ELF Score as a Marker for Systemic Inflammation and Fibrosis

#1

Has anyone had an ELF test done? Although this test is ordered to screen for liver fibrosis, it is a measure of systemic inflammation and fibrosis which, as we know, drives all aspect of aging. Unfortunately, it costs >$ 800 out of pocket, so unaffordable for me. After learning about what ELF measures, I wondered why drugs such as Pirfenidone, montelukast, and others which target fibrotic, inflammatory processes more directly are not discussed as much as others such as rapamycin, as potential healthspan extending drugs. ELF score is based on serum levels of Procollagen III Amino-Terminal Peptide (PIIINP), Hyaluronic acid , and Timp-1. Elevated levels of these markers all indicate systemic inflammation and fibrotic processes which manifest as stiffness in organs and numerous other changes associated with aging and metabolic dysfunction. I am researching which drug to take, since my husband’s ELF score indicated SEVERE fibrosis, although Fibroscan ruled out fibrosis in his liver. Interestingly, his Fibroscan improved over 1 year, while his ELF score worsened. This was puzzling, until I discovered that he has Aortic Regurgitation, which AI confirmed may be the cause of his high ELF score. Another consideration is that such a high level of inflammation may be the body’s attempt to repair damage, therefore reducing it enough that it doesn’t cause damage while keeping it at a level to enable repair is beyond me. Maybe a med bot knows.

1 Like
#2

I’ve never heard of the ELF test, so ask Claude to update me on it. Does this information sync up with what you know about the test?

Claude Prompt:

Summarize the scientific and clinical evidence behind the ELF test. Although this test is ordered to screen for liver fibrosis, it is a measure of systemic inflammation and fibrosis. Would this be a good test to assess healthy aging?

Claude Opus 5.5 Response:

Short answer: no, not as a stand-alone healthy-aging metric. ELF is a well-validated liver test and a useful liver screen for an aging, metabolically at-risk person. As a systemic aging readout, though, it is non-specific, rises with age for reasons that aren’t fully understood, and has never been validated against aging outcomes outside liver disease. One correction to the premise: ELF measures matrix turnover and remodeling (fibrogenesis), not inflammation as such. Only TIMP-1 behaves like an inflammatory or senescence-associated protein.

What it measures

ELF combines three direct fibrosis markers: hyaluronic acid (HA), type III procollagen peptide (PIIINP), and tissue inhibitor of metalloproteinase-1 (TIMP-1). The standard algorithm is 2.494 + 0.846·ln(HA) + 0.735·ln(PIIINP) + 0.391·ln(TIMP-1), so HA carries the most weight. labcorp

Each marker has its own biology:

  • HA is cleared mainly by liver sinusoidal endothelial cells. A high HA therefore reflects reduced hepatic clearance as much as increased production. It is also produced in joints and other connective tissue.
  • PIIINP comes from type III collagen synthesis anywhere in the body: vessels, heart, lung, skin, and growing tissue.
  • TIMP-1 is an acute-phase-like protein that blocks matrix breakdown. It is also a component of the senescence-associated secretory phenotype (SASP).

So the “systemic” signal is real, but it comes from mixed sources.

Clinical evidence

Diagnosis.

  • A 36-study systematic review found good accuracy for advanced fibrosis in viral hepatitis (AUROC 0.69–0.98) and excellent accuracy in NAFLD (0.78–0.97) and alcohol-related liver disease (0.92–0.94). ucl
  • The LITMUS meta-analysis (Vali 2020) found sensitivity above 0.90 for excluding fibrosis at a cutoff of 7.7, but limited specificity at low cutoffs. amsterdamumc
  • Positive predictive value collapses in low-prevalence settings: at the 10.51 threshold with 5–10% disease prevalence, PPV is only 0.26–0.43. labcorp

Prognosis. This is the strongest part of the evidence base.

