Advertisement

Lifestyle interventions for the treatment of non-alcoholic fatty liver disease in adults: A systematic review

  • Christian Thoma
    Affiliations
    Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK

    NIHR Biomedical Research Centre for Ageing & Age-related Disease, Newcastle University, Newcastle upon Tyne, UK

    MRC Centre for Brain Ageing & Vitality, Newcastle University, Newcastle upon Tyne, UK
    Search for articles by this author
  • Christopher P. Day
    Affiliations
    Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK

    NIHR Biomedical Research Centre for Ageing & Age-related Disease, Newcastle University, Newcastle upon Tyne, UK
    Search for articles by this author
  • Michael I. Trenell
    Correspondence
    Corresponding author. Address: MoveLab; 4th Floor William Leech Building, Newcastle University, Newcastle upon Tyne, NE4 6BE, UK. Tel.: +44 191 222 6935.
    Affiliations
    Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK

    NIHR Biomedical Research Centre for Ageing & Age-related Disease, Newcastle University, Newcastle upon Tyne, UK

    MRC Centre for Brain Ageing & Vitality, Newcastle University, Newcastle upon Tyne, UK
    Search for articles by this author
Open AccessPublished:July 04, 2011DOI:https://doi.org/10.1016/j.jhep.2011.06.010
      Non-alcoholic fatty liver disease is a serious and growing clinical problem. Despite lifestyle modification, i.e. diet and physical activity, being the recommended therapy, there are currently no systematic evaluations of its efficacy. This review applies a systematic approach to evaluating lifestyle modifications studied to date.
      Medline (Pubmed), Scopus, and the Cochrane Controlled Trials Register were searched for studies and study groups assessing the effect of diet, physical activity, and/or exercise modification in adult populations with non-alcoholic fatty liver disease. The outcome markers of interest were indicators of steatosis, histological evidence of inflammation and fibrosis, and glucose control/insulin sensitivity.
      We identified 23 studies for inclusion; seven had control groups, but only six were randomised. Eleven groups received diet-only interventions, two exercise-only, and 19 diet and physical activity/exercise. Studies consistently showed reductions in liver fat and/or liver aminotransferase concentration, with the strongest correlation being with weight reduction. Of the 5 studies reporting changes in histopathology, all showed a trend towards reduction in inflammation, in 2 this was statistically significant. Changes in fibrosis were less consistent with only one study showing a significant reduction. The majority of studies also reported improvements in glucose control/insulin sensitivity following intervention. However, study design, definition of disease, assessment methods, and interventions varied considerably across studies.
      Lifestyle modifications leading to weight reduction and/or increased physical activity consistently reduced liver fat and improved glucose control/insulin sensitivity. Limited data also suggest that lifestyle interventions may hold benefits for histopathology.

      Abbreviations:

      NAFLD (non-alcoholic fatty liver disease), IHTAG (intrahepatic triacyglycerol concentration), 1H-MRS (proton magnetic resonance spectroscopy), CT (computed tomography), ALT (alanine aminotransferase), AST (aspartate aminotransferase)

      Keywords

      Introduction

      Non-alcoholic fatty liver disease (NAFLD) encompasses liver conditions ranging from hepatic steatosis through steatohepatitis to cirrhosis [
      • Day C.P.
      Non-alcoholic fatty liver disease: current concepts and management strategies.
      ]. Its prevalence has been estimated at between 20% and 33% of the adult population depending on criteria and country [
      • Barshop N.J.
      • Sirlin C.B.
      • Schwimmer J.B.
      • Lavine J.E.
      Review article: epidemiology, pathogenesis and potential treatments of paediatric non-alcoholic fatty liver disease.
      ]. Prevalence increases with degree of obesity [
      • Fabbrini E.
      • Sullivan S.
      • Klein S.
      Obesity and nonalcoholic fatty liver disease: Biochemical, metabolic, and clinical implications.
      ] and the condition is very common in those with type 2 diabetes [
      • Kotronen A.
      • Juurinen L.
      • Hakkarainen A.
      • Westerbacka J.
      • Cornér A.
      • Bergholm R.
      • et al.
      Liver fat is increased in type 2 diabetic patients and underestimated by serum alanine aminotransferase compared with equally obese nondiabetic subjects.
      ,
      • Kotronen A.
      • Juurinen L.
      • Tiikkainen M.
      • Vehkavaara S.
      • Yki-Järvinen H.
      Increased liver fat, impaired insulin clearance, and hepatic and adipose tissue insulin resistance in type 2 diabetes.
      ]. Rising prevalence of obesity and Type 2 diabetes, particularly in younger people, will ensure that NAFLD remains a growing clinical concern for the future [
      • Fabbrini E.
      • Sullivan S.
      • Klein S.
      Obesity and nonalcoholic fatty liver disease: Biochemical, metabolic, and clinical implications.
      ].
      Elevated intrahepatic triacyglycerol concentration (IHTAG) is the first step in the development of steatohepatitis, liver fibrosis, liver cirrhosis, and hepatocellular carcinoma [
      • Day C.P.
      Non-alcoholic fatty liver disease: current concepts and management strategies.
      ]. Excess liver fat is also linked to insulin resistance [
      • Angelico F.
      • Del Ben M.
      • Conti R.
      • Francioso S.
      • Feole K.
      • Fiorello S.
      • et al.
      Insulin resistance, the metabolic syndrome, and nonalcoholic fatty liver disease.
      ], and is an independent risk factor for Type 2 diabetes [
      • Fraser A.
      • Harris R.
      • Sattar N.
      • Ebrahim S.
      • Davey Smith G.
      • Lawlor D.A.
      Alanine aminotransferase, gamma-glutamyltransferase, and incident diabetes: the British Women’s Heart and Health Study and meta-analysis.
      ] and cardiovascular disease [
      • Targher G.
      • Day C.P.
      • Bonora E.
      Risk of cardiovascular disease in patients with nonalcoholic fatty liver disease.
      ]. Liver lipid is part of the early adaptive response to stress, and as such, is a biomarker of NEFA flux, oxidative-, ER- and cytokine-mediated stress that result in steatosis and progressive liver damage. Lifestyle modification, encompassing diet, physical activity, and/or exercise related behaviours, is the primary recommended therapy for NAFLD [
      • Loria P.
      • Adinolfi L.E.
      • Bellentani S.
      • Bugianesi E.
      • Grieco A.
      • Fargion S.
      • et al.
      Practice guidelines for the diagnosis and management of nonalcoholic fatty liver disease. A decalogue from the Italian Association for the Study of the Liver (AISF) Expert Committee.
      ], especially in the absence of approved pharmaceutical agents.
      Although reviews with NAFLD disease as their focus abound, few have reported using a systematic approach to study selection or reporting, and none, to date, have applied this approach to examining the efficacy or effectiveness of lifestyle management. This systematic approach is necessary to provide clinical care teams with the information to determine whether lifestyle therapy should be used, and if so, what aspects are key to achieving success. Our objective was to perform a systematic assessment of lifestyle interventions in adults with NAFLD to: (i) Define the efficacy of different lifestyle interventions in reducing IHTAG and/or liver aminotransferases; (ii) assess the effect of lifestyle interventions on histological parameters; and (iii) establish the efficacy of different lifestyle interventions on glucose control/insulin sensitivity (Fig. 1).

      Methods

      Eligibility criteria

      The review is restricted to published prospective interventions reporting the effects of lifestyle modification on IHTAG, liver enzymes, and/or insulin sensitivity in adults (⩾19 years) with NAFLD, including non-alcoholic steatohepatitis but not late stage liver diseases i.e. cirrhosis or hepatocellular carcinoma. Eligible publications included: randomised controlled trials or specific arms thereof, and non-randomised interventions. Only full reports were considered to provide sufficient information to allow critical evaluation.
      No specific criteria defining NAFLD were set as the methods of diagnosis and cut-offs vary between studies. It was considered sufficient for reports to provide their own diagnostic criteria based on one or more of the following in order of preference: (1) histological examination of biopsies; (2) proton magnetic resonance spectroscopy (1H-MRS); (3) computed tomography (CT); (4) ultrasound; and/or (5) blood concentrations of alanine aminotransferase (ALT) and/or aspartate aminotransferase (AST).
      Lifestyle modification could include general recommendations or specific diet, physical activity, and/or exercise prescription. Studies or study arms designed to test pharmaceuticals, dietary supplements, or herbal preparations were excluded. Study arms in which pharmaceutical agents were used as part of standard treatment and where participants were receiving these prior to the study, without a reported increase in dose during the study, were eligible for inclusion.
      The primary outcomes of interest were changes in IHTAG assessed by liver biopsy, 1H-MRS, CT, or ultrasound, and histological indicators of inflammation and fibrosis. Blood ALT and/or AST concentrations were also considered. The secondary outcome was glucose tolerance and/or insulin sensitivity as assessed directly by insulin clamp techniques or oral glucose tolerance tests, or inferred by validated formulae.
      Only studies that clearly described or appropriately referenced their intervention, and that provided some direct indicators of protocol adherence, or those conducted under very close supervision, e.g. inpatient protocol delivery, were eligible for inclusion.

