Arch Dis Child 96:1003-1007 doi:10.1136/adc.2010.190975
  • Original articles

Abnormal liver function in children with metabolic syndrome from a UK-based obesity clinic

  1. Julian P H Shield1
  1. 1Department of Paediatric Endocrinology and Diabetes, Bristol Royal Hospital for Children, Bristol, UK
  2. 2Institute of Child Health & Life, School of Clinical Sciences, University of Bristol, Bristol, UK
  1. Correspondence to Professor Julian P H Shield, Professor in Diabetes and Metabolic Endocrinology, Department of Paediatric Endocrinology, Bristol Children's Hospital, Upper Maudlin Street, Bristol BS2 8BJ, UK; j.p.h.shield{at}
  • Accepted 9 October 2010
  • Published Online First 19 November 2010


Aim To examine transaminitis in obese children, its association with glucose metabolism and the metabolic syndrome and the response to weight loss through lifestyle modification.

Methods 216 children (90 male), aged 2.9–17.6 (median 12.4 years) with median body mass index (BMI) SD score (SDS) of 3.36 (range 1.92–6.22) attending a hospital obesity clinic in Bristol (UK) underwent an oral glucose tolerance test (OGTT) with fasting lipid and liver profile. Auxological measures included weight, height, waist circumference, percentage body fat. Parental history of type 2 diabetes (T2DM) was recorded. The 2007 International Diabetes Federation definition of metabolic syndrome was used. 90 children undergoing a trial of lifestyle modification to improve weight were re-assessed at 12 months.

Results 34/216 (16%) children had raised alanine aminotransferase (ALT) (>40 IU/l) with greater prevalence in boys (23% vs 10%, p=0.01) and in those with a parental history of T2DM (30% vs 13.2% p=0.019). Patients with transaminitis were more likely to fulfil the criteria for metabolic syndrome (p<0.001) and have subtle abnormalities in glucose metabolism during an OGTT with elevated glucose levels at 90 (p=0.041) and 120 (p=0.014) min and a greater glucose area-under-the-curve (p=0.014). Improvement in BMI SDS over 1 year correlated with improvement in ALT levels (τ = −0.29, p<0.001).

Conclusion Prevalence of transaminitis is significant in obese children and is associated with components of the metabolic syndrome. Lifestyle-based improvement in BMI SDS offers an effective tool for correcting transaminitis and should remain the central component of therapy.


Non-alcoholic fatty liver disease (NAFLD), an obesity associated morbidity, is a liver disease characterised by histopathological changes without clinical manifestation of liver disease. Previously, only described in obese adults, NAFLD is increasingly being diagnosed in children having been first described in the 1980s.1 The prevalence in childhood is likely to rise in parallel with the increasing prevalence of obesity. It is now believed to be the commonest cause of unexplained, biochemical liver abnormalities in children.2

NAFLD is asymptomatic and probably under diagnosed by clinicians. It is often found incidentally with abnormalities of liver function tests or hepatomegaly. The sensitivity of non-invasive imaging such as ultrasound of the liver is felt to be inadequate for its diagnosis,3 and liver biopsy remains the diagnostic gold standard. However, given the high prevalence of childhood obesity, it is impractical to biopsy every obese patient with abnormal liver biochemistry. At present, there is no consensus as to how these patients should be managed and by whom.

What is already known on the topic

  • The prevalence of obesity in the UK childhood population is high

  • Transaminitis with NAFLD is increasingly being recognised as an association with obesity and the metabolic syndrome

  • Various treatment strategies have been advocated for improving transaminitis

What this study adds

  • The prevalence of transaminitis is high (16%) in a hospital-based cohort of obese children

  • Transaminitis appears to be intimately associated with the metabolic syndrome and dysregulation of glucose metabolism

  • Modest weight improvement of around 0.3 of a body mass index SD score over 1 year is associated with resolution of transaminitis

The natural history and progression of the disease in patients is not fully understood. The spectrum of the disease is broad ranging from benign fatty liver through non-alcoholic steatohepatitis (NASH) to cirrhosis and liver failure in adults. Up to one third of the cases of NAFLD may develop into NASH when there is increased risk of cirrhosis, hepatocellular carcinoma and death from end stage liver failure.4 Worryingly, liver cirrhosis has been reported in children as young as 10 years of age.5

