Nonalcoholic fatty liver disease (NAFLD) is a worldwide medical problem that parallels the obesity epidemic (1) and has become the most common liver disorder at all ages (2,3). Loss of weight and regular physical activity are the mainstay of treatment, but these are difficult to achieve (4–7). Although histological liver abnormalities are not generally a major concern in children, treatment is essential to avoid progression to more severe forms during or after transition into adult age (7,8). However, no proven effective pharmacological therapies are yet available (9,10).
Several studies have suggested that small intestinal bacterial overgrowth (SIBO) may contribute to the pathogenesis of NAFLD (11–13). Manipulation of intestinal flora with probiotics in experimental NAFLD of rodent animal models did indeed have a beneficial effect (14,15). In humans, a preliminary study investigating the outcome of probiotic VSL#3 (a mixture of 4 lactobacilli strains) therapy on liver dysfunction parameters in different categories of adult chronic liver disease including NAFLD also yielded promising results (16); however, the opposite results were reported in a recent article (17). The aim of our pilot study was, therefore, to evaluate whether a short course of therapy with probiotics improves the outcome of liver function tests in children with obesity-related liver disease.
PATIENTS AND METHODS
Obese children followed up at our pediatric obesity and pediatric liver units, with a body mass index (BMI) > 95th percentile for age and sex, who failed to adhere to previous slimming diets and who had not undergone previous pharmacological treatment for obesity, were offered the possibility of entering the present study. Twenty patients with persistent (>3 months) liver abnormalities (alanine aminotransferases [ALT] higher than upper normal values [40 U/L] associated with ultrasonographic [US] liver brightness suggestive of fatty liver) were enrolled.
Exclusion criteria were the coexistence of causes of increased transaminase levels other than obesity (namely muscular disease, viral hepatitis B and C, autoimmune hepatitis, α1-antitrypsin deficiency, cystic fibrosis, Wilson disease, hemochromatosis, hereditary fructose intolerance, amino acid disorders, atypical celiac disease, alcohol abuse, and drug toxicity), which were investigated by appropriate biochemical tests or verified by anamnestic data. Patients receiving concomitant antibiotic treatment were also excluded.
At entry, weight, height, BMI, BMI z score, and visceral fat thickness measured by US (18) were investigated. Standard liver function tests and US brightness were evaluated and graded as described elsewhere (4,19). A quantitative ratio of liver versus renal brightness that referred to specific regions of interest was obtained according to Mancini et al (20). All of the patients underwent a standard oral glucose tolerance test. The glucose H2 breath test (H2BT) was performed after an overnight fast (1 g/kg, maximum 50 g/250 mL) (21). Patients were asked to avoid foods containing beans, bran, and high-fiber cereals on the day before the test. Breath samples were collected every 15 minutes for 2 hours. The test was considered positive for SIBO if the fasting breath H2 concentration was> 20 ppm and/or increased >10 ppm above baseline within the first 2 hours. Serum levels of tumor necrosis factor-alpha (TNF-α) were measured by quantitative sandwich enzyme immunoassay technique (R&D Systems, Minneapolis, MN). Serum antibodies to peptidoglycan-polysaccharide (PG-PS) complexes were determined by classical enzyme-linked immunosorbent assay according to the method of Hazenberg et al (22) and using commercially available PG-PS 100 P (Fluka Kemie GmbH, Buchs, Switzerland) (23). In the case of the hepatorenal US ratio, TNF-α, and PG-PS IgA, we compared the values in 20 age-, sex-, and weight-matched obese nonhepatopathic children.
The 20 children were randomly allocated to 1 of the 2 groups for a pilot randomized, double-blind (sealed envelopes), placebo-controlled study. The study drugs were oral Lactobacillus GG (12 billion CFU/day) and an indistinguishable placebo administered for 8 weeks. The primary outcome measure was improvement and/or normalization of ALT levels (<40 IU/L). Secondary endpoints were treatment-related changes in liver echogenicity, TNF-α values, and, possibly, intestinal tests (H2BT and PG-PS antibodies).
Changes in lifestyle were not encouraged during the brief framework of the study. Parents were instructed to report any adverse effect that occurred during treatment and to supervise the administration of pills. Parents gave written informed consent. The study was conducted in accordance with Good Clinical Practices and the Declaration of Helsinki, and was approved by the ethics committee of our university hospital (number 8404).
