OBJECTIVE. Fatty liver disease is diagnosed increasingly in children, but the prevalence remains unknown. We sought to determine the prevalence of pediatric fatty liver as diagnosed by histology in a population-based sample.
METHODS. We conducted a retrospective review of 742 children between the ages of 2 and 19 years who had an autopsy performed by a county medical examiner from 1993 to 2003. Fatty liver was defined as ≥5% of hepatocytes containing macrovesicular fat.
RESULTS. Fatty liver was present in 13% of subjects. For children and adolescents age 2 to 19 years, the prevalence of fatty liver adjusted for age, gender, race, and ethnicity is estimated to be 9.6%. Fatty liver prevalence increases with age, ranging from 0.7% for ages 2 to 4 up to 17.3% for ages 15 to 19 years. Fatty liver prevalence differs significantly by race and ethnicity (Asian: 10.2%; black: 1.5%; Hispanic: 11.8%; white: 8.6%). The highest rate of fatty liver was seen in obese children (38%).
CONCLUSIONS. Fatty liver is the most common liver abnormality in children age 2 to 19 years. The presence of macrovesicular hepatic steatosis in ∼1 of every 10 children has important ramifications for the long-term health of children and young adults. The influence of the risk factors identified should be taken into consideration in the development of protocols designed to screen at-risk children and adolescents.
The growing epidemic of childhood obesity has prompted studies of the prevalence of obesity-related conditions, such as hypertension, hypercholesterolemia, and the metabolic syndrome.1–4 Reports of pediatric fatty liver disease and steatohepatitis in obese children have been increasing5–7 and include cases of cirrhosis8–11 and liver transplantation.12 However, despite the potentially serious nature of fatty liver in children, the prevalence is unknown.
Efforts to determine the prevalence of fatty liver in children have been limited to indirect measures, such as blood tests or ultrasound, to predict a histologic outcome. In 1989, based on ultrasound, fatty liver prevalence was estimated to be 2.6% in school-aged children in northern Japan.13 Projections for children in the United States are considerably higher,14 although no population-based estimates are available. All of the other studies have been limited to children selected for the conditions of obesity. Taken together, the prevalence of fatty liver in obese children in China, Italy, Japan, and the United States has been reported to be between 10% and 77%.15–18 The wide range in estimates of prevalence is attributable in part to the difficulty in determining the presence or absence of fatty liver using noninvasive tests. Therefore, we conducted the Study of Child and Adolescent Liver Epidemiology (SCALE) to determine the prevalence of fatty liver as diagnosed by histology in a population-based sample of children and adolescents.
SCALE Study Design
Given that fatty liver is a histologic diagnosis, numerous problems exist in trying to estimate prevalence. Children without fatty liver are unlikely to undergo liver biopsy. However, noninvasive surrogate markers lack sufficient sensitivity and specificity. Thus an autopsy-based study design is the only means of assessing population-based histology. The county of San Diego is particularly well suited to such a study based-on its size as the fourth largest county in the United States and its substantial racial and ethnic diversity. The county of San Diego medical examiner investigates the cause of death in all homicides, suicides, and accidental deaths. The medical examiner also investigates natural deaths when the cause of death is unknown. However, the medical examiner does not perform a postmortem examination for those children who have been seen by a physician within 20 days before death and have a known cause of death.
We included all children age 2 through 19 years at the time of death who had an autopsy performed by the medical examiner from 1993 to 2003. We set a minimum age of 2 years, because this is the youngest age reported for biopsy-proven steatohepatitis.10 The upper age limit was set at 19 years to mirror the age brackets used by the US Census Bureau. Exclusion criteria were insufficient data (missing clinical data or the absence of a liver slide for review) or identification of a factor that may unduly influence liver histology (inpatient clinical care or alcohol or drug use). SCALE was conducted with approval from the county of San Diego and the University of California, San Diego.
Clinical Data Collection
Medical examiner records were reviewed for age, gender, race, ethnicity, weight, height, toxicology, and the mode and cause of death. BMI was calculated as the weight (kilograms) divided by the height (meters) squared. Weight status was defined according to the Institute of Medicine as underweight (BMI ≤ 5th percentile), normal weight (BMI > 5th and < 85th percentile), overweight (BMI ≥ 85th and < 95th percentile), and obese (BMI ≥ 95th percentile).19 BMI varies with age and gender; therefore, to compare subjects independently of age and gender, BMI Z-scores were calculated. The BMI z score represents the number of SDs from the national reference mean that a subject is for a given age and gender.