NAFLD is linked to adiposity, especially central obesity, and is commonly associated with features of the metabolic syndrome such as hyperlipidaemia and insulin resistance. Abnormal liver function tests in the form of elevated alanine aminotransferase (ALT) levels have been demonstrated in 10–25% of obese paediatric patients.6 The association between abnormal liver function manifest as elevated ALT in children with type 2 diabetes (T2DM)7 and acanthosis nigricans (a surrogate stigmata of hyperinsulinaemia)6 in patients with NAFLD suggests that hyperinsulinism plays a key role in the aetiology. Treatments aimed at improving insulin sensitivity have been shown to be effective in ameliorating NAFLD. Profound weight loss postbariatric sugery8 and metformin9 increase insulin sensitivity and have been shown to reduce hepatic fat as well as improve liver biochemistry. NAFLD should therefore perhaps be regarded as a feature of the metabolic syndrome and there is a cogent argument for abnormalities in liver biochemistry being incorporated into its definition.1 2

The aetiology of NAFLD in obese patients is multifactorial.10 Increased portal concentrations of free fatty acids (FFAs) are found in patients with obesity and insulin resistance. These increased levels may cause damage to intracellular membranes through lipid peroxidation and injury to mitochondria and reduced β-oxidation of hepatic FFA. Adipocytokines such as adiponectin and, tumour necrosis factor α may also be involved in the development of NAFLD. Oxidative stress is important as products of lipid peroxidation inhibit electron transfer along the respiratory chain, increasing reactive species of oxidation. For example, an increase in iron levels has been identified in patients with NASH and insulin resistant status. Relatively low levels of vitamin E antioxidant levels in obese children have also been described.10

Transaminitis in obese children11,,16 and adults17,,19 has been associated with features of the metabolic syndrome such as obesity, hyperinsulinism and hyperlipiademaia16,,19. In addition, high body mass index (BMI) in patients with transaminitis has been shown to be an independent predictor of fibrosis in the progression of NAFLD.20 Using transaminitis as a surrogate maker of NAFLD, this study examined the prevalence of abnormal liver function tests in obese children and adolescents in a UK, secondary care obesity clinic and the response to weight loss with lifestyle modification alone. In addition, the study examined the associations between transaminitis and components of the metabolic syndrome to examine its role as an additional marker of cardiovascular risk and/or abnormal glucose homeostasis.

Subjects and methods

A cross-sectional descriptive study was undertaken involving 216 (90 male) obese patients under 18 years referred to the Obesity Clinic at the Bristol Royal Hospital for Children from 1997 to 2007. Obesity was defined as a BMI over the 95th centile according to UK BMI centile charts for children. Median age (range) of the group was 12.4 years (12.9–17.6) and median BMI SD score (SDS) was 3.36 (range 1.92–6.22). None of the participants had pre-exiting past medical conditions associated with liver disease or required medication that may have influenced their liver function.

Each participant underwent a standard oral glucose tolerance test (OGTT) with glucose load of 1.75 g/kg (max 75 g) after an overnight fast as per clinic protocol. Glucose levels were reported at 0, 30, 60, 90, 120 min and insulin levels at 0, 30, 120 min. The American Diabetes Association Criteria for the diagnosis of impaired glucose tolerance (IGT) and diabetes was adopted.21 Insulin resistance was calculated by homeostasis model assessment–insulin resistance calculated by fasting insulin (mIU/l) × fasting glucose level (mmol/l)/22.5. A fasting lipid profile and liver function tests (ALT) were also taken at baseline. Transaminitis was defined as an ALT of over 40 IU/l.

Auxological measurements included weight (Seca Scales in light clothing, no shoes), height (Harpenden stadiometer to the nearest 0.1 cm) and maximum waist circumference to the nearest 0.1 cm; percentage body fat by Bioimpendance (Tanita BC 418 MA BIA analyser) as well as blood pressure (BP) (using the random zero sphygmomanometer with BP repeated three times in a sitting position with 5 min rest in between each and the mean of the last two reported in a subgroup of 90 patients and the rest by the Dinamap Critikon 8100 with appropriate sized cuff). BMI was calculated by weight (kg)/height2 (m2). Parental history of T2DM was also recorded. BMI and waist circumference were all converted to a SDS using the British 1990 growth reference data from the Child Growth Foundation.22 Percentage body fat was converted to SDS from age and sex specific centiles by McCarthy et al.23

A subgroup of 90 patients completed a randomised trial in which they received either a 1-year programme based on behavioural modification of dietary and exercise habits delivered within a multidisciplinary clinic24 or a similar framework with the addition of a novel, computerised device to gradually slow speed of eating and reduce portion size.25 As both arms essentially received similar lifestyle modification therapy aimed solely at improving eating and physical activity behaviours, we have used this group to explore the impact of weight improvement (BMI SDS) on transaminitis over the 12-month period.26 The therapeutic study was approved by United Bristol Hospitals Health Care Trust Ethics Committee (E5472) and registered at NCT00407420.

