Pediatric nonalcoholic fatty liver disease: Prevalence, diagnosis, risk factors, and management

Watch a video presentation of this article

Watch the interview with the author

Answer questions and earn CME

Abbreviations

ALT

alanine aminotransferase

BMI

body mass index

MR

magnetic resonance

NAFLD

nonalcoholic fatty liver disease

NASH

nonalcoholic steatohepatitis

PNPLA3

patatin‐like phospholipase 3

RD

registered dietitian

US

ultrasound

WLS

weight‐loss surgery

With the rise in pediatric obesity, nonalcoholic fatty liver disease (NAFLD) has become the most common cause of abnormal serum aminotransferase levels in children. The best estimate of population prevalence was derived from an autopsy‐based study that found fatty liver in 13% of 742 children (2‐19 years old) in San Diego County.1 Rates of fatty liver were higher in adolescents (17%) and obese children (38%) and were disproportionately lower in African Americans versus other ethnicities and races (Table 1). The estimated US prevalence of NAFLD was 9.6% after adjustments for age, sex, race, and ethnicity. Nonalcoholic steatohepatitis (NASH) affects a smaller proportion of children: 23% of patients with fatty liver in the autopsy study had NASH, and this yielded an estimated population prevalence of 3%.1 Pediatric NASH can be histologically severe, with reports of progressive fibrosis and end‐stage liver disease in adolescents.2

Table 1.

Prevalence and Odds Ratio of Fatty Liver by Ethnicity and Race in San Diego County

Group	Prevalence (%)	Odds Ratio (95% Confidence Interval)
African American	1.5	Reference group
Asian	10.2	1.7 (0.5‐7.3)
Caucasian, non‐Hispanic	8.6	2.7 (1.0‐9.4)
Hispanic	11.8	5.0 (1.9‐17.2)

NOTE: This table was adapted with permission from Pediatrics.¹

The racial and ethnic discrepancies indicate that genetic susceptibility plays a role in the development of NASH in the setting of obesity. Two studies in children have identified a single‐nucleotide polymorphism (rs738409) in the patatin‐like phospholipase domain‐containing 3 (PNPLA3) gene as a risk factor for steatosis, inflammation, and fibrosis.5 This PNPL3 polymorphism is also more frequent in Hispanic and Caucasian children versus African American children.5 Two separate polymorphisms in PNPLA3 and in the glucokinase regulatory protein jointly account for 32% of the hepatic fat fraction variance in Caucasian children, 39% of the variance in African American children, and 15% of the variance in Hispanic children.7

Clinical Risk Factors

NAFLD is strongly associated with obesity and insulin resistance, yet these factors alone are not sufficient for the development of NASH. In a large multicenter study of 254 children (6‐17 years old), metabolic syndrome was associated with more severe steatosis, ballooning degeneration, and more advanced fibrosis.8 A study conducted in a cohort of Italian children confirmed that increased waist circumference, elevated triglyceride levels, and older age were associated with a greater risk of NASH.9

Studies in adults have confirmed that greater consumption of sweetened beverages (which is reflective of a high fructose intake) is associated with more advanced fibrosis.10 Although a clear association between fructose intake and NASH severity has not been shown in children, uric acid levels, which are a surrogate for fructose intake, have been found to be significantly higher in children with NASH.11

Diagnosis

NAFLD in youth is often clinically silent and is discovered only upon screening or the incidental testing of serum aminotransferase levels. At present, three sets of practice guidelines based entirely on expert opinion recommend screening for overweight or obese children, but there is no consensus on the populations appropriate for screening, the frequency of screening, or the appropriate screening tests. Although two of these sets of guidelines were developed in the context of the prevention, assessment, and treatment of paediatric obesity, the third set was reported in a position paper issued by the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition.12 Those in favor of screening cite evidence showing that pediatric NASH can be clinically silent yet histologically severe and progressive.2 At present, the rationale for screening, however, is significantly undermined by the inadequacy of currently available screening methods and the lack of well‐validated effective therapies for NASH beyond weight loss. Not only can liver enzymes be normal with NASH, but a wide range of upper limits of normal have been reported by clinical reference laboratories. A recent national study found 95th percentile upper limits of normal for alanine aminotransferase (ALT) of 26 and 22 U/L for lean and healthy male and female adolescents, respectively, and these levels are significantly lower than the values reported by most clinical laboratories.15

If serum aminotransferase levels are persistently elevated, the exclusion of other causes of chronic hepatitis is required (Table 2). In the absence of these conditions, a child who is overweight or obese with chronically elevated liver enzymes is assumed to have NAFLD, but the diagnosis of NASH can currently be made only with liver biopsy. As in adults, percutaneous liver biopsy is limited by sampling error, higher costs, and risks of significant morbidity and mortality. The decision to biopsy must be considered in the context of the risks of complications and the benefits of confirming the diagnosis of NASH and guiding therapeutic decisions. Liver biopsy is helpful if the diagnosis is uncertain, but it also may guide and encourage the escalation of weight management interventions in a patient failing with standard lifestyle counseling to more intensive options (discussed later). The importance of considering liver biopsy is also underscored by the 30% prevalence of pediatric overweight and obesity, which raises the likelihood that some patients with other causes of liver disease may also be overweight and mistakenly be assumed to have only NAFLD. The causes of biopsy‐confirmed steatosis and steatohepatitis are similar to those in adults, but in young children (especially those who lack insulin resistance), inborn errors of metabolism must also be considered (Table 2). Histologically, the majority of pediatric NASH cases exhibit a mixed phenotype of both portal and lobular inflammation with milder ballooning, but a subset of these cases have prominent portal inflammation without ballooning and lobular inflammation.16 The clinical significance and outcomes of these pediatric phenotypes are unclear in the absence of prospective natural history studies (including histology) in children.

