We appreciated the recent study by Fracanzani et al.1 The authors address the need to identify valuable criterion to discriminate patients with suspected nonalcoholic fatty liver disease (NAFLD) who deserve histological assessment of disease severity and progression from those who do not. In a cohort of 458 adults with biopsy confirmed NAFLD, they found patients with normal levels of alanine aminotransferase (ALT) (n = 63 [14% of the whole sample]) being significantly older, having lower body mass index, triglycerides, milder insulin resistance, inflammation, and steatosis, but higher prevalence of hypertension compared with patients with increased ALT.
In a sample of children and adolescents (n = 203) with biopsy-confirmed NAFLD, we report similar results. Normal values of ALT are seen at the time of biopsy in 21% to 23%,2, 3 and 60% of these present with liver fibrosis.2 Severity of necro-inflammation and fibrosis worsens as the number of morbidities clustering in the MetS increases, while there is no association with levels of ALT.2
|Normal ALT (n = 43)||Increased ALT (n = 160)||P Value*|
|Age (years)||12.8 ± 3.1||12.0 ± 3.3||0.5|
|Body mass index (z-score)||1.89 ± 0.45||1.80 ± 0.69||0.3|
|Waist circumference (cm)||93.4 ± 10||89.3 ± 10.9||0.05|
|Systolic blood pressure (mm Hg)||114.8 ± 15.9||112.8 ± 15.2||0.3|
|Diastolic blood pressure (mm Hg)||66 ± 8||68 ± 8||0.2|
|Fasting glucose (mg/mL)||83.6 ± 15||81.6 ± 15||0.2|
|2-Hour glucose (mg/mL)||116 ± 47||123 ± 31||0.004|
|Fasting insulin (μIU/mL)||14.5 ± 11.4||13.9 ± 10.1||0.9|
|2-Hour Insulin (μIU/mL)||78 ± 56||101 ± 61||0.05|
|Cholesterol (mg/mL)||155 ± 19||163 ± 40||0.6|
|Triglycerides (mg/mL)||83 ± 24||112 ± 79||0.05|
|Alanine aminotransferase (IU/L)||25.8 ± 8||97.9 ± 72|
|Aspartate aminotransferase (IU/L)||30.3 ± 12||55.9 ± 30|
|Gamma glutamyl transpeptidase (IU/L)||17.6 ± 8.6||39.9 ± 23.8|
|HOMA-IR||2.96 ± 2.3||2.81 ± 20.9||0.4|
|NAS score||3.91 ± 1.7||3.75 ± 1.6||0.6|
|Steatosis grade (1/2/3)||31.9/40.2/27.9||29.9/41.3/28.8||0.8|
|Fibrosis stage (0/1/2/3/4)||18.6/69.8/7/4.6||22.6/68.8/4.5/4.1||0.9|
|Necro-inflammatory grade (0/1/2/3)||14/81.3/4.7/0||15.1/70/12.4/2.5||0.02|
Children and adolescents with normal ALT levels differed significantly from those with abnormal aminotransferases in waist circumference (P = 0.05), triglycerides (P = 0.05), 2-hour glucose (P = 0.004), and insulin (P = 0.01) after a standard load of 75 g glucose. On univariate analysis, in the group of children with normal ALT levels, age (P = 0.04), impaired glucose metabolism (defined as yes/no including impaired fasting or glucose tolerance and overt type 2 diabetes [P = 0.05]), and 2-hour glucose (P = 0.02) associated significantly with nonalcoholic steatohepatitis. Insulin resistance (as estimated by the Homeostatic Model Assessment [HOMA-IR]) was associated with fibrosis (grade ≥2 [P = 0.04]). In patients with abnormal ALT levels, no variable associated significantly with nonalcoholic steatohepatitis, and ALT was the only variable associated with fibrosis (P < 0.0001). The model variables included age, body mass index, ALT, HOMA-IR, 2-hour glucose and insulin, and presence of glucose metabolism abnormalities (as described above).
In conclusion, there is a strong agreement between data obtained in Italian adult patients with those of children. In agreement with the report of Fracanzani et al.,1 we found a prevalence of necro-inflammation and fibrosis in children with ultrasonographic evidence of steatosis even in absence of ALT abnormalities; in this group of individuals, impaired glucose metabolism and insulin resistance are the factors more closely associated with severe liver disease. The prevalence and the number of the components of the MetS is similar between patients with normal and abnormal ALT values. Therefore, in children as well as in adults, normal ALT does not represent a valuable criterion to exclude from liver biopsy patients with persistently ultrasonographic evidence of steatosis. Because the disease may persist into adulthood and the lifestyle intervention may result in the significant amelioration of histological derangement,4 there is a strong need for identifying individuals among those presenting with risk factors for NAFLD (obesity, insulin resistance, and so forth) at risk for severe histological liver derangement despite normal ALT values.