  • In the original derivation cohort, 457 patients were followed for a median of 7 years, and each one-unit rise in ELF roughly doubled the risk of a liver-related outcome. adelaide
  • In a second cohort, each unit increase carried a 2.53-fold higher risk of a liver event, adjusted for age and fibrosis stage. jci
  • A 2024 Scottish cohort extended this to all-cause mortality: each unit of ELF was associated with an adjusted hazard ratio of 1.86 for death from any cause. Note, however, that this was a referred population of 1,327 people with a median ELF of 10.2. It was not a healthy cohort. jhep-reportsdundee

Regulatory status. ELF was the first prognostic tool for patients with advanced (F3–F4) fibrosis due to NASH to receive FDA De Novo marketing authorization. It is explicitly not cleared for diagnosing NASH or for staging fibrosis. The FDA-indicated risk bands are below 9.8 (low risk), 9.8 to under 11.3 (moderate), and 11.3 or higher (high). bizwire 2021 8 24 fda grants marketing authorization to siemens healthineers elf test for nash prognostic assessment +2

Population screening. The Danish study by Kjaergaard et al. (J Hepatol 2023) is the best general-population comparison:

  • ELF identified advanced fibrosis more accurately than FIB-4 or the NAFLD Fibrosis Score (AUROC 0.85 vs 0.73 and 0.66). nih
  • ELF produced 11% false positives, compared with 35% for FIB-4 and 45% for NFS. sdu
  • However, in participants aged 65 or older, ELF frequently gave false positive results. journal-of-hepatology

Beyond the liver. ELF tracks fibrosis in other organs:

  • In systemic sclerosis, ELF correlated with the extent of skin involvement, lung diffusing capacity (DLCO), disability, and age (r=0.41). bmj
  • In children with NAFLD, higher HA and PIIINP were associated with lower eGFR. Kidney function is therefore a confounder. nih
  • During growth, healthy children have a median ELF of 8.9, well above adults, peaking around puberty because HA and PIIINP rise with growth. physiciansweekly

Why it fails as an aging metric

  1. Strong, poorly explained age gradient. In the Finnish Health 2000 general population cohort, 13.4% had ELF of 9.8 or higher. That figure was under 1% in people younger than 40 but rose to 38% in those over 70. Some of this may be real occult liver fibrosis. However, the Danish data show most of these elevated scores are false positives against liver stiffness. You can’t separate “systemic fibro-aging” from “subclinical liver disease” from “reduced HA clearance” without imaging. tandfonline
  2. Healthy ranges overlap the disease zone. One reference-interval study concluded that the healthy reference range overlaps the range the manufacturer labels as moderate fibrosis. cuni
  3. Noise relative to the changes you’d want to track. In healthy volunteers, within-subject variation was 32.7% for HA, 10.6% for PIIINP, and 4.2% for TIMP-1. The reference change value for the overall ELF score was 10.1%. At a baseline near 9, you need roughly a 0.9-unit shift before a change is meaningful. HA is also sensitive to recent meals and exercise, so standardize collection: fasting, morning draw, no hard training the day before. tacr
  4. No validation against aging outcomes. I found no general-population study linking ELF to frailty, multimorbidity, or all-cause mortality after adjusting for age and liver stiffness. The mortality signal comes from liver-clinic cohorts.
  5. Better tools already capture this biology. TIMP-1 appears in aging proteomics, where it is among the age-associated SASP proteins linked to IL-6, gait speed, and red cell distribution width. Large proteomic clocks also weight matrix biology heavily: proteins involved in extracellular matrix interactions are among the top contributors to the UK Biobank proteomic age clock. ELF is essentially a crude three-protein subset of that signal, with an algorithm optimized for liver histology rather than aging. nihs4me

Where it does fit in a longevity panel

ELF is worth ordering in someone with metabolic risk factors as a second-line liver check after FIB-4. That is exactly the role the Danish data support. Liver fibrosis is a meaningful and modifiable driver of mortality in aging adults.

If you track it over time for research or self-experimentation, a few rules help:

  • Standardize sampling conditions.
  • Look at the three components separately, not just the composite score.
  • Pair it with cystatin C/eGFR to account for kidney function.
  • Confirm any value of 9.8 or higher with elastography before calling it liver fibrosis.

For a systemic aging readout, GDF15, cystatin C, or a validated proteomic clock is the more defensible choice.

Sources: