.

Methods: MEDLINE, EMBASE, and the Cochrane Library were searched for randomized controlled trials (RCTs) both in adults and in children.

Results: Fifteen (2 pediatric patients and 13 adults) RCTs met the inclusion criteria. A significant effect on normalization of alanine transaminase was found in patients treated with metformin compared with vitamin E, and in those treated with high-dose (3 g) carnitine vs diet. In contrast, there was no difference in patients treated with pioglitazone combined with vitamin E versus vitamin E alone, ursodeoxycholic acid (UDCA) combined with vitamin E or alone versus placebo, or UDCA versus combination of vitamin E and vitamin C, and in patients treated with vitamin E, probucol, N-acetylcysteine, low doses of carnitine, or Yo Jyo Shi Ko compared with placebo. Aspartate aminotransferase normalization was significantly higher in those treated with UDCA combined with vitamin E versus UDCA alone or placebo, and in those treated with metformin. Small number of subjects, high drop-out rates, and numerous interventions in 1 study limit the value of many studies. Only 7 RCTs analyzed biopsy specimens, but most of them have significant methodological limitations. Pioglitazone had reduced liver necrosis and inflammation in 1 large study.

Conclusions: Limited data do not allow one to draw firm conclusions on the efficacy of various treatments for NAFLD.

*Department of Gastroenterology, Hepatology and Immunology, Children's Memorial Health Institute, Poland

2nd Department of Paediatrics, Medical University of Warsaw, Warsaw, Poland

Department of Pediatrics, University of Naples “Federico II,” Naples, Italy

§Paediatric Liver Centre, Institute of Liver Studies, Variety Club Children's Hospital, King's College School of Medicine at King's College Hospital, London, UK

Received 10 April, 2008

Accepted 15 September, 2008

Address correspondence and reprint requests to Dr Piotr Socha, Department of Gastroenterology, Hepatology and Immunology, Children's Memorial Health Institute, Al. Dzieci Polskich 20, 04-730 Warsaw, Poland (e-mail:

The authors report no conflicts of interest.

Nonalcoholic fatty liver disease (NAFLD) and/or steatohepatitis (NASH) are increasingly being recognized in children. These conditions usually come to attention when investigating abdominal pain; alternatively, elevated liver function tests get reported from an incidental biochemical profile, and a subsequent ultrasound scan of the liver suggests fatty changes. Recognition of the liver disease may also follow investigational diagnostic protocols of obese individuals. In adults, fatty liver with abnormal liver enzymes is commonly seen with alcohol abuse, hence, the terminology. However, in children, this terminology needs revising because alcohol is less of a problem; moreover it is important to exclude inherited disorders of metabolism.

Nonalcoholic fatty liver disease is a spectrum of liver pathology that encompasses liver histology ranging from bland steatosis to steatosis with hepatitis (nonalcoholic steatohepatitis) and cirrhosis. In children NAFLD was first reported in the early 1980s (1). Since then, a number of case series have been described (2–6), with the following clinical characteristics: occurrence in young children, a male predominance; serum alanine transaminase (ALT) more elevated than serum aspartate aminotransferase (AST), hypertriglyceridemia, and nonspecific symptoms; vague abdominal pain or obesity is often the reason for clinical assessment. Cirrhosis complicating childhood NALFD (3–5) has been encountered and has been seen in patients as young as 9 years (3). Most of the series of children with NAFLD involve some degree of selection bias, either a referral bias or the child having required a liver biopsy because of persistently elevated serum aminotransferases, hence, the precise prevalence remains unknown. Small epidemiological studies using ultrasound or liver enzymes in obese children as diagnostic tests have shown the prevalence to range from 2.6% to 25% (6–9). An autopsy-based series in the United States (10) reports that, for children and adolescents ages 2 to 19 years, the prevalence of fatty liver adjusted for age, sex, race, and ethnicity is estimated to be 9.6%. The highest rate of fatty liver was seen in obese children (38%).

Loss of weight and physical activity are the mainstay treatment, but are difficult to achieve. Therefore, substitute therapies on the basis of pathogenetic mechanisms have been attempted. Treatments of glucose intolerance and hyperlipidemia have been shown to be helpful in improving steatosis and steatohepatitis in adults (11,12). Several pharmacologic agents such as lipid-lowering agents (clofibrate, gemfibrozil) (13,14), ursodeoxycholic acid (UDCA) (15,16), antioxidants (vitamin E, N-acetylcysteine) (17–19), and drugs that improve insulin sensitivity (20,21) have been shown to be promising in the treatment of NASH. However, most of these are open-label, uncontrolled studies. Uncertainty exists regarding the optimal treatment of NAFLD/NASH especially in childhood in which the number of randomized controlled trials (RCT) is limited. The purpose of this review was to evaluate current evidence of the efficacy of pharmacological agents used thus far in the management of both adult and pediatric NAFLD/NASH on gold standard criterion, which is liver histology, and surrogate marker, which is aminotransferase serum activity.