      Search strategy and study selection

      The following databases were searched: Medline (Pubmed), Scopus, and the Cochrane Controlled Trials Register. The search of Scopus, the most comprehensive of the three databases, was done in duplicate by two authors (CT and MT), whereas the other databases were searched by one author (CT). The last search of all three databases was done on June 26, 2010. However, automatic updates of the Scopus search were reviewed up to October 18, 2010. A medical librarian assisted with the selection of the search strategies.
      The selected search terms and related MESH headings were: (NAFLD or “non-alcoholic fatty liver” or “nonalcoholic fatty liver” or “non-alcoholic steatohepatitis” or “nonalcoholic steatohepatitis” or “non-alcoholic steatosis” or “nonalcoholic steatosis” or “non-alcoholic liver steatosis” or “nonalcoholic liver steatosis” or “non-alcoholic hepatic steatosis” or “nonalcoholic hepatic steatosis”) AND (lifestyle or exercise or “diet∗” or diet or training or behaviour or behavior or nutrition or sport or “physical activity” or “weight reduction” or “weight loss” or “energy restriction”). These were restricted to title, abstract, and keyword (Scopus only). The database permitting, the following were excluded: reviews; letters; editorials; commentaries; animal studies; and studies in those aged under 19 years. Review of articles was restricted to those published in English.
      Titles and abstracts of studies identified were evaluated against eligibility criteria. Studies appearing eligible based on their abstract were read in full. The decision to exclude any of these studies was made by the consensus of two authors (CT & MT).

      Data items

      The items of interest from each report included: study type/design; diagnostic criteria for NAFLD; inclusion and exclusion criteria; blinding; similarity of groups at baseline; sex; age; definition of participant adherence; treatment protocol, including professions involved and contact time; reported adherence; criteria for dealing with medication; methods used to assess diet and physical activity; loss to follow-up; intention-to-treat or per-protocol analysis; IHTAG; measures of glucose control; ALT and/or AST concentration.

      Data extraction

      Relevant data from included reports were recorded in itemised tables. Results were converted to SI units or otherwise standardised and changes from baseline converted to percentages to facilitate comparison across studies. Where liver fat is given as a percentage, a change from 10% fat to 5% fat is referred to as an absolute reduction of 5% (10–5%) and a relative reduction of 50%.
      Multiple publications from the same study were identified by comparing author names, sample sizes, and intervention protocols. Where papers noted that other reports of the study existed, these were also obtained to allow consistency between different reports to be assessed and/or missing data to be obtained.

      Risk of intra- and inter-study reporting and publication bias

      Included studies were compared to their published protocols when available to identify omissions of outcome data. Alternatively, the methods section of each report was compared to the results section to assess reporting bias. The International Clinical Trials Registry, EU Clinical Trial Register, and metaRegister of Controlled Trials were searched using the key words fatty liver and steatohepatitis to identify trials described as completed. Studies registered prior to 2009 with records not updated in the past 12 months were assumed to be completed. A literature search using the relevant principle investigator was conducted to identify publications resulting from relevant registered trials.

      Description and critique of primary outcome indicators

      Liver biopsy

      Histological examination of biopsy samples can assess the presence of necro-inflammation and fibrosis, and can differentiate between macro- and micro-vesicular steatosis; it remains the reference standard for the grading and staging of NAFLD [
      • Loria P.
      • Adinolfi L.E.
      • Bellentani S.
      • Bugianesi E.
      • Grieco A.
      • Fargion S.
      • et al.
      Practice guidelines for the diagnosis and management of nonalcoholic fatty liver disease. A decalogue from the Italian Association for the Study of the Liver (AISF) Expert Committee.
      ]. However, it is subject to sampling error due to histological heterogeneity [
      • Merriman R.B.
      • Ferrell L.D.
      • Patti M.G.
      • Weston S.R.
      • Pabst M.S.
      • Aouizerat B.E.
      • et al.
      Correlation of paired liver biopsies in morbidly obese patients with suspected nonalcoholic fatty liver disease.
      ,
      • Larson S.P.
      • Bowers S.P.
      • Palekar N.A.
      • Ward J.A.
      • Pulcini J.P.
      • Harrison S.A.
      Histopathologic variability between the right and left lobes of the liver in morbidly obese patients undergoing Roux-en-Y bypass.
      ], scoring is semi-quantitative limiting its ability to detect modest changes, and scoring systems vary between reports precluding direct comparisons.

      Proton magnetic resonance spectroscopy

      1H-MRS quantitatively measures IHTAG by differentiating between signals from lipids and water [
      • Schwenzer N.F.
      • Springer F.
      • Schraml C.
      • Stefan N.
      • Machann J.
      • Schick F.
      Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance.
      ]. This technique has superior accuracy and sensitivity to CT and ultrasound [
      • Lee S.S.
      • Park S.H.
      • Kim H.J.
      • Kim S.Y.
      • Kim M.-Y.
      • Kim D.Y.
      • et al.
      Non-invasive assessment of hepatic steatosis: prospective comparison of the accuracy of imaging examinations.
      ]. Using 1H-MRS, IHTAG above 5–5.6% is considered elevated [
      • Szczepaniak L.S.
      • Nurenberg P.
      • Leonard D.
      • Browning J.D.
      • Reingold J.S.
      • Grundy S.
      • et al.
      Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the general population.
      ,
      • Korenblat K.M.
      • Fabbrini E.
      • Mohammed B.S.
      • Klein S.
      Liver, muscle, and adipose tissue insulin action is directly related to intrahepatic triglyceride content in obese subjects.
      ].

      Computed tomography

      CT provides a semi-quantitative method for the evaluation of IHTAG based on the change in image intensity, measured in Houndsfield units, between the liver and either the spleen, which stores no fat, or an external lipid standard [
      • Schwenzer N.F.
      • Springer F.
      • Schraml C.
      • Stefan N.
      • Machann J.
      • Schick F.
      Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance.
      ]. An increase in liver:spleen ratio or liver density is indicative of reduced IHTAG.

      Ultrasound

      Ultrasound provides semi-quantitative estimates of hepatic steatosis based on diffuse increases in echogenicity [
      • Schwenzer N.F.
      • Springer F.
      • Schraml C.
      • Stefan N.
      • Machann J.
      • Schick F.
      Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance.
      ]. Reported sensitivity and specificity vary between 60–94% and 66–95%, respectively [
      • Schwenzer N.F.
      • Springer F.
      • Schraml C.
      • Stefan N.
      • Machann J.
      • Schick F.
      Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance.
      ]. A study of inter- and intra-observer variability reported a mean agreement for the presence of steatosis of 72% and 76%, respectively, and intra-observer agreement of severity of 55–68% [
      • Strauss S.
      • Gavish E.
      • Gottlieb P.
      • Katsnelson L.
      Interobserver and intraobserver variability in the sonographic assessment of fatty liver.
      ].

      Blood biomarkers

      Some studies have based their diagnosis of NAFLD on the liver enzymes ALT and/or AST. However, these are non-specific for steatosis or disease stage. In a cohort of 708 individuals with elevated intrahepatic lipid (>5.6% assessed by 1H-MRS) it was found that 79% had normal ALT [
      • Browning J.D.
      • Szczepaniak L.S.
      • Dobbins R.
      • Nuremberg P.
      • Horton J.D.
      • Cohen J.C.
      • et al.
      Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity.
      ]. Further, normal ALT and AST concentrations have been reported in the presence of histological evidence of steatosis, fibrosis, and cirrhosis [
      • Mofrad P.
      • Contos M.J.
      • Haque M.
      • Sargeant C.
      • Fisher R.A.
      • Luketic V.A.
      • et al.
      Clinical and histologic spectrum of nonalcoholic fatty liver disease associated with normal ALT values.
      ,
      • Fracanzani A.L.
      • Valenti L.
      • Bugianesi E.
      • Andreoletti M.
      • Colli A.
      • Vanni E.
      • et al.
      Risk of severe liver disease in nonalcoholic fatty liver disease with normal aminotransferase levels: a role for insulin resistance and diabetes.
      ].

      Assessment of glucose control

      The euglycemic hyperinsulinaemic clamp is the reference standard for the assessment of insulin sensitivity, with the more commonly performed frequently sampled 2-h oral glucose test showing good correlation with the clamp [
      • Borai A.
      • Livingstone C.
      • Ferns G.A.A.
      The biochemical assessment of insulin resistance.
      ]. Measures of glucose control derived from fasting glucose and insulin ratio, such as the homeostasis model assessment (HOMA) are less sensitive options to assess glucose control [
      • Borai A.
      • Livingstone C.
      • Ferns G.A.A.
      The biochemical assessment of insulin resistance.
      ].

