Advertisement

A natural history study of pediatric non-alcoholic fatty liver disease over 10 years

  • Laura Draijer
    Correspondence
    Correspondence: Meibergdreef 91100 DD Amsterdam, the Netherlands +31-205662906
    Affiliations
    Department of Pediatric Gastroenterology and Nutrition, Amsterdam University Medical Centers, Location Academic Medical Center/Emma Children’s Hospital, University of Amsterdam, Amsterdam, the Netherlands

    Amsterdam Reproduction & Development Research Institute, Amsterdam University Medical Centers, Location Academic Medical Center/Emma Children’s Hospital, Amsterdam, the Netherlands

    Amsterdam UMC, University of Amsterdam, Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology Metabolism Research Institute, Amsterdam, Netherlands
    Search for articles by this author
  • Maaike Voorhoeve
    Affiliations
    Department of Pediatric Gastroenterology and Nutrition, Amsterdam University Medical Centers, Location Academic Medical Center/Emma Children’s Hospital, University of Amsterdam, Amsterdam, the Netherlands
    Search for articles by this author
  • Marian Troelstra
    Affiliations
    Department of Radiology and Nuclear Medicine, Amsterdam University Medical Centers, Location Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands
    Search for articles by this author
  • Adriaan Holleboom
    Affiliations
    Department of Vascular Medicine, Amsterdam University Medical Centers, Location Academic Medical Center, Amsterdam, The Netherlands
    Search for articles by this author
  • Ulrich Beuers
    Affiliations
    Department of Gastroenterology and Hepatology, Amsterdam University Medical Centers, Location Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands
    Search for articles by this author
  • Meeike Kusters
    Affiliations
    Department of Pediatrics, Amsterdam University Medical Centers, Location Vrije Universiteit Amsterdam, Amsterdam, the Netherlands
    Search for articles by this author
  • Aart Nederveen
    Affiliations
    Department of Radiology and Nuclear Medicine, Amsterdam University Medical Centers, Location Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands
    Search for articles by this author
  • Marc Benninga
    Affiliations
    Department of Pediatric Gastroenterology and Nutrition, Amsterdam University Medical Centers, Location Academic Medical Center/Emma Children’s Hospital, University of Amsterdam, Amsterdam, the Netherlands
    Search for articles by this author
  • Bart Koot
    Affiliations
    Department of Pediatric Gastroenterology and Nutrition, Amsterdam University Medical Centers, Location Academic Medical Center/Emma Children’s Hospital, University of Amsterdam, Amsterdam, the Netherlands
    Search for articles by this author
Open AccessPublished:January 25, 2023DOI:https://doi.org/10.1016/j.jhepr.2023.100685
      Highlights.
      • One third of the young adults who had childhood obesity develop steatosis.
      • One third of those with pediatric NAFLD have resolution of steatosis at adult age.
      • Progression of steatosis is associated with worsening of metabolic disturbances.
      • Monitoring those with NAFLD for fibrosis is important, as 6% had advanced fibrosis.

      Abstract

      Background & Aims

      The long-term outcome of pediatric non-alcoholic fatty liver disease (NAFLD) has not been well established. Between 2008 and 2012, an unselected cohort of 133 children with severe obesity was screened for NAFLD. The aim of this study was to determine the 10-year natural history of NAFLD in this cohort.

      Methods

      All 133 participants of the original study were approached. Proton Magnetic Resonance Spectroscopy (1H-MRS) and Enhanced Liver Fibrosis (ELF) test were used to assess longitudinal changes in steatosis and fibrosis, respectively. Risk factors for disease progression were explored.

      Results

      Fifty-one of the 133 subjects (38%) from the original cohort were included. The mean follow-up time was 10.3 years (range 7-13 years), 65% was female and 92% had persistent obesity. The proportion of subjects with steatosis remained unchanged (47%). Nine subjects developed steatosis and in 9 subjects steatosis resolved. Predefined relevant individual changes in 1H-MRS were seen in 38% of the subjects. The mean ELF test did not change significantly (8.70 ± 0.58 versus 8.51 ± 0.71, p=0.22). However, 16% had a relevant increase in ELF test and 6% of those with NAFLD developed advanced fibrosis at follow-up. Changes in steatosis correlated with changes in established metabolic risk factors, ALT and bariatric surgery. Change in ELF test was associated with change in triglycerides.

      Conclusions

      This 10-year follow-up study shows that one third of the young adults who had childhood obesity develop steatosis and in one third steatosis resolves. Six percent of those with NAFLD had developed advanced fibrosis at follow-up. These data underscore the importance to screen for NAFLD and monitor for progression to advanced NAFLD in the young with obesity.

      Lay Summary

      Childhood obesity accompanied by fat accumulation in the liver persists into young adulthood in the vast majority, and 6% develops serious liver injury. Worsening of metabolic disturbances increases the risk of liver injury.

      Graphical abstract

      Keywords

      Abbreviations:

      1H-MRS (proton magnetic resonance spectroscopy), ALT (alanine aminotransferase), AST (aspartate aminotransferase), BMI (body mass index), ELF (enhanced liver fibrosis test), HDL (high density lipoprotein cholesterol), HOMA-IR (homeostatic model assessment for insulin resistance), HA (hyaluronic acid), LDL (low density lipoprotein cholesterol), NAFLD (non-alcoholic fatty liver disease), NAS (NAFLD Activity Score), OSAS (Obstructive Sleep Apnea Syndrome), PIIINP (type III procollagen peptide), PNPLA3 (Patatin-like phospholipase domain-containing protein 3), ULN (upper limit of normal), TIMP-1 (tissue inhibitor of matrix metalloproteinase 1), yGT (gamma-glutanyl transferase)

      Conflict of interest

      All authors have nothing to disclose.

      Financial disclosure

      This study was funded by the Louise Vehmeijer Foundation, the Jean Louis Bernardi Foundation and by Siemens Healthineers. All funding sources had no involvement in the collection, analysis and interpretation of data; nor in the writing of the report; and had no involvement in the decision to submit the article for publication.

      Data Statement

      The research data is confidential. The authors do not have necessary permissions to share data.

      Authors contributions

      BK and LD were involved in the design and conduction of the study. LD and MV collected the data. MT performed the 1H-MRS measurements and analyses. LD analyzed the data. BK and LD were involved in data interpretation. LD and BK wrote the manuscript, LD wrote the first draft. MB supervised the study. MB, OH, UB and AN reviewed and edited the manuscript. All authors significantly contributed to the improvement of the manuscript and have approved the final content of this manuscript.