Back to Top | Article Outline


We aimed to identify all RCTs assessing the efficacy of pharmacological and dietary supplement interventions in patients with NAFLD or NASH. The electronic search strategy applied standard filters for identification of RCTs (22). The databases searched were Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library Issue 4, 2006), MEDLINE, and EMBASE (both up to December 2006). Only studies published in the English language and written as full peer-reviewed publications were included. In addition to the electronic search, we checked out cross-references from original articles and reviews. Our topic-specific strategy search included the following:

  • No. 1 (nonalcoholic[All Fields]) and (“fatty liver”[MeSH Terms] or fatty liver[Text Word]) and (“disease”[MeSH Terms] or disease[Text Word]) or (steatohepatitis[All Fields]) or (steatosis[All Fields]) and (“liver”[MeSH Terms] or liver[Text Word]) or (NAFLD[All Fields])
  • No. 2 (“therapy”[Subheading] or (“therapeutics”[TIAB] NOT Medline[SB]) or “therapeutics”[MeSH Terms] or treatment[Text Word]) or (Treatment*[Multi]) or (“vitamin E”[MeSH Terms] or vitamin E[Text Word] or (“tocopherols”[TIAB] NOT Medline[SB]) or “tocopherols”[MeSH Terms] or tocopherol[Text Word]) or (“taurine”[MeSH Terms] or taurine[Text Word]) or (“ursodeoxycholic acid”[TIAB] not Medline[SB]) or “ursodeoxycholic acid”[MeSH Terms] or UDCA[Text Word]) or (“clofibrate”[MeSH Terms] or clofibrate[Text Word]) or (“pioglitazone”[Substance Name] or troglitazone[Text Word]) or (“metformin”[MeSH Terms] or metformin[Text Word]) or (“betaine”[MeSH Terms] or betaine[Text Word])
  • No. 1 and No. 2.
Back to Top | Article Outline

Included and Excluded Studies

Two reviewers (A.H. and P.D.) independently screened the titles and the abstracts identified according to the above-described strategy. All potentially relevant articles were retained, and the full text of these studies was examined to determine which trials satisfied the inclusion criteria. The same reviewers independently carried out data extraction, using standard data extraction forms. Discrepancies between the reviewers' findings were resolved by discussion. The primary outcome measures for the included trials were the normalization of serum ALT and/or AST levels as well as histological changes in a liver biopsy specimen. The secondary outcome measures were the incidence and severity of adverse events. Nonalcoholic fatty liver disease was defined by fatty liver disease related to obesity/overweight, which was diagnosed histologically or by bright liver on ultrasound combined with increased transaminase activity. Authors of the studies were contacted for additional information when applicable.

Back to Top | Article Outline

Study Quality

Two reviewers (A.H. and P.D.) independently assessed the methodological quality of the included trials. Use of the following strategies associated with good-quality studies was assessed using Cochrane definitions: reporting of allocation concealment and description of an adequate randomization method (yes/no/unclear); blinding (yes/no/not reported); intention-to-treat analysis (yes/no); and comprehensive follow-up (yes/no).

Adverse effects and compliance to treatment were evaluated as well.

Back to Top | Article Outline

Statistical Methods

The data were analyzed using Review Manager 4.2.7 (version date 27 May 2004; The Cochrane Collaboration). Data were analyzed for every participant for whom the outcome was obtained (ie, available case analysis). The binary measure for individual studies and pooled statistics is reported as the risk ratio (RR) between the experimental and control groups with 95% confidence intervals (95% CI). We calculated the number needed to treat (NNT) as the inverse of the pooled absolute risk differences and 95% CI. We examined heterogeneity using χ 2 and I 2, which can be interpreted as the percentage of the total variation between studies that is attributable to heterogeneity rather than to chance. A value of 0% indicates no observed heterogeneity and larger values show increasing heterogeneity. In studies that gave results only in figures, but not in numbers, we contacted the corresponding author for clarification; however, unfortunately this was without success. Because of a limited amount of data, we did not test for publication bias.

Back to Top | Article Outline


Trial Flow

Our initial electronic search identified 22 RCTs, 7 studies were excluded, and 15 trials were eligible for evaluation (Table 1) (23–37). One trial was excluded because multifactorial treatments were allowed in the treatment arm (38). Six trials were excluded because the defined outcome measures were not assessed (39–45). The 2 reviewers agreed on the selection of the included studies.



Back to Top | Article Outline

Study Characteristics and Methodological Quality

The 15 selected studies recruited a total of 801 participants. Except for 2 RCTs (30,36), all were performed in adults.

All patients allocated to active and no treatment/placebo group were also put on caloric-restriction diet. Two RCTs involving 86 patients compared metformin with restricted diet (25) or no treatment (34) and 1 RCT (n = 57) compared metformin with vitamin E (25). Four RCTs (n = 216) compared vitamin E with placebo or no treatment (25,28,30,36). One RCT (n = 18) compared pioglitazone combined with vitamin E (33) and another (n = 47) compared pioglitazone versus placebo (24). One 3-arm RCT (n = 40) compared ursodeoxycholic acid (UDCA) with ursodeoxycholic acid combined with vitamin E and placebo (27). One RCT (n = 126) compared UDCA with placebo (31) and another RCT compared UDCA with vitamin E and vitamin C (n = 56) (23). One trial (n = 27) compared probucol with placebo (29). One RCT (n = 8) compared Yo Jyo Hen Shi Ko (YHK) with placebo (26). One trial (n = 30) compared N-acetylcysteine with placebo (32). One RCT involving 101 patients compared carnitine (at 3 different doses: 1 g, 2 g, and 3 g) with placebo (35). Finally, 1 RCT (n = 44) compared orlistat with placebo (37). Not only were there variations in the interventions but also in the duration of the interventions, which lasted from 2 to 24 months.

The methodological quality of the trials also varied (Table 1). Allocation concealment was adequate in 9 trials and unclear in 6 of the trials. Only 10 were double-blind studies, but often it was not stated who was blinded. The remaining trials were open trials. The completeness of follow-up was adequate in 13 trials. Intention-to-treat analysis was performed in 10 trials.

Back to Top | Article Outline

Study Outcomes

Normalization of ALT

Eleven of 15 RCTs provided data on the normalization of the ALT level at the end of the treatment (23,25–27,29,30,32–36). Three of these studies were 3-arm trials (25,27,35), so all together 17 comparisons between the intervention and control groups were available for analysis (Fig. 1).

FIG. 1

FIG. 1

A meta-analysis of 3 RCTs (170 participants) (25,30,36) showed no difference in the percentage of patients with a normal ALT level in those treated with vitamin E compared with placebo (RR 1.05, 95% CI 0.79–1.39). The included studies were homogeneous (χ 2 = 2.75, and I 2 = 27.2).

The pooled results of 2 RCTs (n = 86) (25,34) revealed a significant, although modest, effect on the normalization of serum ALT levels in patients with NAFLD treated with metformin compared with placebo (RR 2.0, 95% CI 1.23–3.24). On the basis of the results of 1 RCT (n = 57), metformin was also significantly more potent than vitamin E (RR 3.14, 95% CI 1.16–8.47) (25).

On the basis of the results of a single trial (n = 51), there was a significant normalization of ALT levels in patients treated with high-dose (3 g/day) carnitine (RR 5.75, 95% CI 1.45–22.74) (35).

In contrast, there was no significant effect on the normalization of ALT levels found in those treated with pioglitazone combined with vitamin E compared with vitamin E alone (33). There was also no significant effect on the normalization of ALT levels found in patients treated with UDCA combined either with vitamin E or alone versus placebo or UDCA versus combination of vitamin E and vitamin C (23,27). We also did not reveal any significant difference in the rate of ALT normalization in patients treated with probucol (29), N-acetylcysteine (32), low doses (1 g and 2 g) of carnitine versus placebo (35), or YHK (26) compared with placebo.

Back to Top | Article Outline

Normalization of AST

Four of 15 RCTs (n = 128), including one 3-arm study, provided data on the effects of pharmacological interventions on the normalization of AST levels (Fig. 2) (27,29,32,34). The rate of AST normalization was significantly higher in those treated with UDCA given simultaneously with vitamin E versus UDCA (RR 1.72, 95% CI 1.04–2.84) or placebo (RR 2.98, 95% CI 1.3–6.85) (27), as well as in those treated with metformin compared with no treatment (RR 2.81, 95% CI 1.16–6.82) (34). Compared with placebo, no effect was found in patients treated with UDCA (27), probucol (29), or N-acetylcysteine (32).

FIG. 2

FIG. 2

Back to Top | Article Outline

Liver Biopsy

Nine RCTs involving 564 participants provided data on liver biopsy histology (24,25,27,28,31,33–35,37). Only 7 RCTs that analyzed pretreatment and posttreatment biopsy specimens between the groups were taken into consideration (24,28,31,33–35,37). Two studies (25,27) compared pretreatment and posttreatment biopsy specimens within the study groups only (not between the groups); these studies were excluded from this review.

The histopathology scoring system for the evaluation of liver biopsy specimens varied between the studies; nevertheless, all of the studies evaluated stages of steatosis, necroinflammation, and fibrosis. One study (24) used Kleiner's criteria (46) and the remaining 6 RCTs (28,31,33–35,37) used Brunt modified criteria (47). In the study by Belfort et al (24), 47 posttreatment biopsy specimens from 55 patients who entered the study were available. In patients treated with pioglitazone compared with placebo, there was a reduced risk of necroinflammation (RR 0.45, 95% CI 0.24–0.74, NNT 3, 95% CI 2–6), ballooning necrosis (RR 0.44, 95% CI 0.2–0.96, NNT 4, 95% CI 2–65), and inflammation (RR 0.44, 95% CI 0.2–0.85, NNT 3, 95% CI 2–13); however, there was not a reduced risk of steatosis (RR 0.58, 95% CI 0.3–1.02) or fibrosis (RR 0.72, 95% CI 0.3–1.5).