      Other sources of bias or confounding

      The following sources of bias are most relevant to this review: inclusion of non-randomised and non-controlled trials; the inability to conceal allocation in lifestyle therapy; methodological heterogeneity for diagnosis; misclassification; and selection bias. Restriction to randomised controlled trials and/or studies using only narrow diagnostic criteria and direct assessment of liver fat would have limited the scope of this review, and thereby its ability to address its aims.
      Misclassification of disease is plausible as elevated IHTAG has several potential causes [
      • Larter C.Z.
      • Chitturi S.
      • Heydet D.
      • Farrell G.C.
      A fresh look at NASH pathogenesis. Part 1: The metabolic movers.
      ], the relative contributions of which are not routinely assessed. A contribution from alcohol is particularly difficult to rule out entirely due to variations in definitions for levels of intake [
      • Gunzerath L.
      • Faden V.
      • Zakhari S.
      • Warren K.
      National Institute on Alcohol Abuse and Alcoholism report on moderate drinking.
      ], and limitations in existing biomarkers and questionnaires, which often focus on diagnosing dependence rather than accurately quantifying intake [
      • Niemela O.
      Biomarkers in alcoholism.
      ]. These do not invalidate a link between intervention and outcome, but do increase the chance of attributing the observed change to the wrong factors.

      Results

      General limitations of studies reviewed

      Specific details about individual studies reviewed can be found in Table 1, Table 2, Table 3, Table 4. There was considerable heterogeneity of assessment methods used, diagnostic criteria for NAFLD applied, and the detail with which exclusion criteria were reported. With one exception, studies employing 1H-MRS either did not report a minimum IHTAG or report including participants <5%. Those employing histological examination of biopsies applied cutoffs, but used different scoring systems. Eligibility based on history of alcohol consumption varied between no alcohol intake and 560 g/week, and assessment methods were seldom cited. Although most studies noted excluding participants with other liver conditions, including potential drug induced steatosis, few provided comprehensive criteria, e.g. of drugs deemed steatogenic, or details of the relevant analytical methods employed. Collectively, these factors suggest considerable heterogeneity in study populations and limit direct inter-study comparisons and extrapolation to patient populations.
      Table 1Diet only interventions.
      Data are mean ± standard errors usually rounded to the nearest full number; changes reported were statistically significant p <0.05 unless noted otherwise. Sample size reflects those in the final analysis.
      ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body max index (kg/m2); BW, body weight; d, day(s); CHO, carbohydrate; CT, computed tomography; d, day(s); E, energy; EGP, endogenous glucose production; E-HC, euglycaemichyperinsulinaemic clamp; GRa, glucose rate of appearance; h, hour(s); HbA1c, glycated haemoglobin; HISI, hepatic insulin sensitivity index; HOMA, homeostasis model assessment; IHTAG, intrahepatic triacyglycerol concentration; IMSGP, insulin-mediated suppression of glucose production; IMGU, insulin-mediated glucose uptake; IS, whole body insulin sensitivity; ITT, intention to treat analysis; LBM, lean body mass; n, sample size of cohort in final statistical analysis; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; n/r, data not reported; n.s., result not statistically significant; PP, per protocol analysis; PRO, protein; PUFA, polyunsaturated fatty acid; RCT, randomised controlled trial; RT, randomised trial with no control; RMR, resting metabolic rate; SAFA, saturated fatty acid; UCT, uncontrolled trial; US, ultrasound; w, week(s).
      Diagnoses by histopathology at baseline only, no repeat biopsies on completion of the intervention.
      Table 2Exercise only interventions.
      Data are mean ± standard errors; changes were statistically significant p <0.05 unless noted otherwise.
      1H-MRS, proton energy magnetic resonance spectroscopy; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body max index (kg/m2); BW, body weight; d, day(s); HOMA, homeostasis model assessment insulin resistance; IHTAG, intrahepatic triacylglycerol concentration; min, minute; n/r, not reported; n.s., not statistically significant change; PP, per protocol analysis; UCT, uncontrolled trial; VO2 peak, peak oxygen consumption ml/min/kg; wk, week(s).
      Age was only reported for the combined study groups – see Table 4 for other study group.
      Table 3Interventions combining diet and broad physical activity/exercise advice.
      Data are mean ± standard errors usually rounded to the nearest full number; changes reported were statistically significant p <0.05 unless noted otherwise. Sample size based reflects those in the final analysis.
      ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body max index (kg/m2); BW, body weight; d, day(s); CHO, carbohydrate; CT, computed tomography; d, day(s); E, energy; E-HC, euglycaemic-hyperinsulinaemic clamp; GRa, glucose rate of appearance; GRd, glucose rate of disappearance; h, hour(s); HbA1c, glycated haemoglobin; HOMA, homeostasis model assessment; IHTAG, intrahepatic triacyglycerol concentration; IS, whole body insulin sensitivity; ITT, intention to treat analysis; MET, metabolic equivalent of task; n, sample size of cohort in final statistical analysis; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; n/r, data not reported; n.s., result not statistically significant; PP, per protocol analysis; PRO, protein; RCT, randomised controlled trial; RT, randomised trial with no control; SAFA, saturated fatty acid; UCT, uncontrolled trial; w, week(s).
      Table 4Interventions combining diet and specific physical activity/exercise advice.
      Data are mean ± standard errors usually rounded to the nearest whole number. Changes reported as percentages where possible. Changes reported were statistically significant p <0.05 unless noted otherwise.
      2 h Glucose, blood glucose concentration at 2 h post-oral glucose tolerance challenge; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body max index (kg/m2); BW, body weight; d, day(s); CHO, carbohydrate; CT, computed tomography; d, day; E, energy; EGP, endogenous glucose production; E-HC, euglycaemic-hyperinsulinaemic clamp; FFQ, food frequency questionnaire; GI, glycaemic index; GRa, glucose rate of appearance; GRd, glucose rate of disappearance; HISI, hepatic insulin sensitivity index; HOMA, homeostasis model assessment; IHTAG, intrahepatic triacylglycerol concentration; ITT, intention to treat analysis; MET, metabolically equivalent tasks; min, minute; mod, moderate; MUFA, monounsaturated fatty acid; NAFLD, non-alcoholic fatty liver disease; n/r, data not reported; n.s., no statistically significant change; NAS, non-alcoholic fatty liver disease activity score; NASH, non-alcoholic steatohepatitis; LBM, lean body mass; NRCT, non-randomised control group; OGTT, oral glucose tolerance test; PA, physical activity; PP, per protocol; PRO, protein; wk, week.
      Monitoring of adherence to diet or exercise was often limited. No studies reported using objective measures of physical activity such as accelerometers, instead questionnaires were used. Dietary assessment methods were not reported in sufficient detail to assess likely accuracy. The greater the complexity of interventions, the less precisely the actions of participants tended to be described. Interventions involving diet, exercise, and behaviour change methods were reported with a focus on outcomes with limited information on diet and physical activity adherence or how behaviour change methods were applied; thus limiting our ability to provide precise recommendations based on these interventions.
      Studies did not report allocation concealment during data analysis except for blinding during the analysis of liver biopsy samples. Some studies doing per protocol analysis did not provide baseline data for the specific group in the final analysis. Most studies did not have a control group; those that did provided this group with some form of limited intervention. IHTAG measured by 1H-MRS appears stable over a 4-week period, in the absence of any intervention [
      • van Werven J.R.
      • Hoogduin J.M.
      • Nederveen A.J.
      • van Vliet A.A.
      • Wajs E.
      • Vandenberk P.
      • et al.
      Reproducibility of 3.0 Tesla magnetic resonance spectroscopy for measuring hepatic fat content.
      ]. Therefore, gross overestimation of efficacy is unlikely.

