Animal Models of Nonalcoholic Fatty Liver Disease

Animal models of nonalcoholic fatty liver disease (NAFLD) are central to understanding how the disease progresses from simple steatosis through to cirrhosis and hepatocellular carcinoma. Because tracking this full spectrum in humans requires decades of follow-up and repeated invasive biopsies, researchers rely on rodent and other animal systems to observe disease mechanisms, test interventions and identify therapeutic targets in a controlled, ethically regulated setting. This article explains why these models are used, which are most common, how well they reflect human liver disease, and what their current limitations mean for translating preclinical findings into clinical practice.

Why do researchers use animal models to study fatty liver disease?

Animal models allow researchers to observe the full progression of NAFLD in a controlled setting, enabling experimental manipulation, tissue sampling at defined time points and mechanistic investigation that longitudinal human studies cannot provide.

Animal models are used to study non-alcoholic fatty liver disease (NAFLD) because they allow researchers to observe the full progression of liver disease in a controlled setting that would be ethically and practically impossible in humans. NAFLD encompasses a spectrum from simple hepatic steatosis through to non-alcoholic steatohepatitis (NASH), fibrosis, cirrhosis and, in some cases, hepatocellular carcinoma. Tracking this continuum in living human subjects requires decades of follow-up and repeated invasive liver biopsies, making longitudinal human studies extremely difficult to conduct.

It is worth noting that the international research community has recently adopted updated terminology: NAFLD is increasingly referred to as metabolic dysfunction-associated steatotic liver disease (MASLD), and NASH as metabolic dysfunction-associated steatohepatitis (MASH). The older NAFLD and NASH terms remain widely used in preclinical and legacy literature, so both sets of terms appear across the scientific literature.

Animal models offer several practical advantages. Researchers can:

• Control diet, genetics, environment and disease duration precisely

• Obtain liver tissue at defined time points, subject to the ethical and regulatory oversight described below

• Test experimental interventions before any human exposure

• Investigate molecular mechanisms at a cellular and subcellular level

In the UK, all animal research is governed by the Animals (Scientific Procedures) Act 1986 (ASPA), which requires a Home Office project licence, personal licences for individuals conducting procedures, and ethical review by an Animal Welfare and Ethical Review Body (AWERB). The 3Rs framework, promoted in the UK by the National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs), requires researchers to replace animal use with alternatives where possible, reduce the number of animals used, and refine procedures to minimise suffering.^[2] Animal tissue is therefore obtained within a tightly regulated framework, not without ethical oversight.

Beyond logistics, animal studies provide mechanistic insight that observational human data cannot. For example, identifying whether a specific inflammatory pathway drives fibrosis, or whether a metabolic intervention reverses steatosis, requires the kind of experimental manipulation that only animal work permits. Rodent models in particular have short reproductive cycles and well-characterised genomes, making them amenable to genetic modification and high-throughput screening.

It is important to note that no single animal model perfectly replicates human NAFLD in all its complexity. The field therefore relies on a range of complementary models, each capturing different aspects of the disease. Understanding the strengths and limitations of each model is essential for interpreting preclinical data and deciding which findings are likely to translate meaningfully to clinical research.

Which animal models are most commonly used in NAFLD research?

Mice, particularly C57BL/6J, ob/ob and db/db strains, are most commonly used; combined dietary models such as AMLN and GAN diets are increasingly preferred for producing progressive NASH with fibrosis in a metabolically relevant context.

Mice and rats are the most commonly used animal models in NAFLD research, with mice predominating because of the availability of well-characterised inbred strains and the relative ease of genetic manipulation. Among mouse strains, C57BL/6J mice are widely used because they develop diet-induced obesity and hepatic steatosis reliably when fed high-fat or Western-style diets, making them a practical starting point for metabolic liver disease studies.^[4]

Several specific models have become established reference standards:

• ob/ob mice: These carry a homozygous mutation in the leptin gene, resulting in severe obesity, insulin resistance and hepatic steatosis. Because they lack functional leptin signalling, they are relatively resistant to developing significant fibrosis without an additional dietary or chemical challenge, and are therefore often used as a platform for studying steatosis and early inflammatory changes.

