Predictors and markers of hepatic fibrosis in obese children and adolescents

INTRODUCTION

Non-alcoholic fatty liver disease (NAFLD) is among the most common comorbidities associated with childhood and adolescent obesity.^[1-3] NAFLD is a progressive condition that spans a spectrum from steatosis to steatohepatitis, fibrosis, and cirrhosis, with an increased risk for hepatocellular carcinoma in later life.^[4] The limited potential for reversal and the high probability of progression to liver cirrhosis highlight the importance of early detection of hepatic fibrosis.^[5] Although a liver biopsy is considered the gold standard for diagnosing hepatic fibrosis, its invasive nature makes it challenging to implement in clinical practice.^[6] While ultrasound can detect the presence of liver fat, it does not provide information about fibrosis, which is the primary concern. Very little information about the relationship between liver enzymes and hepatic fibrosis in children and adolescents is available.

Transient elastography (TE), a non-invasive method for assessing hepatic steatosis and stiffness, correlates with liver biopsy results in adults and is increasingly utilized to identify hepatic fibrosis.^[7] However, there is limited information regarding the correlation of TE-identified hepatic fibrosis in Indian children and adolescents with NAFLD.^[8] The high cost of TE restricts its widespread implementation in resource-poor settings. Serum alanine transaminase (ALT) levels, however, are readily available across resource settings and have the potential to identify obese children and adolescents who are at risk for hepatic fibrosis. However, there is a lack of information regarding the appropriate ALT cut-off to predict hepatic fibrosis in obese children and adolescents.

Identifying determinants and predictors of hepatic fibrosis would help devise early diagnosis and treatment strategies. We have reported a high prevalence of NAFLD in obese children and adolescents presenting to our center.^[9] This study was done to identify the prevalence, predictors, and determinants of hepatic fibrosis in obese Indian children and adolescents attending a tertiary center in urban north India.

MATERIAL AND METHODS

This study was conducted at our hospital from June 2022 to May 2023, following ethical approval from the institutional ethics board. Children and adolescents between 5 and 18 years of age who presented at our pediatric (controls) and pediatric endocrinology clinic (obese subjects) were invited to participate. We excluded subjects with chronic liver disease, pathological obesity, the use of hepatotoxic drugs, recent viral illnesses, and disorders leading to hepatic fibrosis. Controls were selected from children presenting to the pediatric outpatient department with non-gastrointestinal complaints. A sample size of 62 subjects in each group was calculated with 80% power and a significance level of 0.05 to detect a 10% difference in fibrosis prevalence between the two groups.^[10]

Clinical and laboratory assessments were conducted following informed consent from parents and assent from children and adolescents. Height and weight were measured using a Seca digital stadiometer (Model No. 274) and weighing scale (Model No. 813). Body mass index (BMI) was converted to a standard deviation score (SDS) according to the data in the Indian Academy of Pediatrics (IAP) 2015 growth standards.^[11] The IAP 2015 criteria were used to classify individuals into obese and non-obese categories.^[11] The assessment included serum ALT levels and TE measurements in both groups and complication screenings (oral glucose tolerance test [OGTT], lipid profile, and blood pressure assessment) in obese subjects. Hypertension was defined as blood pressure at or above the 95^th percentile for age, gender, and height; a systolic blood pressure of at least 130 mmHg or a diastolic blood pressure of at least 80 mmHg measured on three consecutive occasions.^[12] An OGTT was performed after an overnight fast, with blood glucose (BG) levels measured at 0 and 2 h after administering 1.75 g/kg anhydrous glucose dissolved in water (maximum of 75 g). The American Diabetes Association criteria were used to identify dysglycemia (impaired fasting glucose, impaired glucose tolerance [IGT], or diabetes).^[13] A fasting sample was obtained for the lipid profile. Dyslipidemia was diagnosed according to the Endocrine Society guidelines.^[14] All investigations were carried out in a National Accreditation Board for Testing and Calibration Laboratories (NABL)-accredited laboratory using the Roche Cobas 6000 system.