      Study findings

      Diet only interventions

      We identified 11 eligible study groups including 322 participants (approximately 65% women, 20 controls) prescribing dietary change: six using low-to-moderate fat/moderate-to-high carbohydrate energy restricted diets [
      • Elias M.C.
      • Parise E.R.
      • Carvalho Ld.
      • Szejnfeld D.
      • Netto J.P.
      Effect of 6-month nutritional intervention on non-alcoholic fatty liver disease.
      ,
      • Kirk E.
      • Reeds D.N.
      • Finck B.N.
      • Mayurranjan S.M.
      • Mayurranjan M.S.
      • Patterson B.W.
      • et al.
      Dietary fat and carbohydrates differentially alter insulin sensitivity during caloric restriction.
      ,
      • Yamamoto M.
      • Iwasa M.
      • Iwata K.
      • Kaito M.
      • Sugimoto R.
      • Urawa N.
      • et al.
      Restriction of dietary calories, fat and iron improves non-alcoholic fatty liver disease.
      ,
      • Petersen K.F.
      • Dufour S.
      • Befroy D.
      • Lehrke M.
      • Hendler R.E.
      • Shulman G.I.
      Reversal of nonalcoholic hepatic steatosis, hepatic insulin resistance, and hyperglycemia by moderate weight reduction in patients with type 2 diabetes.
      ,
      • de Luis D.A.
      • Aller R.
      • Izaola O.
      • Sagrado M.G.
      • Conde R.
      • Gonzalez J.M.
      Effect of a hypocaloric diet in transaminases in nonalcoholic fatty liver disease and obese patients, relation with insulin resistance.
      ,
      • Viljanen A.P.M.
      • Iozzo P.
      • Borra R.
      • Kankaanpää M.
      • Karmi A.
      • Lautamäki R.
      • et al.
      Effect of weight loss on liver free fatty acid uptake and hepatic insulin resistance.
      ], one of which also specifically restricted iron intake [
      • Yamamoto M.
      • Iwasa M.
      • Iwata K.
      • Kaito M.
      • Sugimoto R.
      • Urawa N.
      • et al.
      Restriction of dietary calories, fat and iron improves non-alcoholic fatty liver disease.
      ]; three groups were given low carbohydrate ketogenic diets [
      • Kirk E.
      • Reeds D.N.
      • Finck B.N.
      • Mayurranjan S.M.
      • Mayurranjan M.S.
      • Patterson B.W.
      • et al.
      Dietary fat and carbohydrates differentially alter insulin sensitivity during caloric restriction.
      ,
      • Tendler D.
      • Lin S.
      • Yancy W.
      • Mavropoulos J.
      • Sylvestre P.
      • Rockey D.
      • et al.
      The effect of a low-carbohydrate, ketogenic diet on nonalcoholic fatty liver disease: a pilot study.
      ,
      • Benjaminov O.
      • Beglaibter N.
      • Gindy L.
      • Spivak H.
      • Singer P.
      • Wienberg M.
      • et al.
      The effect of a low-carbohydrate diet on the nonalcoholic fatty liver in morbidly obese patients before bariatric surgery.
      ]; and two high protein diets [
      • Viljanen A.P.M.
      • Iozzo P.
      • Borra R.
      • Kankaanpää M.
      • Karmi A.
      • Lautamäki R.
      • et al.
      Effect of weight loss on liver free fatty acid uptake and hepatic insulin resistance.
      ,
      • Gasteyger C.
      • Larsen T.M.
      • Vercruysse F.
      • Astrup A.
      Effect of a dietary-induced weight loss on liver enzymes in obese subjects.
      ]. Two studies employed biopsy [
      • Yamamoto M.
      • Iwasa M.
      • Iwata K.
      • Kaito M.
      • Sugimoto R.
      • Urawa N.
      • et al.
      Restriction of dietary calories, fat and iron improves non-alcoholic fatty liver disease.
      ,
      • Tendler D.
      • Lin S.
      • Yancy W.
      • Mavropoulos J.
      • Sylvestre P.
      • Rockey D.
      • et al.
      The effect of a low-carbohydrate, ketogenic diet on nonalcoholic fatty liver disease: a pilot study.
      ], but only one at follow-up [
      • Tendler D.
      • Lin S.
      • Yancy W.
      • Mavropoulos J.
      • Sylvestre P.
      • Rockey D.
      • et al.
      The effect of a low-carbohydrate, ketogenic diet on nonalcoholic fatty liver disease: a pilot study.
      ], the other used ALT and AST at follow-up [
      • Yamamoto M.
      • Iwasa M.
      • Iwata K.
      • Kaito M.
      • Sugimoto R.
      • Urawa N.
      • et al.
      Restriction of dietary calories, fat and iron improves non-alcoholic fatty liver disease.
      ]; three used 1H-MRS [
      • Kirk E.
      • Reeds D.N.
      • Finck B.N.
      • Mayurranjan S.M.
      • Mayurranjan M.S.
      • Patterson B.W.
      • et al.
      Dietary fat and carbohydrates differentially alter insulin sensitivity during caloric restriction.
      ,
      • Petersen K.F.
      • Dufour S.
      • Befroy D.
      • Lehrke M.
      • Hendler R.E.
      • Shulman G.I.
      Reversal of nonalcoholic hepatic steatosis, hepatic insulin resistance, and hyperglycemia by moderate weight reduction in patients with type 2 diabetes.
      ,
      • Viljanen A.P.M.
      • Iozzo P.
      • Borra R.
      • Kankaanpää M.
      • Karmi A.
      • Lautamäki R.
      • et al.
      Effect of weight loss on liver free fatty acid uptake and hepatic insulin resistance.
      ], two used CT [
      • Elias M.C.
      • Parise E.R.
      • Carvalho Ld.
      • Szejnfeld D.
      • Netto J.P.
      Effect of 6-month nutritional intervention on non-alcoholic fatty liver disease.
      ,
      • Benjaminov O.
      • Beglaibter N.
      • Gindy L.
      • Spivak H.
      • Singer P.
      • Wienberg M.
      • et al.
      The effect of a low-carbohydrate diet on the nonalcoholic fatty liver in morbidly obese patients before bariatric surgery.
      ], three studies relied on ALT and AST [
      • de Luis D.A.
      • Aller R.
      • Izaola O.
      • Sagrado M.G.
      • Conde R.
      • Gonzalez J.M.
      Effect of a hypocaloric diet in transaminases in nonalcoholic fatty liver disease and obese patients, relation with insulin resistance.
      ,
      • Gasteyger C.
      • Larsen T.M.
      • Vercruysse F.
      • Astrup A.
      Effect of a dietary-induced weight loss on liver enzymes in obese subjects.
      ]. Only two studies had a control group [
      • Elias M.C.
      • Parise E.R.
      • Carvalho Ld.
      • Szejnfeld D.
      • Netto J.P.
      Effect of 6-month nutritional intervention on non-alcoholic fatty liver disease.
      ,
      • Yamamoto M.
      • Iwasa M.
      • Iwata K.
      • Kaito M.
      • Sugimoto R.
      • Urawa N.
      • et al.
      Restriction of dietary calories, fat and iron improves non-alcoholic fatty liver disease.
      ]; in one the control group were those with low adherence to the protocol [
      • Elias M.C.
      • Parise E.R.
      • Carvalho Ld.
      • Szejnfeld D.
      • Netto J.P.
      Effect of 6-month nutritional intervention on non-alcoholic fatty liver disease.
      ]. Intervention and outcome details are summarised in Table 1.
      Interventions lasted 1–6 months and achieved mean body weight reductions of 4–14%. All studies using biopsy or imaging techniques to estimate IHTAG reported reductions. The three studies using 1H-MRS reported absolute reductions of 4–10% and relative reductions of 42–81%. The only study to do a post-intervention biopsy (n = 5) reported reduced inflammation and trend towards reduced fibrosis (p = 0.07), as well as the reduction in steatosis, following a ketogenic diet and a mean weight reduction of 14% [
      • Benjaminov O.
      • Beglaibter N.
      • Gindy L.
      • Spivak H.
      • Singer P.
      • Wienberg M.
      • et al.
      The effect of a low-carbohydrate diet on the nonalcoholic fatty liver in morbidly obese patients before bariatric surgery.
      ]. Five out of seven studies reporting liver enzymes showed reductions and one showed no change. The study that found an increase in ALT and AST, but only in women, suggested this might have been due to the analysis being done before weight had stabilised [
      • Gasteyger C.
      • Larsen T.M.
      • Vercruysse F.
      • Astrup A.
      Effect of a dietary-induced weight loss on liver enzymes in obese subjects.
      ]. Five out of six studies reporting glucose control/insulin sensitivity noted improvements.

      Exercise only interventions

      Two studies published contained exercise only groups [
      • Johnson N.A.
      • Sachinwalla T.
      • Walton D.W.
      • Smith K.
      • Armstrong A.
      • Thompson M.W.
      • et al.
      Aerobic exercise training reduces hepatic and visceral lipids in obese individuals without weight loss.
      ,
      • Sreenivasa Baba C.
      • Alexander G.
      • Kalyani B.
      • Pandey R.
      • Rastogi S.
      • Pandey A.
      • et al.
      Effect of exercise and dietary modification on serum aminotransferase levels in patients with nonalcoholic steatohepatitis.
      ] and are summarised in Table 2. A total of 35 participants (approximately 30% women, 7 controls) were included in the two intervention groups and one control group. The interventions involved moderate intensity aerobic activity. Four weeks of stationary cycling three times per week resulted in a reduction in 1H-MRS measured IHTAG of 1.8%, relative reduction of 21%, but no statistically significant change in HOMA relative to either baseline or control [
      • Johnson N.A.
      • Sachinwalla T.
      • Walton D.W.
      • Smith K.
      • Armstrong A.
      • Thompson M.W.
      • et al.
      Aerobic exercise training reduces hepatic and visceral lipids in obese individuals without weight loss.
      ]. Three months of aerobic exercise including brisk walking/jogging or rhythmic aerobic exercise resulted in a 47% and 48% reduction in ALT and AST, respectively [
      • Sreenivasa Baba C.
      • Alexander G.
      • Kalyani B.
      • Pandey R.
      • Rastogi S.
      • Pandey A.
      • et al.
      Effect of exercise and dietary modification on serum aminotransferase levels in patients with nonalcoholic steatohepatitis.
      ]. Exercise only intervention groups in both studies maintained their baseline weight suggesting that weight reduction is not a prerequisite for liver fat or biomarker reduction.