• db/db mice: These carry a mutation in the leptin receptor gene and similarly develop obesity and steatosis, but again require a secondary insult to progress to NASH with fibrosis.^[6]

• Melanocortin-4 receptor knockout (MC4R-KO) mice: These develop obesity, insulin resistance and, importantly, progressive fibrosis on a high-fat diet, making them useful for studying later-stage disease.

• Foz/foz mice: A less commonly used but informative model carrying a mutation in the Alms1 gene; these mice develop spontaneous NASH with fibrosis when fed an atherogenic or high-fat diet.

Rat models, including Zucker fatty rats and Sprague-Dawley rats fed specific diets, are also used, particularly in pharmacological studies where larger tissue volumes are advantageous. Combined dietary models such as the AMLN diet (high fat, fructose and cholesterol) and the GAN diet are increasingly used because they produce progressive NASH with fibrosis in a metabolic context that more closely resembles human disease. The STAM model, which combines neonatal streptozotocin treatment with a high-fat diet, is used to study NASH-associated hepatocellular carcinoma.

Non-human primates have been studied but are rarely used owing to cost, ethical constraints and long disease timelines. Zebrafish are an emerging model for high-throughput genetic and drug screening; their optical transparency, rapid development and amenability to genetic manipulation allow hepatic lipid accumulation to be assessed by fluorescent staining and transgenic reporter systems, making them particularly useful for early-stage compound screening.

How well do these models reflect what happens in human livers?

Rodent models replicate key histological features of human NASH but diverge in portal fibrosis pattern, gut microbiome composition and lipid metabolism, and genetic models do not capture the polygenic complexity of human NAFLD.

Animal models of NAFLD reflect several key features of human liver disease but diverge from the human condition in important ways that researchers must account for when interpreting results. In humans, NAFLD is closely associated with central obesity, type 2 diabetes, dyslipidaemia and insulin resistance, and the disease progresses heterogeneously, with only a minority of patients advancing to NASH and fibrosis. Many rodent models replicate the metabolic context reasonably well, particularly the association between dietary fat intake, visceral adiposity and hepatic lipid accumulation.

Histologically, rodent NASH shares features with human disease, including hepatocyte ballooning, lobular inflammation and pericellular fibrosis. However, there are notable differences:

• Rodents rarely develop the degree of portal fibrosis seen in human NASH

• The gut microbiome composition differs substantially between rodents and humans, which is relevant given the emerging role of the microbiome in NAFLD pathogenesis

• Rodent lipid metabolism differs from human metabolism, particularly in the handling of dietary cholesterol

• Some models require extreme dietary manipulations (such as methionine and choline deficient diets) that produce rapid liver injury but cause weight loss rather than obesity, which is the opposite of the human phenotype

Histological activity in preclinical studies is commonly graded using the NAFLD Activity Score (NAS), developed by Kleiner and colleagues (Hepatology, 2005). It is important to understand that NAS was designed to grade the degree of histological activity, not to diagnose NASH, and it is not a validated surrogate endpoint for clinical outcomes. Many research groups adapt NAS-based scoring for rodent tissue, and results should be interpreted accordingly rather than as direct equivalents of clinical trial endpoints.

Genetic models with single-gene mutations, such as ob/ob mice, do not reflect the polygenic and environmental complexity of human NAFLD. The translational record is mixed: some targets identified in animal models, such as those engaged by pioglitazone and, more recently, resmetirom, have shown benefit in human trials, but many others have not replicated in the clinic. This high attrition rate between preclinical promise and clinical efficacy underscores the importance of using animal data as hypothesis-generating rather than confirmatory evidence. Despite these limitations, well-designed animal studies remain a valuable component of the research pipeline when their findings are interpreted with appropriate caution.

What dietary and genetic methods are used to induce NAFLD in animals?

High-fat and Western diets are the most widely used dietary methods; genetic approaches include hepatocyte-specific PTEN knockout and SREBP-1c overexpression, with combined dietary and genetic models such as DIAMOND offering a broader disease spectrum.

Dietary and genetic methods are both used to induce NAFLD in animals, and researchers frequently combine the two approaches to produce models that more closely resemble the human metabolic phenotype. The choice of induction method determines which stage of the disease spectrum is modelled and how quickly it develops.

Dietary induction methods include:

• High-fat diet (HFD): Typically providing 45 to 60 per cent of calories from fat, this is the most widely used approach. C57BL/6J mice fed an HFD develop obesity, insulin resistance and steatosis within 8 to 16 weeks, though significant fibrosis may take considerably longer.