Relationships between risk factors of the metabolic syndrome and abnormal ALT were analysed statistically using χ2 tests (for categorical variables) and Student t tests or Mann–Whitney U test where variable was not normally distributed (to compare the means of two groups). For these, the statistical software package SPSS versus 14 was used (SPSS, SPSS Headquarters, Chicago, Illinois, USA, 2008). Profiles of mean glucose concentrations during the OGTTs for the groups of children with and without raised ALT were compared using a repeated measures analysis of variance (ANOVA), the main effects being group and time point (0, 30, 60, 90, 120 min). This model was fitted using the ‘mixed model’ facility in SAS (SAS release 9.1, 2002–2003, SAS Institute, Cary, USA), using an ‘unstructured’ covariance structure for the repeated measures and was followed by posthoc comparison of the means of the two groups at each time point. Changes in ALT concentrations over 12 months, available for the 90 patients who participated in the weight management programme, were categorised into four subgroups: normal that remained normal, normal that became abnormal, abnormal that normalised and abnormal that remained abnormal. Changes in mean BMI SDS observed in these four groups were assessed using a repeated-measures ANOVA. A 5% level of significance was used.


ALT were raised (>40 IU/l) in 34 (16%) children with median value of 63.5 IU/l (range 41–109 IU/l). The prevalence was greater in boys with 21/90 (23.3%) versus 13/126 (10.3%) in boys and girls, respectively (p=0.01). Correlation between log (ALT) and BMI SDS at baseline was not significant (r=0.1, p=0.13, 95% CI 0.03 to 0.24). Parental history of T2DM was known in 212 children and there was a greater prevalence of transaminitis in those with a positive family history (9/30 (30.0%) vs 24/182 (13.2%), p=0.019) (table 1).

Table 1

Demographic details of subjects

Data sufficient to diagnose the metabolic syndrome (IDF definition) were available in 152 patients who were over 10 years of age and among these, 22/152 (14.5%) patients were positive. 10 out of 28 (35.7%) of those with an elevated ALT fulfilled the criteria for the metabolic syndrome compared with 12/124 (9.7%) with a normal ALT (p<0.001). Those with abnormal ALT values were more likely to have subtle anomalies in glucose handling during an OGTT with higher mean glucose concentrations at 90 and 120 min and an elevated glucose area-under-the-curve (table 1 and figure 1). Glucose concentration data at 120 min during the OGTT were available in 183 patients of whom 14 had IGT but none had diabetes. IGT was not more frequent among those with a raised ALT (3/28, 10.7% vs 11/155, 7.1%).

Figure 1

Glucose levels at 0, 30, 60, 90 and 120 min during oral glucose tolerance test in participants with and without transaminitis.

Of the individual components of the metabolic syndrome, triglyceride (TG) levels (1.31 vs 1.09 mmol/l, p=0.045) and systolic BP (125 vs 117 mm Hg, p≤0.001) were significantly higher in those with raised ALT at baseline in the whole group (table 1). Improvement in TG was the only feature of the metabolic syndrome that showed a significant improvement with ALT changes from baseline to 12 months (Kendall's rank correlation τ = 0.26, p<0.001) among the 90 patients in the prospective trial.

In the prospective trial, there was a significant inverse relationship between 12-month change in BMI SDS and corresponding change in ALT (see figure 2; Kendall's rank correlation τ = −0.29, p<0.001); 15 (68%) of the 22 in whom BMI SDS increased showed an increase in their ALT compared with only 18 out of 68 (26%) in whom BMI SDS fell (p<0.001). ALT normalised in 10 out of 13 (77%) patients whose ALT levels were abnormal at baseline (table 2). These patients showed a significant improvement in mean BMI SDS −0.30 (−0.52 to −0.08) p=0.009, with corresponding mean BMI change of −1.51 kg/m2. Those in whom ALT remained abnormal (n=3) or became abnormal after being normal at baseline (n=4) had no significant change in BMI SDS, and a mean BMI increases of +1.2 and +1.9 kg/m2, respectively. Patients with persisting transaminitis showed a mean weight increase of 9.73 kg compared with 1.75 kg in those who have normalised at the end of the trial. Although the mean BMI SDS in those with a normal ALT at 12 months was lower than those with abnormal ALT, there was no significant differences between the four groups either at baseline (p=0.96) or at 12 months (p=0.55).

Figure 2

Change in alanine aminotransferase (expressed as a ratio, 12 months to baseline) with the magnitude of the absolute fall in body mass index SD score (Kendall's rank correlation coefficient = −0.29 (p<0.001)).

Table 2

Mean weight, body mass index (BMI) and BMI SD score (SDS) vs change in alanine aminotransferase (ALT) before and after 1-year intervention

There was no association between the presence of transaminitis and age, pubertal status, insulin sensitivity, BMI or fat SDS (table 1).