      Introduction

      Non-alcoholic fatty liver disease (NAFLD) is the most prevalent chronic liver disease in children and adults worldwide
      • Younossi Z.M.
      • Koenig A.B.
      • Abdelatif D.
      • Fazel Y.
      • Henry L.
      • Wymer M.
      Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes.
      . Its spectrum ranges from simple steatosis, to steatohepatitis (NASH), fibrosis and cirrhosis
      • Chalasani N.
      • Younossi Z.
      • Lavine J.E.
      • Charlton M.
      • Cusi K.
      • Rinella M.
      • et al.
      The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the Study of Liver Diseases.
      . It is well established that obesity and other components of the metabolic syndrome, as well as genetic variants, are the main risk factors for NAFLD
      • Patton H.M.
      • Yates K.
      • Unalp-Arida A.
      • Behling C.A.
      • Huang T.T.
      • Rosenthal P.
      • et al.
      Association between metabolic syndrome and liver histology among children with nonalcoholic Fatty liver disease.
      • Anderson E.L.
      • Howe L.D.
      • Jones H.E.
      • Higgins J.P.
      • Lawlor D.A.
      • Fraser A.
      The Prevalence of Non-Alcoholic Fatty Liver Disease in Children and Adolescents: A Systematic Review and Meta-Analysis.
      • Valenti L.
      • Alisi A.
      • Galmozzi E.
      • Bartuli A.
      • Del Menico B.
      • Alterio A.
      • et al.
      I148M patatin-like phospholipase domain-containing 3 gene variant and severity of pediatric nonalcoholic fatty liver disease.
      • Larrieta-Carrasco E.
      • León-Mimila P.
      • Villarreal-Molina T.
      • Villamil-Ramírez H.
      • Romero-Hidalgo S.
      • Jacobo-Albavera L.
      • et al.
      Association of the I148M/PNPLA3 variant with elevated alanine transaminase levels in normal-weight and overweight/obese Mexican children.
      • Romeo S.
      • Kozlitina J.
      • Xing C.
      • Pertsemlidis A.
      • Cox D.
      • Pennacchio L.A.
      • et al.
      Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease.
      . NAFLD is associated with increased morbidity and overall mortality in adults, with liver-related mortality as the third common cause of death, after malignancy and cardiovascular disease
      • Ekstedt M.
      • Hagstrom H.
      • Nasr P.
      • Fredrikson M.
      • Stal P.
      • Kechagias S.
      • et al.
      Fibrosis stage is the strongest predictor for disease-specific mortality in NAFLD after up to 33 years of follow-up.
      ,
      • Adams L.A.
      • Lymp J.F.
      • St Sauver J.
      • Sanderson S.O.
      • Lindor K.D.
      • Feldstein A.
      • et al.
      The natural history of nonalcoholic fatty liver disease: a population-based cohort study.
      . Those with pediatric age onset NAFLD might be particularly at risk of these complications, as they have a longer exposure to NAFLD and its metabolic disturbances during their lifetime.
      It has been suggested that NAFLD in children progresses more rapidly. This assumption is mainly based on case reports and case-series reporting progression of NAFLD to cirrhosis in childhood
      • Molleston J.P.
      • White F.
      • Teckman J.
      • Fitzgerald J.F.
      Obese children with steatohepatitis can develop cirrhosis in childhood.
      ,
      • Kinugasa A.
      • Tsunamoto K.
      • Furukawa N.
      • Sawada T.
      • Kusunoki T.
      • Shimada N.
      Fatty liver and its fibrous changes found in simple obesity of children.
      and young adulthood
      • Feldstein A.E.
      • Charatcharoenwitthaya P.
      • Treeprasertsuk S.
      • Benson J.T.
      • Enders F.B.
      • Angulo P.
      The natural history of non-alcoholic fatty liver disease in children: a follow-up study for up to 20 years.
      ,
      • Suzuki D.
      • Hashimoto E.
      • Kaneda K.
      • Tokushige K.
      • Shiratori K.
      Liver failure caused by non-alcoholic steatohepatitis in an obese young male.
      . However, these highly selective cases do not inform us on the risk of progression to liver fibrosis for children with NAFLD in the general population. Due to this lack of long-term studies with unselected cohorts of children followed into adulthood, the natural history of pediatric NAFLD and predictors for a severe disease course have not been established. This could help to risk stratify patients, to tailor screening and initiate early and intensified treatment in those at risk of severe NAFLD, in order to reduce morbidity and mortality in these patients.
      Between 2008 and 2012, an unselected cohort of 133 adolescents with severe obesity was screened for NAFLD in our center. The aim of the present study is to assess changes in steatosis and fibrosis in this cohort after a period of 10 years and to explore risk factors for disease progression.

      Patients and methods

      Study design

      This was a prospective long-term follow-up study. The study design and in- and exclusion criteria of the original study have been previously published
      • Koot B.G.
      • van der Baan-Slootweg O.H.
      • Tamminga-Smeulders C.L.
      • Rijcken T.H.
      • Korevaar J.C.
      • van Aalderen W.M.
      • et al.
      Lifestyle intervention for non-alcoholic fatty liver disease: prospective cohort study of its efficacy and factors related to improvement.
      . In short, 133 adolescents with obesity were screened for NAFLD prior to inclusion into a lifestyle intervention program using Proton Magnetic Resonance Spectroscopy (1H-MRS) to detect steatosis and the Enhanced Liver Fibrosis (ELF) test as a proxy to detect fibrosis. All participants subsequently underwent a combined lifestyle intervention for six months, as described previously
      • Koot B.G.
      • van der Baan-Slootweg O.H.
      • Tamminga-Smeulders C.L.
      • Rijcken T.H.
      • Korevaar J.C.
      • van Aalderen W.M.
      • et al.
      Lifestyle intervention for non-alcoholic fatty liver disease: prospective cohort study of its efficacy and factors related to improvement.
      . All participants of the original study were eligible for participation in the present follow-up study and were contacted by email or telephone. All measurements in the present study were performed during one single visit. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Medical Ethics Committee of the Amsterdam University Medical Centers of the University of Amsterdam. Written informed consent was obtained from all participants.

      Clinical assessment

      Physical examination included measurement of weight, height, BMI, waist circumference and blood pressure. Waist circumference was measured at the midpoint between the last floating rib and the top of the iliac crest
      • Nishida C.
      • Ko G.T.
      • Kumanyika S.
      Body fat distribution and noncommunicable diseases in populations: overview of the 2008 WHO Expert Consultation on Waist Circumference and Waist-Hip Ratio.
      . Reference values for waist circumference were based on a study performed in Dutch children aged 2-21 years
      • Fredriks A.M.
      • van Buuren S.
      • Fekkes M.
      • Verloove-Vanhorick S.P.
      • Wit J.M.
      Are age references for waist circumference, hip circumference and waist-hip ratio in Dutch children useful in clinical practice?.
      . International age- and gender-specific cutoffs for BMI were used to define obesity and morbid obesity, comparable to an adult BMI of 30 kg/m2 and 40 kg/m2, respectively
      • Cole T.J.
      • Bellizzi M.C.
      • Flegal K.M.
      • Dietz W.H.
      Establishing a standard definition for child overweight and obesity worldwide: international survey.
      . The age- and gender-specific BMI standard deviation (SD), the so called BMI z-score, was calculated

      calculator Tbz-s. https://tnochildhealthstatistics.shinyapps.io/JGZRichtlijnLengtegroei/.

      . Presence of the metabolic syndrome was assessed at follow-up and was defined according to the National Cholesterol Education Program Adult Treatment Panel III
      Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III).
      . Obstructive Sleep Apnea Syndrome (OSAS) was suspected when at least two of the four following criteria were met: occurrence of daytime sleepiness, loud snoring, morning headache, witnessed breathing interruptions or awakenings due to gasping or choking.

      Laboratory tests

      Venous blood was sampled after an overnight fast to determine alanine aminotransferase (ALT), aspartate aminotransferase (AST), gamma-glutamyl transferase (yGT), high density lipoprotein (HDL) cholesterol, low density lipoprotein (LDL) cholesterol and triglycerides using standard laboratory methods. Insulin and glucose were used to calculate the homeostatic model assessment of insulin resistance score (HOMA-IR)
      • Matthews D.R.
      • Hosker J.P.
      • Rudenski A.S.
      • Naylor B.A.
      • Treacher D.F.
      • Turner R.C.
      Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man.
      . In adolescents, insulin resistance was defined as HOMA-IR >2.8 for boys and >3.4 for girls
      • Shashaj B.
      • Luciano R.
      • Contoli B.
      • Morino G.S.
      • Spreghini M.R.
      • Rustico C.
      • et al.
      Reference ranges of HOMA-IR in normal-weight and obese young Caucasians.
      . In adults, insulin resistance was defined as HOMA-IR > 2.7. Dyslipidemia was defined as either elevated LDL, total cholesterol, triglycerides or lowered HDL, according to age- and sex-specific cutoffs

      Kist-van Holthe JE vME, van den Akker ELT et al. NVK Richtlijn basisdiagnostiek cardiovasculair risico bij kinderen met obesitas en behandeling van hypertensie. NVK, 2016.