• Western diet: Combines high fat with high fructose or sucrose to mimic the dietary pattern associated with human NAFLD, producing a more complete metabolic syndrome phenotype.

• Methionine and choline deficient (MCD) diet: Rapidly induces steatohepatitis and fibrosis within weeks but causes weight loss and does not reproduce the metabolic context of human disease, limiting its translational relevance.^[9][10]

• High-fructose diet: Used to model the contribution of dietary sugar to hepatic de novo lipogenesis and steatosis.

• Atherogenic diet: Enriched with cholesterol and cholate, used particularly in mice to accelerate fibrosis.

• AMLN and GAN diets: These high-fat, high-fructose, high-cholesterol formulations are increasingly used because they produce progressive NASH with fibrosis in a metabolic context that more closely resembles human disease than the MCD diet.

Genetic methods include targeted deletion or mutation of genes involved in lipid metabolism, insulin signalling or inflammatory pathways. Hepatocyte-specific PTEN knockout mice develop progressive steatohepatitis and fibrosis owing to dysregulated PI3K-Akt signalling, and mice overexpressing SREBP-1c develop hepatic lipid accumulation through upregulation of lipogenic gene programmes. Chemical induction using low-dose carbon tetrachloride is sometimes added to dietary models to accelerate fibrosis.

The DIAMOND (Diet-Induced Animal Model Of Non-Alcoholic Fatty Liver Disease) mouse model, described by Asgharpour and colleagues (Journal of Hepatology, 2016), uses a specific hybrid genetic background (C57BL/6J x 129S1/SvImJ) combined with a high-fat, high-fructose Western diet. This combined approach produces progressive NASH with fibrosis and, in a proportion of animals, hepatocellular carcinoma, offering a broad disease spectrum that more closely mirrors the human condition than single-diet or single-gene models.

How are animal models used to test new treatments for fatty liver disease?

Candidate therapies are assessed in animal models for histological, biochemical and molecular endpoints before human exposure, with preclinical testing forming a regulatory requirement overseen by agencies including the MHRA and EMA.

Animal models are used to test new treatments for fatty liver disease by providing a controlled environment in which candidate therapies can be assessed for efficacy, safety and mechanism of action before any human exposure. Preclinical testing in animals is a regulatory requirement for investigational medicines and forms a critical step in the drug development pipeline overseen by agencies such as the MHRA and the European Medicines Agency (EMA).

In practice, treatment studies in NAFLD models typically assess several endpoints:

• Histological outcomes: Reduction in hepatic steatosis, lobular inflammation, hepatocyte ballooning and fibrosis stage, graded using systems such as the NAFLD Activity Score (NAS). NAS is a measure of histological activity and is not a diagnostic tool for NASH, nor a validated surrogate for clinical outcomes; its use in rodent studies should be interpreted with this in mind.

• Fibrosis quantification: Collagen deposition is commonly measured by Sirius Red morphometry and hepatic hydroxyproline content, providing quantitative fibrosis endpoints alongside histological staging.

• Biochemical markers: Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), lipid profiles and markers of insulin resistance.

• Molecular targets: Gene expression, protein levels and pathway activation in liver tissue.

• Body composition and metabolic parameters: Body weight, adiposity, glucose tolerance and energy expenditure.

Several drug classes that have reached clinical trials were first characterised in animal models, including farnesoid X receptor (FXR) agonists, peroxisome proliferator-activated receptor (PPAR) agonists and fibroblast growth factor 21 (FGF21) analogues. GLP-1 receptor agonists, which are now licensed for type 2 diabetes and weight management, have also demonstrated hepatic benefits in animal studies that have informed ongoing clinical investigation.

Researchers increasingly use combination treatment protocols in animals to reflect the multifactorial nature of human NAFLD, testing metabolic agents alongside anti-fibrotic or anti-inflammatory compounds. The reproducibility of animal treatment data across independent laboratories remains an important quality consideration. The ARRIVE 2.0 guidelines (published in PLOS Biology, 2020, and supported by the NC3Rs) provide a reporting framework that encourages randomisation, blinding, pre-specified sample sizes and, where possible, pre-registration of animal study protocols, all of which help reduce reporting bias and improve the reliability of preclinical findings.

What are the current limitations of animal-based NAFLD research?