Continuous variables are reported as mean ± standard deviation. Two-tailed tests of significance are reported, and P < 0.05 were considered statistically significant. The unpaired Student t test was used to assess differences in the average variations (8 weeks–baseline) between the treatment and placebo groups, as well as between the baseline means of the 2 groups. The possible confounding effect of BMI z score variations on ALT concentrations was explored by a multiple regression model with ALT as dependent variable and with BMI z score variation and treatment as independent variables. The Fisher exact test was used to calculate an exact probability value for the relation between 2 dichotomous variables.
Demographic data, anthropometric measurements, and biochemical and ultrasonographic features in the 20 patients at entry into the study are shown in Table 1. Baseline results confirmed that isolated hypertransaminasemia was the only abnormality to emerge from the liver tests. All of the children had mild to moderate US fatty liver. Diabetes was excluded in all of the children based on the results of oral glucose tolerance test. H2BT was abnormal (H2 peak of 19 ppm above baseline at 60 minutes) only in 1 patient (of the treated group).
All of the patients completed the 8-week study with good adherence to therapy as documented by pill count and parents' supervision reports. No adverse effects were reported. Anthropometric, laboratory, and instrumental findings in the 2 groups of patients at baseline and after 8 weeks of probiotic or placebo treatment are listed in Table 2. As depicted also in Figure 1, the average variations of ALT levels in the 2 arms of the study were significantly more pronounced in the Lactobacillus GG–treated group (P = 0.03). Eight of the 10 treated patients and 3 of 10 patients in the placebo group attained ALT values <40 U/L, but the difference did not reach statistical significance (P = 0.069 Fisher exact test). PG-PS IgA significantly decreased only in the probiotic-treated group. Anthropometric data including US visceral fat, TNF-α, and US liver brightness evaluated by the hepatorenal ratio remained fairly stable in both groups during the study (Table 2).
This is the first study to investigate the effects of probiotic treatment on obesity-related liver abnormalities in children. The study was designed as a short-term pilot trial to circumvent the frequently reported unpredictable effects of lifestyle changes that usually confound the results of a long-term study in this population. In fact, even minimal weight and lifestyle changes may affect the biochemical parameters of NAFLD (4–6). Here we show that a short course of probiotic treatment significantly improved ALT values, which in most cases became normal, and that this effect was independent of weight changes.
Small intestinal bacterial overgrowth has been implicated in the pathogenesis of NAFLD (11–13). The results of the glucose H2BT, which is a surrogate test for SIBO evaluation, in our children did not confirm the increased prevalence of SIBO reported in morbidly obese adults (12). However, the level of PG-PS IgA antibodies, another surrogate for SIBO, was significantly higher than that of a group of 20 age-, sex-, and weight-matched obese nonhepatopathic controls (Table 1), and moreover they significantly decreased only in the children treated with probiotics and not in the placebo-treated group. Because PG-PS polymers are integral cell wall components of most bacterial species, including intestinal Bacteroides, which are the major determinants of SIBO (24,25), the decrease of PG-PS IgA antibodies after treatment may imply abnormalities in the patients' small intestinal microflora. Wigg et al (11) found that levels of endotoxin (ie, another SIBO marker) did not differ between patients with NAFLD and healthy controls. In a more recent work, endotoxin levels in the peripheral circulation were increased in patients with NAFLD, but they were unrelated to disease severity (26). Given the differences between peripheral and portal blood values and potential methodological difficulties caused by the short half-life of endotoxins, data obtained from the measurement of serum endotoxin levels should be interpreted with caution (11,26). For this reason, we did not measure endotoxin levels in the present study.
TNF-α is believed to be one of the protagonists of the inflammatory response in NAFLD after stimulation via Toll-like receptors by endotoxins, which reach the liver via portal flow through a more permeable intestinal barrier. However, similar to a previous study (11), the TNF-α values in our hepatopathic obese patients overlapped those of the 20 age- and weight-paired obese individuals. They tended toward a mild decrease in both treatment arms without, however, reaching statistical significance. Lactobacillus GG–induced modulation on studied parameters involved in NAFLD pathogenesis will probably require long-term studies. The effects of probiotics on NAFLD have been evaluated in animal models and, although to a lesser extent, in human adults. Probiotic VSL#3 treatment in NAFLD ob/ob mice lowered the transaminase level and affected histological liver inflammation without modifying steatosis (14). The same treatment in mice with diet-induced NAFLD attenuated fibrosis without affecting steatohepatitis (15,27). The unchanged US hepatorenal ratios in our patients are in line with these data. Another study showed that probiotics had a beneficial effect on several types of human liver disease including NAFLD, probably acting on the steatosis component (16). Contrary to what may be expected from the results of animal studies, a recent preliminary report on human NAFLD found a reversible increase in magnetic resonance imaging–calculated fat liver content during probiotic treatment (17). However, details of possible simultaneous weight changes were not reported. Given the above, individuals enrolled in such studies should undergo accurate anthropometric monitoring to ensure that other factors do not confound the effect of treatment on the outcome.