TE was conducted by a single trained observer using the FibroScan 502 Touch (Echosens, France) to obtain the controlled attenuation parameter score (CAP, a marker of hepatic steatosis) and the liver stiffness measurement (LSM). The subjects were instructed to fast for at least three hours before the examination. They were positioned in the dorsal decubitus posture with the right arm fully abducted, and measurements were taken by scanning the right liver lobe through an intercostal space. The small, medium, or XL probe was utilized based on body size and the automatic probe selection tool integrated into the device software. The results reflected the median value of ten valid measurements with an interquartile range of <10%. Based on the LSM score, subjects were classified as having no fibrosis (F0–F1 <7 kPa), significant fibrosis (F2 7.0–8.5 kPa), advanced fibrosis (F3 8.6–10.9 kPa), and cirrhosis (F4 >11 kPa).^[15] CAP values between 240 dB/m and 280 dB/m indicated steatosis, while values exceeding 280 dB/m suggested severe steatosis.^[15] Subjects with hepatic fibrosis, severe steatosis, or elevated ALT levels were referred to a pediatric gastroenterologist for evaluation of alternative causes of chronic liver disease and management. The workup included an assessment for hepatitis B and C infection, immunological cause, and possible genetic etiology.

The data were analyzed using the Statistical Package for the Social Sciences version 24.0 and expressed as percentages, means, and standard deviations. The t-test and Chi-square test were utilized to compare continuous and categorical variables among obese children and controls. The correlation between CAP and LSM values and continuous variables was evaluated using the Pearson test. Multivariate analysis identified predictors of hepatic steatosis and fibrosis. A receiver operating characteristic (ROC) curve was created for obese subjects to identify ALT and BMI SDS cutoffs for detecting hepatic steatosis and fibrosis. A P-value below 0.05 was deemed significant.

RESULTS

Ten of the 133 subjects who met the inclusion criteria approached for the study did not consent to the procedure. TE was successful in all subjects (62 obese and 61 controls). Steatosis was observed in 41 (33.3%; severe steatosis in 25, 20.3%) and fibrosis in seven subjects (5.7%; advanced fibrosis F3 in 4, 3.3%). The diagnostic evaluation of subjects with severe steatosis and fibrosis did not reveal any underlying hepatic pathology.

Obese subjects had a comparable age and gender distribution compared to the controls. BMI SDS, ALT levels, CAP score, and LSM values were greater in obese subjects than in controls [Table 1]. Obese subjects were more likely to have ALT levels exceeding twice the upper limit of normal (50 IU/L in boys and 44 IU/L in girls, 35.5% vs. 3.3%, P = 0.003), steatosis (59.5% vs. 6.5%, P < 0.001), severe steatosis (40.3% vs. none), significant fibrosis (9.6% vs. 1.6%, P = 0.06), and advanced fibrosis (6.5% vs. none) compared to controls. Metabolic complications observed in obese individuals included hypertension in 14 (22.6%), dyslipidemia in 11 (17.7%), and dysglycemia in three (4.8%, impaired fasting BG in one and IGT in two). The proportion of obese subjects with metabolic complications was higher in those with steatosis than in those without it (35.5% vs. 10.5%, odds ratio 4.7, P = 0.05). Obese subjects with steatosis (n = 25) had higher BMI SDS 2.2 ± 0.58 as against 1.9 ± 3.9, P = 0.03) and a trend of higher 2 h BG level (111.7 ± 17.4 mg/dL as against 97.5 ± 1.6 mg/dL, P = 0.06) compared to those without it (n = 37). No difference in blood pressure, fasting BG, and lipid levels were observed in the two groups.