Assessment of Liver Histology
A hepatopathologist (C.B.) with expertise in pediatric fatty liver reviewed a section of liver tissue stained with hematoxylin and eosin for all of the subjects in a blinded fashion. The severity of steatosis from any 1 sample obtained at autopsy has been shown to correlate well with the severity of steatosis for the liver overall.20 Steatosis was recorded as the percentage of hepatocytes containing macrovesicular fat (fat droplet equal to or larger than the size of the nucleus) or microvesicular fat (numerous small fat droplets surrounding a centrally located nucleus). Fatty liver was defined as ≥5% of hepatocytes containing macrovesicular fat.21 Steatohepatitis was defined as reported previously for children; fatty liver along with ≥1 of the following features: ballooning degeneration, perisinusoidal fibrosis, portal inflammation, or portal fibrosis.11 Based on these features, those biopsies with steatohepatitis were subclassified as type 1, type 2, or overlap.11 Type 1 steatohepatitis is defined as fatty liver with ballooned hepatocytes and/or perisinusoidal fibrosis in the absence of portal features. Type 2 steatohepatitis is defined as fatty liver with portal inflammation and/or fibrosis and the absence of ballooning degeneration and perisinusoidal fibrosis.
Means and SEs of the mean are reported for continuous variables, and frequencies and percentages are reported for categorical variables. To estimate the prevalence of fatty liver in the county of San Diego, fatty liver was defined as ≥5% steatosis versus <5% steatosis. To allow for direct comparisons between the study population and county population data, the prevalence of fatty liver, overall and for each subgroup, was directly standardized to the age group, gender, race, and ethnicity distribution of the county. We obtained 95% confidence interval estimates for prevalence using estimation for proportions based on normal distribution assumptions or simulation using 2000 samples when normal distribution assumptions did not hold. Based on our previous work suggesting that age, gender, race, ethnicity, and weight influence fatty liver,10,22 the independent relationship of each of these predictors with fatty liver was tested using 2-tailed t tests for continuous variables and Pearson χ2 tests for categorical variables. In addition, an a priori logistic regression model was developed, and odds ratios for fatty liver were estimated using the lowest-risk subgroup as the reference category. All of the reported P values are 2-sided.
From 1993 to 2003, 1138 children age 2 to 19 years had an autopsy performed by the San Diego County Medical Examiner. Of these, 86 subjects were excluded because of insufficient data: no available liver tissue (n = 36) or missing clinical data (n = 50). An additional 310 subjects were excluded because of factors that may influence liver histology: inpatient hospital care (n = 98), alcohol (n = 128), positive drug toxicology screen (n = 68), and prescription medications (n = 16). The remaining 742 subjects comprise the SCALE population. Motor vehicles were involved in 41% (301 of 742) of all deaths. The decedent was a driver in 96 cases, passenger in 143 cases, pedestrian in 48 cases, and a bicyclist in 14 cases. Other accidents (drowning, falls, and fires) accounted for an additional 115 deaths. Additional common modes of death were homicide (n = 137) and suicide (n = 100). Postmortem examination determined the most common mechanisms of death to be blunt force head trauma (35%), asphyxia (14%), gunshot wound to the head (12%), gun shot wound to the trunk (9%), and blunt force chest trauma (9%). Twenty-six other causes made up the remaining 21% of cases.
The SCALE population (Table 1) had a mean age of 14 years, and the majority of subjects were boys. The racial and ethnic distribution is noted to closely represent that of children in San Diego as determined by the US Census Bureau's 2000 Census.23 The distribution of weight category in the SCALE population (4% were underweight; 54% were normal weight; 17% were overweight; and 26% were obese) also closely matched the weight distribution for children age 5 to 19 years in the county of San Diego as estimated by the Centers for Disease Control and Prevention Pediatric Nutrition Surveillance Study.24
Descriptive Findings of Fatty Liver and Steatohepatitis in SCALE Population
Chronic liver disease other than fatty liver was detected in only a small number of subjects: chronic hepatitis (n = 3), sclerosing cholangitis (n = 1), adenoma (n = 1), granuloma (n = 1), and hemangioma (n = 1). As shown in Table 1, fatty liver was present in 13% (n = 97) of subjects. There were 532 subjects (72%) with no fat evident on liver biopsy and 113 subjects (15%) who had between 1% and 4% of hepatocytes with macrovesicular steatosis. Children and adolescents with fatty liver were significantly older (P < .0001), more likely to be male (P < .02), and had a higher BMI (P < .0001) than children and adolescents without liver steatosis. Fatty liver was absent in underweight children and present in 5% of normal weight children compared with 16% of overweight children and 38% of obese children. Thus, overweight and obese children accounted for 81% of all of the cases of fatty liver.