      Exercise combined with diet

      Seven studies involving 436 participants (approximately 50% women, 98 controls) employed a selection of behaviour change methods to decrease energy intake and increase physical activity/exercise over 3–12 months [
      • Promrat K.
      • Kleiner D.E.
      • Niemeier H.M.
      • Jackvony E.
      • Kearns M.
      • Wands J.R.
      • et al.
      Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis.
      ,
      • Lazo M.
      • Solga S.F.
      • Horska A.
      • Bonekamp S.
      • Diehl A.M.
      • Brancati F.L.
      • et al.
      Effect of a 12-month intensive lifestyle intervention on hepatic steatosis in adults with type 2 diabetes.
      ,
      • Albu J.B.
      • Heilbronn L.K.
      • Kelley D.E.
      • Smith S.R.
      • Azuma K.
      • Berk E.S.
      • et al.
      Metabolic changes following a 1-year diet and exercise intervention in patients with type 2 diabetes.
      ,
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Effect of a lifestyle intervention in patients with abnormal liver enzymes and metabolic risk factors.
      ,
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Independent effects of physical activity in patients with nonalcoholic fatty liver disease.
      ,
      • Kantartzis K.
      • Thamer C.
      • Peter A.
      • Machann J.
      • Schick F.
      • Schraml C.
      • et al.
      High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
      ,
      • Oza N.
      • Eguchi Y.
      • Mizuta T.
      • Ishibashi E.
      • Kitajima Y.
      • Horie H.
      • et al.
      A pilot trial of body weight reduction for nonalcoholic fatty liver disease with a home-based lifestyle modification intervention delivered in collaboration with interdisciplinary medical staff.
      ]; these studies provided general physical activity guidelines, but did not prescribe specific exercise protocols. Key study details are summarised in Table 3. The focus was predominantly on body weight reduction and maintenance with mean reductions of 2.2–8.8%. Only two studies reported an objective measure of physical activity adherence, specifically changes in cardiorespiratory fitness [
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Effect of a lifestyle intervention in patients with abnormal liver enzymes and metabolic risk factors.
      ,
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Independent effects of physical activity in patients with nonalcoholic fatty liver disease.
      ,
      • Kantartzis K.
      • Thamer C.
      • Peter A.
      • Machann J.
      • Schick F.
      • Schraml C.
      • et al.
      High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
      ]. One reported energy intake [
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Effect of a lifestyle intervention in patients with abnormal liver enzymes and metabolic risk factors.
      ], another specifically stated target energy intakes were achieved but did not report any details [
      • Kantartzis K.
      • Thamer C.
      • Peter A.
      • Machann J.
      • Schick F.
      • Schraml C.
      • et al.
      High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
      ], and one reported energy intake reductions but not macronutrient composition. Six of seven reported reductions in IHTAG and/or circulating liver enzyme in the intervention groups, with absolute mean reductions in IHTAG, measured by 1H-MRS, of 2–4.6% [
      • Lazo M.
      • Solga S.F.
      • Horska A.
      • Bonekamp S.
      • Diehl A.M.
      • Brancati F.L.
      • et al.
      Effect of a 12-month intensive lifestyle intervention on hepatic steatosis in adults with type 2 diabetes.
      ,
      • Kantartzis K.
      • Thamer C.
      • Peter A.
      • Machann J.
      • Schick F.
      • Schraml C.
      • et al.
      High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
      ]. Relative mean reductions assessed by biopsy [
      • Promrat K.
      • Kleiner D.E.
      • Niemeier H.M.
      • Jackvony E.
      • Kearns M.
      • Wands J.R.
      • et al.
      Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis.
      ], CT [
      • Albu J.B.
      • Heilbronn L.K.
      • Kelley D.E.
      • Smith S.R.
      • Azuma K.
      • Berk E.S.
      • et al.
      Metabolic changes following a 1-year diet and exercise intervention in patients with type 2 diabetes.
      ,
      • Oza N.
      • Eguchi Y.
      • Mizuta T.
      • Ishibashi E.
      • Kitajima Y.
      • Horie H.
      • et al.
      A pilot trial of body weight reduction for nonalcoholic fatty liver disease with a home-based lifestyle modification intervention delivered in collaboration with interdisciplinary medical staff.
      ], or 1H-MRS were 13–51%. Six out of the seven studies reported improvements in glucose control/insulin sensitivity [
      • Lazo M.
      • Solga S.F.
      • Horska A.
      • Bonekamp S.
      • Diehl A.M.
      • Brancati F.L.
      • et al.
      Effect of a 12-month intensive lifestyle intervention on hepatic steatosis in adults with type 2 diabetes.
      ,
      • Albu J.B.
      • Heilbronn L.K.
      • Kelley D.E.
      • Smith S.R.
      • Azuma K.
      • Berk E.S.
      • et al.
      Metabolic changes following a 1-year diet and exercise intervention in patients with type 2 diabetes.
      ,
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Independent effects of physical activity in patients with nonalcoholic fatty liver disease.
      ,
      • Kantartzis K.
      • Thamer C.
      • Peter A.
      • Machann J.
      • Schick F.
      • Schraml C.
      • et al.
      High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
      ,
      • Oza N.
      • Eguchi Y.
      • Mizuta T.
      • Ishibashi E.
      • Kitajima Y.
      • Horie H.
      • et al.
      A pilot trial of body weight reduction for nonalcoholic fatty liver disease with a home-based lifestyle modification intervention delivered in collaboration with interdisciplinary medical staff.
      ].
      Promrat et al. assessed histopathology, reported significant (p ⩽0.05) reductions in overall NAFLD histological activity score (NAS) and steatosis, but reductions in parenchymal inflammation and ballooning injury were not significant, and there was no mean change in fibrosis in the intervention relative to the control group or baseline [
      • Lazo M.
      • Solga S.F.
      • Horska A.
      • Bonekamp S.
      • Diehl A.M.
      • Brancati F.L.
      • et al.
      Effect of a 12-month intensive lifestyle intervention on hepatic steatosis in adults with type 2 diabetes.
      ]. Huang et al. also assessed histopathology, but reported only significant reductions in hepatitis score (p = 0.06) [
      • Vilar Gomez E.
      • Rodriguez De Miranda A.
      • Gra Oramas B.
      • Arus Soler E.
      • Llanio Navarro R.
      • Calzadilla Bertot L.
      • et al.
      Clinical trial: a nutritional supplement Viusid, in combination with diet and exercise, in patients with nonalcoholic fatty liver disease.
      ]; this study is discussed further below.
      Five studies involving 306 participants (approximately 50% women, 10 controls) prescribed specific diets and aerobic exercise programs for 3–6 months [
      • Sreenivasa Baba C.
      • Alexander G.
      • Kalyani B.
      • Pandey R.
      • Rastogi S.
      • Pandey A.
      • et al.
      Effect of exercise and dietary modification on serum aminotransferase levels in patients with nonalcoholic steatohepatitis.
      ,
      • Fraser A.
      • Abel R.
      • Lawlor D.A.
      • Fraser D.
      • Elhayany A.
      A modified Mediterranean diet is associated with the greatest reduction in alanine aminotransferase levels in obese type 2 diabetes patients: results of a quasi-randomised controlled trial.
      ,
      • Kelley D.E.
      • Kuller L.H.
      • McKolanis T.M.
      • Harper P.
      • Mancino J.
      • Kalhan S.
      Effects of moderate weight loss and orlistat on insulin resistance, regional adiposity, and fatty acids in type 2 diabetes.
      ,
      • Vilar Gomez E.
      • Rodriguez De Miranda A.
      • Gra Oramas B.
      • Arus Soler E.
      • Llanio Navarro R.
      • Calzadilla Bertot L.
      • et al.
      Clinical trial: a nutritional supplement Viusid, in combination with diet and exercise, in patients with nonalcoholic fatty liver disease.
      ,
      • Ueno T.
      • Sugawara H.
      • Sujaku K.
      • Hashimoto O.
      • Tsuji R.
      • Tamaki S.
      • et al.
      Therapeutic effects of restricted diet and exercise in obese patients with fatty liver.
      ]. Key study details are summarised in Table 4. The focus was predominantly on body weight reduction and maintenance with mean reductions of 4.2–10.6%. All studies reported reductions in direct measures of liver fat and/or liver enzymes; none used 1H-MRS. Two studies reported histological endpoints [
      • Vilar Gomez E.
      • Rodriguez De Miranda A.
      • Gra Oramas B.
      • Arus Soler E.
      • Llanio Navarro R.
      • Calzadilla Bertot L.
      • et al.
      Clinical trial: a nutritional supplement Viusid, in combination with diet and exercise, in patients with nonalcoholic fatty liver disease.
      ,
      • Ueno T.
      • Sugawara H.
      • Sujaku K.
      • Hashimoto O.
      • Tsuji R.
      • Tamaki S.
      • et al.
      Therapeutic effects of restricted diet and exercise in obese patients with fatty liver.
      ]. Villar-Gomez et al. had the single largest cohort of biopsy assessed participants of the studies reviewed (n = 30) and reported significant (p <0.5) reductions in inflammation, ballooning injury, and fibrosis, relative to baseline, following a six month intervention with a mean 10.6% weight reduction [
      • Tsai A.G.
      • Wadden T.A.
      The evolution of very-low-calorie diets: an update and meta-analysis.
      ]. Ueno et al. reported significant reductions in steatosis, but reductions in other parameters were not statistically significant [
      • Thomas D.E.
      • Elliott E.J.
      • Naughton G.A.
      Exercise for type 2 diabetes mellitus.
      ]. Relative mean reductions in IHTAG based on biopsy scores were 40–43% [
      • Vilar Gomez E.
      • Rodriguez De Miranda A.
      • Gra Oramas B.
      • Arus Soler E.
      • Llanio Navarro R.
      • Calzadilla Bertot L.
      • et al.
      Clinical trial: a nutritional supplement Viusid, in combination with diet and exercise, in patients with nonalcoholic fatty liver disease.
      ,
      • Ueno T.
      • Sugawara H.
      • Sujaku K.
      • Hashimoto O.
      • Tsuji R.
      • Tamaki S.
      • et al.
      Therapeutic effects of restricted diet and exercise in obese patients with fatty liver.
      ]. The four studies reporting glucose control/insulin sensitivity showed improvements [
      • Kelley D.E.
      • Kuller L.H.
      • McKolanis T.M.
      • Harper P.
      • Mancino J.
      • Kalhan S.
      Effects of moderate weight loss and orlistat on insulin resistance, regional adiposity, and fatty acids in type 2 diabetes.
      ,
      • Vilar Gomez E.
      • Rodriguez De Miranda A.
      • Gra Oramas B.
      • Arus Soler E.
      • Llanio Navarro R.
      • Calzadilla Bertot L.
      • et al.
      Clinical trial: a nutritional supplement Viusid, in combination with diet and exercise, in patients with nonalcoholic fatty liver disease.
      ,
      • Ueno T.
      • Sugawara H.
      • Sujaku K.
      • Hashimoto O.
      • Tsuji R.
      • Tamaki S.
      • et al.
      Therapeutic effects of restricted diet and exercise in obese patients with fatty liver.
      ].
      Additional post-hoc analyses to identify key determinants of liver fat reduction were done in three studies [
      • Promrat K.
      • Kleiner D.E.
      • Niemeier H.M.
      • Jackvony E.
      • Kearns M.
      • Wands J.R.
      • et al.
      Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis.
      ,
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Independent effects of physical activity in patients with nonalcoholic fatty liver disease.
      ,
      • Kantartzis K.
      • Thamer C.
      • Peter A.
      • Machann J.
      • Schick F.
      • Schraml C.
      • et al.
      High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
      ]. The percentage change in body weight was positively correlated with reductions in liver enzymes (r = 0.5), reductions in hepatic steatosis (r = 0.6), and overall NASH disease activity (r = 0.5) [
      • Promrat K.
      • Kleiner D.E.
      • Niemeier H.M.
      • Jackvony E.
      • Kearns M.
      • Wands J.R.
      • et al.
      Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis.
      ]. Cardiorespiratory fitness at baseline was found to be a better predictor of change in liver fat than baseline IHTAG, visceral, or total adipose tissue mass following a combined diet and physical activity intervention [
      • Kantartzis K.
      • Thamer C.
      • Peter A.
      • Machann J.
      • Schick F.
      • Schraml C.
      • et al.
      High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
      ]. Improvements in overall NASH Score, steatosis, inflammation, ballooning injury, and fibrosis were significantly (p <0.05) greater in those achieving weight reductions ⩾7% of baseline body weight compared to those with smaller reductions [
      • Lazo M.
      • Solga S.F.
      • Horska A.
      • Bonekamp S.
      • Diehl A.M.
      • Brancati F.L.
      • et al.
      Effect of a 12-month intensive lifestyle intervention on hepatic steatosis in adults with type 2 diabetes.
      ]. When dividing their group (n = 15) into responders and non-responders based on total NASH Score, Huang et al. reported statistically significantly greater weight reduction (−6.6 vs. +1.8 kg) and questionnaire reported physical activity among responders compared with non-responders [
      • Vilar Gomez E.
      • Rodriguez De Miranda A.
      • Gra Oramas B.
      • Arus Soler E.
      • Llanio Navarro R.
      • Calzadilla Bertot L.
      • et al.
      Clinical trial: a nutritional supplement Viusid, in combination with diet and exercise, in patients with nonalcoholic fatty liver disease.
      ]. Further, increased duration and frequency of physical activity was associated with increasing reductions in liver enzymes [
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Independent effects of physical activity in patients with nonalcoholic fatty liver disease.
      ]. A similar relationship was observed between changes in cardiorespiratory fitness and liver enzymes, but only when results were compared relative to baseline rather than controls [
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Independent effects of physical activity in patients with nonalcoholic fatty liver disease.
      ].