Key limitations include the absence of a single comprehensive model, genetic homogeneity of inbred strains, dietary extremes that do not reflect human disease, publication bias and biological differences in lipid metabolism and immune composition between rodents and humans.

The current limitations of animal-based NAFLD research are substantial and help explain why many treatments that show promise in preclinical studies have not replicated their effects in human clinical trials. The most fundamental limitation is the biological distance between rodents and humans: differences in hepatic lipid metabolism, immune cell composition, gut microbiome, body size and lifespan all affect how disease develops and how the liver responds to intervention.

Key limitations include:

• Lack of a single comprehensive model: No existing animal model reproduces all features of human NAFLD simultaneously, including the metabolic syndrome context, progressive fibrosis, and spontaneous progression to hepatocellular carcinoma at a realistic rate.

• Sex and hormonal differences: Many studies use only male animals, which does not reflect the sex-specific differences in NAFLD prevalence and progression observed in humans. Regulatory and scientific bodies, including the NC3Rs, increasingly encourage the inclusion of both sexes in preclinical study designs.

• Dietary extremes: Some widely used diets, such as the MCD diet, produce liver injury through mechanisms that differ from those driving human disease, reducing their translational relevance.

• Genetic homogeneity: Inbred mouse strains lack the genetic diversity of human populations, potentially overestimating treatment effects that may not generalise across individuals.

• Publication bias: Positive preclinical results are more likely to be published, which can create an inflated impression of a treatment's potential. Pre-registration of animal study protocols, as encouraged by the NC3Rs and ARRIVE 2.0 guidelines, is one measure being adopted to address this.

• Regulatory and ethical constraints: The 3Rs framework (replacement, reduction and refinement) governs animal research in the UK under the Animals (Scientific Procedures) Act 1986, requiring researchers to minimise animal use and seek alternative methods where possible. All UK animal research requires Home Office project licensing and ethical review by an AWERB.

Organ-on-a-chip technology, human liver organoids and in silico modelling are emerging as complementary approaches that may reduce reliance on animal models in future, and active research programmes are exploring their application to MASLD and MASH. However, at present, well-designed animal studies remain an indispensable component of NAFLD research, provided their limitations are transparently acknowledged and findings are interpreted with appropriate caution before clinical translation is assumed.

This page describes research models used in scientific investigation of fatty liver disease. It does not constitute clinical advice on diagnosis or treatment. Anyone with concerns about their liver health should speak to their GP or a specialist.

Scientific References

1. Research and testing using animals: licences and compliance.
2. Replacing animals in science: A strategy to support the development, validation and uptake of alternative methods.
3. A multisociety Delphi consensus statement on new fatty liver disease nomenclature.
4. Nonalcoholic Fatty Liver Disease Induced by High-Fat Diet in C57BL/6 Mice.
5. Butein Alleviates Non-Alcoholic Steatohepatitis in Leptin-Deficient Mice by Modulating the PDE4/cAMP/p-CREB Pathway.
6. Obese and diabetic db/db mice develop marked liver fibrosis in a model of nonalcoholic steatohepatitis: role of short-form leptin receptors and osteopontin.
7. Design and validation of a histological scoring system for nonalcoholic fatty liver disease.
8. A diet-induced animal model of non-alcoholic fatty liver disease (DIAMOND).
9. The GLP-1R agonist liraglutide limits hepatic lipotoxicity and inflammatory response in mice fed a methionine-choline deficient diet.
10. Comparison of murine steatohepatitis models identifies a dietary intervention with robust fibrosis, ductular reaction, and rapid progression to cirrhosis and cancer.
11. Reporting preclinical gene therapy studies in the field of Niemann-Pick type C disease according to the ARRIVE guidelines [references ARRIVE 2.0, PLOS Biology 2020].
12. Role of Resmetirom, a Liver-Directed, Thyroid Hormone Receptor Beta-Selective Agonist, in the Treatment of NASH.
13. Clinical Trial: Phase 3 Trial of Resmetirom Versus Placebo in Metabolic Dysfunction-Associated Steatohepatitis – Reanalysis of Fibrosis Stage 2–3 Subset.
14. Quantitative regression of qFibrosis with resmetirom: Exploratory histologic endpoints from the MAESTRO-NASH phase III clinical trial.

Frequently Asked Questions