      . The upper limit of normal (ULN) for ALT in teenage girls and boys was defined as 22 IU/L and 26 IU/L, respectively
      • Schwimmer J.B.
      • Dunn W.
      • Norman G.J.
      • Pardee P.E.
      • Middleton M.S.
      • Kerkar N.
      • et al.
      SAFETY study: alanine aminotransferase cutoff values are set too high for reliable detection of pediatric chronic liver disease.
      . The ULN in adult women and men was defined as 22 IU/L and 29 IU/L, respectively
      • Ruhl C.E.
      • Everhart J.E.
      Upper limits of normal for alanine aminotransferase activity in the United States population.
      . Hepatitis B and C, Wilson disease and auto-immune hepatitis were excluded using the appropriate tests. The ELF test was performed in subjects with steatosis at baseline and in those who developed steatosis during the follow-up period. To calculate the ELF test, hyaluronic acid (HA), type III procollagen peptide (PIIINP), and (TIMP-1) were measured in serum samples by immune assay on Atellica® analyzer (Siemens Diagnostics, Erlangen, Germany). The ELF test was calculated by the following algorithm: ELF test = 2.278 + .851 ln (concentration HA) + .751 ln (concentration PIIINP) + .394 ln (concentration TIMP-1). Advanced fibrosis was defined as ELF test >9.8
      • Day J.
      • Patel P.
      • Parkes J.
      • Rosenberg W.
      Derivation and Performance of Standardized Enhanced Liver Fibrosis (ELF) Test Thresholds for the Detection and Prognosis of Liver Fibrosis.
      ,
      • Lichtinghagen R.
      • Pietsch D.
      • Bantel H.
      • Manns M.P.
      • Brand K.
      • Bahr M.J.
      The Enhanced Liver Fibrosis (ELF) score: normal values, influence factors and proposed cut-off values.
      . At an individual level, a relevant change in ELF test was defined as an increase or decrease of ≥0.5 unit
      • Gawrieh S.
      • Wilson L.A.
      • Yates K.P.
      • Cummings O.W.
      • Vilar-Gomez E.
      • Ajmera V.
      • et al.
      Relationship of ELF and PIIINP With Liver Histology and Response to Vitamin E or Pioglitazone in the PIVENS Trial.
      . The single nucleotide polymorphism (SNP) in the rs738409 region of the patatin-like phospholipase domain-containing protein-3 (‘adiponutrin’) gene (PNPLA3) was determined in all subjects, wherein the CC genotype is the wildtype (Supplementary CTAT Table).
      Proton Magnetic Resonance Spectroscopy (1H-MRS)Subjects were examined in fasting state using an Ingenia 3 Tesla MR scanner (Philips, Best, The Ne therlands) at both baseline and follow-up. A 20 × 20 × 20 mm voxel was positioned in the right liver lobe in segment VI or VII. Spectra were acquired by using a point resolved spectroscopy sequence (echo time msec/repetition time msec = 35/2000, 64 acquisitions). The liver 1H-MRS spectra containing the water and fat resonance peaks were fitted with the AMARES algorithm of the jMRUI software version 4.0
      • Naressi A.
      • Couturier C.
      • Devos J.M.
      • Janssen M.
      • Mangeat C.
      • de Beer R.
      • et al.
      Java-based graphical user interface for the MRUI quantitation package.
      . The relative fat content was expressed as a ratio of the fat peak area over the cumulative water and fat peak areas. Peak areas were corrected for T2 relaxation
      • de Bazelaire C.M.
      • Duhamel G.D.
      • Rofsky N.M.
      • Alsop D.C.
      MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results.
      . MRI technicians were blinded to clinical data and to 1H-MRS results at baseline. The presence of steatosis was defined as a liver fat percentage of > 1.8% on 1H-MRS, which has been validated to correspond with >5% steatosis on histologic analysis for the 1H-MRS-setting in our center
      • van Werven J.R.
      • Marsman H.A.
      • Nederveen A.J.
      • Smits N.J.
      • ten Kate F.J.
      • van Gulik T.M.
      • et al.
      Assessment of hepatic steatosis in patients undergoing liver resection: comparison of US, CT, T1-weighted dual-echo MR imaging, and point-resolved 1H MR spectroscopy.
      . In subjects with steatosis at baseline, resolution of steatosis was defined as 1H-MRS <1.8% at follow-up. At an individual level an absolute change of >5% in 1H-MRS was considered relevant as it approximates a change in steatosis grade, similar to a >25% change in histological steatosis
      • van Werven J.R.
      • Marsman H.A.
      • Nederveen A.J.
      • Smits N.J.
      • ten Kate F.J.
      • van Gulik T.M.
      • et al.
      Assessment of hepatic steatosis in patients undergoing liver resection: comparison of US, CT, T1-weighted dual-echo MR imaging, and point-resolved 1H MR spectroscopy.
      ,
      • Bohte A.E.
      • Koot B.G.
      • van der Baan-Slootweg O.H.
      • van Werven J.R.
      • Bipat S.
      • Nederveen A.J.
      • et al.
      US Cannot Be Used to Predict the Presence or Severity of Hepatic Steatosis in Severely Obese Adolescents.
      .

      Statistical analyses

      Standard descriptive methods were used to report patient characteristics at baseline and follow-up. To determine any selection bias, baseline characteristics and response to the lifestyle intervention at baseline of the included subjects were compared to those of the whole original cohort using a student t-test (normal distribution) or Mann Whitney U test (no normal distribution) for continuous variables and a Chi square test for nominal variables. Changes between baseline and follow-up were tested with a paired Student-t-test or a Wilcoxon rank test for continues variables and with a McNemar test for categorical data. The significance level was set at p<0.05. In the group of subjects with NAFLD at baseline and subjects who had developed NAFLD at follow-up, univariate linear regression analyses were used to identify clinical and biochemical parameters that were associated with change in 1H-MRS and ELF test. Given the small sample size and explorative character of this analysis, co-variables with a p<0.10 were reported. All statistical analyses were performed using SPSS version 26.0 (IBM Corporation, Chicago, IL).

      Results

      Study population and clinical outcomes

      In total, 51 of the 133 subjects (38%) from the original cohort were included in the present study. We were unable to reach 18 subjects or his/her relatives due to invalid contact details and 63 subjects refused to participate, most often due to a lack of interest or a lack of time. The mean follow-up time was 10.3 years (range 7-13 years) and 65% were female. At follow-up, mean BMI was 40.06 kg/m2, 92% had obesity and the median serum ALT was 21 IU/L. Compared to the whole original cohort, the 51 included subjects had a significantly higher HOMA-IR (4.2 vs 3.4, p=0.02), serum ALT level (35 vs 28 IU/L, p=0.01) and a higher prevalence of steatosis (47% vs 31%, p=0.04) at baseline. Age, gender, BMI and waist circumference were not significantly different. Resolution of 1H-MRS determined steatosis after the 6-months lifestyle intervention was not different in the included subjects compared the whole original cohort (33% and 32%, respectively). Table 1 shows the longitudinal changes in anthropometry, biochemistry, steatosis and ELF test in those with (n=24) and without (n=27) NAFLD at baseline. Sixteen subjects (31%), almost equally distributed among both groups, underwent bariatric surgery during the follow-up period. The mean time interval between surgery and follow-up measurements was 3.9 ± 2.3 years. Twenty-eight subjects (55%) received further lifestyle interventions guided by a health professional at some point during the follow-up period. Only six subjects were still receiving lifestyle support by a professional at follow-up. Fifty-nine percent of the subjects did some form of exercise at follow-up, with a mean frequency of 3 hours per week, most of them without a professional trainer.
      Table 1Longitudinal change in clinical, biochemical and liver parameters.
      NAFLD at baseline (n=24)No NAFLD at baseline (n=27)
      BaselineFollow-upp-valueBaselineFollow-upp-value
      Female, n (%)12 (50)--21 (78)--
      Age, years14.3 ± 2.224.6 ± 3.0<0.00013.8 ± 2.324.0 ± 2.8<0.000
      BMI z-score4.16 (4.01-4.48)4.05 (2.90-4.59)0.413.79 (3.62-4.16)4.13 (3.58-4.45)0.87
      BMI (kg/m2)39.95 ± 4.8140.48 ± 8.280.8137.31 ± 5.6840.01 ± 6.660.08
      BMI > 30 kg/m2, n (%)21 (88)23 (96)0.3924 (89)24 (89)1.00
      Waist circumference (cm)109 ± 9121 ± 200.01101 ± 10113 ± 160.00
      Alcohol use, n (%)NR15 (63)-NR14 (52)-
      Units alcohol per monthNR7 ± 6-NR15 ± 13-
      Smoking1 (4)10 (42)0.001 (4)9 (33)0.01
      PackyearsNR4.1 ± 3.7-NR1.3 ± 1.2-
      OSAS, n (%)7 (29)6 (25)0.7110 (37)3 (11)0.04
      Bariatric surgery, n (%)0 (0)9 (38)-0 (0)7 (26)-
      Ethnicity, n (%)
      • -
        Dutch
      14 (58)--22 (81)--
      • -
        African
      1 (4)--3 (11)--
      • -
        Turkisch
      6 (25)--1 (4)--
      • -
        Other
      3 (13)--1 (4)--
      PNPLA3 GG, n (%)9 (37)--5 (19)--
      PNPLA3 CG, n (%)5 (21)--4 (15)--
      PNPLA3 CC, n (%)10 (42)--17 (66)--
      Biochemical (n=47)
      ALT, IU/L36 (22-49)26 (15-64)0.4027 (16-36)21 (15-24)0.04
      ALT >ULN, n (%)15 (68)13 (54)0.2617 (64)7 (28)0.01
      yGT, IU/L26 (18-36)23 (13-65)0.2918 (16-23)18 (14-26)0.61
      Triglycerides, mmol/L0.89 (0.74-1.35)0.98 (0.77-1.59)0.510.69 (0.49-1.02)0.61 (0.47-1.04)0.82
      LDL, mmol/L2.63 (2.19-2.91)2.75 (2.10-3.17)0.452.39 (1.91-2.79)2.48 (2.25-3.21)0.17
      HDL, mmol/L1.05 (0.94-1.29)1.17 (1.04-1.44)0.041.02 (0.89-1.23)1.24 (1.14-1.44)<0.000
      HOMA-IR4.6 (2.6-6.7)2.3 (1.3-3.7)0.0072.3 (1.4-3.5)1.9 (1.1-3.1)0.18
      Insulin resistance, n (%)16 (67)12 (50)0.107 (26)9 (33)0.76
      Steatosis (n=51)
      1H-MRS (%)4.68 (2.76-9.48)3.47 (1.52-5.84)0.240.83 (0.52-1.25)1.40 (0.94-2.12)0.01
      Δ 1H-MRS >5%, n (%)-9 (37.5)--0 (0)-
      Steatosis, n (%)24 (100)15 (62.5)0.030 (0)9 (33)0.01
      ELF test (n=32)*
      ELF test8.72 ± 0.518.52 ± 0.720.238.62 ± 0.798.45 ± 0.660.70
      ELF test ≥ 9.8, n(%)1 (4)1 (4)1.0001 (4)<0.000
      Δ ELF > 0.5, n (%)-13 (54)--5 (56)-
      Δ ELF > 1.0, n (%)-5 (21)--2 (22)-
      Data are presented as mean ± SD or as median (IQR), as appropriate. Differences between baseline and follow-up were tested with a paired Student-t-test or a Wilcoxon rank test for continues variables and with a McNemar test for categorical data. The significance level was set at p<0.05. * ELF test was performed in subjects with steatosis at baseline (n=24) and in those who developed steatosis (n=8).Abbreviations: 1H-MRS, proton magnetic resonance spectroscopy; ALT, alanine aminotransferase; BMI, body mass index; ELF, enhanced liver fibrosis test; HDL, high density lipoprotein cholesterol; HOMA-IR, homeostatic model assessment for insulin resistance; LDL, low density lipoprotein cholesterol; NR, not reported; PNPLA3, Patatin-like phospholipase domain-containing protein 3; yGT, gamma-glutanyl transferase