One RCT assessed the effect of combined treatment of pioglitazone with vitamin E compared with vitamin E (33). Data from 18 patients who entered the study were available for analysis. Combination therapy was significantly superior to vitamin E with respect to cytologic ballooning, Mallory hyaline, and inflammation, but not with respect to steatosis grade, pericellular fibrosis, or portal fibrosis. However, a comparison made for the percentage change from baseline until the end of the study showed only a significant difference in terms of a change in the degree of steatosis.

One RCT assessed the effectiveness of UDCA on the histological changes in patients with NASH (31). Out of 126 patients who entered the trial, pre- and posttreatment histological material was available from 107 patients. There was no difference in the degree of steatosis, inflammation, ballooning of the hepatocytes, Mallory hyaline, or fibrosis in those treated with UDCA compared with placebo.

In another RCT consisting of 45 enrolled patients, everybody underwent liver biopsy at the end of the treatment (28). Vitamin E and C treatment resulted in no statistically significant improvement in the fibrosis score compared with placebo (at least 1 point of improvement was found in 11 of 23 patients from the vitamins group and 9 of 22 patients from the placebo group).

In a trial assessing L-carnitine therapy for the treatment of NAFLD, a liver biopsy was performed in 73 of 101 patients who entered the study, as well as in 50 patients at the end of the 6 months of treatment (35). No significant change was observed in the grade of hepatic steatosis of the liver at the end of the treatment in any of the groups.

In the RCT comparing metformin with placebo, pre- and posttreatment biopsy data were available from 23 of 33 patients who entered the study (34). The frequency of improvement in the necroinflammatory activity index appeared greater in the metformin group than in the placebo group (46% vs 10%), although the difference was not significant.

In the RCT assessing the effect of orlistat on histological improvement, 40 of 44 randomized patients had liver biopsy and 22 patients underwent a repeat biopsy at the end of the study (11 in each group) (37). There were comparable improvements in the degrees of fibrosis (at least 1 degree of improvement in 2 of 11 patients in the orlistat group and 4 of 11 patients in the placebo group) and steatosis (at least 1 degree of improvement in 5 of 11 patients in the orlistat group and 3 of 11 patients in the placebo group) between the groups.

Back to Top | Article Outline

Adverse Effects

Eleven RCTs provided data regarding adverse effects in their patients. However, the adverse effects were not significantly more frequent in patients in the active groups than in controls. Except for 1 study reporting pioglitazone hepatotoxicity in 1 patient (33), the adverse events were mild. Only 1 RCT revealed a high-withdrawal rate in the UDCA group compared with the placebo group; however, the difference was not statistically significant (20% vs 14%, RR 0.79, 95% CI 0.35–1.17) (31).

Back to Top | Article Outline


Our systematic review of published randomized, double-blind trials on pharmacological and dietary supplement interventions for NAFLD/NASH showed that only a limited number of studies are available; unfortunately, this precludes drawing firm conclusions regarding the efficacy of various treatments. Because most RCTs have been performed in adults, one should be cautious when trying to transfer these data to the treatment of pediatric patients. Still, the pathomechanisms and risk factors of NAFLD both in adults and in children seem to be similar and pediatricians in many clinical situations take lessons from doctors treating adult patients. Adult studies indicate the most promising treatments to be tested in pediatric settings. Pioglitazone, carnitine, metformin, and tocopherol combined with UDCA seem to be promising therapies; tocopherol does not show any effect. All patients were also put on caloric-restriction diet because dietary counseling seems to be the basic therapeutic approach.

There are several limitations to this review that we acknowledge. Although it is well known that noninvasive imaging studies may be insensitive to degrees of steatosis <25% to 30% and that significant liver disease may exist with liver enzymes in the normal range among patients with NAFLD (48), the diagnostic criteria of NAFLD in many studies used surrogate markers such as ultrasonographic bright liver and increased transaminases.

Liver histology is still the gold standard for both diagnosis and follow-up of NAFLD/NASH (49), but firm recommendations of when to perform a liver biopsy in the routine clinical setting have not yet been developed (48). Especially after a short-term therapeutic trial a repeat biopsy may be considered too invasive so that surrogate markers are often necessary to evaluate the effectiveness of the treatment. Steatosis alone is not considered clinically important because it is in the majority of the cases fully reversible (48,49). Because increased transaminase activity is an indicator of hepatic necroinflammation, and necroinflammatory processes lead to fibrosis and cirrhosis, transaminase normalization was regarded to be an important outcome. Irrespective of the treatment, we looked for the therapy that would be able to stop progression of liver disease. Alanine transaminase seems to be more specific for hepatic necroinflammation than AST, and this outcome was used in most of the studies (23,25–27,29,30,32–36). We did not analyze the dietary restrictions in detail because in almost all studies they were not described.

Usually a note was given that a dietary therapy was also performed. Although this may not have a role to play in a large RCT, with the small size of many of the RCTs included, factors such as these may have influenced the outcomes or modified the placebo response sufficiently to decrease the potential for the intervention to show a difference.

