Phosphatidylethanol improves diagnostic accuracy of MASLD, MetALD, and ALD: Results from a prospective cohort study Authors: Somaya Albhaisi, Brian P Lee Published: 2026 Torp N, Bech KT, Schnefeld HL, Johansen S, Hansen CD, Semmler G, et al. Phosphatidylethanol and self-reported alcohol intake to subclassify individuals at risk of steatotic liver disease: an analysis of data from a prospective cohort study. Lancet Gastroenterol. Hepatol. 2025;10:975–985. doi: 10.1016/S2468-1253(2500187-6). The diagnostic landscape of steatotic liver disease (SLD) has undergone a paradigm shift following the introduction of the recent multi-etiological nomenclature.1 By acknowledging the frequent overlap between metabolic dysfunction and alcohol consumption, the current classification – comprising metabolic dysfunction-associated steatotic liver disease (MASLD), metabolic dysfunction- and alcohol-associated liver disease (MetALD), and alcohol-associated liver disease (ALD) – aims to provide a more nuanced clinical framework.1 However, this framework is still heavily based on the accuracy of self-reported alcohol intake. In a prospective cohort study published in Lancet Gastroenterology and Hepatology, Torp et al. demonstrate that this reliance on patient self-report is a potential diagnostic, and they propose phosphatidylethanol (PEth) as the objective biomarker needed to refine SLD subclassification.2 Alcohol consumption is prone to misclassification, particularly underreporting, which can be unintentional (e.g. misunderstanding of what constitutes a standard drink) or intentional (e.g. due to stigma).3 The current study by Torp et al. quantifies this potential misclassification among patients with SLD. Among approximately 3,000 study participants, approximately 40% of those with a history of excessive alcohol consumption underestimated their consumption when compared with blood PEth concentrations (586/1,482 [39.5%]). Alcohol intake was self-reported as both a 1-week recall and a 3-month average, supplemented by the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C). PEth cut-off values were used to categorize participants into groups (<20 ng/ml, 20–79 ng/ml, 80–199 ng/ml, and ≥200 ng/ml). Underestimation was defined as a low self-report (<20 g/day for females, <30 g/day for males) paired with a PEth concentration of ≥80 ng/ml, or a moderate–high self-report (20–49 g/day for females, 30–59 g/day for males) paired with a PEth concentration of ≥200 ng/ml. While a PEth threshold of ≥20 ng/ml is considered to reflect alcohol consumption of at least two standard drinks per day and has been recommended by expert opinion to distinguish MASLD from MetALD and ALD(6), the current study used a higher threshold of PEth ≥80 ng/ml, which has a sensitivity of 91% and specificity of 77% for detecting consumption exceeding four standard drinks per day.2 Even within the context of the metabolic group, more than 10% of individuals had PEth ≥80 ng/ml, indicative of significant alcohol consumption not detected by standard questionnaires. The significance of these findings extends beyond diagnostic accuracy and has implications for patient management. The distinction between MASLD, MetALD, and ALD is not just a semantic difference. These subgroups often exhibit distinct patterns of fibrosis progression as well as differential responses to therapeutic interventions.4 While weight loss and metabolic control are the cornerstones of MASLD treatment, abstinence from alcohol or a significant reduction in alcohol consumption is the primary driver of improvement in MetALD and ALD.5 Misclassifying a patient with significant alcohol-related injury as having MASLD alone may lead to an overemphasis on metabolic targets without addressing the primary driver of liver damage. PEth offers a unique advantage as a direct biomarker. Unlike indirect markers such as gamma-glutamyltransferase, which can be influenced by various non-alcohol-related factors, PEth is only formed in the presence of ethanol.6 With a detection window of 1 to 4 weeks, PEth provides a stable and objective measure of recent consumption that bridges the gap between the immediate and the long-term.7 Torp et al. further enhance the clinical utility of this biomarker by developing a decision tree that integrates PEth with the AUDIT-C questionnaire (Fig. 1). Their algorithm suggests that nearly 40% of patients could forego PEth testing if their self-report and clinical profile are sufficiently low-risk, thereby optimizing resource allocation while catching the majority of underreporters (812/2,042 [39.8%] of patients with SLD). Importantly, the algorithm also identifies patients at the opposite extreme − those already reporting high alcohol intake (≥50 g/day for females, ≥60 g/day for males) − in whom PEth testing offers limited diagnostic benefit. In this group, 81.5% already had PEth ≥200 ng/ml, and fewer than 1% had PEth <20 ng/ml. Consequently, the decision tree assumed ALD in these patients without reflexive PEth testing, reserving it for unclear cases where it provides the most clinical value. Among patients with MASLD and a low AUDIT-C, only a small fraction had elevated PEth levels (20/612 [3.3%] had PEth ≥80 ng/ml). These findings highlight the continued value of validated self-report measures, such as AUDIT-C; PEth is an adjunct tool to self-report, not a replacement. Diagnostic algorithm for PEth testing in steatotic liver disease. AUDIT-C, Alcohol Use Disorders Identification Test-Consumption; ALD, alcohol-related liver disease; MASLD, metabolic dysfunction-associated steatotic liver disease; MetALD, metabolic dysfunction and alcohol-related liver disease; PEth, phosphatidylethanol. Adapted from Torp N, et al. Lancet Gastroenterol Hepatol. 2025. DOI: 10.1016/S2468-1253(25)00187-6. There are several challenges associated with routine PEth testing. From this study, it is clear that although PEth is sensitive for detecting MetALD and ALD, established thresholds to distinguish MASLD, MetALD, and ALD by PEth require further study. For example, among patients classified as MASLD by self-report, 39.0% had PEth levels ≥20 ng/ml, suggesting reclassification to MetALD or ALD [559/1,433]. This is notable because the ≥20 ng/ml threshold used here is the consensus lower boundary for any alcohol exposure, which means that these patients had objective evidence of recent drinking beyond what their self-report suggested, rather than confirmed reclassification per se. This difference was most striking among those with a prior history of excessive drinking: 19.3% of patients from the alcohol group who self-reported MASLD-range intake had PEth ≥200 ng/ml compared with only 3.9% in the metabolic group (72/374 vs. 41/1,059), a reminder that the same diagnostic threshold can carry very different clinical weight depending on the patient population being assessed. In addition, a single assessment of PEth may not capture changes in drinking behavior over time since it is a cross-sectional assessment that does not account for longitudinal changes.8 A significant number of patients may move from one SLD category to another over time as their drinking behavior changes; thus, PEth may be more useful as a longitudinal assessment tool as opposed to a single assessment tool for subclassifying SLD.9 The work of Torp et al. represents a major step forward in the development of an objective and biomarker-based approach to measure alcohol consumption among individuals with liver disease. By comparing PEth to self-reported data, this study critically examines the potential pitfalls of relying on uncorroborated self-report in the assessment of alcohol-related liver damage, particularly with underreporting. With the development of personalized medicine and etiology-specific therapies, the use of objective markers such as PEth will be essential in the management of SLD. This study provides the robust evidence base needed to transition PEth from a specialized tool to part of standard care in the subclassification and management of SLD.10 Financial support NIH T32 Translational Research Training in Hepatology grant (T32DK127977) (SA). National Institute on Alcohol Abuse and Alcoholism of the National Institutes of Health under Award Number K23AA029752 (BPL). Conflict of interest The authors declare no conflicts of interest pertaining to this manuscript. Please refer to the accompanying ICMJE disclosure forms for further details. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.jhepr.2026.101894. Supplementary data The following is the Supplementary data to this article: Multimedia component 1 References Associated Data Supplementary Materials Multimedia component 1