      Change in anthropometry and biochemistry

      As shown in Table 1, most anthropometric and biochemical parameters did not change significantly at follow-up in both those with and without NAFLD at baseline. Most importantly, the BMI z-score showed no change over time. The observed increase in waist circumference and decrease in HOMA-IR are in line with physiological changes after adolescence
      • Fredriks A.M.
      • van Buuren S.
      • Fekkes M.
      • Verloove-Vanhorick S.P.
      • Wit J.M.
      Are age references for waist circumference, hip circumference and waist-hip ratio in Dutch children useful in clinical practice?.
      . Indeed no change in the prevalence of insulin resistance was found based on age-specific HOMA-IR cut-offs. Three subjects, all with NAFLD at baseline, developed type 2 diabetes mellitus. Only in those without NAFLD at baseline, serum ALT decreased significantly. An increase in serum HDL cholesterol was seen in both groups. Dyslipidemia was found in 70% at follow-up compared to 35% at baseline (p=0.001). A history compatible with OSAS was found in 18% of the subjects at follow-up, compared to 33% at baseline (p=0.06).

      Change in 1H-MRS determined liver steatosis

      As shown in Table 1, the proportion of subjects with steatosis in the total cohort at the 10-year follow-up compared to baseline remained unchanged: 24/51 (47%), as NAFLD resolved in 9 subjects and 9 subjects developed NAFLD. The mean absolute change in 1H-MRS was -1.53 ± 6.84% in those with NAFLD at baseline (p=0.24) and 0.79 ± 1.38% in those without NAFLD at baseline (p=0.01). Figure 1a and 1b show the individual variation in change in 1H-MRS in those with and without NAFLD at baseline, respectively. In those with NAFLD at baseline, six (25%) subjects showed a decrease in 1H-MRS of >5%, and three (13%) showed an increase in 1H-MRS of >5%. In those without NAFLD at baseline, none had a change in 1H-MRS of >5%.
      Figure thumbnail gr1
      Figure 1A: Individual change in 1H-MRS in those with NAFLD at baseline .Variation in change in 1H-MRS from baseline to follow-up, in those with NAFLD at baseline (n=24). B: Individual change in 1H-MRS in those without NAFLD at baseline. Variation in change in 1H-MRS from baseline to follow-up, in those without NAFLD at baseline (n=27).
      The subjects that underwent bariatric surgery had a significantly larger improvement in 1H-MRS (-0.94 vs 0.74, p=0.01) compared to subjects without bariatric surgery. This was accompanied by a larger improvement in BMI z-score (-0.64 vs 0.21, p=0.02), triglycerides (-0.16 vs 0.11, p=0.04), HOMA-IR (-2.6 vs 0.0, p=0.001) and yGT (14 vs 24 IU/L, p=0.003).
      PredictorsUnivariate linear regression analyses were performed in 33 subjects consisting of those with NAFLD at baseline (n=24) and those who developed NAFLD (n=9). In the assessment of baseline predictors, a higher BMI z-score and ALT level at baseline were associated with a decrease in 1H-MRS (Supplemental Table S1). The PNPLA3 variant was not associated with a change in 1H-MRS. No multivariate analysis was performed due to multicollinearity.
      In the assessment of longitudinal changes using univariate linear regression analysis, an increase in BMI z-score, triglycerides, LDL-cholesterol and ALT were associated with an increase in 1H-MRS (Table 2). Bariatric surgery was associated with a 4.5% decrease in 1H-MRS.
      Table 2Longitudinal associations between change in 1H-MRS and risk factors.
      Change in parametersp-valueB (95% CI)R2
      Δ BMI z-score (per 1.0 SD)0.013.0 (0.8 to 5.2)0.21
      Δ waist circumference (per cm)0.470.1 (-0.1 to 0.2)0.02
      Δ triglycerides* (per mmol/L)0.092.5 (-0.4 to 5.5)0.10
      Δ HDL (per mmol/L)0.24-4.5 (-12.1 to 3.2)0.05
      Δ LDL (per mmol/L)<0.0006.1 (3.3 to 9.0)0.41
      Δ HOMA-IR (per 1.0 point)0.260.4 (-0.3 to 1.1)0.05
      Δ ALT (per U/L)<0.0000.1 (0.1 to 0.2)0.46
      Δ yGT (per U/L)0.42-0.03 (-0.1 to 0.1)0.02
      Bariatric surgery0.05-4.5 (-9.0 to -0.1)0.12
      Correlations between Δ 1H-MRS and Δ risk factors in those with NAFLD at baseline or at follow-up (n=33). Correlations were assessed using univariate linear regression analyses. The significance level was set at p<0.10. A positive B-value indicates an increase in 1H-MRS. A negative B-value indicates a decrease in 1H-MRS. Abbreviations: 1H-MRS, proton magnetic resonance spectroscopy; ALT, alanine aminotransferase; BMI, body mass index; HDL, high density lipoprotein cholesterol; HOMA-IR, homeostatic model assessment for insulin resistance; LDL, low density lipoprotein cholesterol; PNPLA3, Patatin-like phospholipase domain-containing protein 3; yGT, gamma-glutanyl transferase

