.png)
The condition formerly known as NAFLD has a new name, a stricter definition, and a clearer detection pathway. Most health check panels in Hong Kong have not kept pace.
Approximately one in four adults in Hong Kong has Metabolic Dysfunction - Associated Fatty Liver Disease (MAFLD). Most health checks report it as absent. This is not primarily a technology problem — it is a clinical best practice gap, compounded by panels designed around diagnostic criteria that the major hepatology societies have now substantially revised.
A population study by Wong et al., published in Gut, used proton magnetic resonance spectroscopy — the gold-standard method for quantifying intrahepatic fat — to establish an NAFLD (now MAFLD) prevalence of 27.3% (95% CI 24.5–30.2%) among Hong Kong Chinese adults, drawn from a random sample of the government census database.[1]
The metabolic dose-response in that cohort was direct: prevalence ranged from 4.5% in adults with no metabolic syndrome components to 80.0% in those with all five.[1] Liver disease tracks closely with the cardiometabolic profile that most corporate health check packages already measure.
The downstream consequences extend well beyond the liver. Large meta-analyses of patients studied under prior NAFLD criteria — populations that overlap substantially with MAFLD — show that those with the condition carry:
These are not parallel conditions to be screened for separately. They are overlapping metabolic pathways. A missed MAFLD diagnosis is also a missed opportunity to act on cardiovascular, renal, and metabolic risk simultaneously.
Hong Kong's demographic trajectory sharpens the urgency. According to the Primary Healthcare Blueprint published by the Hong Kong SAR Government, the population aged 65 and over will grow from 1.5 million — 20% of the total — in 2021 to 2.52 million, or 31% of the total, by 2039.[5] Of those currently living with chronic health conditions, 47% are already aged 65 and over.[5] Older adults carry a higher accumulation of the cardiometabolic risk factors that define MAFLD — and, as discussed below, require adjusted diagnostic thresholds under current clinical guidelines.
The term NAFLD has been replaced in clinical guidelines across the globe, though not uniformly. The Asian Pacific Association for the Study of the Liver (APASL) was the first to act: its 2020 guidelines introduced MAFLD — metabolic-associated fatty liver disease — representing a fundamental shift from exclusion-based diagnosis to one grounded in positive evidence of metabolic dysfunction. The 2025 APASL Clinical Practice Guidelines reaffirm MAFLD as the regional standard for the Asia-Pacific.[12] Notably, clinical researchers from the Chinese University of Hong Kong were among the contributing authors — reflecting the significant role that Asia-Pacific hepatology has played in shaping how this disease is defined and diagnosed globally. In 2023, AASLD and EASL adopted a parallel framework under the term MASLD through an international consensus process;[6] the two constructs overlap substantially in most patient populations.
Under MAFLD, a positive diagnosis requires hepatic steatosis plus one of three conditions:[12]
The first two conditions — overweight/obesity and type 2 diabetes — are measured in virtually every standard executive health check in Hong Kong. The third covers lean patients with metabolic risk, drawing on parameters that many comprehensive health check panels already collect. The connection to liver risk is rarely made.
Two compounding detection gaps explain the underdiagnosis:
Standard liver function tests are insufficient. A meta-analysis of 11 studies covering 4,084 patients found that approximately 25% of NAFLD/MAFLD patients have a normal ALT value (95% CI 20–31%).[7] A separate study published in Hepatology confirmed that the full histologic spectrum of the disease — including advanced fibrosis — occurs in patients with normal ALT.[8] Reporting a normal liver panel does not exclude MAFLD.
Ultrasound has a detection threshold. The Hernaez et al. meta-analysis published in Hepatology reported pooled sensitivity of 84.8% and specificity of 93.6% for detecting hepatic steatosis — but only at a hepatic fat content of approximately 20–30% or above.[9] Below that level, sensitivity degrades materially. Patients with early or mild steatosis — precisely those most amenable to intervention — are those most likely to be missed.
Both gaps compound each other. A patient with early MAFLD, normal ALT, and hepatic fat content below the ultrasound detection threshold will receive a normal liver result on a standard health check. The disease continues to progress.
The missing step is risk asseessment, such as FIB-4. The Fibrosis-4 index is calculated from four values already present in a standard blood panel: age, AST, ALT, and platelet count. No additional sample is required. It is endorsed as the primary first-line non-invasive fibrosis stratification tool by the AASLD 2023 Practice Guidance,[10] the EASL-EASD-EASO 2024 Clinical Practice Guidelines,[11] and the APASL 2025 Clinical Practice Guidelines.[12] Most health check panels already collect all four inputs. Almost none calculate the score.
FIB-4 is included in DARA's portfolio of clinically validated risk calculators — meaning health check operators can generate a stratified fibrosis risk output from data they are already collecting, without adding tests or changing their existing panel.
The stepwise pathway endorsed by AASLD, EASL-EASD-EASO, and APASL is sequential and non-invasive:[10][11][12]
This age-adjusted threshold is directly relevant to Hong Kong's demographic trajectory. As the population aged 65 and over approaches one-third of the total by 2039, a growing proportion of health check participants will fall into the cohort where applying standard FIB-4 thresholds without adjustment would systematically over-refer.
The practical implication for health check programme operators is this: the data to apply the current guideline pathway already exists in most standard panels. The diagnostic criteria that define MAFLD are already being measured. FIB-4 uses inputs that are already being collected. What is largely absent is the clinical framework that connects them — and a reporting mechanism that translates those inputs into a stratified, actionable risk assessment.
Updated clinical knowledge does not only change what is documented in a result. It reveals what was always there, unread.
This article summarises key developments in MAFLD guidelines for health check programme operators and is not a substitute for clinical guidance. Programme operators and clinicians should refer directly to the guidelines cited below for diagnostic criteria, clinical thresholds, and implementation detail.
Contact us to see how DARA's portfolio of clinically validated risk calculators supports a current-standard approach to metabolic and liver health assessment.
References used in this post