      Publication and study selection bias

      A search of clinical trial registers indicated three potentially relevant studies that may have been completed but unpublished – based on the records of recruitment status not being recently updated. All other studies either did not fit our criteria or had been published and considered for inclusion. As only two studies considered for inclusion were registered, this information provides little indication of the degree of publication bias.
      Only the Look AHEAD study had a separately published protocol, but this was not specific to the two sub-studies included herein [
      • Albu J.B.
      • Heilbronn L.K.
      • Kelley D.E.
      • Smith S.R.
      • Azuma K.
      • Berk E.S.
      • et al.
      Metabolic changes following a 1-year diet and exercise intervention in patients with type 2 diabetes.
      ,
      • St George A.
      • Bauman A.
      • Johnston A.
      • Farrell G.
      • Chey T.
      • George J.
      Effect of a lifestyle intervention in patients with abnormal liver enzymes and metabolic risk factors.
      ]. A comparison of expected outcome measures in the method section of reviewed papers and those reported in the results showed good agreement. As above, markers of adherence were consistently less well reported.

      Discussion

      The studies reviewed demonstrate that a range of lifestyle modifications are effective in reducing IHTAG and circulating liver enzymes, and improving measures of glucose control and/or insulin sensitivity in patients with NAFLD. Energy restriction, with and without increased physical activity, and weight reduction were the most frequently employed methods to reduce IHTAG. Weight reductions of 4–14% resulted in statistically significant relative reductions in IHTAG of 35–81%. The magnitude of change strongly correlated to degree of weight reduction [
      • Promrat K.
      • Kleiner D.E.
      • Niemeier H.M.
      • Jackvony E.
      • Kearns M.
      • Wands J.R.
      • et al.
      Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis.
      ]. There is also limited evidence that physical activity/exercise can lead to modest reductions in IHTAG without weight change.
      Low (800–1800 kCal/day) and very low-calorie diets (<800 kCal/day), and/or carbohydrate restriction (20–50 g/d) resulted in the most rapid reductions in body weight and IHTAG. The combination of caloric and carbohydrate restriction resulted in a ∼30% reduction in IHTAG and equally substantial improvements in glucose control and insulin sensitivity within 48 h; a time when weight reduction can only be small and largely accounted for by glycogen depletion and water loss [
      • Kirk E.
      • Reeds D.N.
      • Finck B.N.
      • Mayurranjan S.M.
      • Mayurranjan M.S.
      • Patterson B.W.
      • et al.
      Dietary fat and carbohydrates differentially alter insulin sensitivity during caloric restriction.
      ]. The pool of studies in patients with NAFLD is small and long-term follow-up is absent. However, a meta-analysis of randomised controlled trials, not specifically in NAFLD patients, reported comparable long-term (1–5 years) 5–6% body weight reduction from low and very-low calorie diets [
      • Tsai A.G.
      • Wadden T.A.
      The evolution of very-low-calorie diets: an update and meta-analysis.
      ]. Nonetheless, such diets do not constitute long-term lifestyle modification as their use is necessarily time limited and may require medical supervision [
      • Koo B.K.
      • Han K.A.
      • Ahn H.J.
      • Jung J.Y.
      • Kim H.C.
      • Min K.W.
      The effects of total energy expenditure from all levels of physical activity vs. physical activity energy expenditure from moderate-to-vigorous activity on visceral fat and insulin sensitivity in obese Type 2 diabetic women.
      ]. Use of such diets in routine clinical care remains to be tested.
      The best dietary solution for weight maintenance in patients with NAFLD remains unclear. Several of the studies reviewed did not report either actual nutrient intake or physical activity and many used indirect or low accuracy methods to assess IHTAG. Study design and limited reporting of participant adherence preclude firm conclusions being made on differential effects of diet and physical activity. Increased cardiorespiratory fitness was positively related to reductions in IHTAG [
      • Elias M.C.
      • Parise E.R.
      • Carvalho Ld.
      • Szejnfeld D.
      • Netto J.P.
      Effect of 6-month nutritional intervention on non-alcoholic fatty liver disease.
      ] and liver enzymes [
      • Albu J.B.
      • Heilbronn L.K.
      • Kelley D.E.
      • Smith S.R.
      • Azuma K.
      • Berk E.S.
      • et al.
      Metabolic changes following a 1-year diet and exercise intervention in patients with type 2 diabetes.
      ].
      Two reviewed studies reported a reduction in liver enzymes following aerobic exercise without dietary intervention or weight reduction [
      • Johnson N.A.
      • Sachinwalla T.
      • Walton D.W.
      • Smith K.
      • Armstrong A.
      • Thompson M.W.
      • et al.
      Aerobic exercise training reduces hepatic and visceral lipids in obese individuals without weight loss.
      ,
      • Sreenivasa Baba C.
      • Alexander G.
      • Kalyani B.
      • Pandey R.
      • Rastogi S.
      • Pandey A.
      • et al.
      Effect of exercise and dietary modification on serum aminotransferase levels in patients with nonalcoholic steatohepatitis.
      ], but only one study assessed IHTAG directly [
      • Johnson N.A.
      • Sachinwalla T.
      • Walton D.W.
      • Smith K.
      • Armstrong A.
      • Thompson M.W.
      • et al.
      Aerobic exercise training reduces hepatic and visceral lipids in obese individuals without weight loss.
      ]. It showed a modest 1.8% absolute reduction in IHTAG after four weeks, but no change in HOMA assessed insulin sensitivity. This lack of change may be attributable to the insensitivity of the HOMA technique or short study duration. In other conditions with metabolic dysregulation such as type 2 diabetes, physical activity and exercise have been shown to improve glucose control in [
      • Thomas D.E.
      • Elliott E.J.
      • Naughton G.A.
      Exercise for type 2 diabetes mellitus.
      ], and adiposity, in particular visceral adiposity [
      • Koo B.K.
      • Han K.A.
      • Ahn H.J.
      • Jung J.Y.
      • Kim H.C.
      • Min K.W.
      The effects of total energy expenditure from all levels of physical activity vs. physical activity energy expenditure from moderate-to-vigorous activity on visceral fat and insulin sensitivity in obese Type 2 diabetic women.
      ]. In light of the close relationship between IHTAG, glucose control, and adiposity it is likely that exercise improves liver lipid through a combination of lipid redistribution and changes in insulin sensitivity.
      Overall, the studies on lifestyle intervention in patients in NAFLD have several limitations; most notably considerable heterogeneity in the populations studied, and limited detail on adherence to specific aspects of the interventions. We therefore make the following recommendations for reporting:
      • Transparent and comprehensive descriptions of diagnostic markers of NAFLD including inclusion and exclusion criteria;
      • thorough assessment of potential contributors to NAFLD, such as nutrient excess or deficiency where these are likely to be influenced by the intervention; and
      • publication of supplementary material describing complex lifestyle interventions.