      Change in ELF test

      ELF test at baseline and follow-up was available in 32 subjects consisting of all 24 subjects with NAFLD at baseline and in 8 out of 9 subjects who developed NAFLD. The mean ELF test did not change in the 10-year follow-up period: 8.70 ± 0.58 at baseline versus 8.51 ± 0.71 at follow-up (p=0.22). Figure 2a and 2b show the individual variation in change in ELF test in those with NAFLD at baseline and in those who developed NAFLD, respectively. In those with NAFLD at baseline, an increase of at least 0.5 unit in ELF test was seen in 4 (17%) subjects and a drop of at least 0.5 unit was seen in 9 (38%) subjects (Table 1, Figure 2a). In those who developed NAFLD, an increase of at least 0.5 unit in ELF test was seen in 1 (13%) subject and a decrease of at least 0.5 units in ELF test was seen in 4 (50%) subjects (Table 1, Figure 2b). Only one subject had an ELF test ≥9.8 at baseline. At follow-up, 2 out of 32 subjects (6%) had an ELF test ≥9.8 (one with and one without NAFLD at follow-up).
      Figure thumbnail gr2
      Figure 2A: Individual change in ELF test in those with NAFLD at baseline.
      Variation in change in ELF test from baseline to follow-up, in subjects with NAFLD at baseline (n=24).
      B: Individual change in ELF test in those who developed NAFLD.
      Variation in change in ELF test from baseline to follow-up, in subjects who developed NAFLD (n=8).
      PredictorsUsing univariate linear regression analysis, none of the baseline metabolic and genetic characteristics were associated with a change in ELF test (Supplemental Table S2). In the assessment of longitudinal changes using univariate linear regression analysis, an increase in triglycerides was the only factor associated with an increase in ELF test (Table 3). No multivariate analysis was performed due to multicollinearity.
      Table 3Longitudinal associations between change in ELF test and risk factors.
      p-valueB (95% CI)R2
      Δ BMI z-score (per 1.0 SD)0.31-0.2 (-0.6 to 0.2)0.04
      Δ Waist circumference (per cm)0.780.003 (-0.02 to 0.03)0.003
      Δ HOMA-IR (per point)0.750.02 (-0.1 to 0.1)0.004
      Δ Triglycerides (per mmol/L)0.070.4 (-0.04 to 0.8)0.11
      Δ HDL (per mmol/L)0.830.1 (-1.0 to 1.3)0.002
      Δ LDL (per mmol/L)0.27-0.3 (-0.8 to 0.2)0.04
      Δ ALT (per U/L)0.210.01 (-0.004 to 0.02)0.06
      Δ yGT (per U/L)0.190.01 (-0.004 to 0.02)0.06
      Bariatric surgery0.82-0.1 (-0.2 to 0.8)0.002
      Correlations between Δ ELF test and Δ risk factors in those with NAFLD at baseline or at follow-up (n=32). Correlations were assessed using univariate linear regression analyses. The significance level was set at p<0.10. A positive B-value indicates an increase in ELF. A negative B-value indicates a decrease in ELF. Abbreviations: 1H-MRS, proton magnetic resonance spectroscopy; ALT, alanine aminotransferase; BMI, body mass index; ELF, enhanced liver fibrosis test; HDL, high density lipoprotein; HOMA-IR, homeostatic model assessment for insulin resistance; LDL, low density lipoprotein; PNPLA3, Patatin-like phospholipase domain-containing protein 3; yGT, gamma-glutanyl transferase.

