Editorial: Prevalence of Suspected Metabolic Dysfunction-Associated Liver Disease in Adolescents in the United States Using Updated Diagnostic Criteria—Authors' Reply

Abstract

We appreciate the editorial by Bonn and Xanthakos [1] discussing our recent study on the prevalence and predictors of suspected MASLD in US adolescents [2]. Their commentary highlights the importance of MASLD in paediatrics and reinforces the key questions that remain about its clinical and research application. While the transition in nomenclature from NAFLD to MASLD provides a more structured definition, its implementation in children requires further investigation.

A challenge in paediatric MASLD is misdiagnosis, especially in children with obesity whose liver disease is unrelated to steatosis. Hildreth et al. showed how autoimmune hepatitis can present with obesity, elevated ALT, and imaging findings that mimic steatosis, resulting in an erroneous MASLD diagnosis [3]. Conversely, our study found that 20.2% of adolescents with elevated ALT had no cardiometabolic risk factors, raising concerns that some children with liver disease may remain undetected under the updated criteria. Further complicating diagnosis, paediatric MASLD exhibits distinct histologic features, with a greater tendency for periportal inflammation and fibrosis than the pericentral pattern seen in adults [4]. These differences suggest MASLD in children involves unique developmental and metabolic pathways, reinforcing the need for paediatric-specific research rather than assuming direct parallels with adult disease.

Distinguishing association from causation is another key challenge in MASLD. While the definition links hepatic steatosis to metabolic dysfunction, it is unclear whether one drives the other. Hepatic fat accumulation in preschool-aged children, as described by Goyal et al., challenges the assumption that metabolic dysfunction is always the initiating factor [5]. If it were, it should precede hepatic steatosis. However, steatosis can appear before insulin resistance, dyslipidemia, or hypertension, suggesting it may contribute to metabolic dysfunction. Alternatively, both may share early-life determinants, such as prenatal exposures, epigenetic modifications, or perinatal nutrition patterns. Resolving these uncertainties requires longitudinal studies tracking hepatic steatosis from early childhood and would rely on repeated measures of hepatic steatosis. Currently, the only validated non-invasive tool for measuring hepatic steatosis in children is MRI-PDFF [6, 7]. Integrating this technology would allow us to determine whether steatosis precedes, predicts, or parallels metabolic dysfunction and help identify windows for intervention and stratify risk.

The inclusion of waist circumference in MASLD criteria aims to improve specificity by capturing central adiposity [8]. Malki et al. demonstrated that visceral adiposity is the strongest predictor of hepatic steatosis, supporting waist circumference as a potential surrogate [9]. However, our study found that among adolescents with elevated ALT, all those with an elevated waist circumference were also identified by BMI ≥ 85th percentile, meaning waist circumference did not enhance case detection. Furthermore, it is not routinely measured in paediatric practice and is subject to variability due to examiner technique and patient positioning. These factors call into question its value for MASLD diagnosis in children.

Ultimately, criteria for MASLD should be grounded in solid evidence, function well in paediatric practice, and fulfil their intended purpose of informing clinical care for this population, with special attention to ensuring accurate diagnosis and minimising misclassification in children.

Sheila L. Noon: conceptualization, writing – original draft, writing – review and editing. Tin Bo Nicholas Lam: conceptualization, writing – original draft, writing – review and editing. Jeffrey B. Schwimmer: conceptualization, writing – original draft, writing – review and editing.

The authors' declarations of personal and financial interests are unchanged from those in the original article [2].

This article is linked to Noon et al. papers. To view these articles, visit https://doi.org/10.1111/apt.70022 and https://doi.org/10.1111/apt.70067.