The sample sizes in some trials, as well as the number of trials for some comparisons, were small. The pooled sample sizes were also small, thus, there was little statistical power; consequently, we cannot exclude chance as an explanation for the results of many comparisons. Also, true compliance with drug treatments and unpredicted loss of weight and/or change of lifestyle during the trial frame were not always taken into account. Therefore, marked variability among study populations and of the interventions tested may have decreased the sensitivity for detecting possible effects. The methodological quality and the quality of reporting results were varied and sometimes poor. Potential limitations include unclear or inadequate allocation concealment, no intention-to-treat analysis, and no blinding. The findings are, therefore, likely to be affected to a varying degree by selection, attrition, and/or performance biases. In our meta-analysis, the statistical tests of homogeneity (total consistency) of the results were nonsignificant. However, it is important to stress that the power of the statistical methods that investigate heterogeneity is limited, particularly for meta-analyses on the basis of a small number of studies, as in the case of this review. Consequently, the results from our meta-analyses should be viewed with caution. Similarly, given the small number of studies, it is difficult to make firm conclusions.

Several groups of drugs were used in clinical trials according to the pathomechanisms of liver injury in NAFLD—antioxidants, carnitine, and insulin sensitizers (Table 2). Only the use of UDCA does not correspond directly to the mechanism of liver damage related to fatty infiltrations.



Vitamin E therapy seemed to be the most promising, and we were able to identify 3 RCTs that addressed this treatment. However, a meta-analysis performed on these trials did not show any effect with respect to ALT normalization. Authors of these trials also reported the poor effect of vitamin E therapy on ALT normalization (28) and emphasized the major role of diet and physical exercise but not vitamin E (30). However, it has been suggested that poor compliance with vitamin E treatment (as indicated by low vitamin E blood levels) and/or loss of weight in controls should be taken into account when evaluating these studies (36,50). In 1 study, the combination of vitamin E and UDCA was proven to be superior to placebo or UDCA alone as expressed by AST normalization (27). Differences in vitamin E doses among the studies limit the value of meta-analysis performed (400–1000 mg/day).

Another strong antioxidant, probucol, which is also a cholesterol-lowering drug, did not show any effect on ALT and AST normalization (29). Because depletion of glutathione, which is a strong antioxidant, has been observed in many liver diseases, N-acetylcysteine treatment could be regarded as beneficial. Still, the results of 1 identified trial did not show any effect of N-acetylcysteine treatment on ALT and AST normalization (32). One small study with YHK, an herbal medicine, does not allow one to draw any conclusions about its efficacy in treating NAFLD because of the extremely small number of patients included (26).

The idea to use carnitine for NAFLD therapy is based on the pathophysiology of fat accumulation in the liver and the possible role of this drug in improving fatty acid beta-oxidation. Only high-dose beta-carnitine (3 g) seems to be promising, but this conclusion is based on only 1 available trial (35).

Vitamin E, N-acetylcysteine, carnitine, probucol, and YHK are antioxidants from which only high-dose carnitine showed a possible beneficial effect.

Because NAFLD seems to be related to the metabolic syndrome, antidiabetic drugs were administered in some studies. Metformin is a well-known insulin sensitizer whose safety and efficacy for the treatment of diabetes is well established. The results of our systematic review on the basis of 2 trials suggest a beneficial role of metformin with regard to ALT normalization (25,34). The same conclusion was reached when AST normalization was used as the outcome measure in 1 study (34). It is difficult to indicate the dose of metformin, because different doses were used. Another insulin sensitizer—pioglitazone—did not show any effect on ALT normalization in 1 study (33).

Although UDCA therapy is indicated mainly for the treatment of cholestatic disease, improvement of bile flow is thought to have beneficial effects on liver function. Moreover, the safety of UDCA has been documented in many studies, encouraging the use of this drug for NAFLD. However, the results of 1 RCT trial identified did not show any effect of UDCA therapy on ALT and AST normalization. A pilot study suggests that this may be true also in the pediatric age group (51).

Most of the studies reported the effect of treatment on ALT activity, but some studies also included AST activity as an outcome measure. From 4 studies reporting AST, in all but 1 the effect on ALT and AST activity was similiar. In 1 study, the rate of AST but not ALT normalization was significantly higher in those treated with UDCA given simultaneously with vitamin E versus UDCA (27). Because increased ALT activity is used as an indicator of NAFLD also in epidemiological studies and it is a sensitive marker of liver cell damage, for conflicting results ALT should be regarded as a primary outcome measure.

The assessment of liver histology is regarded as a better endpoint than the assessment of liver enzymes; still, the studies available have many limitations. The number of biopsy specimens analyzed at the end of the study was usually significantly smaller than the number of patients at study entry (below 80%), and the criteria used among the studies varied. Still, some of the results should be considered interesting. Pioglitazone, which did not show any effect on transaminase normalization, reduced necrosis and inflammation with no effect on steatosis in 1 large study (24); similar effects were observed in another small study (33). A high dropout rate of patients providing liver biopsy specimens at the end of the metformin study makes conclusions regarding the effects of metformin on liver histology unreliable (34). Ursodeoxycholic acid and carnitine therapy were not effective, but the dropout rates were also high in these studies. Also, vitamin E and C therapy did not influence liver histology. Drawing conclusions about liver histology from the orlistat trial is difficult because of the small number of biopsies performed and a high dropout rate.