Our results do not shed light on the mechanism(s) underlying the improvement of our patients' ALT values. Recent evidence indicates that there are quantitative (ie, SIBO) and qualitative differences in gut microbiota and microbiome between lean and obese individuals (28). The fact that our patients were completely asymptomatic from a gastrointestinal viewpoint may argue against the existence of a true SIBO. Because positive H2BTs in the exceedingly large percentage of Wiggs' patient series has been questioned by several investigators (29), it is feasible that in our study, Lactobacillus GG exerted a beneficial effect on obesity-related liver disease by modulating an altered bacterial composition rather than reducing its quantity (30). However, methodological aspects may have affected the results of the glucose H2BT in detecting SIBO; namely, the dose of the sugar or the duration of the test may not have been sufficient to exclude a bacterial contamination restricted to the distal small bowel. (21) Future large-scale intervention trials investigating the potential benefit of probiotics should include a combination of gastrointestinal dysfunction tests, including lactulose H2BT and intestinal permeability studies.
In view of the excellent tolerance and the remarkable reduction of transaminase after probiotics in our double-blind, placebo-controlled, short-term pilot study, Lactobacillus rhamnosus strain GG could be considered a potential therapeutic tool for pediatric obesity-related liver disease in children who are unable to lose weight. Our promising data provide a rationale for starting a larger study on possibly biopsy-proven pediatric NAFLD.
We are grateful to Patrizia Colicchio, MD, and Brunella Aceto, dietician, for assistance in the clinical management of patients; Maria Passaretti, PhD, and Luciano Rapacciuolo, PhD, for laboratory assistance; and Francesco Manguso, MD, for critical evaluation of the manuscript. We are indebted to Jean Ann Gilder (Scientific Communication srl) for editing the text. Lactobacillus rhamnosus strain GG (ATCC 53103) with verified composition and indistinguishable placebo were supplied by Dicofarm SpA, Rome, Italy.
1. Yanovski SZ, Yanovski JA. Obesity prevalence in the United States—up, down, or sideways? N Engl J Med 2011; 364:987–989.
2. Younossi ZM, Stepanova M, Afendy M, et al. Changes in the prevalence of the most common causes of chronic liver diseases in the United States from 1988 to 2008. Clin Gastroenterol Hepatol
. 2011 Apr 11 [Epub ahead of print].
3. Schwimmer JB, Deutsch R, Kahen T, et al
. Prevalence of fatty liver in children and adolescents. Pediatrics 2006; 118:1388–1393.
4. Franzese A, Vajro P, Argenziano A, et al
. Liver involvement in obese children. Ultrasonography and liver enzyme levels at diagnosis and during follow-up in an Italian population. Dig Dis Sci 1997; 42:1428–1432.
5. St George A, Bauman A, Johnston A, et al
. Effect of a lifestyle intervention in patients with abnormal liver enzymes and metabolic risk factors. J Gastroenterol Hepatol 2009; 24:399–407.
6. Nobili V, Alisi A, Raponi M. Pediatric non-alcoholic fatty liver disease: preventive and therapeutic value of lifestyle intervention. World J Gastroenterol 2009; 15:6017–6022.
7. Bellentani S, Scaglioni F, Marino M, et al
. Epidemiology of non-alcoholic fatty liver disease. Dig Dis 2010; 28:155–161.
8. Caldwell S, Argo C. The natural history of non-alcoholic fatty liver disease. Dig Dis 2010; 28:162–168.
9. Musso G, Gambino R, Cassader M, et al
. A meta-analysis of randomized trials for the treatment of nonalcoholic fatty liver disease. Hepatology 2010; 52:79–104.