Also shown in Table 1, steatohepatitis was observed in 23% (22 of 97) of the subjects with fatty liver. This represents 3% (22 of 742) of the total study population. Steatohepatitis was characterized as type 1 in 4 subjects, type 2 in 17 subjects, and as overlap in 1 subject. The major features of pediatric steatohepatitis were present in the following percentages: portal inflammation (68%), portal fibrosis (59%), perisinusoidal fibrosis (14%), and ballooning degeneration (18%). Bridging fibrosis or cirrhosis was seen in 9% of those with steatohepatitis.
Estimate of Fatty Liver Prevalence
After standardizing for age, gender, race, and ethnicity, the estimated prevalence of fatty liver is 9.6%, which represents ∼70000 children age 2 to 19 years in the county of San Diego (Table 2). Fatty liver prevalence increases with age ranging from 0.7% for ages 2 to 4 years up to 17.3% for ages 15 to 19 years. Fatty liver is present in 11.1% of boys and 7.9% of girls. As shown in Table 3, after controlling for age and gender adjusted BMI z score, race and ethnicity significantly predict the presence of fatty liver. Fatty liver is most prevalent in children and adolescents of Hispanic ethnicity and least prevalent among children and adolescents of black race. As such, the odds of a Hispanic child having fatty liver are 5 times greater than for a black child. The prevalence of fatty liver is estimated to be 10.2% in children and adolescents of Asian race. However, there was substantial variability among the subgroups of Asian children. For example fatty liver was present in 20% of Filipino-American children but only in 4% of children of Cambodian or Vietnamese background.
We conducted the first study of fatty liver epidemiology in the United States to directly assess liver histology. Moreover, we estimate that the prevalence of fatty liver is 9.6% in children age 2 to 19 years in the county of San Diego. If the prevalence is similar for the entire United States, this would represent >6.5 million children and adolescents. In addition, we demonstrate that there are strong effects of race and ethnicity on the presence of fatty liver in children. These findings highlight a public health problem and contribute to the identification of risk factors associated with pediatric fatty liver disease.
Our finding that 38% of obese children have fatty liver is consistent with reports from clinical populations of obese children.15,18 Because the majority of children with fatty liver are obese, effective prevention and treatment of obesity will be important components of addressing the prevalence of fatty liver. However, despite an increased focus on obesity prevention, the rates of pediatric obesity in the United States continue to rise.25 Therefore, we must also consider secondary prevention strategies targeted toward preventing the development of fatty liver in children who are overweight. The current data suggest that a key age group to target for secondary prevention is children age 5 to 9 years.
Despite black children having high rates of risk factors for fatty liver, such as obesity and insulin resistance, few black children are included in pediatric clinical series of fatty liver.26–29 This discrepancy could be because of either lower rates of fatty liver in black children or, alternatively, that fatty liver in black children is more likely to remain undiagnosed. Two recent studies of obese children and adolescents observed much lower rates of a putative surrogate marker of fatty liver, elevated serum alanine aminotransferase activity, in blacks than in whites.22,30 The histologic data in the current study strongly demonstrates that black children, even when obese, are much less likely to have fatty liver. Furthermore, these data refute the contention of a race-based selection bias in clinical studies and support the need for studies to explore the biological basis of the existing differences.
Efforts to curtail obesity in the Hispanic community with a special emphasis on mitigating the development and progression of fatty liver disease should be a public health imperative. In the most recent national report, ∼40% of Mexican-American children were either overweight or obese.25 A high rate of obesity is one important risk factor for fatty liver. However, our data demonstrate that, even after controlling for the severity of obesity, Hispanic boys and girls have higher rates of fatty liver than non-Hispanic peers. Not only are Hispanic children at increased risk for fatty liver, but clinical data suggest they are more prone to advanced liver fibrosis.11 The high rates of fatty liver in Hispanic adolescents may contribute to the higher rate of liver-related morbidity and mortality experienced by Hispanic-American adults compared with white or black Americans.31,32
SCALE is the first study to provide information on fatty liver prevalence within Asian Americans. We estimate that 10% of Asian children and adolescents in the county of San Diego have fatty liver. This is much higher than was reported previously by Tominaga et al13 for the prevalence of fatty liver in Japanese children living in northern Japan. The difference may be explained in part by the high false-negative rate for ultrasound in the context of mild steatosis. In addition, the environments of children living in northern Japan and those living in the United States are likely to be different. Moreover, despite being considered as one common race, there may be considerable differences among children with ancestors from different Asian nations. The Asian subgroup at greatest risk may be Filipino American children, which is consistent with previous reports of a propensity for lipid abnormalities in normal-weight Filipino adolescents33 and for type 2 diabetes mellitus in Filipino-American adults.34 Whether differences between Asian subgroups in the development of fatty liver are related to genetic or environmental factors merits further investigation.