      Selection of methods

      • Quantitative assessments of hepatic steatosis;
      • quantitative assessment of NASH activity via biopsy or closely correlated activity scores such as the NAS Activity Score;
      • objective physical activity monitoring methods;
      • validated dietary assessment methods;
      • comprehensive exclusion of specific causes of steatosis, e.g. nutrient excess or deficiency; and
      • validated methods to assess alcohol consumption.
      For future research:
      • Definition of the dose response relationship for physical activity/exercise intervention and IHTAG reduction;
      • the effect of dietary macronutrient composition on IHTAG, especially carbohydrate restriction and higher protein intakes;
      • a graduated progression from Phase II to Phase III trials; and
      • a focus on how both diet and physical activity/exercise can be used to produce sustained benefit in NAFLD.

      Conclusions

      This systematic review of studies done to date provides consistent evidence that lifestyle interventions designed to reduce energy intake and/or increase physical activity reduce IHTAG and improve insulin sensitivity in patients with NAFLD. A more limited data set indicates a trend for reductions in necroinflammation. The effect on fibrosis is least consistent across studies. Degree of weight reduction is positively correlated with these improvements. However, increased physical activity and/or cardiorespiratory fitness, as well as macronutrient composition, may also act independently to prevent or reverse disease progression. Studies published to date do not allow clear differentiation of the effects of physical activity relative to diet or the importance of diet composition. This is partly due to the paucity of studies, particularly ones reporting histopathology, but also to study design limitations, minimal reporting of specific aspects of intervention adherence, and the use of variable diagnostic criteria.
      Given the clinical impact of NAFLD and the lack of therapies for its management, developing effective, reproducible lifestyle interventions is crucial. Future studies should employ accurate methods to establish the most effective means of producing a sustained reduction in liver fat, necorinflammation, and, if possible, fibrosis, and report their interventions, including objective indicators of adherence, in sufficient detail to be readily translatable to clinical practice.

      Financial support

      The research leading to these results has received funding from the European Community’s Seventh Framework Programme (FP7/2007–2013) under grant agreement n° HEALTH-F2-2009-241762 for the project FLIP. CT is supported by a Ph.D. studentship from Diabetes UK, MIT is supported by a RD Lawrence Fellowship from Diabetes UK and by the Medical Research Council (G0802536).

