HSD17B13 — The Liver's Hidden Shield Against Fat-Driven Inflammation
Deep inside your liver cells, tiny fat droplets accumulate when caloric intake, alcohol, or
metabolic stress overwhelm the liver's processing capacity. Coating the surface of these fat
droplets is a lipid droplet–associated enzyme called
HSD17B13 | Hydroxysteroid 17-beta dehydrogenase 13, a member of the short-chain dehydrogenase/reductase enzyme family exclusively expressed in hepatocytes.
In most people, this enzyme is fully active — and paradoxically, its activity appears to
drive inflammation, fibrosis, and liver disease progression. One of the most striking
discoveries in liver genetics over the past decade is that losing HSD17B13 function is
protective: a naturally occurring insertion variant that silences the enzyme is associated
with dramatically reduced risk of nonalcoholic steatohepatitis (NASH), alcoholic liver
disease, cirrhosis, and hepatocellular carcinoma.
The Mechanism
The rs72613567 variant consists of an adenine insertion (;A) immediately adjacent to the
donor splice site | The nucleotide sequence at the exon-intron boundary where the spliceosome cuts to remove the intron; disruption here causes aberrant or skipped splicing
of intron 6 in HSD17B13. This insertion disrupts normal mRNA processing, producing an
alternative transcript (termed isoform D) that is frameshifted and encodes a truncated,
enzymatically inactive protein. The truncated protein is also unstable — HSD17B13 protein
levels in liver biopsies decrease proportionally to the number of insertion alleles carried.
HSD17B13 normally functions as a retinol dehydrogenase and lipid droplet–associated
oxidoreductase, and it appears to modulate hepatic lipogenesis and lipid droplet expansion.
Its loss alters the liver's phospholipid composition and suppresses inflammation-related
gene expression pathways. Whole-transcriptome profiling of carriers shows downregulation
of 274 genes, predominantly in immune response and inflammatory pathways — explaining
why the loss-of-function state protects against the inflammatory cascade that drives
NASH and fibrosis.
Critically, the variant does not protect against simple
steatosis | Fat accumulation in liver cells without inflammation; the earliest stage of fatty liver disease.
The protection is specific to the progression from steatosis to steatohepatitis
(inflammation) and fibrosis — meaning the enzyme's activity is particularly harmful
during the transition to aggressive liver disease.
The Evidence
The landmark study from Abul-Husn et al. (NEJM 2018) analyzed 46,544 participants in
the Geisinger Health System cohort:
protective effect on alcoholic liver disease: 42% reduction in heterozygotes, 53% in homozygotes; on nonalcoholic liver disease: 17% and 30% reduction; on alcoholic cirrhosis: 42% and 73% reduction | Abul-Husn NS et al. N Engl J Med 2018;378:1096-1106.
These associations were replicated in independent cohorts totaling over 37,000 additional
participants.
At the histology level, Pirola et al. (2019) examined 356 biopsy-proven NAFLD
patients and found that each insertion allele reduces the odds of NASH (OR 0.61),
ballooning degeneration (OR 0.47), and lobular inflammation (OR 0.48) — all hallmarks
of progressive liver disease:
Pirola CJ et al. Splice variant rs72613567 prevents worst histologic outcomes in patients with NAFLD. J Lipid Res, 2019.
A 2020 meta-analysis pooling over 564,000 participants confirmed protection across all
liver disease categories:
pooled OR 0.73 for any liver disease, 0.81 for cirrhosis, 0.64 for hepatocellular carcinoma | Wang P et al. HSD17B13 rs72613567 protects against liver diseases and histological progression of NAFLD: a systematic review and meta-analysis. Eur Rev Med Pharmacol Sci, 2020.
In alcoholic liver disease specifically, a multicenter study of 3,315 European patients
found that TA carriers have significantly lower HCC risk (OR 0.73 in ALD patients, and
OR 0.64 for HCC development among those with ALD):
Yang J, Nault JC et al. Hepatology 2019;71:1099-1108.