      Discussion

      To our knowledge, this is the first longitudinal study assessing the long-term natural history of NAFLD in a non-selected cohort of adolescents with obesity. We showed that after 7 to 13 years, one third of the participants had resolution of steatosis at follow-up, and one third of the subjects newly developed NAFLD. Relevant changes in steatosis were seen in over one third of the subjects (38%) with worsening in 13%. These changes correlated with longitudinal changes in metabolic risk factors, ALT and bariatric surgery. Metabolic worsening was not uncommon as 3 patients developed type 2 diabetes and 35% developed dyslipidemia. At group level the ELF test remained unchanged during follow-up. However, pre-defined relevant changes (>0.5 unit) were seen in 55%, with relevant worsening in 16%. Six percent of the subjects with NAFLD developed advanced fibrosis, defined as ELF test >9.8.
      Progression of fibrosis is a more important predictor of patient outcomes than progression of steatosis
      • Ekstedt M.
      • Hagstrom H.
      • Nasr P.
      • Fredrikson M.
      • Stal P.
      • Kechagias S.
      • et al.
      Fibrosis stage is the strongest predictor for disease-specific mortality in NAFLD after up to 33 years of follow-up.
      ,
      • Angulo P.
      • Kleiner D.E.
      • Dam-Larsen S.
      • Adams L.A.
      • Bjornsson E.S.
      • Charatcharoenwitthaya P.
      • et al.
      Liver Fibrosis, but No Other Histologic Features, Is Associated With Long-term Outcomes of Patients With Nonalcoholic Fatty Liver Disease.
      . In adults the ELF test has proven to be able to predict liver-related outcomes
      • Parkes J.
      • Roderick P.
      • Harris S.
      • Day C.
      • Mutimer D.
      • Collier J.
      • et al.
      Enhanced liver fibrosis test can predict clinical outcomes in patients with chronic liver disease.
      and reflect changes in fibrosis in response to treatment
      • Harrison S.A.
      • Rossi S.J.
      • Paredes A.H.
      • Trotter J.F.
      • Bashir M.R.
      • Guy C.D.
      • et al.
      NGM282 Improves Liver Fibrosis and Histology in 12 Weeks in Patients With Nonalcoholic Steatohepatitis.
      ,
      • Sanyal A.J.
      • Anstee Q.M.
      • Trauner M.
      • Lawitz E.J.
      • Abdelmalek M.F.
      • Ding D.
      • et al.
      Cirrhosis Regression is Associated with Improved Clinical Outcomes in Patients with Nonalcoholic Steatohepatitis.
      . In children, the ELF test is among the best validated markers for fibrosis
      • Draijer L.G.
      • van Oosterhout J.P.M.
      • Vali Y.
      • Zwetsloot S.
      • van der Lee J.H.
      • van Etten-Jamaludin F.S.
      • et al.
      Diagnostic accuracy of fibrosis tests in children with non-alcoholic fatty liver disease: A systematic review.
      . Although the mean ELF test did not change at group level over time, the individual changes in ELF test of >0.5 unit in over half of the participants should be considered relevant as it was recently shown that a 0.3 unit decrease in ELF test was significantly associated with overall NAFLD Activity Score (NAS) based histological improvement and NASH resolution in children, and similar findings were observed in in adults
      • Gawrieh S.
      • Wilson L.A.
      • Yates K.P.
      • Cummings O.W.
      • Vilar-Gomez E.
      • Ajmera V.
      • et al.
      Relationship of ELF and PIIINP With Liver Histology and Response to Vitamin E or Pioglitazone in the PIVENS Trial.
      ,
      • Gawrieh S.
      • Harlow K.E.
      • Pike F.
      • Yates K.P.
      • Wilson L.A.
      • Cummings O.W.
      • et al.
      Relationship of Enhanced Liver Fibrosis Score with Pediatric Nonalcoholic Fatty Liver Disease Histology and Response to Vitamin E or Metformin.
      . Six percent developed advanced fibrosis over a 10 year time span in our study. A meta-analysis of paired biopsy studies in adults showed that the average rate of fibrosis progression in adults with steatosis is 14 years per fibrosis stage and 7 years in those with NASH
      • Singh S.
      • Allen A.M.
      • Wang Z.
      • Prokop L.J.
      • Murad M.H.
      • Loomba R.
      Fibrosis progression in nonalcoholic fatty liver vs nonalcoholic steatohepatitis: a systematic review and meta-analysis of paired-biopsy studies.
      . Although a comparison with these data is cumbersome, our study does not suggest a more rapid progression in the young.
      Our findings are contrary to the faster fibrosis progression rate reported in previous short-term pediatric studies in more selected populations
      • Lavine J.E.
      • Schwimmer J.B.
      • Van Natta M.L.
      • Molleston J.P.
      • Murray K.F.
      • Rosenthal P.
      • et al.
      Effect of vitamin E or metformin for treatment of nonalcoholic fatty liver disease in children and adolescents: the TONIC randomized controlled trial.
      • Schwimmer J.B.
      • Lavine J.E.
      • Wilson L.A.
      • Neuschwander-Tetri B.A.
      • Xanthakos S.A.
      • Kohli R.
      • et al.
      Children With Nonalcoholic Fatty Liver Disease, Cysteamine Bitartrate Delayed Release Improves Liver Enzymes but Does Not Reduce Disease Activity Scores.
      • Cioffi C.E.
      • Welsh J.A.
      • Cleeton R.L.
      • Caltharp S.A.
      • Romero R.
      • Wulkan M.L.
      • et al.
      Natural History of NAFLD Diagnosed in Childhood: A Single-Center Study.
      . Paired liver biopsies from 122 adolescents with a mean interval period of 1.6 years in those in the placebo arms from two pediatric trials showed that 23% progressed in fibrosis stage
      • Xanthakos S.A.
      • Lavine J.E.
      • Yates K.P.
      • Schwimmer J.B.
      • Molleston J.P.
      • Rosenthal P.
      • et al.
      Progression of Fatty Liver Disease in Children Receiving Standard of Care Lifestyle Advice.
      . Selection bias in this study, inherent to biopsy studies in children, is reflected by the severe steatosis (51% with S3 steatosis), as well as the high prevalence of NASH (31%). Similar to our findings, they found a bidirectional evolution of NAFLD, as steatosis resolved in 2.4%, improved in 43% and fibrosis improved at least one stage in 34%. This bidirectional evolution is also found in adults and probably reflects the waxing and waning of disease activity
      • Gawrieh S.
      • Wilson L.A.
      • Yates K.P.
      • Cummings O.W.
      • Vilar-Gomez E.
      • Ajmera V.
      • et al.
      Relationship of ELF and PIIINP With Liver Histology and Response to Vitamin E or Pioglitazone in the PIVENS Trial.
      . Previous retrospective case series of children with biopsy proven NAFLD from tertiary liver centers have suggested an even more worrying natural history with progression of fibrosis in 40-80% at follow-up biopsies and even cases of liver transplantation and death
      • Feldstein A.E.
      • Charatcharoenwitthaya P.
      • Treeprasertsuk S.
      • Benson J.T.
      • Enders F.B.
      • Angulo P.
      The natural history of non-alcoholic fatty liver disease in children: a follow-up study for up to 20 years.
      ,
      • A-Kader H
      • Henderson J.
      • Vanhoesen K.
      • Ghishan F.
      • Bhattacharyya A.
      Nonalcoholic fatty liver disease in children: a single center experience.
      . Although these retrospective studies provide important data showing that NAFLD in children can be progressive, these results are not representative for the general NAFLD population. The relevance of pediatric NAFLD was recently also underscored by a Swedish nation-wide study that showed that young adults with NAFLD had a significantly higher liver-related mortality (HR 16.5) and a 5.9 fold higher rate of mortality than controls, with higher rates in those with NASH or fibrosis
      • Simon T.G.
      • Roelstraete B.
      • Hartjes K.
      • Shah U.
      • Khalili H.
      • Arnell H.
      • et al.
      Non-alcoholic fatty liver disease in children and young adults is associated with increased long-term mortality.
      . Again, the generalizability of these findings is limited by selection bias, as only patients that qualified for liver biopsy were included.
      These findings on the natural history of pediatric NAFLD underscore the importance to regularly monitor for development and progression of NAFLD in this age group. Current expert opinion based guidelines advice to monitor children with obesity for the development of NAFLD every 2-3 years and to monitor those with NAFLD yearly for progression to fibrosis, while in adults monitoring every 2-3 years for progression to advanced NAFLD is advised
      • Vos M.B.
      • Abrams S.H.
      • Barlow S.E.
      • Caprio S.
      • Daniels S.R.
      • Kohli R.
      • et al.
      NASPGHAN Clinical Practice Guideline for the Diagnosis and Treatment of Nonalcoholic Fatty Liver Disease in Children.
      ,
      EASL-EASD-EASO Clinical Practice Guidelines for the management of non-alcoholic fatty liver disease.
      . Our data suggest that screening yearly in adolescents with NAFLD without fibrosis every 2-3 years might also be appropriate. However, larger longitudinal and cost-effectivity studies are needed to determine the optimal follow-up frequency in this age group.
      Previous cross-sectional studies consistently showed that metabolic risk factors are strongly associated with pediatric NAFLD
      • Xanthakos S.A.
      • Lavine J.E.
      • Yates K.P.
      • Schwimmer J.B.
      • Molleston J.P.
      • Rosenthal P.
      • et al.
      Progression of Fatty Liver Disease in Children Receiving Standard of Care Lifestyle Advice.
      ,
      • Schwimmer J.B.
      • Deutsch R.
      • Rauch J.B.
      • Behling C.
      • Newbury R.
      • Lavine J.E.
      Obesity, insulin resistance, and other clinicopathological correlates of pediatric nonalcoholic fatty liver disease.
      ,
      • Peng L.
      • Wu S.
      • Zhou N.
      • Zhu S.
      • Liu Q.
      • Li X.
      Clinical characteristics and risk factors of nonalcoholic fatty liver disease in children with obesity.
      . In our study, these risk factors were also longitudinally associated with change in steatosis; i.e. dyslipidemia and body composition, and were in line with findings from longitudinal studies in adults
      • 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.
      . This suggests that worsening in metabolic parameters should indeed guide the intensity of screening for development and progression of NAFLD as current guidelines also indicate
      • Vos M.B.
      • Abrams S.H.
      • Barlow S.E.
      • Caprio S.
      • Daniels S.R.
      • Kohli R.
      • et al.
      NASPGHAN Clinical Practice Guideline for the Diagnosis and Treatment of Nonalcoholic Fatty Liver Disease in Children.
      ,
      EASL-EASD-EASO Clinical Practice Guidelines for the management of non-alcoholic fatty liver disease.
      . It is notable that important metabolic worsening was not uncommon as 3 subjects still developed type 2 diabetes and 35% developed dyslipidemia at follow-up. One third of the participants in the present study underwent bariatric surgery. In our study, it was associated with long term improvement in BMI, biochemistry and steatosis, in correspondence with its efficacy reported in adults
      • Fakhry T.K.
      • Mhaskar R.
      • Schwitalla T.
      • Muradova E.
      • Gonzalvo J.P.
      • Murr M.M.
      Bariatric surgery improves nonalcoholic fatty liver disease: a contemporary systematic review and meta-analysis.
      . The lack of association with change in ELF test is most possibly explained by the overall low ELF test at baseline in this cohort.
      No longitudinal association between the PNPLA3 polymorphism and change in steatosis or fibrosis was found. Cross-sectional studies showed that the PNPLA3 C>G polymorphism increases the susceptibility to the whole NAFLD spectrum in children and adults
      • Sookoian S.
      • Pirola C.J.
      Meta-analysis of the influence of I148M variant of patatin-like phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity of nonalcoholic fatty liver disease.
      • Tang S.
      • Zhang J.
      • Mei T.T.
      • Guo H.Q.
      • Wei X.H.
      • Zhang W.Y.
      • et al.
      Association of PNPLA3 rs738409 G/C gene polymorphism with nonalcoholic fatty liver disease in children: a meta-analysis.
      • Dai G.
      • Liu P.
      • Li X.
      • Zhou X.
      • He S.
      Association between PNPLA3 rs738409 polymorphism and nonalcoholic fatty liver disease (NAFLD) susceptibility and severity: A meta-analysis.
      . Besides the relatively small sample size, known gene-environment interactions for PNPLA3 variants, in particular carbohydrate intake, might have obscured this correlation
      • Smagris E.
      • BasuRay S.
      • Li J.
      • Huang Y.
      • Lai K.M.
      • Gromada J.
      • et al.
      Pnpla3I148M knockin mice accumulate PNPLA3 on lipid droplets and develop hepatic steatosis.
      ,
      • Davis J.N.
      • Lê K.A.
      • Walker R.W.
      • Vikman S.
      • Spruijt-Metz D.
      • Weigensberg M.J.
      • et al.
      Increased hepatic fat in overweight Hispanic youth influenced by interaction between genetic variation in PNPLA3 and high dietary carbohydrate and sugar consumption.
      . Dietary intake was not recorded in our study.
      The strengths of this study are the long follow-up period (7-13 years) and the evaluation of a cohort of adolescents with obesity wherein NAFLD was assessed through screening. Despite selection through screening, a selection bias in those who participated in follow-up cannot be excluded. The significantly higher HOMA-IR, ALT levels and steatosis prevalence in the included subjects in comparison to the whole original cohort in our study, might point to a selection bias towards those more severely affected at baseline. The high rate of bariatric surgery in our cohort and the fact that many subjects received other lifestyle interventions during the follow-up period, could also suggest a selection bias toward those motivated for treatment. In addition, the lifestyle intervention at the start of follow-up that all children participated in, might have attenuated the progression of NAFLD. However, as most children with severe obesity undergo lifestyle interventions and in young adults bariatric surgery is increasingly performed, we still believe that the findings in this cohort represent the natural history of NAFLD at a population level more accurately than previous pediatric studies. The participation rate of only 38% is a limitation inherent to most long-term follow-up studies. Due to the consequently relatively small sample size, we could not thoroughly analyse the risk factors related to change in steatosis and fibrosis, and these findings need to be validated in a larger cohort. Finally, as the participants were adolescents at inclusion, the changes in metabolic parameters over time can also reflect natural changes after adolescence. This is particularly true for insulin resistance. Also for ELF test some age- and sex-dependent changes have been reported
      • Thusing I.
      • Antonsen S.
      • Hoffmann-Petersen B.S.
      • Rasmussen H.M.
      • Wittenhagen P.
      Can the Enhanced Liver Fibrosis score be used to Diagnose Children with Liver Fibrosis?.
      . More studies are needed in adolescents and young adults to better validate reference values, its longitudinal use and prognostic value.
      In conclusion, this 10-year follow-up study shows that one third of the young adults who had childhood obesity develop NAFLD and one third has resolution of steatosis. The changes in steatosis correlated with changes in established metabolic risk factors. At a group level no significant change in ELF test was observed, but 16% had a relevant increase in ELF test and 6% of those with NAFLD had developed advanced fibrosis at follow-up. These data underscore the importance to monitor for development of NAFLD and progression to advanced NAFLD in the young with obesity.