10. Socha P, Horvath A, Vajro P, et al
. Pharmacological interventions for nonalcoholic fatty liver disease in adults and in children: a systematic review. J Pediatr Gastroenterol Nutr 2009; 48:587–596.
11. Wigg AJ, Roberts-Thomson IC, Dymock RB, et al
. The role of small intestinal overgrowth, intestinal permeability, endotoxaemia, and tumor necrosis factor α in the pathogenesis of non alcoholic steatohepatitis. Gut 2001; 48:206–211.
12. Sabaté JM, Jouët P, Harnois F, et al
. High prevalence of small intestinal bacterial overgrowth in patients with morbid obesity: a contributor to severe hepatic steatosis. Obes Surg 2008; 18:371–377.
13. Miele L, Valenza V, La Torre G, et al
. Increased intestinal permeability and tight junction alterations in nonalcoholic fatty liver disease. Hepatology 2009; 49:1877–1887.
14. Li Z, Yang S, Lin H, et al
and antibodies to TNF inhibit inflammatory activity and improve nonalcoholic fatty liver disease. Hepatology 2003; 37:343–350.
15. Esposito E, Iacono A, Bianco G, et al
reduce the inflammatory response induced by a high-fat diet in the liver of young rats. J Nutr 2009; 139:905–911.
16. Loguercio C, Federico A, Tuccillo C, et al
. Beneficial effects of a probiotic VSL#3 on parameters of liver dysfunction in chronic liver diseases. J Clin Gastroenterol 2005; 39:540–543.
17. Solga SF, Buckley G, Clark JM, et al
. The effect of a probiotic on hepatic steatosis. J Clin Gastroenterol 2008; 42:1117–1119.
18. Sabir N, Sermez Y, Kazil S, et al
. Correlation of abdominal fat accumulation and liver steatosis: importance of ultrasonographic and anthropometric measurements. Eur J Ultrasound 2001; 14:121–128.
19. Vajro P, Fontanella A, Perna C, et al
. Persistent hyperaminotransferasemia resolving after weight reduction in obese children. J Pediatr 1994; 125:239–241.
20. Mancini M, Prinster A, Annuzzi G, et al
. Sonographic hepatic-renal ratio as indicator of hepatic steatosis: comparison with (1)H magnetic resonance spectroscopy. Metabolism 2009; 58:1724–1730.
21. Gasbarrini A, Corazza GR, Gasbarrini G, et al
. Methodology and indications of H2-breath testing in gastrointestinal diseases: the Rome Consensus Conference. Aliment Pharmacol Ther 2009; 29(suppl 1):1–49.
22. Hazenberg MP, de Visser H, Bras MJ, et al
. Serum antibodies to peptidoglycan-polysaccharide complexes from the anaerobic intestinal flora in patients with Crohn's disease. Digestion 1990; 47:172–180.
23. Schrijver IA, De Man YA, Melief MJ, et al
. Reduced systemic IgG levels against peptidoglycan in rheumatoid arthritis (RA) patients. Clin Exp Immunol 2001; 123:140–146.
24. Lichtman SN, Keku J, Schwab JH, et al
. Evidence for peptidoglycan absorption in rats with experimental small bowel bacterial overgrowth. Infect Immun 1991; 59:555–562.
25. Lumsden AB, Henderson JM, Kutner MH. Endotoxin levels measured by a chromogenic assay in portal, hepatic and peripheral venous blood in patients with cirrhosis. Hepatology 1988; 8:232–236.
26. Harte AL, da Silva NF, Creely SJ, et al
. Elevated endotoxin levels in non-alcoholic fatty liver disease. J Inflamm (Lond) 2010; 7:15.
27. Velayudham A, Dolganiuc A, Ellis M, et al
. VSL#3 probiotic treatment attenuates fibrosis without changes in steatohepatitis in a diet-induced nonalcoholic steatohepatitis model in mice. Hepatology 2009; 49:989–997.
28. Turnbaugh PJ, Gordon JI. The core gut microbiome, energy balance and obesity. J Physiol 2009; 587:4153–4158.
29. Riordan SM, Duncombe VM, Thomas MC, et al
. Small intestinal bacterial overgrowth, intestinal permeability, and non-alcoholic steatohepatitis. Gut 2002; 50:136–138.
30. Abu-Shanab A, Quigley EM. The role of the gut microbiota in nonalcoholic fatty liver disease. Nat Rev Gastroenterol Hepatol 2010; 7:691–701.