Table 2.

Differential Diagnosis of Chronic Hepatitis and Steatosis/Steatohepatitis

If steatosis/steatohepatitis is documented, evaluate for the following:

• Viral hepatitis C

• Wilson's disease

• Alcohol consumption (in adolescents)

• Malnutrition

• Medications associated with steatosis (high‐dose estrogen, Depakote, methotrexate, and high‐dose corticosteroids)

• Total parenteral nutrition

• Inborn errors of metabolism (including fatty acid oxidation defects and other disorders of mitochondrial function, glycogen storage diseases, and lysosomal storage disorders)

• Cystic fibrosis

If serum aminotransferases are elevated, also consider the following:

• Viral hepatitis B (consider hepatitis A if it is clinically indicated)

• Autoimmune hepatitis

• Hemochromatosis

• Alpha‐1‐antitrypsin deficiency

• Drug toxicity (statins, metformin, Depakote, methotrexate, and high‐dose corticosteroids)

Noninvasive Assessment

The noninvasive diagnosis of NASH is not clinically feasible yet. Imaging studies, including ultrasound (US), magnetic resonance (MR) imaging or spectroscopy, and computed tomography, cannot at present distinguish NASH from NAFLD. US‐ and MR‐based elastography methods show promise in diagnosing advanced fibrosis, but we are awaiting further validation trials in the pediatric age group.17 Furthermore, US elastography is limited by severely obese body habitus, and MR modalities are limited by cost.

Panels of biomarkers for fibrosis detection in children remain investigational. A pediatric NAFLD index (range = 0‐10) derived from age, triglycerides, and waist circumference had a high positive predictive value of 98.5 (95% confidence interval = 91.8‐100.0) when a value ≥ 9 was used for the diagnosis of NASH in a single‐center Italian pediatric cohort, but this needs to be validated in other diverse cohorts.9 The addition of enhanced liver fibrosis markers to this model predicted the absence or presence of fibrosis in 86% of the patients in the same cohort, but this is not commercially available in the United States.19

Management in Children

The first priority in preventing and treating NAFLD and NASH remains making lifestyle changes to achieve a healthy weight. Improvements in the body mass index (BMI) through exercise and a healthier diet can result in histological and sonographic improvements in pediatric NAFLD.20 Notably, interventions in studies were intensive with frequent visits and multidisciplinary support. At our center, we follow a more feasible management algorithm of visits every 3 months that include counseling with a registered dietitian (RD); this has resulted in significant mean decreases in BMI and ALT levels after 1 year.22 Multidisciplinary care is recommended for the management of pediatric obesity and NAFLD (Fig. 1).

Figure 1.

Management algorithm for pediatric NAFLD. *An at‐risk child is defined as a child who is overweight (BMI ≥ 85–94th percentile) or obese (BMI ≥ 95th percentile), especially if cardiometabolic risk factors (including insulin resistance, diabetes mellitus type 2, central adiposity, and dyslipidemia) are present. **As specific treatments for pediatric NASH, high‐dose vitamin E and bariatric surgery require additional definitive outcome and validation studies in children.

When lifestyle management fails, there are no well‐validated medications for treating pediatric NASH. There has been only one pediatric, multicenter, randomized controlled trial including histological outcomes; that trial included 173 children with biopsy‐proven NAFLD. Although no benefit was shown in lowering ALT levels with either vitamin E or metformin versus a placebo, secondary analyses showed that treatment with 400 IU of vitamin E twice daily for 96 weeks was associated with significantly greater histological resolution of NASH in comparison with a placebo, and there were no significant risks (Fig. 2).23 Despite these encouraging results, further validation of high‐dose vitamin E in children with NASH is warranted before the issuance of any broad recommendations for its use at this age because the long‐term safety of high‐dose vitamin E remains unclear; recent studies in adults have indicated higher risks for morbidity, mortality, and prostate cancer associated with high‐dose vitamin E (>400 IU per day).24 Also, vitamin E was associated with the resolution of NASH in only 58% of the patients (25/43) with borderline or definite NASH at the baseline in the treatment group, and this suggests that it is unlikely to be a panacea for pediatric NASH.

Figure 2.

Resolution of NASH by the treatment group in the Treatment of Nonalcoholic Fatty Liver Disease in Children trial. NASH resolution was defined as no NASH at week 96 among patients with borderline or definite NASH at the baseline.

Lacking other easily implemented and effective treatment options for progressive NASH, some pediatric obesity experts now consider advanced NASH to be an appropriate indication for adolescent weight‐loss surgery (WLS), but they have not defined specific histological criteria for severe NASH.26 The recommendation rests on favorable meta‐analysis results from largely retrospective studies of WLS in adult patients with NAFLD, but controlled trials are lacking for both age groups.27 Because WLS is the only current treatment proven to result in significant and sustained weight loss for severely obese adolescents over a period of up to 10 years, it remains an attractive option for select severely obese adolescents with NASH but requires prospective controlled outcome studies.28 Whether it would be effective for more advanced fibrotic NASH or any positive effects would persist over a longer lifetime (especially with significant weight regained) is unknown.

Future Horizons

Pediatric NAFLD is common and can result in severe and progressive NASH in childhood. Routine screening in obese children is controversial because of the limited knowledge of the natural history of the disease, the lack of evidence for improved outcomes with screening, and the inadequacy of current screening modalities. The prevention and treatment of pediatric obesity remain the cornerstone for reducing the pediatric burden of NAFLD, but significant weight improvement is not attainable for many patients through lifestyle changes alone. Advances must be made in noninvasive diagnosis to improve screening for NASH, and further randomized controlled trials of alternative therapies are needed to expand our treatment options for pediatric NASH.