The LSM score correlated with CAP values (r = 0.5, P < 0.001). All obese subjects with fibrosis exhibited steatosis (mean CAP score 308 ± 25.4 dB/m). The CAP score positively correlated with ALT levels (r = 0.6, P < 0.001), BMI SDS (r = 0.7, P < 0.001), and age (r = 0.3, P = 0.006). A similar correlation was observed between the LSM score and ALT levels (r = 0.5, P < 0.001), BMI SDS (r = 0.3, P < 0.001), and age (r = 0.3, P = 0.002). Obese subjects with steatosis (n = 37) showed a higher BMI SDS (2.2 ± 0.6 vs. 1.9 ± 0.4, P = 0.03) and ALT levels (57.2 ± 34.0 vs. 39.7 ± 24.0, P = 0.02) than those without it (n = 25). Obese subjects with fibrosis (n = 6) had higher BMI SDS (2.4 ± 0.8 vs. 2.1 ± 0.5, P = 0.05) and ALT levels (81.1 ± 42.2 vs. 46.7 ± 27.0, P = 0.05) than those without it (n = 56). BMI SDS (standardized odds ratio 4.1, 95% confidence interval [CI], 1.1–10.4, P = 0.001) and age (standardized odds ratio 1.3, 95% CI: 1.1–1.6, P = 0.01) were significant determinants of hepatic steatosis in the multivariate analysis. BMI SDS (standardized odds ratio 2.3, 95% CI: 1.2–9.8, P = 0.04) was the sole determinant of hepatic fibrosis in the multivariate analysis. The area under the ROC curve for BMI SDS with fibrosis and severe steatosis was 0.769 (95% CI: 0.629–0.887, P = 0.02) and 0.832 (95% CI: 0.742–0.944, P < 0.001), respectively. A BMI SDS cutoff of 1.7 demonstrated 71.4% sensitivity and 62.1% specificity for identifying hepatic fibrosis, whereas a cutoff of 1.4 showed a sensitivity of 92% and specificity of 65.3% [Figure 1a and b].

Close

Figure 1:

Receiver operating characteristic curve for body mass index standard deviation score cutoff for (a) severe steatosis and (b) fibrosis. The area under the curve for (a) Severe steatosis 0.832 (95% confidence interval [CI], 0.742–0.944, P < 0.001) and (b) Fibrosis- 0.769 (95% CI: 0.629–0.887, P = 0.02). Green line represents scenario with no value of the diagnostic test and the blue line represents the value observed in the study.

Export to PPT

Serum ALT levels were below twice the upper limit of normal in 23 (56.1%) subjects with steatosis and 2 (28.7%) with fibrosis (one from the control group and one from the obese group). The ROC curve for ALT associated with fibrosis and severe steatosis in obese subjects showed areas under the curve of 0.756 (95% CI 0.48–1, P = 0.001) and 0.747 (95% CI 0.619–0.875, P = 0.001), respectively. An ALT cutoff of 69 IU/L exhibited 83.0% sensitivity and 82.1% specificity in identifying TE-detected hepatic fibrosis. In comparison, levels of 42.5 IU/L had a specificity of 57% and a sensitivity of 80% for severe hepatic steatosis [Figure 2a and b].

Close

Figure 2:

Receiver operating characteristic curve for alanine transaminase cut-off for (a) severe steatosis and (b) fibrosis. Area under curve for (a) Severe steatosis- 0.747 (95% confidence interval [CI] 0.619–0.875, P = 0.001) and (b) Fibrosis- 0.756 (95% CI: 0.48–1.0, P = 0.001). Green line represents scenario with no value of the diagnostic test and the blue line represents the value observed in the study.