Making the right choice of pharmacotherapy for managing patients with NAFLD is one of the key issues in clinical hepatology. It is the reason why this question was also raised by the Cochrane Collaboration Group. Four systematic reviews were performed, each one concerning a different treatment group (52–55). The publications appeared soon after our systematic review was completed, with a slightly different search strategy and earlier date of search completion. One systematic review concerned antioxidant supplements (52). A positive significant effect on AST activity alone was shown with no effect on ALT activity. Histological data were too limited to allow one to draw any conclusions. The authors' conclusions confirm our findings. Another systematic review by the Cochrane Collaboration Group includes drugs that improve insulin resistance (53). Only 3 RCTs were included, which also appeared in our review. Additionally, we included 1 more recent study (24). The results are in line with our findings: normalization of ALT levels was found in 2 studies and 1 pioglitazone study showed histological improvement. On the basis of these studies and our review, metformin appears to be the first-choice therapy and is considered a safe treatment. Ursodeoxycholic acid therapy was the subject of another Cochrane review (54) in which 4 RCTs were identified, one of which was included in our review. The poor quality of the studies was emphasized, and no effect on transaminase activity was shown. Again, the results of this review confirm our findings regarding UDCA therapy. We did not find any studies in which the efficacy of probiotic therapy for NAFLD/NASH was investigated. This was searched also by the Cochrane Collaboration Group, who identified no RCTs related to this subject (55).

Our systematic review is adding to the Cochrane reviews because it focuses on different outcome measures, which are of practical importance to pediatricians. For example, in the Cochrane review, the biochemical outcome measures were not as strictly defined as in our systematic review. Although the Cochrane Collaboration Group accepted any biochemical presentation of the results, we focused on the “normalization” of transaminases. In our opinion, the changes in transaminase activity seem not to be discriminative enough to prove a significant clinical effect of treatment. In contrast to the Cochrane Group, we regard radiological findings as not reliable enough to include them as outcome measures. These are examples of the differences in our approach that resulted in a slightly different study selection, even if the general conclusions seem to be the same.

To give any advice to clinicians regarding the present state of knowledge and to identify the most promising therapies for further research, it first seems reasonable to review findings related to the pathophysiology of NAFLD. The “two-hit” theory tries to explain the progression of the disease to nonalcoholic steatohepatitis and cirrhosis (56). Insulin resistance seems to play a central role in the liver's accumulation of triglycerides and initiation of the inflammatory cascade. That is why insulin sensitizers can be regarded as beneficial at this stage of liver injury. The second hit may be oxidative stress or increased abnormal cytokine production. That is why antioxidants may ameliorate liver damage due to free radicals. This short review of the liver damage in NAFLD shows the complexity of the pathophysiology involved. One can also expect positive effects of different drugs at different stages of liver injury. Still, insulin sensitizers and antioxidants can be regarded as worthy of being tested. A large collaborative study on vitamin E versus metformin versus placebo in 180 children (TONIC—treatment of nonalcoholic fatty liver disease in children) is close to being concluded and probably will add further important information on treatment of patients with NASH/NAFLD (57). The present review indicates limitations to formulate therapeutic guidelines, pointing to small number of subjects in the studies and numerous interventions, making comparisons difficult.