      Acknowledgements

      We would like to thank the Louise Vehmeijer Foundation, Jean Louis Bernardi Foundation and Siemens Healthineers for their financial support for this study. We also thank Sandra van der Berg-Faay, Annemiek Heijboer and Aldo Grefhorst for their effort in the 1H-MRS, ELF test and genetic analyses, respectively. Lastly, we thank the participants of this study.

      Appendix A. Supplementary data

      The following is/are the supplementary data to this article:

      References

        • Younossi Z.M.
        • Koenig A.B.
        • Abdelatif D.
        • Fazel Y.
        • Henry L.
        • Wymer M.
        Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes.
        Hepatology. 2016; 64: 73-84
        • Chalasani N.
        • Younossi Z.
        • Lavine J.E.
        • Charlton M.
        • Cusi K.
        • Rinella M.
        • et al.
        The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the Study of Liver Diseases.
        Hepatology. 2018; 67: 328-357
        • Patton H.M.
        • Yates K.
        • Unalp-Arida A.
        • Behling C.A.
        • Huang T.T.
        • Rosenthal P.
        • et al.
        Association between metabolic syndrome and liver histology among children with nonalcoholic Fatty liver disease.
        Am J Gastroenterol. 2010; 105: 2093-2102
        • Anderson E.L.
        • Howe L.D.
        • Jones H.E.
        • Higgins J.P.
        • Lawlor D.A.
        • Fraser A.
        The Prevalence of Non-Alcoholic Fatty Liver Disease in Children and Adolescents: A Systematic Review and Meta-Analysis.
        PLoS One. 2015; 10: e0140908
        • Valenti L.
        • Alisi A.
        • Galmozzi E.
        • Bartuli A.
        • Del Menico B.
        • Alterio A.
        • et al.
        I148M patatin-like phospholipase domain-containing 3 gene variant and severity of pediatric nonalcoholic fatty liver disease.
        Hepatology. 2010; 52: 1274-1280
        • Larrieta-Carrasco E.
        • León-Mimila P.
        • Villarreal-Molina T.
        • Villamil-Ramírez H.
        • Romero-Hidalgo S.
        • Jacobo-Albavera L.
        • et al.
        Association of the I148M/PNPLA3 variant with elevated alanine transaminase levels in normal-weight and overweight/obese Mexican children.
        Gene. 2013; 520: 185-188
        • Romeo S.
        • Kozlitina J.
        • Xing C.
        • Pertsemlidis A.
        • Cox D.
        • Pennacchio L.A.
        • et al.
        Genetic variation in PNPLA3 confers susceptibility to nonalcoholic fatty liver disease.
        Nat Genet. 2008; 40: 1461-1465
        • Ekstedt M.
        • Hagstrom H.
        • Nasr P.
        • Fredrikson M.
        • Stal P.
        • Kechagias S.
        • et al.
        Fibrosis stage is the strongest predictor for disease-specific mortality in NAFLD after up to 33 years of follow-up.
        Hepatology. 2015; 61: 1547-1554
        • Adams L.A.
        • Lymp J.F.
        • St Sauver J.
        • Sanderson S.O.
        • Lindor K.D.
        • Feldstein A.
        • et al.
        The natural history of nonalcoholic fatty liver disease: a population-based cohort study.
        Gastroenterology. 2005; 129: 113-121
        • Molleston J.P.
        • White F.
        • Teckman J.
        • Fitzgerald J.F.
        Obese children with steatohepatitis can develop cirrhosis in childhood.
        Am J Gastroenterol. 2002; 97: 2460-2462
        • Kinugasa A.
        • Tsunamoto K.
        • Furukawa N.
        • Sawada T.
        • Kusunoki T.
        • Shimada N.
        Fatty liver and its fibrous changes found in simple obesity of children.
        J Pediatr Gastroenterol Nutr. 1984; 3: 408-414
        • Feldstein A.E.
        • Charatcharoenwitthaya P.
        • Treeprasertsuk S.
        • Benson J.T.
        • Enders F.B.
        • Angulo P.
        The natural history of non-alcoholic fatty liver disease in children: a follow-up study for up to 20 years.
        Gut. 2009; 58: 1538-1544
        • Suzuki D.
        • Hashimoto E.
        • Kaneda K.
        • Tokushige K.
        • Shiratori K.
        Liver failure caused by non-alcoholic steatohepatitis in an obese young male.
        J Gastroenterol Hepatol. 2005; 20: 327-329
        • Koot B.G.
        • van der Baan-Slootweg O.H.
        • Tamminga-Smeulders C.L.
        • Rijcken T.H.
        • Korevaar J.C.
        • van Aalderen W.M.
        • et al.
        Lifestyle intervention for non-alcoholic fatty liver disease: prospective cohort study of its efficacy and factors related to improvement.
        Arch Dis Child. 2011; 96: 669-674
        • Nishida C.
        • Ko G.T.
        • Kumanyika S.
        Body fat distribution and noncommunicable diseases in populations: overview of the 2008 WHO Expert Consultation on Waist Circumference and Waist-Hip Ratio.
        Eur J Clin Nutr. 2010; 64: 2-5
        • Fredriks A.M.
        • van Buuren S.
        • Fekkes M.
        • Verloove-Vanhorick S.P.
        • Wit J.M.
        Are age references for waist circumference, hip circumference and waist-hip ratio in Dutch children useful in clinical practice?.
        Eur J Pediatr. 2005; 164: 216-222
        • Cole T.J.
        • Bellizzi M.C.
        • Flegal K.M.
        • Dietz W.H.
        Establishing a standard definition for child overweight and obesity worldwide: international survey.
        BMJ. 2000; 320: 1240-1243
      1. calculator Tbz-s. https://tnochildhealthstatistics.shinyapps.io/JGZRichtlijnLengtegroei/.

      2. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III).
        Jama. 2001; 285: 2486-2497
        • Matthews D.R.
        • Hosker J.P.
        • Rudenski A.S.
        • Naylor B.A.
        • Treacher D.F.
        • Turner R.C.
        Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man.
        Diabetologia. 1985; 28: 412-419
        • Shashaj B.
        • Luciano R.
        • Contoli B.
        • Morino G.S.
        • Spreghini M.R.
        • Rustico C.
        • et al.
        Reference ranges of HOMA-IR in normal-weight and obese young Caucasians.
        Acta Diabetol. 2016; 53: 251-260
      3. Kist-van Holthe JE vME, van den Akker ELT et al. NVK Richtlijn basisdiagnostiek cardiovasculair risico bij kinderen met obesitas en behandeling van hypertensie. NVK, 2016.