Export to PPT

DISCUSSION

The high prevalence of TE identified severe steatosis and significant fibrosis in asymptomatic obese children and adolescents in our study highlights the need for screening for liver health in obese children and adolescents. These observations are commensurate with studies showing a high prevalence of TE-detected steatosis and fibrosis in children and adolescents with severe obesity.^[16] The prevalence of elevated ALT levels amongst obese children is similar to our previous study (35.5% as against 45%).^[9]

In our study, the significantly higher CAP and LSM scores in obese subjects compared to controls align with previous studies.^[17-19] Identifying BMI SDS as the primary determinant of steatosis and fibrosis reinforces the connection between obesity and liver disease. A significant correlation between markers of liver dysfunction (ALT, CAP, and LSM scores) and elevated levels in obese individuals with steatosis and fibrosis suggests a dose-dependent adverse effect of BMI SDS on the liver. These findings are similar to those observed in the National Health and Nutrition Examination Survey (NHANES) study of 867 adolescents, in which BMI was the key predictor of TE-diagnosed steatosis and fibrosis.^[20] Higher BMI and prandial BG levels in obese individuals with steatosis reiterate the link between obesity, steatosis, and insulin resistance.

The identification of TE-detected hepatic fibrosis in 9.6% of obese subjects (6.5% with advanced hepatic fibrosis) raises concerns. This was significantly higher than the prevalence in controls (1.6%). Hepatic fibrosis signifies an advanced stage of NAFLD progression, posing a high risk for the development of chronic liver disease and cirrhosis.^[4,5] The natural history of NAFLD in children and adolescents is less clear. However, given the rising incidence of liver transplants and NAFLD-related liver disease in adolescents in the developed world, early identification of hepatic fibrosis is crucial. The increased likelihood of hepatic dysfunction in obese individuals with metabolic complications in our study aligns with a previous study that showed an odds ratio of 3.6 for hepatic fibrosis in subjects with metabolic complications.^[21] This underscores the necessity of screening for hepatic fibrosis in individuals with metabolic complications.

The high diagnostic accuracy of ALT in identifying TE-detected severe steatosis and fibrosis underscores its importance in screening obese individuals for liver dysfunction. While TE is a valuable tool for detecting hepatic fibrosis, its high cost limits its use across different settings. ALT levels may remain normal in individuals with mild steatosis, making them unsuitable as screening tests for detecting NAFLD.^[22] However, the primary purpose of biochemical screening is to identify severe steatosis and fibrosis, where ALT levels exhibit a high discriminatory value, as shown by our study. We suggest an ALT cutoff of 42.5 IU/L and 69 IU/L to identify obese children and adolescents at increased risk for TE-detected severe steatosis and fibrosis. This two-step strategy (liver function test in the initial step and fibrosis assessment in high-risk individuals) is expected to reduce workup burden while not missing fibrosis. This aligns with the IAP obesity guidelines that suggest ALT levels as a screening test for obese children, with additional diagnostic tests like fibroscan reserved for levels above 80 IU/L.^[23] A BMI SDS above 1.4 and 1.7 can serve as clinical markers for assessing severe steatosis and fibrosis.

Our study is limited by using TE rather than liver biopsy to identify hepatic fibrosis, having a small sample size, and being conducted in a single center. Liver biopsy is the gold standard for determining NAFLD-associated hepatic fibrosis, but its invasive nature limits its application. The identification of differences between the obese subjects and controls and the demonstration of high accuracy of the ALT level, however, suggests that the study was adequately powered. The sample size was not calculated for ROC or sensitivity analysis; the significance observed in drawing the curve, however, did allow us to suggest cutoffs for ALT levels.

Our study underscores the significance of early identification of adverse hepatic health in obese children and adolescents, facilitating lifestyle changes that can improve their condition. Serum ALT levels above 69 IU/L in this population suggest an increased risk of TE-detected hepatic fibrosis, which calls for prompt evaluation and management. A multicenter study is required to validate our findings. Furthermore, prospective studies are needed to evaluate the impacts of lifestyle interventions on hepatic steatosis and fibrosis.

CONCLUSION

The identification of hepatic fibrosis in apparently healthy obese children and adolescents highlights the importance of careful evaluation for their liver health. The development of appropriate evaluation and management strategy requires prospective multicentric studies.