Back to Top | Article Outline


1. Moran JR, Ghishan FK, Halter SA, et al. Steatohepatitis in obese children: a cause of chronic liver dysfunction. Am J Gastroenterol 1983; 78:374–377.
2. Baldrigde AD, Perez-Atayde AR, Graeme-Cook F, et al. Idiopathic steatohepatitis in childhood: a multi-centre retrospective study. J Pediatr 1995; 127:700–704.
3. Rashid M, Roberts EA. Nonalcoholic steatohepatitis in children. J Pediatr Gastroenterol Nutr 2000; 30:48–53.
4. Molleston JP, White F, Teckmann J, et al. Obese children with steatohepatitis can develop cirrhosis in childhood. Am J Gastroenterol 2002; 97:2460–2462.
5. Kinugasa A, Tsunamoto K, Furukawa N, et al. Fatty liver and its fibrous changes found in simple obesity of children. J Pediatr Gastroenterol Nutr 1984; 3:408–414.
6. Tominaga K, Kurat JH, Chen YK, et al. Prevalence of fatty liver in Japanese children and relationship to obesity. An epidemiological ultrasonographic survey. Dig Dis Sci 1995; 40:2002–2009.
7. Vajro P, Fontanella A, Perna C, et al. Persistent hypertransaminasemia resolving after weight reduction in obese children. J Pediatr 1994; 125:239–241.
8. Tazawa Y, Noguchi H, Nishiomiya F, et al. Serum alanine aminotransferase activity in obese children. Acta Paediatr 1997; 86:238–241.
9. Bergomi A, Lughetti L, Corciulo N, et al. Italian multicentre study on liver damage in pediatric obesity [abstract]. Int J Obes Rel Metab Dis 1998; 22(Suppl 4):S22.
10. Schwimmer JB, Deutsch R, Kahen T, et al. Prevalence of fatty liver in children and adolescents. Pediatrics 2006; 118:1388–1393.
11. Palmer M, Schaffner F. Effect of weight reduction on hepatic abnormalities in overweight patients. Gastroenterology 1990; 99:1408–1413.
12. Anderson T, Gluud C, Franzmann MB, et al. Hepatic effects of dietary weight loss in morbidly obese patients. J Hepatol 1991; 12:224–229.
13. Laurin J, Lindor KD, Crippin JS, et al. Ursodeoxycholic acid or clofibrate in the treatment of non-alcohol-induced steatohepatitis: a pilot study. Hepatology 1996; 23:1464–1467.
14. Basaranogly M, Acbay O, Sonsuz A. A controlled trial of gemfibrozil in the treatment of patients with nonalcoholic steatohepatitis. J Hepatol 1999; 31:384.
15. Guma G, Viola L, Thome M, et al. Ursodeoxycholic acid in the treatment of nonalcoholic steatohepatitis: results of a prospective clinical controlled trial [abstract]. Hepatology 1997; 26:387A.
16. Gulbahar O, Karasu ZA, Ersoz G, et al. Treatment of nonalcoholic steatohepatitis with N-acetylcysteine [abstract]. Gastroenterology 2000; 118:A1444.
17. Lavine JE. Vitamin E treatment of non-alcoholic steatohepatitis in children: a pilot study. J Pediatr 2000; 136:734–738.
18. Hasegawa T, Yoneda M, Nakamura K, et al. Plasma transforming growth factor B-1 level and efficacy of alpha-tocopherol in patients with nonalcoholic steatohepatitis: a pilot study. Aliment Pharmacol Ther 2001; 15:1667–1672.
19. Caldwell SH, Hespenheiden EE, Redick JA, et al. A pilot study of thiazolidinedione, troglitazone, in nonalcoholic steatohepatitis. Am J Gastroenterol 2001; 96:519–525.
20. Marchesini G, Brizi M, Bianchi G, et al. Metformin in non-alcoholic steatohepatitis. Lancet 2001; 358:893–894.
21. Caldwell SH, Hespenheiden EE, Redick JA, et al. A pilot study of thiazolidinedione, troglitazone, in nonalcoholic steatohepatitis. Am J Gastroenterol 2001; 96:519–525.
22. Robinson KA, Dickersin K. Development of a highly sensitive search strategy for the retreival of reports of controlled trials using PubMed. Int J Epidemiol 2002; 31:150–153.
23. Ersoz G, Gunsar F, Karasu Z, et al. Management of fatty liver disease with vitamin E and C compared to urodeoxycholic acid treatment. Turk J Gastroenterol 2005; 16:124–128.
24. Belfort R, Harrison S, Brown K, et al. A placebo controlled trial of pioglitazone in subjects with Nonalcoholic Steatohepatitis. N Engl J Med 2006; 355:2297–2307.
25. Bugianesi E, Gentilcore E, Manini R, et al. A randomized controlled trial of metformin versus vit E or prescriptive diet in nonalcoholic fatty liver disease. Am J Gastroenterol 2005; 100:1082–1090.
26. Chande N, Laidlaw P, Marrota P. Yo Jyo Hen Shi Ko (YHK) improves transaminases in non alcoholic steatohepatitis (NASH) a randomized pilot study. Dig Dis Sci 2006; 5:1183–1189.
27. Dufour JF, Oneta C, Gonvers JJ, et al. Randomized placebo-controlled trial of ursodeoxycholic acid with vit E in nonalcoholic steatohepatitis. Clin Gastroenterol Hepatol 2006; 4:1537–1543.
28. Harrison S, Torgerson S, Hayashi P, et al. Vitamin E and vitamin C treatment improve fibrosis in patients with nonalcoholic steatohepatitis. Am J Gastroenterol 2003; 98:2485–2490.
29. Merat S, Malekzadeh R, Sohrabi M, et al. Probucol in the treatment of non-alcoholic steatohepatitis: a double-blind randomized controlled study. J Hepatol 2003; 38:414–418.