This study demonstrated a high prevalence (16%) of transaminitis especially among boys and those with a positive family history of T2DM in a cohort of 216 obese children from a UK-based secondary care clinic. These patients were more likely to fulfil the criteria for metabolic syndrome and have subtle abnormalities in glucose metabolism during a glucose challenge.

In our prospective trial, patients with persisting transaminitis over 12 months demonstrated a weight gain of 7.98 kg in that time. In those patients with a persisting raised ALT after 12 months we perform a liver ultrasound scan, viral hepatitis screen, copper and caeruloplasmin to exclude other pathological causes of transaminitis: to date no alternate diagnoses have been identified on this screen. What we have demonstrated is that those achieving a modest mean BMI/BMI SDS reduction of 1.51 kg/m2 and 0.3, respectively, achieved normalisation of transaminitis.

The study has limitations. Transaminitis in the form of raised ALT >40 IU/l has been used as a surrogate marker for the diagnosis of NAFLD, while the gold standard for such a diagnosis is by liver biopsy. There is also likely to be some ascertainment bias as this cohort was drawn from those attending a secondary obesity clinic which is more likely to follow-up patients with profound obesity. Although none of the participants has a history of alcohol abuse, data on alcohol consumption among the group is unknown.

Although screening for steatosis has been recommended in the National Health Service care pathway for the screening or management of overweight and obesity in both adults and children,27 there is no universal guideline for the treatment of NAFLD in children. The main stay of treatment recommended by most experts appears to be weight loss via life style modification or pharmacological interventions mostly based on evidence extrapolated from adult data28 29.

In obese adults with NAFLD, a gradual weight loss has shown improvements in transaminitis and reversal of histological abnormalities in its early stages prior to the presence of cirrhotic change.29 30 As little as a 5–10% reduction in body weight has been shown to have beneficial effects on NAFLD.31 Data on the optimal rate of weight loss are lacking and there have been concerns that rapid weight reduction in such patients could exacerbate hepatitis.29 Studies investigating the effectiveness of treatment in NAFLD in obese children and adolescents are scarce. A few studies support weight loss via lifestyle interventions in reversing the progression of NAFLD in children.32,,34 However, the data on the optimal amount of weight loss required to achieve improvement are sparse. In our study, 77% of patients with transaminitis at the beginning of the study normalised after 12 months showing a clinically significant mean BMI SDS reduction of −0.3 (+3.29 to +2.99 SD). Nobili et al32 demonstrated a reduction in levels of fasting glucose, insulin, lipid and liver enzymes as well as liver echogenicity on ultrasonography in 84 children with biochemical and histological diagnosed NAFLD in a 12-month life style intervention programme with diet and physical exercise patients losing more than 5% of body weight had the greatest improvement in ALT. Tock et al33 showed that a 12-week multicomponent therapy with low fat diet and exercise in children with NAFLD diagnosed by changes on ultrasound scan and the presence of transaminitis showed a significant reduction in weight and BMI. In this study, 11% of patients in the subgroup with raised ALT at the beginning of the weight management programme who had normalised their levels after 12 months displayed a 9% reduction in BMI SDS. Reinehr et al34 demonstrated that 109 obese children after a 1-year lifestyle intervention, significantly reduced their weight and the degree of transaminitis compared with 43 without lifestyle intervention. The reduced prevalence of NAFLD was seen with a minimal BMI reduction of less than 0.25 kg/m2 and the greatest changes were seen in those who were most overweight. Studies into the additional effects of pharmacological intervention such as metformin,35 orlistat36 and antioxidants37 38 such as vitamin E have not demonstrated greater effectiveness in the improvement of transaminitis or histological changes of NAFLD in adolescents when compared with weight loss from lifestyle modification alone.

Using transaminitis, as a surrogate marker of NAFLD, we have demonstrated a high prevalence in obese children. Transaminitis is associated with components of the metabolic syndrome especially in those with positive family history of T2DM illustrating the genetic predisposition of these subjects. Our data show that a raised ALT is associated with subtle dysregulation of glucose handing. It is therefore an additional co-morbidity linked to the metabolic syndrome and we believe it should be considered as an additional component of the metabolic syndrome in children. Most importantly we have demonstrated that normalisation of obesity-related transaminitis can be achieved with a reduction in BMI SDS that is achievable with lifestyle modification alone as opposed to additional pharmacotherapy.


This study was supported in part by a grant from the BUPA Foundation. AF was supported by additional grants from The HSA Charitable Trust for Nurses and The Johnson & Johnson/Ethicon Nurse Education Trust Fund. No authors declare any conflict of interests.


  • CW and AF contributed equally to this work

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the United Bristol Hospitals Health Care Trust Ethics Committee (E5472).

  • Provenance and peer review Not commissioned; externally peer reviewed.