Hepatic Steatosis in Infancy: The Beginning of Pediatric Nonalcoholic Fatty Liver Disease?

**Background:** Nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease in children and the fastest rising indication for liver transplantation in young adults in the US. While emerging evidence describes NAFLD in preschool-age children, and fetal hepatic steatosis has been reported in stillborns (42% in one autopsy study), the natural history of early-onset steatosis is unknown. Current pediatric guidelines recommend screening at-risk children starting at 9–11 years of age. This study aimed to determine the prevalence of CT evidence of hepatic steatosis in previously healthy infants ≤3 months of age.

**Methods:** This retrospective study at Cincinnati Children's Hospital Medical Center identified all abdominal CT scans performed between November 1, 2009, and December 1, 2019, in infants ≤3 months corrected age. Inclusion required contrast-enhanced CT adequate for steatosis assessment. Exclusions: noncontrast CT, prematurity <28 weeks gestational age, known diagnoses contributing to steatosis. Of 119 CT scans identified, 65 were in previously healthy infants. Two independent, blinded reviewers measured liver and spleen attenuation in Hounsfield units (HU). Hepatic steatosis was defined using two published criteria: (1) portal venous phase liver-spleen attenuation difference <–20 HU (LS20 criteria), and (2) portal venous phase L–B < 101 HU (LB101 criteria), where L–B = (L – 0.3 × [0.75 × P + 0.25 × A])/0.7. Clinical data were extracted from electronic medical records. Groups were compared using Wilcoxon-Mann Whitney or Fisher exact tests; kappa statistics assessed interobserver agreement.

**Key Results:** The cohort (n=65) was predominantly male (72%), non-Hispanic (97%), White (74%), with median age 60 days (IQR 34–73 d). Most CTs (91%) were for trauma evaluation, including 30 for suspected nonaccidental trauma (NAT). Using LS20 criteria, 23% (radiologist) to 28% (second reviewer) had steatosis; using LB101 criteria, 26% to 28% had steatosis. Prevalence did not differ significantly by age month (17% in month 1, 30% in month 2, 27% in month 3; P=0.657). Maternal overweight/obesity/diabetes prevalence was 11% overall, with no significant difference between steatosis and no-steatosis groups (18% vs 8%; P=0.366). Infants with steatosis had significantly lower ALT (median 25 U/L [IQR 20–52] vs 75 U/L [IQR 37–123]; P=0.018) and AST (median 36 U/L [IQR 26–61] vs 72 U/L [IQR 50–146]; P=0.031) compared to those without steatosis. Alkaline phosphatase and glucose were not significantly different. None of the original clinical radiology reports mentioned steatosis. Interobserver agreement was good (kappa 0.763; 95% CI 0.646–0.880).

**Clinical Implications:** This study demonstrates that CT evidence of hepatic steatosis is present in up to 26% of otherwise healthy infants ≤3 months of age, yet is universally underreported in clinical practice. The finding that ALT and AST were lower in the steatosis group (potentially reflecting extrahepatic aminotransferase sources in the nonsteatosis group, such as from trauma) highlights the limitations of relying on liver enzymes alone for screening in this age group. The lack of association with maternal metabolic risk factors in this cohort may reflect underreporting or the possibility that in utero metabolic effects wane by 2 months of age. The study is limited by its retrospective design, small sample size, predominantly non-Hispanic population, use of CT (not MRI) for fat quantification, and potential bias from the high proportion of NAT workups. Nonetheless, these findings challenge current screening paradigms that begin at 9–11 years and suggest that the origins of pediatric NAFLD may extend into infancy. Prospective birth cohort studies are urgently needed to determine the natural history of infantile steatosis and whether early intervention could prevent progression to advanced liver disease.