30. Nobili V, Manco M, Devito R, et al. Effect of vitamin E on aminotransferase levels and insulin resistance in children with non-alcoholic fatty liver disease. Aliment Pharmacol Ther 2006; 24:1553–1561.
31. Lindor K, Kowdley K, Heathcote E, et al. Ursodeoxycholic acid for treatment of nonalcoholic steatohepatitis: results of a randomized trial. Hepatology 2004; 39:770–778.
32. Pamuk G, Sonsuz A. N-Acetylcysteine in the treatment of non-alcoholic steatohepatitis. J Gastroenterol Hepatol 2003; 18:1220–1221.
33. Sanyal A, Mofrad P, Contos M, et al. A pilot study of vitamin E versus vitamin E and pioglitazone for the treatment of nonalcoholic steatohepatitis. Clin Gastroenterol Hepatol 2004; 2:1107–1115.
34. Uygun A, Kadayifci A, Isik A, et al. Metformin in the treatment of patients with non-alcoholic steatohepatitis. Aliment Pharmacol Ther 2004; 19:537–544.
35. Uygun A, Kadayifci A, Bagci S, et al. L-Carnitine therapy in non-alcoholic steatohepatitis. Turk J Gastroenterol 2000; 11:196–201.
36. Vajro P, Mandato C, Franzese A, et al. Vitamin E in pediatric obesity-related liver disease: a randomized study. J Pediatr Gastroenterol Nutr 2004; 38:48–55.
37. Zelber-Sagi S, Kessler A, Brazowsky E, et al. A double-blind randomized placebo-controlled trial of orlistat for the treatment of nonalcoholic fatty liver disease. Clin Gastroenterol Hepatol 2006; 4:639–644.
38. Athyros V, Mikhailidis D, Didangelos T, et al. Effect of multifactorial treatment on non-alcoholic fatty liver disease in metabolic syndrome: a randomized study. Curr Med Res Opin 2006; 22:873–883.
39. Basanaroglu M, Acbay O, Sonsuz A. A controlled trial of gemfibrozil in the treatment of patients with nonalcoholic steatohepatitis. J Hepatol 1999; 31:384.
40. Daubioul C, Horsmans Y, Lambert P, et al. Effect of oligofructose on glucose and lipid metabolism in patients with nonalcoholic steatohepatitis: results of a pilot study. Eur J Clin Nutr 2005; 59:723–726.
41. Daubioul C, Horsmans Y, Lambert P, et al. Effect of oligofructose on glucose and lipid metabolism in patients with nonalcoholic steatohepatitis: results of a pilot study. Eur J Clin Nutr 2005; 59:723–726.
42. Kugelmas M, Hill D, Vivian B, et al. Cytokine and NASH: a pilot study of the effects of lifestyle modification and vitamin E. Hepatology 2003; 38:413–419.
43. Mendez-Sanchez N, Gonzalez V, Chavez-Tapia N, et al. Weight reduction and ursodeoxycholic acid in subjects with nonalcoholic fatty liver disease. A double-blind, placebo-controlled trial. Ann Hepatol 2004; 3:108–112.
44. Miglio F, Rovati L, Santoro A, et al. Efficacy and safety of oral betaine glucuronate in non-alcoholic steatohepatitis. A double blind, randomized, parallel-group, placebo-controlled prospective clinical study. Arzneim-Forsch 2000; 50:722–727.
45. Morita Y, Ueno T, Sasaki N, et al. Nateglinide is useful for nonalcoholic steatohepatitis (NASH) patients with type 2 diabetes. Hepatogastroenterology 2005; 52:1338–1343.
46. Kleiner DE, Brunt EM, van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005; 41:1313–1321.
47. Brunt EM, Janney G, Di Bisceglie AM, et al. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol 1999; 94:2467–2474.
48. Neuschwander-Tetri BA, Caldwell SH. Nonalcoholic steatohepatitis: summary of an AASLD Single Topic Conference. Hepatology 2003; 37:1202–1219.
49. Brunt EM. Pathology of nonalcoholic steatohepatitis. Hepatol Res 2005; 33:68–71.
50. Vajro P, Mandato C, Franzese A, et al. More on vitamin E therapy. Hepatology 2004; 39:569.
51. Vajro P, Franzese A, Valerio G, et al. Lack of efficacy of ursodeoxycholic acid for the treatment of liver abnormalities in obese children. J Pediatr 2000; 136:739–743.
52. Lirussi F, Azzalini L, Orando S, et al. Antioxidant supplements for non-alcoholic fatty liver disease and/or steatohepatitis. Cochrane Database Syst Rev. 2007:CD004996.
53. Angelico F, Burattin M, Alessandri C, et al. Drugs improving insulin resistance for non-alcoholic fatty liver disease and/or non-alcoholic steatohepatitis. Cochrane Database Syst Rev. 2007:CD005166.
54. Orlando R, Azzalini L, Orando S, et al. Bile acids for non-alcoholic fatty liver disease and/or steatohepatitis. Cochrane Database Syst Rev. 2007:CD005160.
55. Lirussi F, Mastropasqua E, Orando S, et al. Probiotics for non-alcoholic fatty liver disease and/or steatohepatitis. Cochrane Database Syst Rev. 2007:CD005165.
56. Day C, Saksena S. Non-alcoholic steatohepatitis. Definitions and pathogenesis. J Gastroenterol Hepatol 2002; 17(Suppl 3):S377–S384.
57. Lavine JE, Schwimmer JB. Clinical Research Network launches TONIC trial for treatment of non-alcoholic fatty liver disease in children. J Pediatr Gastroenterol Nutr 2006; 42:129–130.

Liver disease; NAFLD; NASH

© 2009 Lippincott Williams & Wilkins, Inc.