        • Schwimmer J.B.
        • Dunn W.
        • Norman G.J.
        • Pardee P.E.
        • Middleton M.S.
        • Kerkar N.
        • et al.
        SAFETY study: alanine aminotransferase cutoff values are set too high for reliable detection of pediatric chronic liver disease.
        Gastroenterology. 2010; 138 (1364.e1-2): 1357-1364
        • Ruhl C.E.
        • Everhart J.E.
        Upper limits of normal for alanine aminotransferase activity in the United States population.
        Hepatology. 2012; 55: 447-454
        • Day J.
        • Patel P.
        • Parkes J.
        • Rosenberg W.
        Derivation and Performance of Standardized Enhanced Liver Fibrosis (ELF) Test Thresholds for the Detection and Prognosis of Liver Fibrosis.
        J Appl Lab Med. 2019; 3: 815-826
        • Lichtinghagen R.
        • Pietsch D.
        • Bantel H.
        • Manns M.P.
        • Brand K.
        • Bahr M.J.
        The Enhanced Liver Fibrosis (ELF) score: normal values, influence factors and proposed cut-off values.
        J Hepatol. 2013; 59: 236-242
        • Gawrieh S.
        • Wilson L.A.
        • Yates K.P.
        • Cummings O.W.
        • Vilar-Gomez E.
        • Ajmera V.
        • et al.
        Relationship of ELF and PIIINP With Liver Histology and Response to Vitamin E or Pioglitazone in the PIVENS Trial.
        Hepatol Commun. 2021; 5: 786-797
        • Naressi A.
        • Couturier C.
        • Devos J.M.
        • Janssen M.
        • Mangeat C.
        • de Beer R.
        • et al.
        Java-based graphical user interface for the MRUI quantitation package.
        Magma. 2001; 12: 141-152
        • de Bazelaire C.M.
        • Duhamel G.D.
        • Rofsky N.M.
        • Alsop D.C.
        MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results.
        Radiology. 2004; 230: 652-659
        • van Werven J.R.
        • Marsman H.A.
        • Nederveen A.J.
        • Smits N.J.
        • ten Kate F.J.
        • van Gulik T.M.
        • et al.
        Assessment of hepatic steatosis in patients undergoing liver resection: comparison of US, CT, T1-weighted dual-echo MR imaging, and point-resolved 1H MR spectroscopy.
        Radiology. 2010; 256: 159-168
        • Bohte A.E.
        • Koot B.G.
        • van der Baan-Slootweg O.H.
        • van Werven J.R.
        • Bipat S.
        • Nederveen A.J.
        • et al.
        US Cannot Be Used to Predict the Presence or Severity of Hepatic Steatosis in Severely Obese Adolescents.
        Radiology. 2011;
        • Angulo P.
        • Kleiner D.E.
        • Dam-Larsen S.
        • Adams L.A.
        • Bjornsson E.S.
        • Charatcharoenwitthaya P.
        • et al.
        Liver Fibrosis, but No Other Histologic Features, Is Associated With Long-term Outcomes of Patients With Nonalcoholic Fatty Liver Disease.
        Gastroenterology. 2015; 149 (97.e10): 389
        • Parkes J.
        • Roderick P.
        • Harris S.
        • Day C.
        • Mutimer D.
        • Collier J.
        • et al.
        Enhanced liver fibrosis test can predict clinical outcomes in patients with chronic liver disease.
        Gut. 2010; 59: 1245-1251
        • Harrison S.A.
        • Rossi S.J.
        • Paredes A.H.
        • Trotter J.F.
        • Bashir M.R.
        • Guy C.D.
        • et al.
        NGM282 Improves Liver Fibrosis and Histology in 12 Weeks in Patients With Nonalcoholic Steatohepatitis.
        Hepatology. 2020; 71: 1198-1212
        • Sanyal A.J.
        • Anstee Q.M.
        • Trauner M.
        • Lawitz E.J.
        • Abdelmalek M.F.
        • Ding D.
        • et al.
        Cirrhosis Regression is Associated with Improved Clinical Outcomes in Patients with Nonalcoholic Steatohepatitis.
        Hepatology. 2021;
        • Draijer L.G.
        • van Oosterhout J.P.M.
        • Vali Y.
        • Zwetsloot S.
        • van der Lee J.H.
        • van Etten-Jamaludin F.S.
        • et al.
        Diagnostic accuracy of fibrosis tests in children with non-alcoholic fatty liver disease: A systematic review.
        Liver Int. 2021; 41: 2087-2100
        • Gawrieh S.
        • Harlow K.E.
        • Pike F.
        • Yates K.P.
        • Wilson L.A.
        • Cummings O.W.
        • et al.
        Relationship of Enhanced Liver Fibrosis Score with Pediatric Nonalcoholic Fatty Liver Disease Histology and Response to Vitamin E or Metformin.
        J Pediatr. 2021; 239: 161-167.e5
        • Singh S.
        • Allen A.M.
        • Wang Z.
        • Prokop L.J.
        • Murad M.H.
        • Loomba R.
        Fibrosis progression in nonalcoholic fatty liver vs nonalcoholic steatohepatitis: a systematic review and meta-analysis of paired-biopsy studies.
        Clinical gastroenterology and hepatology : the official clinical practice journal of the American Gastroenterological Association. 2015; 13 (e40): 643
        • Lavine J.E.
        • Schwimmer J.B.
        • Van Natta M.L.
        • Molleston J.P.
        • Murray K.F.
        • Rosenthal P.
        • et al.
        Effect of vitamin E or metformin for treatment of nonalcoholic fatty liver disease in children and adolescents: the TONIC randomized controlled trial.
        Jama. 2011; 305: 1659-1668
        • Schwimmer J.B.
        • Lavine J.E.
        • Wilson L.A.
        • Neuschwander-Tetri B.A.
        • Xanthakos S.A.
        • Kohli R.
        • et al.
        Children With Nonalcoholic Fatty Liver Disease, Cysteamine Bitartrate Delayed Release Improves Liver Enzymes but Does Not Reduce Disease Activity Scores.
        Gastroenterology. 2016; 151 (e9): 1141-1154
        • Cioffi C.E.
        • Welsh J.A.
        • Cleeton R.L.
        • Caltharp S.A.
        • Romero R.
        • Wulkan M.L.
        • et al.
        Natural History of NAFLD Diagnosed in Childhood: A Single-Center Study.
        Children (Basel). 2017; 4
        • Xanthakos S.A.
        • Lavine J.E.
        • Yates K.P.
        • Schwimmer J.B.
        • Molleston J.P.
        • Rosenthal P.
        • et al.
        Progression of Fatty Liver Disease in Children Receiving Standard of Care Lifestyle Advice.
        Gastroenterology. 2020; 159 (e10): 1731-1751
        • A-Kader H
        • Henderson J.
        • Vanhoesen K.
        • Ghishan F.
        • Bhattacharyya A.
        Nonalcoholic fatty liver disease in children: a single center experience.
        Clin Gastroenterol Hepatol. 2008; 6: 799-802
        • Simon T.G.
        • Roelstraete B.
        • Hartjes K.
        • Shah U.
        • Khalili H.
        • Arnell H.
        • et al.
        Non-alcoholic fatty liver disease in children and young adults is associated with increased long-term mortality.
        J Hepatol. 2021; 75: 1034-1041
        • Vos M.B.
        • Abrams S.H.
        • Barlow S.E.
        • Caprio S.
        • Daniels S.R.
        • Kohli R.
        • et al.
        NASPGHAN Clinical Practice Guideline for the Diagnosis and Treatment of Nonalcoholic Fatty Liver Disease in Children.
        J Pediatr Gastroenterol Nutr. 2016;
      4. EASL-EASD-EASO Clinical Practice Guidelines for the management of non-alcoholic fatty liver disease.
        J Hepatol. 2016; 64: 1388-1402
        • Schwimmer J.B.
        • Deutsch R.
        • Rauch J.B.
        • Behling C.
        • Newbury R.
        • Lavine J.E.
        Obesity, insulin resistance, and other clinicopathological correlates of pediatric nonalcoholic fatty liver disease.
        The Journal of Pediatrics. 2003; 143: 500-505
        • Peng L.
        • Wu S.
        • Zhou N.
        • Zhu S.
        • Liu Q.
        • Li X.
        Clinical characteristics and risk factors of nonalcoholic fatty liver disease in children with obesity.
        BMC Pediatr. 2021; 21: 122
        • 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
        • Fakhry T.K.
        • Mhaskar R.
        • Schwitalla T.
        • Muradova E.
        • Gonzalvo J.P.
        • Murr M.M.
        Bariatric surgery improves nonalcoholic fatty liver disease: a contemporary systematic review and meta-analysis.
        Surg Obes Relat Dis. 2019; 15: 502-511
        • Sookoian S.
        • Pirola C.J.
        Meta-analysis of the influence of I148M variant of patatin-like phospholipase domain containing 3 gene (PNPLA3) on the susceptibility and histological severity of nonalcoholic fatty liver disease.
        Hepatology. 2011; 53: 1883-1894
        • Tang S.
        • Zhang J.
        • Mei T.T.
        • Guo H.Q.
        • Wei X.H.
        • Zhang W.Y.
        • et al.
        Association of PNPLA3 rs738409 G/C gene polymorphism with nonalcoholic fatty liver disease in children: a meta-analysis.
        BMC Med Genet. 2020; 21: 163
        • Dai G.
        • Liu P.
        • Li X.
        • Zhou X.
        • He S.
        Association between PNPLA3 rs738409 polymorphism and nonalcoholic fatty liver disease (NAFLD) susceptibility and severity: A meta-analysis.
        Medicine (Baltimore). 2019; 98e14324
        • Smagris E.
        • BasuRay S.
        • Li J.
        • Huang Y.
        • Lai K.M.
        • Gromada J.
        • et al.
        Pnpla3I148M knockin mice accumulate PNPLA3 on lipid droplets and develop hepatic steatosis.
        Hepatology. 2015; 61: 108-118
        • Davis J.N.
        • Lê K.A.
        • Walker R.W.
        • Vikman S.
        • Spruijt-Metz D.
        • Weigensberg M.J.
        • et al.
        Increased hepatic fat in overweight Hispanic youth influenced by interaction between genetic variation in PNPLA3 and high dietary carbohydrate and sugar consumption.
        Am J Clin Nutr. 2010; 92: 1522-1527
        • Thusing I.
        • Antonsen S.
        • Hoffmann-Petersen B.S.
        • Rasmussen H.M.
        • Wittenhagen P.
        Can the Enhanced Liver Fibrosis score be used to Diagnose Children with Liver Fibrosis?.
        Journal of Pediatric Gastroenterology and Nutrition. 2022;