Normal serum alanine aminotransferase and non-alcoholic fatty liver disease among Korean adolescents: a cross-sectional study using data from KNHANES 2010–2015

The prevalence of non-alcoholic fatty liver disease (NAFLD) is increasing worldwide. [1] Therefore, it is necessary for obese adolescents to be actively examined for NAFLD. And symptoms and clinical signs, laboratory and radiological imaging test, and liver biopsy is needed to make diagnosis finally as NAFLD. [2, 3] But in general population, it is not possible to perform all diagnostic procedures for each people. Therefore, Alanine aminotransferase (ALT) is used to find prevalence in general populations. ALT is an enzyme that is found in the cytosol of hepatocytes, and blood levels of ALT increase after liver injury [4]. Thus, blood testing for ALT is used globally as a minimally invasive and inexpensive tool for detecting chronic liver diseases, such as non-alcoholic fatty liver disease (NAFLD) [1].

However, the reference ranges for normal ALT vary widely across different laboratories and populations [1]. Among adults, the upper normal limit of ALT was derived from healthy Italian blood donors, with values of 30 U/L for men and 19 U/L for women [5]. Among American adolescents, the upper normal limit of ALT was estimated by the National Health and Nutrition Examination Survey to be 25.8 U/L for boys and 22.1 U/L for girls [6, 7].

Nevertheless, population differences indicate that these values may not be the same among Korean adolescents. However, the method used to define NAFLD in this study is only tools to estimated NAFLD in public populations, so this result cannot directly be used in clinic. But result of this study will give clues to evaluate and manage NAFLD in clinic. Therefore, the present study aimed to estimate the upper normal limits of ALT among Korean adolescents, as well as the prevalences of NAFLD based on those values.

The present study evaluated data from the 2010–2015 Korea National Health and Nutrition Examination Survey (KNHANES). These annual cross-sectional surveys are performed using multi-stage probability samples that are representative of the general Korean population. The data of KNHANES surveys are available at http://​knhanes.​cdc.​go.​kr/.

During 2010–2015, 48,482 individuals participated in the KNHANES. The present study included participants who were 10–18 years old, but excluded participants with missing data and metabolic syndrome components, based on the International Diabetes Federation consensus definition (www.​idf.​org). Furthermore, 10–15-year-old participants were excluded if they had a waist circumference of ≥90th percentile for sex and age, triglyceride (TG) levels of ≥150 mg/dL, high density lipoprotein (HDL) levels of < 40 mg/dL, systolic blood pressure (SBP) of ≥130 mmHg, diastolic blood pressure (DBP) of ≥85 mmHg, and glucose levels of ≥100 mg/dL. Moreover, participants who were 16–18-years-old were excluded if they had a waist circumference of ≥90 cm or HDL levels of < 40 mg/dL (boys), a waist circumference of ≥80 cm or HDL levels of < 50 mg/dL (girls), TG levels of ≥150 mg/dL, SBP of ≥130 mmHg, DBP of ≥85 mmHg, and glucose levels of ≥100 mg/dL. Finally, we excluded participants who previously had diseases related to low HDL levels, high TG levels, diabetes, or very low/high body mass index (BMI, <5th percentile or > 85th percentile) [8]. Furthermore, participants who had hepatitis B infection were excluded. Thus, 1785 healthy participants (916 boys and 869 girls) were included in the analysis to determine the upper normal limit of ALT. For the present study, NAFLD was defined as being overweight (≥85th percentile of BMI) plus having elevated ALT (≥95th percentile, 24.1 U/L for boys and 17.7 U/L for girls). A total of 4149 participants (2226 boys and 1923 girls) were included in the analyses to estimate the prevalence of NAFLD (Figure 1) [9].

All statistical analyses were performed using SPSS software (version 23.0; IBM Inc., Armonk, NY). Categorical data were collected for healthy participants (without metabolic syndrome components), and continuous laboratory and anthropometry data were expressed as mean ± standard error. The 95th percentiles for ALT were estimated for each sex using weighted analysis, and those values were used to estimate the prevalences of NAFLD among adolescents who participated in KNHANES and had available BMI and ALT data.

The characteristics of the 1785 healthy participants (916 boys and 869 girls) are shown in Table 1. All characteristics appeared to be within the normal ranges, and the mean ALT levels were 14.2 U/L for boys and 10.9 U/L for girls. The 95th percentiles for ALT among healthy participants were 24.1 U/L for boys and 17.7 U/L for girls (Figure 2a).

A total of 4149 KNHANES participants had available ALT and BMI data. Among all participants during 2010–2015, the estimated prevalence of elevated ALT levels was 10.9% (95% confidence interval [CI]: 9.8–12.1%). During 2015, the overall estimated prevalence of elevated ALT levels was 13.7% (95% CI: 11.0–15.5%). In addition, the estimated prevalences of elevated ALT levels during 2015 were 14.9% (95% CI: 11.0–19.8%) for boys and 12.3% (95% CI: 8.6–17.3%) for girls (Figure 2b).

The prevalence of NAFLD was estimated using the 95th percentile values for ALT (24.1 U/L for boys and 17.7 U/L for girls) plus the age- and sex-specific 85th percentile values for BMI. During 2015, the overall prevalence of NAFLD was 8.9% (95% CI: 6.7–11.6%), with prevalences of 10.8% (95% CI: 7.7–15.0%) among boys and 6.6% (95% CI: 4.0–10.9%) among girls (Figure 3).