      References

        • Day C.P.
        Non-alcoholic fatty liver disease: current concepts and management strategies.
        Clin Med. 2006; 6: 19-25
        • Barshop N.J.
        • Sirlin C.B.
        • Schwimmer J.B.
        • Lavine J.E.
        Review article: epidemiology, pathogenesis and potential treatments of paediatric non-alcoholic fatty liver disease.
        Aliment Pharmacol Ther. 2008; 28: 13-24
        • Fabbrini E.
        • Sullivan S.
        • Klein S.
        Obesity and nonalcoholic fatty liver disease: Biochemical, metabolic, and clinical implications.
        Hepatology. 2010; 51: 679-689
        • Kotronen A.
        • Juurinen L.
        • Hakkarainen A.
        • Westerbacka J.
        • Cornér A.
        • Bergholm R.
        • et al.
        Liver fat is increased in type 2 diabetic patients and underestimated by serum alanine aminotransferase compared with equally obese nondiabetic subjects.
        Diabetes Care. 2008; 31: 165-169
        • Kotronen A.
        • Juurinen L.
        • Tiikkainen M.
        • Vehkavaara S.
        • Yki-Järvinen H.
        Increased liver fat, impaired insulin clearance, and hepatic and adipose tissue insulin resistance in type 2 diabetes.
        Gastroenterology. 2008; 135: 122-130
        • Angelico F.
        • Del Ben M.
        • Conti R.
        • Francioso S.
        • Feole K.
        • Fiorello S.
        • et al.
        Insulin resistance, the metabolic syndrome, and nonalcoholic fatty liver disease.
        J Clin Endocrinol Metab. 2005; 90: 1578-1582
        • Fraser A.
        • Harris R.
        • Sattar N.
        • Ebrahim S.
        • Davey Smith G.
        • Lawlor D.A.
        Alanine aminotransferase, gamma-glutamyltransferase, and incident diabetes: the British Women’s Heart and Health Study and meta-analysis.
        Diabetes Care. 2009; 32: 741-750
        • Targher G.
        • Day C.P.
        • Bonora E.
        Risk of cardiovascular disease in patients with nonalcoholic fatty liver disease.
        N Engl J Med. 2010; 363: 1341-1350
        • Loria P.
        • Adinolfi L.E.
        • Bellentani S.
        • Bugianesi E.
        • Grieco A.
        • Fargion S.
        • et al.
        Practice guidelines for the diagnosis and management of nonalcoholic fatty liver disease. A decalogue from the Italian Association for the Study of the Liver (AISF) Expert Committee.
        Dig Liver Dis. 2010; 42: 272-282
        • Merriman R.B.
        • Ferrell L.D.
        • Patti M.G.
        • Weston S.R.
        • Pabst M.S.
        • Aouizerat B.E.
        • et al.
        Correlation of paired liver biopsies in morbidly obese patients with suspected nonalcoholic fatty liver disease.
        Hepatology. 2006; 44: 874-880
        • Larson S.P.
        • Bowers S.P.
        • Palekar N.A.
        • Ward J.A.
        • Pulcini J.P.
        • Harrison S.A.
        Histopathologic variability between the right and left lobes of the liver in morbidly obese patients undergoing Roux-en-Y bypass.
        Clin Gastroenterol Hepatol. 2007; 5: 1329-1332
        • Schwenzer N.F.
        • Springer F.
        • Schraml C.
        • Stefan N.
        • Machann J.
        • Schick F.
        Non-invasive assessment and quantification of liver steatosis by ultrasound, computed tomography and magnetic resonance.
        J Hepatol. 2009; 51: 433-445
        • Lee S.S.
        • Park S.H.
        • Kim H.J.
        • Kim S.Y.
        • Kim M.-Y.
        • Kim D.Y.
        • et al.
        Non-invasive assessment of hepatic steatosis: prospective comparison of the accuracy of imaging examinations.
        J Hepatol. 2010; 52: 579-585
        • Szczepaniak L.S.
        • Nurenberg P.
        • Leonard D.
        • Browning J.D.
        • Reingold J.S.
        • Grundy S.
        • et al.
        Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the general population.
        Am J Physiol Endocrinol Metab. 2005; 288: E462-E468
        • Korenblat K.M.
        • Fabbrini E.
        • Mohammed B.S.
        • Klein S.
        Liver, muscle, and adipose tissue insulin action is directly related to intrahepatic triglyceride content in obese subjects.
        Gastroenterology. 2008; 134: 1369-1375
        • Strauss S.
        • Gavish E.
        • Gottlieb P.
        • Katsnelson L.
        Interobserver and intraobserver variability in the sonographic assessment of fatty liver.
        AJR Am J Roentgenol. 2007; 189: W320-W323
        • Browning J.D.
        • Szczepaniak L.S.
        • Dobbins R.
        • Nuremberg P.
        • Horton J.D.
        • Cohen J.C.
        • et al.
        Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity.
        Hepatology. 2004; 40: 1387-1395
        • Mofrad P.
        • Contos M.J.
        • Haque M.
        • Sargeant C.
        • Fisher R.A.
        • Luketic V.A.
        • et al.
        Clinical and histologic spectrum of nonalcoholic fatty liver disease associated with normal ALT values.
        Hepatology. 2003; 37: 1286-1292
        • Fracanzani A.L.
        • Valenti L.
        • Bugianesi E.
        • Andreoletti M.
        • Colli A.
        • Vanni E.
        • et al.
        Risk of severe liver disease in nonalcoholic fatty liver disease with normal aminotransferase levels: a role for insulin resistance and diabetes.
        Hepatology. 2008; 48: 792-798
        • Borai A.
        • Livingstone C.
        • Ferns G.A.A.
        The biochemical assessment of insulin resistance.
        Ann Clin Biochem. 2007; 44: 324-342
        • Larter C.Z.
        • Chitturi S.
        • Heydet D.
        • Farrell G.C.
        A fresh look at NASH pathogenesis. Part 1: The metabolic movers.
        J Gastroenterol Hepatol. 2010; 25: 672-690
        • Gunzerath L.
        • Faden V.
        • Zakhari S.
        • Warren K.
        National Institute on Alcohol Abuse and Alcoholism report on moderate drinking.
        Alcohol Clin Exp Res. 2004; 28: 829-847
        • Niemela O.
        Biomarkers in alcoholism.
        Clin Chim Acta. 2007; 377: 39-49
        • van Werven J.R.
        • Hoogduin J.M.
        • Nederveen A.J.
        • van Vliet A.A.
        • Wajs E.
        • Vandenberk P.
        • et al.
        Reproducibility of 3.0 Tesla magnetic resonance spectroscopy for measuring hepatic fat content.
        J Magn Reson Imaging. 2009; 30: 444-448
        • Elias M.C.
        • Parise E.R.
        • Carvalho Ld.
        • Szejnfeld D.
        • Netto J.P.
        Effect of 6-month nutritional intervention on non-alcoholic fatty liver disease.
        Nutrition. 2009; https://doi.org/10.1016/j.nut.2009.09.001
        • Kirk E.
        • Reeds D.N.
        • Finck B.N.
        • Mayurranjan S.M.
        • Mayurranjan M.S.
        • Patterson B.W.
        • et al.
        Dietary fat and carbohydrates differentially alter insulin sensitivity during caloric restriction.
        Gastroenterology. 2009; 136: 1552-1560
        • Yamamoto M.
        • Iwasa M.
        • Iwata K.
        • Kaito M.
        • Sugimoto R.
        • Urawa N.
        • et al.
        Restriction of dietary calories, fat and iron improves non-alcoholic fatty liver disease.
        J Gastroenterol Hepatol. 2007; 22: 498-503
        • Petersen K.F.
        • Dufour S.
        • Befroy D.
        • Lehrke M.
        • Hendler R.E.
        • Shulman G.I.
        Reversal of nonalcoholic hepatic steatosis, hepatic insulin resistance, and hyperglycemia by moderate weight reduction in patients with type 2 diabetes.
        Diabetes. 2005; 54: 603-608
        • de Luis D.A.
        • Aller R.
        • Izaola O.
        • Sagrado M.G.
        • Conde R.
        • Gonzalez J.M.
        Effect of a hypocaloric diet in transaminases in nonalcoholic fatty liver disease and obese patients, relation with insulin resistance.
        Diabetes Res Clin Pract. 2008; 79: 74-78
        • Viljanen A.P.M.
        • Iozzo P.
        • Borra R.
        • Kankaanpää M.
        • Karmi A.
        • Lautamäki R.
        • et al.
        Effect of weight loss on liver free fatty acid uptake and hepatic insulin resistance.
        J Clin Endocrinol Metab. 2009; 94: 50-55
        • Tendler D.
        • Lin S.
        • Yancy W.
        • Mavropoulos J.
        • Sylvestre P.
        • Rockey D.
        • et al.
        The effect of a low-carbohydrate, ketogenic diet on nonalcoholic fatty liver disease: a pilot study.
        Dig Dis Sci. 2007; 52: 589-593
        • Benjaminov O.
        • Beglaibter N.
        • Gindy L.
        • Spivak H.
        • Singer P.
        • Wienberg M.
        • et al.
        The effect of a low-carbohydrate diet on the nonalcoholic fatty liver in morbidly obese patients before bariatric surgery.
        Surg Endosc. 2007; 21: 1423-1427
        • Gasteyger C.
        • Larsen T.M.
        • Vercruysse F.
        • Astrup A.
        Effect of a dietary-induced weight loss on liver enzymes in obese subjects.
        Am J Clin Nutr. 2008; 87: 1141-1147
        • Johnson N.A.
        • Sachinwalla T.
        • Walton D.W.
        • Smith K.
        • Armstrong A.
        • Thompson M.W.
        • et al.
        Aerobic exercise training reduces hepatic and visceral lipids in obese individuals without weight loss.
        Hepatology. 2009; 50: 1105-1112
        • Sreenivasa Baba C.
        • Alexander G.
        • Kalyani B.
        • Pandey R.
        • Rastogi S.
        • Pandey A.
        • et al.
        Effect of exercise and dietary modification on serum aminotransferase levels in patients with nonalcoholic steatohepatitis.
        J Gastroenterol Hepatol. 2006; 21: 191-198
        • Promrat K.
        • Kleiner D.E.
        • Niemeier H.M.
        • Jackvony E.
        • Kearns M.
        • Wands J.R.
        • et al.
        Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis.
        Hepatology. 2010; 51: 121-129
        • Lazo M.
        • Solga S.F.
        • Horska A.
        • Bonekamp S.
        • Diehl A.M.
        • Brancati F.L.
        • et al.
        Effect of a 12-month intensive lifestyle intervention on hepatic steatosis in adults with type 2 diabetes.
        Diabetes Care. 2010; 33: 2156-2163
        • Albu J.B.
        • Heilbronn L.K.
        • Kelley D.E.
        • Smith S.R.
        • Azuma K.
        • Berk E.S.
        • et al.
        Metabolic changes following a 1-year diet and exercise intervention in patients with type 2 diabetes.
        Diabetes. 2010; 59: 627-633
        • St George A.
        • Bauman A.
        • Johnston A.
        • Farrell G.
        • Chey T.
        • George J.
        Effect of a lifestyle intervention in patients with abnormal liver enzymes and metabolic risk factors.
        J Gastroenterol Hepatol. 2009; 24: 399-407
        • St George A.
        • Bauman A.
        • Johnston A.
        • Farrell G.
        • Chey T.
        • George J.
        Independent effects of physical activity in patients with nonalcoholic fatty liver disease.
        Hepatology. 2009; 50: 68-76
        • Kantartzis K.
        • Thamer C.
        • Peter A.
        • Machann J.
        • Schick F.
        • Schraml C.
        • et al.
        High cardiorespiratory fitness is an independent predictor of the reduction in liver fat during a lifestyle intervention in non-alcoholic fatty liver disease.
        Gut. 2009; 58: 1281-1288
        • Oza N.
        • Eguchi Y.
        • Mizuta T.
        • Ishibashi E.
        • Kitajima Y.
        • Horie H.
        • et al.
        A pilot trial of body weight reduction for nonalcoholic fatty liver disease with a home-based lifestyle modification intervention delivered in collaboration with interdisciplinary medical staff.
        J Gastroenterol. 2009; 44: 1203-1208
        • Fraser A.
        • Abel R.
        • Lawlor D.A.
        • Fraser D.
        • Elhayany A.
        A modified Mediterranean diet is associated with the greatest reduction in alanine aminotransferase levels in obese type 2 diabetes patients: results of a quasi-randomised controlled trial.
        Diabetologia. 2008; 51: 1616-1622
        • Kelley D.E.
        • Kuller L.H.
        • McKolanis T.M.
        • Harper P.
        • Mancino J.
        • Kalhan S.
        Effects of moderate weight loss and orlistat on insulin resistance, regional adiposity, and fatty acids in type 2 diabetes.
        Diabetes Care. 2004; 27: 33-40
        • Vilar Gomez E.
        • Rodriguez De Miranda A.
        • Gra Oramas B.
        • Arus Soler E.
        • Llanio Navarro R.
        • Calzadilla Bertot L.
        • et al.
        Clinical trial: a nutritional supplement Viusid, in combination with diet and exercise, in patients with nonalcoholic fatty liver disease.
        Aliment Pharmacol Ther. 2009; 30: 999-1009
        • Ueno T.
        • Sugawara H.
        • Sujaku K.
        • Hashimoto O.
        • Tsuji R.
        • Tamaki S.
        • et al.
        Therapeutic effects of restricted diet and exercise in obese patients with fatty liver.
        J Hepatol. 1997; 27: 103-107
        • Tsai A.G.
        • Wadden T.A.
        The evolution of very-low-calorie diets: an update and meta-analysis.
        Obesity (Silver Spring). 2006; 14: 1283-1293
        • Thomas D.E.
        • Elliott E.J.
        • Naughton G.A.
        Exercise for type 2 diabetes mellitus.
        Cochrane Database Syst Rev. 2006; 3: CD002968
        • Koo B.K.
        • Han K.A.
        • Ahn H.J.
        • Jung J.Y.
        • Kim H.C.
        • Min K.W.
        The effects of total energy expenditure from all levels of physical activity vs. physical activity energy expenditure from moderate-to-vigorous activity on visceral fat and insulin sensitivity in obese Type 2 diabetic women.
        Diabet Med. 2010; 27: 1088-1092