We acknowledge that it is uncertain to what extent the SCALE population mirrors all of the children in the community. In strong support, however, is the finding that the distributions of race and ethnicity, as well as weight classification within the SCALE population, closely match the distributions of these features within the county of San Diego. A limitation of the study design is that the specific cause of fatty liver cannot be determined. In the absence of a liver biopsy, the differential diagnosis for a child with suspected fatty liver is rather broad. However, because we based the diagnosis on histology, the possibilities are substantially reduced. By restricting the definition of fatty liver to require macrovesicular steatosis, we eliminated many possible diagnoses, such as inborn errors of metabolism. The inclusion of only those children who experienced a rapid death outside of the hospital setting eliminated possible causes, such as parenteral nutrition. The possible explanations for fatty liver were further reduced by excluding all of the subjects with any drug or alcohol use. Therefore, the remaining differential diagnosis is rather narrow and can be focused on nonalcoholic fatty liver disease, cystic fibrosis, and Wilson's disease. No children with cystic fibrosis were knowingly included in the data set. The prevalence of Wilson's disease is 1 in 30000, making it an unlikely cause. Thus, although we cannot prove cause on a case-by-case basis, it is likely that most of the cases of fatty liver in the study population are attributable to nonalcoholic fatty liver disease.
We conclude that fatty liver is the most common form of pediatric liver disease. Obesity is a major risk factor and is associated with high rates of fatty liver, especially in Hispanic-American children. The risk factors identified may be useful in the development of protocols designed to screen at-risk children and adolescents. Despite an increasing awareness of fatty liver disease and steatohepatitis, the majority of affected children go undiagnosed. An important goal must be to identify those children with advanced fibrosis, as well as the ones most likely to progress to end-stage liver disease. Given the large number of children affected, it is imperative that we establish a better understanding of the natural history of pediatric fatty liver not only in terms of the progression of liver disease but also regarding its potential relationship with other health outcomes, such as type 2 diabetes mellitus and cardiovascular disease.
SCALE was supported in part by grants from the Rest Haven Foundation and by M01 RR000827 from the National Center for Research Resources of the National Institutes of Health for the General Clinical Research Center at University of California, San Diego.
We thank Susan Korosy for her assistance with data collection.
- Accepted May 22, 2006.
- Address correspondence to Jeffrey B. Schwimmer, MD, Division of Gastroenterology, Hepatology, and Nutrition, Department of Pediatrics, University of California, San Diego, 200 West Arbor Dr, San Diego, CA 92103-8450. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
This work was presented in abstract form at the annual meeting of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition; October 21, 2005; Salt Lake City, UT.
- Sorof JM, Lai D, Turner J, Poffenbarger T, Portman RJ. Overweight, ethnicity, and the prevalence of hypertension in school-aged children. Pediatrics.2004;113 :475– 482
- Ford ES, Mokdad AH, Ajani UA. Trends in risk factors for cardiovascular disease among children and adolescents in the United States. Pediatrics.2004;114 :1534– 1544
- ↵Lavine JE, Schwimmer JB. Pediatric non-alcoholic steatohepatitis. In: Farrell G, George J, Hall P, McCollough A, eds. Non-Alcoholic Steatohepatitis. Oxford, England: Blackwell Publishing; 2004:229–240
- ↵Committee on Prevention of Obesity in Children and Youth. Preventing Childhood Obesity Health in the Balance. Washington, DC: National Academies Press; 2005
- ↵Schwimmer JB, McGreal N, Deutsch R, Finegold MJ, Lavine JE. The influence of gender, race, and ethnicity on suspected fatty liver in obese adolescents. Pediatrics.2005;115 (5). Available at: www.pediatrics.org/cgi/content/full/115/5/e561
- ↵SANDAG. Census data online. Available at: www.sandag.org/resources/demographics_and_other_data/demographics/census/index.asp. Accessed August 18, 2005
- ↵Centers for Disease Control and Prevention. Pediatric nutrition surveillance system. Available at: www.cdc.gov/pedness. Accessed August 20, 2005
- Copyright © 2006 by the American Academy of Pediatrics