The protective effect also modifies the impact of the major liver-risk variant
PNPLA3 rs738409 | The strongest common genetic risk factor for NAFLD and alcoholic liver disease — an I148M missense variant in adiponutrin/patatin-like phospholipase domain-containing protein 3:
the HSD17B13 TA allele substantially attenuates the liver injury associated with
the PNPLA3 G (risk) allele, demonstrating that these two variants interact within
the same lipid droplet biology pathway.
One important nuance: in patients who have already progressed to portal hypertension
and advanced cirrhosis, the protective effect appears attenuated — the variant protects
against developing severe liver disease more than it improves outcomes once severe
disease is established.
Practical Implications
For carriers of one or two insertion alleles, the evidence supports that your liver has
a meaningful biological buffer against alcohol-induced inflammation and metabolic liver
disease progression. However, this protection is not absolute — it operates on a spectrum
and can be overwhelmed by sufficient stressor load (heavy alcohol use, severe obesity,
or co-inherited liver-risk variants like PNPLA3 GG).
For non-carriers (DD genotype), the absence of this protection means that alcohol,
metabolic syndrome risk factors, and other liver stressors carry a higher baseline risk
for inflammatory liver disease progression. Periodic liver function monitoring (ALT/AST,
GGT) and abdominal imaging is particularly valuable for informing decisions about alcohol
consumption and metabolic management.
Interactions
The most important genetic interaction is with PNPLA3 rs738409 (I148M): the GG
high-risk genotype at PNPLA3 markedly amplifies liver disease risk, but HSD17B13 TA
carriers show attenuated ALT levels and less severe histology even when carrying PNPLA3
risk alleles. The ALT-lowering benefit of HSD17B13 TA is greatest among individuals
with three or four steatogenic alleles across PNPLA3 and TM6SF2 (rs58542926).
TM6SF2 rs58542926 (E167K) impairs hepatic VLDL secretion and elevates liver
triglycerides; like PNPLA3, its risk effect operates in the same lipid droplet
biology axis where HSD17B13 exerts protection, making these three variants key
components of any comprehensive liver genetic risk assessment.
All Genotypes
No protective HSD17B13 insertion — standard liver disease risk baseline
You carry the common reference genotype at rs72613567, with no adenine insertion in either copy of HSD17B13. Your HSD17B13 enzyme is fully functional, which is the typical state. About 64% of people of European descent share this genotype. This does not mean elevated liver disease risk in absolute terms — rather, you do not carry the genetic protection against NASH and alcoholic liver disease progression that the insertion allele provides. Your liver disease risk is driven by environmental factors (alcohol use, diet, metabolic health) and other genetic variants rather than by this SNP.
Two protective HSD17B13 insertions — strongest genetic protection against liver disease progression
You carry two copies of the protective adenine insertion in HSD17B13. Both copies of your HSD17B13 gene produce a truncated, non-functional protein, which means you have near-complete loss of HSD17B13 enzyme activity in the liver. This is associated with the strongest protection against inflammatory liver disease: roughly 30–53% reduction in risk of alcoholic and nonalcoholic liver disease, and up to 49–73% reduction in cirrhosis risk compared to non-carriers. About 4–5% of people of European descent and up to 10% of East Asians carry this homozygous protective genotype.
One protective HSD17B13 insertion — meaningfully reduced liver disease risk
You carry one copy of the protective adenine insertion in HSD17B13. This reduces your HSD17B13 enzyme levels by approximately half, conferring significant protection against the inflammatory progression of liver disease. Compared to non-carriers, you have approximately 17–42% lower risk of NASH and alcoholic liver disease, and roughly 26–42% lower risk of developing cirrhosis. About 32% of people share this heterozygous genotype. The protection is dose-dependent — two copies confer greater benefit — but one copy still provides a substantial and clinically meaningful buffer.